Bioactive Natural Products (Part G), Volume 26 (Studies in Natural Products Chemistry)

Studies in Natural Products Chemistry Volume 26 Bioactive Natural Products (Part G)

studies in Natural Products Chemistry edited by Atta-ur-Rahman

Vol. 1 Vol. 2 Vol. 3 Vol. 4 Vol. 5 Vol. 6 Vol. 7 Vol. 8 Vol. 9 Vol. 10 Vol. 11 Vol. 12 Vol. 13 Vol. 14 Vol. 15 Vol. 16 Vol. 17 Vol. 18 Vol. 19 Vol. 20 Vol. 21 Vol. 22 Vol. 23 Vol. 24 Vol. 25 Vol. 26

Stereoselective Synthesis (Part A) Structure Elucidation (Part A) Stereoselective Synthesis (Part B) Stereoselective Synthesis (Part C) Structure Elucidaton (Part B) Stereoselective Synthesis (Part D) Structure and Chemistry (Part A) Stereoselective Synthesis (Part E) Structure Chemistry (Part B) Stereoselective Synthesis (Part F) Stereoselective Synthesis (Part G) Stereoselective Synthesis (Part H) Bioactive Natural Products (Part A) Stereoselective Synthesis (Part I) Structure and Chemistry (Part C) Stereoselective Synthesis (Part J) Structure and Chemistry (Part D) Stereoselective Synthesis (Part K) Structure and Chemistry (Part E) Structure and Chemistry (Part F) Bioactive Natural Products (Part B) Bioactive Natural Products (Part C) Bioactive Natural Products (Part D) Bioactive Natural Products (Part E) Bioactive Natural Products (Part F) Bioactive Natural Products (Part G)

Studies in Natural Products Chemistry Volume 26 Bioactlve Natural Products (Part G)

Edited by

Atta-ur-Rahman H.E.J. Research Institute of Chemistry, University of Karachi, Karachi 75270, Pakistan

2002

ELSEVIER Amsterdam - London - New York - Oxford - Paris - Shannon - Tokyo

ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 211, 1000 AE Amsterdam, The Netherlands

® 2002 Elsevier Science B.V. All rights reserved.

This work is protected under copyright by Elsevier Science, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier Science Global Rights Department, PO Box 800, Oxford OX5 IDX, UK; phone: (+44) 1865 843830; fax: (+44) 1865 853333; e-mail: [email protected]. You may also contact Global Rights directly through Elsevier's home page (http://www.elsevier.com), by selecting 'Obtaining Permissions'. In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (+1) (978) 7508400; fax: (+1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London WIP OLP, UK; phone: (+44) 207 631 5555; fax: (+44) 207 631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of Elsevier Science is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier Science Global Rights Department, at the mail, fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.

First edition 2002 Library of Congress Cataloging in Publication Data A catalog record from the Library of Congress has been applied for.

ISBN:

0 444 51004 4

^ The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Printed in The Netherlands.

FOREWORD The present volume of "Studies in Natural Product Chemistry" presents the chemistry of a large number of exciting natural products. The triterpenoid saponins from Caryophyllaceae family have been reviewed by Jia et al. Marine fatty acids present an unusual array of structural types. Recent developments in the total synthesis of bioactive marine fatty acids are reviewed by Carballeira. The chemistry of secologanin, a monoterpenoid glucoside which is a precursor of about 3000 indole, isoquinoline and related alkaloids, is presented by Szabo. Lignans represent an important and widely distributed class of dimeric phenylpropanold derivatives, many of which have shown antimicrobial, anti-viral or anti-feedant activities, some of which play an important role In plant defense. The chemistry of the lignans of Podophyllum has been reviewed by Moraes et al, while the bioactivity of lignans has been reviewed by Rios et al. Saffron, the yellow orange stigmas from the Crocus sativus flower, is the world's most expensive spice which contains some 34 volatile components responsible for the powerful odor and a number of non-volatile components responsible for its color. The chemistry and biological activities of the constituents of saffron are reviewed by Liakopoulou-Kyriakides et al. The bark of Fraxunus emus has been employed In the traditional systems of medicine for treatment of wounds, inflammation, arthritis and dysentry. The chemistry and biological activity of the constituents of the plant have been reviewed by Kostova et al. The management of undesirable fresh water phytoplankton blooms, which can cause losses in aquaculture, farm livestock and waterfowl, using natural campounds is discussed by Schrader etal. Aromatic plants have been used in folk medicine, cosmetics and food industry, largely because of the essential oils that they contain. The pharmacological activities and applications of essential oils of Salvia sclarea and Salvia desoleana are presented by Peana et al. Plant- and animal-parasitic nematodes are responsible for considerable crop damage each year and they also contribute to malnutrition and disease in humans and domestic livestock. The chemistry and bioactivity of secondary metabolites is reviewed by Ghisalberti. Resveratrol is one of the most widely known stilbenes which undergoes polymerization in plants and fungi to afford a host of complex oligomers. The structure, chemistry and biological activity of resveratrol oligomers is reviewed by Cichewicz et al. Phytopathogenic bacteria and plants produce a number of bioactive metabolites, many of which are plant growth regulators and phytotoxins. The chemistry and pharmacological activity of such compounds is reviewed by Evidente et al. The chemistry and biological activities of the constituents of Rubia tlnctomm L are presented by Derksen et al. The Pteridium species are In a group referred to as the "bracken fern". Recent developments in the chemical behaviour, toxicology and chemical ecology of compounds found in bracken fern are reviewed by Alonso-Amelot. Flavonoids can serve as important neutraceutlcals as they exhibit interesting anti-oxidant properties and play a role on ascorbic acid preservation. This field is discussed by Marin et al. The occurrence, structure and bioactivity of 1,7-diarylheptanoids is reviewed by Claeson et al. The structures and biological activities of natural insecticides from Lauraceae, Compositae, Ranunculaceae and Boraginaceae species are presented by Gonzalez-Coloma et al, while potential anti-parasitic substance from natural sources are presented by Kayser et al. The chemistry and bioactivity of nitric oxide (NO) in plant and animal cells is reviewed by Wendehenne et al. while the role of 0-aminophenol-type tryptophan metabolites in living organisms is discussed by Rescigno et al. Recent developments on structural studies and pharmacology of strychnos alkaloids is reviewed by Rasoanaivo et al. while the chemistry and biological activity of certain compounds present in the pungent principles of radish is presented by Uda et al. The occurrence, structure, properties, metabolism, biological activities and uses of carbohydrates having the hexo-D-manno configuration are reviewed by Matheson. Sponges belonging to the llthistid order have

VI

proved to be spectacular sources of several interesting classes of biologically active compounds. This is discussed in a review by D'Auria et al. The chemistry and biological properties of copper/topa quinone-containing amine oxidases is reviewed by Sebela. The present volume contains articles written by eminent experts in their respective fields. It should prove to be of wide interest to medicinal chemists, phamriacologists and organic chemists working in academia and industry. This 26^^ volume of this encyclopaedic series represents a landmark. The 25 volumes published previously during the last 12 years contain a huge amount of exciting chemistry. It is hoped that it will be received with the same order of enthusiasm as its predecessors. I would like to express my thanks to Dr. Shakeel Ahmad for his assistance in the preparation of the index. I am also grateful to Mr. Muhammad Asif and Mr. Waseem Ahmad for typing and to Mr. Mahmood Alam for secretarial assistance.

Atta-ur-Rahman Ph.D. (Cantab), Sc.D. (Cantab)

July, 2001

Vll

PREFACE It is testimony to the healthy state of Natural Products chemistry over the last decade that the present volume is the 26**^ in a series which is only 12 years old. Professor Atta-urRahman is once again to be congratulated for his great skill and perseverance needed to sustain such a valuable project and for his ability to attract first-rate investigators to write chapters covering the full range of the field. Volume 26 Is of particular Interest since all of the chapters reflect the importance of the biological activity of nature's molecules activity which continues to Inspire the evolution of new drugs against disease. It is significant that a high percentage of clinically approved drugs have structures based on natural product leads. Although the biological receptors for the majority of the structures described in this volume are still unknown, progress at the chemistry- biology interface is now accelerating rapidly and we look forward to future volumes devoted to the molecular description of natural product bioactivity. Also, thanks to modern, rapid spectroscopic methodology, there appears to be an exponential rate of discovery and description of novel structures from natural sources and we can confidently expect this trend to continue. Finally, I would like to express the hope that in spite of his heavy duties as Federal Minister for Science and Technology in Pakistan, Professor Atta-ur-Rahman will continue to find time to produce this very valuable series of reviews on Natural Product Chemistry, which taken as a whole serve as essenfial reading for active researchers in the field.

A. Ian Scott, F.R.S.

Davidson Professor of Science Director of Center for Biological NMR August. 2001

This Page Intentionally Left Blank

IX

CONTENTS Foreword Preface Contributors Triterpenoid saponins from caryophyllaceae family ZHONGHUA JIA, KAZUO KOIKE, NIRANJAN P. SAHU, AND TAMOTSU NIKAIDO

v vii xiii 3

Recent developments in the total synthesis of bioactive marine fatty acids NESTOR M. CARBALLEIRA

63

Some aspects of the chemistry of secologanin LASZLO F. SZABO

95

The lignans of Podophyllum RITA M. MORAES

149

New findings on the bioactivity of lignans JOSE LUIS RIOS, ROSA M. GINER, JOSE M. PRIETO

183

Crocus sativus-biological active constituents M. LIAKOPOULOU-KYRIAKIDES, D.A. KYRIAKIDIS

293

Chemical components of Fraxinus ornus bark-structure and biological activity IVANKA N. KOSTOVA, TANYA lOSSIFOVA

313

Natural compounds for the management of undesirable freshwater phytoplankton blooms KEVIN K. SCHRADER, AGNES M. RIMANDO, STEPHEN O. DUKE

351

Pharmacological activities and applications of Salvia sclarea and Salvia desoleana essential oils ALESSANDRA T. PEANA, MARIO D.L. MORETTI

391

Secondary metabolites with antinematodal activity EMILIO L. GHISALBERTI

425

Resveratrol oligomers: structure, chemistry, and biological activity ROBERT H. CICHEWICZ, SAMIR A. KOUZI

507

Bioactive metabolites from phytopathogenic bacteria and plants ANTONIO EVIDENTE AND ANDREA MOTTA

581

Rubia tinctorum L. GOVERDINA C.H. DERKSEN AND TERIS A. VAN BEEK

629

The chemistry an toxicology of bioactive compounds in bracken fern {Pteridium Sp.), with special reference to chemical ecology and carcinogenesis MIGUEL. E. ALONSO-AMELOT

685

Flavonoids as nutraceuticals: structural related antioxidant properties and their role on ascorbic acid preservation F.R. MARIN, M.J. FRUTOS, J.A. PEREZ-ALVAREZ,

741

F. M A R T I N E Z - S A N C H E Z , J.A. DEL

RIO

Natural products as potential antiparasitic drugs OLIVER KAYSER, ALBRECHT F. KIDERLEN, SIMON L. CROFT

779

Natural insecticides: structure diversity, effects and structure-activity relationships. A case study A. G O N Z A L E Z - C O L O M A , M . REINA, C. GUTIERREZ, B.M. FRAGA

849

Occurrence, structure, and bioactivity of 1,7-diarylheptanoids PER CLAESON, UBONWAN P. CLAESON, PATOOMRATANA TUCHINDA AND VICHAI REUTRAKUL

881

Nitric oxide: chemistry and bioactivity in animal and plant cells DAVID WENDEHENNE, LAURE DUSSABLY, JEAN-FRANCOIS JEANNIN AND ALAIN PUGIN

909

o-Aminophenol-type tryptophan metabolites: 3-hydroxykynurenine, 3Hydroxyanthranilic acid, and theu* role in living organisms ANTONIO RESCIGNO AND ENRICO SANJUST

965

New contributions to the structure elucidation and pharmacology of Strychnos alkaloids P. RASOANAIVO, M- T. MARTIN, E. GUITTET AND F. FRAPPIER

1029

Occurrence of biologically active 2-thioxopyrrolidines and 3,5-disubstituted 2-thiohydantoins from the pungent principle of radish (Raphanus sativus L.) YASUSHI UDA, YOSHIO OZAWA AND KOICHI YONEY AMA

1073

Structure, occurrence and roles of carbohydrates with the hexo-D-manno configuration NORMAN K. MATHESON

1113

The chemistry of lithistid sponge: a spectacular source of new metabolites M. VALERIA D'AUPUA, ANGELA ZAMPELLA AND FRANCO ZOLLO

1175

XI

Copper/topa quinone-containing amine oxidases - recent research developments MAREK SEBELA, IVO FREBORT, MAREK PETRIVALSKY AND PAVEL PEC

1259

Subject Index

1301

This Page Intentionally Left Blank

Xlll

CONTRIBUTORS

Miguel. E. AlonsoAmelot

Grupo de Quimica Ecologica, Departamento de Quimica, Universidad de Los Andes, Merida, Venezuela

Koichi Yoney Ama

Center for Research on Wild Plants, Universit)^ Utsunomiya, 321-8505, Japan

Teris A Van Beek

Laboratory of Organic Chemistry, Phytocheniical Section, Wageningen University, Dreijenplein 8, 6703 HB Wageningen, The Ntherlands

Nestor M. Carballeira

Department of Chemistry, University of Puerto Rico, P.O. Box 23346, San Juan, Puerto Rico 00931-3346, Division of Basic Pharmaceutical Sciences, College of Pharmacy, University of Louisiana at Monroe, Monroe, Louisiana 71209, USA

Robert H. Cichewicz

Utsunomiya

Per Claeson

Division of Pharmacognosy, Department of Pharmacy, Upssala University, Biomedical Centre, Box 579, SE-751 23 Uppsala, Sweden

Ubonvvan P. Claeson

Institute for Bioactive Natural Products, Uppsala Science Park, SE-751 83 Uppsala, Sweden

Simon L. Croft

London School of Hygiene and Tropical Medicine Department of Infectious and Tropical Diseases Keppel Street London, WCIE 7HT, United Kingdom

M. Valeria D'Auria

Dipartimento di Chimica Delle Sostanze Naturali, Via D. Montesano 49 80131 Napoli, Italy

Goverdina C.H. Derksen

Laboratory of Organic Chemistry, Phytochemical Section, Wageningen University, dreijenplein 8, 6703 HB Wageningen, The Netherlands

Stephen O. Duke

United States Department of Agriculture, Agricultural Research Service, Natural products Utilization Research Unit. P.O. Box 8048, University MS 38677-8048, USA

Laure Dussably

EPHE/INSERM, Laboratory of Cancer Immunotherapy and Unit 517, Faculty of Medicine, 7 Bd Jeanne d'Arc, 21033 Dijon Cedex, France

Antonio Evidente

Dipartimento di Scienze Chimico-Agrarie, Universita di Napoli "Fderico \V\ Via Universita 100, 80055 Portici, Italy

XIV

B.M. Fra",a

Institute de Productos Naturales y Agrobiologia, CSIC, Avda. Asti'ofisico F. Sanchez, 38206 La Laguna, Tenerife, Spain

F. Frapp ier

Laboratoire de Chimie des Substances Maturelles, ESA 8041 CNRS, Museum National d'Histoire Naturelle, 63 rue Buffon, 75231 Paris Cedex 05, France Department of Biochemistry, Faculty of Science, Palaci
Ivo Frebort

M.J. Frutos

Departamento de Tecnologia Agroalimentaria (Division de Tecnologia de Alimentos), escuela Politecnica Superior de Ingenieros Agronomos, Universidad Miguel Hernandez de Elche, 03312, Orihuela (Alicante), Spain

Emilio L. Ghisalberti

Department of Chemistry, Universit}/ of Western Australia, Nedlands 6907 W.A., Austi^alia

Rosa M. Giner

Departament de Farmacologia, Facultat de Farmacia, Universitat de Valencia, Vicent Andres Estelles, 46100 Burjassot, Valencia, Spain

A. Gonzalez-Coloma

Centro de Ciencias Medioambientales, CSIC, Serrano 115dpdo., 28006 Madrid, Spain

E. Guittet

Laboratoire de RMN, Institut de Chimie des Substances Naturelles, CNRS, 91198-Gif-sur-Yvette, France

C. Gutierrez

Centro de Ciencias Medioambientales, SCIC, Serrano 115dpdo., 28006 Madrid, Spain

Tanya lossifova

Institute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of Sciences, Bg-1113 Sofia, Bulgaria

Jean-Francois Jeannin

EPHF7iNSE)RM, Laboratory of Cancer immunotherapy and Unit 517, Faculty of Medicine, 7 Bd Jeanne d'Arc, 21033 Dijon Cedex, France

Zhonghua Jia

School of Pharmaceutical Sciences, Toho University, Miyama 2-2-1, Funabashi, Chiba 275-8510, Japan

Oliver Kayser

Freie Universitat berlin, Institut fiir Pharmazie, Pharmazeutische Biotechnologie, Kelchstrape 31, 12169 Berlin, Germany

XV

Albrecht F. Kidelen

Robert Koch-Institut, Nordufer 20, 13353 Berlin, Germany

Kazuo Koike

School of Pharmaceutical Sciences, Toho University, Miyama 2-2-1, Funabashi, Chiba 275-8510. Japan

IvankaN. Kostova

Institute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of Sciences, Bg-1113 Sofia, Bulgaria

Samir A. Kouzi

Division of Basic Pharmaceutical Sciences, College of Pharmacy, University of Louisiana at Monroe, Monroe, Louisiana 71209, USA

D.A. Kyriakidis

Faculty of Chemistry, Aristotle Universit)^ of Thessaloniki, Greece 54006

M. LiakopoulouKyriakides

Dept. Chem. Engineering, Section of Chemistry and Faculty of Chemistry, Aristotle University of Thessaloniki, Greece 54006

F.R. Marin

Departamento de Tecnologia Agroalimentaria (Division de Tecnologia de Alimentos), escuela Politecnica Superior de fngenieros Agronomos, Universidad Miguel Hernandez de Elche, 03312, Orihuela (Alicante), Spain

M-T. Martin

Laboratoire de RMM, Institut de Chimie des Substances Naturelles, CNRS, 91198-Gif-sur-Yvette, France

F. Martinez-Sanchez

Departamento de Tecnologia Agroalimentaria (Division de Tecnologia de Alimentos), escuela Politecnica Superior de Ingenieros Agronomos, Universidad Miguel Hernandez de Elche, 03312, Orihuela (Alicante), Spain

Norman K. Matheson

Department of Agricultural Chemistry and Soil Science, The University of Sydney, N.S.W. Australia, 2006

Rita M. Moraes

National Center for Natural Products Research, The Research Institute of Pharmaceutical Sciences, School of Pharmacy, University Mississippi, University, MS 38677, USA

Mario D.L. Moretti

Dipartimento di Scienze del Framaco, Universita degli Studi di Sassari 1-07100 Sassari, Italy

Andrea Motta

Istituto per la Chimica di Molecole di Interesse Biologico (Istituto Mazionale di Chimica di Sistemi Biologici) del CNR, Via Toiano 6, 80072 Arco Felice, Italy

XVI

Tamotsu Nikaido

School of Pharmaceutical Sciences, Toho University, Miyama 2-2-1. Funabashi, Chiba 275-8510, Japan

Yoshio Ozawa

Department of Food Nutrition, Gunma Women's Junior College, Takasaki, 370-0033, Japan

Alessandra T. Peana

Dipartimento di Scienze del Fraraaco, Universita degli Studi di Sassari 1-07100 Sassari, Italy

Pavel Pec

Department of Biochemistry, Facult)/ of Science, Palacky University, Slechtitelu 11, 783 71 Olomouc, Czech Republic

J.A. Perez-Alvarez

Departamento de Tecnologia Agroalimentaria (Division de Tecnologia de Alimentos), escuela Politecnica Superior de Ingenieros Agronomos. Universidad Miguel Hernandez de Elche, 03312, Orihuela (Alicante), Spain

Marek Petfivalsky

Department of Biochemistry, Faculty of Science, Palacky Universit>% Slechtitelu 11, 783 71 Olomouc, Czech Republic

Jose M. Prieto

Departament de FaiTnacologia, Facultat de Famiacia, Universitat de Valencia, Vicent Andres Estelles, 46100 Burjassot, Valencia, Spain

Alain Puein

Unite Mixte INRA/Universite de Bourgogne, BBCE-IPM Laboratory, INRA BV 1540, 17 rue Sully, 21034 Dijon Cedex, France

P. Rasoanaivo

Laboratoire de Phytochimie et de Pharmacologic Cellulaire et Parasitaire, Institut Malgache de Recherches Appliquees, B.P. 3833, 101-Antananarivo, Madagascar

M. Reina

institute de Productos Naturales y Agrobiologia, CSIC, Avda. Astrofisico F. Sanchez, 38206 La Laguna, Tenerife, Spain

Antonio Rescigno

Cattedra di Chimica Biologica, Dipartimento di Scienze Mediche, Universita di Cagliari, Cittadella Universitaria, 09042 MonseiTato, Cagliari, Italy

Vichai Reutrakul

Department of Chemistry, Faculty of Science, Mahidol Universit)/, Bangkok 10400, Thailand

Agnes M. Rimando

United States Department of Agriculture, Agricultural Research Service, Natural products Utilization Research

XVll

Unit. P.O. Box 8048, University MS 38677-8048, USA J.A. Del Rio

Departamento de biolgia Vegetal (Unidad de Fisiologia Vegetal), Facultad de Biologia, Univesidad de Murcia, 30100, Espinardo (Murcia), Spain

Jose Luis Rios

Departanient de Farmacologia. Facultat de Farmacia, Universitat de Valencia, Vicent Andres Estelles, 46100 Burjassot, Valencia, Spain

Miranjan P. Sahu

Indian Institute of Chemical Biology, 4 Raja S.C. Mullick Raod, Calcutta-700 032, India

Kevin K. Schrader

United States Department of Agriculture, Agricultural Research Service, Natural products Utilization Research Unit. P.O. Box 8048, University MS 38677-8048, USA

Marek Sebela

Department of Biochemistry, Faculty of Science, Palacky University, Slechtitelu 11, 783 71 Olomouc. Czech Republic

Enrico Sunjust

Cattedra di Chimica Biologica, Dipartimento di Scienze Mediche, Universita di Cagliari, Cittadella Universitaria, 09042 Monserrato, Cagliari, Italy

Laszlo F. Szabo

Institute of Organic Chemistry, Semmelweis University, Hogyes u. 7. H-1092 Budapest, Hungary

Patoomratana Tuchinda

Department of Chemistry, Faculty of Science, Mahidol University, Bangkok 10400, Thailand

Yasushi Uda

Department of Bioproductive Sciences, and Center for Research on Wild Plants, Utsunomiya University, Utsunomiya, 321-8505, Japan

David Wendehenne

Unite Mixte I>JRA/Universite de Bourgogne, BBCE-IPM Laboratory, INRA BV L540, 17 rue Sully, 21034 Dijon Cedex, France

Angela Zampella

Dipartimento di Chimica Delle Sostanze Naturali, Via D. Montesano 49 80131 Napoli, Italy

Franco Zollo

Dipartimento di Chimica Delle Sostanze Naturali, Via D. Montesano 49 80131 Napoli, Italy

This Page Intentionally Left Blank

Bioactive Natural Products

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

TRITERPENOID SAPONINS FROM CARYOPHYLLACEAE FAMILY ZHONGHUA JIA"'^, KAZUO KOIKE", NIRANJAN P. SAHU^ AND TAMOTSU NIKAIDO' ^School of Pharmaceutical Sciences, Toho University, Miyama l-l-l, Funabashi, Chiba 275-8510, Japan. ^Indian Institute of Chemical Biology, 4 Raja S, C. Mullick Road, Calcutta-700 032, India ABSTRACT: Triterpenoid saponins from the Caryophyllaceae family are reviewed with special emphasis on recent developments in purification and structural elucidation aspects. Structural characterization using high field NMR techniques is examplified using saponarioside A, a triterpenoid saponin isolated from Saponaria officinalis and possessing nine sugar units in complex chains. A brief report on the biological activities of the saponins of Caryophyllaceae family is also included.

INTRODUCTION The family Caryophyllaceae is a group of mostly herbaceous flowering plants occurring mainly in temperate and warm-temperate regions of the Northern Hemisphere with its center of diversity in the Mediterranean region. The family consists of more than 2000 species with about 90 genera and is fairly easy to recognize because of its uniformity, most species having swollen nodes and opposite leaves. A salient feature of the plants of this family is their capacity to produce voluminous stable froth when shaken with water, indicating the presence of saponins. In fact, the name saponin was derived from the Caryophyllaceae plant Saponaria officinalis (soapwort), the root of which was used as soap since time immemorial {sapo—hdiXm for soap). Other well-known saponinbearing plants in the family are Gypsophila paniculata/muralis (baby's breath) and Acanthophyllum gypsophiloides (Turkestan soap root). Many plants of the family viz. Dianthus chinensis, Vaccaria segetalis, Silene species are well documented as herbal drugs in Asia, especially in traditional Chinese medicine. Saponaria officinalis and Gypsophila species have also been used as folk medicine in Europe. Besides, the family is the source of many ornamental plants, the most widely known species being Dianthus caryophyllus (carnation), which is cultivated around the world. Others in^Present address: Complex Carbohydrate Research Center, The University of Georgia, 220 Riverbend Road, Athens, Georgia 30602-4712, USA

elude Gypsophila, Saponaria, and Silene genera. The roots of several Gypsophila species (commercially known as Saponariae alba radix) are used for the production of a number of industrial products viz detergents and chemicals, film emulsion, and as ingredients of fire extinguishers. Saponins of this family are based on the P-amyrin type of triterpenoid aglycones (Fig. 1) with one (monodesmoside) or two (bisdesmoside) sugar chains. The oligosaccharide chains are mostly branched rather than linear, and possess as many as 9 to 10 monosaccharide units. Most of the saponins have sugars attached to the hydroxy group at C-3 and the carboxy group at C-17. The linkage of a glucuronic acid to C-3 and of a fucose unit (often acetylated) to C-28 of the aglycones is a common feature of the Caryophyllaceae family. Unlike most of the triterpenoid saponins found in other families, Caryophyllaceous saponins have the tendency to have more sugar constituents at C-17 than at C-3. The presence of only an acyloside sugar chain, i.e. one attached to the carboxy group at C-28 of the aglycone, is not uncommon. The last review on the Caryophyllaceae family was published almost two decades ago [1]. Since then, purification and structural elucidation techniques have changed enormously. Many of the earlier chemical studies have been reinvestigated and new findings published. The present review deals with the chemistry of the triterpenoid saponins of Caryophyllaceae family with special emphasis on recent developments in purification techniques and structural study aspects. GENERAL METHOD OF ISOLATION The defatted, air-dried plant material is extracted with methanol, either cold or hot, or with 50% aqueous methanol at ambient temperature. Fresh plant materials have to be processed immediately after collection to prevent fungal growth and enzymatic hydrolysis of the saponins. During extraction, care should be taken as saponins may undergo transformation, e.g. esterification of acidic saponins, hydrolysis of labile ester groups or transacylation, etc. Often the crude extract is suspended in water and partitioned using ethyl acetate and «-butanol successively. Most of the saponin constituents are found in the w-butanol solublefi-action.However, the low-polar saponin constituents may be present in the ethyl acetate part while the most highly polar saponins may be found in the aqueous layer, as in the case oi Saponaria officinalis [2]. Saponins are usually highly polar compounds occurring as complex mixtures, and their separation into individual components is a formidable and time-consuming task. The traditional purification method consists of repetitive chromatography on silica gel columns using chloroform-methanol-water

as eluent, and the method is being widely used even today to get rid of pigments and other non-saponin constituents. Such process may yield pure products in few cases; but usually it separates the crude saponin mixture into different fractions according to their polarity, final purification being achieved by high performance liquid chromatography (HPLC). Following is the method of purification for saponins which is usually adopted by various workers including us. The crude saponin mixture is passed over a column of Diaion HP-20 using water-methanol as eluent in various ratios (0, 10, 30, 50, 70 and 100% in methanol). Often the saponin fraction is obtained from the 70-100% methanol eluates. The saponin fractions thus obtained sometimes afforded pure compounds on repeated silica gel and ODS chromatographic purification. However, in most of the cases, fractions found to be homogeneous on TLC turned out to be mixtures on HPLC analysis. Final purification is usually done by HPLC over a reversedphase column (Cig or so-called ODS) with methanol-water or acetonitrilewater as eluent. Saponins of Caryophyllaceae family usually contain a free carboxyl group either in the aglycone part or in the monosaccharide chain (the so-called acidic saponins), and suppression of ion formation is necessary otherwise peak broadening may occur and separation becomes impossible. This can be achieved by adding a small amount of trifluoroacetic acid (TFA) or acetic acid to the solvent system. We usually use 60-75%) methanol containing 0.06%) TFA and the system works alright in most of the cases. However, care has to be taken while recovering the saponins from the acidic medium. Removal of the acid under reduced pressure and at a relatively lower temperature (<40^C) was found to be safe even for saponins with acetate substituents [2]. GENERAL STRATEGY FOR STRUCTURE ELUCIDATION Saponins can be regarded as composed of two units, an aglycone and an oligosaccharide moiety. Saponins of Caryophyllaceae family are almost wholly based on the P-amyrin skeleton with a hydroxy group at C-3 and a carboxy group at C-17 (Fig 1). In all the aglycones reported so far, C-23 exhibited different degrees of oxidation varying from -CH3 to -CHO (gypsogenin, II), -COOH (gypsogenic acid, IV) or even -OH (nor C-23, segetalic acid, IX). Besides, the aglycones may have an additional a-hydroxy group at C-16 (quillaic acid. III). The most common aglycones found in this family are gypsogenin, gypsogenic acid and quillaic acid. Aglycones of lupane skeleton have also been encountered in this family, but are limited to a single plant only [3-5]. Other types of aglycones, viz. 3,4-secogypsogenic acid (XI), and 3P-hydroxy-oleana-ll,13(18)-dien-23,28-dioic acid (VIII) were also isolated

[6].

COOH

CHO Gypsogenin R = H(II) Quillaicacid R = OH(III)

p-Amyrin (I)

COOH

COOH

COOH

COOH

Gypsogenic acid R = H (IV) 16a-Hydroxygypsogenic acid R = OH (V)

Medicagenic acid R = H (VI) Zanhicacid R = OH(VII)

COOH

COOH

COOH 3p-Hydroxy-oleana-l 1,13(18)dien-23,28-dioic acid (VIII)

Segetalic Acid R = OH(IX) Vaccaric Acid R = H (X)

COOH

COOH 3,4-Secogypsogenic acid R = H (XI) 16a-Hydroxy 3,4-secogypsogenic acid R = OH (XII)

Lupanoid type aglycones (XIII) R,,R3 = HorOH R2 = CH20HorCOOH

Fig. (1). The aglycones of the triterpenoid saponins from Caryophyllaceae family.

However, the chances of finding new sapogenins are becoming less and less, the new saponins usually varying only in the component sugars and nature of the sugar sequence. Therefore, in most of the cases, the structure determination of saponins has amounted to identifying the sapogenins and establishing the sugar sequence. The sugar units found in these saponins are D-glucuronic acid (GluA), D-glucose (Glc), D-galactose (Gal), D-fucose (Fuc), Dquinovose (Qui), L-rhamnose (Rha), D-xylose (Xyl), or L-arabinose (Ara) (in pyranose or furanose forms). Chemical characterization of the trierpenoid saponins of the Caryophyllaceae family started as early as in 1930s. Historically, the structural study of saponins can be roughly divided into three stages: (A). Traditional chemical methods (before the 70s); (B). Spectroscopic methods (^H and ^^C NMR techniques) combined with traditional chemical means (from 1970s to 1980s); and (C). Modem multipulse NMR approaches (since 1980s). All these techniques have been employed in the chemical investigation on this family of saponins. Traditional chemical methods The early chemical investigation on the saponins of the Caryophyllaceae family usually began with complete hydrolysis of the isolated saponin, leading to identification of the sapogenin and the individual sugar units. The sugars were identified by PC or TLC comparison, or by GLC analysis of their TMS derivatives. The interglycosidic linkages were established through permethylation of the saponin prior to complete hydrolysis. The nonmethylated sites of the hydrolysis products revealed the points of linkages. The sequence information could be deduced by stepwise hydrolysis of the saponin, identifying each prosapogenin thus obtained. Such processes are tedious and need large amounts of pure samples. Sequence and linkage information could also be inferred from Smith degradation [7, 8]. Sugars with a glycol structure are susceptible to periodate oxidation. If such glycol segments were present in an oligosaccharide chain, cleavage of the chain could be effected at these positions by treatment with periodate followed by reduction and mild acid hydrolysis. In a phased manner, the long oligosaccharide chain could be dissected into smaller and more readily analyzed units and the structure of the oligosaccharide chain thus established. Spectroscopic and chemical methods With the advent of modem spectroscopic methods, especially FT-NMR and

soft ionization MS methods (FD and FAB MS), the structural study of saponins shifted away fi-om the time-consuming and sample-demanding chemical methods. Mass spectroscopy can provide information about molecular weight and the numbers of the monosaccharides, and in some cases information about the sequence of the oligosaccharide chain. Analysis of the MS and ^^C NMR spectra, and comparison of the data with similar structurally related compounds available in the literature sometimes furnished information about the oligosaccharide moiety as well as the aglycone also. The most important information obtainedfi*om^^C NMR is regarding the linkages between the sugar units. The so-called 'glycosidation-induced shift' method is quick and easy to apply, and is widely used even today. Usually the glycosylation shifts in ^^C NMR are relatively regular. The glycosylated carbon shifts to lower field by 4-10 ppm (the a-effect). The resonances of the adjacent carbons are often shifted upfield by a smaller magnitude, while other carbons in the same molecule remain virtually unaffected. These glycosylation effects depend on the configuration at the anomeric center of the glycosidating pyranose and the absolute configuration of both the residues. However, we came across an exception to this rule during our investigation on the saponins from Ardisia crenata [9], namely the chemical shift of glucose C-2 remains virtually unchanged when it is glycosylated by a rhamnose. The difference is too small and abnormal to predict a glycosidation site. Literature survey revealed that similar situation existed in some glycosides possessing rhamnose 1-^2 glucose or galactose units. Therefore, application of the glycosylation shift rule should be treated with caution in such cases. Our molecular modeling study showed that the lack of glycosylation shifts in C-2 of Glc in the Rha-al->2 Glc-p fragment could be explained by the changes in the corresponding torsion angles. For simple oligosaccharides, use of ^^C NMR and MS techniques in combination with some chemical reactions is good enough to establish the structure. However for branching, complex oligosaccharide chains, these methods cannot provide exact structure at the branching point, for the glycosylation shifts can only indicate that a substitution occurred but provides no information about the nature of the substituents. This problem can be overcome by examining the products from mild acidic (partial) hydrolysis or enzymatic degradation. Under such condition, the oligosaccharide chain will usually be cleaved at either of the branching points and lead to the formation of corresponding linear products. Comparison of the ^"^C NMR data of the degraded products with the original one immediately establishes the oligosaccharide structure. The anomeric configurations can also be deduced from Klyne's rule of molecular rotation [10] and ^H, ^^C NMR data.

Modern NMR approaches The advent of modern multidimensional NMR techniques based on high field magnet and versatile computer programs has greatly advanced the structural study of saponins. The structural characterization of saponins involves determining: (i) the structure of the sapogenin; (ii) the nature and number of the constitutent sugar units including the ring size and anomeric configuration; (iii) the interglycosidic linkages and the sequence including the location of the sugar chain attachment to the sapogenin, and (iv) the 3D structure. The following discussion mainly focuses on the structural study of the sugar moiety. The first step in the structural analysis of a saponin is to obtain the ID ^Hand ^^C-NMR spectra. Saponins are usually investigated as deuterium exchanged samples and the most commonly used solvent is pyridine-^is although use of DMS0-(i6 or methanol-(i4 is reported in the literature in few cases. The hydroxyl protons can be exchanged by adding several drops of D2O or deuteriated methanol. The ID ^H NMR spectrum displays only some recognizable signals, especially those for the anomeric protons at 4.5-6.4 ppm, methyl doublets of 6-deoxy sugar units at 1.3-1.5 (e.g. H-6 of rhamnose), and the methyl groups of the aglycones. The ^^C NMR spectra give a better dispersion over a 200 ppm range; additionally the protonation levels are deducible fi-om a DEPT experiment [11]. The sugar anomeric carbons can be located at 6 90-110 and the non-anomeric carbons at 8 60-83 ppm. Most resonances of the aglycone part appear below 6 55 in the higher field region, the oxo-substituted carbon signals appearing in the same range as the non-anomeric sugar carbons. The signal for C-3 is usually found at about 5 84 ppm and that for C-28 at about 5 176 for the acylosides. For the (3amyrin type sapogenins, the signals for the A^^ double bonds are typically observed at ca 5 122 (C-12) and 144 (C-13). ID ^H and ^^C NMR spectra usually give partial structural information only, such as the nature of the sapogenin and the number (by counting the anomeric protons and carbons) of sugars in the molecule. The identification of constituent sugars The identity of sugars and the sugar sequence of the oligosaccharide can be determined by a combination of COSY [12], HOHAHA [13, 14] or TOCSY [15], HETCOR [16] or HMQC [17], HMBC [18] and NOESY [19] techniques. The prerequisite for the identification of the component sugars is the unambiguous assignment of the ^H signals. The through-bond connectivties of the ring protons within each sugar residue can be established by COSY

10

and HOHAHA experiments. Since the anomeric protons of each residue resonate in a characteristic region well isolated from that of the nonanomeric protons, they are usually the starting points for analyzing the spectra. In favorable cases (i.e. less than four or five sugar units), a COSY, preferably DQF-COSY [20] spectrum, which reveals the connectivities for each pair of vicinal and geminal protons, may map out all the spin systems. For saponins with more complex sugar moiety, a HOHAHA experiment can be of further help in the assignment. The individual spin-system can be discerned from the sub-spectra corresponding to the anomeric or the methyl groups (for the deoxy sugars) in the HOHAHA spectrum. Fig. (2) is the HOHAHA spectrum of saponarioside C isolated from Saponaria officinalis [21]. As seen from the example, HOHAHA has the potential of reducing the composite oligosaccharide ^H-NMR spectrum into a subset of spectra. Sometimes, several HOHAHA experiments with different mixing times may be necessary to trace the spin systems from the anomeric to the terminal proton step by step. However, the distribution of magnetization within the spin system can be impeded by small coupling, such as typically found between H-4 and H-5 in a galactosyl residue (system G'" in Fig. 2), though it has been reported that ID-HOHAHA has the potential to obtain correlation past this threshold [22].

-V

i_A

X

XiLj

Q B

n K IS

'W

-i-

^»

Fig. (2). The HOHAHA spectrum (!„, = 150 ms) for the sugar moiety of saponarioside C from Saponaria officinalis showing the sub-spectra corresponding to each anomeric proton. Note that for the glucose (G, G', G") and xylose (X) units, the correlations were observed from the anomeric protons till terminals. However, the correlation stopped at H-4 of the galactose (G"') moiety due to the small coupling between H-4 and H-5. Spectrum measured at 500 MHz in pyridine-tis.

11

Additionally, the intra-residue NOEs from 2D-N0ESY or ROESY experiments could furnish further information concerning the proton assignment. In general, 1,3-diaxial and 1,2-eq- ax proton pairs in pyranosyl rings produce intra NOE cross peaks, viz. for P-galactopyranosyl residue NOEs are observed between H-1 and H-3 (and H-5), while only the H-1 to H-2 intraresidue NOE is noticed in a-galacto- pyranosides; these greatly simplify the mapping of the spin systems. After all the spin systems are mapped out and the number of different spin systems established by the above methods, one can go ahead to identify the sugars as well as their anomeric stereochemistry. In pyranosides, the six-membered ring generally forms a stable conformation providing a classification of protons as axial or equatorial. Therefore, the coupling patterns are characteristic of the stereochemistry for this type of sugars. The VHH coupling constants are large (> 6 Hz) for the anti-periplanar orientation of the vicinal ring protons (Z H-C-C-H « 180°), and small (< 5 Hz) for the gauche orientation (Z H-C-C-H « 60°). Several methods have been proposed to measure the proton coupling constants [23, 24]. For conformational study, the coupling constant should have a high accuracy. Such accuracy is not required for the identification of the sugar residues. But it is necessary to distinguish between small and large J values. In order to prevent misassignment, it is recommended that the sugar composition be confirmed by chemical analysis, and the molecular weight confirmed by FAB or MALDI mass spectroscopy. Sugar identity and stereochemistry can also be inferred from ^"^C NMR data. If the *H spectrum be completely assigned, a HETCOR (^^C detected) or HMQC/HSQC (^H detected) experiment would be enough to establish the ^•^C signal identity. Sugars can be identified by comparing their ^^C data with those of the standard methyl glycosides or the published data of saponins. As glycosidation tends to shift the a-carbon to lower field and P-carbon to somewhat higher field positions, careful attention should be paid to the determination of the ring size and anomeric configuration on the basis of chemical shifts as these may lead to ambiguity. In addition, a reliable criterion for determining the anomeric configuration of D-sugars in the "^Ci pyranose form is from one- bond ^^C-^H couplings (^JCH) [25]. Determination of the inter-glycosidic linkage and sugar sequences Once each sugar residue has been identified and its anomeric configuration determined using a combination of the above-mentioned techniques, all that is required is to identify the sugar sequence and the inter-glycosidic linkage. Proton-proton scalar couplings that are four bonds apart (VHH) are usually too small to observe. Therefore, it is necessary to make use of either homo-

12

nuclear dipolar coupling (NOE measurements), or the long-range heteronuclear coupling constant VCH across the glycosidic linkages. Once the ^H assignment is complete within each individual sugar spin system, measurement of NOE enhancement provides a very powerful means for determining linkage and sequence. The presence of an inter-glycosidic NOE from the anomeric proton of a particular sugar residue to a proton of the other sugar residue or non-sugar residue (sapogenin) defines the glycosidic linkage between the two residues. NOE connectivities are most often observed between the anomeric proton and the proton connected to the carbon atom of the linkage. 1

II glc ^^^^--^^^

r 1

B.2

93'

^--

6.0

5.8

5.6

H.12

S'c

xyl

glc

^—U__AJI_

5.4

5.2

5-0

H-3Q G'"-2

/f"^r'S^%o

G-3

" ^ ^

G.5

0 G"-6

G-3 Q G'-3 0 x-3 G'.S

y

^-

^ ^G".S

X.5 ©

1

if

1

HOr^" p-D-Glc'(G') OH

H GOGH

H ^ .H ^ SC i ( >L^ S HO P-D-Xyl (X)

H^;. Q^Y^Qj/^OH's^Q

P-D-GIc (G)

' OH a-D-Gal (G'")

P-D-GIc" (G")

Fig. (3). Phase-sensitive NOESY spectrum of saponarioside C from Saponaria officinalis (the mixing time was 600 msec). Note that both the inter-residue (bold labeled) and intra-residue NOEs were observed. For the a-D-galactose moiety, the only intra-residue NOE noted is between Hi and H2. For all other sugar units, the intra-residue NOEs between H1/H3 and H1/H5 are observed.

13 This has been found to be of wide appHcabihty for the structure determination of the naturally occurring glycosides. An example of a NOESY spectrum is given in the section on the structural study of saponarioside C isolated from Saponaria officinalis (Fig. 3) [21]. Both the inter-residue NOEs and the intra-residue NOEs were observed. The conventional NOEs in the laboratory frame can be positive or negative and pass through zero when cooXc, the product of spectrometer angular frequency and molecular rotational correlation time (which depends on the size and shape of the molecules and on the viscosity of the rotating medium) is approximately equal to unity. Such a problem, which is typical of middlesized molecules like saponins, can be solved by performing the experiment in the rotating frame, the so-called ROESY [26, 27]. In NOE-based linkage and sequence analysis, it is necessary to be certain that all resonances are unequivocally assigned. Even then, the observed NOEs may be inconclusive, if the chemical shift for the aglyconic proton located at the glycosylated carbon coincides with the chemical shifts for protons of other sugar residues. This situation is encountered very often with complex glycosides. Another serious drawback is that the interglycosidic NOE between the anomeric proton (H-1) and the proton on the attachment site of the sugar or the aglycone is not necessarily the only and the largest inter-residue NOE effect. Therefore, NOEs should not be used as the sole criterion for establishing the position of a glycosidic linkage, especially in the branching center. An example is the saponin mimusopin isolated from the seeds of Mimusops elengi [28]. The structure was established by a combination of COSY, HOHAHA, HETCOR, HMBC and phase-sensitive NOESY. The preferred conformation calculated from molecular modeling study for the sugar chain at C-28 of the sapogenin is shown in Fig. (4). The NOE relationship from NMR is indicated by arrows. Besides the NOEs between protons across the glycosidic bonds (R1/A2, RVX3, X1/R4, R'VRs), NOEs between R1/A3, R"i/R2, Xi/R"2 as well as between A1/R5 were observed. Before looking at the molecular modeling result, we were not able to make the inter-glycosidic assignment by using NOE data only. By careftilly examining the preferred conformation, we found that the arabinose adopts a ^C^ conformation placing H-3 in the equatorial position (vicinal to the linkage site). NOEs from such environment (another example is H-4 in 3-glycosylated galactose residue) have been reported and well discussed before [29, 30]. Such protons may exhibit NOEs of approximately the same magnitude as the rest and determination of the linkage might appear equivocal. The NOEs from the pair R"i/R2 is of the same origin. R2 is in an equatorial position vicinal to the linkage site at C-3. However, the observed NOEs between Xi/R"2 and A1/R5 appear somewhat unusual. Molecular modeling results showed that the distance between these two pairs are 2.64 and 2.41 A, re-

14 spectively, within the effective NOE distance. This means that substitutions at both C-3 and C-4 cause the branching center to be very crowded and distort the torsion angles. In fact, anomaly is also found in the ^"^C NMR data of the disubstituted rhamnose (R). Molecular modeling results showed that the corresponding dihedral angles were distorted compared to another saponin, isolated from the same source but lacking the substitution at C-3 of rhamnose (R) [28]. From the above discussion, we can see that using only the observed inter-residue NOEs may lead to wrong conclusions about the primary structure of the sugar chain.

The Optimized Conformation of Mimusopin from Molecular Modeling Study

2.29 A

'R"ha NOEs observed between the fragments Rhal~2Ara and R"hal-3 Rha of Mimusopin Fig. (4). The calculated preferred conformation (part) of mimusopin and NOEs observed from phasesensitive NOESY.

15 Another, more effective way to determine the sugar Hnkage and sequence is to detect the long-range VCH coupling across the glycosidic bond. The most practical technique is HMBC (heteronuclear multibond correlation). An HMBC experiment can furnish inter-glycosidic multi-bond correlation between the anomeric proton and the aglycone carbon, and thus serve to identify the linkage. Besides, HMBC also furnishes intra-residue multi-bond correlations, which are very valuable for confirming the H and C assignments. As with proton-proton vicinal coupling constants, the three-bond carbon-proton couplings as observed in HMBC also conform to the Karplus relationship, the maximum of VCH being usually observed at a dihedral angle of 180° and the minimum near 90°. Therefore, HMBC also furnishes information concerning anomeric configurations and the overall conformations. For example, for a-L-rhamnose possessing an equatorial anomeric proton, intra-residue multi-bond correlations between the anomeric proton and C-3, C-5 are observed (the dihedral angles between H-1 and C-3, H-1 and C-5 are about 180°). The same is true for the a-D-galactose unit of saponarioside C found in the plant Saponaria officinalis [21]. 3D-Structure by NMR and Molecular Modeling studies It is generally believed that biological activity is determined by the 3D structure (conformation) of the compounds. For saponins, the most flexible and biologically important part is the sugar unit. Thus, conformational analysis of the sugar part is crucial for the elucidation of the biological action of these kinds of compounds. The conformational properties of sugars in solution have been largely determined by ^H NMR NOE techniques. It originates in dipolar cross relaxation between protons, which is a function of distances and molecular motions. Because of the inverse sixth power dependence of the NOEs on the interproton distances (r), a highly sensitive conformational "ruler" is available when r is small (< 5A). The most important NOE contacts in sugars are those between protons across the glycosidic bonds. Sometimes other through-space contacts, which are very important in probing local conformations, are also observed. The latter situation is not uncommon in a branching center with attachment of two or more sugars to the same aglycone. Another probe for the three-dimensional structure is the ^JCH value across the glycosidic bonds [31, 32]. This value is an indication of the dihedral angle between the sugars. Thus the local conformation can be determined from the VcH magnitudes. Additionally, ^"^C chemical shifts of the carbons involved in the interglycosidic linkages can also be used as an indication con-

16 ceming the local conformation. But the specific relationship between conformation and glycosidation shifts has not been fully understood yet [33]. In recent years, with the development in computational methods, molecular mechanics (MM) and molecular dynamics (MD) calculations have provided a powerful tool for evaluating intermolecular distances and dihedral angles for comparing with the results obtained from NOE data and VHH measurements respectively. Analyzing 2D NMR data using the energyminimized structures usually leads to more reliable assignments [28]. Fig. (5) is a summary of the general guideline of the modem NMR techniques employed in the structural study of triterpenoid saponins in our laboratory.

'HNMR

13

CNMR

COSY, HOHAHA: Mapping out all of the proton J-coupled spin system of each sugar units

DEPT: assign ^^C multiplicity HETCOR/HMQC; assign the C signals

NOE/ROESY (through space interactions)

HMBC (long range heteronuclear ^H,^-decouplings)

Assign the inter-residue connectivites using NOE and HMBC

Establish the stereochemistry from NOEs and molecular modeling study

Fig. (5). Structure elucidation strategy of saponins from 2D-NMR approaches.

17 TRITERPENOID SAPONINS FROM SAPONARIA OFFICINALIS Saponaria officinalis L., popularly known as fuller's herb or soapwort, is native to Europe and western to central Asia, and is presently cultivated in different parts of the world for its medicinal properties. This plant was well known for its detergent property and was used as soap in ancient times. In folk medicine it is used for various skin diseases, rheumatic disorders and as an expectorant for bronchitis [35]. In Europe, different parts of the plant, viz. the roots, rhizomes and shoots are used for the preparation of herbal medicines [36]. It has long been known that the bioactivity of the plant is due to the triterpenoid saponin constituents. Earlier chemical investigation during 70s led to the isolation of four saponins; the structures of two major ones, saponasides A and D, were established partially, with the latter containing up to ten sugar units (Fig. 6) [37, 38]. The structures were elucidated by various chemical methods such as permethylation, hydrolysis, periodate oxidation, partial hydrolysis and LiAlH4 reduction. During the same period, one more saponin designated saponaroside was isolated from the same source and its structure was established too (Fig. 6) [39]. Since then, no further reports are available on the chemical investigation on the saponin constituents of the plant.

'COOR2 R,0' CHO

Saponaside A Saponaside D

D-GluA

D-Glc-(l-6) I D-Glc-( 1-3)1 D-Glc

D-Gal-(l-2) I L-Ara-(l-4) I D.Xyl-(l-3) I ^ , ^ L-Rha-( 1 -4) I D-GluA

Saponaroside

D-Glc-(1-2)| D-Gal-(1-4)| D-XyKl-S) I L-Rha-( 1 -4) | D-Fuc

COOH

COOH O H ^ H

Fig. (6). Saponasides A, D and saponaroside from Saponaria officinalis.

18 Chemical investigations on the sapogenin component have been carried out as early as 1910s and gypsogenin was proposed as the sapogenin [40]. Some confusion concerning the structures of the sapogenins lasted until the 1980s, when Kubota et al. could isolate only quillaic acid and gypsogenic acid as the sapogenin constituents from S. officinalis [41]; very recently Henry et al. confirmed the presence of quillaic acid in the rhizomes using ^^C NMR techniques [42]. The medicinal and commercial importance of the plant coupled with our continuing interest in the chemistry of triterpenoid saponins prompted us to reinvestigate the saponin constituents of S. officinalis [2, 6, 21]. The MeOH extract of the freshly collected whole plant of S. officinalis was suspended in water and partitioned successively with EtOAc and nBuOH. The aqueous part, on chromatography over Diaion HP-20 followed by repeated MPLC and HPLC purification afforded two major triterpenoid saponins, saponariosides A and B. Similarly, the «-BuOH soluble fraction afforded six triterpenoid saponins designated as saponariosides C-H. Investigation of the plant material collected from different geographical locations led to the isolation of previously reported saponariosides C, E, F and G along with five more new saponins, saponariosides I-M (Fig. 7), from which two new sapogenins have been characterized as VIII and XII (Fig. 1). However, saponarioside D, the major constituent of the «-BuOH soluble fraction as reported previously by us [21], could not be detected. Saponariosides A and B, the major constituents isolated from the aqueous part, are highly soluble in water and water-MeOH mixture but have poor solubility both in MeOH and EtOH. They are bisdesmosides of quillaic acid, glucuronic acid being linked to C-3 of the genin. The other sugar constituents are galactose, xylose, and three 6-deoxy hexoses (rhamnose, fucose, and quinovose). It is of interest to note that both the saponins possessed one acetyl function appended to the quinovose residue at C-4. The saponins isolated from the «-BuOH soluble fraction have aglycones other than quillaic acid, and displayed a distinct and characteristic color (sky blue) on silica gel TLC plates when sprayed with 5% H2SO4 and heated. The aglycones are either gypsogenic acid or 16a-hydroxy gypsogenic acid, with a free carboxyl group at the a-face of C-4. Saponarioside C has a somewhat different substitution pattern and possesses an unprecedented a-D-galactose moiety linked to C-6 of the inner glucose. The minor compound saponarioside H, with only one glucose unit at C-28, is presumed to be the biosynthetic intermediate for the other saponins. Due to the acidic nature of these saponins, a small amount of TFA (0.05% v/v) had to be added to the mobile phase for HPLC purification of the saponins.

19

Saponariosides

GIcA (I)

HO' " ^ ^ ^ ^ " HO Xyl(D) HO '

o OH

HO

Rha (A)

Gal (C) A R = Xyl(H) B R=H HO HO Xyl (F) HOfOH

HO OH

,X:^°'

HO HO

p.Gic(G)

^^^^^^s:;^z:oH/0

OH

COOH

OH

P-Glc (G")

a-Gal (G"')

P-Xyl C

R=H

I

R = OH HCK-OH ^ ^ n ^ O H

"N/

f^^^^^^Y^^

^^

/

/

G' "^OH

^ R2O'

G tOOH R2

\

RaO-T--^ ^ O H O ^ ^ O ^ ^ O H

R3

D

H

Xyl

Glc (G'")

E

H

Glc

Glc

F

OH

Xyl

Glc

G

OH

Xyl

H

20

Minor saponin

OH OH OH

COOH

COOR

' " H ^ ^ : ^

COOH

^^^^^

HO/fOH _ OH O

COOR

HCV^li^r^ HO^^'^OH

COOH

R=

Ho:

^

G

OH .OH HOT^---^^^'' \2-»0,^^^^/^0H G"

WO'^ST^^ir^

' COOH

OH

OH OH

Fig. (7). Triterpenoid saponins from Saponaria officinalis.

Saponarioside A is a white, bitter-tasting powder that induces sneezing. It produces a stable froth when shaken with water. The molecular formula C82H128O45 was determined from its negative ion HRFAB-MS ([M-H]" peak at m/z 1831.7649) in combination with ^^C DEPT NMR data. Of the 82 carbons, 30 were assigned to the aglycone part, 50 to the oligosaccharide moiety and the remaining two to an acetoxy group. The IR spectrum showed absorptions at 3406 cm"^ (-0H) and 1728 cm"^ (ester carbonyl). The six methyl

21

carbon signals at 5 11.1, 15.8, 17.4, 24.5, 27.0 and 33.2 ppm, and the two sp^ carbon signals at 5 122.2 (d) and 144.4 (s), coupled with the information from ^H NMR (six methyl proton singlets and a broad triplet for a vinyl proton at 5 5.54), indicated that the aglycone had an olean-12-ene skeleton. After an extensive 2D-NMR analysis, the aglycone was identified as quillaic acid [43, 44]. The downfield shift of C-3 signal (6 84.4) and upfield shift of C-28 signal (5 175.9) compared to those of the aglycone indicated that it was a bisdesmoside. The ^H and ^"^C NMR spectra displayed signals for nine anomeric sugar protons [8 4.89 d (J= 7.3 Hz), 4.99 d (2H, J = 7.6 Hz), 5.13 d (2H, y = 7.1 Hz), 5.32 d (J= 7.7 Hz), 5.55 d (J= 7.3 Hz), 5.93 d (J= 8.2 Hz), 6.29 s] and carbons (5 94.4, 100.9, 103.8, 104.2, 104.9, 105.5, 105.8, 106.2, 106.9), Fig. (8). Alkaline hydrolysis furnished a prosapogenin, identified as quillaic acid 3-0-P-D-galactopyranosyI-(l->2)[P-D-xylopyranosyl(l->3)]-P-D-glucurono-pyranoside from its spectral data. Acid hydrolysis afforded quillaic acid, and the monosaccharide components were identified as fucose, galactose, xylose, quinovose and rhamnose in the ratio 1:1:4:1:1 from GLC analysis of their TMS derivatives. The other monosaccharide was identified as glucuronic acid by co-TLC analysis (both saponarioside A and the authentic sugar were applied to the TLC plate and then hydrolyzed under HCl vapor at 65^C for 1 h; developing solvent: CHCb-MeOH-HsO, 10:5:1). From the above evidences, it was concluded that saponarioside A was a bisdesmosidic triterpenoid glycoside with glucuronic acid, galactose and xylose linked to the C-3 position of the aglycone, the other six monosaccharides being linked to C-28 of the aglycone through an ester bond. The sequence of the oligosaccharide chains was established using the NMR protocol as discussed in the previous section. In order to facilitate proton assignments, the nine monosaccharides were labeled by the letters A to I. The resonance position for anomeric protons corresponding to E and F as well as G and H appeared overlapped, thus making the proton assignment difficult. The individual spin-systems were discerned from the subspectra corresponding to the anomeric protons or the methyl groups (for the deoxy sugars) in the HOHAHA experiment (Xm = 150 ms) (Fig. 9). Interpretation of the COSY and 2D-H0HAHA spectra revealed the presence of nine monosaccharide units. As shown in the HOHAHA spectrum, three monosaccharide units (A, B and G) were 6-deoxy sugars. Spin-system of A displayed two broad singlets (H1: 5, 6.29 s and H-2: 4.68 s), a double doublet (H-3: 5 4.60, dd, 8.5, 3.2 Hz), two overlapped signals for H-4 and H-5 protons and a terminal doublet for a methyl group (6 1.53, d, 5.8 Hz) (Fig 9). Furthermore, the large ^CH (171 Hz) and strong three-bond HMBC correlations from the anomeric proton to C-3 and C-5 indicated the equatorial orientation of the anomeric proton, thus suggesting a a configuration.

22

-T—I—I—i—r

V W V * l i n - ^ V V'rf^i'Ww^^i^vn^'N ' \V^)^^»^»»>M[g.v^lV»^««»^WWrt^M»>•^^>V^^

""'|""""'l

210

200

"'I'

I'"

I'"

190

180

170

|"M.iii.|

160

150

|.

140

|M.mM.|....nM.|

130

120

110

'^/•f

|.nmni|mM.iM|immM|rimiiM|Mmim|mmm|immM|MiMim|iiiiiiiM|.M

100

Fig. (8). The ^H and ^^C spectra of saponarioside A.

90

BO

70

50

50

40

30

20

10

23 E(Xyl') G(Qui) F(Xyl")-H(Xyl"')"••*^'»\7G4(0«I)/ I (GJcA)

I A (Rham) B(FllC)

6.0

IM2(agl)

CU(Gan (GJ

5.8

^ 4

%

%2

F5

t$

1^3

3&4"

8

«"^i3

A

* ¥.

<4^ 4

7^

.iF^'

G5 H5'

E5'-SS

_E(Xyr) G(Qui) _ F(Xyl") H(Xyl"')

0 1116/15 (agl)

Fig. (9). HOHAHA (Tm=150 ms) spectra (part) of saponarioside A (500 MHz in pyridine-^5) showing the subspectra corresponding to each anomeric proton (upper) and methyl group (lower). Note that the resonances for anomeric protons of E and F, G and H overlapped.

24

The above information coupled with ^^C NMR data identified the spinsystem of A as a-rhamno-pyranose. For the B-system, only three crosspeaks could be traced from the subspectrum corresponding to the anomeric proton (6 5,93, d, 8.2 Hz). However, a careful examination of the sequence corresponding to the methyl proton doublet at 6 1.48 suggested that the system consisted of a terminal methyl group with five methine protons. As indicated in the HOHAHA spectra, the coupling constants between H-5 and the methyl group was high but the coherence transfer was barely propagated through H-5 and H-4 due to the small JH4,H5 coupling. Combination of the above facts with NOE relationships (between H-1 and H-3, H-1 and H-5) as well as ^^C NMR data allowed us to conclude that B was (i-fucopyranose (6deoxy-(3-galactopyranose). The other deoxy sugar (G) showed couplings from the anomeric proton (8 4.99, d, 7.6 Hz) through the system including the terminal methyl group (6 1.26, d, 6.1 Hz). An acetyl group attached to C4 of the system was evident as the H-4 signal (5.08, dd, 9.4, 9.8 Hz) was well isolated from other methine proton signals due to the shift effect of the acetoxy. Moreover, the observed long range HMBC correlation between H-4 and the acetyl carbonyl carbon (170.2 ppm) also supported the attachment of the acetoxy function at C-4. The coupling patterns of H-4 (dd, 9.4, 9.8 Hz) suggested axial disposition for H-3, H-4 and H-5. NOESY showed significant through space interaction between H-1 and H-3 as well as H-5. The above information suggested that the G-system was p-6-deoxy glucose, i.e. p-quinovopyranose. Proceeding in a similar manner, the other six sugar units were identified as p-glucopyranuronic acid (I), p-galactopyranose (C), and four p-xylopyranoses (D, E, F, H). It is noteworthy that in addition to fiirnishing the information concerning the inter-sugar linkages as discussed below, HMBC also significantly helped to locate some of the closely related proton and carbon signals especially for monosaccharides E and F, G and H in which the anomeric proton and carbon signals appeared overlapped. The linkage of the sugar units at C-3 was established from the following HMBC correlafions: H-1 of C spin-system (galactose) with C-2 of I (glucuronic acid); H-1 of D (xylose) with C-3 of I (glucuronic acid). The attachment of the trisaccharide moiety to C-3 of the aglycone was confirmed by the observed long-range correlation between H-1 of glucuronic acid and C-3 of the aglycone. The sugar chain at C-28 was established from the following HMBC correlations: H-1 of F (terminal xylose) with C-3 of E (inner xylose); H-1 of E with C-4 of A (rhamnose); H-1 of A with C-2 of B (fiicose); H-1 of H (terminal xylose) with C-3 of G (quinovose); and H-1 of G with C-4 of B (fucose). The attachment of the hexasaccharide chain to C-28 of the aglycone was based on observed correlation between H-1 of fticose (B) and C-28 of the aglycone. The same conclusion with respect to the sugar sequence was also drawn from the NOESY experiment. However, due to the

25 highly overlapped nature of the proton spectrum, NOEs could not be used as the sole source of evidence for the inter-sugar linkage. The linkage was also supported by the fragmentation patterns observed in the ESI-MS/MS experiment. MS/MS analysis of the deprotonated molecular ion [M-H]" {m/z 1831) gave a daughter ion at m/z 1699 [(M-H)"-132] due to the loss of one of the terminal pentoses (xylose). The most prominent fragment observed at m/z 955 was due to the loss of the hexasaccharide unit linked to C-28 of saponarioside A (Fig. 10). All monosaccharides were in the pyranose form as determined from their '^C NMR data. The p anomeric configurations for the fucose, galactose, glucuronic acid, quinovose and xylose were evident from their VHI,H2 (7-8 Hz) and VCI,HI coupling constants as well as from NOE information. The broad singlet observed for the anomeric proton, the large ^JCI,HI (171.7 Hz), and the three-bond HMBC couplings noticed between the anomeric proton and C-3 or C-5 of the rhamnose indicated a orientation of this monosaccharide. The absolute configurations of these sugars were chosen in keeping with those mostly encountered among other plant glycosides. Thus, the structure of saponarioside A was established as 3-0-PD-galactopyranosyl-(l ->2)-[P-D-xylopyranosyl-(l ->3)]-p-Dglucuronopyranosyl quillaic acid 28-0-p-D-xylo-pyranosyl-(l ->3)-P-Dxylopyranosyl-(l ->4)-a-L-rhamnopyranosyl-( 1 -^2)-[P-D-xylo-pyranosyl(l~>3)-P-D-4-(9-acetylquinovopyranosyl-(1^4)]-P-D-fucopyran-oside.

[M-H]-; 1831.7 [M-H]"-H20: 1813.6 [M-H]"-CH2C0: 1789.6 [M-2Hf': 915.4

'-^6 HOOC Xyl(D) \ H O ^

955.45 1699.6

HOHO HO

.-?£^" .

Xyr(H)^.y HO-

OHkHO HO

'

^

•

-

HO in-^^ OHHHO Xyl"(F)

-

^

^

^

Xyr(E)

Fig. (10). Key HMBC correlations and ESI-MS/MS fragmentation pattern of saponarioside A.

26 Saponarioside C was assigned a molecular formula of C59H94O29 from its MALDI-TOF MS (m/z 1289 [M+Na]^, 1305 [M+K]^) and ^^C, DEPT NMR analysis. The pentasaccharide nature was evident from its ^H [5 4.90 d (J = 7.7 Hz), 4.99 d ( J = 7.3), 5.24 d (7= 7.9), 5.47 d ( J = 3.6), 6.19 d ( J = 8.2)] and ^^C [8 95.0, 100.6, 105.3, 105.7, 106.3] NMR spectra. Among the five monosaccharides, three were identified as glucose, one as xylose, and the remaining one as galactose. The p anomeric configurations for the glucose and xylose units were determined from their VHI,H2 coupling constants (7-8 Hz). Only four protons were traced from the subspectrum corresponding to the remaining anomeric proton at 5 5.47 (d, J = 3.6 Hz) in the HOHAHA spectrum (Fig. 2). The galactosyl unit showed a typical distribution of the scalar coupling around the system that was impeded by the small coupling between H-4 and H-5. The a configuration for the galactosyl unit was assumed from the small VHI,H2 coupling constant (3.6 Hz), which demonstrated the gauche relationship between H-1 and H-2, the two vicinal protons. Further evidence supporting the a configuration was obtained from the long-range HMBC correlation from the anomeric proton to C-3 and C-5 (the dihedral angles being about 180°), and the existence of NOE only between H-1 and H-2 (Fig. 3). Additionally, the one-bond ^^C-^H coupling CJCH) of 167 Hz for the anomeric carbon (163 for the p-anomer), as well as the relative upfield shifts of C-1, C-2, C-3, and C-5 compared to those of the panomer, also confirmed the a-anomer (Table 1). The absolute configurations of these sugars were determined from the HPLC analysis of the l-[(5)-A^acetyl-a-methylbenzyl-amino]-l-deoxyalditol acetate derivatives [45, 46]. Saponarioside C contains a a-D-galactopyranosyl moiety, which is a common component for glycolipids [47, 48] but rare for triterpenoid saponins. To the best of our knowledge, this is the first example of a triterpenoid saponin containing a a-D-galactose unit. Table 1. The NMR Data for the a and p-D-galactose Moieties (500 MHz in pyridine-^5) Carbon

a

a-D-galactose

p-D-galactose

1

100.6 (167 Hz)

5.46 d (3.7 Hz)

104.2 (163 Hz)

5.54 d (7.6 Hz)

2

70.6

4.64 dd (3.6,9.8)

73.7

4.45 dd (7.6, 9.8)

3

71.6

4.55 dd (9.8, 3.2)

75.4

4.15 dd (9.8, 3.0)

4

71.0

4.59 d (3.2)

70.2

4.56 d (3.0)

5

72.5

4.60 m

76.7

4.03 d (6.1)

6

62.6

4.40 (2H) d (6.4)

61.7

4.44,4.50

The data in this column were taken from saponarioside A.

27

The saponin fraction of 5. officinalis has shown antiinflammatory activity in vitro against carrageenan-induced rat-paw oedema and inhibited prostaglandin synthetase [49]. Purified saponins have been recently shown to possess hypocholesterolaemic effects in vivo, which is believed to be due to the ability of the saponin to form an insoluble complex with cholesterol, preventing its absorption from the small intestine [50]. The saponins also demonstrated spermicidal activity, which may come from their hemolytic property [51]. TRITERPENOID SAPONINS FROM VACCARIA SEGETALIS The plant Vaccaria segetalis (Neck.) Garcke (syn. V. pyramidata Medik) is an annual herb widely distributed in Europe, Asia, and other parts of the world. In Japan, it has been cultivated as a garden plant for several centuries. The seeds of the plant, popularly known as Wang-Bu-Liu-Xing, have been prescribed frequently in China to cure the diseases associated with women after childbirth. According to the theory of traditional Chinese medicine, Wang-Bu-Liu-Xing has the capacity to activate the flow of blood and to promote milk secretion. Besides, it is also used in the treatment of amenorrhoea and breast infections [52]. Early chemical investigation of the seeds of this species led to the isolation of several triterpenoid saponins having as many as 10 monosaccharide units (Fig. 11) [53-55]. Reinvestigation of the seeds of the plant in our laboratory led to the isolation of eight new triterpenoid saponins, designated as vaccarosides A-H (Fig. 12) [56, 57]. However, we could not

COOR. RiO

VacsegosideB

Vacsegoside C

>; COOH

D-Gal-(1.2) D-GaKl-3) D-GluA L-Ara-{l-6)

D-Xyl-( 1 -4)-L.RIia-( 1.3>D-Fuc

D-Gal-(l-2) D-Gal-(l-3) D-Xyl-( 1 -4) D-GluA L-Ara-(l-6)

L-Rha-( 1 -3)-D-Xyl-( l-4)-L-Rlia-( 1 -3) D-Fuc D-Glc-(l-4)

Fig. (11). Vacsegosides B and C from Vaccaria segetalis.

28 identify any of the earlier reported structures even by using the most modern chromatographic and spectral techniques. It is doubtful whether the saponins isolated earlier were pure or mixtures of saponins.

Vaccarosides A

R,=H

B

Rl=H

C

Rl=glc

R2 = H

CH3 O

R2 = H

OH

O

D

HO OH

COOH

OH

GlcA HOOC HO' HO HO

^

K2

o.

^

"' /

F"c

"^^-^ HO^^-"'^ Gal

Rha HO

OH

Fig. (12). Vaccarosides A-H from Vaccaria segetalis.

E

CHO

OH

F

OH

OH

G

CHO

H

H

OH

H

29 The eight saponins isolated from V. segetalis can be divided into two groups. Vaccarosides A-D, structurally related to saponariosides isolated from the A?-BUOH soluble fraction, are less polar with either gypsogenic acid or 3,4-secogypsogenic acid as aglycones. All the four saponins possess the same oligosaccharide moiety attached to the carboxyl group at C-28 of theaglycones. Vaccarosides E-H were bisdesmosides possessing the same sugar arrangement at both C-3 and C-28, but differing in the aglycone part. The aglycones of vaccarosides F and H were new nor-triterpenoids, designated as segetalic acid and vaccaric acid respectively. Vaccarosides A, E, and F are the major saponin constituents of the plant. Vaccaroid A, a compound having identical structure with vaccaroside A was isolated from the same source by another group [130]. Vaccaroside D, unlike gypsogenic acid, showed five methyl proton singlets [5 0.90, 0.92, 0.99, 1.19 (x 2)] and two vinyl proton singlets (5 5.60, 6.58). The characteristic t-like signal at 6 5.64, coupled with the two sp^ carbon signals at 6 122.8 (d) and 143.9 (s) indicated that the aglycone possessed an olean-12-ene skeleton. Detailed analysis of COSY, HOHAHA, HMQC, HMBC and NOES Y spectra indicated that structural features in the C, D, and E-rings of the aglycone are similar to those of gypsogenic acid. Further, the presence of a -CH2CH2CH2OH group and a a-substituted-a,p-unsaturated carbonyl group, which originated from the A-ring of gypsogenic acid, could be deduced from the one-bond and long-range couplings (Fig. 13). The above information suggested that the aglycone was a derivative of 3,4secogypsogenic acid. It may be pointed out that peaks for C-4 (8 146.0 ppm), C-23 (5 171.0) and C-24 (8 124.0) could not be observed in the ID ^^C NMR spectrum, but were located from the crosspeaks in HMQC and HMBC experiments. The 3,4-seco structural feature in the A-ring has strong influence on the chemical shifts of the B-ring carbons. The C-5 and C-9 resonances were shifted to higher fields while those of C-6 and C-10 shifted to lower fields.

•Carbons not observed in '•'C-NMR but locatedfromHSQC or HMBC

Fig. (13). Key HMBC correlations for the aglycone part of vaccaroside D.

30

The C-9 signal moved to 5 38.3 ppm, 10 ppm upfield from its counterpart (8 48.3 ppm) in vaccaroside A. 3,4-Secogypsogenic acid was first reported as the sapogenin of the triterpenoid saponin dianoside I, isolated from the whole plant of Dianthus superbus L. var. longicalycinus Williams (Caryophyllaceae) in 1984, and the structure was established by UV, IR, ^H-, ^^C-NMR and chemical studies [58]. However, most of the ^^C assignments of A and Brings needed revision. Another compound, 16a-hydroxy 3,4-secogypsogenic acid encountered in the present investigation, was identified as the aglycone of saponarioside K of Saponaria officinalis [6]. Vaccaroside F was assigned the molecular formula C65H102O33, a carbon less than that of vaccaroside E. The ^H and ^"^C NMR spectra revealed that the sugar moiety was identical to that of vaccaroside E but the two differed in the aglycone part. Of the 65 carbons, 29 were assigned to the aglycone part, 34 to the oligosaccharide moiety and the remaining two to an acetoxy group. The spectrum of vaccaroside F lacked the signal due to the aldehyde group (C-23, 5 209.4) of vaccaroside E and its C-3 signal was observed at 6 92.4, 9 ppm downfield compared to that (6 83.4) in vaccaroside E. After extensive 2D-NMR analyses, the aglycone was characterized as 3p,4a,16a-trihydroxy23-norolean-12-en-28-oic acid, a new genuine nor-triterpenoid sapogenin designated as segetalic acid. Alkaline hydrolysis resulted in a prosapogenin, which on enzymatic hydrolysis with p-glucuronidase [59] afforded the genuine aglycone, segetalic acid (Fig. 14). Acid hydrolysis of vaccaroside F fiarnished a rearranged aglycone, identified as 3-keto-16a-hydroxy-24-noroleanolic acid. Vaccaroside F

COO-sugar chain ''OH

GIcA

Gal An amorphous solid C29H46O5 Mp: 208-210 ° C [a]23^29.0(MeOH,c0.2) IR:v'KBr ^428, 2928, 1693,1359,1200 cm--1 FAB-MS: m/z 497 [M+Naf

P-glucuronidase (0.25 ml, Sigma) Acetate buffer (pH 5.0,0.5 ml) 37''C, 4 h

^ Segetalic acid

Fig. (14). Alkaline and enzymatic hydrolysis o f vaccaroside F.

31 The 4P-methyl group took up to the energetically more favored 4a position in course of the migration (Fig. 15). During acid hydrolysis, the aglycone having a 3p,4a glycol structure apparently underwent protonation of the 4ahydroxyl group and subsequent elimination of a molecule of water forming a cation at C-4; loss of H-3 then led to an enol which immediately tautomerised to the thermodynamically favorable 4a-methyl-3-keto compound. A minor pinacol rearrangement product also resulted from the migration of the alkyl group (C-2) forming a five-membered A-ring and an aldehyde group. A structurally related saponin squarroside A (vide supra) has been isolated from Acanthophyllum squarrosum [60]. This saponin of gypsogenin contains one more xylose linked to C-3 of the glucuronic acid. This molecule showed a concentration dependent immunomodulatory effect in the in vitro lymphocyte transformation test.

'COOR2

^^ HCl/80°

rCOOH TT n

R,0^ OH vaccaroside F

HO'

Fig. (15). Acid induced rearrangement of the sapogenin of vaccaroside F.

COOH

32

TRITERPENOID SAPONINS FROM DIANTHUS GENUS The genus Dianthus is composed of more than 200 species of both annual and perennial herbs, native to Europe and temperate Asia. Many of them are being cultivated as ornamental plants but the most horticulturally valued species are the carnation, D. caryophyllus, and sweet William, D. barbatus, A number of plants of this genus are well documented in Oriental medicine, but only three species are being used in traditional Chinese Medicine as diuretic and anti-inflammatory agents [61]. Chemical investigation in early 1970s led to the isolation and characterization of three triterpenoid saponins, dianthosides A, B, and C, from the whole plants of Z). deltoides. Fig. (16) [62, 63]. Later, nine more triterpenoid saponins, dianosides A-I, were reported from the whole plants of D. super bus var. longicalycinus, a crude drug used in Japan, Fig. (17) [58, 64-65]. Some of these compounds exhibited significant analgesic activity. The structures of dianosides A-I, were established by ^H and ^C NMR as well as some chemical transformations. Three more gypsogenic acid based saponins, viz. the 3-0-glucopyranoside, 3,28-0-digluco- pyranoside (dianoside A), and 3-0-glucopyranosyl 28-0-[glucopyranosyl (l-^6)glucopyranoside] (azukisaponin) were identified from D. caryophyllus var remontent (carnation) [66]. Recently, we reported the isolation and structure elucidation of eight triterpenoid saponins, dianchinenosides A-H, from the aerial parts of D, chinensis, documented in traditional Chinese medicine (Fig. 18) [34, 67, 68]. It is

COOR

RoO'

R,

R2

Dianthoside A

-COOH

D-GIc

DianthosideB

-COOH

D-Glc(l-6)-D-GIc

Dianthoside C

-CHO

D-Gal-( 1 -3)-D-Xy l-( 1 -3)-D-GlcA-

L-Ara-(l-2)-L-Rha-(l-3)

Fig. (16). Triterpenoid saponins from Dianthus deltoides.

D-Gal-(1.3)| D-Xyl-(l-4) L-Rha-(l-2)

D-Fuc-

33

'COOR2

'C00R2

COOH Ri

R2

Dianoside A

D-Glc

D-Glc

Dianoside B

D-GIc

D-Glc-(l-6) 1^^'^

Dianoside G

H

Dianoside H

H

D-Glc

Azukisaponin

D-Gic

D-Glc-(l-6)-D-Glc

D-Glc-(I-6) 1 ^

Ri

R2

Dianoside C

D-GIc

D-Glc

Dianoside D

D-Glc

Dianoside E

D-Glc

D-Glc-(l-6)-D-Glc

Dianoside F

H

D-Glc

D-Glc-(1.3) L . Q ,

D-Glc-(l-6) l^*^'^

^J^^- "•

Dianoside I

0-J/

r^^^

•OH \ ^

Glc"

Glc

Fig. (17). Triterpenoid saponins from Dianthus superbus var. longicalycinus.

of interest to note that dianchinenosides E and F, and also dianchinenosides G and H are diastereomers, differing only in the stereochemistry of the 1,2propanediol residues attached to C-23 carboxylic acid of the aglycones. The chromatographic behaviors of the pairs were more or less the same. Repeated chromatographic purification over silica gel column and medium-pressure chromatography (silica gel and ODS) led to the isolation of two TLC homogeneous compounds which were found to be mixtures when subjected to NMR analysis (Fig. 19). However, pure compounds of both the stereoisomeric pairs could be isolated by repeated HPLC purification over an ODS column with retention time prolonged to as much as 120 min for dianchinenosides E and F and 140 min for G, H (PEGASIL ODS, Senshu Pak, 10 mm i.d.x 250 mm, MeOH-H20 75:25, 0.5 ml/min; detector: UV 210 nm) (see Fig. 20 for the ^H-NMR spectra of the pure dianchinenosides G and H). Dianchinenoside E was assigned a molecular formula of C57H92O26 from its positive ion FABMS ([M+Na]"" at m/z 1215) and ^^C DEPT NMR spectra. Of the 57 carbons, 30 were assigned to the aglycone part, 24 to the oligosaccharide moiety, and 3 to a 1,2-propanediol group. The IR spectrum showed absorptions at 3405 cm"^ (-0H) and 1718 cm'^ (ester). Acid hydrolysis of the

34

saponin afforded an aglycone, characterized as 23-0-1,2-propanediol gypsogenic acid from its MS, ID NMR (Fig. 21, upper) and 2D-NMR analyses. The existence of the 1,2-propanediol fragment was revealed from the COSY and HETCOR spectra. The linkage of this fragment to C-23 of gypsogenic acid was established from HMBC correlations between the signals at 6 4.14, 4.35 (C1-H2 of the 1,2-propanediol) and 8 177.7 (C-23). The sugar units were identified as glucose from GLC analyses. The sequence of the sugar chain was determined by a combination of COSY, HOHAHA, HETCOR, HMBC and phase-sensitive NOESY experiments.

Dianchinenosides

'COOR3 '

R,0'

Ri

R2

R3

A

D-Ara

H

D-Glc

B

D-Xyl

H

D-Glc

C

H

D-Glc

D-Glc

D

H

H

D-Glc-(l-6)-D-Glc

COOR

G'

0

^.^"^

G

QT-OH

^ O H

0~i>^ G"

R)

/-e^cH3

OH

R2 D-Glc (G'")

H OH /^CH3

D<J1C (G'")

Hd^H (^Ux

.-^V^"3

H

H^OH ^ C H 3 H(> H

Fig. (18). Triterpenoid saponins from Dianthus chinensis.

H

35

Fig. (19). The 'H-NMR of the mixture of dianchinenosides G and H.

iDianchincnoside Gl

u I I

X ^ M M A . I)

..

...J\AM

IMiU

pianchinenoside H

u 1 I 6.0

b;3

5.1

A/MkA. ' I''' 4.b

I' 4JJ

1411

II 3;^5

Fig. (20). The ' H - N M R of dianchinenosides G and H.

3^0

2^_5

g^J

i_^

1.0

36 The 'II-NMR spectrum of the AKiyciiiie oT Dianchinenoiiides K und G

The *H-NMR Spectrum of the Aglycone of Dianchincnosides F and H

UJL_

WU

Fig. (21). The 'H-NMR spectra of the aglycones.

The stereochemistry of the propanediol moiety was estabUshed from the following experiments. The ester could not be hydrolyzed with IN HCl or 2N KOH in aqueous MeOH or even with 3% sodium methoxide in MeOH. Finally, on reduction with lithium aluminum hydride, the aglycone yielded propanediol and the corresponding reduced aglycone of gypsogenic acid, A -

37

oleanen-3p,23,28-triol (Fig. 22). The reduced aglycone was identical (coTLC, NMR, MS) with the product derived from the known triterpenoid glycoside 3-0-a-L-arabinopyranosyl hederagenin 28-0-P-D-glucopyranosyl(l-->6)-P-D-gluco-pyranoside (isolated from the same plant) [68] via hydrolysis with IN HCl and then reduction with LiAlH4. The absolute configuration of the derived propanediol was deduced via HPLC analysis of the primary tosylates using a CHIRALCEL OC column and a hexane-2-propanol (95:5, 2.5 ml/min.) developing system. The propanediol fragment in dianchinenoside E was thus shown to possess the R configuration. Dianchinenoside F was found to have the same molecular formula and the same oligosaccharide chain as that of E but differed in the configuration of the 1,2-propanediol part and the stereochemical difference was clearly reflected in

COOH Aglycone of dianchinenoside E

'coo-^^"3 HO^'H

LiAlH4A'HF/reflux^ ;

INHCl/reflux

CH2OH

HO t j

"

^ 2N KOHMeOH(ag.) I 3%NaOMe

R-(-)-1,2-propanediol CH2OH p-toluenesulphonyl chloride/pyridine

"3^ \\

/ / ^ ' ^ ^ ^ CH3

HO /?-(-)-1,2-propanediol 1 -tosylate

A^^-oleanenOP, 23, 28-triol

HPLC analysis: CHIRALCEL OC (Daicel Chemical, 0.46 x 25 cm) Hexane 2-propanol (95/5), 2.5 ml/min Rt: 29.49 min for /?-(-> 1,2-propanediol 33.48 min for5-("'')-l>2-propanediol

Fig. (22). Identification of the propanediol from dianchinenoside E.

38 their ^H NMR spectra (Fig. 21, Table 2) though the ^^C NMR data were virtually the same. The configuration of the 1,2-propanediol moiety was established as S using the same procedure as described for dianchinenoside E. In the in vitro study, dianchinenosides E and F showed weak inhibitory activity towards cAMP phosphodiesterase but no activity against HL-60 cells (unpublished results). Table 2. ^H and ^^C NMR Data for the Propanediols of Dianchinenosides E and F'' Dianchinenoside E 23-0-1,2propanediol

'H

Dianchinenoside F

•^c

•H

'^C

1

4.35 m, 4.14 m

69.6 (t)

4.24 m (2H)

69.6 (t)

2

4.17m

64.9 (d)

4.20 m

64.9 (d)

3

1.31 d(J=6.1 Hz)

20.0 (q)

1.30 d (6.1 Hz)

20.0 (q)

"Assignments are based on COSY, HOHAHA, DEPT, HETCOR, phase-sensitive NOESY, and HMBC experiments.

TRITEPENOID SAPONINS FROM GYPSOPHILA GENUS The genus Gypsophila has more than 100 species and shows an Eurasic distribution with its center of diversity in South Russia and Asia Minor. The roots of several Gypsophila species including G. paniculata, G, arrostii, G. fastigiata, and G. perfoliata known in trade as Saponariae albae radix or White Soapwort Root contain saponins. The commercial Merck saponin, which has been widely utilized as a standard for hemolytic test, was obtained from the roots of G. paniculata (baby's breath), G. arrostii, and several other Gypsophila species. A number of species are used in folklore medicine as remedies for coughs, colds and ailments of the upper respiratory tract, and in industry as a source of a variety of other products, viz cleaning chemicals, film emulsion, and as ingredients for fire extinguishers. Gypsophila species and cultivars are also popular for floral decorations and bouquets, and are commonly grown as ornamentals. Gypsophila species is one of the earliest members of Caryophyllaceae family that has been investigated for saponins. Efforts were made to identify the sapogenin constituent of the saponins in the later part of 1930s, and a sapogenin named gypsogenin was established as the aglycone [69]. The saponin gypsoside was first isolated from G. pacifica in 1963 and later on from G paniculata, and its structure was established by extensive chemical studies. Fig. (23) [70]. Literature survey revealed that gypsoside occurs in

39 most of the Gypsophila species and in some more plants of Caryophyllaceae family, and seems to be a taxonomic marker for the family [71, 72]. However, such generalization is of doubt since the identification was based merely on the comparison of the glycosides with gypsoside via the derived aglycones and the sugar composition, determined by paper chromatography. Several saponins related to gypsoside were also reported from the genus during the period [73-75].

COOR R.O' CHO

D-Gal-(l-4)-D-GIc-(l-4)l ^, ^ L-Ara-( 1 -3) p-OluA

D-Xyl-(l-3)-D-Fuc-(l-4) I D-XyH 1 -3)-D-Xyl-( 1 -2) | ^"^^

Fig. (23). Gypsoside from Gypsophila pacifica.

R|

R2

R3

R4

R5

Gl

D-Xyl

D-Gal

OH

D-GIc

H

G2

D-Xyl

D-Gal

OH

H

D-Ara-(l-4) -D-Ara*

G3

H

D-Glc

H

D-Glc

H

G4

D-Xyl

D-Gal

H

D-Glc

H

GlcA HOOC, /O. HO' > / ^ R R2O

^^^^^7^

5^^^^^

HO R5'

R4O

HO

Rha *Note that the absolute configuration of the arabinoses was given as D by the authors.

Fig. (24). Triterpenoid saponins from Gypsophila paniculata and G. arrostii.

In early 1990s, Frechet et al. reported the isolation and structure elucidation of four new triterpenoid saponins from the roots of G. paniculata and G. arrostii (Fig. 24) [76], the roots being imported from Hebei, P. R. China. The

40

saponins were purified by HPLC on an RP-18 column and the authors used modern spectroscopic techniques (a combination of homo- and heteronuclear 2D NMR plus FAB-MS) to elucidate the structures without going through chemical degradation. This seems to be a pioneering work in the application of modern NMR techniques for establishing the structures of complex saponin molecules. As can be seen from Fig. (24), none of the compounds was identified as gypsoside and it is presumed that gypsoside is absent in the roots of G. paniculata and G. arrostii. In 1992, Kim et al. reported the isolation and structure elucidation of two major saponins, MS-1 and MS-2 from commercial Merck saponin, a crude saponin fraction from G. paniculata. Fig. (25) [77]. The separation of the saponins was achieved by a combination of normal silica gel column, reversed-phase (C-8) column, and Sephadex LH-20 column chromatography. The structures were established by the hydrothermal degradation method (hydrothermolysis) in combination with methylation studies and ID ^H and ^"^C NMR techniques. For hydrothermolysis, the compound was heated at lOO^C or at a higher temperature for 10-30 h, when the complex oligosaccharide chain was dissected into smaller fragments (Fig. 25). The technique is helpful for 3,28-0-bidesmosides as the ester glycosidic bond is selectively cleaved under such condition. The method is somewhat similar to traditional methods described earlier and needs a large amount of sample.

HO o^^^^^^^^ HO HO^-^'^^H

^ O"^

MS-1: R=H MS-2: R=OH

ii-x:^

HO' HO

OH Qui

Xyl

Xyl

O) CHy^^llN^U CH

HjO 100°C, 30h

H3cz^: ^^-ST;;^

HHO ^ O - '^ X ^ / OH

^^

0 CH3/O. \ f < ^ O W HO

" ^ HO' „"^^^^^-^^ OH

HO^ ^ ^ - ^ ur>^ OH HO

HO

HO

0rH3/O

"?.o 24 h 140*^0, H3C HO HO

HO ^ HO

Fig. (25). The hydrothermolysis of MS-1 and 2 from Merck saponin.

^^^ OH

41 In 1995, Liu et al. reported the isolation of three new saponins from the roots of G. oldhamiana [78], which has been used as a substitute for the wellknown traditional Chinese medicinal herb Sterllaria dichotoma var. lanceolata in the treatment of fever, consumptive disease, and infantile malnutrition. The most polar saponin was a bisdesmoside and its structure was established as shown in Fig. (26). The other two saponins were monodesmosides derived from C-28 methyl esters of quillaic acid and gypsogenin and having identical sugar moieties at C-3.

Rha HO'

HOCH3 /0>

HOHO' X > ^

HO

"OH

Glc

Fig. (26). Saponin from Gypsophila oldhamiana.

Very recently. Delay et al. reported the isolation and characterization of two highly hemolytic saponins from the commercial Merck saponin (roots of G. paniculatd) (Fig. 27) [24]. The two compounds (SAPO 50 and SAPO 30), which possessed 25% and 1% of the total hemolytic activity of the crude Merck saponin extract, were purified by dialysis, reverse phase C-18 column chromatography, and HPLC with monitoring of hemolytic activity. The structures were determined by high-field gradient-enhanced NMR methods (600 MHz for proton and 125 MHz for carbon) including DQF-COSY, ID and 2D z-TOCSY, ID-NOE, ge-PEP-HSQC, 2D-HSQC-T0CSY, and geHMBC experiments. The proton and carbon chemical shifts of each sugar residue were assigned by a combination of DQF-COSY, z-TOCSY and 2DHSQC-TOCSY. The authors made extensive use of the HSQC-TOCSY technique, which has the potential to locate all the proton and carbon resonances within each sugar residue. When recorded with a small mixing time (16 ms) the HSQC-TOCSY displayed only direct and vicinal correlations as in the TOCSY/HOHAHA experiments; the connectivities disappeared when the VHH value was small (e.g. between H4 and H5 of galactose). But complete proton

42

and carbon assignments could only be made by using a longer mixing time (80 ms). Both compounds had the same aglycone, gypsogenin, with eight and seven monosaccharide units in SAPO 50 and SAPO 30, respectively. The latter had the same structure as G4 described by Frechet et ai. Fig. (24) [76]. It may be mentioned that SAPO 50 had similar oligosaccharide moiety as found in saponarioside A isolated from the whole plant of Saponaria officinalis [2]. The only difference is that the quinovose unit has one more xylose unit at C-3 and an acetoxy group at C-4 in saponarioside A.

Gal HO

H OQ.:^ -X/^H OH

•X/^

HO HO

HO

Rha

Xyl

Xyl

Fig. (27). Triterpenoid saponin from commercial Merck saponin (roots of Gypsophila paniculata).

Apart from the above-mentioned complex bisdesmosides, saponins with simple sugar structures were also reported from this genus. Elgamal et al reported the isolation of seven new saponins (Fig. 28) with quillaic and gypsogenic acids as aglycones from the whole plant of G. capillaris grown in the eastern part of Egyptian desert [79, 80]. The sugar moieties contained 2 to 5 units of glucoses and/or galactoses. The structures were determined by ID and 2D-NMR study. The ^H and ^"^C data for both the glucoses and galactoses presented in the paper were almost identical. The authors had mentioned that it is usually not possible to distinguish glucose and galactose from the ^H and ^^C chemical shifts. They had identified the monosaccharides from the change in shape and multiplicity of the H-3 signals observed on irradiation of the respective H-1 signals in the NOB difference experiment.

43

COOR,

HO' CHO

COOR2

Aglycone

Glc-(l-6)-Glc

OH

Glc-(l-2) I Gal-(1-6) Glc

OH

Glc-(l-3)-Glc-(l-2)| OH

Gal-(l-6) |<^'c Glc-(l-2)-Gal-(l-3)| Glc-(l-6) Gal Glc.(l-2>Gal-(l-3) Glc-(l-6) Gal

OH

Gic-(l-2)-Gal-(l-3)| Glc

Gk-il-e) Gal

OH

n HO'

HO OH

HO

D-GIc COOH D-Gal

OH

Fig. (28). Triterpenoid saponins from Gypsophila capillaris.

Very recently, the same group of authors also reported a novel triterpenoid saponin from the same source (7, Fig. 28) [81] and identified the galactose unit from the particular shape of the H-3 and H-5 resonances. However, evidence from the 400 MHz TOCSY spectrum (mixing time 180 ms) presented in the paper indicated that the terminal sugar was most probably a glucose since both the sugar units displayed consecutive scalar-couplings from the anomeric protons to the terminals of each systems. Furthermore, the relative positions of identically numbered protons of the sugar units were almost the same, thus revealing the identity of the terminal sugar. It is well known that in

44

a 2D-H0HAHA or TOCSY experiment, the distribution of magnetization between the H-4 and H-5 in a galactosyl type residue can be impeded due to the small coupling between them. It is noteworthy that the C-3 and C-5 of glucose are usually found to resonate at ca 8 78 ppm (recorded in pyridine-tis at room temperature) while the same carbons for galactose resonate at a higher field around 5 75 or 76 ppm. Also, H-4 of the galactose moiety usually resonates in the lower field region compared to the other ring methines of galactose or glucose. Acebes et al. reported the isolation of three new gypsogenin saponins from the roots of G. bermejoi, a Spanish species [82, 83] (Fig. 29). Two of them were sulfated saponins. All the three compounds possessed the same sugar chain, P-D-G1C-(1->2)-[P-D-G1C-(1-^6)]-P-D-G1C at C-28 of the aglycones.

HO;r%H D-Glc

D-Glc Ri

R2

1

H

CHO

2

SOs"

CHO

^

H

CH20S03-

Fig. (29). Triterpenoid saponins from Gypsophila bermejoi.

Gypsophila saponins have been shown to possess hypocholesterolemic effects in rat [84]. The serum cholesterol levels were significantly lower in saponin-fed rats than in control. In another study, Gypsophila saponins were tested in young rats and the results showed that Gypsophila saponins may reduce Fe absorption and have an adverse effect on the Fe status in human and monogastric animals. However, Gypsophila saponins had no effect on Zn absorption [85, 86]. Another report showed that the purified saponins from G. capitata were responsible for the hypocholesterolemic effects in rabbits [87].

45 TRITERPENOID SAPONINS FROM SILENE GENUS The Silene genus consists of about 400 species and shows an Eurasic distribution. Plants of this genus are annual, biennial or perennial herbs or small shrubs of varied habitats. While some plants are omamental, utility of several species in folk medicines have been well documented. For example, the roots of 5. jenisseensis found its name in traditional Chinese medicine as one of the substitutes for the Chinese herbal medicine Yin-Chai-Hu (the roots of Stellaria dichotoma var. lanceolata Bge.). Chemical investigations on this species have been limited. In 1965, Tegisbaev et al reported a saponin named silenoside from the roots of S. latifolia [88], but only the aglycone (gypsogenin) and the sugar components (D-GluA, D-Glc, D-Fuc, L-Rha, D-Xyl, L-Ara) could be identified. Later, Gan et al identified quillaic acid from the acid hydrolyzate of the saponins from S. jenisseensis using NMR and MS techniques [89]. In early 1990s, Karawya et al reported the identification of the sapogenins quillaic acid and oleanolic acid from S. succulenta [90]. In another report from the same group, two new 28-nortriterpenoids, villosagenins I and II, were isolated from S. villosa after acid hydrolysis of the saponin fraction (Fig. 30) [91] and their structures were established by ID and 2D-NMR techniques.

I Villosagenin I J

CHO

CHO

Fig. (30). Triterpenoid sapogenins from the acid hydrolyzate of Silene villosa.

Several reports on full structure determination of saponins from Silene species have appeared since 1990s. Tan et al. reported the isolation and characterization of rubicunosides A-D from S. rubicunda (Fig. 31) [92-94]. Quillaic acid was identified as the sapogenin. The structure elucidations of the saponins were based on partial acidic hydrolysis, enzymatic cleavage (using glycyrrhizinic acid hydrolase), FAB-MS and ID NMR analysis. All the four saponins contained either one or two acetoxy groups in the fucose or in the quinovose moieties of the oligosaccharide chain attached to C-28 of the aglycone. The structures of the oligosaccharide moieties were determined by comparing

46 13/

the C NMR data of the partially hydrolyzed or enzymatically cleaved products with the original saponins. However, chemical and enzymatic degradation did not afford any linear products from cleavage at either C-2 or C-4 of the fucose, i.e. the branching centers. Thus there was lack of direct evidence in supporting the substitution pattern since the C glycosylation shifts can only indicate the presence of substitution but provide no information about the nature of the substituents.

Ara

HOJ ^

1

HO

HO

Rha

"K,o^^ Xyl OH Ri

R2

R3

Rubicunoside A

CH3CO

AcO-2Qui-

Xyl

Rubicunoside B

CH3C0

AcO-2Qui-

Xyl

Rubicunoside C

CH3C0

GlcMQui-

Xyl

Rubicunoside D

H

AcO-2Qui-

H

Fig. (31). Rubicunosides A-D from Silene ruhicunda.

Lacaille-Dubois et al, [95, 96] reported a series of saponins from two Silene species, S. jenisseensis and S. fortunei collected from Hebei province, P. R. China. The roots of both plants find their place in traditional Chinese medicine as a substitute for the well-known Chinese herbal medicine YinChai-Hu (roots oiStellaria dichotoma var. lanceolata), used for the treatment of various types of fevers. From the roots S. jenisseensis, four stereoisomeric acylated triterpenoid saponins, jenisseensosides A-D (Fig. 32), were isolated and their structures were established both by chemical strategies and spectroscopic methods (COSY, HMQC and HMBC). The saponins were obtained as two inseparable pairs with the acyl groups attached to C-4 of the fucose moiety [existing as geometrical isomers {trans- or c/i'-p-methoxycinnamoyl)]. Jenisseensosides A and B showed weak inhibitory effect in in vitro cyclooxygenase inhibition assay and jenisseensosides C and D exhibited a significant enhancement of the granulocyte phagocytosis in vitro.

47

Jenisseensosides

RiI Glc

Glc

H

R-

-C-C=C-\ >-OMe H ^—^ O H H /=^ - C - C = C - ^ j>-OMe OH /=x -C-C=C-\ >-OMe H — O H H

HO

^=\

H

Gal

Fig. (32). Jenisseensosides A-D from Silene jenisseensis.

From the roots of S. fortunei din octasaccharide glycoside of quillaic acid (Fig. 33) has been isolated along with two other reported earlier, jenisseensosides C and D [97]. The structure was established by chemical degradation, FAB-MS, and 2D-NMR analysis including DQF-COSY, HMQC and HMBC experiments. It has an acetoxy group attached to C-4 of the fucose moiety rather than a cinnamoyl group as in case of other jenisseensosides. In in vitro bioassay, this saponin showed a significant enhancement of granulocyte phagocytosis.

> ; ^ : : ^ ^ ! ^ « ^ ^ > ^ r Fuc

Oal

HO

O

HO'X/'WI ^'"

HO Rha

HO Ara HO-

Fig. (33). Triterpenoid saponin from Silene fortunei

Quite recently, Glensk et al. reported the isolation of three new saponins, silenosides A-C (Fig. 34) from the roots of iS. vulgaris (syn. S. inflata) [98], a perennial herb of Europe used in folk medicine for the treatment of anemia. The structures of the silenosides were established both by chemical and spec-

48 troscopic evidences. The molar ratios of the monosaccharides were determined by GC-MS analysis of TMS derivatives of the methanolysis products. The absolute configurations w^ere determined by GC analysis of the respective trimethylsilylated methyl 2-(polyhydroxyalkyl)-thiazolidine-4(i?)-carboxylate derivatives. The glycosidic and inter-glycosidic linkages were established from ESI-MS, 2D NMR (COSY, HMQC and HMBC) as well as GC-MS analysis of the permethylated products.

Silenosides A

R, H

R,

R3

R.

H

H

Xyl

B

OH

H

Glc

H

C

OH

Gal

Glc

H

Fig. (34). Silenosides A-C from the roots of Silene vulgaris.

TRITERPENOID SAPONINS FROM OTHER CARYOPHYLLACEAE PLANTS Besides the above-discussed saponin-containing genera, triterpenoid saponins were also reported from other Caryophyllaceae plants. Agrostemma githago (commonly called com cockle), an annual plant native to Eurasia, is now naturalized throughout the world. The seeds are said to be poisonous for both animals and human beings due to its saponin contents. Siepmann et al. reported the isolation and structure elucidation of two acylated saponins from the seeds of A. githago var. githago (Fig. 35) [99]. The structures were determined by ESI-MS, MS-MS, methylation analysis, and 2D homo- and heteronuclear correlated NMR (COSY, HMQC, HMBC) experiments. The absolute configuration of the sugars were determined by conversion of the enzymatically cleaved sugars into (-)-i?-but-2-yl-glycosides followed by GC-MS analysis. Spergularia ramosa is a herbaceous plant used in Peru as a remedy for respiratory ailments, tuberculosis, and rickets. Quite recently, De Tommasi et al. reported the results of chemical investigation on the aerial parts of this plant [100]. Six new saponins (Fig. 36) with gypsogenin or quillaic acid as agly-

49 cones were isolated, and the structures were established on the basis of extensive NMR experiments including DQF-COSY, ID-TOCSY, 2D HOHAHA, ROESY, HSQC, and HMBC. These compounds occur in pairs, which differ in the aglycone part. Saponins 1-4, like most of the saponins reported so far from Caryophyllaceae family, possess a glucuronic acid linked to C-3 of the aglycones. The other two saponins (5, 6) had a P-D-Gal-(l->3)-P-D-Glc moiety at the C-3 position. The pentasaccharide moiety at C-28 consists of arabinose, fucose, glucose (or galactose), rhamnose and xylose with the arabinose unit linked to the carboxyl group.

Xyl

1

R=H

2

R = Xyl

Fig. (35). Triterpenoid saponins from Agrostemma githago var. githago.

1

H

(}-D-Gal-
P-D-Glc

2

OH

P-D-Gal-(l-3)-P-D-GluA

p-D-Glc

3

H

p-D-Glc-
p-D-Gal

4

OH

P-D-Glc-(l-3)-p-D-GluA

p-D-Gal

5

H

p-D-Gal-(l-3)-p-D-GIc

P-D-Glc

6

OH

P-D-Gal-(l-3)-P-D-Glc

p-D-Glc

Fig. (36). Triterpenoid saponins from Spergularia ramosa.

50

The genus Acanthophyllum consists of about 50 species distributed in Central Orient. There were several reports on the chemical investigation of the saponins of this genus in the 1970s [1] using standard chemical strategies. In 1993, Lacaille-Dubois et al reported a new triterpenoid saponin named squarroside A from the roots of A. squarrosum [syn. A. pungens var. squarrosuni] (Fig. 37) [60]. The structure was established by extensive NMR analysis coupled with some chemical degradation and permethylation studies. Squarroside A showed a concentration dependent immunomodulatory effect in in vitro lymphocyte transformation test. Saponins with structures similar to squarroside A were also reported from the seeds of Vaccaria segetalis [57].

Fig. (37). Squarroside A from Acanthophyllum squarrosum.

Several Herniaria species are used in folk medicines as diuretic agents. H. glabra (smooth rupturewort) and H. hirsuta (hairy rupturewort) are being used as herbal drugs in Germany in the treatment of kidney disorders, in infections of urinary and respiratory tracts, in neuritis and neural catarrh, in arthritis and rheumatism, and for "purifying the blood" [101]. From the aerial parts of//, glabra seven saponins were isolated (Fig. 38) [23, 102-104]. Their structures were established by a combination of 2D-NMR techniques (COSY, TOCSY, ROESY, HMQC and HMBC). The saponins can be divided into two groups according to their respective genins. The first group consisting of herniaria saponins 1, 3, 5, and 7 had 16a-hydroxymedicagenic acid (or zanhic acid) based bisdesmosides with a glucuronic acid linked to C-3 and an acetylated fucose to C-28 of the genin, a feature common to the saponins of the Caryophyllaceae family. Herniaria saponins 5 and 7 contain a p-Dapiofuranose, very rare for the saponins of this family.

51 The other group (hemiaria saponins 2, 4, and 6) were acylosides of medicagenic acid. The sugar components were relatively simple with only glucose and rhamnose as monosaccharides. It is of interest to note that ID-TOCSY was successfully employed in the assignment of the glucuronic acid spin system in herniaria saponin 3, which clearly displayed each and every proton from H-1 to H-5 by selective excitation of the anomeric proton (H-1) of the glucuronic acid moiety.

GluA HOOC HO n-^^^ HO

OH

Hemiaria saponin

R2

Rl

R2

COCH,

H

1

H

COCH,

3

H

p-D-glc

H

H

H

COCH,

5

OH

|i-D-Api(0

7

H

3-D-Api(r)

Fig. (38). Triterpenoid saponins from Herniaria glabra.

K fontanesii is a widely distributed plant in the Mediterranean area and its aerial parts are used in Moroccan folk medicine as a diuretic and in the treatment of lithiasis. In a series of studies, M'bark et al. isolated and characterized four bisdesmosides, hemiaria saponins A-D from the plant (Fig. 39) [105-107] with medicagenic acid or 16a-hydroxymedicagenic acid as aglycones. The structures were determined on the basis of methyiation studies coupled withlD- and 2D-NMR (^H-*H COSY, HOHAHA, ^^C-^H COSY) experiments. Herniaria saponins C and D possessed a glucuronic acid linked to C-3 and a fucose unit to C-28, conforming to the common structural feature of the saponins from the Caryophyllaceae family. However, hemiaria saponins A and B with a GluA-(1^^4)-Rha linked to C-3 of the aglycones are very rare not only for the saponins from this family but also for triterpenoid saponins in general. When tested in dog in which the bile canal was trans-

52 planted with human bile stones, the hemiaria saponins showed an ability to reduce the biliary secretion of cholesterol and to accelerate the dissolution of the bile stones. This experiment confirmed the claimed litholytic properties of H. fontanesii in folk medicine.

S ^ '"'•

HO' HO

P HOOC

^

o

^ /

"'^

HO

HO

H3c7ro^

HiC HO

HO-'^V

1

.1

HO

L

Rha

Rha

Rha

Hemiaria saponin A Hemiaria saponin B

H

Hemiaria saponin C

OH

Hemiaria saponin D

Fig. (39). Triterpenoid saponins from Herniaria fontanesii.

The rhizomes of Arenaria filicaulis Boiss (syn. Gypsophila filicaulis Boiss., Borm.) has been used in Syrian traditional medicine for the treatment of rheumatism, bladder ailments and constipation. Elgamal et al reported the isolation and structure elucidation of five saponins named snatzkeins A-E (Fig. 40) from the plant [3-5]. It is very interesting to note that the aglycones of these saponins are of lupane skeleton, different from the p-amyrin based skeleton mostly encountered within this family. Snatzkein B possessed a hydroxyethyl group at C-2 of glucose, while snatzkein E carries a sulfate group at this position. This is the first report of sulfated saponin from the Caryophyllaceae family. Sulfated saponins were also isolated from the roots of Gypsophila bermejoi [83]. Snatzkein

Ri

R2

R^

R4

A

OH

CH2OH

H

H

B

OH

CH2OH

H

CH2CH2OH

C

H

COO-P-D-Glc

H

p-D-GIc

D

H

COO-p-D-Glc

H

P-D-Gal

E

H

CH2OH

OH

SO2OH

Ri

'*R2

%

^&^^.

Fig. (40). Triterpenoid saponins from Arenaria filicaulis.

53 BIOLOGICAL ACTIVITY Although the biological activities of saponin bearing plants have been known for a long time, earlier studies were limited only to crude extracts, which contained both saponins and other constituents. With the advent of advanced methods of isolation, purification and structure determination, there has been a growing interest to study the bioactivity of homogeneous constituents or of mixtures of structurally related saponins of individual plants. A large number of saponins are toxic and cause hemolysis when administered intravenously, but their toxic effects are much less when given orally [108], as they cannot enter the blood stream crossing the gut and the presence of blood plasma reduces the hemolytic effect. As such, many saponins are being used nowadays as food additives and feedstuffs. A number of reviews have been published [109-113] describing the varied biological activities of saponins, viz. analgesic, antiviral, cardiovascular, antifungal, antibacterial, hypocholesterolemic, antitumor, hypoglycemic, spermicidal etc.; however, the application of these secondary metabolites as successful therapeutic agents is still very much limited. Some interesting results have been reported in the literature on the biological activities of the saponins of Caryophyllaceae family, which are summarized below. Spermicidal activity Elbary et al [114] have correlated the spermicidal efficacy and the hemolytic index of the saponin fraction of Gypsophila paniculata and saponins derived from other plants. Gypsophila saponins were relatively superior when used either as aqueous solution or as formulation in polyethylene glycol ointment base.

Effect on cholesterol level and intestinal transport The effect of Gypsophila saponins on iron and zinc availability and plasma cholesterol concentration has been studied elaborately by different groups [85, 86, 115]. Saponin feeding of rats of low-Fe group resulted in a significantly lower Fe status when compared with controls. The groups fed with low Zn and saponin also showed reduced iron concentration, although the Zn status was not adversely affected. But in all the cases, consumption of saponins resulted in a significant reduction of blood cholesterol concentration, with rats in low-Fe groups having significantly lower concentrations

54 than their basal and low-Zn counterparts. The authors suggested that the consumption of saponins should be encouraged as they have the ability to lower the blood cholesterol level, but the possible effects on Fe metabolism should be investigated further. Gypsophila saponins facilitated transmucosal transfer of p-lactoglobulin in isolated jejunal loop assay, suggesting that saponins may permeabilize the mammalian intestinal barrier [116]. Further, carrier mediated galactose transport was inhibited whereas uptake of passively transported L-glucose increased in the enterocj^es, indicating that saponins may inhibit active transport in the intestine [117]. A saponin mixture from Gypsophila, like potato and tomato glycoalkaloids, induced changes in the membrane integrity as evident from the lactate dehydrogenase leakage in cell culture [118]. The interaction between bile salts and saponins isolated from Gypsophila has been investigated by Gee et al. [84]. When rats were fed with Gypsophila saponins (1.5% by weight) for 7 days, there was no evidence of inflammation or functional damage to the jejunal mucosa of rats but changes in villous morphology were observed. Serum cholesterol level was significantly lowered compared to controls, while the level of cholesterol of the cecal contents was raised. The results suggested that despite the protective effect of bile salts, Gypsophila saponins interact with the mucosa of the proximal small intestine in vivo, but at the dietary level the mucosa was protected by an enhanced rate of cell replacement. The loss of cholesterol via exfoliated mucosal cells might have contributed to the hypocholesterolemic effect of saponins in rats. Sidhu et al. [50] have studied the effect of saponins from different sources with bile acids. Administration to rats of purified saponins from soap wort (Saponaria officinalis), soybean and quillaia {Quillaia saponarid) mixed with diet significantly reduced the rate of absorption of the bile salts; soybean and soapwort saponins did so substantially, but quillaia saponin to a much lower extent. The authors have explained that due to the formation of large mixed micelles by bile acid and saponin molecules in aqueous solutions, bile acids were not available for absorption thus causing a reduction in the plasma cholesterol level. Saponins A and B isolated from Gypsophila capitata showed significant hypocholesterolemic effects [87] in rabbits. Ulloa and Nervi [119] reported that male Wistar rats fed diets containing Gypsophila saponins showed enhanced biliary cholesterol secretion, which is independent of the input of cholesterol from diet. Herniaria saponin B (Fig. 39) isolated from the aerial parts of Herniaria fontanesii showed [106] litholytic properties. Administration of the compound thrice a day for two months reduced the biliary secretion of cholesterol

55

and accelerated the dissolution of bile stones transplanted into the bile canal of dogs, confirming the litholytic properties attributed to the plant. Immunomodulatory Activity Effects of saponins from Gypsophila, guar, alfalfa, Quillaia, clover and liquorice, as also of glycoalkaloids from potato and tomato, were examined [120] on transmural potential difference in mammalian small intestine. Though an immediate reduction in potential difference was noticed in all cases, variation was observed for particular compounds. The factors affecting the nature and magnitude of the depolarizing effect were pH, solubility and chemical nature of the saponin. It was found that glycyrrhizic acid from liquorice root had a protective effect against the activity of a more potent saponin. The effect of soybean and Gypsophila saponins on the morphology of colonic epithelial carcinoma cells (HCT-15) in culture has been studied [121, 122]. Incubation with different concentrations of soybean saponins (SS) or Gypsophila saponins (GS) for 24 hr and scanning under electron microscope revealed a very rough and granular cell surface in cells treated with 40 ppm of GS, whereas minor changes were observed in the cell surface of soybean saponin-treated cells at a concentration of 1200 ppm. Examination by electron microscopy indicated a dose-dependent effect of the saponins on intracellular morphology of colonic epithelial carcinoma cells (HCT-15). Further study revealed that soybean and Gypsophila saponins interact with cell membranes, induce significant dose-dependent cell morphological changes, and appear to act through different mechanisms. Adjuvant activity and immunostimulating complex formation by a number of saponins and glycoalkaloids differing in the structures of their aglycones and oligosaccharide moieties were examined [123]. The only two saponins apart from Quillaia that possess adjuvant activity were Gypsophila and Saponaria, which resemble Quillaia saponins in their chemical structures, i.e. branched sugar chains attached to C-3 and C-28 of the aglycone. Glycoalkaloids, which differ from the triterpenoid saponins in their aglycone part, lacked adjuvant activity even when having branched sugar chains. Saponaria saponins formed irregular immunostimulating complex (ISCOM)-like structures and Gypsophila saponins produced a sheet of joined pore-like structures. Though alfalfa hederagenin saponin and quinoa formed pore-sheets, they lack adjuvant activity. Saponins from the roots of Acanthophyllum squarrosum showed a concentration dependent immunomodulatory effect [60] in the in vitro lymphocj^e transformation test. An acetylated triterpene saponin, isolated together with jenisseensosides C and D from the roots of Silene fortunei, enhances granulocyte phagocytosis in

56 vitro. Both jenisseensosides A and B, isolated from the roots of Silene jenisseensis, exhibited only a weak inhibitory effect in the cycloxygenase inhibition assay for anti-inflammatory agents [95-97]. Cordell et al [124] have reported that barbatosides A and B isolated from the aerial parts of Dianthus barbatos (China Dollu) possess both analgesic and anti-inflammatory activities. Both are saponins of quillaic acid with rhamnose, arabinose, fucose, xylose, galactose, glucose and an unidentified sugar present in the former, whereas arabinose, fucose, xylose, mannose, galactose, glucose and three unidentified sugars were present in the latter. The saponin fraction of the 80% MeOH extract of Dianthus superbus var. longicalycinus has been found to have antibleeding and antitumor activities [125] at a concentration of 50 to 100 mg/ml. Oral administration of 1.5 mg of Gypsophila saponin/ml enhances the allergic response towards ovalbumin in Brown Norway rats [126]. Effects on membrane Saponins of Gypsophila plant facilitate the release of intracellular K"^ and lactate dehydrogenase in rat hepatocytes, suggesting that they destabilize the plasma membrane [127] by removing membrane lipids [128]. Furthermore, Gypsophila plant derived saponins induce differential permeability of the intracellular membrane of rat hepatocytes [129]. Vaccaroid A, a triterpenoid saponin from Vaccaria segetalis [130], induces contractivity of rat uterus. CONCLUSION Triterpenoid saponins from Caryophyllaceae family have drawn intensive attention since 1970s due to important industrial and medicinal applications [1]. Most of the saponins from the family possess very complex and highly branched oligosaccharide moieties, presenting a formidable challenge for their isolation and structure elucidation. Purification remains a timeconsuming but key step in the study of saponins. HPLC has become one of the indispensable and versatile purification methods for saponins. Structurally closely related saponins, even diastereomers, have been successfully separated by modem HPLC technique [34]. Modem NMR techniques are well suited for providing primary and secondary stmctural information for saponins. 2D-C0SY, along with HOHAHA and NOESY techniques, can be employed to provide complete assignment of the sugar protons. A HETCOR or HMQC can thereafter provide full assignment for the corresponding carbons. The identity of the sugars may be determined from the coupling pattems of

57 the sugar protons and confirmed using information from chemical analysis and FAB-MS. An HMBC experiment can then be used to establish the linkages. Most types of appended groups, e.g. O-acyl substituents, can be located from the same experiment. Additionally, NOE techniques can also be utilized to deduce the sugar linkage and sequence; however, these techniques should not be employed as the only evidence if there is any ambiguous correlation involved. Fig. (5) is a summary of the general guideline of the modem NMR techniques employed in the structural study of triterpenoid saponins in our laboratory. As an example, details of the structural study on saponarioside A, a saponin with complex oligosaccharide moiety isolated from Saponaris officinalis, is given. A brief report on the biological activities of the saponins of Caryophyllaceae family is also included.

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22]

Kondratenko, E. S.; Putieva, Z. M.; Abubakirov. N. K. Khim. Prir. Soedin., 1981, 417. Jia, Z.; Koike, K.; Nikaido, T. J. Nat. Prod., 1998, 61, 1368. Elgamal, M. H. A.; Soliman, H. S. M.; Elmunajjed, D. T.; Toth, G.; Simon, A.; Duddeck, H. Magn. Reson. Chem., 1997, 35, 637. Elgamal, M. H. A.; Soliman, H. S. M.; Elmunajjed, D. T.; Toth, G.; Simon, A.; Duddeck, H. Phytochemistry, 1998, 49, 189. Toth, G.; Simon, A.; Elgamal, M. H. A.; Soliman, H. S. M.; Elmunajjed, D. T.; Horvath, G.; Duddeck, H. Magn. Reson. Chem., 1998, 36, 376. Koike, K.; Jia, Z.; Nikaido, T. J. Nat. Prod., 1999, 62, 1655. Smith, F.; Unrau, A. M. Chem. Ind., 1959, 881. Goldstein, I. J.; Hay, G. W.; Lewis, B. A.; Smith, F. Methods Carbohydr. Chem., 1965,5,361. Jia, Z.; Koike, K.; Nikaido, T.; Ohmoto, T. Tetrahedron, 1994, 50, 11853. Klyne, W. Biochem. J., 1950,47, XLI. Doddrell, D. M.; Pegg, D. T.; Bendall, M. R. 1 Magn. Reson., 1982, 48, 323. Braunschweiler, L.; Ernst, R. R. J. Magn. Reson., 1983, 53, 521. Kessler, H.; Gehrke, M.; Griesinger, C. Angew. Chem. Int. Ed, Engl, 1988, 27, 490. Davis, D.; Bax, A. J. Am. Chem. Soc, 1985, 107, 2821. Homans, S. W.; Dwek, R. A.; Fernandes, D. L.; Rademacher, T. W. Proc. Natn. Acad Sci. U. S. A., 1984, 81, 6286. Bax, A.; Morris, G. A. J. Magn. Reson., 1981,42, 501. Summers, M. F.; Marzilli, L. G.; Bax, A. J. Am. Chem. Soc, 1986, 108, 4285. Bax, A.; Summers, M. F. J. Am. Chem. Soc, 1986, 108, 2093. Macura, S.; Ernst, R. R. Mol. Phys., 1980,41, 95. Ranee, M.; Sorensen, O. W.; Bodenhausen, G.; Wagner, G.; Ernst, R. R.; Wuthrich, K. Biochem. Biophys. Res. Commun., 1983, 117, 479. Jia, Z.; Koike, K.; Nikaido, T. J. Nat. Prod., 1999, 62,449. Inagaki, F.; Shimada, I.; Kohada, D.; Suzuki, A.; Bax, A. J. Magn. Reson., 1989, 81,

58

[23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41 ] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56]

186. Schroeder, H.; Schubert-Zsilavecz, M.; Reznicek, G.; Cart, J.; Jurenitsch, J.; Haslinger, E. Phytochemistry, 1993, 34, 1609. Delay, C; Gavin, J. A.; Aumelas, A.; Bonnet, P.-A.; Roumestand, C. Carbohydr. Res., 1997, 302, 67. Bock, K.; Pedersen, C; Pedersen, H. Adv. Carbohydr. Chem. Biochem., 1984, 442, 193 Bothner-By, A. A.; Stephens, R. L.; Lee, J. M.; Warren, C. D.; Jeanloz, R. W. J. Am. Chem. Soc, 19S4, 106,811. Bax, A.; Davis, D. G. J. Magn. Reson., 1985, 63, 207. Sahu, N. P.; Koike, K.; Jia, Z.; Nikaido, T. Tetrahedron, 1995, 51, 13435. Dabrowski, J.; Hauck, M.; Romanowska, E.; Gamian, A. Carbohydr. Res., 1988, 180, 163. Van Halbeek, H.; Vliegenthart, J. F. G.; Winterwerp, H.; Blanken, W. M.; Van den Eijnden, D. H. Biochem. Biophys. Res. Commun., 1983, 110, 124. Hamer, G. K.; Balza, F., Cyr, N.; Perlin, A. S. Can. J. Chem,, 1978, 56, 3109. Mulloy, B.; Frenkiel, T. A.; Davies, D. B. Carbohydr. Res., 1989, 188, 39. Morat, C; Taravel, F. R.; Vignon, M. R. Carbohydr. Res., 1987, 163, 265. Koike, K.; Li, H.; Jia, Z.; Muraoka, H.; Fukui, S.; Inoue, M.; Ohmoto, T. Tetrahedron, 1994, 50, 12811. Bisset, N. G., Herbal Drugs and Phytopharmaceuticais; Medpharm Scientific Publisher; CRC Press: Stuttgart, Germany, 1994; pp. 453. Henry, M. In Biotechnology in Agriculture and Forestry; Bajaj, Y. P. S. Ed.; Springer-Verlag: Berlin Heidelberg, 1989; Vol. 7, pp. 431-442. Chirva, V. Y.; Kintya, P. K. Khim. Prir. Soedin., 1969, 188. Lazur'evskii, G. V.; Kintya, P. K.; Chirva, V. Y. Dokl. Akad Nauk USSR., 1969, 188, 1060. Bukharov, V. G.; Shcherbak, S. P. Khim. Prir. Soedin., 1969, 389. Kon, G.; Scoper, H. R. J. Chem. Soc, 1940, 34, 617. Kubota, T.; Kitatani, H.; Hinoh, H. Tetrahedron Lett., 1969, 771. Henry, M.; Brion, J. D; Guignard, J. L. Plant Med Phytol, 1981, 15, 192. Nagao, T.; Okabe, H.; Mihashi, K.; Yamauchi, T. Chem. Pharm. Bull, 1989, 37, 925. Liu, Z.; Li, D.; Owen, N. L.; Grant, D. M.; Gates, R. G,; Jia, Z. J. Nat. Prod., 1995, 58, 1632. Oshima, R.; Kumanotani, J. Chemistry Lett., 1981, 943. Oshima, R.; Yamauchi, Y.; Kumanotani, J. Carbohydr. Res., 1982, 107, 169. Jung, J. H.; Lee, H.; Kang, S. S. Phytochemistry, 1996, 42, 447. Hohmann, J,; Toth, L.; Mathe, L; Gunther, G. Fitoterapia, 1996, 64, 381. Cebo, B.; Krupinska, J.; Sobanski, H.; Mazur, J.; Czamecki, R. Herba Pol, 1976, 22, 154; C.A. 1978, 88,31943. Sidhu, G. S.; OakenfuU, D. G. Br. J. Nutr., 1986, 55, 643. Abd Elbary, A.; Nour, S. A. Pharmazie, 1979, 34, 560. Jiangsu New Medical College, Dictionary of Traditional Chinese Medicine, Jiangsu, China, 1977, pp. 311. Baeva, R. T.; Karryev, M. O.; Amanmuradov, K.; Abubakirov, N. K. Khim. Prir. Soedin., 1974,258. Baeva, R. T.; Karryev, M. O.; Abubakirov, N. K. Khim. Prir. Soedin., 1974, 804. Baeva, R. T.; Karryev, M. O.; Abubakirov, N. K. Khim. Prir. Soedin., 1975, 658. Koike, K.; Jia, Z.; Nikaido, T. Phytochemistry, 1998, 47, 1343.

59

[57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90]

Jia, Z.; Koike, K.; Nikaido, T. Phytochemistry, 1998, 48, 529. Oshima, Y.; Ohsawa, T.; Hikino, H. Planta Med 1984, 50, 254. Sasaki, Y.; Morita, T.; Kuramoto, T.; Mizutani, K.; Ikeda, R.; Tanaka, O. Agric. Biol Chem., 1988, 52, 207. Lacaille-Dubois, M.-A.; Hanquet, B.; Rustaiyan, A.; Wagner, H. Phytochemistry, 1993, 34, 489. Jiangsu New Medical College, Dictionary of Traditional Chinese Medicine, Jiangsu, China, 1977, pp. 2702. Bukharov, V. G.; Shcherbak, S. P.; Beshchekova, A. P. Khim. Prir Soedin., 1971, 33. Bukharov, V. G.; Shcherbak, S. P. Khim. Prir Soedin., 1971, 420. Oshima, Y.; Ohsawa, T.; Oikawa, K.; Konno, C; Hikino, H. Planta Med., 1984, 40. Oshima, Y.; Ohsawa, T.; Hikino, H. Planta Med., 1984, 43. Gumnicka, O.; Oleszek, W. Acta Soc. Bot. Pol, 1998, 67, 65. Li, H.; Koike, K.; Ohmoto, T.; Ikeda, K. J. Nat. Prod., 1993, 56, 1065. Li, H.; Koike, K.; Ohmoto, T. Phytochemistry, 1993, 35, 751. Ruzicka, L.; Giacomello, G. Helv. Chim. Acta, 1936, 19, 1136. Kochetkov, N. K.; Khorlin, A. J.; Ovodov, Ju. S. Tetrahedron Lett., 1963,477. Amanmuradov, K.; Tanyurcheva, T. N. Khim. Prir. Soedin., 1969, 326. Yukhananov, D. Kh.; Sokol'skii, I. N.; Zinkevich, E. P.; Kubaev, V. B. Rast. Resur., 1971, 3, 3S6. Bukharov, V. G.; Shcherbak, S. P. Khim. Prir Soedin., 1966, 291. Bukharov, V. G.; Karlin, V. V.; Bukharova, I. L.; Surkova, L. N. Khim. Prir Soedin., 1974, 598. Luchanskaya, V. N.; Kondratenko, E. S.; Abubakirov, N. K. Khim. Prir. Soedin., 1970, 434. Frechet, D.; Christ, B.; Monegier du Sorbier, B.; Fischer, H.; Vuilhorgne, M. Phytochemistry, 1991, 30, 927. Kim, Y. C; Higuchi, R.; Komori, T. Liebigs Ann. Chem., 1992, 941. Liu, Z.; Li, D.; Owen, N. L.; Grant, D. M.; Gates, R. G.; Jia, Z. J. Nat. Prod., 1995, 58, 1632. Elgamal, M. H. A.; Soliman, H. S. M.; Karawya, M. S.; Duddeck, H. Nat. Prod Lett., 1994, 4, 211. Elgamal, M. H. A.; Soliman, H. S. M.; Karawya, M. S.; Mikhova, B.; Duddeck, H. Phytochemistry, 1995, 38, 1481. Elgamal, M. H. A.; Soliman, H. S. M.; Toth, G.; Halasz, J.; Duddeck, H. Magn. Reson. Chem., 1996, 34, 697. Acebes, B.; Diaz-Lanza, A. M.; Bemabe, M. Phytochemistry, 1998, 49, 2077. Acebes, B.; Bemabe, M.; Diaz-Lanza, A. M.; Bartolome, C. J. Nat. Prod., 1998, 61, 1557. Gee, J. M.; Johnson, I. T. J. Nutr, 1988, 118, 1391. Southon, S.; Wright, A. J. A.; Price, K. R.; Fairweather-Tait, S. J.; Fenwick, G. R. Br. J. Nutr., 1988, 59, 389. Southon, S.; Johnson, I. T.; Gee, J. M.; Price, K. R. Br. J. Nutr., 1988, 59,49. Velieva, M. N.; Iskenderov, G. B.; Amiralieva, U. A.; Gazhieva, G. Y. Azerb. Med Z/z.,1988,65,40. Tegisbaev, V. G.; Chumbalov, T. K.; Abubakirov, N. K. Rest. Resur., 1965, 1, 102. Gan, Y.; Li, Y. Zhongyao Tongbao, 1988, 13, 35. Karawya, M. S.; Elgamal, M. H. A.; Shalaby, N. M. M.; Soliman, H. S. M.; Duddeck, H. Egypt. J. Pharm. Sci., 1991, 32, 879.

60

[91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121]

Elgamal, M. H. A.; Soliman, H. S. M.; Karawya, M. S.; Duddeck, H. Nat Prod. Lett, 1994, 4, 291. Tan, N.; Zhao, S.; Zhuo, J.; Cheng, C; Wang, D.; He, Y. Chem. J. Chin. Univ., 1994, 15,859. Tan, N.; Zhao, S.; Zhuo, J.; Cheng, C. Acta Chim. Sinica, 1995, 53, 1024. Tan, N.; Zhuo, J.; Zhao, S.; Cheng, C. Acta Chim. Sinica, 1996, 54, 722. Lacaille-Dubois, M.-A.; Hanquet, B.; Cui, Z.; Lou, Z.; Wagner, H. Phytochemistry, 1995,40,509. Lacaille-Dubois, M.-A.; Hanquet, B.; Cui, Z.; Lou, Z.; Wagner, H. Phytochemistry, 1997,45,985. Lacaille-Dubois, M.-A.; Hanquet, B.; Cui, Z.; Lou, Z.; Wagner, H. J. Nat Prod., 1999,62, 133. Glensk, M.; Wray, V.; Nimtz, M.; Schopke, T. J. Nat Prod, 1999, 62, 717. Siepmann, C; Bader, G.; Hiller, K.; Wray, V.; Domke, T.; Nimtz, M. Planta Med., 1998,64,159. De Tommasi, N.; Piacente, S.; Gacs-Baitz, E.; De Simone, F.; Pizza, C; Aquino, R. J. Nat Prod, 1998,61,323. Bisset, N. G. Herbal Drugs and Phytopharmaceuticals, MedPharm Scientific Publisher, CRC Press: Stuttgart, Germany, 1994; pp. 263-265. Reznicek, G.; Cart, J.; Korhammer, S.; Kubelka, W.; Jurenitsch, J.; Haslinger, E. Pharmazie, 1993, 48, 450. Freiler, M.; Reznicek, G.; Jurenitsch, J.; Kubelka, W.; Schmidt, W.; SchubertZsilavecz, M.; Haslinger, E.; Reiner, J. Helv. Chim. Acta, 1996, 79, 385. Freiler, M.; Reznicek, G.; Schubert-Zsilavecz, M.; Reiner, J.; Haslinger, E.; Jurenitsch, J.; Kubelka, W. Sci. Pharm., 1996, 64, 359. Nait-M'bark, A.; Charrouf, Z.; Wieruszeski, J.-M.; Leroy, Y.; Kol, O. Nat Prod Lett., 1995, 6,233. Charrouf, Z.; Nait-M'bark, A.; Guillaume, D.; Leroy, Y.; Kol, O. In Saponins used in Food and Agriculture; Waller, G.R. and Yamasaki, K., Ed.; Plenum Press: New York, 1996; pp. 241-245. Nait-M'bark, A.; Guillaume, D.; Kol, O.; Charrouf, Z. Phytochemistry, 1996, 43, 1075. George, A.J. Food Cosmet Toxicol., 1965, 3, 85. Alder, C; Huller, K. Pharmazie, 1985, 40, 676. Shibata, S. Adv. Med Phytochem., 1985, 159. Sokolov, S.Ya. Adv. Med Phytochem., 1985, 173. Anisimov, M.M.; Chirva, V. Ya. Pharmazie, 1980, 35, 731. Ohura, H. Mod Med Jpn., 1981, 10, 21; CA, 1981, 94, 149797. Elbary, A.A.; Nour, S.A. Pharmazie, 1979, 34, 560. Price, K.R.; Southon, S.; Fenwick, G.R. Chem. Biol Aspects, 1989, 155. Gee, J.M.; Wal, J.M.; Miller, K.; Atkinson, H.; Grigoriadou, F.; Wijnands, M.V.; Penninks, A.H.; Wortley, G.; Johnson, I.T. Toxicology, 1997, 117,219. Johnson, I.T.; Gee, J.M.; Price, K.R.; Curl, C; Fenwick, G.R. J. Nutr., 1986, 116, 2270. Gee, J.M.; Wortley, G.M,; Johnson, I.T.; Price, K.R.; Rutten, A.A.J.J.L.; Houben, G.F.; Penninks, A.H. Toxicol, in Vitro, 1996, 10, 117. Ulloa, N.; Nervi, F. Biochem. Biophys. Acta, 1985, 837, 181. Gee, J.M.; Price, K.R.; Ridout, C.L.; Johnson, I.T.; Fenwick, G.R. Toxicol. In Vitro, 1989,3,85. Rao, A.V.; Sung, M.K. J. Nutr., 1995, 717.

61

[122] [123] [124] [125] [126] [127] [128] [129] [130]

Sung, M.K.; Kendall, C.W.; Rao, A.V. Food Chem. Toxicol, 1995, 33, 357. Bomford, R.; Stapleton, M.; Winsor, S.; Beesley, J.E.; Jessup, E.A.; Price, K.R.; Fenwick, G.R. Vaccine, 1992, 10, 572. Cordell, G.A.; Lyon, R.L.; Fong, H.H.; Benoit, P.S.; Farasworth, N.R. Lloydia, 1977,40,361. Kim, C.J.; Kang, B.H.; Ryoo, I.J.; Park, D.J.; Lee, H.S.; Kim, Y.H.; Yoo, I.D. Han'gukNonghwaHakhoechi, 1996, 39, 409; CA., 1997, 126, 169081. Atkinson, H.A.; Johnson, I.T.; Gee, J.M.; Grigoriadou, F.; Miller, K. Food Chem. Toxicol., 1996,34,21. Wassler, M.; Westman, J.; Fries, E. Eur. J. Cell Biol, 1990, 51, 252. Lee, S.W.; Lee, J.S.; Lee, S.K.; Ok, C.K.; Kim, Y.H. Yakhak Hoechi, 1989, 33, 15; C^., 1989, 111,33664. Wassler, M.; Jonasson, I.; Persson, R.; Fries, E. Biochem. J., 1987, 247, 407. Morita, H.; Yun, Y.S.; Takeya, K.; Itokawa, H.; Yamada, K.; Shirota, O. Bioorg. Med Chem. Lett, 1997, 7, 1095.

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

63

RECENT DEVELOPMENTS IN THE TOTAL SYNTHESIS OF BIOACTIVE MARINE FATTY ACIDS NESTOR M. CARBALLEIRA Department of Chemistry, University of Puerto Rico, P. O. Box 23346, San Juan, Puerto Rico 00931-3346 ABSTRACT: Recent developments in the total synthesis of bioactive marine fatty acids are described. For the purpose of this review marine fatty acids have been divided into three main groups, namely methylated, methoxylated, and diunsaturated fatty acids. Within the methylated fatty acids the Wittig coupling of methylketones has been the principal synthetic strategy used to introduce the methyl ramification, while allylic methylation has been achieved through the a-methylation of carbonyls. Vinylic methylated fatty acids, on the other hand, for the most part have been synthesized stereospecifically, using as the key step either a Claisen orthoester rearrangement or a stabilized phosphorane, both methodologies favoring the E isomer. a-Methoxylated marine fatty acids were synthesized via the methylation of the corresponding ct-hydroxy acids, which were prepared using Mukaiyama's trimethylsilyl cyanide addition to aldehydes. In addition, some optically active methoxylated fatty acids have been made through the regiospecific ring opening of optically active oxiranes with lithioalkynes. Most recent work with diunsaturated marine fatty acids has been concentrated on the synthesis of A5,9 fatty acids. The key strategy has been to use either acetylenic coupling or the selective ring opening of one of the double bonds of (lZ,5Z)-l,5-cyclooctadiene to generate the A9 double bond, followed by Wittig coupling to generate the A5 double bond.

INTRODUCTION Marine fatty acids continue to occupy a central role in modem marine natural products research because of the unusual array of structural motifs constantly being discovered. Most novel fatty acids were identified in marine invertebrates or marine microorganisms. Although excellent reviews of the syntheses of terrestrial fatty acids are published frequently and several outstanding reviews about the chemistry of fatty acids have appeared recently [1], there has been no review of the synthetic methodology used in the preparation of marine fatty acids. Therefore, this review presents a compilation of what has been published recently regarding the synthesis of unusual marine fatty acids, some of

64

which have interesting biological activities. Special emphasis has been placed in the synthetic methodology employed searching for common strategies used in the construction of the unusual features of these interesting marine lipids. Methylated Fatty Acids Saturated Monomethylation Calyx podatypa Van Soest (Phloeodictyidae) is a common Caribbean sponge rich in cyanobacteria [2]. Several unusual metabolites have been isolated from C. podatypa, including proline-derived diketopiperazines [2], 1,3-diphenylbutanoid compounds [3], and antimicrobial A^methylpyridinium salts related to the xestamines [4]. Recently we

HO

H3CO

H3CO

^OH

^OH

H3C0

.P'^PhaBr

H3C0

H3CO

0 CH3OH/HCI (99%); //) PBr3 (96%); ///) PPhs/CgHe (92%); /v) (CH3)3COK /THF, 2-hexanone; v)H2/PtO2(100%). Fig. (1). Total synthesis of methyl 11-methylpentadecanoate (1)

reported several novel methylated fatty acids from C. podatypa, with particular focus on the unusual 11-methylpentadecanoic acid whose structure was confirmed through the total synthesis of its methyl ester [5]. The 11-methylpentadecanoic acid is interesting because similar methylated hexadecanoic acids, in particular the 13-methylpentadecanoic acid, display antimicrobial activity against the cariogenic bacterium Streptococcus mutans [6]. The synthesis of methyl 11-

65

methylpentadecanoate (1), shown in Fig. (1), started with commercially available 10-hydroxydecanoic acid, which was converted to methyl 10hydroxydecanoate through acid-catalyzed methanolysis. Reaction of the methyl ester with phosphorus tribromide resulted in methyl 10bromodecanoate, which subsequently reacted with triphenylphosphine in benzene, affording methyl (9-methoxycarbonylnonyl)triphenyl phosphonium bromide. This phosphonium salt was used in the Wittig coupling with 2-hexanone affording, using potassium ^butoxide as base and THF as solvent, a 2:1 cis-trans mixture of methyl 11-methyl-10pentadecenoate. Final hydrogenation over PtOi resulted in the desired methyl 11-methylpentadecanoate (1). In this case, the internal methyl branching was introduced using the appropriate methylketone in a Wittig reaction. Saturated Dimethylation Not many syntheses of marine dimethylated fatty acids have been reported. However, two rather interesting syntheses have recently been accomplished, namely that of the novel dimethylated fatty acid 9,13dimethyltetradecanoic acid (2), and the synthesis of the rare 10,13dimethyltetradecanoic acid (3), with both acids being identified in the sponge C podatypa [7]. In these two syntheses the desired dimethylation was obtained using /^o-branched methylketones in a Wittig reaction in order to introduce the typical /.so-ramification of bacterial fatty acids. Therefore, the synthesis of 9,13-dimethyltetradecanoic acid (2) was achieved through the Wittig coupling of 6-methyl-2-heptanone and (7methoxycarbonylheptyl) triphenylphosphonium bromide, Fig. (2). The 6methyl-2-heptanone originated from the catalytic hydrogenation of 6methyl-5-hepten-2-one. The salt (7-methoxycarbonylheptyl) triphenylphosphonium bromide was obtained from commercially available 8-bromooctanoic acid, which was methylated with HCl/MeOH and then reacted under standard conditions with triphenylphosphine [7]. Coupling of the methylated heptanone with the Wittig salt resulted in a 73% yield of a 1.4:1 mixture of the ZIE isomers of methyl 9,13-dimethyl8-tetradecenoate, which upon catalytic hydrogenation followed by saponification, resulted in the desired 9,13-dimethyltetradecanoic acid (2).

66 The first synthesis for the 10,13-dimethyltetradecanoic acid (3), a fatty acid first identified in the sponge Ectyoplasia ferox [8], was also recently accomplished [7]. This synthesis began with commercially available 9bromo-1-nonanol, which upon reaction with triphenylphosphine resulted

A/xA H3C0

.PPha^Br'

/) H2/Pt02; //) PPhs/benzene; ///) /-BuOK/THF/a; iv) H2/Pt02; v) KOH/EtOH Fig. (2). Total synthesis of 9,13-dimethyltetraclecanoic acid (2)

in the (9-hydroxynonyl)triphenylphosphonium bromide as shown in Fig. (3). Reaction of the latter Wittig salt with 5-methyl-2-hexanone, using potassium r-butoxide in THF, resulted in a 1:1 mixture of the ZIE isomers of 10,13-dimethyl-9-tetradecen-l-ol. Catalytic hydrogenation of the dimethylated 9-tetradecenol afforded, in a quantitative yield, 10,13dimethyltetradecan-1-ol. Interestingly the related trimethylated alcohol, namely the 6,10,13-trimethyltetradecan-l-ol, is the aggregation pheromone of the predatory stinkbug Stiretrus anchorago [9-10]. The 10,13-dimethyltetradecan-l-ol was fiirther oxidized with pyridinium dichromate in DMF to the desired 10,13-dimethyltetradecanoic acid (3). These syntheses made possible the characterization of these two acids in the sponge. Saturated Trimethylation Two rather interesting syntheses for marine trimethylated fatty acids have recently appeared in the literature [11-14]. In this family of fatty acids the synthetic scenario is different from that of other methylated fatty acids.

67

Partial syntheses have been preferred because the desired trimethylated marine fatty acids can be related to similar isoprenoid natural products that could be used as starting materials for the syntheses. The isoprenoid fatty acid 5,9,13-trimethyltetradecanoic acid (4) is a rather intriguing fatty acid, which has been isolated from several sponges .PPha'^Br-

HO,

0 PPhs/benzene; //) «-BuLi/THF-DMSO and 5-methyl-2-hexanone; Hi) H2/Pt02; iv) PDC/DMF

Fig. (3). Total synthesis of 10,13-dimethyltetradecanoic acid (3)

such as Cinachyrella alloclada Uliczka [11]. Its reported synthesis started with famesol, which upon catalytic hydrogenation over 10% Pd/C

^COjH

.C02H

0 H2/10% Pd-C/ EtOH; ii) PCC/ CH2CI2; ///) malonic acid/pyridine/piperidine; /v)H2/10%Pd-C/MeOH. Fig. (4). Partial synthesis of 5,9,13-trimethyltetradecanoic acid (4)

afforded hexahydrofamesol, Fig. (4). Subsequent oxidation with pyridinium chlorochromate (PCC) afforded 3,7,11-trimethyldodecanal that upon malonic acid condensation in the presence of pyridine and with

68 catalytic amounts of piperidine, afforded (£)-5,9,13-trimethyltetradec-2enoic acid as shown in Fig. (4). In the last step catalytic hydrogenation over 10% Pd-C afforded the desired acid 4 in excellent yields [12]. A more complex example was recently reported in the seven-step synthesis, Fig. (5), of (3i?,7i?5)-3,7,ll-trimethyl-10-oxododecanoic acid (5), a novel sesquiterpenoid fatty acid originally isolated from the green alga Caulerpa racemosa [13-14]. This particular synthesis started with (i?)-citronellyl acetate that was oxidized at the terminal allylic position with selenium dioxide. The allylic alcohol was then transformed into the corresponding allylic iodide with triphenylphosphine and iodine and then alkylated, in what is the key step in this synthesis, with the lithium enolate of isopropyl methyl ketone, affording the acetate of (3i^)-3,7,lltrimethyl-10-oxododec-6-enol, Fig. (5). After basic hydrolysis of the acetate, catalytic hydrogenation, and a couple of oxidation steps, the naturally occurring (3i?,7J?iS)-3,7,ll-trimethyl-10-oxododecanoic acid (5) was obtained. The stereochemistry at C-7 was never elucidated in the natural acid; therefore, the synthesis was also designed to afford a racemic compound at C-7.

^OAc

^

^

->^

-^

-^

Q^j.

OH

OAc

OH

^

^

^

-^

-^

-^

^^

Qjj

0 Se02/ /-BuOOH, CH2CI2, O^'C; //) PhaP, imidazole, I2, CH3CN-Et20 (3:2 v/v) 0°C; /•//) (CH3)2CHCOCH3/LDA, THF, .78°C; /v) K2C03/MeOH, rt; v) H2, 5% Pd/CaC03, EtOH, rt; vi) PCC/CH2CI2, rt; v//) Ag20/NaOH-H20, SO^^C. Fig. (5). Partial synthesis of (3^,7i?5)-3,7,l 1-trimethyl-lO-oxododecanoic acid (5)

69

A Hylic Methylation Two rather interesting racemic syntheses for aliylic methylated fatty acids have been recently accompUshed [15]. One of these corresponds to the intriguing ( 12JE)-11-methyl-12-octadecenoic acid (6), a marine bacterial fatty acid that in most reported identifications from natural sources has been shown to have the E double bond stereochemistry, but both Z,E isomers have been synthesized [15]. This mainly bacterial fatty acid was initially isolated from Byrsocarpus coccineus seed oil [16], but later it was reported in a bacterium associated with cat scratch disease [17], in Mycobacterium fallax [18], and most recently in a Pseudomonas sp. {Alteromonas) associated with both the toxic dinoflagellate Ostreopsis lenticularis and several Caribbean Palythoa species [15]. More recently, acid 6 has been identified as an intermediate in the biosynthesis of the bacterial acid 10,13-epoxy-ll-methyloctadeca-10,12dienoic acid, a furan fatty acid identified in several marine bacteria such as Shewanella putrefaciens [19-20].

o o o ^OH

0 LDA, CH3I, THF, -78°C (79%); //) DIBAL-H, toluene, -TS^C (85%); ///) BrTh3P''(CH2)5CH3, w-BuLi, THF, O^^C (69%); iv) PDC, DMF, rt (62%). Fig. (6). Total synthesis of racemic 11-methyl-12-octadecenoic acid (6)

A four-step synthesis for racemic acid 6 was accomplished starting with the cyclic lactone oxacyclotridecan-2-one, which was a-methylated with lithium diisopropylamide (LDA) and methyl iodide, as shown in Fig. (6). Reduction of 3-methyloxacyclotridecan-2-one with diisobutylaluminum hydride (DIBAL) in toluene at ~78°C afforded

70

initially a lactol, which then opened to 12-hydroxy-2-methyldodecanal. The aldehyde was subsequently coupled, in a Wittig reaction, with 1hexyltriphenylphosphonium bromide and «-butyllithium, affording (Z)and (£)-11-methyl-12-octadecen-l-ol in a 9:1 ratio, respectively. Final oxidation with pyridinium dichromate, in dimethyl formamide, afforded a 9:1 mixture of (Z) and (£)-11-methyl-12-octadecenoic acid (6). We must emphasize that the stereochemistry at C-11 in acid 6 has never been elucidated [15-18]. Key for the introduction of the allylic methyl group in this synthesis was the a-methylation of a lactone, which was then opened and the carbonyl transformed into a double bond through a Wittig reaction. The fatty acid 9-methyl-lO-hexadecenoic acid (7) is the short-chain analog of 6 which was first identified, in trace amounts, in the marine bacterium Vibrio alginolyticus [21]. Both acids might share a similar biogenesis inasmuch as 7 was postulated to arise from the S-adenosyl methionine (SAM) methylation of (Z)-9-hexadecenoic acid, another acid

cis.-ci;'...^.-..„^.. IV

OTBS

^-

OTBS

OH 4 7

/•) TBSCl, imidazole, DMF; //) CH3I, LDA, THF, 0°C; ///) DIBAL-H, ether, -78°C; iv) BrTh3P''(CH2)5CH3, w-BuLi, THF, 0°C; v) TBAF, THF, rt; v/) PDC, DMF, rt. Fig. (7). Total synthesis of racemic 9-methyl-lO-hexadecenoic acid (7)

that was particularly abundant in this V, alginolyticus, followed by hydrogen elimination [18]. The six-step synthesis of 7 started with commercially available methyl 10-hydroxydecanoate, which was initially protected with r^rZ-butyldimethylsilyl chloride (TBSCl) as shown in Fig.

71

(7) [22]. The silylated methyl ester was then a-methylated with Hthium diisopropylamide and methyl iodide in tetrahydrofuran. Reduction of methyl 10-(r^rr-butyldimethylsilyloxy)-2-methyldecanoate with DIBAL in ether at -78°C afforded the corresponding aldehyde. The lO-itertbutyldimethylsilyloxy)-2-methyldecanal was subsequently coupled in a Wittig reaction with 1-hexyltriphenylphosphonium bromide and «butyllithium affording (Z)- and (£)-l-(/er^butyldimethylsilyloxy)-9methyl-10-hexadecene in a 9:1 ratio, respectively. Deprotection with tetrabutylammonium fluroride (TBAF) in tetrahydrofuran and final oxidation with pyridinium dichromate (PDC) in dimethylformamide resulted in a 9:1 mixture of (Z)- and (£)-9-methyl-10-hexadecenoic acid as shown in Fig. (7). As was also the case with acid 6, the stereochemistry at C-9 in 7 is not known. The key step in the synthesis of the allylic methyl group was a-methylation of a methyl ester, followed by reduction to the corresponding aldehyde, which was used in the subsequent Wittig reaction. Vinylic Methylation The 7-methyl-6-hexadecenoic acid is a ubiquitous marine fatty acid first reported from whale oils by both Pascal and Ackman, and Sano [23-24]. However, most recently both Z and E stereoisomers have been identified

i-9til9

^-

8

0 C9Hi9Br/Mg/THF; ii) CH3C(OCH3)3/EtC02H; ///) LAH; iv) PCC; v) (EtO)2P(0)CH2C02Et/ NaH/THF; vi) Mg/MeOH; vii) KOH/EtOH, Fig. (8). Synthesis of 7-inethyl-6(E)-hexadecenoic acid (8)

72

in sponges such as Desmapsamma anchorata, anemone such as Stoichactis helianthus, and gorgonians. This myriad of possible marine sources for this fatty acid is a clear indication that its real source is bacterial [25]. A stereoselective synthesis for the 7-methyl-6(F)hexadecenoic acid (8) was accomplished by Kulkami et al. [26] who started with commercially available methyl vinyl ketone as shown in Fig. (8). Methyl vinyl ketone was reacted with w-nonylmagnesium bromide, resulting in 3-methyldodec-l-en-3-ol. This allylic alcohol was then submitted to a Claisen orthoester rearrangement, the key step in this synthesis that furnished the E double bond stereochemistry, affording ethyl (£)-5-methyltetradec-4-enoate. The ethyl ester. Fig. (8), was then reduced to the alcohol and re-oxidized to (£r)-5-methyltetradec-4-enal, which was then elongated two carbons to the a,p-unsaturated ethyl ester using Wittig-Homer olefination conditions. Regioselective reduction of the conjugated ester with Mg/MeOH, followed by basic hydrolysis, afforded the expected 7-methyl-6(£)-hexadecenoic acid (8). Therefore, in this synthesis the Claisen orthoester rearrangement was key to obtaining the E methylated vinyl group. A novel two carbon analog of 8, namely the 7-methyl-6(Z)octadecenoic acid was recently identified in the holothurian Holothuria mexicana [27]. However, both Z and E stereoisomers were later shown to originate from the bacterium Vibrio alginolyticus [27]. Therefore, the real source of 7-methyl-6-octadecenoic acid is also bacterial. Both Z and E isomers of 7-methyl-6-octadecenoic acid were synthesized as shown in Fig. (9). In this short synthesis (little more than one step) a Wittig coupling of (6-carboxyhexyl)triphenylphosphonium bromide with 2tridecanone readily afforded a 1:1 mixture of 7-methyl-6(Z)-octadecenoic ^^^ ^A.x'-v.x'-v^iBr

' ^

^^k^x-X^x-^PPh/Br

0 PPha/benzene (85%); //) 1.2 M t-BuOK/ THF, a, 0°C (53%)

Fig. (9). Synthesis of 7-methyl-6-octadecenoic acid

73

acid and 7-methyl- 6(£)-octadecenoic acid [27]. A Wittig reaction of a methylketone was used to assemble the methylated vinylic functionality but with no stereoselectivity. A more complex synthesis was recently reported for elenic acid, an inhibitor of topoisomerase II, which was initially isolated from the Indonesian sponge Plakinastrella sp. by Scheuer et al. [28]. The elegant and convergent synthesis by Mori et al. [29] is shown in Fig. (10). The synthesis started with hexadecane-l,16-diol which was converted to 16bromohexadecan-1-ol with hydrobromic acid, followed by dilithium tetrachlorocuprate (Li2CuCl4) catalyzed coupling with pmethoxybenzylmagnesium chloride, which, after reaction with aqueous hydrobromic acid and dihydropyranyl (DHP) protection, afforded the tetrahydropyranyl protected p-hydroxyphenyl-1-bromoheptadecane, one of the two coupling units needed to assemble the molecule. Fig. (10). The other coupling unit was obtained from methyl (iS)-3-hydroxy-2-methyl propanoate, which was transformed into the ^er/-butyldimethylsilyl (TBDMS) protected (5)-3-hydroxy-2-methylpropanol [29]. The latter was oxidized, under Swem conditions, followed by reaction with a H0(CH2),60H

^^-sj^(CH2),70H ii'HV

Br(CH2),60H

.JU

MeO

JUL

HO^^xk/ OTBDMS

Et02C

OTBDMS

OTBDMS THPO

SOzPh

,ja

THPO

(CH2)nBr a

PhOiSv.x^^::^^'^^^ OTBDMS

OH THPO

OH THPO

/•) aq. HBr, Q^\{^ (57%); //)p-MeOC6H4CH2MgCl, LijCuC^, THF (77%); ///) aq. HBr, AcOH (76%); iv) DHP, TsOH (97%); v) Swem; vi) Ph3P=C(CH3)-C02Et; vii) DffiAL, hexane, CHjCIj (quant.); viii) «-BuLi, TsCI, LiBr, EtjO, HMPA (quant.); ix) PhS02Na.2H20, DMF; x) w-BuLi, THF, HMPA, a (78%); xi) PdCl2 (dppp), LiEtjBH, THF (93%); xii) TBAF, THF (94%); xiii) Dess-Martin periodinane, C5H5N, CH2CI2 (quant.); xiv) NaC102, NaH2P04, DMSO, MeCN, H2O; JCV) aq. HCl, THF (74%). Fig. (10). Mori's synthesis of elenic acid

74

methylated ethoxycarbonylphosphorane, which afforded the desired 2,4dimethylated ethyl ester, with the predominantly E double bond stereochemistry in a 19:1 £/Z ratio [29]. Reduction of the dimethylated ester with DIBAL resulted in the allylic alcohol, which was then converted into the corresponding bromide under Stork's conditions [30]. Further reaction with sodium benzenesulfmate afforded the phenyl sulfone, the other coupling unit needed. Coupling of this sulfone with the phenylated bromoheptadecane proceeded smoothly with «-BuLi. Reductive cleavage of the phenylsulfonyl group was then achieved with LiEtsBH and PdCl2(dppp)2 in THF. Final deprotection of the TBDMS group with TBAF and a series of oxidation steps afforded the desired elenic acid. The key synthetic methodology used to stereoselectively introduce the vinylic methylation was a stabilized phosphorane, which favors the E stereoisomer. Methoxylated Fatty Acids a-Methoxylated Fatty Acids a-Methoxylated fatty acids are natural acids most commonly found in the phospholipids of marine sponges with the R configuration at the chiral center [31]. Recent examples include the (Z)-2-methoxy-6-hexadecenoic acid (9) and (Z)-2-methoxy-5-hexadecenoic acid (10), identified in several Caribbean sponges and shown to possess antimicrobial activity against Gram-positive bacteria [32-33]. Recently, the more classical amethoxyhexadecanoic acid was identified for the first time in nature, being found in an Amphimedon sponge [34]. The first synthesis for methyl (Z)-2-methoxy-6-hexadecenoate was reported by Soderquist et al. [35]. The idea behind the construction of the a-methoxy functionality was the addition of a carboxy synthon to the corresponding aldehyde, in this case 5- pentadecenal, followed by methylation of the hydroxy group. This synthesis started with the SuzukiMiyaura cross coupling of 4-bromo-l-butyl-9-borabicyclononane with (Z)-l-bromo-l-undecene, prepared as shown in Fig. (11).

75 n-C9H]9 n-CpHip

——— H

H

n-C9Hi9>. «__»^

n-C9H j 9 v,^^^.

Br

n-C9Hi9—=—^Br

n-C>9H}9>^

^

H

CHO

n-C9H}9>^

OMe OMe

0 NBS/Me2C0, AgNOs; //) 9-BBN-H, HOAc, HO(CH2)2NH2; ill) 4-bromo-l-butyl-9-borabicyclononane/ NaOH, THF/ Pd(PPh3)4; /v) PCC/CH2CI2; v) LiC(SMe)3/H20/NaH/ Mel; vz) HgCl2/HgO/MeOH/H20 Fig. (11). Soderquist synthesis of methyl (Z)-2-methoxy-6-hexadecenoate

The resulting 5-pentadecen-l-ol was oxidized with PCC to 5pentadecenal. This aldehyde was then reacted with tris(methylthio)methyllithium, affording the corresponding a-hydroxy ortho(trithio)ether, which in turn was methylated in situ using NaH/DMF and Mel to obtain the methoxylated adduct, Fig. (11). Final HgCVHgO catalyzed hydrolysis afforded the desired methyl (Z)-2-methoxy-6hexadecenoate. In this synthesis, the tris(methylthio)methyl group was used as the carboxy synthon. A more recent synthesis by our group utilized Mukaiyama's trimethylsilyl cyanide addition to aldehydes as the key step to introduce the carboxy functionality [33,36]. This approach also required the preparation of (Z)-5-pentadecenal as the key intermediate, Fig. (12). In this case, commercially available decyl aldehyde was coupled with 4carboxybutyltriphenylphosphonium bromide under Wittig conditions, resulting in a 10:1 mixture of the known (Z)- and (F)-5-pentadecenoic acids. The acids were then reduced to the desired (Z)-5-pentadecenal via (Z)-5-pentadecen-l-ol, a known pheromone. Addition of trimethylsilyl cyanide to (Z)-5-pentadecenal, under triethylamine catalysis, yielded

76

OSiMcs

HO

H3CO' OCH3

0 H02C(CH2)3CH2PPh3^Br', n-BuLi, THF/DMSO (1:1), -lO^C; //) IN HCl-MeOH, 3h; //•/) LiAlH4-THF, -78°C; iv) PCC (1.5 eq.) CH2CI2, rt; v) TMS-CN, EtgN, CH2CI2, -lO'^C, 2h; vO HCl (cone.) rt; vii) 50% NaOH, heat; viii) NaH/DMSO, CH3I; ix) KOH-EtOH. Fig. (12). Carballeira's synthesis of (Z)-2-methoxy-6-hexadecenoic acid (9)

2-trimethylsilyloxy-6-hexadecenonitrile. The trimethylsilyl cyanide was then transformed into the intermediate a-hydroxy amide with concentrated HCl, and then hydrolyzed to (Z)-2-hydroxy-6-hexadecenoic acid with 50% NaOH. Double methylation was successfully accomplished with NaH and methyl iodide in DMSO, resulting in the ester that was saponified with KOH in ethanol, affording the desired acid 9. The synthesis of (Z)-2-methoxy-5-hexadecenoic acid (10) was done in a similar fashion, but it first required the preparation of (Z)-4pentadecenal. In this case, the aldehyde was made starting with commercially available 1-dodecyne that was coupled with 2-(2bromoethyl)-l,3-dioxolane and A7-BuLi in tetrahydrofuranhexamethylphosphoramide, resulting in the expected 2-(3-tetradecyne)1,3-dioxolane, Fig. (13). Subsequent catalytic hydrogenation using Lindlar's catalyst afforded the expected 2-(3-tetradecenyl)-l,3-dioxolane. The dioxolane was removed with 5% HCl in acetone-water (1:1), and the equilibriimi favored (Z)-4-pentadecenal. Addition of trimethylsilyl cyanide to (Z)-4-pentadecenal under triethylamine catalysis as described by Mukaiyama ifor other shorter-chain analogues [36] resulted in 2trimethylsilyloxy-5-hexadecenonitrile. Under basic conditions the

77

trimethylsilyloxynitrile easily reverts to the original aldehyde. Therefore, the trimethylsilyloxynitrile had to be first transformed into the corresponding a-hydroxy amide under concentrated acid conditions (HCl) and then hydrolyzed to the a-hydroxy acid with 50% NaOH. Under these conditions, the intermediate (Z)-2-hydroxy-5-hexadecenoic acid was obtained. Double methylation was then successfully accomplished with NaH and methyl iodide in DMSO, resulting in methyl (Z)-2-methoxy-5-hexadecenoate. Final saponification with KOH in ethanol afforded the desired (Z)-2-methoxy-5-hexadecenoic acid (10). We can see that in these two syntheses the nitrile group, delivered using trimethylsilyl cyanide, was used as the carboxy synthon. These two amethoxy fatty acids displayed similar antimicrobial activity against the Gram-positive bacteria Staphylococcus aureus (MIC = 0.35 |Limol/mL) and Streptococcus faecalis (MIC = 0.35 |amol/mL) [33].

/) w-BuLi, THF-HMPA, -70°C; //) Hj, Lindlar; in) 5% HCl, MczCO-HzO, 60''C; /v) TMS-CN, EtjN, CH2CI2, -lO'^C; v) HCl cone, rt; vz) 50% NaOH, heat; vii) NaH/DMSO, CH3I; viii) KOH-EtOH. Fig. (13). Synthesis of(Z)-2-methoxy-5-hexadeeenoicaeid (10)

78

All of the above syntheses resulted in racemic 2-methoxy acids, but the stereospecific synthesis of the naturally occurring (i?)-2methoxyhexadecanoic acid was also recently reported [35]. This synthesis started with commercially available (±)-2-hydroxyhexadecanoic acid for which the S enantiomer was selectively acetylated (47 % yield out of a maximum 50%, > 95% ee) with vinyl acetate in THF using the lipase Pseudomonas fluorescens from Aldrich, Fig. (14). In order to facilitate the separation of the acetylated acid from the non-acetylated acid, both compounds were methylated with diazomethane, thus avoiding cleavage of the acetate functionality. The methyl (i?)-2hydroxyhexadecanoate and the methyl (S)-2-acetoxyhexadecanoate were then separated by silica gel column chromatography. The purified methyl (i?)-2-hydroxyhexadecanoate was further methylated with NaH/Mel in DMSO and finally saponified with KOH in ethanol, which afforded the (i?)-2-methoxyhexadecanoic acid with good enantiomeric purity (>95 % ee). This represented the first synthesis for the (i?)-2methoxyhexadecanoic acid [35].

OH

10

OH

10

OAc

10

O

OCH3 10

/) Pseudomonasfluorescens,CH2=CH0Ac, BHT, THF; //) CH2N2, ether; ///') silica gel column chromatography; /v) NaH, Mel, DMSO; v) KOH/ethanol. Fig. (14). Synthesis of (^)-2-methoxyhexadecanoic acid

Other Methoxylated Fatty Acids Another methoxylated fatty acid of marine origin that has attracted the attention of synthetic chemists has been the (45,7S)-(-)-7-methoxy-4tetradecenoic acid (11), a metaboUte of the marine cyanophyte Lyngbya majuscula [38-39]. This methoxylated compound also displays

79 antimicrobial activity towards the Gram-positive bacteria Staphylococcus aureus and Bacillus sub tills [39]. In a convergent synthetic approach, the synthesis started with (i?)-(2,2-dimethyl-l,3-dioxolan-4-ylmethyl) ptoluenesulfonate that was coupled with 1-hexylmagnesium bromide under LiiCuCU catalysis, Fig. (15). The resulting dioxolane was converted into the desired optically active epoxide by reaction with HBr/ V^o

p-Ts-0^3^0

-h

H3C{H2C)6'i ^ _ J -

of-

° - r ^ -^

H3C(H2C)6^

B.V^^ ^

H3C(H2C)6 OH

H3C(H2C)6

O ' ^

OHH

H

u

H3C(H2C)6^^

C02H

H3C' 11

/) Li2CuCl4, THF, hexylmagnesium bromide; //) HBr/HOAc; ///) KOH/MeOH; /v) CBr4/CH2Cl2/PPh3; v) w-BuLi, 2 eq.; vi) «-BuLi, HMPA, a; vii) YiJ Pd/C; viii) hu, PhSSPh, cyclohexane; ix) A?-BuLi, Mel; jc) MeOH/HCl; xi) KOH/ethanol; xii) H2SO4. Fig. (15). Synthesis of (4£, 75)-(-)-7-methoxy-4-tetradecenoic acid (11)

AcOH followed by ring closure with potassium hydroxide in methanol, which afforded the epoxide. The other coupling unit was made from the reaction of 3-(2,4,10-trioxatricyclo[3.3.1.1"^' ]dec-3-yl)propanal with tetrabromomethane and triphenylphosphine, affording the terminal dibrominated alkene which was subsequently transformed into the terminal alkyne with «-BuLi [40]. Coupling of the lithiated alkyne with the epoxide proceeded stereoselectively to the alcohol with no racemization. The resulting alkyne was then hydrogenated under

80 Lindlar's catalyst into the cis double bond that was subsequently photochemically isomerized to the trans double bond with phenyldisulfide [40]. Final methylation with w-BuLi/Mel and hydrolysis of the orthoester group resulted in the desired (4£,75)-(-)-7-methoxy-4tetradecenoic acid (11) as shown in Fig. (15). Therefore, the key synthetic step for the introduction of the optically active methoxy functionality was the ring opening of an optically active oxirane with a lithioalkyne [40]. Diunsaturated Fatty Acids Straight-Chain A5,9'Fatty Acids The A5,9-diunsaturation is probably the most characteristic diunsaturation in fatty acids from sponge phospholipids [41]. One of the most characteristic sponge fatty acids is the (5Z,9Z)-5,9-hexacosadienoic acid since it is found in the phospholipids of most sponges [41]. More interesting is the published observation that model membranes with symmetrical phosphatidylcholines and phosphatidylethanolamines containing the (5Z,9Z)-5,9-hexacosadienoic acid tend to exclude cholesterol [42]. Other interesting examples of naturally occurring A5,9 fatty acids include the shorter-chain analog (5Z,9Z)-5,9-hexadecadienoic acid (12) that was originally reported from the cellular slime mold Dictyostelium discoideum [43] but later identified in several marine sponges [44]. Another naturally occurring A5,9 fatty acid is the (5Z,9Z)5,9-eicosadienoic acid (13), a major constituent of the phospholipids of the sponge Erylus formosus [45]. Several reports focus on the synthesis of straight-chain A5,9 fatty acids. A recent four-step synthesis was based on acetylide coupling to generate the A9 double bond and the Wittig coupling to generate the A5 double bond [46]. In the synthesis of (5Z,9Z)-5,9-hexadecadienoic acid (12) the 2-(2-bromoethyl)-l,3-dioxolane was coupled with 1-octyne using «-BuLi in tetrahydrofiiran/hexamethyl phosphoric acid triamide, affording the desired 2-(3-decynyl)-l,3-dioxolane, Fig. (16). For the synthesis of (5Z,9Z)-5,9-eicosadienoic acid (13), 1-dodecyne was employed. The advantage of this synthetic scheme is that it is possible to control the chain length of the target fatty acid by simply changing the length of the initial alkyne. Hydrogenation of the alkynyldioxolane in dry

81 hexane under Lindlar's catalysis afforded the desired alkenyldioxolanes, as exemplified by 2-(3-decenyl)-l,3-dioxolane in the synthesis of 12. Removal of the dioxolane was readily accomplished with 5% HCl in acetone/water (1:1) affording the desired aldehydes, namely (Z)-4undecenal as the last synthetic intermediate for 12, or (Z)-4-pentadecenal in the case of 13. Final Wittig olefmation of the aldehydes with 4carboxybutyltriphenylphosphonium bromide and ^7-butyllithium in tetrahydrofuran/dimethylsulfoxide (1:1) afforded the desired (5Z,9Z) diunsaturated acids 12-13, generally in 10:1 ratios favoring the 5Z stereochemistry over the 5E.

C ^ B r * „^

12, n=1 16:2 13, n=5 20:2

0 A?-BuLi, THF-HMPA, -TO'^C; //) H2, Lindlar; ///) 5% HCl, acetone-H20, 60°C; iv) Ph3P^(CH2)4C02H, Br", 2.5 M, w-BuLi, THF/DMSO. Fig. (16). Synthetic sequence for normal-chain A5,9 fatty acids

The first straight-chain A5,9 fatty acid to be synthesized was actually the (5Z,9Z)-5,9-hexacosadienoic acid [47]. The methodology developed by Djerassi et al. is shown in Fig. (17). This six-step synthesis started with the unusual controlled ozonolysis of one of the double bonds of (lZ,5Z)-l,5-cyclooctadiene to secure a 100% Z stereochemistry for what eventually was going to be the A9 double bond of the target molecule. Transformation of the product to the tosylated acetal, followed by Grignard coupling under dilithium tetrachlorocuprate catalysis afforded (Z)-4-heneicosenal, which after Wittig olefmation with 4carboxybutyltriphenylphosphonium bromide resulted in the desired A5,9

82 acid, Fig. (17). This same methodology was used in the synthesis of the corresponding l-^'^C-labeled acid that was used in feeding experiments to elucidate the unusual biosynthesis of A5,9 fatty acids. The C-label was introduced by reaction of K^\4rCN with the mesylate of 4,8-pentacosadienl-ol, followed by basic hydrolysis of the nitrile [48]. Interestingly, these radiolabeling experiments targeting the (5Z,9Z)-5,9-hexacosadienoic acid 0CH3

0CH3

0CH3

H

C13H27

C13H27

^COjH

/) leq. O3, MeOH, -78°C; //) TsOH, Ih; Hi) 1.2 eq. NaBH4, -10°C; iv) w-CnHsTMgBr, THF, Li2CuCl4, 0°C, 2h; v) acetone, cone. HCl (20:1), rt, Ih; v/) Ph3P^CH2CH2CH2CH2C02H Br', KH, Me2S0, rt, 2h.

d;

C13H27

CO2H

^OMs

yf,7/.yx ^Ci3H27

vii) BrTh3P"'CH2CH2CH2C02H; viii) LiAlH4; ix) MsCl, NEt3; ;c) K ' ^ C N , D M F ; xi) KOH, EtOH.

Fig. (17). Synthesis of (5Z,9Z)-5,9-hexacosadienoic acid and its 1- C-labeled analog

in Microciona prolifera suggested that the double-bond introduction follows a random order, i.e., either the A5 double bond or the A9 double bond can be introduced first in the acyl chain, and then either the associated A9 or A5 double bonds are introduced [48]. This contrasts with the order of double bond introduction in fatty acids from animals in which the first double bond tends to be introduced at the A9 position, and the second is inserted between the first bond and the carboxyl group.

83 Monomethylated A5,9'Fatty Acids /yo-branched 5,9-dienoic acids with chain-lengths between 16 and 29 carbons have also been identified in the phospholipids of several marine organisms [49-50]. Recently, the shortest member of the series, namely the (5Z,9Z)-14-methylpentadeca-5,9-dienoic acid (14), was isolated from the Caribbean gorgonian Eunicea succinea and shown to be specifically inhibitory against pathogenic Gram-positive bacteria, such as Staphylococcus aureus and Streptococcus faecalis, but inactive against Gram-negative bacteria [49]. On the other hand, longer-chain isO'/S5,9 analogs, such as the (5Z,9Z)-24-methylhexacosa-5,9-dienoic acid, inhibit DNA topoisomerase I (topo I) with ICso's of 1.1 |aM [51]. The reported synthesis for (5Z,9Z)-14-methylpentadeca-5,9-dienoic acid (14) started with commercially available 4-methylpentan-l-ol, which upon reaction with phosphorous tribromide afforded l-bromo-4methylpentane [52]. Commercially available pent-4-yn-l-ol was also protected as the tetrahydropyranyl ether as shown in Fig. (18). Formation of the lithium acetylide with w-BuLi in THF and subsequent addition of l-bromo-4-methylpentane in hexamethylphosphoric acid triamide resulted in the isolation of the tetrahydropyranyl protected 9-methyldec4-yn-l-ol. Hydrogenation of the alkyne with Lindlar's catalyst and quinoline in dry hexane afforded the cis hydropyranyl-protected 9methyldec-4-en-l-ol. Deprotection of the alcohol with/?-toluenesulfonic acid afforded (Z)-9-methyldec-4-en-l-ol. Pyridinium chlorochromate oxidation of the alcohol resulted in the isolation of the labile (Z)-9methyldec-4-enal. Final Wittig reaction with (4-carboxybutyl) triphenylphosphonium bromide in THF/DMSO resulted in the desired (5Z,9Z)-14-methylpentadeca-5,9-dienoic acid (14). The synthesis of (5Z,9Z)-24-methylpentacosa-5,9-dienoic acid (15) was also reported in the same communication [52], and it started with commercially available 10-bromodecan-l-ol that was oxidized with PCC to 10-bromodecanal, followed by Wittig coupling with 4-methyl-lpentyltriphenylphosphonium bromide resulting in the (Z)- and {E)A' bromo-14-methylpentadec-lO-ene, as shown in Fig. (18). Catalytic

84

OH

»

..^V^^^v^Br a

OH p-^^^«

Hi

p^^CHO

Br

;V

p^/^-^Br.

"'^ ^"^ ^Br

Br

OTHP

.^^J^Xx-'^/OH

Vr CO2H

14 n = 1 15. n = 11

0 PBr3 (70%); //) PPhj/CgHe (95%); ///) PCC/CH2CI2 (83%); iv) 2.5 M /i-BuLi, b, THF/DMSO; v) H2, 5% Pd/C, hexane (92%); vi) 2.5 M w-BuLi, HMPA/THF, a or c; vii) Hz/Lindlar; vm)p-TsOH, MeOH, 45^C; /x) PCC/CH2CI2 (84-89%); x) Ph3P^(CH2)4C02H Bf, 2.5 M w-BuLi, THF/DMSO. Fig. (18). Synthesis of Z50-A5,9-16:2 and /5o-A5,9-26:2 fatty acids

hydrogenation afforded 1 -bromo-14-methylpentadecane. Lithium acetyiide coupling in THF/HMPA resulted in the isolation of the tetrahydropyranyl-protected 19-methyleicos-4-yn-l-ol. In this case, the alcohol was deprotected first with /7-toluenesulfonic acid in methanol, affording 19-methyleicos-4-yn-l-ol that was further hydrogenated with Lindlar's catalyst and quinoline to (Z)-19-methyleicos-4-en-l-ol, Fig. (18). Pyridinium chlorochromate oxidation of the alcohol resulted in (Z)19-methyleicos-4-enal. Final Wittig reaction with (4carboxybutyl)triphenylphosphonium bromide resulted in the desired (5Z,9Z)-24-methylpentacosa-5,9-dienoic acid (15). The synthesis of an iso-anteiso pair of 5,9-21:2 fatty acids, namely the 19-methyleicosa-5,9-dienoic acid (16) and 18-methyleicosa-5,9-dienoic acid (17), prominent constituents of the phospholipids of the sponge

85 Erylus formosus [45], was also reported by Kulkami et al. [53]. As shown in Fig. (19), a common bifunctional nine-carbon aldehyde was used for both syntheses. Both syntheses started with the DHP monoprotection of 1,4-butanediol, followed by subsequent PCC oxidation to the corresponding aldehyde. Reaction of this butanal with (4-carboxybutyl)triphenylphosphonium bromide resulted in the corresponding THP protected olefmic acid that was further deprotected and esterified to the methyl ester. A second oxidization with PCC afforded the desired intermediate methyl 8-formyloct-5-enoate which secured the cis A5 double bond in the final targets, Fig. (19). The

H0(CH2)40H

THPO(CH2)3CHO

O

Me02C(CH2)3

n

A,

(CH2)8C02Me

w7/-x

^'^(CH2)8CH20H

(CH2)2CHO

1^

>

X/-X//

''^^(CH2)9PPh3^Br-

•

^COzH

H ZZ

Jl

(CH2)80H

I ^"'^--'''^(CH2)8PPh3*Br-

^"^(CH2)80H

x/V-xv *•

1^ '^^^^'^(CH2)80H

xvi-xvii ^

^COjH

xviii-xix ^

/) DHP/H"'/CH2Cl2; //) PCC/NaOAc/CHjCIj; ///) Ph3P=CH(CH2)3C027dimsyl/DMSO; /v) MeOH/H"; v) PCC/CH2CI2; v/) CHgPPhjy dimsyl/ DMSO; v/7) LAH/EtjO; viii) H2/ 10%Pd-C/ EtOH; ix) PPh3-Br2; x) PPh3/CH3CN/ heat; xi) KOH/EtOH, H^; xii) a, dimsyl, DMSO; xiii) UgSO^, H^/70% aq. EtOH; xiv) EtPPh3l/dimsyl/DMS0; JCV) H2/10% Pd-C/ EtOH; ;cv/) Ph3P-Br2/CH2Cl2; xvii) Ph3P/CH3CN/heat; xviii) a, dimsyl, DMSO; xix) KOH, EtOH, H"^.

Fig. (19). Synthesis of iso-anteiso 5,9-21:2 fatty acids

86 /50-ramification in 16 was made from methyl 10-oxoundecanoate that was reacted with methylenetriphenylphosphorane followed by LAH reduction to 10-methyleneundecan-l-ol. The further hydrogenation of the methylene group with 10% Pd-C and subsequent bromination with Ph3P*Br2 resulted in 1-bromo-lO-methylundecane, from which the corresponding triphenylphosphonium bromide salt was prepared. Final Wittig reaction of this phosphonium salt with methyl 8-formyloct-5enoate and basic hydrolysis resulted in the synthesis of the 19methyleicosa-5,9-dienoic acid (16). In this particular case, three Wittig reactions were utilized to introduce both double bonds as well as the isoramification. The synthesis of the anteiso isomer, namely the 18-methyleicosa-5,9dienoic acid (17), started from a primary decynol that was converted to the hydroxylated methyl ketone through Hg(II) catalyzed hydration. Condensation of this ketone with ethylidenetriphenylphosphorane and further hydrogenation resulted in the necessary 9-methylundecan-l-ol, with the necessary anteiso functionality. After conversion to the corresponding Wittig salt, coupling with methyl 8-formyloct-5-enoate, and saponification, the expected 18-methyleicosa-5,9-dienoic acid (17) was obtained. Fig. (19). In this case, three Wittig reactions were also used. One of the earlier syntheses of branched A5,9 fatty acids was accomplished by Djerassi's group in the elucidation of the biosynthesis of the unusual sponge fatty acid (i?)-(-)-22-methyl-5,9-octacosadienoic acid (18), a phospholipid fatty acid first identified in the sponge Aplysina fistularis [54]. They established the stereochemistry of the carbon bearing the methyl group in 18 as R, This synthesis started with commercially available (^)-(+)-pulegone that was transformed in ten steps to (J?)-l-bromo-3-methylnonane, Fig. (20). The sequence involved conversion of (i?)-(+)-pulegone to (i?)-(+)-citronellic acid, which was then reduced with lithium aluminum hydride and the resulting alcohol protected with DHP. Ozonolysis of the double bond, NaBH4 reduction, and tosylation was followed by the dilithium tetrachlorocuprate (Li2CuCl4) catalyzed cross coupling with «-propylmagnesium bromide. Final deprotection with acid, tosylation, and reaction with lithium bromide in acetone afforded the stereochemically fixed (i?)-l-bromo-3methylnonane. The A9 cis double bond in (i?)-(-)-22-methy 1-5,9octacosadienoic acid (18) was made, as in previous strategies, from

87

(lZ,5Z)-l,5-cyclooctadiene that was carefully monoepoxidized. This epoxide was opened with H5IO6, and after NaBH4 reduction, the 1,8-oct4-enediol was obtained, monoprotected with DHP, tosylated, and then elongated six carbons in a Li2CuCl4 catalyzed coupling with Me3SiO(CH2)6MgCl. This step was followed by deprotection of the trimethylsilyl group with base, further tosylation, and a third Li2CuCl4

CO2H OTHP

OTHP OH

OTHP

0-^0° IJ-'^^-^'^OTHP

^ CC™ — CC:

OTHP ^OH

xvii-xviii

(i

CO2H

i) HCl; ii) KOH; Hi) LAH, EtjO; iv) DHP, H^; v) O3/ NaBH4; v/) TsCl, Py; v//) LizCuC^, «-PrMgBr; v/7/) TsOH, MeOH; ix) TsCl, py; x) LiBr, acetone; xi) /w-CPBA, CHjClj; xii) HsIG^; xiii) NaBU^; xiv) DHP, H"^, EtzO; xv) TsCl, pyridine; xvi) LijCuCU, Me3SiO(CH2)6MgCl; xvii) K2CO3, MeOH; viii) TsCl, pyridine; xix) LizCuC^, Grignard reagent of a; xx) TsOH, MeOH; xxi) PDC, CH2CI2; xxii) Br'Ph3P''(CH2)4C02H, KH, Me2S0, THF. Fig. (20). Synthesis of (/?)-22-methyl-5,9-octacosadienoic acid

catalyzed coupling with (i?)-l-bromo-3-methylnonane, affording (i?)-17methyltricos-4-en-l-ol after deprotection, Fig (20). Final oxidation of the alcohol to the aldehyde, and Wittig coupling with (4-carboxybutyl)

triphenylphosphonium bromide afforded the desired (i?)-22-methy 1-5,9octacosadienoic acid (18). In this synthesis, the 9Z double bond was obtained from (lZ,5Z)-l,5-cyclooctadiene, the 5Z double bond from a Wittig reaction, and the 22R methyl group from naturally occurring (i?)(+)-pulegone [54]. Cyclopropylidene A5,9-Fatty Acids A recent report describes the isolation of a new class of A5,9 fatty acids from the Australian sponge Amphimedon sp., dubbed amphimic acids. They possess the unusual cyclopropylidene functionality [51]. These compounds were particularly interesting because they displayed DNA topoisomerase I (topo I) inhibition at ICso's of 0.47-3.0 \xM [51]. The HO^^^'"^(CH2)i5CH3

•

HO''^^''^(CH2),5CH3

•

L JL (CH2),5CH3

9"

Ph2(0)P

^^-^(CH2),5CH3

' ^ - \

OH

^^-^(CH2),5CH3

""'^

Ph2(0)P^"^(CH2),5CH3

OH

TBDPSO^^.^v^^A7\ Ph2(0)p'^^^"2^»^^"^

TBDPSO

-

-

-

-

-

-

-

-

-!!^=^

-(CH2).5CH3

HO^''^-^""'^^^^""'^(CH2),5CH3

—

_

^ ._

_

_

_

_

-(CH2),sCH3

'*(CH2),5CH3

19

/) /-BuOOH, (+)-DET, Ti(0-/-Pr)4, CH2CI2, -20°C and then -10°C (76%); //) Red-Al, THF, 0°C (94%); ///) I2, PPhj, HMPA, toluene, rt (72%); /v) PPhj, CaCOj, CH3CN, 80°C; v) NaOH, THF/HjO, SO^'C (72%, 2 steps); v/) MsCl, EtjN, CH2CI2, 0°C (100%); vii) NaN(TMS)2, THF, 0°C (87%); viii) TBDPSO (CH2)3CHO, LDA, THF, -78°C then 0°C; ix) BU4NF, THF, rt (71%); jc) NaH, DMF, 60-70% (80%); xi) DMSO, (C0C1)2, EtjN, CH2CI2, -78°C then 0°C (91%); xii) TBDPS0(CH2)5P'^Ph3r, NaN(TMS)2, toluene, rt (100%); xiii) BU4NF, THF,rt(100%); xiv) DMSO, (C0C1)2, EtjN, -78°C then 0°C (79%); ;cv) NaC102, NaH2P04,2-methyl-2-butene, THF,rt(97%). Fig. (21). Synthesis of amphimic acid A

89 synthesis of one of these acids, namely amphimic acid A (19), started with the Sharpless asymmetric epoxidation of (£r)-2-nonadecen-l-ol, affording the corresponding epoxide, as shown in Fig. (21). Regioselective reduction of the epoxide with Red-Al provided (5)-l,3nonadecanediol. Controlled monoiodination of the diol afforded (iS)-liodononadecan-3-ol, which was subsequently converted into a mesylated phosphine oxide as shown in Fig. (21). Cyclization of the mesylate with NaN(TMS)2 resulted in the optically active cyclopropylphosphine oxide, which after a Wittig-Homer reaction with A'{tert' butyldiphenylsiloxy)butanal resulted in a diastereomeric mixture of cyclopropyl P-hydroxyphosphine oxides. This diastereomeric mixture was separated using silica gel column chromatography. Deprotection of the correct stereoisomer and sodium hydride catalyzed elimination of the P-hydroxyphosphine oxide resulted in a cyclopropylidene with the E stereochemistry, Fig. (21). After Swem oxidation of the alcohol, Wittig reaction, deprotection of the /^rr-butyldiphenylsiloxy group, a second Swem oxidation, and oxidation of the aldehyde to the acid with NaC102, amphimic acid A (19) was obtained. Non-methylene Interrupted Fatty Acids In the family of non-methylene interrupted fatty acids, the total syntheses of the unusual marine acids (IIZ, 15Z)-ll,15-eicosadienoic acid (20) and (lOZ, 15Z)-10, 15-eicosadienoic acid (21), have been reported [55-56]. The (IIZ, 15Z)-11,15-eicosadienoic acid (20) was identified for the first ^ ^

/-//•

<^^(CH2)8C02H

^^^v^OH

i^ _

.

///

•

Bi
^ ^ ^

_

•

BrPh3P*(CH2)ioC02CH3

.xv^^OH

15

20 /) HBr, BrzOz; //) MeOH/H^; ///) PPhj; rv) LiNHz/w-BuBr; v) Hj/ P(2)-Ni; v/) FCC; v/0 NaCH2S(0)CH3, a; viii) KOH (ale), H"^.

Fig. (22). Synthesis of (IIZ, 15Z)-11,15-eicosadienoic acid

90

time in the Caribbean sponge Amphimedon complanata [57], while the (lOZ, 15Z)-eicosadienoic acid (21) originated in the Mediterranean opisthobranch Haminaea templadoi [58]. The stereochemical control in the synthesis of the (IIZ, 15Z)-ll,15-eicosadienoic acid (20) was achieved by a combination of acetylide coupling and Wittig reaction, as shown in Fig. (22). This synthesis started from one side with the preparation of (4Z)-non-4-enal from a five carbon acetylenic alcohol that was alkylated with «-butylbromide and LiNH2 to 4-nonyn-l-ol. Hydrogenation under P(2)-Ni afforded (4Z)-nonen-l-ol, which was oxidized with PCC to the expected (4Z)-non-4-enal, Fig. (22). From the other side, the necessary 10-carbomethoxydecyltriphenylphosphonium bromide salt was prepared from 10-undecenoic acid via the addition of hydrogen bromide and subsequent reaction with triphenylphosphine in refluxing acetonitrile. Final Wittig Z-olefination was accomplished by reacting the 10-carbomethoxydecyltriphenylphosphonium bromide salt with (4Z)-non-4-enal, which resulted in the expected acid 20 after saponification [55]. The stereochemical control in the synthesis of (lOZ, 15Z)-10,15eicosadienoic acid (21) was achieved by selective hydrogenation of a double acetylide as shown in Fig. (23). This synthesis started with the conversion of 4-nonyn-l-ol to l-bromo-4-nonyne with PPh3Br2 that was coupled with 10-undecynoic acid, resulting in the 10,15-eicosadiynoic acid. This double acetylide was further hydrogenated to the (lOZ, 15Z)10,15-eicosadienoic acid (21), as shown in Fig. (23). CH3(CH2)3 —

n

ZZ

(CH2)30H

(CH2)8C02H

15

- L ^

—!-•

CH3(CH2)3-=-(CH2)3Br

CH3(CH2)3-^r-(CH2)3-^r-(CH2)8C02H

10

21

0 PhsP Brj/pyridine/CHsClz; ii) n-BuLi/HMPAn'HF,a; in) H2/ P-2 Ni/EtOH/HjNCHjCHsNHj; iv) MeOH/H"^. Fig. (23). Synthesis of (lOZ, 15Z)-10,15-eicosadienoic acid

91

CONCLUSIONS The Wittig reaction continues to occupy a pivotal role in many of the syntheses of the twenty-six different marine fatty acids presented in this review. As more fatty acid structural motifs continue to be identified in marine organisms, a synthetic response to such an activity will continue to rise in parallel, since it has become an integral part of the characterization of marine fatty acids. In most cases isolated marine fatty acid mixtures are complex, normally resolvable by capillary gas chromatography, and the fatty acids of interest are typically found in small amounts. The synthesizing of new marine fatty acids and obtaining sufficient amounts of them are essential to fully characterizing their structures and understanding their fascinating biological activities.

ABBREVIATIONS 9-BBN BHT DHP DIBAL DMF DMSO HMPA LAH LDA MIC PCC PDC SAM TBAF TBDMS TBSCl THF TMS-CN

9-Borabicyclo[3.3.1 Jnonane Butylated hydroxytoluene 3,4-Dihydro-2//-pyran Diisobutylaluminum hydride //,A^-Dimethylformamide Dimethyl sulfoxide Hexamethylphosphoric triamide Lithium aluminum hydride Lithium diisopropylamide Minimum inhibitory concentration Pyridinium chlorochromate Pyridinium dichromate S-adenosylmethionine Tetrabutylammonium fluoride rer/-Butyldimethylsilyl /er/-Butyldimethylsilyl chloride Tetrahydrofuran Trimethylsilyl cyanide

92

ACKNOWLEDGEMENTS The author gratefully acknowledges grants from the National Science Foundation (NSF-MRI and NSF-MRCE programs); the National Institutes of Health (NIH-MBRS and NIH-MARC programs); and the University of Puerto Rico (FIPI program) that supported most of the author's research for the last fifteen years.

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

S.F. Lie Ken Jie, M.; Pasha, M.K.; Nat. Prod Rep., 1998, 75, 607-629. Adamczeski, M.; Reed, A.R.; Crews, P.; J, Nat. Prod., 1995, 58,201-208. Rodriguez, A.D.; Cobar, O. M.; Padilla, O. L.; J. Nat. Prod., 1997, 60, 915-917. Stierle, D.B.; Faulkner, D.J.; J. Nat. Prod, 1991, 54, 1134-1136. Carballeira, N.M.; Pagan, M.; Rodriguez, A.D.; J. Nat. Prod,, 1998, 61, 10491052. Hattori, M.; Miyachi, K.; Hada, S.; Kakiuchi, N.; Kiuchi, P.; Tsuda, Y.; Namba, T.; Chem. Pharm. Bull, 1987, 35, 3507-3510. Carballeira, N.M.; Pagan, M.; J. Nat. Prod., 2000, 63, in press. Carballeira, N.M.; Maldonado, M.E.; Lipids, 1989,24,371-374. Mori, K.; Wu, J.; Liebigs Ann. Chem., 1991, 8, 783-788. Kochansky, J.; Aldrich, J.R.; Lusby, W.R.; J. Chem. Ecol, 1989, 75, 1717-1728. Gilles, B.; Joseph, M.; Emile, G.; Nicole, B.; Jean-Michael, K.; Lipids, 1992, 27, 779-784. Rane, S.S.; Chadha, M. S.; Mamdapur, V.R.; Bioorg. Med Chem., 1993, 7, 399401. Li, Y.; Huang, C; Li, W.; Li, Y.; Synth. Commun., 1997, 27,4341-4348. Anjaneyulu, A.S.R.; Prakash, C.V.S.; Mallavadhani, U.V.; Phytochemistry, 1991, i(?, 3041-3042. Carballeira, N.M.; Emiliano, A.; Sostre, A.; Restituyo, J.A.; Gonzalez, LM.; Colon, G.M.; Tosteson, C. G.; Tosteson, T.R.; Lipids, 1998, 33,627-632. Spencer, G.F.; Payne-Wahl, K.; Plattner, R.D.; Kleiman, R.; Lipids, 1979, 14, 72-74. Moss, C.W.; Holzer, G.; Wallace, P.L.; Mollis, D.G.; J. Clin. Microbiol., 1990, 28, 1071-1074. Courdec, F.; Lipids, 1995,30, 691-699. Shirasaka, N.; Nishi, K.; Shimizu, S.; Biochim. Biophys. Acta, 1995, 1258, 225227. Shirasaka, N.; Nishi, K.; Shimizu, S.; Biochim. Biophys. Acta, 1997, 1346, 253260. Carballeh-a, N.M.; Sostre, A.; Stefanov, K.; Popov, S.; Kujumgiev, A.; Dimitrova-Konaklieva, S.; Tosteson, C.G.; Tosteson, T.R.; Lipids, 1997, 32, 1271-1275.

93

[22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53]

Carballeira, N.M.; Sostre, A.; Restituyo, J. A.; Chem. Phys. Lipids, 1999, 97, 8791. Pascal, J.C; Ackman, R. G.; Lipids, 1975,10, 478-482. Sano, Y.; Yukagaku, 1967,16, 606. Carballeira, N.M.; Shalabi, F.; J. Nat. Prod., 1994, 57, 1152-1159. Kulkami, B.A.; Chattopadhyay, A.; Mamdapur, V.R.; Synth. Commun., 1992, 22,2921-2925. Carballeira, N. M.; Cruz, C; Sostre, A.; J. Nat. Prod, 1996, 59, 1076-1078. Juagdan, E.G.; Kalidindi, R.S.; Scheuer, P.J.; Kelly-Borges, M.; Tetrahedron le/r., 1995, id, 2905-2908. Takanashi, S.; Takagi, M.; Takikawa, H.; Mori, K.; J. Chem. Soc. Perkin Trans. 1,1998, 1603-1606. Stork, G.; Grieco, P.A.; Gregson, M.; Tetrahedron Lett., 1969, 1393-1395. Ayanoglu, E.; Komprobst, J.M.; Aboud-Bichara, A.; Djerassi, C; Tetrahedron Lett., 1983,2^,1111-1114. Carballeira, N.; Sepulveda, J.; Lipids, 1992,27,12-1 A. Carballeira, N.M.; Emiliano, A.; Hemandez-Alonso, N.; Gonzalez, F. A.; J. Nat. Prod, 1998,(57,1543-1546. Carballeira, N.M.; Colon, R.; Emiliano, A.; J. Nat. Prod, 1998, 61,675-676. Soderquist, J.A.; Rosado, I.; Marrero, Y.; Tetrahedron Lett., 1998, 39, 31153116. Kobayashi, S.; Tsuchiya, Y.; Mukaiyama, T.; Chem. Lett., 1991, 537-540. Carballeira, N.M.; Colon, R.; Tetrahedron: Asymmetry, 1999,10, 3785-3790. Cardellina II, J.H.; Dalietos, D.; Mamer, F.-J.; Mynderse, J.S.; Moore, R.E.; Phytochemistry, 1978,17, 2091-2095. Gerwick, W.H.; Reyes, S.; Alvarado, B.; Phytochemistry, 1987,26, 1701-1704. Muller, C; Voss, G.; Gerlach, H.; Liebigs Ann. Chem., 1995, 673-676. Djerassi, C ; Lam, W.-K.; Ace. Chem. Res., 1991,24,69-75. Ayanoglu, E.; Huiting, L.; Djerassi, C; Duzgunes, N.; Chem. Phys. Lipids, 1988, 47, 165-175. Davidoff, F.; Kom, E.D.; Biochem. Biophys. Res. Commun., 1962, 9, 54-58. Carballeira, N.M.; Maldonado, L.; Lipids, 1986,21,470-471. Carballeira, N.M.; Negron, V.; J. Nat. Prod, 1991, 54, 305-309. Carballeira, N.M.; Emiliano, A.; Guzman, A.; Chem. Phys. Lipids, 1999, 100, 33-40. Mena, P.L.; Pilet, O.; Djerassi, C; J. Org Chem., 1984, 49, 3260-3264. Hahn, S.; Stoilov, I.L.; Tarn Ha, T.B.; Raederstorff, D.; Doss, G.A.; Li, H.-T.; Djerassi, C ; J. Am. Chem. Soc, 1988,110, 8117-8124. Carballeira, N.M.; Reyes, E.D.; Sostre, A.; Rodriguez, A.D.; Rodriguez, J.L.; Gonzalez, F.A.; J. Nat. Prod, 1997, 60, 502-504. Ayanoglu, E.; Walkup, R.; Sica, D.; Djerassi, C; Lipids, 1982,17, 617-625. Nemoto, T.; Yoshino, G.; Ojika, M.; Sakagami, Y.; Tetrahedron, 1997, 53, 16699-16710. Reyes, E.D.; Carballeira, N.M.; Synthesis, 1997,1195-1198. Kulkami, B.A.; Chattopadhyay, A.; Mamdapur, V.R.; Nat. Prod Lett., 1993, 3, 251-255.

94

[54] [55] [56] [57] [58]

Raederstorff, D.; Shu, A.Y.L.; Thompson, J.E.; Djerassi, C; J. Org. Chem., 1987, 52, 2337-2346. Kulkami, B.A.; Chattopadhyay, A.; Mamdapur, V.R.; Org. Prep. Proc. Int., 1993, 25, 333-336. Kulkami, B.A.; Chattopadhyay, A.; Mamdapur, V.R.; J. Nat. Prod., 1994, 57, 537-538. Carballeira, N.M.; Restituyo, J.; J. Nat. Prod., 1991, 5^, 315-317. Carballeira, N.M.; Anastacio, E.; Salva, J.; Ortega, M. J.; J. Nat. Prod, 1992, 55, 1783-1786.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

95

Some Aspects of the Chemistry of Secologanin LASZLO F. SZABO Institute of Organic Chemistry, Semmelweis University, Hogyes u. 7. H-] 092 Budapest, Hungary ABSTRACT: Secologanin, the representative of more than 650 iridoids and precursor of about 2500 indole, isoquinoline and related alkaloids, is a monoterpenoid glucoside. Its aglucone part is rich in functional groups and stereogenic elements. Our purpose was to study some aspects of its chemistry, and to prepare a bioorganic background for the understanding of its biochemical role. Our work directed mainly toward the reactions of N nucleophiles (including biogenic amines) and the chemical deglucosylation, which opened the way for further chemical transformations. During our work, several new heterocyclic ringsystem were constructed, all the three carbon bonds attached to the single stable chiral center of the molecule could be selectively cleaved, and a new biomimetic aromatization was discovered. NMR technics and graph analysis mainly contributed to the interpretation of the stereochemistry and formation of the reaction products from single educts.

INTRODUCTION Secologanin (1) of Fig. (1) is the most important representative of the iridoids (la), a special type of cyclopentanoid monoterpenes. (1) belongs the subclass of secoiridoids (lb), in which the bond C-7-C-8 is cleaved [1]. The number of known iridoids is more than 650 [2]. The biosynthesis of iridoids Uke that of other monoterpenes starts jfrom two C5 units (Ic) whose precursor was considered to be mevalolactone [3]. However, recently it was proved that the biosynthesis of some classes of terpenoids [4,5], and specially of secologanin [6] rurmed from 1-deoxy-D-xylulose5-phosphate as an alternative or even main route. The biochemical importance of secologanin is indicated by the fact that the coupling reaction with tryptamine and dopamine is the first step in the biosynthesis of about 2500 indole, isoquinoline and related alkaloids [2, 7], the most typical of which are shown in Fig. (2). From the late 1960s several research groups worked on the chemistry of secologanin. The aim of our work was to study some aspects of its chemistry, and to prepare a bioorganic background for the understanding of its biochemical role. Our work was directed mainly toward the

96 reactions of N nucleophiles (including biogenic amines) and the chemical deglucosylation, which was expected to open the way for further chemical transformations. The present paper is not a comprehensive review on the chemistry of secologanin but an account about our work done up till now, and completed by some results of other research groups.

iridoid skeleton

secoiridoid skeleton

2nd o.

arrangement of the cyclopentane subunits

3 centers of chirality in the aglucone subunit

• -m primary aglucone

1 stable center of chirality in the aglucone

cleavage of bonds and formation of rings

Fig. (1). Overview about the structure and properties of secologanin.

For our investigations, secologanin was isolated from Lonicera xylosteum L. by a method elaborated in our Institute [8]. General chemical characterization of secologanin The basis of the great variety of the compounds derived from secologanin (1) is its exceptional structure. Although the aglucone part of the molecule contains only 10 carbon atoms having different oxidation levels (except 4^^'), each of them can undergo chemical reactions. (1) contains four carbonyl groups (in free state, in acetal form, as a part of an enolether system or in ester group, respectively) and two C=C bonds, one of them asymmetrically substituted (possibility of E-Z isomerism) and in conjugation with the ester group. The hypothetic compound immediately formed after deglucosylation is called primary aglycone (Id) and it may exist in the equilibrium of many structural and stereoisomers (see later). These are represented by the all-oxo form (le), in which the end-standing second C=C bond can easily be shifted from isolated to conjugated

97 position (le->lf). The aglucone part of the molecule has three centers of chirahty (possibility of RS isomerism), two of which (C-1 and C-9) are easily epimerized by acid or base after removal of the p-Dglucopyranosyloxy group, however, the third one (C-5) keeps its configuration not only in the aglucones of simple secologanin derivatives, but also in all such alkaloids, in which subsequent reactions did not involve this center [9]. These structural elements give many possibilities not only for transformation of the functional groups, but formation of new rings, as well as cleavage and rearrangement of C-C bonds, too (Ig).

.iiOgIc

H3CO

H3CO

geissoschyzine OCH3

(substructures derived from secologanin are in bold lines)

Fig. (2). Alkaloids derived from secologanin.

Simple chemical transformations of secologanin Already its dissolution in protic solvents initiated changes in the structure of secologanin and resulted multicomponent equilibria, therefore the NMR spectra of secologanin and some of its derivatives should be recorded in aprotic solvents (chloroform, benzene, acetonitril, etc). The

98 equilibria were set in at room temperature in 1-2 hr. Although the compounds of the solvent interactions could not be isolated without shifting the equilibria, the transformations were easily followed in deuterated solvent by ^H NMR spectroscopy as shown in Fig. (3). in deuterium oxide

in deuterated methanol

ilOgIc

ilOgIc

D

(-D3COD)

H NMR chemical shifts in 6. Figures in plain are measured in D2O, those in italics in D3COD

Fig. (3). Acid-base and solvent equilibria of secologanin.

The signals of H-3 and H-7 were sensitive for changes both in the aldehyde and the ester group. The position of the equilibrium of hydrate and hemiacetal (5) depended on the structure of the solvent [10]. The simple chemical transformations of secologanin are summerized in Fig. (4). It is well known from the early literature that secologanin (1) can easily be hydrogenated at position 8,10 to (2) without saturation of the conjugated C=C bond [1]. The presence or absence of the 8(10) double bond (vinyl and ethyl, natural and dihydro derivatives, respectively) often

99 resulted in considerable difference of the chemical properties of the appropriate derivatives.

r~\ Ogle

H;,CO

,,ilOglc(Ac)4

,,lOglc(Ac)4

H""T O

^* X = O o r H N

O

T

9a

MCPBA=3-cliloroperben2oic acid

Fig. (4). Simple chemical transformations of secologanin.

Enzymatic deglucosylation with p-glucosidase gave not the simple primary aglucone, but its bicychc derivative both in the natural (3a) and the dihydro series (3b) [11, 12]. However the acetal (4a), in which the formyl group is protected gave the expected primary aglucone (analogously to Id). Removal of the glucosidic unit either by acid or by base resulted in fragmentation of the aglucone, as will be discussed later.

100

The formyl group could be protected reversibly by acid-catalytic acetalation with ethylene glycol (4a) [13] or methanol (4b) [14]. By temporary protection of the formyl group, methanol could be added to the conjugated C=C bond (6) in aqueous-methanolic sodium hydroxide solution (and with subsequent reesterification) [15]. Acetylation of secologanin with acetic anhydride in the presence of a few drops of pyridine gave the 0',0',(9',0'-tetraacetylsecologanin (7), which could be oxidized in aqueous acidic medium with chromium trioxide to secoxyloganic acid and reesterified with diazomethane to (8). The acetyl groups could be removed by the Zemplen method without changes in the methoxycarbonyl group [16]. After the protection of the formyl group in acetal form (7a), the isolated double bond could be oxidied. 3-chloroperbenzoic acid (MCPBA) epoxidized at first the isolated (9), later also the conjugated (9a) double bond [17]. Osmium tetroxide oxidized only the isolated double bond into a glycol unit (10) [18]. The reactivity of secologanin derivatives against N nucleophiles was influenced mainly by the four potential carbonyl groups. As two of them are protected (C-1) or masked (C-3), in secologanin the most reactive point is C-7. C-11 is rather resistant to external nucleophiles, but sensitive for internal nucleophilic attacks coming from the substituents of C-7. C-3 is sufficiently reactive only in the ester and lactone derivatives, and not reactive in the amides and lactams or in compounds where the cyclic 0-2 is exchanged against N. C-1 is resistant to external nucleophiles in the glucoside form, but reactive in the aglucones. As a further general rule it was observed that the exocyclic alkylamino group had the tendency to take spontaneously the endocyclic position by ring opening and ring closure (ANRORC reaction), according to the higher stability of the new azacycle (tetrahydropyridine ring) over the original oxacycle (dihydropyran ring). The following details support these general remarks. With primary amine, the first product was a labile Schiff base (11) having a signal of H-7 at -7.7 6. However, the excess of the amine (especially in the case of methyl amine) easily transformed the Schiff base into a 7-substituted lactam (12). The substituent of C-7 (methylamino, methoxy or hydroxy) entered first into P equatorial orientation, but then rapidly epimerized to the a axial one, which is stabihzed by a conjugation with the non-bonding orbital of the lactam nitrogen. The methylamino group could easily by exchanged for other groups (methoxy,

101 hydroxy, or even other amino groups) (e. g. 13) [19]. As with secondary amine the formation of the lactam was not possible, it gave 7-substituted lactone (14). Both the lactam and the lactone were formed by intramolecular interaction of the nucleophile of C-7. When the formyl group was protected in form of ethylene acetal or by coupling with the appropriate biogenic amine, the methoxycarbonyl group proved to be rather resistant to nucleophiles, and reactions runned at C-3 (see later). Tertiary amine could not be incorporated into secologanin, but in the presence of a protic solvent (water or methanol), catalyzed the lactonization (15) [20]. Bifunctional amines gave the possibility to annellate a further five- or six-membered ring with a second heteroatom (oxygen, nitrogen, etc.) to the secologanin subunit (16) [20,21]. Chemical transformations of bakankosine derivatives The reactions are summerized in Fig. (5). Bakankosine (15, R=H) , a natural lactam derivative was prepared by Tietze in the reaction of secologanin with benzylamine, followed by reduction with sodium tetrahydridoborate, and after spontaneous lactamization by removal of the benzyl group by catalytic hydrogenation [22]. Further A^-substituted bakankosine derivatives (17a, R is alkyl) were obtained by reduction of the appropriate Schiff bases (11) likewise with sodium tetrahydrido borate, accompanied again by spontaneous lactamization [23]. The vinyl group (in the natural series, e.g. 17a) was easily saturated into an ethyl group (dihydro series, e. g. 17b). The bankosine derivatives proved to be rather stable compounds and could be attacked by nucleophiles neither at C-1, nor C-3. However, after deglucosylation with P-glucosidase or with acid catalysed hydrolysis, the aglucone underwent nucleophilic reactions (see later). The deglucosylation with enzyme resulted in the formation of the primary aglucone (18a, 18b) in both cases in the form of an epimeric pair at C-1. The primary aglucone was easily transformed into an acetal (19a, 19b) in methanol with acidic catalysis. The acidic deglucosylaton gave in the natural series a rearranged epimeric pair (20), in the dihydro series an unrearranged one (23) with inversion of the configuration at C-9. The complicated process was analyzed in detail (see later) [24]. In the presence of a strong base (sodium ethanolate in ethanol, sodium hydride in dimethylsulfoxide) the oxacycle could be cleaved in both (18a) and

102

(23) (strong bathochromic shift in the UV spectrum, see e. g. 21), and alkylated on 0-2 into the enol ether (22) and (24), respectively, having Z geometry in chloroform but E geometry in dimethylsulfoxide. (18a) was isomerized in triethylamine to (20). The dihydro aglycone (18b) gave with ethanolic sodium ethanolate a double lactone (25). Its formation was interpreted as the result of an intramolecular Claisen-Tishchenko reaction.

0

30 X=OorNH

Fig. (5). Chemical tranformations of bakankosine derivatives.

As (20) gave with amines rather unstable products, the reactions with N nucleophiles were investigated in the dihydro series in details. Primary amine attacked the aglucone (23) at C-1, but in the excess of the reagent the oxacycle of the first product (26) was transformed into the azacycle (27) by ring-opening and ring-closure. Like at position C-7, the exocyclic amino group of C-1 could be exchanged for hydroxy or methoxy group (or even another amino group) as well (e. g. 28). The secondary amine

103

piperidine attacked the aglucone in the same position, but the oxacycle remained intact (29). By using bifunctional primary amine, further fiveor six-membered rings containing a second heteroatom could be annellated (30) [results partly not pubHshed yet]. Chemical transformations of sweroside derivatives The reactions are shown in Fig. (6). Reduction of secologanin (1) and its dihydro derivative (2) with sodium tetrahydridoborate in protic solvent gave sweroside (31a) and its dihydro derivative (31b), respectively, [25] which could smoothly be deglucosylized by {i-glucosidase (e. g. to 32), however, the acidic hydrolysis gave no well-defined product. The dihydropyrane ring of (32), Hke that of the bakankosine derivative (23) was cleaved by strong nucleophile in dimethylsulfoxide, and alkylated on 0-2 (33). In ethanol the lactone ring was opened by sodium ethanolate, and after acidification the bridged ester aglucone (34) was formed [unpublished results]. For N nucleophihc attack three sites could be found: C-1, C-3 and C11. Their reactivity is interrelated. Interpretation of the reactions with secondary amine was easy. Unlike to bakankosine, sweroside could be deglucosylized by piperidine, but the amino group was attached to C-3 instead of C-1 (35). It means that the ester group sufficiently activated C3, and the primary adduct was stabilized by deglucosylation and tautomerization of the C=C bond fi-om isolated to conjugated position. Although in the dihydro series the glucoside did not give a well defined product with piperidine, it could be incorporated into the aglucone of dihydrosweroside, but, according to the expectation, at C-1 (36). Propylamine as a primary amine attacked C-3 of the glycosides (31a) and (31b), and afforded the first adduct (37a, 37b), which was stabilized by deglucosylation and spontaneous isomerization into the more stable azacycle (38a) and (38b, respectively) by ring opening and ring closure. In the natural series, even a dehydrated product (39) could be isolated. In the dihydro series, in addition to (38b), a further, bridged bicycUc compound (42) was isolated fi-om the same reaction mixture. (42), in which the ester group was transformed into an amide group, could be obtained from the aglucone of the dihydrosweroside (32), too, but not from (38b). Evidently, N-2 deactivated the C-11 of the ester group.

104

O

35 t^

\

\

...•Ogle p-glucosidase ^ from 31b*

I H, - Pd NaBH4 H2O

0-

f

>^l.C,HsONa +

0-

31b8,10-dihydro H2NCH2CH2CH: from 31a

6v^

J ^

N>^

H,NCH,CH,CH. from 31b

H2NCH2CH2CH3

from 37a

Fig. (6). Chemical transformations of sweroside derivatives.

The rather compHcated situation is not completely clear. From the aglucone (32), the formation of the amide (42) should start with two independent nucleophilic attacks on the same molecule, i. e. at C-1 (to 41) and subsequently at C-11 (to 41a). In the case of glucoside (31b), the two attacks runned on different molecules affording the two intermediates (37b) or (40). This latter should give by a second amino attack the same intermediate (41a), but the mechanism of the deglucosylation is not clear in this case. The common closing steps through (41b) to the final product (42) are usual transformations [26].

105 Acidic deglucosylation of simple bakankosine derivatives The reactivity of secologanin and related compounds is partially blocked by the glucosidic linkage, which effectively regulates the reactivity of secologanin at C-1 and C-3 as well as on the vinyl group. After removal of the P-D-glucopyranosyloxy unit, a cascade of reactions can start. The free aglucone may be represented as an all-carbonyl tautomer, e. g. (le), in which active H atoms are at C-6, C-9 and C-4, i. e. in a position to one (C-7 or C-1, respectively) or two (C-3 and C-11) carbonyl groups. In the aglucone the isolated C=C bond can easily be tautomerized to conjugate position, e. g. (If), which fact has important consequences: also this second C=C bond will be asymmetrically substituted (possibility of E-Z isomerism), a further electrophilic center will be developped at C-9, and the H-10 atoms become activated (H-D exchange). As in acidic hydrolysis, secologanin (1) and 7-substituted bakankosine derivatives (e. g. 13) underwent fragmentation (see later), and sweroside (31a) was destroyed without giving a well defined product, A^-methyl bakankosine (17a) and its dihydro derivative (17b) were selected as model compounds for investigation of the acidic deglucosylation of secologanin derivatives. In the stable educts, C-7 and C-11 had been inactivated by lactamization [24]. As it was shown in Fig. (5), in boiling aqueous diluted hydrochloric acid, (17a) and (17b) gave aglucones (20) (epimeric pair at C-8) and (23) (epimeric pair at C-1), respectively. In (20) the dihydropyran ring was cleaved, the C4 unit rotated around the bond C-5-C-9 and the cycle was reformed to C-8 with simultaneous appearance of a new formyl group. (23) proved to be a simple primary aglycone epimer pair. Treatment of either the educts (17a and 17b) or the products (20 and 23), under the conditions of the deglucosylation in deuterated solvent, resulted in D exchange of H-9 in both the natural and dihydro series, and in addition of H-10 in the natural one. These facts suggested the cleavage and closing of the dihydropyran ring in ring-chain tautomerisms, as well as changes in the configuration of C-9 by oxo-enol and C=C bond tautomerism.

106 possible transformations both in "natural" and dihydro aglucones in "natural" aglucons only

a

/

LrfH*

\

\

Estimated relative free enthalpies of formation in kJ/mol. Energy values in parentheses concern possible aglucones derived from the 8,10-dihydro aglucones * indicates H atoms and C atoms having H atoms which could be exchanged for D under the reaction conditions of deglucosylation both in educts and products Fig. (7). Energy graph of the deglucosylation of bakankosine derivatives.

The possible intermediates were organized into an energy graph shown in Fig. (7), which indicates also the relative enthalpies of formation (in kJ/mol) estimated on simple model reactions. The arrows show the direction of the decrease of the free enthalpy of formation. As each structure has two stereogenic elements (unsymmetrically substituted double bond and/or center of chirality), each formula represents four stereoisomers.

107

However, the configurational and conformational changes could not be considered at the estimation of relative free energies. The structure graph (published in the original paper [24]) involved 48 stereoisomers.

from natural glucoside

Fig. (8). Structure graph of the deglucosylation of bakankosine derivatives.

The simpHfied version of the structure graph presented in this paper as Fig. (8) indicates only those structures, which are on the shortest path from the educts to the products. (In the graph the stereochemical descriptors R, S, Z and E show the configuration of the eventual centers of chirahty at C-1, C-3, C-4, C-8 and C-9, and/or the conformation of the eventual C=C bond at C-3, C-8 and C-9, respectively. Indication of the configuration of a center of chirality precedes that of the conformation of

108 the C=C bond, and a stereogenic element in the "lower" region (C-1, C4) of the structure that in the "upper" region (C-1, C-8, C-9) of it. In both graphs H indicates structures, which are relevant in 8,10-dihydro derivatives, too. Structure numbers in Fig. (7) were italicized. In the interpretations of the changes it was considered that H-D exchange and epimerization at C-9 should run through enol-oxo equilibrium involving C-1 and C-9 at the level of structure 4(H) and/or 5(H), which have the highest energy in the reaction matrix. Equilibrium should go back till this level. As one can see from the energy graph (Fig. (7)), compounds isolated from the reaction mixture have the lowest relative free enthalpy of formation both in the "dihydro" and "natural" series ((20) corresponds to {12), (23) to {IH)), Actually, in the "natural" series, the estimated free enthalpy values of aglucone type 10 and 12 have nearly the same values. Although it could not be isolated or demonstrated, a small amount of (20a) corresponding to {10) can not be excluded to be formed during deglucosylation. However, it should be noted, that R. T. Brown and S. B. Pratt did isolate derivatives according to {10) under slightly different conditions of enzymatic deglucosylation [27]. Moreover, this type of substructure is well known in indole alkaloids of the yohimban type. The results are interpreted as follows (see Fig. (8)): According to experimental data, the enzymatic deglucosylation of (17a) and (17b) is accompanied only by simple epimerization at C-1, which can easily be explained by cleavage of the primary aglucone IRR(H) to 3ER(H) and recyclization either back into IRR(H) {=IRM' 18a-b) or into 1SR(H) (=15',9i?-18a-b) in a ratio of 9:1. The acidic deglucosylation is under thermodynamic control. In the case of the dihydro glucoside (17b), the events could likewise easily be interpreted. After hydrolytic removal of the glucosyloxy unit, the primary aglucone IRRH was spontaneously cleaved to 3ERH and tautomerized to 6RRH, which is already a member of the isomerization cycle, in which, through 5RZH and/or SREH, the configuration of C-9 could be epimerized. Retautomerization of 6RSH to 3ESH and recyclization to IRSH (=li?-23) and ISSH (=15'-23) in a ratio of 8:2 complete the reaction sequence. Trans orientation of C-6 and C-8 in the final product is favoured over the original cis one. In the acidic deglucosylation of the "natural" glucoside 17a, the mechanism is more complicated, because the vinyl group of structure {3) can easily be tautomerized into the more stable ethylidene group of

109 structure (7) (see the appropriate values of free enthalpy of formation in Fig. (7)), which could be recyclized not only into structure (9) but (after rotation around bond C-5-C-9) into structure {11), too. As the energy values of Fig. (7) show, formation of (72), through tautomerization of {11), is most favorable, being the most stable compound. Therefore, the shortest way from the educt to the products was explained as follows: the primary aglucone IRR isomerizes through 3ER to 6RR, and then further into the other members {5RZ, 5RE, 8RZ, 8RE, lORR, lOSR and 6RS) of the isomerization cycle. Finally 7EZ and 7EE could be formed, and recyclization of either of them to IISZ, USE, IIRZ and IIRE gave, after a second tautomerization, the most stable final aglucones 12SS (=85-20)and 12RS (=8i?-20) in a ratio of 6:4. The epimer ratio of the product components can approximately be estimated from the ^H-NMR spectrum. These ratios were interpreted by considering stabilizing anomer-type and destabilizing gauche interactions of the hgands of C-1 and C-9 in {IR'23=1RS) and {IS'23=1SSH) as well as of C-8 and C-9 in {8R'20=12RS) and {SS-20=12SS). Really, the aglucone epimers having the minimum number of gauche and maximum number of anomer-type effects were formed and isolated in higher amount (15'-23 and 85-20). As all steps (equilibria) are reversible (forwards and backwards) and catalized by proton (involving protonations and deprotonations), the complicated reaction matrix involves in the "natural" series 48 aglucones, 92 equilibria and 368 elementary steps, in the "dihydro" series 24 aglucones, 40 equilibria and 160 elementary steps. Although any of these structures and transformations can not be excluded, by graph analysis the shortest rational pathways were found for the interpretation of the events. The results could be used for the investigation of other cases in the bioorganic chemistry of indole and related alkaloids, too (see later). Fragmentation of simple secologanin derivatives As mentioned at the beginning of the previous chapter, in the acidic or basic deglucosylation, secologanin and some of its derivatives underwent deeper structural changes, which will be discussed in this chapter. The structural basis of these fragmentations is the special position of C-5, which has a stable S configuration with three different carbon ligands. The C2 and C4 ligands have one (C-7 and C-1, respectively), the C3 ligand two

no (C-3 and C-11) free or masked carbonyl groups all of them in p position to C-5 (see (If) of Fig. (1)). According to this special arrangement, each of the ligands could be detached from C-5 under different circumstances, depending on the other structural elements and the reaction conditions.

k^x:^ Fig. (9). Deglucosylation and fragmentation of secologanin and its ethylene acetal.

The cleavage of the C-4-C-5 bond is shown in Fig. (9). By acidic deglucosylation, secologanin (1) and its ethylene acetal (4a) gave surprisingly benzaldehyde (49). Probably, in the first phase, all acetalic groups as well as the ester group were hydrolized with formation of the aglucone represented by structure (43) in this case. In the second step, this intermediate underwent afragmentationalong bond C-4-C-5 according to a retro-Michael reaction (indicated by curved arrows in 43) and afforded a Cj (44) and a C3 (45) unit. Unfortunately, the latter fragment could not be detected, because, most probably, it was decomposed into carbon dioxide (46) and acetaldehyde (47). The sum of the oxidation levels of the carbon atoms in (44) already corresponded to that of benzaldehyde (49), which could be formed through (48) in an intramolecular aldol-type reaction (indicated likewise by curved arrows in 44) and aromatization by spontaneous elimination of water. It was mentioned previously that from sweroside no well defined product could be obtained, and from A^-

Ill

methylbakankosine (17a) and its dihydro derivative (17b) epimeric aglucone pairs (20) and (23), respectively, were formed under analogous conditions. As likewise mentioned previously, secologanin with primary amine gave the 7-substituted bakankosine derivative (12), with secondary amine the 7-substituted sweroside derivative (14). In both cases, the first site of the nucleophilic attack was the free formyl group. In secologanin ethylene acetal (6) where the formyl group was protected against nucleophilic attack, the site of the reaction was at C-3. The consequences are shown in Fig. (9). The adduct (50) of primary amine was automatically transformed by deglucosylation and subsequent ring opening and ring closure through (51) into a dihydropyridine derivative (52), which was analogous to (39). With secondary amine the result was again surprising, because from the reaction mixture methyl 3piperidinopropenoate (55) was isolated as the only product. For rationalization of the process, it was taken up that in the primary adduct (53) the p-D-glucopyranosyloxy group was removed in a base-catalyzed elimination, and (54) was formed, in which the structural conditions were appropriate for a re/ro-Diels-Alder reaction. This latter gave (55) and a C7 unit (56) again by the cleavage of bond C-4-C-5. Unfortunately, the hypothetic fragment (56) could not be obtained from the reaction mixture [28]. However, in tryptamine derivatives of secologanin, analogous reactions were observed, and both fragments of the same substrate were isolated (see later). Fig. (10) shows that in acid-catalyzed hydrolysis, 7-substituted bakankosine derivatives underwent an other type of fragmentation, in which bond C-5-C-6 or C-5-C-9, i. e. one of other two-carbon hgands of C-5 were cleaved. The reaction was studied on the oxazolo compound (57a) and its dihydro derivative (57b), but analogous changes were observed in other 7-substituted (hydroxy, methoxy or methylamino) A^alkylbakankosines as well.

112

U

|.,||H

U .iiOglc

Q

I ,,|H

^^X. = ^ ^

.J.OH

R=hydroxyethyl, methyl

3

57a 57b 8,10-dihydro

58b 8,10-dihydro |H30^

Fig. (10). Deglucosylation and fragmentation of 7-substituted bakankosine derivatives.

The reaction products depended on the presence (in the vinyl derivative 57a) or absence (in the ethyl derivative 57b) of the double bond of the side-chain. In both cases a pyridone derivative, (61) or (66, respectively, was formed, and from (57a) a C4 unit (probably crotonic aldehyde (62)), from (57b) a C2 unit (probably acetaldehyde (65)) was eliminated. In addition, from the vinyl compound (57a) afiirtherpyridone derivative (69) was formed as side product, which retained the C4 unit in the form of a C5 side chain. The educts could be deglucosylized by enzyme or aqueous acid to (58a) and (58b), and fiirther acidic treatment cleaved the oxazolidine ring giving the 7-hydroxy-aglucones (59a) and

113 (59b), respectively. These structures represented 12 (in the vinyl derivatives) or 6 (in the ethyl derivatives) structural isomers (and four times as many stereoisomers!), which were analogous to the structures shown in Fig. (7) and probably present in the reaction mixture. Our analysis started from these intermediates. The three most stable isomeric aglucones of them, (60), (63) and (67) corresponded to those of the bakankosine derivatives (20), (23) and (20a), respectively, and each of them were considered as a starting point of the formation of a pyridone aglucone (61), (66) and (69), respectively. The main driving force of the fragmentations is the formation of a stable pyridone system. However, in details the differences were important. The simplest case is the fragmentation of the aglucone (60) (corresponding to aglucone type (12) and formed from (57a)), in which the dihydropyran ring is already preformed for a r^/ro-hetero-Diels-Alder reaction, which could directly give the end product (61) (and probably 62). This fragmentation is impossible in the aglucone (63) (corresponding to aglucone type (IH) and formed from 57b). However, the most basic point of the system is the lactam O atom. It might be supposed that association of a proton to this site could polarize the total ring system (according to the curved arrows in (63)) and facilitate the fragmentation with elimination of a C2 unit (probably (65)). In the intermediate (64), by rotation (involving tautomerization and change in the geometry of the C=C bond), the C-1 atom could obtain a position, which is favourable to cyclization and formation the another pyridone system (66). The formation of the ring-opened side-product (69) might have been interpreted by a subsequent vinylogous intermolecular aldol reaction. However, (69) could not be prepared from the pyridone derivative (61) with, neither its amount increased by addition of crotonic aldehyde under the same conditions. Therefore, an intramolecular mechanism was looked for its formation. Structure (67) (corresponding to the third relatively stable aglucone type (10)) seemed to be a good starting point for the formation of (69). In (67), after elimination of water, the double bonds could be arranged into a position according to (68), which seemed to be favourable for a cycloreversion indicated by the curved arrows. Finally, in the strongly acidic medium the conjugated double bonds took up the trans-trans orientation to give (69) [results to be published].

114 Reinvestigation and completion of the coupling reaction of secologanin with biogenic amines One of the first reactions carried out with secologanin was its coupling reaction with tryptamine and dopamine. The pioneering work of Battersby and his group [29, 30] was later completed mainly by that of R. T. Brown and his co-workers [31-33]. It was demonstrated that the coupling reaction of secologanin (1) with tryptamine (70a) gave strictosidine (3iS'74a) and vincoside (3i?-74a) with low stereoselectivity [29]. Analogously, in the reaction of secologanin and dopamine 2-deacetyl-isoipecoside (IS81) and 2-deacetyl-ipecoside (li?-81) were formed [30]. However, the contemporary experimental techniques were limited in this respect, and some discrepancies and uncertainties were the consequences of this fact. Unfortunately, in both series the configuration of the newly formed center of chirahty (C-3 and C-1, respectively) were given to be opposite. Later, the assignments were corrected on the base of X-ray diffraction analysis of derivatives of vincoside [34] and ipecoside [35], but no direct proof was presented in the S series. The biosynthetic studies gave similar problems. Originally, vincoside (3R) and deacetylipecoside (IR) were considered as the precursors of the indole and ipecac alkaloids, respectively. However, Zenk and his group proved that the exclusive precursor of all indole alkaloids (including the 3R epimers of the la class) is strictosidine (35) [36, 37]. Moreover, it was found that the precursor of the alkaloid emetine (liS) is deacetylisoipecoside (IS), and that of the alkaloid glucoside alangiside (IR) is deacetylipecoside (li?) [38]. The coupling reaction with tryptamine has a single enzyme strictosidine synthase, in that with dopamine the activity of two enzymes was suggested. However, during purification, one of them disappeared [39]. In order to get insight into the stereochemistry of the coupling reactions, it was decided to reinvestigate and extend the reactions. Four type of coupling reactions were carried out, which are shown in Fig. (11) and Fig. (13). (The cyclic skeletons were numbered according to the biogenetic numbering [40]). For the description of the stereogenic elements of the products the following system was used: configuration of the new center of chirality (C-3 in tryptamine and oxotryptamine derivatives, C-1 in dopamine and histamine derivatives): R or S; type of conformation around C-14 (in dopamine derivatives C-11, in histamine derivatives C-10): 11, ... 33, (see Fig. (14)); conformation of the

115 dihydropyran ring (N or P); conformation of the tetrahydropyridine ring: N or P (in the tryptamine, dopamine and histamine derivatives) or configuration of C-7: R and S (in the oxotryptamine derivatives).

-H20. + H20 ,,iOglc(R2)^

I 71aR'=R2=H ^ 71b R'=4-bromobenzyl, R-=acctyl

21 in 3.V-75:S31NP ill 3/<-75: R12NN glc(R2),

72a R' = R2=H

72b R'=4"-broniobcnzyl. R2=acctyl

75 R2=H 3\:3/e=50:50

\,>^ NHR' N / 70aR'=H " 70b R'=4-bromobenzyl iOglc(R2)^

yOglc(R2)^

x,'-^

NHR'

76a(R'=H)

77aR'=R2=H 77bR'=benzyl, R2=acciyl

0 73aR'=R2=H 73b R'=4"-broniobcnzyl, R2=acciyl

3.S\7^-78b: S22NR 25% 78aR'=R2=H 3.S:3/J=? 78b R'=bciizyl, R2=acetyl 3S:3R=>95:<5

74aR'=R2=H

3A-74a:SllPP [3/J-74a:RllNN] 3.V-74b:S12NN 0 3/J-74b:RllNP 3.V:3^=50:50

74b R'=4"-broniobcnzyl. R2=acctyl 2S:3R=<5:>95

Oglc(R2)^ 3RJS-19: R12NS 75% 3RJR-79: R12NR 25% 79 R2=acclyl 3 .V:3y?=<5 : >95

Fig. (11). Coupling reaction of secologanin with tryptamine and oxotryptamine.

1. The coupling reaction of secologanin (1) or 0\0\0\0'' tetraacetylsecologanin (7) with tryptamine (70a) and Nb-(4bromobenzyl)tryptamine 70b was first studied according to Fig. (11). In the case of tryptamine, the first resuhs of [29] (except the configuration of C-3) could be confirmed, i. e. the formation of strictosidine and vincoside in an approximative epimeric ratio of 3iS':3i?=50:50, as well as the fact that lactamization of vincoside (37?-74a) to vincosamide (3/?-75) was considerably faster than that of strictosidine (35'-74a) to strictosamide {3875). By appropriate regulation of the coupling reaction, a good method

116 was elaborated for preparation of strictosidine because the lactam vincosamide could be removed from the acidic reaction mixture by simple filtration [41]. It seemed to be necessary for us to confirm the configuration and establish the conformation of strictosidine (3S'142i) [42]. The starting point of the analysis was the experimental fact that the compound was prepared from secologanin, the stereochemistry of which was known to be S at C-5 (corresponding to C-15 in strictosidine) [43]. Then, 2D NMR methods were used to determine unambigously the ^H- and ^^C-NMR chemical shifts, the ^H-^H and ^^C-^H coupling constants, and the 'H-^H NOE interactions in strictosidine (3)S'-74a). These spectroscopic parameteres and some theoretical considerations have made possible to select the single existing structure of the 648 possible stereoisomers (see later) i. e. to confirm the S configuration at the newly formed center of chirality at C-3, to establish the P helicity of the dihydropyran and tetrahydropyridine rings, as well as the S l l conformation around C-14 and the G l conformation around the glucosidic bridge [42]. The stereochemistry of the other products was established likewise by detailed 'H- and ^^C-NMR measurements. The basis for this analysis was the proved S configuration of C-15 (as mentioned previously), [43] and the R configuration of C-3 estabhshed by Hatchinson et aL in 0',0',0',0'-tetraacetyl-4-(4"-bromobenzyl)vincoside (3i?-74b) by X-ray diffraction analysis [34]. In order to prove the configurational relations among the ester and lactam derivatives, chemical correlations were carried out according to Fig. (12) in the four directions. 1. With sUght modification of the method of Hutchinson and co-worker, (37?-74b) was prepared as main product from tetraacetylsecologanin (7) and Nb-4-bromobenzyl tryptamine (70b) and transformed into 37? epimer of 0',0',0',0'-tetraacetyl-18,19dihydro-(75). 2. Vincosamide (3i?-75) was isolated from the coupling reaction of secologanin (1) and tryptamine (70a) by filtration, and transformed into the same compound. 3. From the mother liquid of the previous coupling reaction, strictosidine (35'-74a) was isolated, lactamized into strictosamide (3iS'-75), and transformed into the 3S epimer of 0',0',0',0'-tetraacetyl-18,19-dihydro-(75). 4. Finally, strictosidine {3374a) was 4-bromobenzylated and acetylated to (3iS'-74b), than transformed with appropriate reactions into the same lactam.

117

pH=5

35-74a 3/?-74a

' 1. 4-Br-C6H4-CHO, benzene . 2. NaBH4

separation from the 3a-H izomer by ciystallization

1. H2-Pd ,2. lactamization

' (CH3CO)20 pyiridine

spontaneous lactamization, the precipitate filtered out

l.(CH3CO)20 pyridine . 2. H2-Pd

> 3R-0',0',0',0'-tetraacetyl-18,19-dihydro-75

after evaporation of

NazCOs

l.(CH3CO)20 pyridine , 2. H2-Pd ,

1.4-Br-Cf,H4CH2CI, base , 2. (CH3C0);Q, pyridine ,

l.H2-Pd ,2. lactamization

• 35-0',0',0',0'-tetraacetyl-18,19-dihydro-75

Fig. (12). Chemical correlation of strictosidine and vincoside derivatives.

Detailed NMR studies confirmed the appropriate configurations at C-1, moreover the SI 2d and Rll conformations around C-14 in the 4-(4"bromobenzyl) strictosidine and -vincoside derivatives, respectively. The fixed conformation around C-14 was found to be S31 in the strictosamide series, and in R12 in the vincosamide series. The conformation of the dihydropyran ring was negative in all derivatives except strictosidine itself, that of the tetrahydropyridine ring negative in the benzylstrictosidine and the vincosamide derivatives, positive in the strictosamide and the benzyl vincoside derivatives. Finally, the investigations supported the theoretical suggestion that the benzyl group of N-4 and the secologanin unit of C-3 are in trans diaxial orientation [44]. 2. Next, the coupling reaction of tetraacetylsecologanin (7) with 2-oxo2,3-dihydrotryptamine (76a) (in the following: oxotryptamine) and its N^benzyl derivative (76b) was studied according to Fig. (11). Previously, Brown and co-workers investigated the reaction with secologanin

118 aglucone [45]. Our results were surprising from stereochemical point of view. With oxotryptamine (76a), the reaction could not be stopped at the ester level, and gave exclusively the 3R stereoisomer lactam (3i?-79) in an epimeric mixture of 75:7i?=75:25. With Nb-benzyloxotryptamine (76b) it was stopped, of course, at the ester level and gave the same epimeric ratio at C-7 but, again exclusively, the opposite configuration at C-3 (3S-7Sh). These observation helped to interprete the steroselectivity of the coupling reaction with tryptamine as well [46]. 3. It was found by Battersby and co-workers that in the coupling reaction of secologanin (1) and dopamine (80a) (Fig. (13), 2deacetylisoipecosid (15'-81) and -ipecoside (17?-81) were formed, but the opposite configurations at C-1 were assigned to them [29]. The reaction was studied by Zenk and its group as well, and the configurations were corrected [47]. However, the formation of the appropriate neo compounds (82) and (84), in which the primary cyclization would happen at C-2 rather than at C-6 of the benzene ring of the biogenic amine, was not mentioned, although natural representatives of them had already been isolated from plants [48]. We investigated the reaction of tetraacetylsecologanin (7) with dopamine (80a) and its //-benzyl derivative (80b). In the coupling reaction with dopamine, five compounds of the possible eight (normal/neo, ester/lactam, IR/IS) could be isolated (IS-Sla, 15-82a, 15'-83, li?-83 and li?-84). It was clear that the Nunsubstituted esters were already partially lactamized in the coupling reaction mixture, and again with higher rate in the 17? epimers than in the IS ones. When A^-benzyldopamine was used in the coupling reaction, it was stopped, of course, at the ester level, and three compounds (liS'-81b, 17?-81b and li?-82b) of the possible four were isolated. In the case of the dopamine derivatives too, the formation of the \R epimers was in preference over the 15" ones but the stereoselectivity was lower and could be influenced weakly. The normal compounds formed in higher amount than the neo isomers. Moreover, in the normal derivative (li?-81b), slow epimerization was observed, and the equilibrium, which could be approached also from the opposite epimer (liS'-81b), set in after several days at a ratio of IS:\R= 3:7.

119

.,|0glc(R2),

O 1S-81a:SXXPY 12% 81R'=H,R2=H 1S-81b:SXXNN 16% 1/?-81b: R11NP 70% 81a R'=H, R2=acetyl 81b R' =benzyl, R2=acctyl

81c R'=benzyl, R2=acetyl

H^COH —^, • ifR'=H

80a R'=H 80b R'=benzyl

"•• 1S-83: S31NP 8% 1R-83: R12NN 40% 83 R2=acetyl

HO

x C ^ ^ ^ \

glc(R2)4

O tn1S-82a:SXXPY 12% 82 R'-H, R2-HI .^ ^ ^ g2b: R11NP 14% 82a R"=H, R2=acetyl 82b R'=benzyl, R2=acetyl

glc(R^)4

^^^^2^0

I H' OH "^^.^^^ -T ^ ^

0\

1or7

o^,

.MQ in lR-84 R12NN28% ^glc(R2), 84 R2=acetyl

H glc(R^)4

85a R'=H H3CO

85b R'=benzyl

Q

1S-86a:S11PP 3 8 % ( 3 0 % ) *

1S-86b:? 10%(?) 86a R'=H, R2=acetyl) ^/^-86b: R11NP 90%

,41-

86b R'=benzyl, R2=acctyl

total ratio of 15: 1/? in the R'=H, R ' = benzyl in equilibrium

dopamine series: 32:68 16:84 30 : 70

total ratio of 15 : 17? in the histamine series R'=H, 38:62 R'= benzyl 10(?): 90

'N

Q

1S-88:S31NP (8%)^glc(R^)4 1R.88: R12NN 62% {33%y 88 R2=acetyl

> ^ ^ 3

-H.COH,

' I ^^ ^ - ^ ^ \ / F ' > < ^ « ' ^ 'Y " [ 1...''"

ifR'=H

/ M ^ ^ ^N

' ' ^^N^^N^.8^0

glc(R^)4 H3CO 1R-89: R12NN {29%)*

H

-" 87a R'=H, R2=acetyl 87b R'=benzyl, R2=acetyi

89 R2=acetyl

i

o^

glc(R2)4

% figure indicate the composition of the reaction products. *data in italics and parentheses concern reactions carried out in acid-free solvent.

Fig. (13). Coupling reaction of secologanin with dopamine and histamine.

As the half time of the epimerization was about fifty times longer than that of the coupling reaction, the epimeric ratio formed under kinetic control was only slightly influenced by it. Therefore, the stereoselectivity at C-1 of the coupling reaction in the non-benzylated series was

120 characterized by the ratio of the sum of the IS epimeric ester and lactam (lS-81a + 15-82a + 15-83) and that of the IR ones (li?-83 + IR-S4) [49]. 4. From natural sources no coupled products of secologanin with histamine (85a) or A/^-benzylhistamine (in the following: benzylhistamine) (85b) were known. However, it was logical to extend our investigations to this important biogenic amine according to Fig. (13). It was expected that the coupling reaction showed the properties of both the tryptamine and dopamine derivatives. The coupling reaction of tetraacetylsecologanin with benzylhistamine proved to be rather selective: from the reaction mixture tetraacetylbenzylhisteloside as an analogon of tetraacetylbenzylvincoside could be isolated in 90% yield (the other product in 10% was probably its IS epimer, but its amount and purity was not sufficient to prove its structure). If histamin was used in the coupling reaction, one ester (tetraacetylisohisteloside (liS'-86a)) and three lactams (tetraacetylhistelosamide (li?-88), -isohistelosamide (15'-88) and neohistelosamide (li?-89)) could be isolated. The last compound proved to be structural (neo) isomer of the normal lactam which was analogous to tetraacetylvincosamide. (li?-89) had one of the cyclic nitrogens of the histamine unit in bridgehead position and was formed evidently through (li?-87a), which, however, could not be isolated from the reaction mixture. The neo compound appeared only, if the coupling reaction was carried out in totally acid-free solvent, and its amount was increased at the cost of the normal compound (li?-88). Evidently, the acid catalyzed the isomerization of the neo into the normal izomer [paper submitted for publication]. Interpretation of the stereochemistry of the coupling reactions The interpretation should find the explanation of the following facts: 1. The coupling reaction with biogenic amines is selective, if it is at all, in favour of the 3R or IR epimer (R series), and the selectivity could be increased (especially in the vincoside series) by increasing the bulkiness of the ligand attached to N-4 (or N-2). 2. The A^-benzyl oxotryptamine derivative gave the ester (35-785) as main product, the oxotryptamine derivative the (3i?-79) lactam. 3. The lactamization of the ester alkaloids is much faster in the R series, than in the S series. 4. The neo compounds are formed in smaller amount than the normal products.

121 The following analysis was carried out in the tryptamine series, but the results could be generalized to the other series. In the tryptamine series the stereogenic elements mentioned above (R/S configuration at the new center of chirality, 9 staggered conformation around the methylene bridge, N/P conformation of the dihydropyran ring, N/P conformation of the tetrahydropydirine ring) define 72 stereoisomers. In the oxotryptamine series, the flexible conformations of the pyrrolidine ring were represented by a single envelope structure having the C-14 atom at the flap of the ring, and instead of the helicity N or P of it, a second center of chirality at C-7 (R or S) was considered. In the dopamine and histamine series the number of possible isomers was doubled by formation of normal and neo regioisomers. In the benzyl derivatives, the N-ligand may have an equatorial or axial orientation, with further increase of the number of stereoisomers. Each of the conformers mentioned so far may exist in nine staggered conformers by rotation around the glucosidic oxygen. However, at this last stereogenic element only a single conformer was considered according to our NMR measurements and argumentation detailed later. For interpretation of the results described in the previous section, the most probable stereostructure should be selected. Possible structures were generated by the ALCHEMY [50] computer program, interatomic through-space distances between selected structural elements (atoms) of the generated structures were measured, and expressed in the percentage of the sum of their van der Waals radii (relative van der Waals distances). It was supposed that the higher the number of steric interferences and the shorter the through-space distances between given structural elements, the less probable is its formation. In the first phase of our procedure, the C-14 conformers in the tryptamine series were investigated. The appropriate Newman formulae are shown in Fig. (14). In these formulae, bond C-3-C-14 is orthogonal to the plane of the paper, C-14-C-15 is shown sidewise, and the ligand of C15 pointing to us is placed in a smaller circle. As shown in the formulae, in five of the C-14 conformers of the S and R series the smallest ligands H-3 (in 31, 32, 33) and/or H-15 (in 13, 23, 33) are in synclinal orientation to both H atoms of C-14. Consequently, large (non-H) ligands of C-3 (represented by C-2 or N-4) and C-15 (represented by C-16 and C-20) are in steric interferences. Moreover, in two C-14 conformers (in 21 and 23), the N-1 atom strongly approaches Ugand(s) of C-16 or C-20.

122 S series

TTUR

C^o

H'^!

C'\/—x/C^^

TO"

»^TO" H^

»'1
14R

H'^V-^^N^H'"'^

-C2

/ . R33

R32

^ R31

Fig. (14). Conformations around C-14 in strictosidine and vincoside derivatives.

Therefore, all these stereoisomers are less probable (although not improbable or even not impossible, see later) in the equilibrium mixtures.

123 Table la. Steric Interferences in Selected Stereoisomers of Strictosidine and its Benzyl Derivative data for A^-H compounds 1

isomer

B

a

1^

b

c

1,7 1,9 1,6 1,9

0,670 0,952 1,991 1,847

75 87 76

E

2,089 1.988

H15-HBR 022-HBS H6pa-HBR H14R-HBS H15-HBS

1,7 1,9 1,6 1,6 1,7

0,895 0,590 1,999 1,642 2,105

37 23 85 68 88

1,6 1,6 1,6 1,6 1,6

1,780 1,906 2,028 2,089 2.489

66 80 72 80 96

H20-HBR H20-HBS H6aa-HBS

1,8 1,8 1,6

1,093 0,682 1,992

46 28 83

N4-H20 H14S-022 09-HllS*

1,6 1,6 1,6

1,783 2,088 2.001

66 80 77

N1-H15 H3-H20 H5Pa-H14S H14S-022 09-H17*

1,6 1,6 1,6 1,6 1,8

2,264 1,885 1,967 2,088 1.391

84 79 82 80 54

H20-HBR H20-HBS H6|3a-HBR H14R-HBS C19-HBS H14R-HBR H6aa-HBS

1,8 1,8 1,6 1,6 1,8 1,6 1,6

0,519 1,317 1,999 1,642 1,298 1,716 2,048

22 55 83 68 46 72 85

c

\SJJPP

N4-022 H14S-H21 09-HllS*

|l,7 1,6 1,6

12,165

N4-H20 H5[3a-H14R H5pa-C19 H14S-022 09-HllS*

1 S22NN

e

H15-HBR 022-HBS A H6aa-HBS 022-HBR

1

N4-022 H5Pa-H14R H14S-H21 09-HllS*

|S12NP

1 ^I

additional data for A^-benzyl compounds

75 79 87 96

a

ISUPN

|S12NN

1

1,7 1 6 1,6 1,6

2,176 1,906 2,089 2,483

E

A

E A

E A

E A

28 37 83 71

71 E H14R-HBR 1,925 1,894 79 1 1,6 1,6 79 A H6Pa-HBR 1,888 1,999 83 1,6 1,6 80 2,088 H14S-HBS 1,672 70 1,6 1,6 0.934 1,7 36 1 a) interferring atoms; b) their relative position along bonds; c) their distance through space in A; d) their distance through space in % of the sum of Van der Waals atomradii ("relative Van der Waals distance"). Van der Waals radii of the atoms in A: H: 1,2; C: 1,6 (estimated); N: 1,5; 0: 1,4. e) Values shorter than 60% are given in bold; e) equatorial (E) or axial (A) position of the eventual benzyl group. * Additional interferrence in neo isomers of dopamine derivatives. NB. In dopamine and histamine derivatives the analogous numbering should be considered. S22NP

N1-H15 H3-H20 H14S-022 109-H12*

124

Table lb. Steric Interferences in Selected Stereoisomers of Vincoside and its Benzyl Derivative* data for N-H compounds

additional data for //-benzyl compounds

isomer \RJJNN

a H3-H20 H14S-022 09-HllR*

B 1,6 1,6 1,6

c 1,887 2,089 1.930

d 79 80 74

e a E H15-HBS H20-HBR A H6aa-HBS H14S-HBR H15-HBR

IRUNP

H3-H20 H5aa-H14S H14S-022 09-HllR*

1,6 1,6 1,6 1,6

1,885 2,090 2,089 2.425

79 87 80 93

|R12PN

N4-022 H14S-H21 09-HllR*

1,7 1,6 1,6

1,886 2,089 1.912

65 87 74

1,6 1,6 1,6 1,6 1,6 1,8 1,6 1,7 1,6 1,8 1,6 1,6 1,7

1,889 2,081 1,985 2.421 1,885 2,042 2,089 0.918 2,183 1,692 2,030 2,089 0.933

65 87 83 93 70 70 87 35 81 58 85 87 33

N4-022 H5aa-H14S H14S-H21 09-H11R* N1-H15 |R22PN N1-022 H14S-H21 09-H12* N1-H15 |R22PP N1-022 H5aa-H14R H14S-H21 09-019* *Foot see at Table (la). |R12PP

b 1,7 1,8 1,6 1,6 1,7

c 0,953 1,574 2,055 1,809 2,144

d 40 66 86 75 89

E

H15-HBS H20-HBR A H6Pa-HBR

1,7 1,8 1,6

0,773 1,609 1,999

32 67 83

E

C22-HBR C22-HBS A H6aa-HBS H14S-HBR C16-HBR

1,8 1,8 1,6 1,6 1,7

1,332 0,723 2,048 1,838 2,114

48 26 85 77 76

E

C22-HBR C22-HBS A H6pa-HBR 022-HBS E H14S-HBS A H6aa-HBS H14R-HBR

1,8 1,8 1,6 1,9 1,6 1,6 1,6

0,644 1.305 1,986 1,750 1,896 2,048 1,773

23 47 83 67 79 85 74

E H14S-HBS A H6pa-HBR

1,6 1,6

1,778 1,953

74 83

In the second phase, the final three C-14 conformers (11, 12, 22) in both series were analyzed concerning the helicity of the dihydropyran and tetrahydropyridine rings (N or P). In ring conformers, in which the dihedral angle of bond C-14-C-15 has a negative value (—52°) with bond C-19-C-20 (positive conformation of the dihydropyran ring in S12, S22

125 and R l l ) or slightly positive (~+38^) value with bond C-16-C-22 (negative conformation of the dihydropyran ring in S l l , R12 es R22), steric interferences exist between H-3 or N-4 and the methoxycarbonyl or the vinyl group, respectively. The formation of these stereoisomers is likewise less probable. In all structures mentioned till now, the smallest relative van der Waals distances are equal to or in most cases definitely less than 67% of the sum of the appropriate van der Waals radii. All relative van der Waals distances below 100% of the rest conformers (PN and PP ring conformers of S l l , R12, and R22, as well as NN and NP ring conformers of S12, S22, and R l l ) were Hsted in Table (la) and Table (lb). In the dopamine and histamine series, very similar interferences were measured (In these series their own numbering systems should be considered. For the neo isomers of the dopamine series see a special remark later). On the basis of these considerations, the stereoselectivity of the coupling reactions and lactamizations can be explained as follows: 1. The coupling reaction of tetraacetylsecologanin (7) with benzyl oxotryptamine (76b) can most easily be interpreted as it is stopped at the ester level (Fig. (11)). The Mannich type reaction is reversible (through 77b), and gives exclusively the 3S isomer of the tetracyclic spirooxindole ester (78b) in the form of an epimeric pair at C-7. Conformational analysis of the computer-generated models showed that in order to avoid serious interaction with the oxindole subunit, the benzyl group should have pseudoequatorial position, which is possible only in the conformers (R/S)22. However, in conformers (R)22, 0-22 has strong interference with 0-2 (see e. g. bold arrow in structure (3i?-78be) of Fig. (15)) or with with H-9. Therefore the only favoured conformers belong to the S series, i. e. S22NS (see e. g. structure (3^-78be) of Fig. (15)) and S22NR. The NMR experiment really proved the existence of them. 2. The coupling reaction of (7) with oxotryptamine (76a) (Fig. (11)) is likewise reversible (through (77a)). Unfortunately, the tetracyclic spiroester (78a) could not be investigated, because it was spontaneously cyclized, again exclusively, but into 3R isomer of the lactam (79) in the form of an epimeric pair at C-7. In (78a), in absence of the benzyl group, steric interferences disfavour (R)22 conformers, only. The other stereoisomers of (78a) can be (and probably are) present in the coupling reaction mixture. However, the stereochemical structure of the end product (79) is determined by the lactamization step, which is much faster in the R than in the S series (more details later). Therefore, parallel to

126 lactamization, the primary equilibrium could continuously be shifted in the direction (35-78a)^(77a)-^(3i?-78a)^(3i?-79) giving the more favourable product of the R series. The NMR experiment really proved the formation of the lactam epimers R12NR and R12NS of (3i?-79).

R=acetyl, a and e indicate axial and equatorial orientation of the substituent of

Fig. (15). Selected three-dimensional formulae of spiro compounds.

3. In the coupling reaction of (7) and 4-bromobenzyltryptamine (70b) (Fig. (11)), the reaction was stopped again at the ester level, but unlike to the spirooxindole series, the 3R isomer of (74b) was the main product. As

127 it is well known that the p position of the indole ring is more reactive in electrophilic reactions than the a, it was supposed that at first the intermediate spiroindolenine 72b was formed (through (71b)) in a reversible reaction. Unfortunately, it could not be isolated. However, from steric point of view, (72b) may be considered as a close analog of the spirooxindole (78b), because the only difference between them is an H atom in the former (72b) vs. an O in the latter (78b) at position 2. Therefore, the most favoured conformers should be analogous to them i. e. the C-7 epimeric pair S22NR and S22NS of (35'-72b). The subsequent 1,2-rearrangement of the 5/?/ra-indolopirrolidine system of (3iS'-72b) into the more stable fused P-carboline system (73b) needs the n-s* conjugative assistance of the non-bonding electron pair of N-4, which should be in pseudoequatorial, i. e. antiperiplanar oriention to the migrating C-3~C-7 bond (arrow in 72b). Unfortunately, in the favoured (S)22 conformers this position is occupied by the 4"-bromobenzyl group, and its the pseudoaxial position would result in serious steric interferences (see structure 35-72ba of Figure (15)). The conformational analysis showed that the only relatively favoured conformers having the 4"-bromobenzyl group in axial orientation are R12PS (see structure 3i?-72ba) of Figure (15)) and RUNS in which the smallest relative van der Waals distance between two H atoms is already 68 % (in the other isomers this value is 50 % or even less). It may be supposed that the rearrangement can run relatively smoothly from one of these axial conformers of the R series with retention of the configuration of C-3. The unstable 3i?-73b intermediate gives, by loss of the proton of C-2, the final product 3i?-74b. Of course, the rearrangement and stabilization involved parallel or subsequent changes of the other stereogenic elements. According to the conformational analysis, the most favourable steric pattern in the benzylated p-carboline series is RllNP having least number of least unfavourable interferences (see Table (lb)). The NMR spectra really proved this stereochemistry of the product (3i?-74b). As the coupling reaction in the benzylated series is highly selective for the vincoside derivative (3iZ-74b), epimer 0',0',0',0'-tetraacetyl-4-(4"bromobenzyl)strictosidine (35'-74b) could not prepared by direct coupling. It was prepared from strictosidine {3S-74si) by subsequent benzylation and acetylation. As mentioned previously, NMR data proved its stereostructure according to S12dNN as well as a axial orientation of the 4"-bromobenzyl group. Unfortunately, according to the data of Table

128 (la), all the preferred axially benzylated conformers have 5 interferences. However, the indicated stereostructure seems to be the best, because it has a single interference involving the benzyl group and no interference between N-1 and H-15. A somewhat unfavoured internal strain between N-4 and H-20 was eliminated, according to the NMR data, by partial clockwise turning around the bond C-14-C-15 ("distorted" S12dNN). 4. In the coupling reaction of secologanin (1) with tryptamine (70a) (Fig. (11)), strictosidine (3iS'-74a) and vincoside (3i?-74a) were formed in approximately 1:1 ratio, but lactamization of vincoside (3i?-74a) to vincosamide (3i?-75) already started in the coupling reaction mixture, whereas that of strictosidine (35'-74a) to strictosamide (35-75) was much slower (see explanation later). It may be supposed in this case, too, that the first intermediate of the coupling reaction is a spiroindolenine (72a). As in absence of the 4"-bromobenzyl group, no selected conformers of 72a (except R22PR and R22PS) have serious steric interferences, the proton-catalyzed 1,2-rearrangement into the p-carboline structure 73a can easily run from a conformer of (3i?-72a) or (35-72a) without preference of the S or R series. Subsequent irreversible deprotonation gave strictosidine (35'-74a) and vincoside (3i?-74a), which could take up their best conformation S l l P P and R l l N N according to Table (la) and (lb). In fact, it was proved by our NMR measurements that strictosidine really had the expected stereostructure. Because of the easy lactamization, vincoside (3i?-74a) could not be isolated in sufficiently pure state to confirm this prognose. It should be noted that, unlike to the dopamine derivatives, in the P-carboline system no subsequent epimerization of C-3 was observed under the circumstances of the coupling reaction. 5. The coupHng reaction of (7) with dopamine 80a and its benzyl derivative 80b (Fig. (13)) is more complicated, because epimerization was observed at C-1, and both normal and neo regioisomers were formed. During epimerization the normal:neo ratio did not change, which fact suggested the cleavage of bond C-l-N-2 rather than that of C-1-C-10. No epimerization was observed in the 0,0-dimethyl derivatives or the lactams. Therefore it was concluded that it runned (at least formally) through the structure 81c by deprotonation of the 0-7 and protonation of N'2 (in the case of the non-benzylated and neo compounds likewise a slow epimerization was observed). The stereoselectivity at C-1 was low in the non-benzylated compounds, and slightly increased in the benzylated derivatives. It means that the

129 energy difference is small among the stereoisomers, and the interpretation is difficult. It may be supposed that the isoquinoline ring is directly formed in a Pictet-Spengler reaction, without the possibility of stereoselection at a spiro intermediate level. In the selected conformers of the R and S series, the steric interferences are very similar to those found in the analogous tryptamine derivatives, and the preferred conformers are likewise the same (SllPP and RllNN in (81a) as well as S12NN and R U N ? in (81b)). Although the number of interactions in the nonbenzylated series are equal, 0-22 interacts in the ?>S epimer with a N atom, in the 3i? epimer with an other H atom. In the benzylated series an extra interference was measured between H-3aa of the tetrahydropyridine ring and C-19 of the vinyl group, which is absent in the \R epimer. Moreover, the H-3-H-20 interaction in R U N ? has a more tight counterpart N-4-H-20 interaction in S12NN. These small differences may slightly favour the \R over the \S epimer. In the neo isomers (neoisoipecoside and neoipecoside derivatives) each selected conformer has one strong interference involving 0-9, which is absent in the normal isomer, and can explain the less favoured formation of the neo derivatives. The faster lactamization in the R series will be explained later. 6. In the coupling reaction of (7) with histamine (85a) and its benzyl derivative (85b) (Fig. (13)), the ratio of the normal and neo products depends on electronic and not on steric factors. In this respect, the nucleophilicity of C-2 vs. N-3 is deciding. The higher stereoselectivity in the formation of the benzyl compounds could be interpreted analogously to that of the tryptamine derivatives. 7. It was generally observed that the lactamization was faster in the R than in the S series. Two factors seem to be responsible for this phenomenone. The interpretation will be shown in the tryptamine series. The first factor is the structure of the transition state from the most probable ester educt (e. g. strictosidine and vincoside) to the tetrahedric intermedier of the lactamization (Fig. (16)). Concerning the educts, the most probable C-14 conformer of strictosidine was proved by NMR measurements to be S l l , that of vincoside was prognosed according to Table (lb) to be R l l . The conformation of the tetrahedric intermediates as products should be close to that of the appropriate lactams, i. e. S31 and R12. In the R series, both R l l and R12 are favoured conformers, and the transition from R l l to R12 can pass a single favoured eclipsed

130 conformer (R12t), in which the more bulky, non-H ligands are in synperiplanar orientation only with H atoms, consequently easy lactamization was expected. In the S series, only SI 1 is a favoured conformer, but S31 not. The transition from SI 1 to S31 had to pass such an unfavoured, eclipsed conformer (S3It), in which two non-hydrogen ligands (N-2 or N-4 and C-12) are in syn periplanar orientation (indicated by arrow), consequently, lactamization needs a higher activation energy, and runs more slowly. R series H'5

(-20

5

.<^^" H». Cl6

2 A.

^'\ca.

^

N

Hl4S R12t transitio n state

4

M 14-15, P 14-15

^

2

-

^ Hl4S

R12 immediate precursor of lactamization

S series

S31t transition state

immediate precursor of lactamization

Fig. (16). Transition states toward lactamization in the R and S series.

The second factor is the structure of the tetrahedric intermedier, in which the leaving methoxy group should have axial orientation (structures (35'-75t) and (3i?-75t) in Fig. (17)), since in this case the double stereoelectronic effect coming from the non-bonding electronpair of both the N atom and the geminal hydroxy group (see curved arrows at the reaction center) could facilitate the elimination. However, this intermediate is rather crowded in the S31 conformer because, like in the final lactam, the benzene ring is in axial orientation. R12 is flat in both cases. The same factors operate in the lactamization of the other series, too. 8. As mentioned previously, in the stereochemical analysis only one of the nine staggered conformers around the glucosidic O atom was taken into consideration.

131

35-75

H

3/?-75 R=acetyl, t=tetrahedric intermediate

Fig. (17). Three-dimensional formulae of lactam derivatives and their tetrahedric intermediates.

This assumption was justified by the observation that in ^H-NMR spectrum of acetylated strictosamides, e. g. (S^-O'jO'jO'^O'-tetraacetyl18,19-dihydro-75) the signal of one of the four acetyl methyl protons appeared at -1.2 ppm whereas the usual value is -2,0 ppm. In the analogous vincosamides the signals of all these protons appeared at -2,02,2 ppm. The "anomalous" shift was observed in the analogous dopamine and histamine derivatives as well (although at 1.5 ppm and 1.7 ppm, respectively). Detailed NMR measurements proved that the "anomalous" acetyl group is attached to 0-2' atom of the P-D-glucopyranosyl unit. This anomalous shift is caused by the anisotropic effect of the more or less aromatic ring of the biogenic amine subunit, which is close to this acetyl group. (Aimi et al came to the same conclusion on different experimental base [51]). Molecular modelling and carefuU conformational analysis showed that the proximity of the structural elements is possible only in G l conformer of S series(e. g. in structure S^S-TSa of Fig.(17)) where two n-s* conjugative interactions stabilize this conformer (curved arrows). In

132 the R series the same stabihzation might be expected but can not be demonstrated as the aromatic ring is far to any of the acetyl groups (e. g. 37Z-75b of Fig. (17)). As this double anomeric effect is not disturbed by other structural elements, the same conformational arrangement around the glycosidic bridge is supposed to exist in the coupling products of the other biogenic amines, too. 9. The observations and their interpretation contributed to estimate the possible role of the strictosidine synthase enzyme in the stereoselectivity of the coupling reaction between secologanin (1) and tryptamine (70a). In a comparative study, the reaction was carried out in the presence and absence of the enzyme. The unpublished data of Table (2) demonstrated not only the rapidity but the complete stereoselectivity of the reaction in the presence of the enzyme. It seems to us that it plays the same role as the benzyl group and prefers the formation of the spiroindolenine intermediate S22NR-S22NS with high stereoselectivity. However, while the benzyl group is permanently attached to N-4 atom, the enzyme is attached only temporarily to it, and, after having catalyzed the first step of the coupling reaction, leaves it. In this state, the 1,2-rearrangement can easily run from the predominant indolenine intermediate of the S series with retention of configuration at C-3 Table 2. Comparative Study of the Formation of Strictosidine 1

reaction conditions

in the absence

in the presence

of strictosidine synthase substrate concentration, mol/L

2x10-2

6x10-^

temperature, °C

100

37

t,/2, min

-30

<1

yield of isolated strictosidine %

35

98

stereoselectivity

<5

>99

In summary, on the base of our NMR studies and conformational analysis, the stereochemistry of the biomimetic key step of the biosynthesis of these alkaloids based on secologanin has been demonstrated. Moreover, it was even possible to argue for the stereostructures of the isolated compounds, intramolecular motions and changes of conformations.

133 Deglucosylation of tryptamine derivatives of secologanin Enzymatic deglucosylation was well known from the previous work of R. T. Brown and co-workers [31]. We studied the acidic deglucosylation in vincoside derivatives, which could more easily be obtained in the appropriate coupling reactions. The educts and products are shown in Fig. (18). The educts (3i?-90a)-(3i?-93a) and (37?-90b)-(3i?-93b) were selected for the preparation of vincoside derivatives unprotected or protected by benzyl or methyl group on neither, either or both (N-1, N-4) nitrogen atoms in the natural as well as in the dihydro series. Because 1-methyl vincoside and its dihydro derivative were spontaneously lactamized, for studies of the subsequent cyclization their aglucones (3i?-93a)=(3i?-4-debenzyl-96a) and (3if-93b)=(3i?-4-debenzyl96b) were prepared in situ by catalytic hydrogenolysis of their benzyl derivatives (3i?-96a) and (3i?-96b). It was supposed that first the hydrolytic removal of the p-Dglucopyranosyloxy unit gave the "alkaloid aglucone", which is represented by formulae (98a)-(98e) in Fig (19). It should be remembered that the "alkaloid aglucone" can exist (at least in principle) in the equilibrium of all structures analogous to the bakankosine aglucone types represented in Fig (7). Then, the "alkaloid aglucone" was further transformed, by cyclizations and subsequent other reactions into the final aglucones (3i?-94a)-(3i?-97a) and (3i?-94b)-(3i?-97b). Two types of cyclizations were expected as shown in Fig. (19). The oxa- and (carba)cyclizations runned inside of the secologanin subunit, the azacyclizations between the secologanin subunit and one of the N atoms of the tryptamine subunit. The possible aglucone types are summarized in Fig. (20). In the graph 25 points (actually circles) represent 81 aglycone types (natural, dihydro and ethyliden isomerized types in the A and B series), and 40 arrows show 131 possible cyclizations. These cyclizations are represented by structure type numbers in italics and giving the direction of the electron flow from the nucleophilic to the electrophilic center. VE shows aglucone types, in which the vinyl group can be isomerized into vinylidene position. H indicates aglucon types, which can be formed (at least in principle) in the dihydro series as well. Small circles containing H, X or a minus sign on the arrows of the graph mean that the appropriate reactions can not run because of the absence of some structural elements necessary for them.

134 D i h y d r o series

^=^

/^=^

-/l^ \. J. ' /

\

iT \

14

^

1

N'

Bll

/ Mc

4I ^ B u

Jl

IK

M^r N

H

H J.OH [2 ,

+ H2O

"

M/ + H2O

> ^ 0 H

3/<-94b 21R(a);2lS(c)=4:l

3/{-95b 20S(a):20R(e)=4:l

3^-97b 17Z:17£=1:1

cpimeric ratio on C-20=S:1

N-I->C-21 0-17-+C-2I N-4-+C-22 0-l7-»C-21

0-17->C-21 N-1->C-21

3/?-90b 18,19-diliydro

0 3/?-90a (=3«-74a)

N-4-+C-22 0-17-»C-l9

3«-94a 19R(e):19 S(a)=3:2

3/?-92b 18.l9-dihydro

3/?-91b 18,19-dihydro

0

0

3;?-91a

3«-92a

0-17-».C-19

3/;-93b=3«-4-dcbciizyl-96b

3/?-93a=3«-4-dcbcn2yl-96a

0

N-J-+C-21

3/?-96a 2 1

3/f-95a 19R(e):l9 S(a)=2:l

+ H3COH

natural series

Fig. (18). Acidic deglucosylation of vincoside derivatives.

0-17^C-19 N-4-»C-21

3/?-97a 1 cpinicr

135

R'=HorMe R*=H or Bn

0-21->C-17toJr//> 0-17->C-21 10 5(H)

0-17^C-19to7 C-18->C-17to9

N( 1 or 4)-^C-17 to type 2(H) N(l or4)^C-21 to type ^ W N(I or 4HC-22 to type 6(H) N(l o r 4 H C - 1 9 t o t y p e »

Fig. (19). Possible interactions in vincoside primary aglucone.

The tetracyclic structures formed by oxa(carba)cyclization (types 3(H), 5(H), 7 and 9), are analogous to structures of the bakankosine derivatives {1(H), 2(H), 12 and 10, respectively in Fig. (7) and Fig. (8)). The tetracyclic structures formed by azacyclization (types 2(H), 4(H), 6(H) and 8) have series A and B according the nucleophiles N-1 and N-4, respectively. Combination of the types of cyclizations by two different ways gave the 16 pentacyclic structures in series A and B. Further explanations can be found in the original paper [52]. Evidently, in most cases the transformations are under thermodynamic control. The structure of secologanin subunit of all isolated aglucones corresponded to the energetically more favoured structures in agreement with the results obtained on the bakankosine derivatives. It means that the favoured direction of the oxacychzation is 0-17-^C-21 in the dihydro derivatives (according to type IH of Fig. (7)), and 0-17->C-19 in the natural ones (according to type 12 of Fig. (7)). The result of azacyclization depended on the atom N-1 and N-4. When both N atoms were protected (3i?-92a) and (3i?-92b), the process stopped at the tetracyclic level (3J?-96a) and (3i?-96b). When both of them were unprotected (in (3i?-90a) and (3J?-90b)), the azacyclization preceded the

136

eglucosylation, and afforded the thermodynamically more stable N-4 lactamized products (3/?-94a) and (3i?-94b). alkaloid types derived from the dihydro glycosides, too

alkaloid types derived from the natural glycosides, only

Fig. (20). Structure graph of the possible aglucones of strictosidine and vincoside derivatives.

When only N-1 was free and N-4 protected (in (3i?-91a) and (3i?-91b)), azacyclization took place in both series with participation of C-21 instead of C-22 and gave the products (3i?-95a) and (3i?-95b). Evidently, the tetrahedric intermediates formed by participation of C-22 could not be stabilezed into the planar lactam group at N-1. It was mentioned above that the compound, in which N-1 was protected and N-4 free could not be prepared. Therefore, only the subsequent azacyclization of their hypothetic aglucones could be studied. It was found that in the natural series N-4 of the aglucone (3i?-93a)=(3/?-4-debenzyl-96a) cyclized to the C=0 in position 21, and in a subsequent reaction the 21-hydroxy group

137 was removed by catalytic hydrogenation to (3i?-97a). In the dihydro series, the cyclic hemiacetal of the aglucone (3i?-93b)=(3i?-4-debenzyl96) was cleaved spontaneously into the hydroxy-oxo compound, and the azacyclization took place again with participation of N-4 and C-21; finally, the subsequent removal of the hydroxy group by hydrogenolysis gave the end product (3i?-97b) [52]. In the vincoside series the reactivity order of the functional groups can be summarized as follows. The preferred nucleophilic site is N-4 over N-1 and 0-17 over 0-21 (C-18 has an intermediate position in the natural series). The preferred electrophile is C-19 over C-21 and C-17 in the natural series, and C-21 over C-17 in the dihydro series. C-22 is the preferred electrophile only in the vincoside series when lactamization preceded the deglucosylation. However, it seems from the work of De Silva and co-workers [53] that in the strictosidine series, in which the lactamization is slow, the preferred electrophile is C-17 over the others. Fragmentations in tryptamine derivatives of secologanin Because in the first product of the coupling reactions of secologanin or its tetraacetyl derivative with biogenic amines, the secologanin subunit remains intact (except the formyl group of C-7), fragmentations analogous to secologanin and its simple derivatives were expected, and really observed. The reactions were investigated in 3S epimer of the oxotryptamine and the 3R epimer of the tryptamine derivative. In order to prevent lactamization, the A^-4 atom was benzylated. In aqueous acid the coupled product (3iS'-78c) of A^-benzyl oxotryptamine (76c) and secologanin (1) gave back the educt A^^-benzyl oxotryptamine (76c), and benzaldehyde (49) (Fig. (21)). As it is known that the coupling reaction with 2-oxotryptamine derivatives is reversible, the fragmentation was interpreted as a retro version of the coupling reaction, followed by the fragmentation of secologanin or its aglucone according to Fig. (9). It is not known, on which level the glucosidic unit was removed. In the tryptamine series, acidic deglucosylation of (3i?-74c) (Fig. (22)) gave the pentacyclic compound (3i?-95a) (see in Fig (18), according to the most stable secologanin aglucone type and cyclized between C-21 and N1 without fragmentation.

138

piperidine

35-78C (R.=benzyl, R2=H)

p^ Q^

76cR'=benzyl(Bn)

^

Fig. (21). Fragmentation of the spirooxo derivative of benzylstrictosidine.

The basic deglucosylation of (35-780) (Fig. (21) and (3i?-74c) Fig. (22) gave methyl 3-piperidinopropenoate (55), and, in addition TV^^^b-dibenzyl2-oxotryptamine (105) or A^b^^b'dibenzyl tryptamine (112), respectively. It was supposed that in both series the first changes in the reaction sequence were analogous to that observed in the case of secologanin ethylene acetal (4a) (Fig. (9)), i. e. the primary adduct (99) and (106) underwent a basecatalyzed deglucosylation followed by a re/ra-Diels-Alder reaction, which afforded (55) and formally the hypothetic fragment (100) and (107), respectively. Evidently, the second benzyl group of (105) and (112) was formed from the C7 subunit of these fragments. The necessary requirement for this transformation was the cleavage of bonds C-3-N-4 and C-3-C-7 in (100) or C-3-C-2 in (107), i. e. the reversal of the coupling (Mannich-type) reaction combined with the formation of the new bond N-4-C-21. In the oxotryptamine series the reaction sequence could easily been interpreted by cooperation of the solvent (and catalytic effect of the base) as indicated by arrows in structures from (100) to

139 (105). However, it is not known, at which point was the reduction step at C-21 inserted. In the triptamine series, the final steps from (107a) to (112) were supposed to be analogous, but no experimental support could be obtained for the existence of the indolenine structure (107a), which should have be formed by reversal of the rearrangement of the spiro into the fused system.

Fig. (22). Fragmentation of benzylvincoside.

A final remark was made concerning the step of the reduction of C-21. It was observed that the yield of the final products (105) and (112), respectively, was less than 50%. However, in the tryptamine series, the formation of a side product with a supposed structure (109) was observed. Unfortunately, it could not be obtained in pure state, but the ^H NMR spectrum of the crude product was close to that of the adduct (106) except the signal of C-17, which appeared at the olefinic region. This fact suggested the dehydrogenation of the bond C-16-C-17 at the cost of the reduction of C-21 in the hypothetic intermediate (108). Thin layer chromatography showed that the main product (112) and side-product

140 (109) were formed in approximately 1:1 ratio. Moreover, if the basic deglucosylation of (3i?-74c) was carried out in the presence of sodium tetrahydridoborate, the total amount of the educt could be transformed into (112), however, in the presence of potassium hexacyanoferrate (112) was not formed at all. The formation of (102) analogous to (109) was expected in the oxindol series, too, but it could not be demonstrated [results to be published].

0

COCH3

stemmadenine derivative (I)

H3CO "

O

114 dehydrosecodine

H3CO

\

115 catharanthine (II)

CO2CH3

H3CO "

O

116 tabersonine (III)

"corynanthean-type derivative"

118 "sarpagan-type derivative"

Fig. (23). Bioorganic significance of the fragmentation of secologanin derivatives.

The bioorganic significance of these fragmentations (together with those observed in simple secologanin derivatives) may be evaluated as follows (Fig. (23)): 1. The cleavage of bond C-15-C-16 is the same, which was supposed to be cleaved in biosynthethic pass from the acetyl derivative (113) of the type I indole alkaloid stemmadenin through a secodine derivative (114) toward the type II (ibogan, e. g. catharanthine (115)) and type III (plumeran, e. g. aspidospermidine (116)) alkaloids, and it may be served as a model in the biogenesis of these alkaloids [54]. 2. Thefragmentationsof Fig. (9), Fig. (10), Fig. (21) and Fig. (22) involved C-5 (in simple secologanin derivatives) or C-15 (in tryptamine derivatives of secologanin), respectively, as only unfixnctionalized carbon atoms. However, the reactivity of C-4 has parallel in the formation of the sarpagan (e. g. (118)) (and ajmalan) derivatives from corynanthean

141

derivatives (117) 55, 56]. 3. A new, possibly biogenetic and really biomimetic mode of aromatization has been detected in the cychzation of a linear subunit of terpenoid origine. Two interesting structures Finally, two special compounds are shown in Fig. (24). They represent the several isolated and many possible structures, which may have interesting physical, chemical and probable biological properties. Compound (119) is a trioxadamantane derivative, which was prepared from the methanol adduct (6) of secologanin dimethylacetal (4a), in which all masked carbonyl groups were in acetal form. Aqueous acidic hydrolysis, under termodynamic control, transformed them groups of (6) into totally cyclized acetal structure. The backbone of the molecule has a high symmetry, which is broken by the functional groups [15]. Compound (120) was prepared from 8,10-dihydrosweroside aglucone (32), which could be alkylated in aprotic solvent in the presence of a strong base to (33). However, under the same conditions, but in the absence of the alkylating agent, a sequence of four base-catalyzed reactions (dimerization by a vinylogous Michael addition, intramolecular aldolization, lactonization and cyclization by hetero Michael addition) gave the final product (120) . Detailed analysis of its formation would exceed the frame of this paper [result to be pubHshed].

Fig. (24). Two interesting structures prepared from secologanin.

The bioactive properties of secologanin and their derivatives Secologanin and their congeners as well as their alkaloid derivatives are bioactive compounds. In the literature many iridoids and secoirdoids were mentioned to have bioactive, pharmacological and sometimes even therapeutical properties.

142 Some of them are shown in Table (3). However, it seems that detailed investigations have not been carried out, the reported activities were not strong or characteristic, and neither of the compounds applied in clinical praxis. Table 3. Biological Activity of some iridoids 1 Compound

Biological activity

lAUamdin

antileukemic

Elenolic acid

antiviral

Geniposide

laxative

Harpagoside

analgetic, antiinflammatory

pieuropeine

hypotensive

Plumericine

antimicrobial

Sweroside

hepatoprotevtive

Swertiamarine

bitter taste

Valepotriates

sedative

The situation is quite opposite in the nitrogen containing derivatives (i. e. alkaloids) of secologanin. Many of them have marked, in most cases strong biological activities, which could even be applied for therapeutical purposes. Table (4) shows some of the representatives of the indole and related alkaloids having more or less expressed biological or pharmacological activity. Although, many of these activities were not well documented neither investigated in details, several of the alkaloids proved to be highly active, and used not only in physiological and pharmacological research but also in clinical praxis. A selection of them is shown in Fig. (25). Strychnine (from Strychnos nux vomica L.), one of the first alkaloids isolated in crystallin form and a notorious poison, has a very strong stimulating effect (LD50: 0.96 mg/kg) on the motorial inhibitory neurons of the spinal cord as a consequence of the competitive antagonism at the glycine receptor. It can not be used as a therapeutic agent, but useful in pharmacological experiments. Ctoxiferine-I (from Strychnos toxifera F. Schomb.), a bis-quaternary dimer of a strychnan type subunit, is a very potent non-depolarizing, curare-like, nicotinic antagonist at the neuromuscular junction (LDjoo* 23 |ag/kg). Its hemisynthetic bisnor-bisallyl derivative Alcuronium was developped as a clinically potent skeletal muscle relaxant, which is used in the therapeutical praxis as well.

143 Table 4. Biological Activity of some alkaloids derived from secologanin Compound

Structure Type

[Source

Ajmalicine

l a (corynanthean)

ICotharantus roseus

Pharmacological Activity Hypotensive

Ajmaline

la (corynanthean)

IRauvolfia sp.

Antiarrythmic

Alcuronium

Ip (strychnan, "dimer")

hemisynthetic

Skeletal muscle relaxant

Camptothecine

special

Camptotheca acuminata

Anticancer

Catharathine

11(3 (ibogan)

Catharanthus roseus

Hypoglycemic

Coronaridine

up (ibogan)

Catharanthus roseus

Diuretic, cytotoxic

Ellipticine

special

Ochrosia elliptica

Antitumor

Emetine

isoquinoline skeleton

Cephaelis ipecacuanha

Antiamoebic, emetic, expectorant

Deserpidine

la (corynanthean)

Rauvolfia canescens

Hypotensive

Ibogain

lip (ibogan)

Tabernanthe iboga

Psychotomimetic

Mitraphylline

Ip (strychnan)

Mitrgyna rotundifolia

Weak depressant, hypotensive

Olivacine

special

Aspidosperma

Antileukemic

Quinidine

ruban

Remijia sp.

Cardiac depressant

Quinine

ruban

Cinchona sp.

Antimalarial

Rescinnamine

la (corynanthean)

olivaceum

Rauvolfia sp.

Hypotensive

Rhynchophyllin Ip (strychnan)

Uncaria rynchophyalla

Neuroprotective

e

la (corynanthean)

Rauvolfia vomitoria

Hypotensive

Reserpine

ip (strychnan)

Strychnos nux vomica

CNS depressant

Strychnine

Ip (strychnan, "dimer")

Strychnos toxifera

Skeletal muscle relaxant

Toxiferine I

IIP+IIIP ("dimer")

Catharanthus roseus

Antineoplastic

Vinblastine

Iip+IIip ("dimer")

Catharanthus roseus

Antineoplastic

Vincristine

Ilia (ebuman)

Vinca major

Hypotensive

Vincamine

Iip+IIIp ("dimer")

hemisynthetic

Antineoplastic

Vindesine

IIip (plumeran)

Catharanthus roseus

Substructure of vinblastine

Vindoline

Iip+IIIp ("dimer")

hemisynthetic

Antineoplastic

[a (corynanthean)

Corynanthe johimbe

a-adrenergic blocker

Vinorelbine Yohimbine

1

Yohimbine (from Corynanthe johimbe K. Schum.) is an a2 adrenoreceptor antagonist and applied in the pharmacological research. Its aphrodisiac activity was not proved, although it is used in the treatment of impotentia of men. Reserpine (from Rauwolfia serpentina L. Benth.) was introduced into the therapy in the middle of the 20*^ century as a hypotensive and neuroleptic agent. It depletes the biogenic amines (among others dopamine and serotonine) in the brain. Because of some unfavoured side effects it is already not used in the therapy, but as a lead compound had an important influence on the treatment of hypertension, strongly catalysed the progress of the physiology of neuronal transmission and (among others) initiated the golden age of the psychotherapy. All these factors contributed to the explosional development of the chemistry of the indole and related alkaloids. Vincamine (from Vinca minor L.) is

144 also an alkaloid, which was first applied in the treatment of hypertension. As a lead compound it stimulated the preparation and investigation of many related natural, hemi- or totally synthetic compounds. Its close derivative vinpocetine is even now used as a cerebral vasodilator. Two "dimer" indole alkaloids, vinblastine and vincristine (from Catharanthus roseus G. Don.), are still in the center of interest of the pharmacologists. They have antineoplastic activity and used in the therapy of certain types of leukemia. In the cells they inhibit the formation of microtubulles and by that way hinder the cell division. Among many newer, hemisynthetic derivatives of them are Vindesine and Vinorelbine.

RO2C' strychnine

vincamine, R=Me vinpocetine, R=Et, 16-dehyclroxy, 16,17-didehydro

H3CO

vinblastine, R'=CH3, R2=C00Me, R^=0C0CH3 vincristine, R'=CHO, R2=C00Me, R3=OCOCH3 vindesine, R'=CH3, R2=CONH2, R^=OH vinorelbine, R'=CH3, R2=C00Me, R3=OCOCH3, 20'-dehydroxy-15',20'-didehydro

R2

C-Toxiferine-I, R=Me Alkuronium, R=allyl

yohimbine, 3a-H, 203-H, lea-MeOiC R^=OH, R^=H rescrpine, 3P-H, 20a-H, 1 l-OMe, lep-MeO^C, R2=OMe, R3=3',4',5'-trimethoxybenzoyloxy

Fig. (25). Some selected therapeutically active derivatives of secologanin.

Two further antineoplastic compounds derived from coupling of secologanine and tryptamine but having a quinoline ring are camptothecine. Fig (2), (fi'om Camptotheca acuminata Decsne.) and olivacine {Aspidosperma olivaceum Miill. Arg.). Both of them are currently being investigated as anticancer drugs, but up to now they were not widely be used in the therapy. One of the oldest medicaments used against malaria was the Cinchona bark whose main active agent is quinine (see in Fig. (2)) (from Cinchona ledgeriana Moens). While it is still used as such, it was the most important lead compound in the search for more potent antimalarial agents during World War II. In a huge international

145 cooperation, several thousands of derivatives and other compounds were prepared and tested, some of which proved to be therapeutically useful. Quinidine, a diastereomer of it, is occasionally used as an antiarrhythmic agent. Among the isoquinoline derivatives derived from secologanin and dopamine, emetine (see in Fig. (2)), (from Cephaelis ipecacuanha (Biot.) A. Rich), one of the first alkaloids isolated from plants, was used as an antiamoebic, emetic and expectorant agent, but it had unfavourable side effects. This brief summary shows convincingly that even now, derivatives of secologanin should be considered as potent and potentially rich source of bioactive and terapeutically useful compounds. CONCLUSIONS The experimental work carried out on secologanin by us and several other groups during the last 30 years as well as the large number of alkaloids with high biological activity synthesized by plants from secologanin proved that this terpenoid glucoside is one of the most versatile synthons of nature. The work with alkaloids derived from secologanin largely contributed to the chemistry of indole and isoquinoline compounds and to the development of important physicalchemical methods as mass spectrometry and nuclear magnetic resonance spectroscopy. It catalyzed the elaboration of new isolation techniques and the biosynthetic research. In the last half century, many of these alkaloids were targets of huge and successful synthetic efforts (although the total synthesis of secologanin still fails). The total syntheses of quinine, strychnine, reserpine, emetine, vincamine, and many others were milestones of organic synthesis and elaboration of new, versatile synthetic methods and conceptions. It may be expected, that the investigation of these topics can give surprising and important results in the future too. Some of the results presented in this paper may contribute to the interpretation of the biogenetic chemistry of secologanin as well. Moreover, the chemistry of secologanin clearly demonstrates one of the important principles of Nature: to construct a large variety of more or less complicated molecular architectures by combination of simple building stones. It is a special type of bioorganic combinatorial chemistry.

146 Acknowledgements The results presented in this paper could not have be realized without the experimental and intellectual work of my co-workers and friends Gy. Beke, L. Karolyhazy, A. Kocsis, A. Patthy-Lukats, A. Schwartz and K. Szabo-Pusztay. I express my special gratitude to all of them. Most of the NMR investigations were carried out by B. Podanyi (Chinoin Pharmaceutical and Chemical Works, Ltd, - a member of the Sanofi-Synthelabo group, Budapest). I am indebted very much for his excellent work. The contributions of K. Bqjthe-Horvath, F. Hetenyi, G. Krajsovszky, Z. Pal and M. Varga-Balazs in different phases of this work are also thankfully appreciated. I express my sincere thanks to Prof. A. I. Scott for the generous supply of the plasmid containing the gene of strictosidine synthase and for the practical instructions concerning the work with it. The financial support of this work by the National Scientific Research Foundation (OTKA) and the Research Institute of Medicinal Plants (Director: P. Tetenyi sen.) is gratefully acknowledged.

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9]

Tietze, L.-F. Angew. Chem. 1983, 95, 840-853. Dictionnary of Natural Products. Buckingam, J. Ed.; Chapman and Hall, 1994. Cordell, G. A. Lloydia, 1974, 37, 219-298. Rohmer, M.; Seemann, M.; Horbach, S.; Bringer-Meyer, S.; Sahm, H. J. Amer. Chem. Soc. 1996,118, 2564-2566. Lichtenthaler, H. K.; Schwender, J. Annu. Rev. Plant Physiol. Plant Mol. Biol. 1999,50,47-65. Contin, A.; Heijden, R. van der; Lefeber, A. W. M.; Verpoorte, R. FEES Letters, 199S, 434, 413-416. Atta-Ur-Rahman, A.; Basha, A. Biosynthesis of Indole Alkaloids; Clarendon Press, Oxford, 1983. Dabi-Lengyel E.; Kocsis, A.; Bojthe-Horvath, K.; Szabo, L.; Mathe, I.; Zambo, I.; Varga-Balazs, M. Hung. Teljes HU 28415, Dec. 28, 1983; Chem. Abstr. 1984, 100, 180105q. M. V. Kisaktirek, A. J. M. Leeuwenberg, M. Hesse, in: Alkaloids: Chemical and Biological Perspectives; S. W. Pelletier, Ed., J. Wiley, New York, 1983, pp. 213376.

147

[10] Szabo, L.; Bojthe-Horvath, K.; Hetenyi, F.; Kocsis, A.; Pal, Z.; Varga-Balazs, M.; Tetenyi, P. F. E. C S. Int. Conf. Chem. Biotechnol Biol Act. Nat. Prod., F\ 1981, Short Communications, 1981, 3/1, 87-91. [11] Kinast, G.; Tietze, L-F. Chem. Ber. 1976,109, 3640-3645. [12] Brown, R. T.; Chappie, C. L, Tetrahedron Lett. 1976, 787-790. [13] Tietze, L.-F., Henke, S. Angew. Chem. 1976, 93, 1005. [14] Kawai, H.; Kuroyanagi, M.; Ueno, A. Chem. Pharm. Bull. 1988, 36, 3664-3666. [15] Krajsovszky, G.; Kocsis, A.; Szabo, L. F.; Podanyi, B. Tetrahedron, 1997, 53, 11659-11668. [16] Brown, R. T.; Chappie, C. L.; Duckworth, D. M.; Piatt, R. J. Chem. Soc. Perkin I, 1976, 160-162 [17] Purdy, J. R.; Hamilton, R. G.; Akhter, L.; McLean, S. Can. J. Chem. 1981, 59, 210214. [18] Saunders, G. N.; Purdy, J. R.; McLean, S. Can .J. Chem. 1983, 61, 276-281. [19] Pal, Z.; Varga-Balazs, M.; Szabo, L. F.; Tetenyi, P. In Bio-organic Heterocycles. H. C. van der Plas, L. Otvos, M. Simonyi, Ed.; Akademiai Kiado, Budapest, 1984, pp. 221-224. [20] Kocsis, A.; Pal, Z.; Patthy, A, Mrs.; Szabo, L.; Varga, J. Mrs.; Tetenyi, P. Hung Teljes HU 47304, Febr. 28, 1989; Chem. Abstr. 1990,112, 70033tq. [21] Tietze, L.-F.; Baertels, C; Fennen, J. Liehigs Ann. Chem. 1989, 1241-1246. [22] Tietze, L.-F. Tetrehedron Lett. 1976, 2535-2538. [23] Kocsis, A.; Pal, Z.; Patthy, A. Mrs.; Szab6, L.; Varga, J., Tetenyi, P. Teljes, HU 48,640, June 28, 1989; Chem. Abstr. 1990, 112, 132478c. [24] Schwartz, A.; Szabo, L. F.; Podanyi, B. Tetrahedron, 1997, 53, 10489-10502. [25] Tietze, L. F.; Baertels, C. Tetrahedron, 1989, 45, 681-686. [26] Szabo-Pusztay K.; Szabo L. F.; Podanyi B. Acta Chimica Hungarica. Models in Chemistry, 1994,131, 475-488. [27] R. T. Brown, S. B. Pratt, J. Chem. Soc. Chem. Commun. 1980, 165-167. [28] Karolyhazy, L.; Patthy-Lukats, A.; Szabo, L. F.; Podanyi, B. Tetrahedron Lett. 2000,^7,1575-1578. [29] Battersby, A. R.; Burnett, A. R.; Parsons, P. G. J. Chem. Soc. (C) 1969, 1193-1200. [30] Battersby, A. R., Burnett, A. R., Parsons, P. G. J. Chem. Soc. (C) 1969, 1187-1192. [31] Brown, R. T. In Indole and Biogenetically Related Alkaloids. Phillipson, J. D.; Zenk, M. H. Eds.; Academic Press, 1980, 171-184. [32] R. S. Kapil, R. T. Brown, in: The Alkaloids; R. H. F, Manske, R. Rodrigo, R., Eds.; Academic Press, New York, 1979; Vol. 18, pp. 545-588. [33] Brown, R. T.; Curless, D. Tetrahedron Lett. 1986, 27, 6005-6008. [34] Mattes, K. C; Hutchinson, C. R.; Springer, J, P.; Clardy, J. J. Am. Chem. Soc, 1975,97,6210-6271. [35] Kennard, O., Roberts, P. J., Isaacs, N. W., Allen, F. H., Motherwell, W. D. S., Gibson, K. H., Battersby, A. R. Chem Commun. 1971, 899-900. [36] Stockigt, J; Zenk, M. H. Chem Commun. 1977, 646-648. [37] Rueffer, M.; Nagakura, N.; Zenk, M. H. Tetrahedron Lett. 1978, 1593-1596. [38] Nagakura, N.; Hofle, G.; Coggiola, D.; Zenk, M. H. Planta Med. 34, 381-389. [39] De-Eknamkul, W., Ounaroon, A., Tanahashi, T., Kutchan, T. M., Zenk, M. H. Phytochemistry, 1997, 45, 477-484. [40] Le Men, J.; Taylor, W. L, Experientia, 1965, 21, 508.

148

[41] Kocsis, A.; Pal, Z.; Szabo, L.; Tetenyi, P.; Varga, B. M. Eur. Pat. Appl. EP 156267, Oct. 2, 1985; Chem. Abstr. 1986,104, 149345q. [42] Patthy-Lukats, A.; Karolyhazy, L.; Szabo, L. F.; Podanyi, B. J. Nat. Prod. 1997, 60, 69-75. [43] Lentz, P. J., Jr.; Rossmann, M. G. Chem Commun., 1969, 1269. [44] Lukats-Patthy, A.; Kocsis, A.; Szabo, L. F.; Podanyi, B. / Nat. Prod. 1999, 62, 1492-1499. [45] Brown, R. T.; Chappie, C. L.; Piatt, R. J. Chem. Soc. Chem. Commun. 1976, 14011402 [46] Patthy-Lukats, A.; Beke, Gy.; Szabo, L. F.; Podanyi, B. J. Nat. Prod. 2001, 64, (accepted for pubhcation). [47] Hoefle, G., Nagakura, N., Zenk, M. H. Chem. Ber. 1980,113, 566-576. [48] Itoh, A., Tanahashi, T., Nakagura, N. Chem. Pharm. Bull. 1989, 37, 1137-1139. [49] Beke Gy., Szabo, L. F., Podanyi, P. J. Nat. Prod 2001, 64, 332-340. [50] ALCHEMY II molecular modeling system for the IBM PC. Tripos Associates, Inc., St. Louis, MO. [51] Takayama, H., Ohmori, O.; Subhadhirasakul, S.; Kitajima, M.; Aimi, N. Chem. Pharm. Bull. 1997, 45, 1231-1233. [52] Karolyhazy, L.; Patthy-Lukats, A.; Szabo, L. F.; J. Phys. Org. Chem., 1998, 11, 622-631. [53] De Silva, K. T. D.; Smith, G. N.; Warren, K. E. H. Chem. Commun. 1971, 905-907. [54] Scott, A. I.; Wei, C. C. J. Amer. Chem. Soc. 1972, 94, 8266-8267. [55] Tamelen, E. E. van; Oliver, L. K. J. Amer. Chem. Soc. 1970, 92, 2136-2137. [56] Scott, A. I.; Bioorg. Chem. 1974, 3, 398-429.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

149

THE LIGNANS OF PODOPHYLLUM RITA M. MORAES National Center for Natural Products Research, The Research Institute of Pharmaceutical Sciences, School of Pharmacy, University of Mississippi, University, MS 38677, USA FRANCK E. DAY AN and CAMILO CANEL USDA, ARS, NPURU, National Center for Natural Products Research, University, MS 38677, USA ABSTRACT: Lignans are a widely distributed class of dimeric phenylpropanoid derivatives, many of which have strong antimicrobial, antiviral, or antifeedant activity and thus play important roles in plant defense. Of more restricted taxonomic distribution, the aryltetralin lignans have been found in highest abundance in plants of the genus Podophyllum (Berberidaceae). Foremost among these lignans, podophyllotoxin is a particularly cytotoxic inhibitor of microtubule assembly and a strong antiviral agent. Semisynthetic epimeric derivatives of podophyllotoxin having inhibitory activity against DNA-topoisomerase II have been developed as effective antineoplastic drugs. Current work on Podophyllum lignans is focused on two fronts: 1) Structure optimization to generate derivatives with superior pharmacological profiles and broader therapeutic use, and 2) Development of alternative sources of podophyllotoxin. Numerous variations of the basic aryltetralin structure have been created. Some of the new compounds have shown promising activity profiles, but practically little has been achieved besides improvement in solubility. Interest in new derivatives remains strong, which, along with the formulation of existing drugs for new indications, is increasing the demand for podophyllotoxin. While intense collection has severely reduced the natural stocks of Indian Podophyllum, the primary source of podophyllotoxin, a North American species has emerged as a rich and renewable source of this compound.

HISTORICAL BACKGROUND The genus Podophyllum (Berberidaceae) is most prominently represented by P. peltatum L., Fig. (1), in the United States and P, emodi Wall. (syn. P. hexandrum Royle) in northern India and Nepal. Extracts of dried rhizomes of mayapple and bankakri have a long history of use by the

Fig. (2). Crlstal \tructurc and ahwlute conliguration of I‘-broinopodoph) llt#o\in tlIu\trating the mostly planai tetralin hachhone. the envelop confhrination ofthc lactwic ring. and the freely rotating aryl ring. Whitc - carbon. red oxygen: grcen - hrointne; light blue - hydrogen

Fig. (I). f’odoph) Ilutn pcltatuin in natiiral habitat

151 indigenous populations of North America and the Himalayas, respectively, who valued their cathartic and cholagogue properties. European settlers of North America adopted the use of extracts of P. peltatum, which led to its introduction in Western pharmacopoeias. However, podophyllotoxin, Fig. (2), was shown to possess colchicine-like toxicity in 1891 [1], causing a decline in the therapeutic use of the extracts. The demonstration, in 1942, of the resin's efficacy in the treatment of condylomata acuminata [2], rekindled interest in podophyllotoxin and raised hopes for its use in the treatment of malignant tumors. The resin was later shown to possess potent antimitotic properties [3]. A report describing the results of preliminary screening of hundreds of compounds, performed at the National Cancer Institute, highlighted the capacity of a single dose of podophyllotoxin to grossly damage tumors in mice [4]. Subsequent clinical tests showed that podophyllotoxin caused severe abdominal pain, vomiting, and diarrhea, which made the compound unacceptable as a drug. The belief that glycoside derivatives of podophyllotoxin would exhibit superior pharmacological profiles led researchers to extract and test lignan glycosides from Podophyllum tissues. Being more water-soluble, the glycosides indeed proved less toxic, but showed reduced antitumor activity. Although not immediately successful, this approach eventually led to the serendipitous discovery of glycosidic derivatives of podophyllotoxin that proved to be potent anticancer agents. An insightful account of the development of podophyllotoxin into the antineoplastic drug etoposide has been published [5]. DISTRIBUTION AND BIOLOGY OF PODOPHYLLUM By 1970, in the wake of efforts to capitalize on the newly realized medicinal value of podophyllotoxin for Western medicine, more than 130 tons of Podophyllum rhizomes were needed in the US annually to satisfy the demand for the compound [6]. This plant material was wild-harvested from the understory of oak and hickory forests of the eastern and central US. In 1974, mayapple was considered a common plant with a natural habitat extending from South Carolina to Maine along the Atlantic coast in the East, to eastern Texas, Oklahoma, and Kansas beyond the Mississippi basin in the West, and as far North as Lake Michigan [6]. Today, mayapple is commonly seen along roadsides and can still be abundantly found in other readily accessible areas within its natural range. Several

152 colonies have been found near roads in the town of Oxford, Mississippi, where our laboratories are located. Mayapple is a perennial herb that forms highly branched rhizome systems, with new rhizome segments appearing each year. Consequently, colonies of mayapple often consist of hundreds of closely spaced shoots that are interconnected through extensive rhizome systems covering several hundred square meters. Each shoot supports the growth of one or more new rhizome segments, each of which has a terminal bud that remains dormant over the winter. The shoots senesce at the end of the growing season and detach from the rhizome, leaving a dormant bud at the node. The terminal buds develop into annual aerial shoots the following spring. New shoots may also develop from older dormant nodes, especially if apical dominance is affected by damage to the terminal bud. Even though mayapple is self-incompatible, the genetic diversity within mayapple colonies is highly restricted [7]. This is the result of a combination of factors such as high rates of flower and fruit abortion, low rates of germination, and high seedling mortality. Nevertheless, sexual shoots, which can be easily identified by the presence of two or more leaves, are very common in mayapple colonies. Because they carry at least twice the number of leaves, the photosynthetic area of sexual shoots is approximately twice as large as that of asexual shoots [8]. Studies of the relation between rhizome growth and shoot development indicate that photosynthetic productivity, a function of the size of the photosynthetic area and its longevity, strongly affects the development of new rhizome segments. For instance, late senescing shoots produce more numerous, longer, and heavier new rhizome segments than early senescing ones [9]. Similarly, sexual nodes are twice as likely to branch as vegetative nodes [8] and produce more massive rhizomes than asexual, single-leafed shoots [9]. In contrast, successful fruiting results in less rhizome growth; new rhizome segments produced by fruiting nodes are shorter than those produced by vegetative or non-fruiting sexual nodes [7]. Stored photosynthesis-derived resources appear destined to support fruit development, but, upon flower or fruit abortion, are used instead to produce larger and more numerous new rhizome segments. In view of the high cost of fruit production, the low success rate of sexual reproduction, and the added vigor and increased shoot number that result from flower and fruit abortion, the question arises why P. peltatum has not abandoned propagation by seed altogether. The persistence of sexual development in mayapple must have demographic implications that

153

are still not understood. Interestingly, longer and heavier rhizomes are more likely to give rise to sexual, multi-leafed shoots [9]. A model has been proposed, according to which fruit failure increases the chance of success of sexual reproduction in the future and therefore decreases the probability of the colony's extinction [7]. Given the wide geographic distribution of mayapple, seedling establishment, albeit infrequent, appears to be more than just insurance for survival against possible colony destruction by pathogens and severe weather. The colonization of such an extensive area must have resulted from production and dispersal of viable seeds. Interest as a commercial source of raw material turned to Indian Podophyllum when its rhizomes were found to contain more podophyllotoxin than those of the American species [10]. P. emodi is a perennial rhizomatous herb found in the Himalayas at altitudes of 2,000 to 4,000 m. It grows in the understory of subalpine forests, and near boulders and in open meadows above the tree line [11]. The suggestion that P. peltatum is conditioned to a winter rest period induced by coldness and requires a mean January temperature of no less than -5°C and a growth season of at least 150 frost-free days [6] seems to apply to P. emodi as well, and may explain why the latter is not found in the Indian plains or above 4,000 m in the Himalayas. Commercial interest in P. emodi has driven the species to an endangered status. The size of natural populations of P. emodi has declined sharply in recent years due to intensive collection, lack of cultivation, and loss of habitat to grazing. The present distribution of natural populations is relatively restricted. In Garhwal Himalaya, for instance, several populations first studied in 1982 have virtually disappeared [11]. During a recent visit to the Himalayas, specimens of P. emodi were found in the understory of a protected cedar forest with a distribution of approximately one every 100 m^ [Canel, personal observations]. Based on the number of leaves and the presence of reproductive structures, many of these plants were several years old. In all cases, the plants had a rhizome no more than 10 cm long with a profuse root system and no evidence of branching. In the face of declining natural populations of P. emodi, a valiant effort is being made by Indian research institutions to rescue the species. The study of the population biology and genetic diversity of P. emodi is considered essential for the success of these efforts. For instance, under the direction of Prof. S. N. Raina, researchers of the Department of Botany of the University of Delhi are collecting specimens of Indian Podophyllum

154 from natural populations in the Western Himalayas. These populations are being characterized at the cytogenetic level to assess the existing genetic diversity. Representative specimens are being transplanted to experimental plots in the Himalayan foothills. Similarly, a group at the GB Pant Institute of Himalayan Environment and Development in Garhwal has established a demonstration garden for the cultivation of P. emodi and is engaged in replenishing the disappearing natural stocks with rhizome cuttings, viable seeds, and plants regenerated from embryogenic calli. A fairly high level of genetic diversity seems to exist among and within populations of P. emodi. Isozyme analysis indicated that there is considerable inter- as well as intra-population variation [11]. This offers hope for the success of conservation efforts. The seed biology of Indian Podophyllum is also being studied in order to enhance the availability of this plant. At the Institute of Himalayan Bioresource Technology in Palampur, researchers have subjected seeds to various dormancy-breaking treatments such as chilling, application of hormones, and mechanical scarification. Chilling followed by scarification resulted in a germination rate of 80% within 7 days, whereas embryo rescue produced germination rates nearing 100% [12]. These efforts are especially laudable given the remoteness and difficult accessibility of the natural habitat of P. emodi, and the low level of funding with which some of these institutions operate. STRUCTURE AND BIOSYNTHESIS Lignans are a diverse and ubiquitously distributed group of phytochemicals that are biosynthetically derived from the phenylpropanoid pathway. The structural diversity and taxonomic occurrence of lignans and related compounds have been reviewed [13, 14]. Podophyllum tissues contain lignans of the fran^-aryltetralinlactone type and their 4-0-y9-D-glucopyranosides, among which podophyllotoxin is of most interest. The approved nomenclature for compounds having the aryltetralin carbon skeleton is described in Fig. (3). Podophyllum lignans have 5 rings designated A, B, C, D, and E. Ring A consists of a 1,3dioxolane cycle, rings B and C together form the 1,2,3,4tetrahydronaphthalene substructure, ring D is composed of a y- lactone, and ring E, connected at CI, is an aryl side chain. Carbon 4 carries the hydroxyl group involved in the glycosidic linkage, whereas C2 and C3 are involved in the lactone ring D [15].

155

Rj = H, R2 = OH: podophyllotoxin Rj = OH, R2 = H: epipodophyllotoxin Fig. (3). Nomenclature of aryltetralins

Bonds projecting below and above the plane of the tetracyclic backbone are referred to as a and p, respectively. The a diastereomer at C4 is knovm as podophyllotoxin, whereas the fi diastereomer is called epipodophyllotoxin, Fig. (3). The lactone ring always has the a,/?configuration at the C2,C3 bond, except for the unnatural picropodophyllotoxin obtained as an artifact during extraction [16]. The X-ray analysis of crystalized 2'-bromopodophyllotoxin demonstrated that rings A, B, C, and D form a nearly planar structure with the aryl ring connected to CI being pseudo-axial to ring C, Fig. (2) [17]. The dioxolane ring A is coplanar with ring B, and the lactone ring D exhibits the expected envelope conformation with C3 bending down to form a 'flap' and the remaining atoms of the lactone ring being coplanar. The biosynthesis of lignans has received comprehensive coverage in the recent literature [18]. Lignans are formed by the enzymatic stereospecific linkage of monolignol precursors, most commonly through 8-8' bonds. Although the pathway from monolignol to podophyllotoxin has not been unambiguously demonstrated, the experimental evidence

156 accumulated thus far strongly indicates that podophyllotoxin and congeners found in Podophyllum are synthesized from coniferyl alcohol, Fig. (4). For instance, roots of P. hexandrum incorporated isotopically labeled coniferyl alcohol into podophyllotoxin [19]. Similarly, watersoluble forms of coniferyl alcohol, such as its complex with Pcyclodextrin and the glucoside coniferin, greatly increased podophyllotoxin production by cultured cells of P. emodi [20, 21]. Two biosynthetically distinct groups of Podophyllum lignans have been identified, one having a 3,4,5-trimethoxy substituted aryl group, which includes podophyllotoxin and y^peltatin, and the other having a 4hydroxy-3,5-dimethoxy substituted aryl group, which includes a-peltatin, Fig. (4). The efficient incorporation of (~)-[^'^C]matairesinol into podophyllotoxin, /?-peltatin, 4'-demethylpodophyllotoxin, and a-peltatin in Podophyllum demonstrated that matairesinol is a common intermediate, and probably the branch-point compound, leading to the formation of both groups of Podophyllum lignans [22]. Matairesinol, a 4-hydroxy-3methoxy substituted compound is converted to the 3,4,5-trimethoxy substituted yatein or to 4-demethylyatein, which respectively give rise to podophyllotoxin/)5-peltatin or 4-demethylpodophyllotoxin/a-peltatin, via quinonemethide and desoxy intermediates [23]. Carefully designed labeling experiments have conclusively established the biosynthetic origin of (--)-matairesinol in Forsythia spp. Cell-free enzyme preparations of F. suspensa stems catalyze the stereoselective coupling of two molecules of [8-^'^C]coniferyl alcohol to form (+)pinoresinol in enantiomeric excess [24]. This reaction likely involves the coupling of free radicals of coniferyl alcohol and is mediated by two enzymes: an oxidase and a dirigent protein that controls the stereospecificity of the reaction. Variously labeled coniferyl alcohol was also incorporated into optically pure (-)-secoisolariciresinol and (-)matairesinol by tissues of F. intermedia [25]. It is probable that the series of stereospecific and strictly enantioselective reactions that transform two achiral molecules of coniferyl alcohol into (~)-rotatory matairesinol in Forsythia are also responsible for matairesinol production in Podophyllum, Fig. (4). Little is known about the tissue and subcellular compartmentation of lignan biosynthesis in Podophyllum. In vitro cultured cells of P. emodi are capable of synthesizing podophyllotoxin from coniferyl alcohol, implying

V OH O C H ,

'

H3c0m H3CO

OH

HO

"',,//

I

-

H3C0

Yatein

U

R,=H, R,=OH: Podophyllotoxin R,=OH, R2=H: /%Peltatin

OCH3 (+)-Pinoresinol

HO

O O C H , HO (-)-Secoisolariciresinol

HO Demethyl yatein

HO R, =H, R,=OH 4'-Demethylpodophyllotxin R,=OH, R2=H a-Peltatin

Fig. (4). Proposed lignan biosynthetic pathway in Podophyllum (steps 1-3 have been characterized in Forsythia spp.). 1, Oxidase and dirigent protein create an 8-8' bond between two molecules of coniferyl alcohol; 2, NADPH reductase catalyses two sequential reactions; 3, NADP+ dehydrogenase. Dotted arrows indicate multiple uncharacterized steps.

t; 4

158

that the underground parts of this plant have full biosynthetic capacity [20]. The leaves of P. peltatum, in turn, can incorporate matairesinol into podophyllotoxin and the peltatins [22]. To our knowledge, there are no published reports of lignan content in leaves of P. emodi. Analysis of an accession originating from Nepal showed that only minute amounts of J5peltatin were present [Canel, unpublished data]. Whether this is representative of the species as a whole remains to be established. If so, the possibility that transport of water-soluble monolignol precursors, such as coniferin, from the leaves to the roots merits investigation. Lignan accumulation appears to differ in the American and Indian species. Studies involving differential extraction with aqueous and organic solvents indicate that lignans are stored mostly in the form of 4-0-y^Dglucopyranosides in P. peltatum leaves, the efficient extraction of which requires hydrolytic cleavage of the sugar moiety [26]. In contrast, notwithstanding possible conversion during processing and storage of the tissues, most podophyllotoxin present in the rhizomes of P. emodi exists as the aglycone, and can be extracted directly with ethanol [26]. BIOLOGICAL ACTIVITY The wide taxonomic distribution of lignans and their pronounced toxicity and antimicrobial activity suggest that these phytochemicals play important roles in protecting plants from pathogens and herbivores. For instance, in leaves of P. peltatum, the accumulation of large amounts of podophyllotoxin glucoside, which is rapidly hydrolyzed into the more toxic aglycone upon disruption of the structural integrity of the leaves, points to the possible role of podophyllotoxin as a feeding deterrent. Because of the medicinal importance of podophyllotoxin, the biological activity of Podophyllum lignans has been studied extensively. Most attention has been given to the potential use of these compounds and their derivatives as antineoplastic and antiviral agents. The successful derivatization of podophyllotoxin into potent and safe antineoplastic drugs has generated considerable interest and led to numerous studies by research groups throughout the world. Thousands of compounds have been generated in the last several years. The study of the biological activity and medicinal properties of Podophyllum lignans is therefore tightly linked to the manipulation of the basic aryltetralin backbone and is presented accordingly in this section. In general, efforts to

159 optimize the structure of podophyllotoxin to improve its pharmacological properties and expand its application have focused on one of the following aspects: 1) manipulation of the linkage at the C4 carbon, including derivatization of the sugar moiety, 2) modification of the CI aryl group, and 3) modification of the aryltetralin structure. Anticancer Activity Most of the work on the derivatization of podophyllotoxin has targeted the C4 position. These studies can be classified according to the fate of the carbon-oxygen bond, which may be either retained or replaced by a carbon-nitrogen or a carbon-carbon bond. When retaining the C-0 bond, derivatization has involved either the modification of the sugar moiety or its replacement with groups of different chemical nature. Modification of the Sugar Group

The early work on structural optimization of podophyllotoxin was done in the mid 1950's by chemists at Sandoz, in Switzerland, who sought to eliminate the toxic effects of the compound while retaining its antitumor activity. The first attempts to generate pharmacologically useful derivatives of podophyllotoxin involved the use of the glucoside, since the presence of the sugar moiety made the compound more water-soluble and reduced its toxicity [5]. Modification of the sugar group initially involved alkyl derivatization at the 4" position, which did not produce superior compounds, and later acylation at the same position, which reduced the activity. Further work to stabilize the glucoside was done by blocking positions 4" and 6" with various groups. Derivatives having a 4",6"benzylidene group showed much improved oral bioavailability while retaining the cytostatic properties of the unaltered glucoside. Among the benzylidene derivatives, a compound was fortuitously generated that inhibited cellular proliferation via a mechanism other than spindle poisoning and concomitant mitotic arrest. The unexpected activity was traced to a contaminant found in minute amounts in preparations of derivatives of the total glucoside fraction of Podophyllum rhizomes. The new compound differed from other derivatives by having a p configuration at C4, instead of the a configuration of podophyllotoxin.

160 and by inhibiting DNA-topoisomerase II. Extensive studies ensued on the synthesis of C-4/? analogs, referred to as epipodophyllotoxins, resulting in the discovery of teniposide and etoposide, which were launched commercially in 1976 and 1980, respectively. All podophyllotoxin-based anticancer drugs commercially available today are epipodophyllotoxins that have modified sugar moieties. Teniposide has a thionylidene glucoside, whereas etoposide and etoposide phosphate, a water-soluble phosphate ester prodrug of etoposide launched in 1996 under the trade name Etopophos, have ethylidene glucosides, Fig. (5). Etoposide phosphate was developed to overcome the limitations associated with the considerable hydrophobicity of etoposide. The prodrug can be administered in higher doses than etoposide as a short intravenous injection, after which it is rapidly converted to the parent compound by plasma phosphatases, and thus constitutes an improved formulation of etoposide [27].

Ri

•Q

Tenoposide:

R2

OH

Etoposide:

CH3

OH

NK-6n:

CH,

N(CH3)2

Fig. (5). Various glucosides of podophyllotoxin-derived antineoplastic drugs

Numerous other glucosidic substitutions have been tested [28]. One of the most promising compounds with such modification, NK-611, has a dimethylamino side chain instead of a hydroxyl group at the R2 position of the ethylidene-type sugar moiety. This substitution greatly increased water solubility and improved in vitro antitumor activity. Clinical tests of intravenous and oral formulations of NK-611 suggest that the compound

161 has better bioavailability than etoposide [29, 30]. The synthesis and evaluation of a series of 42 new amino- and alkylamino-glycosides of epipodophyllotoxin has recently been reported. Four of these derivatives showed at least as much activity as etoposide on in vivo leukemia models [31]. This study demonstrated the importance of the 4",6"-acetal substitution in the sugar moiety for biological activity. Surprisingly, none of these compounds inhibited DNA-topoisomerase II or bound to tubulin. Replacement ofC4 Sugar Moiety with Other Groups

Studies using compounds having non-sugar groups attached at C4 indicate that the sugar moiety is not essential for anticancer activity, but its replacement affects the mechanism of action of the lignan. Several C4 esters have been generated [32]. These compounds behaved more like inhibitors of microtubule polymerization than like DNA topoisomerase inhibitors. Most compounds of a series of twenty nine C-4/? derivatives possessing linear alkyl, hydorxyalkyl, methylthioalkyl, acetonide, or dihydroxyalkyl side chains were not very active on DNA-topoisomerase II, except for those having dimethylamino and methylpyrrolidino groups [33]. Replacement of the C4 Carbon-Oxygen Linkage with Amino Functionality

The introduction of a C-N bond at the C4 position yields relatively simpler structures and has resulted in compounds having promising therapeutic profiles, Fig. (6).

-ORi HN

I

5'dl Fig. (6). Podophyllotoxin derivatives possessing novel C4 substitution with a carbon-nitrogen bond replacing the carbon-oxygen bond

162

For example, the new etoposide analog GL-311, possessing a C-Apaminoaniline substitution, is currently in phase II clinical trials for various anticancer treatments. This compound is a more potent inhibitor of proliferation of refractory cancer cells than etoposide [34]. Attempts to increase the biological activity of GL-311 by di- and tri- substitution of the aniline ring have produced derivatives up to 10 times as potent as etoposide [35]. Other such compounds have proved at least as active as etoposide as anticancer agents; these include alkylamino-substituted derivatives [36, 37] and synthetic acetyl salicylic acid, 2mercaptobenzothiazole and 2-aminobenzothiazole derivatives [38]. In all cases, it has been necessary to maintain the P stereochemistry at C4 to obtain acceptably high antitumor and anti-topoisomerase activity. A comparison of the activities of several a and /? analogs, including simple alkylamino derivatives, showed that compounds with a configuration were consistently more potent inhibitors of tubulin polymerization than their counterparts with p configurations [39]. An extension of this approach was evaluated, which combined the C-4yS-aminoaryl substitution with the presence of an o-quinone side chain at CI, Fig. (7). The doubly substituted compounds were more potent than etoposide at either inhibiting DNA-topoisomerase II or inducing DNA breaks [40].

,r-^%. H3CO

Y

"O

o Fig. (7). Podophyllotoxin derivatives with 1-o-quinone and 4-aminoaryl substitutions

163

Carbon-Carbon

Linkages

One of the main objectives of structural derivatization of podophyllotoxin has been the stabilization of the C4 linkage. Therefore, it is surprising that relatively few studies have aimed at substituting the C4 carbon-oxygen bond with carbon-carbon bonds, given the likelihood that these molecules would be less sensitive to degradation. One such study involved the synthesis of 17 analogs having alkene, alkyl, alkyl alcohol, and ether groups at C4. Some of these molecules, including a hydroxypropyl derivative, had activities somewhere between etoposide and tenoposide [41]. TOP-53, another highly active podophyllotoxin derivative possessing a C4 carbon to carbon linkage, is in phase-I clinical trials. Fig. (8). This molecule was selected from a series of C-4/?-alkyl derivatives containing hydroxy, amino, and amido substitutions, among which it was the most active [42]. More detailed studies showed that TOP-53 was twice as active as etoposide at inhibiting DNA-topoisomerase II and is as effective as etoposide at 5 to 6 times lower concentration [43].

Fig. (8). Podophyllotoxin derivatives possessing novel C4 substitutions with a carbon-carbon bond replacing the carbon-oxygen bond

Modification of Ring E

Studies of the structure-activity relationship of CI-substituted derivatives of etoposide, albeit few, have shed light on the relative importance of the nature of the substitution. Most compounds generated have either a hydroxy 1 or a methoxy group at C4' of the aromatic E ring. Analysis of the biological activity of these compounds has clearly demonstrated that

164 the hydroxyl group at C4' is instrumental in the ability of the derivative to inhibit DNA topoisomerase [28]. Novel ring-E analogs having o-quinone, ketal, and spyroketal functionalities, Fig. (9) have been synthesized, but their biological activity has not been tested [44].

r-H MeO

\ OMe OMe

Ti^

MeO

Fig. (9). Modified aryl E-ring possessing o-quinone, ketal, and spiroketal functionalities

Modification of the Aryltetralin Backbone

While the greatest emphasis has been placed on the synthesis of derivatives at the C4 position, several research groups have generated sets of analogs with modified tetralin backbones. One recurrent theme has been the opening of the lactone ring. For example, acid hydrazide and C3(2,2,5,5-tetramethylpyrrolinenyloxy) semicarbazide derivatives have been generated from podophyllotoxin, Fig. (10). The latter compound was as active as etoposide and seemed to have weaker immunosuppressive effects [45]. In contrast, several naphthalene lignans either with or without the lactone ring, such as some junaphthoic acid derivatives, were found to be significantly less active than their aryltetralin counterparts [46]. Derivatives of the naturally occurring podophyllotoxin and its C4-deoxy analog lacking the lactone ring were synthesized and tested on neoplastic systems [47]. The set of twenty two compounds included carbonyl, hydrazone, and oxime derivatives, most of which were active on several tumor cell lines, albeit generally less potently than the natural parent compounds. Computer modeling of the new derivatives suggested that the introduction of a double bond in ring C led to more rigid, as well as more active, structures. Fig. (10). It was hypothesized that the less flexible unsaturated ring C caused a better orientation of the ester group and more closely

165

CHO

\^-"i ^r

COOCH3 O

V

Ar / <

"OH

/

'\' '"VNHNHs >t ' " / - I ^ N H N M L / ^ ^ ^

Fig. (10). Modifications of podophyllotoxin involving opening of the lactone D-ring and introduction of unsaturation in the C-ring

resembled podophyllotoxin. The preparation of simpler triol and ether derivatives obtained by opening the lactone ring has been reported [16], but the biological activity of these compounds was not tested. Compounds with a catechol-type instead of the dioxolane ring A were poorly active [33]. More radical alterations of the basic natural structure, involving the opening of ring A, have been performed. A recent study reported the 11-step total synthesis of an etoposide analog having modifications in rings A and D, Fig. (11).

D o

Fig. (11). Modifications of podophyllotoxin picropodophyllotoxin lactone configuration

involving

the

A-ring

and

using

the

166 This compound possesses a c/^'-fused-lactone ring D, as in picropodophyllotoxin, and a pyridazine ring as cycle A [48]. Diphyllinlike derivatives where ring A was opened, the tetrahydronaphthalene structure was unsaturated to its naphthalene form (rings B and C), and the aryl ring E of podophyilotoxin was retained, Fig. (12) have also been synthesized [49]. The introduction of the double bonds in rings C caused the entire system to be planar, including the hydroxyl group at C4 and the ;r-lactone ring D. The inhibitory activity of these compounds against the

D

O

Fig. (12). Hybrid backbone possessing characteristics of podophyilotoxin and diphyllin natural products

DNA-topoisomerase II was much lower than commercially available compounds. Some attention is being focused on extending the number of cycles beyond ring A in order to generate derivatives that would have better DNA intercalation [50]. A series of synthetic flavonoid/podophyllotoxin hybrid molecules were synthesized with the aim of combining the well-known antitumor activity of many flavonoids and the structural features of the podophyilotoxin considered to be important for activity, such as the dioxolane ring A and the aromatic ring B, Fig. (13). As a whole, these compounds were only weakly cytotoxic towards cultured tumor cells [51]. Another unsuccessful attempt to improve the anticancer properties of podophyilotoxin involved the replacement of the C2 carbon by a nitrogen atom [52]. The aza derivative of

167

deoxypodophyllotoxin inhibited microtubule assembly but lacked inhibitory activity toward topoisomerase; on the other hand, oxidation of ring E of the aza-analog yielded a quinone having non-intercalative antitopoisomerase activity.

-T-R

Fig. (13). Hybrid backbone possession characteristics of podophyllotoxin and flavonoid natural products

Lessons Learned from Studies of Structure-Activity

Relationships

A study that correlated the cytotoxicity of a set of compounds with their mode of action (inhibition of either tubulin polymerization or DNAtopoisomerase II) demonstrated that the effects of podophyllotoxin derivatives could not be consistently predicted [33]. Compounds that inhibited neither site were still highly cytotoxic, whereas some of the most active compounds inhibited both sites, and some strong topoisomerase inhibitors were not necessarily potent anticancer agents. Similarly, several highly active glucosidic derivatives did not show activity against either site [31]. Yet, DNA-topoisomerase II remains the preferred target of podophyllotoxin-derivatization work. Because cytotoxicity and, particularly, low therapeutic index are equated with non-specific induction of mitotic arrest, compounds that inhibit tubulin polymerization are not desired. In fact, all drugs approved thus far and the most promising new derivatives exert their activity through inhibition of the DNA-modifying enzyme. Some features that differentially affect the inhibitory properties of podophyllotoxin analogs have been identified. A graphical summary of these features is shown in Fig. (14). As a rule, compounds with 4a configuration tend to be quite cytotoxic and inhibit microtubule

168 polymerization, whereas those with the Ap configuration are less cytotoxic and inhibit DNA-topoisomerase II.

4p configuration is essential

frans-lactone with 2a,3p configuration is very important

free rotation of ring E is required

4' hydroxy is essential with esterification tolerated

Fig. (14). Diagram summarizing the key features of the podophyllotoxin backbone obtained from structure-activity relationship studies

Commercial podophyllotoxin-based drugs all possess modified 4y^Dglucosides. However, highly active structures have been synthesized that have either amino or carbon-carbon linkage at position C4, suggesting that the sugar structure of etoposide can be simplified considerably while retaining or even improving biological activity. Although many derivatives lacking the lactone ring have shown high anticancer activity, the ;K-lactone ring is critical for topoisomerase inhibition. With few exceptions, the C-2a, C-3y5 trans configuration has been maintained in most studies; typically, /? cis analogs are much less 7active. The importance of the 4-hydroxyphenyl character of the E ring was recognized early on. Derivatization of the hydroxyl group yields significantly less active compounds. Only minor changes such as esterification at 4'-0(E) can be made without diminishing activitiy. Furthermore, free rotation of the E ring appears to be required to maintain topoisomerase inhibitory

169 activity. The dioxolane ring A is considered optimal for anticancer activity. Studies of ring-A opening yielded relatively inactive compounds. In spite of the large amount of work done on the derivatization of podophyllotoxin, a clear picture of the structural features not already present in etoposide that determine whether the molecules inhibit DNAtopoisomerase or microtubule polymerization has not emerged. Etoposide and teniposide are among a group of compounds that bind directly to DNA-topoisomerase II [53]. A recent study has shown that the binding of etoposide to topoisomerase was dependent on the presence of a specific tyrosine residue found in the active site of the enzyme involved in DNA scission [54]. More detailed knowledge than is available today about the enzyme's catalytic site and the binding site of the epipodophyllotoxins is needed to advance drug development. By and large, efforts to rationally design a better antineoplastic drug based on the aryltetralin structure have failed; this fact, together with the large number of possible structural variations, makes the accidental discovery of etoposide all the more remarkable. Antiviral Activity Lignans were first shown to possess antiviral activity over 50 years ago when the successful use of Podophyllum extracts in the treatment of venereal warts, caused by the human papilloma virus, was reported (2). Since then, numerous studies involving the evaluation of podophyllotoxin and its congeners and derivatives as antiviral agents have been performed. The results of these studies and the antiviral activity of other lignans have been reviewed recently [55]. Podophyllum lignans have been shown to inhibit the pathogenic effects of both DNA and RNA viruses. Podophyllotoxin, in particular, showed significant inhibitory activity against herpes simplex virus [56], vesicular stomatitis virus [57], and murine cytomegalovirus [58] at concentrations of 24, 100, and 241 nM, respectively. However, the potency of these lignans as antiviral agents is considered to be insufficient, and their cytotoxicity too pronounced, for therapeutic use. The similarity between the mode of action of aryltetralin lignans as anticancer and antiviral agents is noteworthy. Because of their ability to bind tubulin, these lignans disrupt the cellular cytoskeleton and thus interfere with viral replication. The development of successful

170

podophyllotoxin-based antiviral drugs is likely to require further derivatization work, as was the case for etoposide. Analysis of the results obtained thus far point to structural features, such as the presence of the lactone ring and the a configuration at C2, that positively influence antiviral activity. In addition to tubulin binding, synthetic podophyllotoxin analogs have shown inhibition of reverse transcriptase [59], which may be exploited to selectively combat RNA viruses such as the human immunodeficiency virus (HIV). These studies showed that aryltetralins having a methoxy group at C-8(B) and a 4-benzyloxy-5-demethoxy substituted E-ring had highest selectivity indices against HIV-1. Other Potential Uses The administration of some podophyllotoxin-based drugs causes complex physiological reactions beyond inhibition of tubulin polymerization and DNA-topoisomerase. Benzylidinated podophyllotoxin glucosides are currently in late clinical trials for the treatment of psoriasis and rheumatoid arthritis. Arthritis patients treated with CPH82, a mixture of two semisynthetic benzylidinated podophyllotoxin-based glucosides show reduction of the inflammatory process within three months of therapy [60]. CPH82 stimulates macrophage growth and inhibits lymphokine, this immune response being mediated by levels of cyclic AMP. The results of phase-I and -II trials on CPH82 for rheumatoid arthritis were promising regarding safety and efficacy, however some gastrointestinal inconveniences in 28% of patients were reported. As inhibitors of mitosis, podophyllotoxins have various potential medicinal applications, including uses as immunosupressive, antimalarial and antifungal agents [61]. The biological activity of podophyllotoxins may also have agricultural applications. Podophyllotoxin has been shovm to inhibit larval development [62] and root growth [63]. Phytotoxicity assays carried out in our labs have shown that rye seedUngs are more sensitive than lettuce seedlings to the presence of Podophyllum extracts in the soil, podophyllotoxin being the most active among the lignans [Moraes, unpublished data].

171

CURRENT DEMAND FOR PODOPHYLLOTOXIN AND THE SEARCH FOR ALTERNATIVE SOURCES While the preferred source of podophyllotoxin becomes scarcer, the demand for the compound continues to increase. US sales of etoposide tripled in 1995, after the commercial patent covering the drug expired, and have risen at an annual rate of more than 10% since 1997 [64]. Etoposide is used in combination therapy in refractory testicular, lymphoid and myeloid leukemia, stomach, ovarian, brain, breast, pancreatic, and small and non-small cell lung cancers. Approximately 60 clinical trials are under way to test etoposide for new indications, and another 140 trials use the drug as positive control. In addition, numerous new podophyllotoxin derivatives are currently under development and evaluation as topoisomerase inhibitors and potential anticancer drugs. Some of these compounds have shown improved therapeutic properties over etoposide [30, 65]. Podophyllotoxin-based pharmaceutical products have also been formulated for the treatment of rheumatoid arthritis, showing both efficacy and low toxicity [60]. Furthermore, the application of creams and solutions containing podophyllotoxin is still a common and effective treatment for genital warts, the incidence of which is rapidly increasing [66]. The amounts of podophyllotoxin and Podophyllum rhizomes required to satisfy the current demand of the pharmaceutical industry are difficult to estimate because of lack of published data. Demand is apparently increasing and will probably continue to do so as new podophyllotoxin derivatives become useful drugs and new therapeutic uses are approved for older ones. The destructive collection of wild-growing rhizomes of P. emodi has drastically reduced the size of natural populations. Even a rapidly growing plant such as mayapple may become endangered if the harvest of natural stocks to satisfy the current and anticipated need for podophyllotoxin occurs without proper replenishment. Consequently, scientists have considered it essential to develop alternative or more reliable sources of aryltetralin lignans, including chemical synthesis, bioprospecting for plant species that produce aryltetralin lignans, in vitro culture, and domestication of Podophyllum. Even though a long-term solution appears to have been found in the utilization of leaves of P. peltatum, alternative sources may still be useful for the identification of novel lignans and biological activities.

172 Chemical Synthesis The therapeutic success of semisynthetic podophyllotoxin derivatives such as etoposide has prompted researchers to explore synthetic routes to podophyllotoxin and analogs, both as an alternative to reUance on natural sources and a direct approach to alternative aryltetralin structures. Synthetic work reported up to 1991 has been categorized into four general approaches according to the key intermediate (the oxo ester and dihydroxy ester routes) and the key step (the tandem conjugate addition and DielsAlder reaction routes) involved [67]. With few exceptions, the synthesis of podophyllotoxin and related compounds leads to products having undesired relative configurations at CI, C2, and C3. Lengthy procedures are usually required to obtain the 2,3-trans configuration. More recently, the 8-step total synthesis of podophyllotoxin, with an optical purity of 98% and in an overall yield of 15% was reported [68]. Biotechnological and enzymatic approaches have also been tried. Cell cultures of P. peltatum transformed a synthetic precursor mostly into a podophyllotoxin analog with the targeted 2,3'trans configuration [69]. As an extension of the use of biological systems to establish the stereochemistry of the resulting lignan, cell-free enzyme preparations were used to catalyze the oxidative cyclization [70] and the selective deacetylation [71] of synthetic substrates. Although interesting, these approaches still involve lengthy and low-yielding multi-step procedures, and suffer from limitations typical of biotechnological processes, namely the high costs of enzymes and in vitro culture, as well as the instability of the enzymes. Alternative Natural Sources Aryltetralin lignans are widely distributed in the plant kingdom. In addition to plants of the Berberidaceae, most saliently in the genera Podophyllum, Diphylleia, and Dysosma, these compounds have been found in plants of numerous genera of other families, such as Hyptis, Teucrium, Nepeta, Thymus, Phomis and Salvia, in the Lamiaceae [72, 73], and Linum and Hugonia in the Linaceae [74, 75, 76]. These compounds are also present in gymnosperms of the genera Juniperus and Thujopsis [11, 78, 79]. For the most part, however, the quantities of aryltetralin

173

lignans obtained from plants other than Podophyllum are quite small (Table 1). With the exception of some species of Linum that are widely cultivated, the productive potential of these other sources appears to be very limited. The biotechnological approach to podophyllotoxin production, namely in vitro culture, has not been limited to biotransformation of complex intermediates, but has included attempts to obtain the compound from sugars and other simple precursors. Podophyllum callus has been induced to differentiate into tissues containing 1.6% podophyllotoxin, based on dry weight [85]. Callus, cell, and root cultures of Podophyllum have been shown to accumulate small amounts of podophyllotoxin [86, 87, 88]. Cell cultures of Linum flavum [80] and Linum album [89] have also been tested. The characteristically low productivity of undifferentiated cultures can be improved by the addition of precursors to the culture medium. For example, feeding 0.8 to 2.1 mM coniferin to cell suspension cultures of Indian Podophyllum resulted in an increase in podophyllotoxin production of up to 12.8 times relative to that obtained from control cultures [21]. Coniferin is not commercially available and earlier precursors such as phenylalanine, caffeic acid, and ferulic acid did not improve production. The use of coniferyl alcohol in feeding experiments is limited by its poor solubility, a problem that has been addressed by complexation with pcyclodextrin, which increases the water solubility of the monolignol and results in higher podophyllotoxin production [20]. Leaves of P. peltatum as a Renewable Source of Podophyllotoxin Until recently, the rhizomes and roots of P. emodi were considered the richest source of podophyllotoxin, reportedly containing 4.3% of the compound on a dry-weight basis [81]. It is now known that, if properly extracted, the leaves of P. peltatum may yield similar or larger quantities of podophyllotoxin [26]. Our research group chose to investigate the productive potential of P. peltatum because it is a known source of podophyllotoxin and its natural habitat covers a large area where it grows vigorously and abundantly. In April 1998, a survey of much of the natural range of P. peltatum was carried out to evaluate the biosynthetic capacity of the plants. Highyielding accessions were identified and collected both for extraction and propagation purposes. Encouraged by the identification of naturally

174 occurring chemotypes of mayapple that accumulate large amounts of lignan glucosides [90] our group recently developed an extraction method that takes advantage of the presence of highly active lignan y^Dglucosidases in dried tissues of Podophyllum [26]. This extraction method consists of rehydrating powdered tissues to allow the endogenous glucosidases to convert podophyllotoxin glucoside into the aglycone and thus makes it available for extraction using a water-immiscible organic solvent such as ethyl acetate. Lignan glycosides, including podophyllotoxin 4-y5-D-glucopyranoside were first extracted from Podophyllum rhizomes in 1954 by researchers who hoped that these compounds, in analogy to cardiac glycosides of Digitalis, would exhibit pharmacological properties superior to those of the aglycones [91]. Enthusiasm for the natural glucosides soon waned because they showed less potency than the aglycone as well as poor bioavailability. The leaves of P. peltatum were sufficiently rich in /?-Dglucosides to permit labeling studies, but the possibility of using endogenous enzymes to produce the aglycones was not recognized. Instead, the leaves were macerated with water and treated with almond figlucosidase [22]. Similarly, to liberate the aglycone from tissues of F. intermedia, the plant material was first extracted with methanol, and the crude extract then reacted with almond y5-glucosidase in acetate buffer [19]. The new extraction procedure does not require the addition of catalytic agents, can be carried out at ambient temperature and makes use of inexpensive buffering agents. Aqueous extraction of leaves of glucosiderich chemotypes of P. peltatum yielded over 5% of podophyllotoxin, based on dry weight, exceeding levels previously reported from any source. The significance of these developments cannot be overstated. Leaves are entirely renewable and their harvest does not destroy the plant. The use oi Podophyllum leaves opens the door for the managed utilization of a natural resource, avoiding the destruction of wild populations and virtually guaranteeing the long-term availability of raw material. Cultivation of American Podophyllum

The cultivation of mayapple as a medicinal cash crop was proposed a quarter of a century ago on the basis of its high demand as a source of

175 bioactive phytochemicals [6]. The demonstration that leaves of mayapple can be a highly productive source of podophyllotoxin has made the cultivation of this plant rather attractive. We are currently pursuing the domestication of P. peltatum in order to develop a reliable supply of extractable raw material of consistently high quality at a reasonable and predictable cost. Much work is still needed, however, to propagate highly productive chemotypes and establish the agronomics of mayapple cultivation. To this end, we are carrying out work on several fronts: 1) Identification of highly productive genotypes, 2) Establishment of efficient propagation protocols, and 3) Determination of the environmental factors that affect productivity. Chemical analysis of leaves harvested from 17 naturally occurring mayapple colonies suggested that, if collected randomly from the wild and extracted with an aqueous agent, the podophyllotoxin yield of the combined leaf biomass would exceed 3% of dry weight. Notwithstanding the costs of research and the initial investment needed to plant the required acreage, cultivation of mayapple genotypes that produce over 5% podophyllotoxin may be an economically superior alternative than the managed utilization of natural populations. We have thus far collected six wild mayapple accessions, the leaves of which can accumulate between 4 and 6% podophyllotoxin [82]. Specimens from these colonies are being propagated in our laboratories. We have considered three options for the propagation of mayapple; i.e., by seed, by rhizome cuttings, and micropropagation. The use of seeds is hampered by a number of factors such as low seed availability, poor germination rates, high seedling mortality, and a long juvenile stage. Assuming that the rate of germination and seedling establishment in mayapple can be improved, there may still be a delay of four to five years, before rhizome growth and colony expansion occur. Micropropagation offers the advantage of bypassing the difficult germination and seedling stages. A protocol for the in vitro propagation of P. peltatum has been established [92]. Following this protocol, the formation of adventitious buds can be easily induced, and a large number of rooted buds that adapt well to soil conditions under a controlled environment can be generated. Success rates for the establishment of micropropagated buds are as high as 85%. Unfortunately, micropropagated buds follow a developmental pattern similar to that of buds produced sexually, taking four to five years to initiate rhizome branching. In contrast, rhizome cuttings derived from a mature rhizome system retain their adult character in terms of their ability

176

to grow vigorously and branch. Rhizome cuttings containing a node or a terminal bud root readily in sandy soils. Roots grow at the base of the terminal bud. For cuttings containing a node formed during past growing seasons, the rooting process begins after a new branch has developed and given rise to a new terminal bud. Cuttings should therefore be made in early fall at leaf senescence, forcing the new terminal bud to develop before the temperature drops. Although the number of propagules that can be generated by micropropagation is far larger than that obtained from rhizome cuttings, the latter option appears to offer the most direct route to the replication of highly productive genotypes. Besides bypassing the juvenile stage of development, propagules derived from rhizome cuttings are more likely to be true to type than those obtained sexually or from in vitro culture. Currently, our group is carrying out multisite and multiyear evaluation experiments to determine to what extent highly productive phenotypes are the result of genotypic factors, and how environmental factors affect the productivity of our accessions. Some basic agronomic issues are also being addressed. Mayapple requires a well-drained soil, but can grow in a variety of different soil pHs. Our accessions were collected in sites where soil pH ranged from 4.3 to 7.2. No correlation was found between soil pH and podophyllotoxin content. Being a shade-tolerant plant and one of the first plants to grow in early spring, weed competition may not affect P. peltatum as much as other crops. Issues related to pest and disease control are very important. The presence of podophyllotoxin may afford protection from herbivores, but is not effective against the specialist fungus Puccinia podophylli. The incidence of this pathogen, to which sexual progeny are more vulnerable than are shoots produced vegetatively [93], is not only a contributing factor to seedling mortality, but also greatly decreases leaf area and therefore causes the loss of productive biomass and inhibition of rhizome growth. Preliminary observations suggest that there is variability among our accessions regarding their susceptibility to the fungus. In addition, plants growing in the shade appear to be more susceptible than those growing in open spaces. Shade conditions may also affect the onset of leaf senescence, which appears to occur earlier in the shade, and rhizome branching, which seems to be promoted under full sun. The determination of the optimal time of leaf harvest is considered essential. On the one hand, leaves should be harvested when podophyllotoxin content and leaf area are greatest, but on the other, leaves

177

should remain in place sufficiently long to allow the development of new rhizome segments. Manipulation of biosynthetic pathway Efforts to increase the production of a particular metabolite by manipulation of the biosynthetic pathway primarily target those biochemical branch points at which competition for precursors is likely to occur. Availability and partitioning of coniferyl alcohol is of direct relevance for production of podophyllotoxin. In addition to its importance for lignan biosynthesis, coniferyl alcohol is also a key intermediate in lignin production [94, 95]. Lignin biosynthesis is a temporally and spatially regulated process that may involve the action of pathway-specific enzymes and specialized subcellular structures [96]. For instance, in nonwoody plants, the bulk of lignin biosynthesis is associated with the formation of vascular tissues. Therefore, even though both lignan and lignin biosynthesis utilize monolignol precursors, the two pathways may not be directly competing for these. Nevertheless, the demonstration that suppression of lignin biosynthesis does not compromise structural integrity in transgenic aspens [97], offers the possibility to explore the effect of down-regulating the lignin pathway on lignan biosynthesis. The possible use of species outside the genus Podophyllum for podophyllotoxin production must take into account the characteristic stereospecificity of the monolignol coupling mechanism, which is known to vary among plant taxa [98]. The reduction of pinoresinol is thought to be a key branch point in lignan biosynthesis since it gives rise to precursors for various types of 88' lignans. A prerequisite for the genetic manipulation of the pathway at this point has been fulfilled by the recent cloning of the gene encoding the NADPH-dependent enzyme (+)-pinoresinol/(+)-lariciresinol reductase in F. intermedia [99]. Rhizomes and leaves of P, peltatum show different patterns of lignan accumulation that point to other regulatory controls within the lignan biosynthetic pathway. The ratio of podophyllotoxin to peltatin content ranged widely in leaves of various genotypes of wildgrowing P. peltatum, revealing a strongly negative correlation between the quantities of these compounds [82]. This observation suggests that precursor utilization at the matairesinol level and branch points downstream from matairesinol are differentially regulated in roots and

178

leaves. Directing metabolic flux away from peltatin end products by controlling these key biochemical steps, either genetically or by modifying growth conditions, may bring about an increase in podophyllotoxin production. CONCLUDING REMARKS More than fifty years after the first medicinal application of the antimitotic activity of podophyllotoxin was proposed, the aryltetralin lignans of Podophyllum continue to be subjects of intense interest and research. Derivatization of podophyllotoxin has produced the potent and widely used antineoplastic drugs etoposide and teniposide. Numerous other analogs have been synthesized in the hope of increasing the usefulness of these drugs, but little improvement has been achieved. These efforts, nevertheless, have generated large amounts of data that may facilitate the design of podophyllotoxin-derived drugs having novel and improved pharmacological properties. On the other hand, significant progress has been made towards securing a reliable source of podophyllotoxin. This compound is, still today, primarily obtained from wild-harvested rhizomes of P. emodi, a species considered in danger of extinction due to overcollection and loss of habitat. The discovery that the leaves of P. peltatum, a North American species, produce large amounts of podophyllotoxin not only promises to make this plant the preferred source of the compound, but also offers the possibility to establish it as a new high value cash crop. ACKNOWLEDGEMENTS This work was partly funded by USDA grant 97-35501-4886. We thank Dr. Daneel Ferreira for his help in the nomenclature of some of the compounds. REFERENCES [1] [2] [3] [4]

Neuburger, J. Arch. Exp. Pathol Pharmakol 1891, 28, 32. Kaplan, I.W. New Orleans Med. Surg. J. 1942, 94, 388. King, L.S.; Sullivan, M. Sci. 1946, 104, 244. Hartwell, J.L.; Shear, M.J. Cancer Res. 1947, 7, 716.

179

[5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32]

Stahelin, H.F.; von Wartburg, A. Cancer Res. 1991, 51, 5. Meijer, W. Econ. Bot. 1974, 28, 68. Sohn, J.J.; Policansky, D. Ecol. 1977, 58, 1366. Geber, M.A.; de Kroon, H.; Watson, M.A. J. Ecol 1997, 85, 211. de Kroon, H.; Whigham, D.F.; Watson, M.A. Functional Ecol 1991, 5, 360. Nadkami, M.V.; Hartwell, J.L.; Maury, P.B.; Leiter, J. J. Amer. Chem. Soc. 1953,75, 1308. Bhadula, S.K.; Anoop, S.; Lata, H.; Kuniyal, C.P.; Purohit, A.N. Plant Gen. Resources Newslet. 1996, 106,26. Institute of Himalayan Bioresource Technology, Palampur, India, Annual Report, 1999. Whiting, D.A. Nat. Prod. Rep. 1985, 2, 499. Wallis, A.F.A. In Lignin and lignan biosynthesis; Lewis, N.G., Sarkanen, S., Eds.; American Chemical Society: Washington, 1998; ACS Symposium Series 697, pp. 323-333. Schrecker, A.W.; Hartwell, J.L. J. Org. Chem. 1956, 21, 381. Castro, M.A.; Gordaliza, M.; del Corral, J.M.M.; San Feliciano, A. Org. Prep.Procedures, 1994, 539. Petcher, TJ.; Weber, H. P.; Kuhn, M.; von Wartburg, A. J. Chem. Soc, Perkin Trans. 2, 1973, 288. Lewis, N.G.; Davin, L.B.; Sarkanen, S. In Lignin and lignan biosynthesis; Lewis, N.G.; Sarkanen, S. Eds.; American Chemical Society: Washington, 1998; ACS Symposium Series 697, pp. 1-27 Rahman, M.M.A.; Dewick, P.M.; Jackson, D.E.; Lucas, J.A. Phytochemistry, 1990,29, 1841. Woerdenbag, H.J.; van Uden, W.; Frijlink, H.W.; Lerk, C.F.; Pras, N.; Malingre, T.M. Plant Cell Rep. 1990, 9, 97. van Uden, W.; Pras, N.; Malingre, T.M. Plant Cell Tis. Org Cult. 1990, 23, 217. Broomhead, A.J.; Rahman, M.M.; Dewick, P.M.; Jackson, D.E.; Lucas, J.A. Phytochemistry, 1991, 30, 1489. Kamil, W.M.; Dewick, P.M. Phytochemistry, 1986, 25, 2089. Davin, L.B.; Bedgar, D.L.; Katayama, T.; Lewis, N.G. Phytochemistry, 1992, 31,3869. Umezawa, T.; Davin, L.B.; Lewis, N.G. J. Biol. Chem. 1991, 266, 10210. Canel, C; Dayan, F.E.; Ganzera, M.; Khan, I.A.; Rimando, A.; Burandt, C.L. Jr.; Moraes, R.M. Planta Med. 2000, in press. Schacter, L. Sem. Oncol. 1996, 23, 1. Long, B.H.; Musial, S.T.; Brattain, M.G. Biochem. 1984, 23, 1183. Mross, K.; Huttmann, A.; Herbst, K.; Hanauske, A.-R.; Schilling, T.; Manegold, C; Burk, K.; Hossfeld, D.K. Cancer Chemother. Pharmacol. 1996, 38, 217. Pagani, O.; Zucchetti, M.; Sessa, C; de Jong, J.; DTncalci, M.; de Fusco, M.; Kaeser-Frohlich, A.; Hanauske, A.; Cavalli, F. Cancer Chemother. Pharmacol. 1996,38,541. Daley, L.; Guminski, Y.; Demerseman, P.; Kruczynski, A.; Etievant, C; Imbert, T.; Hill, B. T.; Monnert, C. J. Med Chem. 1999, 41, 4475. Gupta, R.S.; Chenchaiah, P.C. Anti-cancer Drug Design, 1987, 2, 13.

180

[33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56]

Terada, T.; Fujimoto, K.; Nomura, M.; Yamashita, J.-I.; Kobunai, T.; Takeda, S.; Wierzba, K.; Yamada, Y.; Yamaguchi, H. Chem. Pharm. Bull 1992, 40, 2720. Huang, T.-S.; Lee, C.-C; Chao, Y.; Shu, C.-H.; Chen, L.-T.; Chen, L.-L.; Chen, M.-H.; Yuan, C.-C; Whang-Peng, J. Pharmaceut. Res. 1999, 16, 997. Zhu, X.-K.; Guan, J.; Tachibana, Y.; Bastow, K. F.; Cho, S. J.; Cheng, H.-H.; Cheng, Y.-C; Gurwith, M.; Lee, K.-H. J. Med Chem. 1999, 42, 244L Yan-Guang, W.; Jian-Lin, P.; Yao-Zu, C. Curr Sci. 1996, 71,312. Ying-jie, C; Xuan, T. Cur Sc. 1998, 75, 1383. Jian-Iin, P.; Yan-guang, W.; Yao-zu, C. Cur Sci. 1997, 72, 268. Liu, S.-Y.; Hwang, B.-D.; Haruna, M.; Imakura, Y.; Lee, K.-H.; Cheng, Y.-C. Mol. Pharmacol. 1989, 36, 78. Zhang, Y.-L.; Shen, Y.-C; Wang, Z.-Q.; Chen, H.-X.; Guo, X.; Cheng, Y.-C; Lee, K.-H. J. Nat. Prod 1992, 55, 1100. Lear, Y.; Durst, T. Can. J. Chem. 1996, 74, 1704. Terada, T.; Fujimoto, K.; Yamashita, J.; Wierzba, K.; Yamazaki, R.; Shibata, J.; Sugimoto, Y.; Yamada, Y. J. Med Chem. 1993, 36, 1689. Utsugi, T.; Shibata, J.; Sugimoto, Y.; Aoyagi, K.; Wierzba, K.; Kobunai, T.; Terada, T.; Oh-hara, T.; Tsuruo, T.; Yamada, Y. Cancer Res. 1996, 56, 2809. Pelter, A.; Ward, R.; Ma, W.-Y. J. Nat. Prod 1994, 57, 1598. Wang, J.-Z.; Tian, X.; Tsumura, H.; Shimura, K.; Ito, H. Anti-cancer Drug Design, 1993,8, 193. Ward, R.S. Nat. Prod Rep. 1993, 10, 1. Gordaliza, M.; Castro, M.A.; del Corral, J.M.M.; Lopez-Vazquez, M.L.; Garcia, P.A.; San Feliciano, A.; Garcia-Davalos, M.D.; Broughton, H. Tetrahedron, 1997,53, 15743. Meresse, P.; Bertounesque, E.; Imbert, T.; Monneret, C Tetrahedron, 1999, 55, 12805. Kamal, A.; Atchinson, K.; Daneshtalab, M.; Micetich, R.G. Anti-cancer Drug Design, 1995, 10,545. MacDonald, T.L.; Lehnert, E.K.; Loper, J.T.; Chow, K.C; Ross, W.E. In DNA Topoisomerase in Cancer, Potmesil, M.; Kohn, K.W., Eds; Oxford University Press; New York, 1991; pp. 119-214. Gonzalez de Peredo, A.; Leonce, S.; Monneret, C; Dauzonne, D. Chem. Pharm. Bull. 1998, 46, 79. lida, A.; Kano, M.; Kubota, Y.; Koga, K.; Tomioka, K. Bioorg. Med. Chem. Lett. 1997,7,2565. Gieseler, F.; Glasmacher, A.; Kampfe, D.; Nubler, V.; Valsamas, S.; Kunze, J. Clark M. Acute Leukemias VI1997, 538. Kingma, P.; S.; Burden, D. A.; Osheroff, N. Biochem. 1999, 38, 3457. Charlton, J.L. J. Nat. Prod 1998, 61, 1447. Markkanen, T.; Makinen, M.L.; Maunuksela, E.; Himanen, P. Drugs Exptl. Clin.

ResAm,l,l\\. [57] [58]

San Feliciano, A.; Gordaliza, M.; del Corral, J.M.M.; Castro, M.A.; GarciaDavalos, M.D.; Ruiz-Lazaro, P. Planta Med 1993, 59, 246. MacRae, W.D.; Towers, G.H.N. Phytochemistry, 1984, 23, 1207.

181

[59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88]

Hara, H.; Fujihashi, T.; Sakata, T.; Kaji, A.; Kaji, H. AIDS Res. Human Retrov. 1997, 13, 695. Bjomeboe, O.; Moen, F.; Nygaard, H.; Haavik, T.K.; Svensson, B. Scand. J. Rheumatol 1998, 27, 26. Figgitt, D.P.; PhD. Dissertation, 1990. University Nottinghan, Nottinghan, UK. Russel, G.B.; Singh, P.; Fenmore, P.G. Aust. J. Biol Sci. 1976, 29, 99. Cronin, K. E.; Ellis, J.R.; Hussey, P.I.; Ray, J.A.; Waldin, T.R. 1999, US patent 5,888,818. IMS Health 1999 Leeson, P.A.; Castaner, J. Drugs of the Future, 1996, 21, 1136. Claesson, U.; Lassus, A.; Happonen, H.; Hogstrom, L.; Siboulet A. Int. J. STD. AIDS. 1996,1,429. Ward, R. S. 1992. Synthesis, 1992, 719. Bush, E.J.; Jones. D.W. J. Chem. Soc, Perkin Trans. I, 1996, 1, 151. Kutney, J.P.; Chen, Y.P.; Gao, S.; Hewitt, G.M.; Kuri-Brena, F.; Milanova, R.K.; Stoynov, N.M. Heterocycles 1993, 36, 13. Kutney, J.P.; Du, X.; Naidu, R.; Stoynov, N.M.; Takemoto, M. Heterocycles 1996,42,479. Berkowitz, D.B.; Maeng J.; Dantzing, A. H; Shepard R. L.; Norman. B.H. J. Am. Chem. Soc. 1996, 118, 9426. Kuhnt, M; Rimpler, H.; Heinrich, M. Phytochemistry, 1994, 36, 485. Novelo, M.; Cruz, J.G; Hernandez, L.; Miranda, R.P.;.Chai, H; Mar, W.; Pezzuto, J.M. J. Nat. Prod. 1993, 56, 1728. Broomhead, A. J; Dewick. P.M Phytochemistry, 1990, 29, 3831. Konuklugil, B. Fitoterapia, 1996, LXVII, 379. Weiss, S.G.; Tin-Wa, M.; Perdue, Jr., R.E.; Farnsworth. N.R.J. Pharm. Sci. 1975,64,95. San Feliciano, A.; Medarde, M.; Lopez, J. L.; Puebla, P.; Miguel del Corral, J. M.; Barrera, A.F. Phytochemistry, 1989, 28, 2863. San Feliciano, A.; Miguel del Corral, J.M.; Gordaliza, M.; Castro, A. Phytochemistry, 1990,29, 1335. Akohori, A.; Yasuda, F.; Ando, M.; Hori, K.; Okanishi, T. Chem. Pharm. Bull 1972,20, 1150. van Uden, W.; Bos, J.A.; Boeke, G.M.; Woerdenbag, H.J.; Pras, N. J. Nat Prod 1997,60,401. Jackson, D.E.; Dewick, P.M. Phytochemistry, 1984, 23, 1147. Moraes, R.M.; Burandt, C.L. Jr.; Ganzera, M.; Khan, I.A.; Canel, C. Econ. Bot. 2000, in press. Choi, Y.H.; Kim J.Y.; Ryu, J.H.; Yoo, K.P.; Chang Y.S.; Kim, J. Planta Med. 1998, 64,482. Jackson, D. E.; Dewick, P.M. Phytochemistry, 1985, 24, 2407. Sakata, K.; Morita, E.; Takezono, T. European Patent Office, EP 424,928,1991. Kadkade P.G. Plant Sci. Lett. 1982, 25, 107 van Uden, W.; Pras, N.; Visser, J.F.; Malingre, T. M. Plant Cell Rep. 1989, 8, 165. Arumugan N.; Bhojwani S.S. Can. J. Bot. 1990, 68, 487.

182

[89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99]

Smollny, T.; Wichers, H.; Kalenberg, S.; Shahsavari, A.; Petersen, Alfermann, M. Phytochemistry, 1998, 48, 975. Bastos, J. K.; Burandt, C. L. Jr.; Nanayakkara, N. P. D.; Bryant, L.; McChesney, J. D. J. Nat. Prod. 1996, 5P, 406. Stoll, A.; Renz, J.; von Wartburg, A. J. Am. Chem. Soc. 1954a, 76, 3103. Moraes-Cerdeira, R.M.; Burandt Jr., C.L.; Bastos, J. K.; Nanayakkara, N.P.D.; McChesney, J. D. Planta Med. 1998, 64, 42. Parker, M.A. Amer. Natur. 1998, 132, 538. Humphreys, J.M.; Hemm, M.R.; Chappie, C. Proc. Natl Acad Sci. U.S.A. 1999, 96, 10045. Osakabe, K.; Tsao, C.C.; Li, L.; Popko, J.L.; Umezawa, T.; Carraway, D.T.; Smeltzer, R.H.; Joshi, C.P.; Chiang, V.L. Proc. Natl. Acad Sci. U.S.A. 1999, 96, 8955. Lewis, N.G.; Davin, L.B. In Lignin and lignan biosynthesis', Lewis, N.G., Sarkanen, S., Eds.; American Chemical Society: Washington, 1998; ACS Symposium Series 697, pp. 334-361. Hu, W.J.; Harding, S.A.; Lung, J.; Popko, J.L.; Ralph, J.; Stokke, D.D.; Tsai, C.J.; Chiang, V.L. Nature Biotechnol. 1999, 17, 808. Umezawa, T.; Okunishi, T.; Shimada M. In Lignin and lignan biosynthesis', Lewis, N.G.; Sarkanen, S. Eds.; American Chemical Society: Washington, 1998; ACS Symposium Series Vol 697, pp. 377-388. Dinkova-Kostova, A.; Gang, D.R.; Davin, L.B.; Bedgar, D.L.; Chu, A.; Lewis, N.G. J. Biol. Chem. 1996, 271, 29473.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

183

NEW FINDINGS ON THE BIOACTIVITY OF LIGNANS JOSE LUIS RIOS, ROSA M^ GINER, JOSE M^ PRIETO Departament de Farmacologia, Facultat de Farmdcia, Universitat de Valencia. Vicent Andres Estelles. 46100 Burjassot, Valencia, Spain ABSTRACT: This chapter reviews the lignans with biological and pharmacological activity, and includes new natural products and some synthetic or semisynthetic compounds. Lignans are widespread in plants and in many cases are tlieir effective principles. They play an important role as phytoestrogens in preventing menopausial symptoms, osteoporosis, cancer and heart diseases. Lignans possess anticancer and antiviral properties and specifically inliibit certain enzymes and mediators involved in inflammation and immunity processes. Tliey affect the cardiovascular system by different mechanisms including tlie modification of phosphodiesterase activity and platelet activating factor function. Lignans can modify liver function, inhibit hpid peroxidation and free radical-mediated processes and exhibit antioxidant activity. Toxicity to fiingi, bacteria, parasites and insects has also been documented. In several cases the most relevant structure-activity relationship established for certain synthetic lignans in different activities have been summarised.

FOREWORD Lignans represent a widespread, structurally diverse class of phenylpropanoid derivatives of considerable pharmacological interest. Increasing research on new^ structures and interesting biological activities has greatly enlarged our knowledge of this area of phytochemistry and pharmacognosy. An overview of the traditional importance of these metabolites offers diverse examples of crude drugs containing lignans as effective constituents. These crude drugs have been used extensively for their medicinal properties including antiviral, antitumor, anti-inflammatory and immunosuppressive actions among many others. The Podophyllum species (Berberidaceae) include antitumor lignans such as podophyllotoxins, from which synthetic analogues currently in clinical use in the treatment of many cancers are obtained. They also exhibit strong activity as potent anti-HIV agents. Schisandra (Schisandraceae) fruits, used successfully for the treatment of hepatitis, contain lignans effective against chronic viral hepatitis and with protective action against liver damage by hepatotoxins. Several of them also affect the cardiovascular system, decrease heart rate

184 and display platelet activating factor antagonist activity. Sesame seed lignans, used to prevent ageing, possess antioxidative properties and those contained in flaxseed, modified by intestinal bacteria and found in relatively large amounts in urine and other physiological fluids, afford protection against hormone dependent cancers. A great many reviews focusing on the distribution [1-4], biosynthesis pathways [1,2], chemistry and synthesis [5-8] and biological activities of lignans [9-11] have been published. Several refer to antiviral [12,13] and cardiovascular [14] activities, while others centre on podophyllotoxins [15,16], which are particularly important in cancer chemotherapy. Periodic reviews covering most of these aspects are published in Natural Product Reports [5-8,17-19]. The present review covers lignans whose biological properties have been reported from 1995 to 1999. In addition, the relevant activities and semisynthetic and synthetic analogues and the established structure-activity relationships are included. AN INTRODUCTION TO LIGNANS The lignans are secondary plant metabolites biosynthetically derived from the shikimate pathway. They possess diverse structures, usually dimers of phenylpropanoid units, although compounds containing three, four or even five such units, have been reported. The number of higher oligomers identified is currently increasing. Two classes of these natural products arising from the coupling of two cinnamate-derived units can be distinguished: lignans and neolignans. The term lignan, coined by Harworth, is used to define those metabolites formed via direct coupling of at least two C6-C3 units, exclusively derived from phenylalanine or tyrosine, and linked by the central carbon atoms (P,P' or 8,8') of their side chains. The term neolignan, introduced by Gottlieb, implies dimeric and oligomeric products of C6-C3 units linked otherwise than P,p'. Neolignans have been redefined as the products of oxidative coupling of allyl- or propenyl-phenols, while lignans are those resulting from the coupling of cinnamyl alcohols. Some authors [20] have recognised that in addition to simple lignans, there are different structural classes of oxygenated lignans: the lignanolides, which possess a lactone moiety, monoepoxy and bisepoxy lignans; cyclolignans, which contain a third six-member ring; and neolignans, obtained by unsymmetrical carboncarbon binding in the side chains. Lewis et al, [2] examined dimeric and

185 oligomeric products containing lignan or neolignan linkages classified into subgroups based on basic structural parameters. In view of this multiplicity it seems necessary to adopt a rational system of nomenclature. Occurrence Lignans are fairly widespread throughout the plant kingdom and have been documented in species belonging to up to seventy plant families. They have been identified in pterydophytes, gymnosperms and angiosperms. Their functions and ubiquitous distribution evidences their role in plant evolution, as the structure of lignans increases in complexity with the evolution of gymnosperms and angiosperms. They have been considered one of the earliest forms of defence to evolve in vascular plants [21]. Even within the bryophytes, the simple 8,8' linked lignans (+)megacerotonic and (+)-anthocerotonic acids, lacking regiospecific (9methylation at C3, have been identified in some hornworts (Anthocerotae). However, the lignans found in the Blechnaceous tree ferns, within the Pterydophytes, have 0-specific methylation and reductive modifications [2]. Castro et al [3] reviewed the distribution of lignans in the six families of the Coniferae order of Gymnosperms. They found different structural types in each, with the exception of Cephalotaxaceae for which no lignans had been reported. Cupressaceae and Pinaceae contain the largest variety of compounds, including neolignans, which are not detected in the other families. Norlignans are common in Taxodiaceae but are absent in Pinaceae. Further relevant features were the presence of dibenzylbutanediols in all the families except Taxodiaceae. Kadsuranin, identified in Larix leptolepis (Pinaceae) is the unique dibenzocyclooctene reported in conifers. Lignans are also found in Taxaceae, Ephedraceae and Ginkgoaceae [2]. Angiosperms constitute the most varied source of lignans. In the dicotyledons, 8,8'-linked lignans containing oxygenated functional groups at C9/C9' and lignans possessing methyl or methylenic groups at C9/C9' are common constituents. The structural variety in the monocotyledons is quite restricted, represented essentially by cyclobutanes and 5,5' or 3,3'-linked dimers [2]. The angiosperm lignans, as well as the gymnosperm lignans, show structural progressions in the superorders and their skeletal type is closely correlated with certain superorders.

186 Lignans occur in all parts of plants, such as heartwood, bark, roots, rhizomes, stems, leaves, flowers, fruits and seeds, as well as in secreted products such as resins. They may be found in variable content in both the free state and glycosidic combination. Lignans have also been detected in mammalians. They are either of dietary origin or are the metabolic products of the gut microflora. The first two lignans identified from these sources were the dibenzylbutyrolactone enterolactone and the dibenzylbutane enterodiol. Both arise from gut microflora action on the lignans matairesinol and secoisolariciresinol contained in whole grains, vegetable fibre, vegetables and finaits. Physiological significance The lignans play an important part in the defence mechanism of many plant species, against pathogens and predators [22]. Various roles in chemical defence, such as fungicidal, bactericidal and insecticidal, have been demonstrated for these secondary metabolites. In their defence strategy some Asteraceae and Piperaceae species include lignans that are inhibitors of the polysubstrate monooxygenases, polyphagous insect enzymes responsible for the biotransformation of plant metabolites. The planar methylenedioxyphenyl ring binds to the heme of the polysubstrate monooxygenases forming stable adducts that are not easily displaced by carbon monoxide or by most other ligands. Diasesartemin, dillapiol, sesamolin and cubebin inhibit polysubstrate monooxygenases in the European corn borer, Ostrinia nubilalis, an important crop pest and highly polyphagous insect [23]. The inhibition of the Phanerochaete chrysosporium lignin peroxidase by bis(methylenedioxy)cinnamic acid lends support to the hypothesis that the accumulation of lignans at the site of a wound may be a plant strategy to inhibit the fungal enzymes involved in lignin degradation [24]. Lignans from the Myristicaceae Virola sebifera and Otoba parvifolia were toxic to the symbiotic fungus of leaf-cutting ants of the species Atta sexdens rubropilosa [25]. Lignans together with other allelochemicals may play a role in synergising antifeedant or insecticidal properties. Sesamin and asirin showed synergistic properties with pyrethrum insecticides [2]. Pinoresinol, bis-epipinoresinol, the hemicetal and the diacid isolated from the Brazilian

187 Melia azedarach (Meliaceae), displayed phago-inhibitory activity against XhtYitmvpttvdiRhodniusproUxus [26]. Relatively few lignans have been recorded as allelopathic agents, and it is therefore interesting to find that dibenzylbutane nordihydroguairetic acid (NDGA), isolated from the creosote bush Larrea tridentata (Zygophyllaceae), is responsible for the growth inhibition of surrounding grasses [27]. A tetrahydrofurofuran derivative from Aegilops ovata (Gramineae) showed inhibitory activity on achene germination of Lactuca sativa (Compositae) in white light [2]. Various podophyllotoxin derivatives reduce the root growth oiBrassica campestris (Cruciferae). This effect has been ascribed to the presence of the methylenedioxy group and the configuration of the lactone ring. However, the activity of matairesinol was enhanced by methylation of the two phenolic groups [7]. CHEMISTRY Biosynthesis The biosynthesis of lignans is a rather unexplored area. Experimental evidence of all the postulated logical biosynthetic sequences has still not been found. Research has focused on enzymatic transformations occurring in lignan biosynthesis, with particular emphasis on the stereoselectivity and enantiospecificity of such conversions. Lignans arise by oxidative coupling of two coniferyl alcohol units. It has been accepted that a monolignol such as coniferyl alcohol is dehydrogenated to phenoxy radicals, and coupling of these radicals or mesomeric forms of them leads to quinone methide structures, which suffer nucleophil additions to give dilignols (lignans and neolignans). It has been demonstrated that two achiral molecules of ^-coniferyl alcohol suffer direct stereoselective coupling to give (+)-pinoresinol, which undergoes a highly enantiospecific NADPH-dependent reduction that sequentially yields (+)lariciresinol and (-)-secoisolariciresinol, the precursor of (-)-matairesinol [28,29]. 0-Methylation reactions of (+)- and (-)-matairesinol catalysed by Forsythia intermedia (Oleaceae) cell-free extracts, are neither highly enantioselective nor regiospecific. Regiospecific glucosylation first occurs to yield matairesinoside, and subsequent methylation yields actiin, which is then converted into arctigenin via the action of a p-glucosidase [30]. The first natural benzylic ether reduction (or quinone methide analogue) in plant

188 systems was demonstrated in Forsythia cell-free extracts, which catalyse the conversion of pinoresinol into secoisolariciresinol via lariciresinol [29]. The biosynthesis of (+)-lariciresinol via stereospecific reduction of (+)pinoresinol has also been demonstrated in Zanthoxylum ailanthoides (Rutaceae). This lignan is an intermediate on the pathway from coniferyl alcohol to arctigenin and is formed by reductive cleavage of one of the benzylic ether linkages in (+)-pinoresinol in presence of NADPH [31]. The biosynthetic pathway to the furoforan lignans from quinone methides by stereospecific nucleophilic additions was investigated in feeding experiments with Forsythia species. Dibenzylbutyrolactones require reduction and formation of the lactone ring. Methylation and glycosylation reactions occur late in the sequence [32]. Podophyllum aryltetralin lactone lignans are produced biosynthetically by oxidative cyclization of dibenzylbutyrolactone lignans. Feeding experiments with matairesinol in three different Podophyllum systems demonstrated this dibenzylbutyrolactone to be the common precursor of the 3',4',5'trimethoxy and 4'-hydroxy-3',5'-dimethoxy groups oiPodophyllum lignans [33]. Most lignans are obtained optically active and, presumably, enantiomerically pure. However, the phenolic coupling processes catalysed by H202-dependent peroxidases, 02-requiring laccases or phenol oxidases yield racemic products. A protein without an active centre has been isolated that in presence of an oxidase produces stereoselective bimolecular phenoxy radical coupling reactions in in vitro lignan biosynthesis. Its mechanism of action is presumed to involve capture of E'-coniferyl alcoholderived free radical intermediates [34]. Two distinct classes of putative pinoresinol-lariciresinol reductase cDNA clones from Thuja plicata (Cupressaceae) that encoded a pinoresinol-lariciresinol reductase with different enantiospecificity have been reported. One of the proteins converted (+)-pinoresinol into (-)secoisolariciresinol, whereas the other utilised the opposite (-)-enantiomer to give the corresponding (+)-form [35]. Classification In the present review, we classify these metabolites following closely the classification based on structure designed in previous lignan research [5,6,7,8], but we cover only those with pharmacological activity.

189 These natural products are divided into two major groups, lignans and neolignans. Oligomeric lignans and neolignans, hybrid lignans and norlignans constitute three minor groups. The first of them comprises oligomeric phenylpropanoids. Hybrid lignans refer to a single C6-C3 unit bound to another structure and thus yielding flavanolignans, isoflavanolignans, xanthonolignans, coumarinolignans, stilbenolignans, sesqui-, di- and triterpenelignans. The lignans are classified into six subgroups based on general structures: the dibenzylbutanes and dibenzylbutanediols (formed by two P-P' linked C6-C3 units); dibenzylbutyrolactones (with two P-p' linked C6-C3 units and an additional y-y' lactonic bridge); substituted tetrahydrofurans (two P-P' linked C6-C3 units and an oxygen bridge); 2,6-diaryl-3,7dioxabicyclo[3.3.0]octanes, also called tetrahydrofiirofurans, (two p-P' linked C6-C3 units and two oxygen bridges); arylnaphthalene derivatives (two P-P' linked C6-C3 units and a C2,Ca' bond that yields a tetralin or a naphtalene); and dibenzocyclooctadiene derivatives (two P-p' linked C6-C3 units and a €2,02' bond that forms a biphenyl). The neolignans possess many different structures and are divided into more than fifteen subgroups. They include benzoftiran, dihydrobenzofuran, diarylethane, benzodioxin, alkyl aryl ether and bicyclo[3.2.1]octane derivatives among others The principal skeletons of lignans are listed in Table 1 and the chemical structures of the most relevant pharmacologically active lignans are listed in Tables 2-10.

190 Table 1.

General Skeletons of Lignans Cited in This Review

Dibenzylbutane

Arylnaphthalene

Tetrahydrofiiran

Dfcenzylbutyrolactone

Dibenzocycboctadiene

Aryltetrahydronaphthaleiie

Cycbbutane

Furo&ran

Table 2.

Active Dibenzylbutane Lignans

B

.A

Burseran 2,3-Dibenzylbutane-l,4-diol Dihvdrocubebin ,~~eso-Dihvdroguaiarelic acid (-)-3,4-Divanillyltetrahydrofuran Enterodiol Enterohran hIachilin .A Nordihydroguaiaretic acid (NDGA) Phyllantin Rhinacanthin E Rhinacanthin F -)-Secoisolariciresinol (-)-Secoisolariciresinol diglucoside 2,6,8,10-Tetra-O-rnethyl NDG.A

.A .-I .-I

B A

RI Rz -0CH:OH H -0CH:OOCH3 OH OCH3 OH OH H OH H -0CH;OOH OH OCH3 OCH3 -OCH;.O-0CH:OOCH3 OCH3 0-Glu OCH, OCH,

R3

OCHp H H H OCHs OH OH H H H OCH, OCH3

C R4

Rs

OCH3 OCH3 H H -0CH:OOH OCH3 OH H H H H H -0CH:OOH OH OCH3 OCHS -0CH:O-0CHzOOH H 0-Gh H OCH, OCH,

R-

Rs -CH:OCHz-

OH OH CHtOH CH:OH CH3 CHI -CH:OCHy CHzOH CHPH -CHzOCH:CHI CH3 CH3 CH3 CHiOCHt CH20CH3 COOCH3 COOCH, COOCHj COOCH3 CH-OH CH:OH CH:OH CH:OH CH3 CH3

-A

.A

B C

.A

1’ X B .A,B .A .A

B

OCH3 H

*

2

192

Table 3.

Active Dibenzylbutyrolactone Lignans

XN

r^^^

R5 R4

R4

Arctigenin Arctigenin methyl ether (-)-Arctiin Bursehemin 4'-Demethylyatein (-)-Deoxypodorhizone Enterolactone Epipodorhizol (-)-Hinokinin (-)-6'-Hydroxyyatein (-)Matairesinol (-)Matairesinol dimethyl ether Nortrachelogenin Podorhizol Prestegane B Trachelogenin (-)-rra«5-2-(3",4",5"-trimethoxybenzyl)3-(3\4'-ethylenedioxybenzyl)butyrolactone (-)-traAw-2-(3",4"-dimethoxybenzyl)-3-(3' 4' -methylenedioxybenzyl) butyrolactone Wikstromol (-)-Yatein

A A A A A B B A A B B B

Ri R2 OCH3 OH OCH3 OCH3 OCH3 0-Glu OCH3 OCH3 -OCH2O. -OCH2OOH H -OCH2O-OCH2O-OCH2OOCH3 OH OCH3 OCH3 OH OCH3 -OCH2OOH OCH3 OH OCH3 -OCH2O-

OCH3 OCH3 OH OCH3 H OCH3 OCH3 OCH3 OCH3 OCH3 OH OCH3 OCH3

R4 Rs H OCH3 H OCH3 OCH3 OCH3 -OCH2OOH OCH3 OCH3 OCH3 H H OCH3 OCH3 -OCH2OOCH3 OCH3 OH H H OCH3 OH H OCH3 OCH3 H OCH3 H OCH3 OCH3 OCH3

B

-OCH2O-

OCH3

OCH3

B A

OCH3 OH -OCH2O-

H OCH3

OH OCH3

B B B B A

R.. OCH3 OCH3

H H

R7

R6 H H H H H H H H H OH H H H H H OH H

H H H H H H H aOH H H H H H pOH H H H

R8 H H H H H H H H H H H H OH H H H H

H

H

H

H

OCH3 OCH3

OH H

H H

H H

193 Table 4.

Active Aryltetrahydronaphtalene Lignans ?5

Ri-

R2

R5

R^_

R2

0CH3

Austrobailignan-1 p-Apopi cropodophyll in 4 '-Demethyl deoxypodophy 11 otoxin 4'-Demethylepipodophyllotoxin 4'-DemethyIpodophyllotoxin Deoxypi cropodophyll in Deoxypodophyllotoxin (anthricin) 4-Epipodophyllotoxin Isopicropodophyllone Isodeoxypodophyllotoxin 4-0-Methylpicropodophyllotoxin 5-Methoxypodophyll otoxin 5-Methoxypodophyllotoxin-4-P-D-glu Morelensin a-Peltatin p-Peltatin p-Peltatin A methyl ether Picropodophyllone Picropodophyllotoxin Podophyll otoxin Po^ophyllotoxin-p-D-glucoside

A A A A A B A A D C B A A A A A A B B A A

Ri H H(A') H H OH H H H

=o

R2 H H H OH H H H OH

H H H OCH3 OH H 0-Glu H H H H H H H H H =0 OH H OH H OGlu H

R3 R4 -OCH2OOCH3 OCH3 OH OCH3 OH OCH3 OH OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 H OCH3 OH OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3

R5 H H H H H H H H H H H OCH3 OCH3 H OH OH OCH3 H OCH3 H H

194

Table 5.

Active Arylnaphtalene Lignans

R3 Ri R2 A -OCH2OH A -OCH2OH A -OCH2OOCH3 A -OCH2OH A -OCH2OH -OCH2OH A Deoxydehydropodophyllotoxin A OCH3 OCH3 Diphyllin H A OCH3 OCH3 H Diphyllin acetate A OCH3 OCH3 H Diphyllin-0-apioside Diphyllin-O-apioside-5-acetate A OCH3 OCH3 H A 0CH3 0CH3 Diphyllinin H A OCH3 0CH3 Diphyllinin monoacetate H A 0CH3 OCH3 Diphyllinin crotonate H B Helioxanthin H -OCH2OA OCH3 0CH3 Justicidin A H A OCH3 OCH3 Justicidin B H B H -OCH2OJusticidin D (Neojusticin A) H -OCH2OB Justicidin E (Taiwanin C) 5-M ethoxy dehydropodophyllotoxin -OCH2OA OCH3 B Neojusticin B H OCH3 OCH3 Phyllamyricin E A OCH3 OCH3 H B Retrojusticidin B H OCH3 OCH3 -OCH2OH Taiwanin E A A -OCH2OH Taiwanin E methyl ether H Tuberculatin A OCH3 OCH3 c a =-O-apiose-S-acetate; b =-OC4H40(OH)2CH20H; d =-OC4H40(OH)2CH20COCHCHCH3 Chinensinaphthol Chinensinaphthol methyl ether Dehydro-p-peltatin methyl ether Dehydropodophyllotoxin 4'-(9-Demethyldehydropodophyllotoxin

R6 R? R8 R5 R4 H H OH OCH3 OCH3 H H OCH3 OCH3 OCH3 H H OCH3 OCH3 OCH3 H OH OCH3 OCH3 OCH3 OH H OCH3 OH OCH3 H H OCH3 OCH3 OCH3 H -OCH2OH OH H OAc -OCH2OH H -OCH2OO-api H -OCH2OH H a b H -OCH2OH c -OCH2OH H d H -OCH2OH H -OCH2OH H -OCH2OH H OCH3 -OCH2OH H H H -OCH2OH OCH3 H -OCH2OH H H OH OCH3 OCH3 OCH3 H -OCH2OH OCH3 H -OCH2OH OCH3 H -OCH2OH H -OCH2OH H OH H -OCH2OH OCH3 H -OCH2OO-api H =-OC4H40(OH)2CH20COCH2

Table 6.1.

Active Aryltetralin Lignans

RI (+)-Dimethylisolariciresinol-2-a-~lose H>pophyllantin lsolariciresinol (+)-Lvoniresinol3a-O-P-D-glucop\ranoside hlethyl9-deo~-9-oso-a-apopicropodoph~llate Nirtetralin Nudiposide 2,2,2-Trifluoroethylhydrazoneof methyl -9-deosy-9-0x0-aapopicropodophvllate Phenylhydrazone of methyl 9-deo!q+OXO-~~apopicropodophvllate Podophyllic acid (+)-Tsugacetal

.\

B B

B .A B A .A

R2

R3

OCH3 OH H OCH3 -0CHzOOCH, OH H OCH3 OH OCH, -0CH20H -0CH-0OCH3 OCHl OH OCH3 -0CH20H

Rs

R6

R4 H H H OCH, OCHJ H OCH3 OCH,

OH CH3 OH OH OCH, CHI OH OCH3

CH2O-Xy1 CH2OCH3 CHzOH aCH20glu PCOOCH, CHzOCHi PCHZOXYI PCOOCH, PCOOCH,

.A

-0CHzO-

H

OCH3

OCH3

.A B

-OCH20OCH3 OH

H H

OCH3 H

OCHa OH

RCHzOH CHzOCH, CHzOH PCHzOH CHO CHzOCH, aCH2OH CH=N-NHCH:CF, CH=N-NH-Ph

aCOOH BCH20H -CH2OCH(POCH,)-

R8

H H H H

HJH H HJ’ H.1aOH

H

196 Table 6.2.

Active Aryltetralin Lignans H3C0.

H2C0

2,3-Bis(methoxycarbonyl)-l -(3,4dimethoxyphenyl)-4-hydroxy6,7,8-trimethoxynaphthalene Phyllamycin B Phyllamyricoside A Phyllanthostatin A

Table 7.1.

Ri

R2

R3

R4

R5

R6

R7

OH

CO2CH3

CO2CH3

OCH3

OCH3

H

OCH3

H 0-GIu H

CH2OCH3 CH2OH CH2OCOCH3

CHO CH3 COOGlu

H

-OCH2OH -OCH2C>H -OCH2O-

Active Dibenzocyclooctadiene Lignans

H3C0

0CH3

B R2

Ri Episteganangin

H

A 00c

CH3

Isopicrostegane Steganacin

B A

H H

Steganangin

A

H

H OAc

OOC

CH3

II H H

197 Table 7.2.

Active Dibenzocyclooctadiene Lignans CH3 R,0.

R9O

Angel oylgomi sin H

Ri

R2

CH3

CH3

R3

HjC

Benzoyl gomisin O Deoxygomisin A Gomisin A Gomisin B

R4

R5

Re

R7

Rs

R9

CH3

CH3

H

H

OH

CH3

OCOC6H5 H H

H H H

H H OH

CH3 CH3 CH3

H

CH3

OH

H H H H H H H

H H H H H H H

CH3 CH3 CH3 CH3 CH3 CH3 CH3

H

CH3

OH

H H H H

OH H OH CH3

CH3 CH3 CH3 OH

OH

CH3

H

CH3

CH3

CH3 CH3 -CH2-CH2-

CH3 CH3 CH3

-CH2CH3 CH3 CH3 CH3

-CH2-

CH3

CH3

CH3 H3C

Gomisin J Gomisin K3 Gomisin L2 Gomisin Mi Gomisin N Kadsuranin Interiotherin A

H CH3 CH3 CH3 H CH3 CH3 CH3 -CH2CH3 CH3 -CH2-

CH3 H CH3 H CH3 CH3 CH3

Interiotherin B

-CH2-

CH3

CH3 H CH3 CH3 -CH2-CH2CH3 CH3 -CH2-CH2-

CHj

H H H H H H OCOC6H5

-CH2H,C

Schizandrin Schisandrin A Schisandrol A Schisantherin D

CH3 CH3 CH3 CH3 CH3 CH3 -CH2-

CH3 CH3 CH3 CH3

CH3 CH3 CH3 CH3 CH3 CH3 -CH2-

Tigloylgomisin P

-CH2-

CH3

CH3

CH3

H

H H H OCOC6H5 H H5C

Wuweizisu C

-CH2-

CH3

-CH2-

H

CH3

H

198

Active Tetrahydrofuran Lignans

Table 8.1.

0CH3

0CH3

Grandisin L-652,731 L-659,989 L-662,025 Nectandrin B (+)-Neoolivil

C A A A B C

Ri

R2

R3

R4

OCH3 OCH3 SO2CH3

OCH3 OCH3 OCH2CH2CH3 OCH2CH=CH2 OH OH

OCH3 OCH3 OCH3 OCH3 OH OH

OCH3 OCH3 OCH3 OCH3 H H

N3

H H

R5 CH3

R6 CH3

H H H

H H H

CH3

CH3

CH2OH

CH2OH

199

Active Tetrahydrofuran Lignans

Table 8.2.

H3C0,

CH2OH

HO

R2

Ri Acuminatin Lariciresinol LariciresinoI-4p-D-glucoside

-OCH2OOH O-Glu

OCH3 OCH3

HOHjC

CH30' OCH3

(-)-Hernone 7-S,8i^,8'/?-(-)-Lariciresinol-4,4'-bis-0-|3-D-glucopyranoside Magnone A Magnone B 75,8/?,8'/?-(-)-5-Methoxylariciresinol-4,4'-bis-(9-p-D-gIu (-)-Nymphone

A B A A B A

Ri OCH3 H H H H OCH3

R2 H H H OCH3 OCH3 H

R3 R4 OCH3 OCH3 O-Glu OCH3 OCH3 OCH3 OCH3 OCH3 O-Glu OCH3 -OCH2O-

200

Table 9.1.

Active Furofuran Lignans

(+)-1 - Acetoxypinoresinol (+)-l -Acetoxypinoresinol dimethylether (+)-l -Acetoxypinoresinol-p-D-glu (-)-Eudesmin (+)-Eudesmin = (+)-Pinoresinol dimethyl ether (+)-l -Hydroxypinoresinol (+)-Isogmelinol Kobusin (+)-Magnolin (+)-Medioresinol (-)-Pinoresinol (-)-Pinoresinol-p-D-glucoside (+)-Pinoresinol (+)-Pinoresinol-p-D-glucoside (+)-Pinoresinol di-p-D-glucoside (+)-Pinoresinol monomethyl ether (-)-Prinsepiol (+)-Sesamin (-)-SyringaresinoI (-)-Syringaresinol di-p-D-glucoside (-)-Syringaresinol diacetate (+)-Syringaresinol = (+)-Lirioresinol B (+)-Syringaresinol di-p-D-glucoside = Liriodendrin (+)-Syringaresinol dimethyl ether = (+)-Yangambin

B B B A B

Ri OCH3 OCH3 OCH3 OCH3 OCH3

R2 OH OCH3 OH OCH3 OCH3

R3 H H H H H

R4 OCH3 OCH3 OCH3 OCH3 OCH3

R5 OH OCH3 O-Glu OCH3 OCH3

R6 H H H H H

R7 OAc OAc OAc H H

Rs H H H H H

H H OCH3 H H H H H H H H H H OCH3 OCH3 OCH3 OCH3

OH OCH3 OCH3 OCH3 -OCH2OOCH3 OCH3 OH OCH3 OH OCH3 OH OCH3 OH OCH3 OH OCH3 OCH3 O-Glu OH OCH3 OH OCH3 -OCH2OOH OCH3 OCH3 O-Glu OAc OCH3 OH OCH3

H H H OCH3 OCH3 H H H H H H H H OCH3 OCH3 OCH3 OCH3

OH

H

A A A B

OH OCH3 0CH3 OCH3 H OCH3 OCH3 OCH3 OH OCH3 OH OCH3 OCH3 O-Glu OH OCH3 OCH3 O-Glu OCH3 O-Glu OCH3 OCH3 OH OCH3 -OCH2OOH OCH3 OCH3 O-Glu OAc OCH3 OH OCH3

OH H H H H H H H H H OH H H H H H

H H H H H H H H H H OH H H H H H

B

OCH3

O-Glu

OCH3

OCH3

O-Glu

OCH3

H

H

B

OCH3

OCH3

OCH3

OCH3

OCH3

OCH3

H

H

B B B B B A A B B B B B

B

201

Table 9.2.

Active Furofuran Lignans

Hi'

(-)-Asarinin = Episesamin (+)-Epiaschaiitin (-)-Epieudesmin (+)-Epimagnolin (+)-Epipinoresinol (+)-Epipinoresinol dimethyl ether (+)-Epiyanganibin (+)-Fargesin Isomagnolin (-)-Phylligenin Phillyrin

A C B C A A C C A B A

Ri R2 -OCH2O-OCH2OOCH3 OCH3 OCH3 OCH3 OCH3 OH OCH3 OCH3 OCH3 OCH3 -OCH2OOCH3 OCH3 OCH3 OH OCH3 OCH3

R3 H H H H H H OCH3 H H H H

R5 R4 -OCH2OOCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OH OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 O-Glu H

R6 H OCH3 H OCH3 H H OCH3 H OCH3 H OCH3

202 Table 10.1.

Active Benzofuran Neolignans

Conocarpan Dehydroconiferyl alcohol 3 ',4-(9-Dimethylcedrusin Licarin A Licarin B Licarin D Obovatifol Perseal C

A B A B B B B B

Ri CH=CHCH3 CH=CHCH20H CH2CH2CH2OH CH=CHCH3 CH=CHCH3 CH=CHCH3 CH=CHCH3 CHO

R2 H OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 OCH3

R3 R4 H OH OCH3 OH OCH3 OCH3 OCH3 OH -OCH2OOCH3 OCH3 OH OH -OCH2O-

R5 H H H H H H OCH3 H

OH

Eupomatenoid-5 Obovaten Perseal D

Ri CH=CHCH3 CH=CHCH3 CHO

R2 H OCH3 OCH3

R3 OCH3 OH OH

R4 H OCH3 OCH3

R6 CH3 CH2OH CH2OH CH3 CH3 CH3 CH3 CH2OH

203

Table 10.2.

Active Benzofuran Neolignans OCH

HsCO-

4'-Demethoxy-3',4'-methyIenedioxyniethylrocaglate 4'-Demethoxy-3' ,4'-methylenedioxyrocaglaol l-0-Formyl-4'-demethoxy-3',4'-methylenedioxymethylrocaglate Methylrocaglate 1 -Oxo-4'-demethoxy-3 ',4'-methylenedioxyrocaglaol

Ri

R2

R3

OH OH

H H H H

COOCH3

OCHO

OH

=o

H COOCH3 COOCH3

H

R4

R5

-OCH2O-OCH2O-OCH2OOCH3 H -OCH2O-

Isolation and identification Research is increasingly focused on developing adequate techniques for analysing lignans from plant sources and physiological fluids. Four reviews contain a section dedicated to analysing and identifying lignan mixtures [58]. The efficiency and selectivity of supercritical fluid extraction (SFE) make this method an alternative to conventional extraction of lignans with organic solvents. Supercritical carbon dioxide was used to extract lignans from Schisandra chinensis (Schisandraceae) fruit, seeds and leaves [36,37]. Choi et al [38] have established the optimal SFE conditions of temperature and pressure for the main bioactive lignans schisandrol A, schisandrol B, schisandrin A, schisandrin B, and schisandrin C. The bioactive neolignans from Magnolia virginiana (Magnoliaceae) flowers were extracted sequentially with supercritical CO2 alone or modified with solvents, and

204

were quantified by high performance liquid chromatography (HPLC) using photodiode array detection (DAD) [39]. Magnolol and honokiol were extracted from Magnoliae cortex using supercritical-fluid chromatography (SFC) on-line coupled with SFE by an amino column trapping [40]. Wolf ^/ al [41] resolved the atropisomers of arylnaphthalene lignans by cryogenic subcritical fluid chromatography. Both normal phase and reverse phase HPLC with an UV detector or in combination with DAD were found to be rapid, reliable methods for qualitative and quantitative analysis of lignans. Quantification of the antioxidant lignan glucosides sesamin, sesamolin, sesaminol and sesamolinol in some varieties of sesame was achieved by HPLC determination [42]. On-flow liquid chromatography/proton nuclear magnetic resonance (LC/^H-NMR) and LC/UV/MS analysis was performed to identify the antioxidant lignans (-)-phylligenin, (-)-eudesmin and (-)epieudesmin from Orophea enneandra (Annonaceae) [43]. LC/mass spectrometry (MS), fast-atom bombardment (FAB)-MS and FAB-tandem MS techniques were applied to the rapid detection of lignans with radical scavenging activity from Krameria triandra (Krameriaceae) roots [44]. The isolation of the lignans eudesmin, magnolin, yangambin and kobusin from Polygala gazensis (Polygalaceae) and subsequent analysis by thermospray LC-MS and HPLC-DAD of two other related Polygala species have been reported [45]. This analysis was also applied successfully to the identification of lignan constituents from Schisandra chinensis fruit [46]. HPLC coupled with either ionspray or continuous-flow FAB-MS or gas chromatography (GC)-MS using single ion monitoring allowed rapid detection and identification of secoisolariciresinol diglycoside in flaxseed. Two diastereoisomers of secoisolariciresinol diglycoside were detected by this method [47,48]. The application of a solid-state fluorodensitometric technique after thin layer chromatography allowed highly reproducible, sensitive quantification of the toxic constituents of Cleistanthus collinus (Euphorbiaceae), the arylnaphthalene lignan lactone diphyllin, and its glycosides cleistanthin A and cleistanthin B in samples of forensic and clinical value [49,50]. Cleistanthin A was determined by an enzyme-linked immunoassay in cases of C. collinus poisoning [51]. Analytical high-speed countercurrent chromatography (CCC) has also been used to study lignans [52]. Interfacing CCC coupled with thermospray MS provides an analytical methodology for identifying bioactive lignans

205 from Schisandra rubriflora [53]. Anti-HIV lignans from Larrea tridentata were isolated by assay-guided CCC and their chemical structures determined by GC-MS and NMR [54]. More recently Ma et al [55] purified these compounds by pH-zone-refming CCC. LC-MS and GC-MS have been used to analyse samples containing the mammalian lignans enterolactone and enterodiol and their precursors matairesinol and secoisolariciresinol. Fotsis et al [56] described a capillary GC method for quantifying lignans in human urine after appropriate derivatization. The plant lignans lariciresinol, isolariciresinol and secoisolariciresinol were identified by GC/MS in human urine [57]. A quantitative method based on ion-exchange chromatography and isotope dilution GC-MS has been described for the determination of plasma lignans [58]. Secoisolariciresinol and matairesinol in plant-derived foods were measured by an isotope dilution GC-MS method [59] that was also applied to quantify enterolactone and enterodiol in human and animal sources [6062]. Schisandrin from Schisandra fruits has been determined in human plasma by selected-ion monitoring with GC-MS using a ftised-silica capillary column [63]. Atkinson et al [64] presented a GC coupled with an ion mobility spectrometric (IMS) method that makes it possible to quantify selected mammalian lignans in biological fluids. The need for selective, sensitive analytical techniques to measure these metabolites in human plasma, tissue and urine has led to the development of an HPLC with coulometric array detection, which uses a series of flow-through electrochemical sensors, each providing 100% electrolytic efficiency, with satisfactory resolution [65]. A method based on time-resolved fluoroimmunoassay using a europium chelate as a label has been described to measure plasma enterolactone [66]. Yoo and Porter [67] developed a sensitive immunoassay for the natural Podophyllum lignan, podophyllotoxin. Reverse-phase KPLC separation followed by radioimmunoassay (RIA) for human plasma samples and immunoaflfinity extraction (lAE) followed by RIA for human urine samples were used to investigate a lignan-related hypocholesterolemic agent. This analysis proved to be very usefiil in quantifying extremely low drug concentrations in body fluids [68].

206 PARMACOLOGICAL ACTIVITIES Introduction The principal pharmacological activity of lignans can be summarised in the following relevant groups. The lignans play an interesting role as phytoestrogens in mammals, including humans; they have anticancer and antiviral properties; they modify the cardiovascular function by different mechanisms including the modification of phosphodiesterase activity or platelet activating factor function; they have antioxidant properties and can modify the liver function; and some of them modify the activity of enzymes and mediators implicated in inflammation and immunity process. Lignans as phytoestrogens Phytoestrogens are a group of naturally occurring diphenolic compounds present in legumes, whole grains, fruits and vegetables that exert estrogenic effects on the central nervous system (CNS). The physiological effects are similar to those of endogenous estrogens mediated by binding to the estrogen-receptor (ER), followed by gene activation and its specific generation of products and physiological effects, like induction of estrus and stimulation of the female genital tract [69]. High consumption of phytoestrogen-rich foods has been linked to a reduced incidence of estrogen hormone-dependent cancers. Phytoestrogens were described for the first time by Setchell, Axelson et al [70,71], and their effects have been widely studied by Adlercreutz et al since 1980, and reviewed by them [72-77] and by Potter and Steinmetz [78], Knight et al [79], Kurzer and Xu [69], Rickard and Thompson [80], Mazur and Adlercreutz [81] and Tham et al [82]. They described three main classes of phytoestrogens, one of which are lignans. These compounds are not natural products present in plants but are produced by bacterial metabolism of dietary components in the animal and human gut. Matairesinol, secoisolariciresinol and its diglycoside are the major precursors of the mammalian lignans enterodiol and enterolactone and they are present in oilseeds, such as Linum usitatissimum (Linaceae) seeds (flaxseed), soybeans, sunflower seeds and peanuts but are also present in foods such as whole dried legumes (lentils, kidney, navy and pinto beans), whole grain cereals (triticale, wheat, oats, brown rice, corn, rye, barley),

207

cereal brans (oat, wheat and rice bran), vegetables (garlic, asparagus, carrots, sweet potatoes, broccoli, ...) and fruit (pears, plums, ...). In fact, enterolactone excretion may be an indicator of consumption of grains and legumes. In the case of glycosides, metabolism occurs after a previous hydrolysis of the sugar moiety by the gastric juice and p-glucosidases from foods or bacteria gut. After absorption in the small intestine, lignans are conjugated in the liver and eliminated in the kidney by urine and in liver by bile, after which they undergo enterohepatic circulation with a later reabsorption after new bacterial hydrolysis. In humans, the principal lignans identified are enterolactone, enterodiol and matairesinol in serum, semen, urine, faeces and bile. In addition, they have been identified in cow's milk. The values of excretal lignans increase 3- to 285-fold depending on the organic fluid. However, the urinary content of enterodiol and enterolactone originating from secoisolariciresinol diglycoside-fed rats represented only twenty per cent of the levels of flaxseed-fed rats. This indicates the presence of other precursors or the incomplete conversion of secoisolariciresinol diglycoside to enterodiol and enterolactone [83]. The most relevant physiological effects described for the phytoestrogens are: a) effects on human reproductive hormones, including sexual differentiation, fertility, binding to the estrogen receptor, synthesis and availability of estrogens, and effects on grov^h of estrogen dependent cells; b) tumor cell differentiation and mitogenesis, focusing on protein tyrosine kinase activity and DNA topoisomerase; c) angiogenesis; and d) antioxidation. In addition, other physiological processes may be modified, such as menstrual cycles and menopausal symptoms. On the other hand, the pathological effects are principally those related to sexual hormone dependent diseases, such as polycystic ovary syndrome, osteoporosis, sex organ cancer, and cardiovascular disease as well as menopausal symptoms. In the case of cardiovascular disease, the phytoestrogens decrease lowdensity lipoprotein (LDL) cholesterol and increase high-density lipoprotein (HDL) cholesterol. Gansser et al [84] described a test system designed to search for compounds that interfere with human sex hormone-binding globulin (SHBG) even in complex plant extracts, and report that the lignan secoisolariciresinol reduced the binding activity of human SHBG. After studying the effects of a short-term phytoestrogen-rich diet on menopausal symptoms and serum levels of phytoestrogens and SHBG, Brzezinski et al [85] concluded that a 12-week partial substitution of an omnivorous diet for

208 phytoestrogen-rich food increases the SHBG semm levels in postmenopausal women, and may alleviate symptoms such as hot flashes and vaginal dryness. However, the long-term effects on bone density and heart disease remain to be determined. Except (-)-pinoresinol, all the compounds tested by Schottner et al [86] had binding affinity to human SHBG in the in vitro assay. Whereas (±)enterodiol, (+)-neoolivil, (-)-isolariciresinol and dehydrodiconiferyl alcohol developed only weak affinities, (-)-secoisolariciresinol and enterolactone were moderate, enterofuran showed higher affinity (73% displacement of specific binding), and the affinity of (-)-3,4-divanillyltetrahydrofuran was the highest of all, with 95% ["^H]-dihydrotestosterone displacement. As lignans may displace active steroid hormones from the SHBG binding site, they may influence the blood level of these hormones and have potential beneficial effects on benign prostatic hyperplasia. In a complementary study, the same authors [87] investigated, in the same experimental system, a wide variety of synthetic lignans derived from the most active one, (-)3,4-divanillyltetrahydrofuran, with different substitution patterns in the aromatic and aliphatic part of the molecule. They concluded that (±)diastereoisomers are more active than meso compounds, that the 3methoxy-4-hydroxy substitution pattern in the aromatic part is the most effective, and that activity increased with the decrease in polarity of the aliphatic part of the molecule. Finally, they reported the effects of the previously studied natural lignans cited above, and also NDGA, on the binding of [^H]-5a-dihydrotestosterone to SHBG [88]. The lignan (-)-3,4divanillyltetra-hydrofiiran showed the highest binding affinity again {Kdi = 3.2 luiM). The results suggest a dose dependent competitive inhibition of the SHBG-dihydrotestosterone interaction. When hydrophobicity increases in the aliphatic part of the butane-1,4-diol-butanolide-tetrahydrofuran lignans, these compounds have higher binding affinity, whereas the 3-methoxy-4hydroxy substitution in the aromatic ring is the most effective for binding to SHBG. Enterolactone had been described as a moderate competitive inhibitor of human estrogen synthetase (aromatase) and it binds to or near the substrate region of the active site of the P-450 enzyme [61]. In another study [89], seven lignans were evaluated for their abilities to inhibit aromatase enzyme activity in a human preadipose cell culture system. The lignan enterolactone and its precursors, didemethoxymatairesinol and 3'-demethoxy-3-0demethylmatairesinol, decreased aromatase enzyme activity, with K\ values

209 of 14.4, 7.3 and 5.0 |iM, respectively, whereas the aromatase inhibitor aminoglutethimide had a K\ value of 0.5 |LIM. On the other hand, the lignan enterodiol and its precursors, 0-demethylsecoisolariciresinol, demethoxysecoisolariciresinol, and didemethyl-secoisolariciresinol, were less active. The inhibition of human preadipocyte aromatase activity by lignans suggests a mechanism by which consumption of lignan-rich plant foods may contribute to a reduction in estrogen-dependent disease. In addition, the endogenous lignans are inhibitors of 5a-reductase. This can justify the effects on the development of prostatic carcinoma and the inhibition of other growth-promoting steroid hormones. In fact, it was reported [90] that 5a-reductase and 17P-hydroxysteroid dehydrogenase are inhibited by enterodiol and enterolactone in human genital skin fibroblast monolayers and homogenates, in benign prostatic hyperplasia tissue homogenates, and in genital skin fibroblast monolayers, but enterolactone was the most potent inhibitor. To elucidate whether enterolactone has antiestrogenic properties Mousavi and Adlercreutz [91] studied the w vitro effect of relatively low concentrations of enterolactone (0.5-2 |Limol/L) added both alone and in combination with estradiol (1 nmol/L), using MCF-7 breast cancer cells in culture. By itself, enterolactone stimulated the proliferation of MCF-7 cells, but in combination with estradiol it either resulted in lower stimulation or had no effect at all as compared with the control. The concentrations of enterolactone needed for the elimination of the proliferative effect of estradiol are physiological and similar to those used in experiments with the antiestrogen tamoxifen. Recently [92], the antiestrogenic effects of the enterolactone precursor secoisolariciresinol diglycoside was compared with tamoxifen by monitoring rat estrus cycling, and it proved to be antiestrogenic but without having gross tissue toxicity. Phytoestrogens may modulate the activity of the human sex steroid binding protein and so influence the role of this protein in the delivery of hormonal information to sex steroid-dependent cells, as reported by Martin et al [93] in research with NDGA, enterolactone and enterodiol. These phytoestrogens had different dose-dependent inhibitory effects on the steroid binding protein. It is not clear whether there is a positive association between the rate of bone loss in postmenopausal related osteoporosis and the enterolactone level in human [94]. Experimental results on thirty-two selected women with an annual rate of radial bone loss did not demonstrate that a low.

210 unsupplemented dietary intake of phytoestrogens has a preventive effect on postmenopausal cortical bone loss. However, no conclusions can be drawn about effects of higher doses of phytoestrogens. Tou et al [95] studied the effects of the mammalian lignan precursor secoisolariciresinol diglycoside as an estrogen agonist or antagonist, and examined whether feeding flaxseed or lignan to rats during a hormonesensitive period can have effects on reproduction. The results demonstrated that flaxseed had no effect on pregnancy outcome, as compared with other treatments. The female offspring had shortened anogenital distance, higher uterine and ovarian relative weights, began puberty at an earlier age and weighed less at puberty, had lengthened estrus cycle and persistent estrus. The male offspring had reduced postnatal weight gain, greater sex gland and prostate relative weights, suggesting estrogenic effects. On the other hand, when compared with the basal diet, lignans reduced immature ovarian relative weight by 29%, delayed puberty, and tended to lengthen diestrus, indicating an antiestrogenic effect. Because flaxseed affects the reproductive development of offspring, caution is suggested when consuming flaxseed during pregnancy and lactation. Anticancer and cytotoxic effects The resin of Podophyllum, called podophyllin, has been known for its antitumoral properties since the end of the last century. Podophyllotoxin was isolated from it, and during the present century its pharmacological properties have been established. Its gastrointestinal toxicity, antimitotic properties, binding to tubulin and relationships with colchicine have been studied and developed over the last few years. Other closely related compounds were isolated from species of Podophyllum, such as the podophyllotoxin-P-D-glucoside and 4'-demethylpodophyllotoxin, 4'-demethylepipodophyllotoxin, a-peltatin and P-peltatin. In addition, some semisynthetic derivatives of 4-epipodophyllotoxin, such as etoposide and teniposide, are used in therapy. In 1998, Imbert [15] reviewed the history and isolation of the naturally occurring podophyllotoxins, hemisynthesis and use in cancer chemotherapy of this interesting group of lignans. He compiled all the information on the mechanism of action and cited the possibility of obtaining new hemisynthetic derivatives with different mechanisms of action to treat cancer. The podophyllotoxins are inhibitors of microtubule assembly with a capacity to inhibit topoisomerase II by

211

inducing double-stranded breaking of DNA molecules. The structural features responsible for this activity were broadly determined. Two hundred and seventy-four natural and semisynthetic derivatives of podophyllotoxins are reviewed by Damayanthi and Lown [16], including synthesis, pharmacological activity, clinical studies, mechanism of action, structureactivity relationships and data on cytotoxicity, inhibition of DNA topoisomerase II activity and cellular protein-DNA complex formation of many of them. Recently, Kinghorn et al [96] reviewed the anticancer agents discovered by activity-guided fractionation and they include seventy compounds from twenty-two plants. Only one lignan is cited in the review; guaiacylglycerolp-0-6'-(2-methoxy)cinnamyl alcohol ether isolated from Brucea javanica (Simaroubaceae) exhibited weak activity against the induction of cell differentiation of human promyelocytic leukemia (HL-60) cells [96,97].

0CH3

H3C0

Etoposide (R=CH3) Teniposide (R=tienyl)

Guaiacylglycerol-p-0-6'-(2-methoxy)cinnamyl alcohol ether

All the antitumor and antimitotic lignans discovered before 1984 have been compiled by MacRae and Towers [9]: NDGA, (+)-dimethylisolariciresinol-2a-xyloside, burseran, podophyllotoxins and peltatins, austrobailignan-1, diphyllin and its acetate, justicin-D, diphyllinin and its derivatives, stegnacin, stegnangin, stegnanol, stegnanone, (-)-/ram'-2(3",4"-dimethoxybenzyl)-3-(3',4'-methylenedioxybenzyl)butyrolactone and the 3",4",5"-trimethoxy derivative. Other interesting compounds, however, are not included in it, such as phyllanthostatin A (ED50 4 |Ltg/mL) isolated from Phyllanthus acuminatus (Euphorbiaceae) [98]; megaphone acetate.

212 dysodanthin A and dysodanthin B from Endlicheria dysodantha (Lauraceae), which showed activities in the brine shrimp lethality test, inhibited the growth of crown gall tumors on potato discs and were cytotoxic to human tumor cells in culture [99]; 3',4-(9-dimethylcedrusin from Croton spp. (Euphorbiaceae), which did not stimulate cell proliferation, but rather inhibited thymidine incorporation while protecting cells against degradation in a starvation medium [100]; schiarisanrin C from Schisandra arisanensis, which showed cytotoxic activity against tumor cells, with ED50 values between 0.36-7.1 |Lig/mL [101]; (-)syringaresinol from Annona montana (Annonaceae) and its acetate, which showed significant cytotoxicity (ED50 values of 0.67 and 3.78 M-g/mL, respectively) against P-388 cells [102]; brevitaxin, the first terpenolignan isolated from Taxus brevifolia (Taxaceae), which showed a selective cytotoxicity against the prostate cancer cells, with an ED50 value of 6.8 |LIM [103]; asarinin and xanthoxylol from Asiasarum heterotropoides var. mandshuricum (Aristolochiaceae), which exhibited remarkable inhibitory effects on a two-stage carcinogenesis test of mouse skin and pulmonary tumors [104]; (-)-6'-hydroxyyatein, (-)-hernone and (-)-nymphone isolated from Hernandia nymphaeifolia (Hernandiaceae), which exhibited cytotoxic activities against four cell lines with ED50 values were < 4 |ig/mL [105]. The in vitro cytotoxicity of 5-methoxypodophyllotoxin, obtained from a root culture derived from Linum flavum, against EAT and HeLa cells was determined and compared with those of podophyllotoxin, etoposide, teniposide and 5-methoxypodophyllotoxin-4-3-D-glucoside [106]. The tested lignans had about the same cytotoxic potency as podophyllotoxin (ED50 of 32 and 22 |Lig/mL versus 42.8 and 20.5 M^g/mL, respectively against EAT and HeLa cells). However, in comparison with etoposide and teniposide they were clearly less potent (1.1 and 7.9 |Lig/mL, and 0.06 and 0.3 iLig/mL, respectively). Nine lignans were isolated from Hyptis verticillata (Lamiaceae) by bioactivity directed fractionation using the brine shrimp lethality test [107]. 5-Methoxydehydropodophyllotoxin (P-388, ED50 4 |ig/mL) and the dehydro-P-peltatin methyl ether (P-388, ED50 18 |Lig/mL) were reported for the first time as isolated natural products. All the isolated compounds were evaluated against a panel of cell lines comprising a number of human cancer cell types (breast, colon, fibrosarcoma, lung, prostate, KB, and KBVI) and murine lymphocytic leukemia (P-388). 5-Methoxydehydropodophyllotoxin, dehydro-P-peltatin methyl ether, dehydropodophyllotoxin and

213 deoxydehydropodophyllotoxin showed marginal cytotoxic activity against the human cell lines tested. In contrast, (-)-yatein, 4'-demethyldeoxypodophyllotoxin, isodeoxypodophyllotoxin, deoxypicropodophyllin and Papopicropodophyllin demonstrated a general nonspecific activity comparable to that of podophyllotoxin (ED50 < 10 ng/mL).

0CH3 H3C0, '''0CH3

H3C0 OCH3

0CH3

Megaphone acetate

Dysodanthin A(Ri+R2=-OCH20-) Dysodanthin B (Ri=R2=OCH3)

Brevitaxin

Etoposide is the most potent lignan knov^n in chemotherapy. However, it is often applied to treat tumor cells that have developed resistance to previously used chemotherapeutic agents. It is still limited in its applications and common problems found in chemotherapy such as multidrug resistance and toxicity problems are also apparent [108]. Middel et al [109] synthesised eleven lignans and studied their cytotoxicities in a human small cell lung carcinoma cell line (GLC4) using the microculture tetrazolium assay. Furthermore, five naturally-occurring podophyllotoxin related compounds were tested and were compared with the clinically applied anticancer agents etoposide, teniposide and cisplatin. Most of the compounds showed moderate to high activities against GLC4, and two of

214

them containing a menthyloxy group showed activities comparable to the reference cytotoxic agents. Podophyllotoxin, 4'-demethylpodophyllotoxin, 5-methoxypodophyllotoxin, and compound-1 showed IC50 values in the same range as the reference drugs etoposide and teniposide (0.03-0.83 JLIM), except in the case of deoxypodophyllotoxin, which had an IC50 of 8 nM. A large number of cyclolignans isolated from Juniperus sahina (Cupressaceae) leaves and some semisynthetic derivatives were evaluated for their antineoplastic activity against P-388 murine leukemia, A-549 human lung carcinoma, and HT-29 colon carcinoma [110,111]. Some of them were active in both types of assays at concentrations below 1 |Lig/mL; deoxypodophyllotoxin (2.5-4 ng/mL) and (3-peltatin A methyl ether (4 ng/mL) were the most potent compounds in all cases, for the three neoplastic systems, and /raw5-tetralinelactones were clearly more potent than the c/^-form or non-lactonic compounds. In a complementary study, the same authors synthesised a series of fused pyrazole derivatives of cyclolignans, evaluated them for their cytotoxic activity against the same cell lines, and showed similar ranges of potency [112]. In addition, several cyclolignans lacking the lactone moiety were prepared from podophyllotoxin and deoxypodophyllotoxin by simple chemical transformations, and their cytotoxicity was studied. Most of the compounds show similar effects in all the. neoplastic systems tested, except methyl 9-deoxy-9-oxo-aapopicropodophyllate (IC50 12 ng/mL), 2,2,2-trifluoroethyl-hydrazone of methyl 9-deoxy-9-oxo-a-apopicropodophyllate (IC50 48 ng/mL), and phenylhydrazone of methyl 9-deoxy-9-oxo-a-apopicropodophyllate (IC50 39 ng/mL), which showed highly selective cytotoxicity towards HT-29 human colon carcinoma [113]. On the other hand, 4*-demethyldeoxypodophyllotoxin, P-apopicropodophyllin, and the isoxazopodophyllic acid demonstrated immunosuppressive activity [114]. The same group [115] synthesised a new type of cyclolignans with an isoxazoline ring fused to the cyclolignan core from 7-ketolignanolides. The eleven compounds prepared were evaluated for their cytotoxic activities on four cell lines and showed an IC50 range of 23-0.2 |iiM, clearly lower than the reference drug tested, podophyllotoxin (IC50 12 nM). Only one synthetic lignan, 7-oxime, obtained from podophyllotoxone, gave an IC50 of 0.2 |LIM.

215 OCHj

O

O Menthyl

OCH3

Compound-1 =((37?,4i?,5/?)7a/?)-3-[(3,4,5-TrimethoxyphenyI)[(ethoxyethyl)oxy]methyl]-4-[2-methoxy-3,4(methylenedioxy)-phenyl]oxomethyl)-5-(l-menthyloxy)dihydro-2(3//)-furanone

H3CO

H3CO' '•'COOH H3CC)

H3C0

y

0CH3

0CH3

Isoxazopodophyllic acid

Nemerosin

Several D-rings (the lactonic group) were obtained from podophyllotoxin by semisynthesis, and their in vitro anti-cancer activity was determined. Most of the analogues showed activity against breast, ovarian, prostate, colon, renal and lung cancer cell lines, and some of them were active in the nM range [116]. Compounds with an open D-ring had increased activity, and the podophyllic acid methyl ether was the most interesting, with higher potency than podophyllic acid and etoposide against different cell lines. The 50% growth inhibition (GI50) range was 10 to 60 nM (breast cancer cells), 10 to 40 nM (colon cancer cells) and 10 to 40 nM against other human cell lines. Deoxypodophyllotoxin (anthricin), (-)-deoxypodorhizone, nemerosin, morelensin, and (-)-hinokinin from Anthriscus sylvestris (Umbelliferae) showed antiproliferative activities in vitro against MK-1, HeLa, and

216 B16F10 cells. The ED50 range was 0.5-1.0 ng/mL for anthricin, 0.3-0.8 |Lig/mL for deoxypodorhizone and 0.4-07 \xg/mL for nemerosin [117], 1.672.72 \xg/mL for hinokinin and 53-87 ng/mL for morelensin [118]. Hattalin is a synthetic lignan with affinity for the ouabain receptor. A study of the effect of twelve mammalian lignans on the growth of human mammary tumor ZR-75-1 cells revealed that the suppressive effect on tumor cell growth did not occur through inhibition of DNA synthesis but rather partly by inhibition of the plasma membrane ATPase other than Na^ and K^-dependent ones [119]. The effect of hattalin, which inhibited growth at an EC50 of 2.1 |Lig/mL, was the strongest of all the assayed lignans. Hirano et al [120] demonstrated its specific activity against ATPase from cancer tissue, which was inhibited by more than 50% by 2.0 mM hattalin, whereas at the same dose there was only 33% inhibition of the specific activity of ATPase from normal gastric mucosa. Hattalin inhibited both enzymes more strongly than did ouabain. From these data it is evident that the sensitivity of plasma membrane ATPase associated to lignan increased in gastric cancer. The effects of steganacin analogues and the stereochemical requirements for them to be active against KB cells were established after studying nineteen isomer lignans in their (+)-, (-)- and (±)-forms [121]. The compounds showing higher activity were that of the (-)-series, with the same R absolute configuration around the pivotal biphenyl bond. In addition, a-acetoxy seems to be clearly better than the P-acetoxy form. Steganangin, steganacin, steganolide A, episteganangin, and steganoate B from Steganotaenia araliacea (Umbelliferae) demonstrated cytotoxic activity against a panel of eleven human tumor cell lines, with an ED50 range between 0.4 and 17.3 |Lig/mL for active compounds [122]. However, they were clearly less potent than colchicine, which had an ED50 range between 1 ng/mL and 3.5 |ig/mL, against the same cell lines tested. In a second paper they studied the effects of deoxypodophyllotoxin and Ppeltatin methyl ether, from Bursera permoUis (Burseraceae). In this case the lignans showed strong cytotoxic effects against the eleven human cancer cell lines assayed, with a ED50 range from 0.2 ng/mL to 11.3 |Lig/mL; deoxypodophyllotoxin was the most active and potent of the lignans tested [123].

217

H3C0

H3C0

HjCO

Steganoate B

H3CO

OCH3

Steganolide A

Phyllanthin and hypophyllanthin from Phyllanthus amarus had no cytotoxic activity against different cultured mammalian cells, but both enhanced the cytotoxic response mediated by vinblastine with multidrugresistant KB cells, and phyllanthin displaced the binding of vinblastine to membrane vesicles derived from this cell line, probably by an interaction with the P-glycoprotein [124]. Hirano et al [125] studied the anti leukemic cell efficacy of eleven naturally occurring lignans against the human promyelocytic leukemic cell line HL-60, and four of them showed considerable suppressive effects, with IC50 ranging from 10-940 ng/mL. However, they showed little inhibitory activity on mitogen-induced histogenesis of human peripheral blood lymphocytes. Of the active compounds, honokiol, machilin A, matairesinol and arctigenin had the strongest effects. Their IC50 values of less than 100 ng/mL were close to those of current anti-cancer agents. The lignans, however, had little or no cytotoxicity against the tested cells as assessed by dye exclusion tests (LC50 > 2.9 |Lig/mL), whereas the standard anti-cancer agents had potent cytotoxicity. These lignans inhibited HL-60 cell growth with a non-toxic mechanism, probably by interfering with the DNA, RNA and/or protein synthesis of the leukemic cells, because they strongly suppressed the incorporation of [^H]-thymidine, [^H]-uridine, and [^H]leucine into HL-60 cells. In addition, machilin A and arctigenin inhibited the growth of human T lymphocytic leukemia cells. Arctiin and arctigenin, isolated from Arctium lappa (Compositae), showed strong cytotoxicity against HepG2 cells, but little toxicity against Chang liver cells [126]. The cytotoxicity of arctigenin against Chang liver cells was markedly potentiated by treatment with the glutathione (GSH) synthesis inhibitor L-buthionine-(5',/?)sulfoximine, and the cytotoxicity of arctigenin against HepG2 cells increased in an exposure time-dependent

218 manner, but was hardly changed by the GSH inhibitor. Arctigenin and some of its aliphatic esters, especially ^/-decanoate, transformed more than half of the mouse myeloid leukemia M 1 cells into phagocytic cells at a concentration of 2 |iM. The y-lactone ring appears to have a relevant role in this activity, whereas cyclohexene and cyclooctadiene skeletons showed negative effects [127]. In another study, Kato et al [128] demonstrated that arctiin has weak protective potential against hepatocarcinogenesis in rats induced by 17p-ethinyl estradiol and 2-acetylaminofluorene. Only two of the five lignans isolated from Hernandia nymphaeifolia showed significant cytotoxic activities against P-388, KB 16, A549 and HT29 cell lines. (-)-Deoxypodophyllotoxin and (-)-yatein had ED50 values < 5 ng/mL, whereas in the rest of the isolated lignans, (+)-epimagnolin, (+)epiaschantin, (+)-epiyangambin, the activity range was 2.777-0.389 |Lig/mL [129]. Medarde et al [130,131] obtained a new class of heteroaromatic analogues of lignans, called heterolignanolides, with diverse heterocyclic rings instead of the trimethoxyphenyl ring found in the natural lignans yatein and podorhizol. They showed moderate antineoplastic activity against the tumor cell lines P-388, A-549 and HT-29 (IC50 > 2.5 |Lig/mL) when compared with yatein (IC50 range 25-50 ng/mL) [130]. The same authors [131,132] synthesised several bromo-, chloro- and iodolignanolides from yatein and 4'-0-demethylyatein, which were assayed as antineoplastics but only showed a modest antineoplastic activity. These results confirm that demethylation and halogenation decrease the activity of the dibenzylbutyrolactone type lignans. Five lignans isolated from Aglaia elliptica (Meliaceae) inhibited the growth of twelve human cancer cells in culture [133,134]. The IC50 values for methyl rocaglate, 4'-demethoxy-3',4'-methylenedioxy-methyl rocaglate, 4'-demethoxy-3 ',4'-methylenedioxyrocaglaol, 1 -oxo-4'-demethoxy-3 ',4'-methylenedioxyrocaglaol and 1 -(9-formyl-4'-demethoxy-3 ',4'-methylenedioxymethyl rocaglate were in the 0.8-30 ng/mL range. In a further assay, 4'demethoxy-3',4'-methylenedioxy-methyl rocaglate (IC50 = 0.9 ng/mL) induced accumulation in the Gi/Go phase of the cell cycle afler 24 or 32 h of incubation and inhibited cell proliferation in a dose-dependent manner, but colony formation was not reduced. In addition, the lignan strongly inhibited the protein biosynthesis (IC50 = 25 ng/mL), but did not affect the nucleic acid biosynthesis, even at a high concentration (1 |uig/mL). The growth of BCl in athymic mice was delayed by treatment with 10 mg of

219 lignan per kg three times per week, and there were no toxic effects nor was the body weight affected. The effects of dihydroguaiaretic acid on the initiation stage in a rat multiorgan carcinogenesis model were examined by Takaba et al [135]. Histological examination revealed no inhibitory effects in terms of the multiplicities and/or incidence of neoplastic lesions in any of the organs examined at the experimental doses assayed. The same lignan was isolated and identified as the cytotoxic principle of Pycnanthus angolensis (Myristicaceae), which inhibited the growth of six human tumor cell, with an ED50 range of 1.63-3.10 |Lig/mL [136]. Dihydroguaiaretic acid and NDGA inhibited complex formation of the fos'jun dimer and the DNA consensus sequence, with IC50 values of 0.21 \iM and 7.9 nM, respectively, and suppressed in vitro leukemia, lung cancer and colon cancer cell lines [137]. The growth and metabolism of four prostate cell lines were inhibited with different doses of NDGA over 3 days in culture. Differences in the pattern of the results suggest that different mechanisms were operating, but there was no evidence of any synergistic activity on the inhibition of cell proliferation [138]. As part of a search for chemopreventive agents from natural products, nearly 400 plant extracts were tested for their potential to induce human promyelocytic leukemia (HL-60) cell differentiation. Seventeen of the plant extracts were active, with ED50 values of 4 |j.g/mL [139], One of most potent was an extract derived from Dirca occidentalis (Thymelaeaceae), which gave an ED50 of 0.14 |Lig/mL, and from it genkwanin, (±)lariciresinol and sitoindoside II were isolated as active principles by a bioassay-guided fractionation. The ED50 values of the compounds were 18.3, 1.1 and 0.069 |j.M, respectively. Three new formyl neolignans, obovatinal, perseal A and perseal B, isolated from the leaves of Persea obovatifolia (Lauraceae) were assayed as cytotoxic agents by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide colorimetric method against mouse lymphocytic leukemia P-388, human nasopharyngeal carcinoma KB 16, human lung adenocarcinoma A549, and human colon adenocarcinoma HT-29 cell lines [140]. The activity range measured by the ED50 values was 0.266 to 1.493 |ig/mL. Perseal A was five fold more potent against A549 adenocarcinoma than its isomer, perseal B. Continuing with this research, the same authors [141] completed the study with four additional neolignans isolated from the same source, obovatifol, obovaten, perseal C and perseal D. They showed

220

significant cytotoxic activities against the same cancer cell lines used in the former experiment. Obovatifol was the most active compound, with ED50 values close to those of the reference drug, mithramycin, against the KB 16 cell line. The acetylation of obovatifol and obovaten gave two compounds that were more potent than the reference drug against the KB 16 cell line (ED50 = 0.075 and 0.049 |Lig/mL, for obovatifol diacetate and obovaten diacetate, respectively). The bis-glucoside derivatives isolated from Galinum sinaicum (Rubiaceae), IS, 8/?, 8'i?-(-)-lariciresinol-4,4'-bis-0-p-D-glucopyranoside and IS, Si?, 8'/?-(-)-5-methoxylariciresinol-4,4'-bis-0-p-D-glucopyranoside, exhibited weak cytotoxic activity against the P388 cell line, with IC50 values of 100 and 42 |itg/mL, respectively [142]. Rocaglaol, pyrimidinone and aglaiastatin isolated from Aglaia odorata (Meliaceae) were potent inhibitors of the growth of K-ra^-NRK cells, with IC50 values of 1-10 ng/mL, and induced normal morphology in K-ra^-NRK cells at 10-30 ng/mL. They also specifically inhibited protein synthesis. Aglaiastatin was the most effective in inhibiting cell growth. Aglaiastatin reduced the amount of Ras, possibly by inhibiting its de novo synthesis [143]. OH

H3CO

Phytoestrogens have a direct relationship with the reduction in the risk of breast, colon and prostate cancer, but these effects are due to mammalian lignans that reach the human body from a vegetable diet. These effects are extensively dealt with in the corresponding section. We now focus on the direct cytotoxicity of mammalian lignans on cancer cells, their effects on cell growth and cell metabolism. The anticancer mechanisms of phytoestrogens, lignans included, seem to be due to the antagonism of

221 estrogen metabolism, antioxidant activity, and modulatory effects on key control points of the cell cycle [80]. Landstrom et al [144] demonstrated that soy flour inhibited implanted prostate cancer growth and suggest that the lignans present may participate in the delayed prostate tumor growth. They observed a significant increase in daily urinary excretion of enterolactone and enterodiol during the metabolic period. These findings support the relevance of phytoestrogens in the prevention of sex cancer. On the other hand, flaxseed reduces metastasis and inhibits the growth of the metastatic secondary tumors in animals, and it may be a useful nutritional adjuvant to prevent metastasis in cancer patients [145]. Sung et al [146] investigated whether enterolactone and enterodiol have growth inhibitory effects against colon tumor cells and whether these effects are mediated through their antiestrogenic activity. At 100 |LiM concentration both lignans significantly reduced the proliferation of all cell lines, though enterolactone was more than twice as effective as enterodiol at this concentration. Growth was not affected by the presence of 17P-estradiol, which implies that these cells are not estrogen-sensitive. Thompson et al. [147] demonstrated that secoisolariciresinol diglycoside has an antitumor effect when taken at the early promotion stage of tumor genesis and may contribute to the health benefits of high-fiber foods. This lignan reduced pulmonary metastasis of melanoma cells in mice and inhibited the growth of metastatic tumors that formed in the lungs [148]. However, while secoisolariciresinol diglycoside appears to be beneficial throughout the promotional phase of carcinogenesis, the oil component seems to be more effective at the stage when tumors have already started to develop [149]. In an in vivo study, Jenab et al. [150] demonstrated the protective effects of flaxseed lignans against colon cancer. This was associated with an increase in p-glucuronidase activity. The total activity of this enzyme correlated positively with total urinary lignan excretion, especially secoisolariciresinol diglycoside, and negatively with the total number of aberrant crypts and the total number of aberrant crypt foci in the distal colon. Different lignans, including some known phytoestrogens, were evaluated for their effects on DNA synthesis in estrogen-dependent and independent human breast cancer cells. At 0.1-10 \iM, enterolactone induced DNA synthesis by 150-235%, and at 20-90 |iM it inhibited DNA synthesis by 50%. Inhibition of estrogen-dependent and independent breast cancer cells

222

at high concentrations suggests additional mechanisms independent of the ER [151]. To study the mechanisms of phytoestrogen effects on estrogen action and tyrosine kinase activity, a number of phytoestrogens and related compounds were evaluated for their effects on DNA synthesis, estimated by thymidine incorporation analysis, in estrogen-dependent cells in the presence of estradiol, tamoxifen, insulin or epidermal growth factor. At 10 |LiM, enterolactone enhanced estradiol-induced DNA synthesis [152]. The lignans enterodiol, enterolactone and NDGA and the lignan metabolite methyl /?-hydroxyphenyllactate interfere with mitogenic and tumor promotional signal transduction pathways [153]. NDGA and methyl p-hydroxyphenyllactate did not inhibit 12-0-tetradecanoylphorbol-13acetate (TPA) mediated c-fos transcription, and enterolactone and enterodiol had only a weak inhibitory effect. NDGA at 0.1-10 |Limol/L increased c-fos mRNA levels. A potential mechanism of dietary anticarcinogenesis involves the induction of detoxifying phase II enzymes such as NADPH:quinone reductase. Wang et al. [154] examined the ability of six prominent phytoestrogens, including enterolactone, to affect cellular expression of NADPH:quinone reductase in colonic cells. The concentrations required to double the specific activity of the reductase for enterolactone was 0.04 \xM. The results demonstrated that these phytoestrogens are capable of inducing the synthesis of the enzyme in Colo205 cells by promoting its specific mRNA expression, and suggest a novel mechanism by which dietary phytoestrogens may be implicated in colorectal cancer chemoprevention. Antiviral eflfects MacRae et al [12] and more recently Charlton [13] reviewed the antiviral activity of lignans. Since the effect of resin containing lignans (podophyllin) against venereal warts {Condyloma acuminatum), an ailment caused by a papilloma virus, was first described, a large number of studies on the antiviral activity of these compounds have been reported. Charlton reviews forty-seven lignans up to 1997, belonging to the aryltetrahydronaphtalenes (or aryltetralins), arylnaphtalenes, dibenzocyclooctadienes and dibenzylbutanes. The first group is the most relevant and best known, and its compounds have the most potent antiviral effects, because some of them are active at nM concentrations, such as podophyllotoxins and peltatins. Several viruses have been assayed; there are reports on the

223

herpes simplex virus (HSV), measles virus, Sindbis virus, murine cytomegalovirus (MCMV), vesicular stomatitis virus (VSV), human papilloma virus (HPV) and human immunodeficiency virus (HIV). The mechanism of actions involves tubulin binding, reverse transcriptase inhibition, integrase inhibition and topoisomerase inhibition. Podophyllotoxins bind to tubulin and are able to disrupt the cellular cytoskeleton and interfere with some vital virus processes. There is no relationship between the inhibition of reverse transcriptase (RT) and chemical structure in the case of lignans, because all the chemical antiviral structures are able to bind to the enzyme. As to the rest of the mechanism, there is not much information available. The effects of rabdosiin may be due to its topoisomerase inhibitory effects. Charlton concluded that the antiviral activity of lignans is not strong and that except for podophyllotoxin, which is used topically to treat various warts caused by HPV, none of them are of interest for commercial application.

HO'

Rabdosiin

Many compounds of plant origin have been identified that inhibit different stages in the replication cycle of HIV. Some of them are lignans and have been reviewed by Vlietinck et al [155] and Matthee et al [156], along with their possible mechanism of action. The lignans (-)-arctigenin from Forsythia intermedia and (-)-trachelogenin from Ipomoea cairica (Convolvulaceae) suppressed the integration of proviral DNA into the cellular genome, and interiotherin A and schisantherin D from Kadsura interior (Schisandraceae) inhibited HIV replication at |j.g/mL range. Nine anti-HIV lignans were isolated from K. interior, of which gomisin-G was the most potent, with EC50 and therapeutic index values of 6 and 300

224

ng/mL, respectively. Schisantherin-D, kadsuranin, and schisandrin-C showed good activity, with EC50 values of 0.5, 0.8, and 1.2 |ag/mL, and therapeutic index values of 110, 56, and 33.3, respectively [157]. Interiotherin A and schisantherin D had EC50 values of 3.1 and 0.5 |Lig/mL, and a therapeutic index of 13.2 and 50.6 respectively [158]. Sometimes the antiviral activity is not due to a single lignan but to a mixture of them. This is the case of anolignan A and anolignan B, which were isolated and identified by bioassay-guided fractionation as the active HIV-1 RT inhibitory constituents of Anogeissus acuminata var. lanceolata (Combretaceae). Anolignan B had weak activity when tested alone (IC50 = 60.4 M
Anolignan A (Ri=OH, R2+R3=-0CH20-) Anolignan B (Ri=R2=H, R3=OH)

A synthetic bromine derivative of gomisin J [(6/?,7/LSVS-biar)-4,9-dibromo -3, lO-dihydroxy-1,2,11,12-tetramethoxy-6,7-dimethyl-5,6,7,8-tetrahydrodibenzo[a,c]cyclooctene], was found to be a potent inhibitor of the cytopathic effects of fflV-1 on MT-4 human T cells (ED50, 0.1 to 0.5 |LIM). Comparison of various gomisin J derivatives with gomisin J itself showed that iodine, bromine, and chlorine in the fourth and ninth positions increased RT inhibitory activity as well as cytoprotective activity [161].

225 0CH3

OCHj

Termilignan

Thannilignan

Of the six lignans isolated from Phyllanthus myrtifolius, only phyllamyricin E inhibited HIV-1 RT activity, whereas phyllamyricoside A increased the activity of HIV-1 RP by 65% at a concentration of 1.89 JLIM [162]. Phyllamycin B and retrojusticidin B isolated from the same source had a strong inhibitory effect on HIV-1 RT, but inhibited human DNA polymerase-a activity much less [163]. The IC50 of phyllamycin B and retrojusticidin B were 3.5 and 5.5 |LIM for HIV-1 RT, and 289 and 989 ^iM for human DNA polymerase-a, respectively. They act by a non-competitive inhibition mechanism with respect to template-primer and triphosphate substrate. In a later study, Liu et al [164] established the quantitative structure-activity relationships for both lignans against both enzymes. Phyllamycin B and retrojusticidin B have the methylenedioxyphenyl two ring system and the naphthylenelactone three ring system dissects at 90° forming a T-shape. By an overlapping of the 3-D models of both lignans along the C10-C20-C19 backbones, a nearly perfect superimposition is obtained. In the case of the HIV-1 RT, dipolar interactions {\\) appear to be more important than hydrogen bonding, while the opposite is true in the case of the human DNA polymerase-a inhibition. These characteristics are fundamental for the selectivity of the lignans as antiviral agents, and these finding may be useful in further molecular modifications. From a weak antiviral lignan, NDGA, Hwu et al [165] obtained eight 0-methylated derivatives. M6?5o-l,4-bis(3,4-dimethoxyphenyl)-(2i?,3*S)-dimethylbutane (or tetramethyl-NDGA) was the strongest as an anti-HIV on the in vitro test, with an IC50 of 11 ^M. The rest of the derivatives had an IC50 range between 14 and 38 \xM. The degree of methylation seems to improve the antiviral activity. In a second study, they examine whether tetramethyl-0-NDGA is able to inhibit the replication of HSV [166]. The results showed that in Vero cells, the lignan inhibits the expression of the herpes immediate early gene (a-ICP4), which is essential for HSV

226 replication at IC50 = 43.5 |iM. For tetramethyl-0-NDGA, drug sensitivity (IC50) varied between 11.7 and 4 ^iM in 10 passages of HSV-1 and 4 passages of HSV-2, and there is no indication that a higher drug concentration is needed. In the case of acyclovir, the IC50 increased from 7 |LiM in the first passage to 444 |iM in the tenth passage of HSV-1, and was > 88 |LiM for the fourth passage of HSV-2, thus indicating a rapid build-up of drug resistance against acyclovir. The selective index (relationship between toxicity and activity, TC50/IC50) remained constant for tetramethylO-NDGA and no cross-resistance between the lignan and acyclovir in their anti-HSV effects was detected. Justicidins A and B, diphyllin, diphyllin apioside and diphyllin apioside5-acetate from of Justicia procumbens var. leucantha (Acanthaceae) showed strong antiviral activity, with a minimum inhibitory concentration (MIC) of 0.13, > 0.06, 0.25, 0.25 and 0.13 |Lig/mL, respectively, against VSV, and they had low cytotoxicity [167]. The lignans asarinin and xanthoxylol isolated from Asiasarum heterotropoides var. mandshuricum [104], inhibited Epstein-Barr virus (EBV) early antigen activation induced by TPA. Rhinacanthin E and rhinacanthin F isolated from the aerial parts of the plant Rhinacanthus nasutus (Acanthaceae) showed significant antiviral activity against influenza virus type A but no effect on HSV-2. The EC50 obtained in a hemadsorption-inhibition assay were 1.7 and < 0.94 |ig/mL, and in a viral cytopathic effect assay were 7.4 and 3.1 |Lig/mL, respectively [168]. From a water extract of the fruits of Helicteres isora (Sterculiaceae), Tezuka et al [169] isolated three new neolignans, helisorin, helisterculins A and B, with weak inhibitory activity against RT from avian myeloblastosis virus (IC50 0.46, 1.6 and 1.0 mM, respectively). Antibacterial, Antifungal, Parasiticidal and Insecticidal EtTects Antibacterial Effects

Myristica Jragrans (Myristicaceae), Magnolia obovata and M officinalis had antibacterial activity against the cariogenic bacterium Streptococcus mutans. The active principles from M jragrans were dihydrodiisoeugenol and 5'-methoxydehydrodiisoeugenol [170] which were bactericidal and

227

inhibited Streptococcus mutans growth at an MIC of 12.5 |Lig/niL. When the authors compared the effects of these compounds with the activity of isoeugenol, they observed that dihydrodiisoeugenol is much more potent than isoeugenol (MIC > 200 |ag/mL). The active principles from Magnolia species were magnolol and honokiol, the MIC of which were of 6.3 |ug/mL, but they did not inhibit the adherence of the heat-treated bacteria to glass or smooth surfaces in the presence of sucrose and glucosyltransferase, the synthesis of soluble and insoluble glucans from glucose, and the agglutination of bacteria in the presence of high molecular weight dextran [171]. In a research program on biologically active plants, the extract of Sassafras randaiense (Lauraceae) roots was active against Staphylococcus aureus, Mycobacterium smegmatis, Saccharomyces cerevisiae and Trichophyton mentagrophytes. El-Feraly et al [172] isolated two active neolignans, magnolol and isomagnolol from this species. Isomagnolol was slightly more potent than magnolol, with MICs of 3.12, 1.56, 25 and 1.56 l^g/mL, respectively for each microorganism tested.

xX

R2' HO

CH2—CH==CH2 Magnolol (Ri=H, R2=OH) Honokiol (R,=OH, R2=H)

Antifungal Effects

Two aryltetralin lignans, 4'-(9-demethyldehydropodophyllotoxin and picropodophyllone isolated from Podophyllum hexandrum showed strong antifungal activity against Epidermophyton fuccosum, Curvularia lunata, Nigrospora oryzae, Microsporum canis, Allescheria boydii and Pleurotus ostreatus. Picropodophyllone also showed activity against Drechslera rostrata [173].

228

Lignans isolated from Terminalia hellerica inhibited the growth of Penicillium expansum. The activity range was 1-5 fig with a direct bioautographic method, but against Candida albicans the activity on an agar overlay technique was 10 \ig for termilignan, 80 |ig for termilignan and > 200 |Lig for anolignan B [160]. Eighteen synthetic racemic 8.0.4'-neolignans with six different substitution patterns in rings A and B, in their ketone and in their erythro/threO'forms, and three hydroxy-derivatives were evaluated for antifungal activity by the agar dilution method against the dermatophytes Microsporum canis, M. gypseum, Tricophyton mentagrophytes, T. rubrum, and Epidermophyton floccosum [174]. Only the hydroxy lignans exhibited a broad spectrum of activities against the five fungus strains assayed. The neolignan compound-2 (Table 11) was the most active against E, floccosum, with an MIC of 5 |Lig/mL. The neolignans compound-3 to 5 (Table 11) were the compounds that inhibited the assayed fungi, with MIC under 100 |ig/mL. None of the tested compounds had activity against the yeasts Candida albicans, Saccharomyces cerevisiae and Cryptococcus neoformans, nor against the filamentous fiingi Aspergillus niger, A. fiimigatus and A. flavus. The racemic antifungal alcohols of the 8.0.4'neolignan type were evaluated for inhibitory activity on the fungal cell wall using the whole cell Neurospora crassa hyphal growth inhibition assay [175]. The experimental results suggest that these compounds could act by inhibiting cell wall polymer synthesis or assembly. Verrucosin, oleiferin-B, 3,4,3',4'-tetramethoxylignan-7-ol, oleiferin-F and oleiferin-G (Table 12) isolated from Virola oleifera showed antifungal activity against Cladosporium sphaerospermum and C cladosporoides at 25 lag, but only oleiferin-B and oleiferin-G inhibited C. cladosporoides at 10|Lig[176]. HO,

H3CO'

OC:H3

Verrucosin

229 Parasiticidal Effects The toxic lignans isolated from Bupleurum salicifolium (Umbelliferae) were assayed against Globodera pallida and G. rostochiensis, and bursehernin and matairesinol inhibited the hatching of the two nematodes tested. The authors conclude that lignans may play a role in the defence mechanisms of potato plants, as allelopathic substances acting against cystforming nematodes [177]. The in vitro antileishmanial activity of lignans from Doliocarpus dentatus (Dilleniaceae) against amastigotes of Leishmania amazonensis was reported by Sauvain et al [178]. The lignans isolated were (+)-pinoresinol, (+)-medioresinol and (-)-lirioresinol B. Table 11.

Chemical Structure of Compounds 2-5 ^5

()CH3

^ ' \ ^ , ^ ^ \ ^ ^ ^ OH

K^^^Y 2 (±)-£'rvr/7ro-3,4-methylenedioxy-7-hydroxy-1' allyl-3',5'-dimethoxy-8.0.4'-neolignan 3 (±)-r/7reo-3,4,5-trimethoxy-7-hydroxy-l'-allyl3 ',5 '-dimethoxy-8.0.4' -neolignan 4 (±)-r;2reo-3,4-dimethoxy-7-hydroxy-l '-allyl-3'niethoxy-8.0.4'-neolignan 5 (±)-£'rvr/?ro-3,4,5-trimethoxy-7-hydroxy-l '-(£)propenyl-3'-methoxy-8.0.4'-neoIignan

Ri R2 -OCH2O-

R3 H

R4 OCH3

Rs -CH2CH=CH2

OCH3

OCH3

OCH3

OCH3

-CH2CH=CH2

OCH3

OCH3

H

H

-CH2CH=CH2

OCH3

OCH3

OCH3

H

-CH=CHCH2

230

Table 12.

Active Dibenzylbutane Lignans OH

Oleiferin B Oleiferin F Oleiferin G 3,4,3 ',4'-Tetramethoxylignan-7-ol

Ri Veratryl(3,4-dimethoxyphenyl) Pip erony 1(3,4-m ethyl enedi oxypheny 1) Veratryl(3,4-dimethoxyphenyI) Veratryl(3,4-dimethoxyphenyl)

R2 Piperonyl(3,4-methylenedioxyphenyl) Guaiacyl(4-hydroxy-3-methoxyphenyl) Guaiacyl(4-hydroxy-3-methoxyphenyl) Veratryl(3,4-dimethoxyphenyl)

The lignan (H-)-nyasol, isolated from the roots of Asparagus africanus (Liliaceae), proved to be an antiprotozoal agent. It potently inhibits the growth of Leishmania major promastigotes, with an IC50 of 12 |LIM, and moderately inhibits Plasmodium falciparum schizonts, in this case with an IC50 = 49 |LiM. At the assayed concentrations, the proliferation of human lymphocytes was only moderately affected [179]. Although the potency was clearly lower than that of the reference drug used in the experiment, the lignan inhibited both the chloroquine-sensitive and the chloroquineresistant strains. This feature indicates a non cross resistance between the two drugs, and for this reason the authors recommend using this compound to develop antiprotozoal drugs. Among the compounds isolated from Virola surinamensis, the lignans presented the highest trypanosomicidal activity in vitro against the trypomastigote form of Trypanosoma cruzi [180]. They were identified as veraguensin and grandisin, two tetrahydrofiiran lignans. At 5 [ig/mL they produced total parasite lysis after 24 h, without red cell damage. After reinoculation in healthy mice, nofiirtherinfections were observed. Whereas burchellin was only partially active, NDGA drastically reduced the number of epimastigotes and metacyclic trypomastigotes of Trypanosoma cruzi in the excreta of infected Rhodnius prolixus, the insect vector feeding on epimastigotes of T cruzi. However, the number of parasites in the gut decreased in both cases [181].

231

Termilignan and anolignan B isolated from Terminalia bellerica inhibited the chloroquine-susceptible strain of Plasmodium falciparum assayed, with IC50 of 9.6 and 20.5 ^iM respectively [160].

0CH3

0CH3 H3C0

0CH3

Veraguensin

(+)-Nyasol

Insecticidal A ctivity

In their review of lignans MacRae and Towers [9] cited the activity of kobusin and sesamin against silkworm larvae, sesamin and sesamolin against juvenile hormone activity in the milkweed bug, and podophyllotoxins against insect larval growth. On the other hand, sesamin, asarinin, savinin and hinokinin enhanced the toxicity of a wide variety of insecticides. The powdered dried leaves and leaf extracts of Libocedrus bidwilli (Cupressaceae) were toxic to the larvae of the housefly (Musca domestica) and the codling moth (Laspeyresia pomonella), but the powdered material was not toxic to the light-brown apple moth {Epiphyas postvittana). The most active compound present in the leaves was the p-peltatin-A methyl ether, which at a concentration of 100 ppm in a chemically defined diet gave 98% mortality among housefly larvae [182]. The neolignan (+)haedoxan A present in Phryma leptostachya (Phrymaceae) and (+)haedoxan D, its synthetic demethoxy derivative, exhibited insecticidal activity against Musca domestica, whereas (-)-haedoxan D was inactive [183].

232

H3C0.

OCH3 (+)-Haedoxan A (R=OCH3) (+)-Haedoxan B (R=H)

Four known neolignans isolated from Piper decurrens (Piperaceae) showed insecticidal activity against mosquito larvae. Eupomatenoid-6 was the toxic lignan, followed by conocarpan and eupomatenoid-5, whereas decurrenal showed weak activity [184]. The neolignans magnolol, 4methoxyhonokiol and 4,4'-diallyl-2,3'-dihydroxybiphenyl ether, from Magnolia virginiana were toxic to brine shrimp and mosquito larvae and showed strong fungicidal and bactericidal activities [185], and helioxanthin from Taiwania cryptomerioides (Taxodiaceae), inhibited the yellow fever mosquito larvae growth with an LC50 of 3.0 |ig/mL [186]. Cardiovascular effects Recently, in 1997, Ghisalberti [14] published a review on the cardiovascular activity of naturally occurring lignans, including the effects on cyclic adenosine monophosphate (cAMP) phosphodiesterase (PDE), Ca^^ channel movement, hypertension, and as platelet activating factor (PAF) and endothelin antagonists. Antioxidant and vasorelaxant effects were also covered. The most relevant finding on this subject fi*om 1996 until 1999 are dealt with below.

233

Myocardial Ischemia

Human correlation and animal studies suggest that consuming of foods rich in phytoestrogens and lignans may result in reduced risk of cardiovascular disease. The postulated antiatherogenic mechanisms are hypocholesterolemic effects as well as inhibition of platelet activation [80]. Pretreatment with a lignan-enriched Sheng-Mai-San, a, traditional Chinese formulation used for the treatment of coronary disease, was effective in protecting against isoproterenol-induced myocardial injury in rats, and in ischemia-reperfusion injury in isolated perfused hearts prepared from pretreated animals [187]. The major myocardial protective compound seems to be the lignan-enriched extract of Fructus Schisandrae. In a complementary study with the same crude drugs, Li et al. [188] studied the effects of Sheng'MaUSan and the lignan-enriched extract of Fructus Schisandrae, its antioxidant component, in different experimental models of myocardial infarction and myocardial ischemia-reperfusion injury in rats. Their results indicate that the myocardial protection afforded by Sheng-MaUSan pretreatment may be mediated by the antioxidant activity of the Fructus Schisandrae lignans. Calcium Channel Blocker

Arctigenin, matairesinol, trachelogenin, nortrachelogenin and their glucosides isolated from the Compositae Trachelospermum jasminoides and Arctium lapa, hypophyllanthin from Phyllanthus niruri, and liriodendrin (but not syringaresinol-p-D-glucoside) from Boerhaavia diffusa (Nyctaginaceae) exhibited significant calcium channel antagonistic effects. Experimental data on these substances are compiled by Ghisalberti [14]. In addition, fargesone A, B and C, denudatin B, pinoresinol dimethyl ether, lirioresinol-B dimethyl ether, magnolin and fargesin from Magnolia fargesii were tested by Chen et al, [189] and were found to possess Ca^^antagonistic activity in taenia coli of the guinea pig. Denudatin B and fargesone A and B had over 50% antagonism in the 10"^ M range, whereas the rest of the substances needed concentrations of over 10""* M. Denudatin B seems to be the most active: 100% inhibition at 2.8 x 10'^ M and 50%) at 1.4 X 10"^ M. Graminone B from Imperata cylindrica (Gramineae) at 0.1

234

|j-M inhibited the KCl-induced contraction on rabbit aorta without affecting NA-induced contractions [190], but its mechanism was not elucidated. 2,3-Dibenzylbutane-l,4-diol is a Hgnan found in human urine, but its physiological significance is unknown. Abe et al [191] studied the effects of this Hgnan on vascular smooth muscle of rabbit aorta. It had no influence on the tension and the "^^Ca^^ uptake on the contractile response in vascular smooth muscle but completely inhibited the "^^Ca^^ uptake induced by high KCl or noradrenaline (NA). In the resting state, it inhibited high KCl- and CaCb-induced contractions in the partially depolarised muscle strip. The lignan inhibited NA- and angiotensin Il-induced contractions, but not the transient contraction evoked by NA in Ca^^-free solution. The results indicated that 2,3-dibenzylbutane-l,4-diol induces the relaxation of vascular smooth muscle by inhibiting Ca^^ influx but not Ca^^ release from the intracellular stores. In later research [192], the same authors studied the effects of this lignan on high K^-induced contraction in rabbit femoral artery. The lignan inhibited both the transient and the sustained contraction induced by high K^ and inhibited the transient contraction remaining in the verapamil- or the nifedipine-pretreated preparation in a concentrationdependent manner. Moreover, 2,3-dibenzylbutane-l,4-diol inhibited sustained contractions elicited by the Ca^^ channel activator Bay K 8644. As the most relevant feature of these experimental results, the mammalian lignan affected both the sensitive and the insensitive calcium channels. When 2,3dibenzylbutane-l,4-diol was assayed in rat aorta, it inhibited NA-induced contractions in a concentration-dependent manner [193]. This inhibitory effect was similar to that obtained in the high K^ solution containing verapamil. The lignan also inhibited the contractions and intracellular Ca^^ elevation induced by NA with the same potency and increased the cyclic guanine monophosphate (cGMP) levels. From these results the authors conclude that the inhibitory effects of 2,3-dibenzylbutane-l,4-diol on NAinduced contraction in rat aorta may be caused by an increase in cGMP levels.

235

OCH3

Graminone B

Antihypertensive Effects

Various reports on the antihypertensive activity of lignans were published before 1997 [14]. (±)-Pinoresinol diglucoside from Eucommia ulmoides (Eucommiaceae), pinoresinol monoglucoside from Forsythia spp., sesamin from sesame oil, neojusticin B from Justicia procumbens, ephedradine B (a hybrid neolignan-alkaloid) from Ephedra spp.(Ephedraceae), danshensuan B from Salvia miltiorrhiza (Labiatae) and magnolol from different sources were all reported to be hypotensives in different animal models. The antihypertensive effect of sesamin was studied by Matsumura et al [194] using hypertensive and unilaterally nephrectomised rats. The sesamin feeding group showed an improved development of deoxycorticosterone acetate salt-induced vascular hypertrophy in both the aorta and mesenteric artery. In a second study [195], these authors examined the antihypertensive effect of sesamin using two-kidney, one-clip renal hypertensive rats. In this experiment, hypertension was markedly reduced by the sesamin-containing diet. There were significant increases in the left ventricle plus septum weight to body weight ratio in the control group compared with the shamoperated rats. This rise was also significantly reduced in the sesamin group. The histochemical analysis of the thoracic aorta indicated that vascular hypertrophy occurred in the control group but not in the sesamin group. The sesamin diet tended to ameliorate this vascular hypertrophy, although its effect was not statistically significant. A third study [196] was performed

236 using salt-loaded and unloaded stroke-prone spontaneously hypertensive rats. Systolic blood pressure, cardiovascular hypertrophy and renal damage were evaluated. In the salt-loaded group, sesamin feeding suppressed the development of hypertension, and an efficient suppression was maintained. The left ventricle plus septum weight-to-body weight ratio was slightly but significantly lowered by sesamin feeding. In the histochemical evaluation of the degree of vascular hypertrophy, the aorta and superior mesenteric artery, wall thickness and wall area of these vessels decreased significantly in the sesamin feeding group. On the other hand, in the salt-unloaded group, only a slight, nonsignificant suppressive effect of sesamin on the development of hypertension was observed. Sesamin feeding was, then, more effective as an antihypertensive regimen in the salt-loaded strokeprone spontaneously hypertensive rats than in unloaded ones, which suggests that sesamin is more useful as a prophylactic treatment in the malignant status of hypertension and/or hypertension followed by water and salt retention. Digoxin-like activity

Some lignans with estrogenic properties have been described as "endogenous ouabain-like" factors. They were studied by Braquet et al [197-199], Hirano et al [200] and Fagoo et al [201]. The animal lignans (enterolactone, prestegane B and 3-(9-methyl enterolactone) inhibited the Na^/K^-ATPase pump activity in human red cells and human and guineapig heart cell membranes, with IC50 ranging from 5 to 9 x 10'"^ M, but the IC50 for ouabain (7 x 10'^ M) was not modified by addition of the lignans. These results suggest a non-competitive inhibition of ouabain receptor. In addition, the apparent affinity for internal Na^ and the maximal rate of cation translocation diminished, and the inhibition of the NaVK^-pump was obtained at doses higher than those required for ouabain. In a complementary study [201], enterolactone displaced [^H]-ouabain from its binding sites on cardiac digitalis receptor and inhibited dose dependently the Na^/K^-ATPase activity of human and guinea-pig heart in the 10""^ M range. The cross-reactivity of enterolactone with antidigoxin antibodies was low, even in man. In conclusion, lignans may contribute to the putative digitalis-like activity found in tissues, blood and urine of several mammals and may be endogenous digitalis-like substances.

237

Sixteen mammalian-type lignan derivatives, including enterolactone and enterodiol, were studied for their potential endogenous digoxin-like activity [200]. M(^5'o-2,3-dibenzylbutane-l,4-diol showed the most potent crossreactivity against the antidigoxin antibody, but many of the lignans had activity in the three experimental models tested, inhibiting NaVK^-ATPase activity with IC50 lower than 5 x 10"^ M and [^H]-ouabain displacement activity with an IC50 range of lO""* to 10*^ M. No study on the digoxin-like activity of hattalin has been reported though it has affinity for the ouabain receptor. Endothelin antagonism

Hussain et ah [202] isolated three lignans from Phyllanthus niruri which were identified as phyllanthin, hypophyllanthin and nirtetralin. They were also found to inhibit [^^^I]-ET-1 binding to the recombinant human ETA receptor expressed in Chinese hamster ovary cells (CHO-ETA), but were inactive against the recombinant ETB receptor. Hypophyllanthin was the most potent compound, with an IC50 value of 40 |LIM. Cyclic AMP phosphodiesterase (PDE) inhibition The cAMP-PDEs are a group of enzymes capable of hydrolysing cAMP to non-cyclic AMP (5'-AMP). Inhibition of PDE activity increases the cAMP levels, which produces stimulation of cardiac contractility and vasodilatation, and decreases vascular resistance and arterial pressure. Different subtypes of PDEs have been described, with different and selective effects. Since Nikaido et al [203] first described the inhibition of cAMP-PDE by lignans, different synthetic analogues have been described [204-206]. On a screening with 250 aqueous extracts, thirty-four of them gave inhibitory activity, and the active principles of the Anemarrhena asphodeloides (Anthericaceae) were norlignans such as m-hinokiresinol [203]. The same year, Nikaido etal [207] studied different anti-inflammatory plants, such as Forsythia suspensa, F. viridissima and F. koreana. (-f)Pinoresinol, (+)-pinoresinol-P-D-glucoside, (-)-matairesinol and arctigenin showed strong inhibitory effects against PDE, with IC50 values of 7.5, 14.2, 9.8 and 13.9 x 10"^ M respectively, whereas phillygenin, phyllirin, arctiin

238 and (-)-matairesinol-p-D-glucoside had no activity at 50 x 10"^ M. Stmctural modifications of these active principles showed that in general the synthetic derivatives did not improve the activity. However, some special cases should be mentioned. The presence of two/7-hydroxyl groups or substituted hydroxyl groups is essential for the PDE inhibitory activity of lignans. 1 - Acetoxy-(+)-pinoresinol and 1 -acetoxy-(+)-pinoresinol-(3-Dglucoside were the most potent of all the assayed lignans, with IC50 values of 3.2 and 4.4 x 10"^ M respectively. Other interesting compounds were (+)pinoresinol monomethyl ether, (+)-pinoresinol-di-P-D-glucoside and syringaresinol-di-P-D-glucoside. Different kind of lignans that inhibit PDE activity have been described, and are reviewed by Ghisalberti [14]. They are dibenzylbutyrolactones, fliranofurans and tetrahydroftirans. In the first group, (+)-matairesinol is the main principle, with an IC50 of 9.8 x 10"^ M. (+)-Pinoresinol is the most potent in the second group, which has an IC50 of 7.5 x 10"^ M, but acetylation clearly increases the potency. In addition, the stereochemistry of the fiiran-phenyl bond can clearly modify the activity, and the C-TS / C-IS compounds or series (+) are more active than the C-TSI C-7R compounds or series (-). A third group, the tetrahydrofiirans, is less relevant because there is a clear loss of activity. The dibenzocyclooctadiene lignans have a biphenyl characteristic, similar in part to that of the bipyridin group present in the standard inhibitors amrinone, milrinone and others. The activity of these lignans improves with the introduction of a methylenedioxy group or a bromide, and decreases with the introduction of a hydroxyl radical, whereas the enantiomers have the same range of activity. Other interesting compounds are (+)-dehydrocaffeic acid dilactone (from fungal origin), schisandrin and NDGA. NDGA is included in the dibenzylbutane lignan group, which can adopt a conformation similar to that of the alkaloid papaverine, according to the three dimensional modelling studies [14]. Inhibition of PDE IV contributed to relaxation of airway smooth muscle and the prevention of proinflammatory cell activation. However, important side effects limited their therapeutic effects. In order to obtain new data on lignans as PDE inhibitors, Iwasaki et ai [204] synthesised a series of 1aryl-2,3-bis(hydroxymethyl)naphthalene lignans and evaluated their ability to selectively inhibit PDE IV isolated from guinea pig and the histamineinduced and antigen-induced bronchoconstriction in the guinea pig. The more potent and selective compounds were the N-alkylpyridone derivatives (ring at C-1), and the most potent of these was 6,7-diethoxy-2,3-bis

239 (hydroxymethyl)-1 -[ 1 -(2-methoxyethyl)-2-oxo-pyrid-4-yl]naphthalene with ED50 values of 0.08 and 2.3 mg/kg /.v., against histamine-induced and antigen-induced bronchoconstriction in the guinea pig respectively. In addition, replacement of the 1-phenyl ring by a pyridone ring led to marked improvement in their selectivity for PDEIV over PDE III. Ukita et al [206] studied the structural requirements of a 1-pyridylnaphthalene series, and the best compound was compound-6 (PDE IV inhibition IC50 = 0.13 nM, selectivity PDE III/PDE IV ratio = 14000). This compound showed potent antispasmogenic activities, reducing the bronchoconstriction in guinea pigs induced by antigen and histamine, with ED50 values of 63 and 33 |Lig/kg, respectively, with few cardiovascular effects. Moreover, this synthetic lignan induced significantly weaker emetic effects than the reference drug after p.o, and /.v. administration in dogs, which is compatible with the lower affinity for the high-affinity rolipram binding site (2.6 nM). This may imply that the synthetic lignan has an improved therapeutic ratio because of a broad margin between the K\ value of binding affinity and the IC50 value of PDE IV inhibition (ratio = 0.050). In a complementary work, Ukita et al. [205] studied the structural requirements for potent, specific PDE V inhibition (cGMP-specific PDE) by a 1-arylnaphthalene lignan series. CH3CH20.

CH3CH2O

2HC1

Compound-6 2,3-Bis(hydroxymethyl)-6,7-diethoxy-l-{2[l(2//>.phthalazinon-4-(3-pyridyI)-2.yl]-4pyridyl} naphthalene hydrochloride

Compound-7 l-(3-bromo-4,5-dimethoxyphenyl)-5-chloro-3-[4-(2hydroxyethyl)-l-piperazinylcarbonyl]-2(methoxycarbonyl)naphthalene

This enzyme is distributed in different smooth muscle tissues, and its inhibition may be involved in the treatment of hypertension, angina, congestive heart failure and impotence. In this study, the synthetic lignan

240

compound-7 was the most potent specific inhibitor, with a PDE V inhibition of IC50 = 6.2 nM, and selectivity for PDE V against PDE I, II, III, and IV > 16,000. This compound had the best selectivity against the different enzyme isoforms among the known PDE V inhibitors. In addition, it showed relaxant effects on rat aortic rings, with an EC50 = 0.10 |LAM, and should be evaluated for use in the treatment of cardiovascular diseases. Platelet activating factor antagonism and coagulation Platelet activating factor (PAF) is a biologically active phospholipid released from activated basophils to induce platelet aggregation. It is a potent mediator that acts through a specific membrane receptor and its effects are exerted as nM concentrations. It is involved in physiological and pathological processes such as asthma, allergies, inflammation, respiratory and cardiovascular events, blood coagulation, immunity, etc. The isolation of specific PAF antagonists may be of interest in lignan research because some of them are clearly active. Ghisalberti [14] reviewed many of the lignans described before 1997. Kadsurenone from Piper futokadsura seems to be the most active as an inhibitor of PAF binding and of PAF-induced platelet aggregation, and kadsurenin H is the most potent of the bicyclo [3,2,l]octane neolignans. Different structural modifications have been made, and their principal features are included in the Ghisalberti review [14]. Other natural lignans of interest as PAF antagonists are honokiol and magnolol from Magnolia species; (-)-denudatin B from Piper kadsura, cinnamophilin and /wt^^o-dihydroguaiaretic acid from Cinnamomum philippinense (Lauraceae), and gomisin M from Kadsura heterocollita.

0CH3

H3C0 0CH3 HO

^.^^ OCH3 OH

Kadsurenone

Cinnamophilin

241 Shen and Hussaini [208,209] analysed the role of kadsurenone and other related natural and synthetic lignans as antagonists of PAF. Kadsurenone did not inhibit platelet aggregation induced by ADP, collagen, AA or thrombin even at 50 |LIM, but it inhibited the platelet aggregation induced by PAF. Among the structural analogous studied only the 9,10-dihydro derivative was similar in potency and site of activity to the original lignan. In addition, futoquinol, fiitoenone and fiitoxide, present in the same source, showed inhibitory effects, and the synthetic diaryltetrahydrofurane lignans, called L-652,731, L-662,025 and L-659,989, and the diaryltetrahydrothyophene called L-653,150, seem to be active in the same range and have the same mechanism as the natural lignans. The direct displacement of [^H]-PAF by the antagonists is usually used to study the specific PAF receptor antagonist. The dose-response curves of displacement of ["^H]-PAF showed a competitive antagonism in all cases, with an increase in potency from futoenone (less active), L-652,731, kadsurenone to L-659,989 (more potent). As inhibitors of PAF-induced platelet aggregation, the range of potency goes from the IC50 = 0.80 |LIM of L-659,989 to IC50 = 9.50 JLIM for futoenone. These values are in agreement with the previous results. Kadsurenone gave an IC50 = 3.50 |iM in this assay. These results demonstrate that kadsurenone and related lignans are specific, potent and reversible inhibitors of PAF receptor and PAF-induced platelet aggregation. In addition, they induce PAF-aggregation and degranulation of neutrophils. Thirty lignans from extracts of Forsythia suspensa and Arctium lappa were tested for their inhibitory effects on PAF binding to rabbit platelets, and nine of them were active [210]. Tetrahydrofurans had no activity. The active compounds as inhibitors of PAF receptor binding were bistetrahydrofiiran and butanolide type. Bistetrahydrofurans also inhibited in vitro platelet aggregation induced by PAF, which indicates that lignans of this type are PAF antagonist. The IC50 (l^iM) values for PAF binding obtained were 1.2 for (+)-pinoresinol dimethyl ether, 0.67 for (+)-acetoxypinoresinol dimethyl ether, 2.8 for (+)-isogmelinol, 1.6 for phillygenin, 0.91 for (+)epipinoresinol dimethyl ether, 0.66 for fargesin, 0.42 for isomagnolin, 2.9 for arctigenin and 0.56 for arctigenin methyl ether. When assayed for their capacity to inhibit the platelet aggregation induced by PAF the IC50 (|LAM) values were 31 for pinoresinol and 38 for fargesin, whereas for the reference drugs the value was 9.8.

242

H3C0.

H3C0'

MachilinG

L-653,150

The furofiiran lignan, epiyangambin, competitively inhibited in a dosedependent manner PAF-induced rabbit platelet aggregation in vitro, without modifying the amplitude of the maximal response, but it had no effect upon the platelet aggregation induced by collagen, thrombin or ADP [211]. The IC50 value for epiyangambin was 6.1 x 10'^ M for PAF (10'^ M)-induced aggregation, and the Schild analysis provided a pA2 of 6.91 with a slope of 0.98 and a pKb of 6.94. In vivo administration of the lignan at 20 mg/kg significantly inhibited PAF-induced thrombocytopenia in rats. These results confirm that epiyangambin is a potent and selective antagonist of PAF both in vitro and in vivo. In an analogous study, the same authors [212] demonstrated that (+)-yangambin, an isomer of isoyangambin, is an antagonist that selectively blocks PAF receptors on platelets. They studied its activity on rabbit platelet aggregation and binding of [^H]-PAF to rabbit platelet plasma membranes. Yangambin inhibited PAF-induced platelet aggregation competitively in platelet-rich plasma in a dose-dependent manner (pA2 of 6.45 with a slope of 1.17) and at 10"^ M did not affect the platelet aggregation induced by ADP, collagen, or thrombin. In a complementary study, yangambin competitively displaced [^H]-PAF binding, with an IC50 of 1.93 |iM, and did not prevent PAF-induced in vitro chemotaxis at 10"^ M. The results confirm that this lignan is an antagonist of PAF receptors on platelets, but as it had no effect on PAF-induced neutrophil chemotaxis, the authors hypothesise the presence of differences between PAF receptors expressed in platelets and neutrophils. In a later experiment, Herbert et al [213] demonstrated that yangambin can discriminate between two different types of PAF receptors on platelets and PMNL and can be considered the first PAF receptor antagonist described to date that exhibits this effect. Yangambin competitively displaced [•^H]-PAF from its high affinity binding sites on washed human platelets, with a K\ value of 1.1 (iM, and inhibited PAF-induced aggregation of human platelets in vitro (IC50 = 10 |LIM), but it had no effect on

243

PAF-induced oxidative burst in human PMNL. In guinea pigs, yangambin inhibited PAF-induced thrombocytopenia but did not affect leukocytopenia. The bistetrahydrofuran lignan (-)-syringaresinol isolated from the stem of Annona montana was tested on antiplatelet aggregation [102]. It showed significant inhibitory effects on the aggregation of rabbit platelets induced by collagen and PAF, and especially against the aggregation induced by AA. However, when the aggregation induced by thrombin was studied, no effect was observed. Puberulins A, B, and C (Table 13) from Piper puberulum were tested for in vitro inhibition of PAF binding on isolated rabbit plasma platelet membrane [214]. Only puberulins A and C inhibited specific binding of [^H]-PAF to its receptor site, with IC50 values of 7.3 and 5.7 |LIM, respectively. On the other hand, piperbetol, methylpiperbetol, piperol A, and piperol B (Table 13), isolated from Piper betle, selectively inhibited the washed rabbit platelet aggregation induced by PAF in a concentrationdependent manner [215]. The IC50 values of the four lignans were 18.2, 10.6, 114.2, and 11.8 |imol/L, respectively, whereas for ginkgolide B (reference drug) the IC50 was 4.8 |amol/L. The concentration-response curve of PAF-induced platelet aggregation was shifled to the right by 50 |Limol/L of lignans, and the EC50 of PAF was increased by these compounds from 1.5 nmol/L to 14.3, 23.1, 2.4 and 20.6 nmol/L respectively, and 47.2 nmol/L for the reference drug. The compounds also inhibited the binding of ["^H]PAF to washed rabbit platelets, with IC50 values of 8.7, 5.3, 8.8, and 6.2 |imol/L, and 1.8 |Limol/L for the reference drug. However, the aggregation of washed rabbit platelets induced by threshold ADP and AA were not affected by the lignans. Furthermore, piperbetol, methylpiperbetol, piperol A, and piperol B had no effects on the cAMP contents in the rest of the washed rabbit platelets. Neojusticin A, justicidin B, taiwanin E methyl ether and taiwanin E isolated from Justiciaprocumbens inhibited platelet aggregation induced by AA with an IC50 range of 1.1-8.0 inM. These values are higher than that of aspirin (20.3 |LIM) but lower than that of indomethacin (0.21 |LIM) [216]. Based on this structural model of active lignans, Tanabe et al. [217] synthesised and studied the expected anti-platelet activity of justicidin E, but using a ^H-C16-PAF binding assay. The IC50 value of justicidin E was 100-150 |LiM, whereas the reference drug gave an IC50 of 3.2 |LIM.

244

Table 13.

Active Neolignans

Puberulin A Puberulin B Puberulin C

Ri OCH3 OH

R2 OCH3 OCH2O OCH3

R3 OCH3 OCH3 CH2CH=CH2

R4 CH2CH=CH2 CH2CH=CH2 OCH3

H2C

CH30,

OCH3

OCH3

CH3O

RiO

Methylpiperbetol Piperbetol Piperol A Piperol B

A A B A

Ri CH3 H CH3 CH3

R2 Ac Ac H H

From the fruits of Schisandra chinensis Jung et al [218] isolated three lignans with antagonistic activities of ["^H]-PAF to washed rabbit platelet receptor binding. Schisandrin A was the most potent isolated compound. From the same source, Lee et al. [219] isolated pregomisin and chamigrenal, two lignans with PAF antagonistic activity. The IC50 values were 4.8 x 10"^ M and 1.2 x 10'"^ M, respectively. Recently, the same authors [220] studied the structure-activity relationships of Schisandra chinensis' lignans and their derivatives as PAF antagonists, and some general features were observed. The introduction of an ester group at C-6, a methylene dioxy moiety or a hydroxyl group at C-7 tends to decrease the activity, and lignans with an /?-biphenyl configuration are more active than

245

those with an iS-form. A derivative of schisandrol A, 6(7)dehydroschisandrol A, was the most potent of the compounds tested, with an IC50 of 2.1 |LIM, whereas schisandrol A had an IC50 of 0.25 mM. The authors suggested that these Hgnans could be the active principles of this crude drug in PAF-related inflammatory disorders, such as asthma, allergy, atopic dermatitis and other inflammatory diseases. Two new Hgnans isolated from the flower buds of Magnolia fargesii and called magnone A and magnone B showed antagonistic activity against PAF in the [^H]-PAF receptor binding assay, with IC50 values of 3.8 x 10"^ M and 2.7 x 10"^ M, respectively [221]. Although the antagonistic activities of both Hgnans were weaker than that of the reference drug (ginkgolide B IC50 1.9 X 10'^), they are in the range of other related compounds.

H3C0'

Futoquinol

Futoenone

Effects on Metabolism and Cholesterol Level Sesame oil affects cholesterol mobility in the human organism. To study the potential activity of other kinds of compounds, Hirose et al [222] analysed the effects of sesamin, a lignan present in the oil, on various aspects of cholesterol metabolism, and they observed that a diet with sesamin reduced the concentration of serum and liver cholesterol except in the group free of cholesterol. Sesamin decreased lymphatic absorption of cholesterol and increased the fecal excretion of the neutral but not the acidic form. At the liver level, there was a significant reduction in the activity of microsomal 3-hydroxy-3-methylglutaryl coenzyme A reductase, but the activity of hepatic cholesterol 7a-hydroxylase and alcohol dehydrogenase were not affected. Microscopic histological examination did

246

not show any abnormalities, and the activity of serum transaminases (GOT and GPT) remained unchanged. The authors proposed further studies to ratify the possible use of sesamin as a hypocholesterolemic agent. In 1995, Ogawa et al [223] completed the previous study, and analysed the effects of sesamin and its epimer, episesamin, on cholesterol and lipid metabolism using normo- and hypercholesterolemic stroke-prone spontaneously hypertensive rats. In normocholesterolemic rats, both lignans increased the concentration of total cholesterol in serum by increasing HDL and decreased the serum concentration of very low density lipoprotein (VLDL). In the liver, only episesamin affected the microsomal enzymatic activity. In hypercholesterolemic rats fed a high-fat, high-cholesterol diet, only episesamin improved serum lipoprotein metabolism. In the liver both sesamin and episesamin significantly suppressed cholesterol accumulation, but only episesamin increased the activity of microsomal cholesterol 7ahydroxylase. These results confirm that sesamin may be effective in preventing cholesterol accumulation in the liver and episesamin may be effective in regulating of cholesterol metabolism in serum and liver. On the other hand, sesamin exerts its hypotriglyceridemic effect at least in part through an enhanced metabolism of exogenous-free fatty acid to oxidation at the expense of esterification in rat liver, as was demonstrated by Fukuda et al [224] working on ketone body production and lipid secretion in isolated perfused liver from rats. In fact, ketone body production increased in the livers perfused with oleic acid, and sesamin feeding caused a stimulation of ketone body production. In addition, the ratio of Phydroxybutyrate to acetoacetate, an index of mitochondrial redox potential, was increased in the oleic acid-perfused livers with respect to free fatty acid, though it was lowered by dietary sesamin. The cumulative secretion of triacylglycerol, but not of cholesterol, by the livers from sesamin-fed rats decreased, especially in the oleic acid administered group; this suggests an inverse relationship between the rates of ketogenesis and triacylglycerol secretion.

247

H3Q

.H

H3CO,

OCH3

H3CO

H3CO" OCH3

HsCC)

H3CO

Pregomisin

OCH3

6,7-Dehydroschisandrol A

Fifteen lignans isolated from the fruits of Schisandra chinensis, the leaves of Machilus thunbergii (Lauraceae), and the flower buds of Magnolia denudata were studied for their capacity to inhibit the microsomal acyl-coenzyme A (CoA):cholesterol acyltransferase. They inhibited the enzyme activity at different ranges of concentration. In order of decreasing potency the IC50 (l^M) values were gomisin N (25), gomisin K3 and licarin D (37), gomisin L2 (38), benzoylisogomisin 0 (47), gomisin J (51), schisantherin D (57), wuweizisu C (65), licarin A (75) and machilin G (81). The rest of the tested compounds had IC50 values higher than 100 |iM (gomisin A and B, angeloylgomisin H, tigloylgomisin P and schisandrin), and in the case of the reference drug used in this experiment the IC50 was 42 |aM. Some of these lignans may be useful lead compounds for the design of new acyl-CoA:cholesterol acyltransferase inhibitors, but further studies are necessary to verify the in vivo cholesterol lowering activity of these principles [225]. Sixteen synthetic arylnaphthalene lignans were evaluated for hypolipidemic activity in diet-induced hypercholesterolemia in rats. The total cholesterol and HDL cholesterol in plasma were measured as a reference of the activity [226]. 2,3-Bis(methoxycarbonyl)-l-(3,4dimethoxyphenyl)-4-hydroxy-6,7,8-trymethoxynaphtalene was the most active of the series assayed; it reduced serum cholesterol and increased HDL cholesterol. The effect was about 100 times higher than that of cholestyramine. The comparative studies of the results obtained with all of the assayed lignans made it possible to establish that the most relevant structure-activity relationships are as follows: 1) the introduction of a methoxycarbonyl group at the C-3 position increases the activity; 2) the conversion of the 2,3-dimethoxycarbonyls to a lactone diminishes the activity; 3) the elimination of a methoxy in A-ring decreases activity; 4) the

248 hydroxy aromatic group at C-4 of the B-ring is necessary for activity; 5) modifications of the 3,4-dimethoxy groups at the C-ring decreases the activity; and finally 6) the presence of a biphenyl group is essential for activity. A mechanistic study indicated that 2,3-bis(methoxycarbonyl)-l(3,4-dimethoxyphenyl)-4-hydroxy-6,7,8- trimethoxynaphtalene inhibits the intestinal absorption of both cholesterol and bile acids. The same research group [227] completed the study with a series of diesters of the arylnaphthalene lignan and their heteroaromatic analogues. Modifications at C-3 in the diesters improve the hypocholesterolemic and HDL cholesterol-elevating activities. Structure-activity analysis indicated that the 2-pyridylmethyl ester of 2-(methoxycarbonyl)-1 -(3,4-dimethoxy-phenyl)-4hydroxy-6,7,8-trymethoxy-3-naphtoic acid has the optimum activity both in hypocholesterolemic and HDL cholesterol-elevating properties. Flaxseed is the richest source of co-3 fatty acid and lignans and can reduce the levels of oxygen free radicals and hence prevent the development of hypercholesterolemic atherosclerosis. Prasad [228] studied the effects of flaxseed on a high cholesterol diet, which increases the serum level of total cholesterol without altering the levels of serum triglycerides. These changes are associated with a marked development of atherosclerosis in the aorta. Flaxseed reduced the development of aortic atherosclerosis by 46% without significantly lowering the serum cholesterol. Modest dietary flaxseed supplementation was effective in reducing hypercholesterolemic atherosclerosis markedly without lowering serum cholesterol. Its effectiveness against this disorder could be due to suppression of enhanced production of oxygen free radicals by PMNL in hypercholesterolemia. As the anti-atherogenic activity of flaxseed could be due to its a-linolenic acid and/or lignan content, Prasad et al [229] studied the different effects of two kinds of flaxseed: type I flaxseed with 51-55% a-linolenic acid in its oil and rich in plant lignans, and type II flaxseed, with similar oil and lignan content but a very low a-linolenic acid content (2-3% of the total oil). The results indicate that the anti-atherogenic activity of the type II flaxseed is not due to a-linolenic acid. For this reason it is possible to hypothesise that the role of lignans is more important than that of the unsaturated fatty acid. In a complementary study [230], the same authors investigated the effects of secoisolariciresinol diglycoside on various blood lipid and aortic tissue oxidative stress parameters and on the development of atherosclerosis in rabbits fed a high-cholesterol diet. The results are in agreement with the hypothesis about the effects of lignans, because secoisolariciresinol

249 diglycoside reduced hypercholesterolemic atherosclerosis and this effect was associated with a decrease in serum cholesterol, LDL-cholesterol, and lipid peroxidation products and an increase in HDL-cholesterol and antioxidant reserve. Antioxidant role of Lignans and Their Effects on Liver Function Antioxidative Properties and Antiperoxidative Effects.

In an interesting review on antioxidants and free radical scavengers of natural origin, Potterat [231] covers the principal antioxidative lignans. The exceptional stability of sesame oil from Sesamum indicum (Pedaliaceae) to oxidative deterioration led to the isolation of antioxidative lignans such as sesamol, together with several bisfuranyl lignans such as sesamolinol, sesaminol and pinoresinol. In addition, other antioxidative lignans were isolated and identified from other sources, such as NDGA from Larrea tridentata, dihydroguaiaretic acid, guayacasin and isopregnomisin from Porlieria chilensis (Zygophyllaceae), gomisin N from Schisandra chinensis, cinnamophilin from Cmnamomum philipense, kadsurin, kadsurenone and burchellin from Kadsura heteroclita, and magnolol and honokiol from Magnolia officinalis. The antioxidant activity of a series of lignans was studied by Faure et al [232] in a rat brain homogenate autoxidation test, and lipid peroxidation was evaluated from the luminescence intensity and thiobarbituric reactive product accumulation. Isopregomisin (0.7 JLLM) was the most active lignan, followed by guayacasin (I.IJLIM) and dihydroguaiaretic acid (2.8 |LtM). The difference in the effect can be explained by the degree of methylation, which increases the activity. Enterolactone, prestegane B and 2,3-dibenzylbutane-l,4-diol [233,234] were studied as antioxidatives on superoxide production and luminoldependent chemiluminescence response in human PMNL. None of the three lignans had a direct effect on the responses of human PMNL. Prestegane B and 2,3-dibenzylbutane-l,4-diol enhanced the superoxide production and luminol-dependent chemiluminescence response induced by formyl-Met-Leu-Phe (fMLP), but the effects of the latter were stronger than those of the former. Enterolactone inhibited fMLP-induced effects. The results suggest that activation of phospholipase A2 (PLA2) and Ca^^-

250 calmodulin-pathways may be involved in the effect of 2,3-dibenzylbutanel,4-d^ol but that activation of Hpoxygenase (LOX), cyclooxygenase (COX) or protein kinase C (PKC), and PAP release are not [235]. CH3 HaCO,,^^^

y

1 HI-

H3CO

X/-^"^

i /-=r\ H3C0

H3CO

o •"•

o

""-yj H3C0

OCH3

HO Sesamolinol

Schisanhenol

Seven of the nine lignans isolated from Schisandra chinensis, S. rubriflora and Kadsura longipedunculata inhibited the ascorbic acid/ NADPH induced lipid peroxidation of rat liver microsomes, iron/cysteineinduced lipid peroxidation of rat liver microsomes and superoxide anion production in the xanthine/xanthine oxidase system [236,237]. Schisanhenol, iS'-(-)-schizandrin C and iS'-(-)-schizandrin B were more potent than vitamin E at the same concentration (1 mM). Schisanhenol and schizandrin B inhibited gossypol-induced superoxide anion generation in rat liver microsomes. Oral administration of schisanhenol and schizandrin B reduced liver malondialdehyde (MDA) formation induced by ethanol and increased superoxide dismutase (SOD) and catalase activities in rat liver cytosol. The authors concluded that this kind of lignan has strong antioxidant activity. The effects of schisanhenol and schizandrin on the peroxidative damage of ageing and ischemic rat brain were studied by Xue et al. [238]. Schisanhenol at 10""* M completely inhibited the peroxidative damage to rat brain mitochondria membrane induced by Fe^^-cysteine, and prevented swelling and disintegration of brain mitochondria as well as reduction in brain membrane fluidity. On the other hand, schisanhenol significantly impeded production of MDA and loss of ATPase activity induced by reoxygenation following anoxia in an experiment on ischemia and reperfusion of brain mitochondria and membrane in vitro. In vivo, it increased cytosol GSH-peroxidase in mouse brain during reoxygenation following anoxia. Schisandrin had similar activity but was less potent.

251 These results demonstrate that both compounds have protective action against oxidative stress. In a later study [239], schisanhenol was tested in rat pancreatic islets against the inhibitory action of alloxan on glucosestimulated insulin release and had partially protective effects. In islets reincubated for 30 min in the presence of schisanhenol, the ratio between islets exposed to alloxan during the last 15 min of pre-incubation and those not exposed to the diabetogenic agent with respect to insulin output over 90 min incubation in the presence of D-glucose averaged 142% of that recorded under the same experimental conditions in islets pre-incubated in the absence of schisanhenol. The results demonstrated the potential benefit of schisanhenol as a preventive or therapeutic tool in situations characterised by oxygen radical-induced damage. Wuweizisu B from Schisandra chinensis showed strong protective effects on lipid peroxidation damage to the surface of cultured hepatocytes of rats treated with Fe^Vcystein and also scavenger free radical properties [240]. On the other hand, the oral administration of wuweizisu B increases the activities of antioxidant enzymes. Three new benzofiiran lignans were isolated from Schizonepeta tenuifolia (Labiatae) and called schizotenuins A, Ci and C2. Based on these chemical structures, Maeda et al [241] synthesised a series of related compounds that were then tested for their inhibitory effects on lipid peroxidation in rat brain homogenate and rat liver microsomes. Three of the twelve compounds tested, showed prominent inhibitory activity in rat brain homogenate, with IC50 values of 1.20 (compound-8), 0.70 (compound-9) and 0.77 ^iM (compound-10), respectively (Table 14). Compounds 8 and 10 were tested in rat liver microsomes and gave IC50 of 3.66 |j,M and 4.49 |LIM respectively, and they were found to be more potent than schizotenuin A (IC50 = 36.26 |j,M) and much more potent than that of (±)-a-tocopherol (IC50 = 976 |LiM). In a second study [242], the same authors obtained a new series of related compounds by an oxidative coupling reaction of methyl (£)-3-(4,5-dihydroxy-2-methoxyphenyl) propenoate and obtained four active principles, compounds 11-14. The IC50 obtained on lipid peroxidation in rat liver microsomes were 1.13, 1.20, 0.95 and 0.89 |j.M respectively, and in rat brain homogenate were 0.32, >1.00, 1.02 and 0.58 |iM respectively, which means that all of them are more potent than the compounds previously described.

252

OH Schizotenuin A

HOOC.

OH

OH Schizotenuin Ci HO.

COOH

OH

OH

OH

Schizotenuin C2

Gomisin C has an inhibitory effect on the respiratory burst of rat neutrophils in vitro [243]. The mechanism of action may be mediated partly by the suppression of NADPH oxidase and partly by the decrease in cytosolic Ca^"^ released from an agonist-sensitive intracellular store. In fact, gomisin C attenuated the activity of TPA-activated neutrophil particulate NADPH oxidase in a concentration-dependent manner and reduced the increase in cytosolic free Ca^^ in neutrophils stimulated by fMLP in presence or absence of ethylenediaminetetraacetic acid (EDTA). In addition, this study suggests that the gomisin C mechanism is not mediated by changes in cellular cAMP or in inositol phosphates, or by scavenging

253

superoxide anion O2" released from neutrophils, because it had no effect on O2' generation and uric acid formation in the xanthine-xanthine oxidase system and failed to alter O2" generation during dihydroxyfumaric acid autoxidation. Table 14.

Active Benzofuran Lignans C00CH3 H3C00C.

8 9

10

Methyl (£)-3-[2-(3,4-diacetoxyphenyl)-7-acetoxy-3methoxycarbonyl-benzofiiran-5-ylJpropenoate Methyl (£)-3-[2-(3,4-dihydroxyphenyl)-7-hydroxy-3methoxycarbonyl-benzofuran-5-yl]propenoate Methyl (£)-3-[2-(3,4-dihydroxyphenyl)-7-methoxy-3methoxycarbonyl-benzofiiran-5-yl]propenoate

Ri OCOCH3

R2

R3

OCOCH3

OCOCH3

OH

OH

OH

OCH3

OH

OH

The lignan glucosides obtained from germinated sesame seeds inhibited in a dose-dependent manner the lipid peroxidation reaction of linoleic, linolenic and arachidonic acids induced by an H202/FeCl2 system, and also suppressed enzymatic (ADP/NADPH/FeS04) and nonenzymatic (H202/FeCl2) lipid peroxidation by microsomes [244]. In an in vivo experiment, the lipid peroxidation activity measured as 2-thiobarbituric acid reactive substances v^as significantly lower in the kidneys and liver of sesamolin-fed rats than in the controls [245]. In addition, the amount of 8hydroxy-2*-deoxyguanosine excreted in the urine was significantly lower in the sesamolin-fed rats. These results suggest that sesamolin and its metabolites, sesamol and sesamolinol, may contribute to the antioxidative properties of sesame seeds and oil and support our hypothesis that sesame lignans reduce susceptibility to oxidative stress. In a recent study Ashakumary et al [246] demonstrated that dietary sesamin increase both mitochondrial and peroxisomal palmitoyl-CoA oxidation rates. In addition, sesamin increased the hepatic activity of fatty acid oxidation enzymes, including carnitine palmitoyltransferase, acyl-CoA dehydrogenase, acyl-

254

CoA oxidase, 3-hydroxyacyl-CoA dehydrogenase, enoyl-CoA hydratase, and 3-ketoacyl-CoA thiolase. Dietary sesamin also increased the activity of 2,4-dienoyl-CoA reductase and A^,A"^-enoyl-CoA isomerase, enzymes involved in the auxiliary pathway for P-oxidation of unsaturated fatty acids [246]. On the other hand, the results obtained by Fukuda et al, [247] suggest that increased fatty acid oxidation by dietary sesamin leads to decreased esterification of fatty acids and reduces the synthesis and secretion of triacylglycerol. 0CH3

^COOCHj

CH3COO"

^

>r

"^COOCHs

OCOCH3 OCOCH, Compound-11 (R=H) Compound-12 (R=CH3CO) CHaOOC^ ^ ^ ' V .

Compound-13

^^N.

^0^.

^C00CH3

OCOCH3

CH3O

^-^

OCOCH3

Compound-14

Microsomal and mitochondrial lipid peroxidation induced by Fe^^ADP/NADH were inhibited by honokiol and magnolol, neolignans isolated from Magnolia obovata. Honokiol was more potent than magnolol as an antioxidative (IC50 8.7 versus 53.4 |iM on microsomal lipid peroxidation, and 7.0 and 11.3 |LIM on mitochondrial lipid peroxidation, respectively). The neolignans protected mitochondrial respiratory chain enzyme activity against NADPH-induced peroxidative stress and protected red cells against oxidative hemolysis [248].

255

From an active extract from Betula platyphylla var. japonica (Betulaceae), Matsuda et al [249] isolated four known lignans together with other compounds. They were tested as hepatoprotective agents on the liver injury induced by D-galactosamine/lipopolysaccharide or CCI4 and for O2 "-scavenging and antioxidative activities. The glucoside of (+)lyoniresinol and nudiposide had marked inhibitory activity on lipid peroxidation, giving inhibition values of 97 and 87% at 200 |Lig/mL, respectively, and 67 and 59% at 20 |Lig/mL, respectively. In the hepatoprotective test, nudiposide gave a weak effect only at 100 \xM (27% inhibition), whereas (+)-lyoniresinol 3a-0-P-D-glucopyranoside gave a weak effect but it was not dose-dependent (22-27%o inhibition at doses of 3100 laM). Other antioxidative compounds of interest are syringaresinol and medioresinol, which were determined by the ferric thiocyanate and thiobarbituric acid methods [235]; gomisin J inhibited Fe^Vascorbic acid and ADP/NADPH-induced lipid peroxidation in rat liver mitochondria, with IC50 of 5.5 and 4.7 |Limol/L, respectively [250]; vibsanol and 9'-0methylvibsanol isolated from the wood of Viburnum awabuki (Caprifoliaceae) had antioxidative properties in rat brain homogenate [251]; (-)-phylligenin from Orophea enneandra displayed antioxidative and radical scavenging properties against the 2,2-diphenyl-l-picrylhydrazyl radical in bioautographic TLC assays [43]; 3,4-/>/.v(4-hydroxy-3-methoxybenzyl) tetrahydrofiiran from Pandanus odoratissimus (Pandanaceae) exhibited strong antioxidative activities in the thiocyanate method [252]. Effects of Lignans on Liver Function

The flavolignans of Silybum marianum (Compositae) fruit are known antihepatotoxic agents used therapeutically. Hikino et al [253] used CCI4and galactosamine-induced cytotoxicity in primary cultured rat hepatocytes and a transaminase activity inhibition test to demonstrate the hepatoprotective effects of silandrin, silybin, silychristin, 3-deoxysilychristin, silydianin and silymonin. All of them reduced the cytotoxic effects of CCI4 and galactosamine at 1 mg/mL, but only 3-deoxysilychristin was significantly active at 0.01 mg/mL. Using the same experimental model, Hikino et al, [254] studied twenty-two lignans isolated from Schisandra chinensis and Kadsura japonica for their antihepatotoxic activity.

256

Wuweizisu C and schisantherin D were the most active on the CCU-test, with significant activity at 0.01 mg/mL, and deoxygomisin A, gomisin N, wuweizisu C, gomisin C and schisantherin D proved to be the most active on the galactosamine test at the same dose. The authors concluded that a methylenedioxy group in the dibenzocyclooctane skeleton is the principal feature in the antihepatotoxic activity of this kind of lignan. None of the tested compounds modified transaminase activity directly. In addition, gomisin A showed activity at 1 mg/mL on the CCU test and at 0.01 mg/mL on the galactosamine test. Maeda et al [255] reported the effects of gomisin A on liver functions in various experimental liver injuries and on bile secretion in CCU-induced liver injury. Gomisin A facilitated the liver function in normal and liver injured rats, and protected the liver against CCU-induced cholestasis, maintaining the bile flow and biliary output of each electrolyte nearly at the level of the vehicle-treated group, but did not affect biliary output of total bile acids. On the other hand, gomisin A inhibited the increase in the serum bile acid concentration induced by administration of deoxycholic acid, but hardly influenced the serum bile acids in the phenobarbital combined group [256]. The authors conclude that the inhibitory effect of gomisin A on the promoting action of deoxycholic acid is due to its improving bile acid metabolism, but the effect could not be elucidated from the metabolism of bile acids. In another study Ohkura et al [257] explained that the inhibitory effect of gomisin A on liver injuries is the result of its preventing the AA release and decreasing the LT production (see section about inflammation). In a study on immunologically induced liver injuries in vivo and in vitro [258], gomisin A inhibited dose-dependently the hepatotoxic chemicalinduced liver injuries and the mortality of mice with acute hepatic failure. It also suppressed necrosis, but the infiltration of non-specific inflammatory cells was not affected. It inhibited the isolated liver cell injuries induced by antibody-dependent cell-mediated cytotoxic reactions or activated macrophages in vitro. The results suggest that gomisin A can be protective against immunological liver injuries. In a related work, Nagai et al. [259] studied the hepatoprotective effect of gomisin A on three immunological liver injury models in mice, and they demonstrated that gomisin A inhibited the elevation of transaminase (GOT and GPT) activities and showed a tendency to inhibit histopathological changes in the liver in all the models. Moreover, the lignan inhibited deoxycholic acid-induced release of transaminase from cultured rat hepatocytes in vitro, but did not affect the

257

formation of hemolytic plaque forming cells in immunised mice spleens and the hemolytic activity of guinea pig complement in immunohemolysis reactions. Therefore, the hepatoprotective effect of gomisin A could have to do With the protective effect of hepatocyte plasma membrane rather than the inhibition of the antibody formation and complement activity. Gomisin A stimulated liver regeneration after partial hepatectomy by enhancing ornithine decarboxylase (ODC) activity [260], which is an important biochemical event in the early stages of liver regeneration. In addition, gomisin A enhanced the mitotic index and the level of DNA synthesis increased after partial hepatectomy. Moreover, the ODC activity increased in the early stages of liver regeneration, as did hepatic putrescine. On the other hand, in the liver from rats simultaneously treated with 3'-methyl-4dimethylaminoazobenzene (3*-MeDAB), gomisin A inhibited both the increase in number and size of glutathione S-transferase placental form (GST-P)-positive foci, a marker enzyme of preneoplasm, and the population of diploid nuclei, as a proliferative state of hepatocytes. The lignan inhibited the hepatocarcinogenesis induced by 3'-MeDAB by enhancing the excretion of the carcinogen from the liver and by reversing the normal cytokinesis [261]. Two years later, the same authors [262] investigated the relationship between the serum concentration of bile acids, the appearance of preneoplastic change, GST-P-positive foci in the liver of rats fed 0.06% 3'-MeDAB, and the effects of gomisin A. The increase in serum bile acids, especially deoxycholic acid, and the appearance of preneoplastic lesions, and the number and area of GST-P-positive foci in the liver were significantly inhibited by simultaneous oral administration of gomisin A (30 mg/kg). The results confirmed that deoxycholic acid is an endogenous risk factor for hepatocarcinogenesis and suggest that the antipromoter effect of gomisin A is based on its improving the metabolism of bile acids, including deoxycholic acid. These results were corroborated by Nomura et al [263,264], who suggested that the effects of gomisin A are related to improved liver fimction and reversal of abnormal ploidization, and that gomisin A may be a candidate for a chemopreventive drug that would inhibit the promotion process in hepatocarcinogenesis. Gomisin A protects the liver from injury by acetaminophen. One of its possible mechanisms involves the suppression of lipid peroxidation [265]. It inhibited not only the elevation of serum aminotransferase activity and hepatic lipoperoxide content, characteristic of acetaminophen administration, but also the appearance of histological changes such as degeneration

258

and necrosis of hepatocytes. However, gomisin A did not affect the decrease in liver GSH content. A histological analysis indicates that the massive necrosis and vacuolisation in the liver of rats treated with both acetaminophen and gomisin A were reduced in comparison with rats treated with acetaminophen only [266]. Gomisin A quickly induced the hepatocyte growth factor mRNA expression through mechanisms different from those involved in acetaminophen-induced liver injury. In three experiments, Ko et al [267-269] demonstrated the protective effect of a lignan-enriched extract of Fructus Schisandrae {Schisandra chinensis fruits) against physical exercise-induced muscle damage in rats and the liver damage induced by different agents. Pre-treatment of rats with the lignan-enriched extract caused a moderate enhancement of hepatic GSH regeneration capacity in control rats treated with /-butyl hydroperoxide, but this effect was greatly increased after CCU challenge. This was manifest in the increase in the hepatic GSH level and activities of hepatic glucose-6phosphate and GSH reductase, as well as by a decreased susceptibility of hepatic tissue homogenates to in vitro peroxide-induced GSH depletion. This effect was more evident after CCU challenge, because the lignanenriched extract caused a dose-dependent protection against the CCUinduced impairment in hepatic GSH status, thus decreasing the activities of MDA levels in tissue and alanine aminotransferase plasma activities. On the other hand, pre-treatment with the lignan-enriched extract protected against physical exercise-induced muscle damage in rats. This protection was associated with enhancement of the hepatic antioxidant status, whereas pre-treatment with a-tocopherol acetate decreased the MDA level in skeletal muscle but did not protect against exercise-induced muscle damage or improve hepatic antioxidant status in exercised rats. The protective effect that pre-treatment of rats with the lignan-enriched extract has on the muscle damage produced by physical exercise may be due to an enhancement of hepatic GSH status that provides GSH for effective antioxidant protection of skeletal muscle during exercise. Li [240] reviewed the activity of lignans from the Schisandraceae family {Schisandra and Kadsura) and cited the effects of eighteen lignans against CCI4 induced hepatotoxicity, including the relationship between the functional group and the stereostructures. The activities of ^S-enantiomers of wuweizisu B and C were stronger than their i?-forms, where the twist boat chair configuration of cyclooctadiene seems to be more active than the twist boat conformation. The compound with two methylenedioxyl groups is

259 more active than the ones with only a methylenedioxyl group. The compound with a hydroxyl in the cyclooctadiene ring showed only weak inhibition of MDA formation. H3CO.

Tribulusamide A (R=H2) Tribulusamide B (R=0)

Tribulusamides A and B, new lignanamides isolated from the fruits of Tribulus terrestris (Zygophyllaceae) prevented cell death induced by Dgalactosamine/tumor necrosis factor-a (TNFa) after addition to primary cultured mouse hepatocytes [270]. This experimental model is known to cause fatal liver failure, and in the case of the lignamide groups the cell survival rate rose to 70-106% that of the normal group at a concentration range of 50-200 |LIM. However, at higher doses the lignamides did not improve the cell survival rate because of their cytotoxicity. Two new dibenzocyclooctadiene type lignans called heteroclitins F and G and related compounds isolated from the stems of Kadsura heterodita inhibited lipid peroxidation in the rat liver homogenate stimulated by Fe^^ ascorbic acid, CCI4, NADPH and ADP-NADPH [271]. 7>a«5-kielcorin and ^ram-isokielcorin B were effective in preventing perturbation of cell GSH homeostasis, as revealed by measuring reduced and oxidised GSH, lipid peroxidation and cell viability after inducing toxicity with or without the studied compounds at different concentrations [272]. The lignan (+)-sesamin and related lignans present in sesame seeds or its oil are specific inhibitors of A^-desaturase in polyunsaturated fatty acid biosynthesis in both microorganisms and animals [273,274]. The results obtained in experiments with both a cell-free extract of the fungus Mortierella alpina and rat liver microsomes demonstrated that (+)-sesamin

260

is a non-competitive inhibitor of A^-desaturase at low concentrations (K\ for rat liver, 155 |iM), but does not inhibit A^-, A^- and A^^-desaturases. (+)Sesamolin, (+)-sesaminol, (+)-episesamin, (-)-asarinin and (-)-epiasarinin also inhibited only A^-desaturase of the fungus and liver. In an analogous study Umeda Sawada et al [275] demonstrated that sesamin inhibits the A^desaturation of n-6 fatty acid (dihomo-y-linolenic acid to AA), but not that of n-3 fatty acid to eicosapentaenoic acid in rat livers. Sesamin administration decreased incorporation of eicosapentaenoic acid and simultaneously increased the AA content in the liver. Changes in the hepatic concentration of eicosapentaenoic and linolenic acids (n-3) were significantly reduced by their simultaneous administration with sesamin, but sesamin had no effect on the n-6 and n-9 fatty acid concentrations. However, no significant differences in lymphatic absorption between eicosapentaenoic acid (n-3) and AA (n-6) were observed in presence or absence of sesamin [276]. In addition, sesamin, reduced the A^-desaturase activity in non-neoplastic cell cultures and could be used to manipulate the content of series 1 and 2 PGs derived from dihomo-y-linolenic acid and AA. However, it had little effect on the content and ratio of dihomo-ylinolenic acid and AA in neoplastic cells, which means that A^-desaturase activity was low in these cells [277]. In a complementary study, Umeda Sawada et al. [278] demonstrated that dietary sesame lignans promote ketogenesis and reduce polyunsaturated fatty acid esterification into triglyceride. They demonstrated that sesame lignans inhibited changes in the n-6/n-3 ratio by reducing the hepatic polyunsaturated fatty acid content, and this reduction may occur because of the effects of sesame lignans on polyunsaturated fatty acid degradation (oxidation) and esterification. Sesamin and episesamin can improve liver function. In studies on experimental models in rodents, Akimoto et al. [279] demonstrated the protective effects of these lignans against liver damage caused by alcohol and e c u . They improved mouse blood parameters, such as aspartate aminotransferase and alanine aminotransferase activities, and the concentrations of total cholesterol, triglycerides and total bilirubin that had been pathologically increased as a result of continuous inhalation of ethanol. In addition, sesamin showed a significant protective effect against the accumulation of fat droplets and vacuolar degeneration in the mouse liver, as confirmed by histological examination. The effects of a mixture of sesamin and episesamin (1:1) on ketone body production and lipid secretion were studied by Fukuda et al. [224] in isolated perfused liver from rats.

261 They demonstrated that dietary lignans exert their hypotriglyceridemic effect at least in part through enhanced metaboHsm of exogenous free fatty acid in rat Hver. The cumulative secretion of triacylglycerol, but not of cholesterol, by the liver rats decreased, especially when exogenous oleic acid was provided, suggesting an inverse relationship between the rates of ketogenesis and triacylglycerol secretion. Sesame seed contains y-tocopherol, a compound that has vitamin E activity 20% lower than that of a-tocopherol, but a sesame seed diet has high vitamin E activity. Yamashita et al [280] studied this paradox using changes in red blood cell hemolysis, plasma pyruvate kinase activity, and peroxides in plasma and liver as indices of vitamin E activity, and groups of rat which were fed with four different diets: vitamin E-free, a-tocopherol, Y-tocopherol and sesame seed. In a second series, two diets containing sesame lignan (sesaminol or sesamin) and y-tocopherol were tested. The global results of the experiments indicate that the vitamin E activity of ytocopherol in sesame seed is equal to that of a-tocopherol because of a synergistic interaction with sesame seed lignans. In a second paper and using the same methods, Yamashita et al [281] demonstrated that sesame seed and its lignans induce a significant increase in a-tocopherol content in the blood and tissue of rats. Supplementation with even 5% sesame seed increased the a-tocopherol content of plasma and liver produced by a low a-tocopherol diet. These results indicate that lignans enhance vitamin E activity in rats fed a low a-tocopherol diet and increase the concentration of a-tocopherol in blood and tissues. In a complementary study, Kamal et al [282] demonstrated that the bioavailability of y-tocopherol is enhanced in sesamin-containing diets as compared with purified diets. Sesamin feeding increased y-tocopherol and the y-/a-tocopherol ratios in plasma, liver and lung, but the increase in a-tocopherol was non-significant. Thus, sesamin appears to conserve y-tocopherol in rat plasma and tissues, and this effect persists in the presence of a-tocopherol, a known competitor of ytocopherol.

262 Effects on Inflammation, Immunity and Their Mediators Anti-inflammatory Effects

Inflammation involves a series of events that can be elicited by physical injuries, infectious agents, antigen-antibody reaction, and other agents. There are different phases in the inflammatory process, depending on the agents, time or intensity: the acute phase characterised by vasodilatation and edema, the subacute phase with leukocyte and phagocytic cell infiltration, and the chronic phase with tissue degeneration and fibrosis. Depending on the phase, there are different kinds of mediators: metabolites of AA such as prostaglandins (PGs) and leukotrienes (LTs), lipidic mediators like PAF, peptides such as interleukines (ILs), TNF, granulocyte/macrophage colony stimulating factor (GM-CSF), and others. Some of these increase gene expression, probably by the activation of transcription factors, such as the nuclear factor KB (NF-KB). Anti-inflammatory compounds can act at different levels. They can, for example, decrease vasodilatation and edema, as in the case of antihistaminic drugs; inhibit PGs synthesis as do the non-steroidal antiinflammatory drugs; modify synthesis or block the receptors of mediators such as PAF or LTs; interfere with the functions of peptide mediators; block the transcription factors or scavenge free radicals generated in the process. Natural products as anti-inflammatory agents may act at different levels, some by non specific pathways as antioxidants do or by a specific mechanism via receptor antagonism. The lignans can act in both cases. Few reports on the experimental in vivo anti-inflammatory activity of lignans have been published. In their 1984 review of the pharmacological activities of lignans MacRae and Towers [9] included no references to antiinflammatory activity. Kimura et al [283] studied the effects of lignans and neolignans on adjuvant-induced inflammation in mice. When administered into the pouch, the neolignans magnoshinin and magnosalin inhibited the granuloma tissue formation, but had no effect on the fluid volume. However, the lignans assayed had no activity. These results suggest that neolignans may act on the mechanism of granuloma tissue formation rather than on blood vessel permeability. Magnoshinin had stronger activity than the other neolignan, but its effects were not enhanced when the dose was increased. After oral administration, magnoshinin showed a stronger effect.

263 similar to that obtained with hydrocortisone at 60 mg/kg/day, but the neolignan only affected granuloma formation. 0CH3

H3C0.

H3C0.

^()CH3

H3C0 0CH3

OCH3

H3C0'

OCH3

Magnosalin

Magnoshinin

From the methanol extract of Haplophyllum hispanicum (Rutaceae) two topical anti-inflammatory aryl naphthalide lignans were isolated and identified as diphyllin acetyl apioside and tuberculatin [284]. The methanol extract was only active against the acute TPA-induced ear edema, whereas it was not active against the chronic inflammation induced by TPA or oxazolone-induced contact-delayed hypersensitivity in mouse ears. Diphyllin acetyl apioside was the most active of the isolated compounds on the acute TPA, with an ID50 of 0.27 |amol/ear, whereas tuberculatin gave an ID50 of 1.23 |Limol/ear. Yasukawa et al [285] tested seven dibenzocyclooctadiene lignans against inflammation induced by application of 1 \ig of TPA/mouse ear, and gomisin A, gomisin J, and wuweizisu C inhibited the edema. The ED50 range of these compounds for TPA-induced inflammation was 1.4-4.4 jiimol, and gomisin A was the most active compound. Furthermore, at 5 lamol/mouse, gomisin A suppressed the promoter effect of TPA on skin tumor formation in mice following initiation with 7,12-dimethylbenz[a]anthracene. The authors concluded that the inhibition of tumor promotion by gomisin A is related to its anti-inflammatory activity. Baba et al [286,287] studied the structure-activity relationships of some lignans as anti-inflammatory agents and in a second paper gave detailed information on the bone resorption inhibitory effects of these compounds. Chemical modification of the potent bone resorption inhibitor justicidin was performed and various naphthalene lactones, quinoline lactones and

264 quinoline derivatives bearing an azole moiety at the side chain were prepared. Antiallergenic Effects

In a study published in 1991 it was reported that Magnolia salicifolia flower bud extracts had an antiallergenic effect on the passive cutaneous anaphylaxis test. By monitoring their activities with an in vitro bioassay system measuring inhibitory effects on induced histamine release from rat mast cells, Tsuruga et al [288] isolated different active constituents, including lignans, which were evaluated for their biological activities using the in vitro bioassay. Although magnosalicin was not so active as an inhibitor of histamine release, the authors tested a large number of lignans from different structural types. Bistetrahydrofiirans and butanolides had activity, whereas tetrahydrofuran type lignans had no activity. The IC50 range was between 18 and 70 ^iM, but no chemical structure and pharmacological relationship was established. HjCO^

^^N.

.OCH3

H3CO ^0CH3

H3CO"

""^

*'0CH3

Magnosalicin

In an experiment on the ethanol-induced modulation of immune indices related to food allergy, sesamin and sesaminol affected the plasma levels of immunoglobulins (Ig) and eicosanoids. Sesamin increased the IgG level, but IgE was not affected. On the other hand, sesaminol decreased the plasma PGE2 concentration [289]. Yangambin is a flirofuran lignan with PAF antagonist properties that seems to also have anti-allergic properties. In a study with actively sensitised or normal rats, Serra et al [290] demonstrated that yangambin inhibited the pleural neutrophil and eosinophil infiltration evoked by injection of PAF or antigen and the pleural neutrophil infiltration triggered

265 by LTB4 in both group of rats, as well as the blood thrombocytopenia and intestinal anaphylaxis elicited by antigen in rats. However, plasma exudation evoked by both stimuli was unaffected. In addition, yangambin attenuated the hemoconcentration, thrombocytopenia, and leukocytosis observed after /.v. administration of PAF. At 10" and 10""^ M this lignan inhibited the anaphylactic contraction of longitudinal jejunal segments in response to antigen challenge and the contraction of jejunal segments from normal rats to PAF, but the response to serotonin (5-HT) was not affected. These findings indicate that yangambin is an antagonist of receptors other than those of PAF, such as LTB4, and it may be an important therapeutic tool in the management of some allergic diseases. Arachidonic Acid Metabolism

Gomisin A, a lignan present in Schisandra chinensis fruits, suppressed the production of LTB4 but did not affect 5-LOX or PLA2 activities [257]. The release of AA from macrophages stimulated with fMLP or the Ca^^ ionophore A23187 was suppressed by treatment with gomisin A. The results obtained in this work indicate that gomisin A inhibits the biosynthesis of LTs by preventing the release of AA, and the authors explain its inhibitory effect on liver injuries as the result of its preventive effect on the AA cascade due to the role of LTs in inflammatory liver diseases. Justicidin E from Justicia procumbens is a potent non-redox inhibitor of 5-LOX activity that inhibits the biosynthesis of LTB4 by human leukocytes with an IC50 = 70 nM [291]. Of the different semisynthetic principles, only the 5-phenylpyridyl derivative was more potent (about three times more) than justicidin E as a LOX inhibitor. The lignan compound-15 inhibits the oxidation of AA to 5-hydroperoxyeicosatetraenoic acid (5-HPETE) by 5-LOX (IC50 = 14 nM) and the formation of LTB4 in human PMNL (IC50 = 1 5 nM) as well as in human whole blood (IC50 = 50 nM). This lignan is a selective, direct nonredox 5LOX inhibitor (FLAP) showing no significant inhibition of human 15LOX, porcine 12-LOX or binding to human 5-LOX-activating protein up to 10 |iiM. The open form of the lactone (carboxylate) obtained by the lignan is well absorbed in the rat and is transformed into the active form. In addition, the carboxylate is orally active in the rat pleurisy model (ED50 =

266 0.6 mg/kg) and in the functional model of antigen-induced bronchoconstriction in allergic squirrel monkeys. Pre-treatment at 0.3 mg/kg of active carboxylate produced a 95% inhibition of the increase in airway resistance and a 95% inhibition of the decrease in dynamic compliance [292].

^-'"X

Compound-15 7-[[3-(4-Methoxy-tetrahydro-2//-pyran-4-yl)phenyl]methoxy]-4-phenylnaphtho[2,3-c]fijran-l(i//)-one

The tetrahydropyranyl naphthalenic lignan lactone called L-702,539 is a potent, selective 5-LOX nonredox inhibitor transformed in vivo from the hydroxy acid L-702,618 by an enzymatic process [293]. Studies with microsomes from genetically engineered human cell lines expressing individual cytochrome P-450 indicate that the isozyme responsible for the metabolism at the tetrahydropyran ring was P-450(3 A4). Another derivative was synthesised from this compound and called L-708,780 [294]. It inhibited the oxidation of AA to 5-HPETE by 5-LOX (IC50 = 190 nM) and the formation of LTB4 in human PMNL (IC50 = 3 nM) as well as in human whole blood (IC50 = 150 nM). In addition, it was orally active in the functional model of antigen-induced bronchoconstriction in allergic squirrel monkeys (95% inhibition at 0.1 mg/kg). Cinnamophilin proved to be a selective thromboxane A2 (TXA2) receptor antagonist especially in rat aorta with voltage-dependent Ca^^ channel blocking properties [295]. Cinnamophilin (10 |LIM) protected against the irreversible vasoconstriction of rat aorta caused by 9,11dideoxymethanoepoxy-9a,lla-PGF2a (U-46619) at 0.05 |j,M and showed voltage-dependent Ca^^ channel blocking action. At 30 |LIM it produced a slight relaxation of NA-induced tonic contractions, but this relaxation was abolished in presence of nifedipine. At 100 |j,M, cinnamophilin did not

267 affect the aortic contraction induced by endothelin-1, angiotensin II, carbachol or serotonin, but cAMP and cGMP levels did not increase in rat aorta. In an in vitro study [296] with human platelets, rat isolated aorta and guinea-pig isolated trachea and in vivo in mice and guinea-pigs, cinnamophilin inhibited in a dose dependent manner the human plateletrich plasma aggregation induced by AA, collagen and U-46619, with IC50 of 5.0, 5.6 and 3.0 |LIM, respectively. Cinnamophilin blocked the TXA2 receptor of human platelets, rat aorta and guinea-pig trachea in a study on the competitive antagonism of U-46619-induced aggregation of human platelet-rich plasma, contraction of rat aortic rings and guinea-pig tracheal rings. The pA2 values were 7.3, 6.3 and 5.2, respectively. The lignan also suppressed ["^HJ-inositol monophosphate formation and the rise of intracellular Ca^^ induced by U-46619 in human platelets, induced a dosedependent inhibition of TXB2 formation, and increased PGE2 formation. The authors conclude that cinnamophilin is a novel dual TX synthase inhibitor and TXA2 receptor antagonist and propose its use as a tool in research and treatment of diseases involving TXA2. Tetrahydrofurofiiran lignans have been described as 5-LOX inhibitors in addition to having the previously reported anti-COX properties. In fact, ten lignans were isolated from Zanthoxylum armatum and tested against COX from sheep seminal vesicles and 5-LOX from pork leukocytes [297]. (+)Spinescin, (+)-sesamin and (+)-asarinin were active, with IC50 of 55, 17 and 56 |iM, respectively, when assayed against 5-LOX. These values are clearly higher than that of NDGA used as the positive control (IC50 =1.5 |aM). Only (+)-fargesin and planinin inhibited COX activity and the inhibition was not significant. In order to establish more data on structure activity relationships, these authors tested other synthetic-related lignans. The lignan (±)-pinoresinol was the most active inhibitor of 5-LOX. Other active compounds against 5-LOX were (+)-epiashantin, (±)-syringaresinol, (-)prenylpiperitol and (-)-prenylpluvia-tilol. On the other hand, (-)prenylpiperitol was the most active lignan in inhibiting COX, with an inhibition percentage of nearly 50%, while the rest gave only about 30% inhibition or less. As the compounds had no activity in the peroxidation assay, the mechanism of inhibition may have to do with a direct inhibition of the enzyme. The authors summarise some of the structural features for the anti-LOX activity of these lignans: a) diequatorial-substituted derivatives are more active than the endo-exo substituted one; b) piperonyl moieties are more active than veratryl substituted ones; c) methylenedioxy

268 function enhances their activity, but the methoxy groups and hydroxy substituents at the bridge carbons decrease it. Eudesmin, magnoHn and lirioresinol-B dimethylether, the three Hgnans isolated from the flower buds of Magnolia fargesii by means of bioassayguided isolation, showed inhibitory effects on TNFa production in lipopolysaccharide (LPS)-stimulated murine macrophage cell lines. Eudesmin showed the strongest activity of the three isolated compounds, with an IC50 of 51 |iM [298]. Using the same experimental protocol, Cho et al [299] isolated a series of lignans from the rhizomes of Coptis japonica var. dissecta (Ranunculaceae), which were assayed as inhibitors of TNFa production. The active lignans, pinoresinol, woorenoside-V and lariciresinol glycoside, with an activity range from 37 to 55% at the concentration of 25 |J,g/mL, may partly participate in the anti-inflammatory and the antiallergenic effect of this species by inhibiting TNF-a production. Seven lignans and two neolignans were assayed on the concanavalin Ainduced proliferation of human peripheral blood lymphocytes in vitro [300]. All the tested compounds showed inhibitory activity, with IC50 ranging from 0.02 to 4.30 \x§JrvL. Machilin A was the strongest inhibitor, with an IC50 of 1.6 x 10'^ M, which is lower than that of the immunosuppressive glucocorticoid prednisolone (1.7 x 10"^ M). Other interesting lignans are /wt^w-dihydroguaiaretic acid and (-)-sesamin, which showed IC50 values of 4.8 x 10"^ and 9.3 x 10"^ M, respectively. The viability of lymphocytes before and after treatment indicated no changes, and therefore the lignans are not toxic against lymphocytes but may inhibit DNA synthesis. The authors suggest the possible value of plant lignans as immunosuppressive agents. Other Effects Effects on Smooth and Skeletal Muscle

The lignans of (+)-pinoresinol, (+)-epipinoresinol, (+)-lariciresinol and (+)isolariciresinol isolated from Fagraea racemosa (Loganiaceae) were tested using a bioassay of the relaxation effect on NA-induced contraction in rat aortic strips [301]. The analgesic properties of the plant extract in the acetic acid-induced writhing and tail pressure tests in mice are due to the lignan

269 fraction. (+)-Pinoresinol showed dose-dependent analgesic effect on writhing symptoms in mice, and local anaesthesia in guinea pigs. Cinnamophilin has protective activity against reperflision injury of the ischaemic skeletal muscle in rats. Ischaemia was induced in one hind limb by application of a tourniquet on the proximal thigh; the contralateral limb served as an internal control. Of the four-reperfiision groups, only the cinnamophilin group had a smaller triphenyltetrazolium chloride reduction and lower muscle weight gain [302]. The dihydrobenzofliran lignan, 3',4-0-dimethylcedrusin improved wound healing in vivo by stimulating the formation of fibroblasts and collagen [303]. Renal and Diuretic Effects

Hall et al [304] tested the effects of a diet supplemented with flaxseed, rich in a-linolenic acid and plant lignans, on a murine model of lupus nephritis, and observed that the percentage of flaxseed-fed mice with proteinuria was lower than the control and that spleen lymphocyte proliferation was significantly higher in the control group than in the flaxseed group. In addition, mortality was lower in the flaxseed-fed mice versus the control. The diuretic properties of arctigenin and hattalin were studied by Hirano et al. [305], who demonstrated that arctigenin had no effect on urine volume, but that hattalin increased it in rats and mice. This compound decreased Na^, K^ and CI' excretion in rats, but serum Na^ and K^ levels did not change. However, serum CI' levels in these animals decreased with respect to the control group. The authors propose that the mechanism of the diuretic effects of hattalin was different from that of known diuretics. The natriuretic effect of prestegane B observed in vivo [306] could have to do with the inhibition of NaVK^-ATP activity demonstrated in vitro in previous studies. This synthetic lignan probably acts beyond the proximal tubule, as urinary phosphate was not altered. Prestegane B mimics the effects of other endogenous diuretic and natriuretic hormones, but its site of action and its effect on renal hemodynamics are obviously different.

270

Central Nervous System Effects

Few studies on the activity of lignans on CNS have been reported. In 1979 was reported the central depressive effects in rabbit of (+)-nortrachelogenin from Wikstroemia indica [9]. Watanabe et al [307] isolated magnolol and honokiol from Magnolia officinalis and studied their central depressant effects. In the animal experiments both compounds produced this kind of effect as well as muscle relaxation by a central action. In addition, magnolol depressed both the hypothalamic and the reticular formation ascending activating system in rat, according to the results obtained in the EEG experiments. Yamazaki et al [308] studied the effect of (+)- and (-)-syringaresinol, (+)-syringaresinol glucosides and syringin on neurite outgrowth of a cultured cell line of paraneuron, PC12h cells, in order to know if they can induce the neuronal differentiation in cells. Only (+)-syringaresinol diglucoside was found to be a promoter of the neurite outgrowth and stimulated responses to a high concentration of KCl and to carbachol in the cells, as demonstrated by the increase in the concentration of cytosolic free calcium. Schisandrin has extensive inhibitory effects on the CNS, which is characteristic of neuroleptic drugs [240], and isoamericanol A, americanol A and americanin A enhanced choline acetyltransferase activity at 10"^ M in a cultured neuronal cell system derived from fetal rat hemisphere [309]. Four lignans isolated from the roots of Valeriana officinalis (Valerianaceae) were identified as (+)-l-hydroxypinoresinol, (+)-pinore-sinol-p-Dglucoside, (+)-pinoresinol and (-)-prinsepiol. They were tested as potential ligands of the 5-HTIA-, GABAA-, benzodiazepin- and |Li-opioid-receptors, but only (+)-l-hydroxypinoresinol showed affinity for the 5-HTiA-receptor, with an IC50 of 2.3 |LIM [310]. Genotoxicity and Mutagenicity

The effect of NDGA on the production of sister-chromatid exchanges and on the level of the mitotic index in cultured human lymphocytes and in mouse bone marrow cells in vivo was evaluated by Madrigal Bujaidar et al [311], and they observed that in both models NDGA produced genotoxic and cytotoxic effects. In a later study [312], the same author observed

271 contradictory effects when they tested the capacity of NDGA to inhibit the rate of sister chromatid exchanges induced by methyl methanesulfonate. They used cultured human lymphocytes from two female donors for the experiment, and the results indicated that NDGA may be an antigenotoxic agent in mammalian cells in vitro and in vivo. The methanol extract from Machilus thunhergii [313] showed a suppressive effect on the Salmonella typhimurium test against the mutagen 3-amino-l,4-dimethyl-5H-pyrido[4,3-b]indole which requires liver metabolising enzymes. The fractionation of the methanol extract gave a suppressive compound in the chloroform fraction that inhibited the activity of the mutagenic agent and was identified as w^^o-dihydroguaiaretic acid. In addition, it suppressed the gene expression and aflatoxin B-1 effects. Kulling et al [314,315] studied the potential genotoxicity of enterolactone and enterodiol, and their precursors, matairesinol and secoisolariciresinol, at concentrations of 200 mM in the cell-free system and 100 mM in cultured cells, and the effects were compared with those of the aneuploidogen diethylstilbestrol and the clastogen 4-nitroquinoline-A^oxide. As none of the four lignans had any activity at the endpoints studied, the authors concluded that the four lignans are devoid of aneuploidogenic and clastogenic potential under the experimental conditions used in this study. CONCLUSIONS The lignans are an interesting group of natural products than can serve as a reference in the search for pharmacologically active drugs in different therapeutical groups. In addition, they have phytoestrogen properties which justify their interest in dietetic. Lignans present in food have no direct effects as phytoestrogens but some of their metabolites do. They are not potent as phytoestrogens, but some of them play a relevant role in sex differentiation and function, including different pathologies related with sexual hormone dependent diseases such as postmenopausical troubles, osteoporosis, estrogen dependent cancers, cardiovascular and liver diseases. A diet rich in proestrogenic lignans, such as flaxseed or soybean or whole dried legumes and cereals, can be a complementary treatment in the above-cited diseases. Lignans have been described as the active principles in different antiviral and anticancer medicinal plants and crude extracts. However, few of them

272

are potent enough to be used as therapeutical agents. Only some podophyllotoxins and their derivatives showed clear efficacy and potency as anticancer and antiviral agents. Of the cited properties, the activity on PAF and cAMP-PDE can be of interest in the selection of series to synthesise new and potent lignans. Isolation of compounds with activity in the 10'^ M range as PAF inhibitors can provide a good starting point for obtaining new compounds. On the other hand, the selectivity as inhibitors of different PDEs can serve to obtain new synthetic derivatives in future research. ABBREVIATIONS AA ADP AMP cAMP ATP CCC CNS CoA COX DAD DNA ED50 ER FAB FLAP FMLP GC OCR GMP cGMP GSH GST-P HDL HIV ^H-NMR

= Arachidonic Acid = Adenosine Diphosphate = Adenosine Monophosphate = Cyclic Adenosine Monophosphate = Adenosine Triphosphate = Counter Current Chromatography = Central Nervous System = Coenzyme A = Cyclooxygenase = Diode Array Detector = Deoxyribonucleic Acid = Effective Dose-50 = Estrogen Receptor = Fast-Atom Bombardment = 5- Lipoxygenase Activating Protein = Formyl-Met-Leu-Phe = Gas Chromatography = Glucocorticoid Receptor = Guanine Monophosphate; = Cyclic Guanine Monophosphate = Glutathione = Glutathione iS-Transferase Placental Form = High-Density Lipoprotein = Human Immunodeficiency Virus = Proton Nuclear Magnetic Resonance

273

= High Performance Liquid Chromatography = Human Papilloma Virus = Herpes Simplex Virus = Serotonin, 5-Hydroxytriptamine = Inhibitory Concentration 50 IC50 = Immunoglobulin Ig IL;IL-1 = Interleukin; Interleukin-1 /.v. = Intravenously LC = Liquid Chromatography = Lethal Dose 50 LD50 LDL = Low-Density Lipoprotein LOX = Lipoxygenase LPS = Lipopolysaccharide LTs; LTB4 = Leukotrienes; Leukotriene B4 MDA = Malondialdehyde 3'-MeDAB = 3'-Methyl-4-dimethylaminoazobenzene MIC = Minimum Inhibitory Concentration MPO = Myeloperoxidase MS = Mass Spectrometry = Noradrenaline, Norepinephrine NA NADPH = Nicotinamide Adenine Dinucleotide Phosphate NDGA = Nordihydroguaiaretic Acid ODC = Ornithine Decarboxylase PAF = Platelet Activating Factor PDE = Phosphodiesterase PGs; PGE2 = Prostaglandins; Prostaglandin E2 PKC = Protein Kinase C PLA2 = Phospholipase A2 PMNL = Polymorphonuclear Leukocytes P.O. ^^ per OS, orally RT = Reverse Transcriptase SFC = Supercritical Fluid Chromatography SFE = Supercritical Fluid Extraction SHBG = Sex Hormone-Binding Globulin SOD = Superoxide Dismutase TNF;TNF-a = Tumor Necrosis Factor; Tumor Necrosis Factor -a TPA = 12-0-Tetradecanoylphorbol-13 -Acetate TX; TXA2 = Thromboxane; Thromboxane A2 HPLC HPV HSV 5-HT

274

VLDL VSV

= Very Low Density Lipoprotein = Vesicular Stomatitis Virus

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22]

Lewis, N.G.; Davin, L.B.; ACSSymp. Ser, 1994, 562, 202-246. Lewis, N.G., Kato, M.J., Lopes, N.; Davin, L.B.; ACS Symp. Ser., 1995, 588, 135-167. Castro, M.A.; Gordaliza, M.; Miguel del Corral, J.M.; San Feliciano, A., Phytochemistry, 1996, 4J, 995-1011. Massanet, G.M.; Pando, E.; Rodriguez-Luis, F.; Zubia, E.; Fitoterapia, 1988, 60, 3-35. Ward, R.S.; Nat. Prod. Rep., 1993,10, 1-28. Ward, R.S.; Nat. Prod. Rep., 1995,12, 183-205. Ward, R.S.; Nat. Prod. Rep., 1997, 14, 43-74. Ward, R.S.; Nat. Prod Rep., 1999,16, 75-96. MacRae, W.D.; Towers, G.H. Phytochemistry, 1984, 6, 1207-1220. Ayres, D.C.; Loike, J.D.; Lignans. Chemical, Biological and Clinical Properties, Cambridge University Press: Cambridge, 1990. Nishibe, S. In Polyphenolic Phenomena, Scalbert, A. ed.; INRA: Paris, 1993, 247-255. MacRae, W.D.; Hudson, J.B.; Towers, G.H.; Planta Med, 1989, 55, 531-535. Charlton, J.L.; J. Nat. Prod, 1998, 61, 1447-1451. Ghisalberti, E.L.; Phytomedicine, 1997, 4, 151-166. Imbert, T.F.; Biochimie, 1998, 80, 207-222. Damayanthi, Y.; Lown, J.W.; Curr. Med. Chem., 1998, 5, 205-252. Whiting, D A ; Nat. Prod Rep., 1985, 2, 191-211. Whiting, D.A.; Nat. Prod Rep., 1987, 4, 499-525. Whiting, D.A.; Nat. Prod Rep., 1990, 7, 349-364. Harbome, J.B.; Baxter, H. Phytochemical Dictionary, Taylor & Francis: London, 1993. Dinkova-Kostova, AT.; Gang, DR.; Davin, LB.; Bedgar, D.L.; Chu, A; Lewis, N.G.; J. Biol. Chem., 1996, 271, 29473-29482. Gang, DR.; Dinkova-Kostova, AT.; Davin, LB.; Lewis, N.G.; ACS Symp. Ser., 1997, 658, 58-89.

275 [23] [24] [25]

[26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42]

Bernard, C.B.; Arnason, J.T.; Philogene, B.J.R.; Lam, J.; Wadwell, T.,Phytochemistry, 1989, 28, 1373-1377. Frias, I.; Trujillo, J.M.; Romero, J.; Hernandez, J.; Perez, J.A.; Biochimie, 1995, 77, 707-712. Pagnocca, F.C.; Ribeiro, SB.; Torkomian, V.L.V.; Hebling, M.J. A.; Bueno, O.C; Da Silva, M.F.D.F.; Ferreira, A.G.; J. Chem. Ecol, 1996, 22, 1325-1330. Cabral, M.M.; Garcia, E.S.; Kelecom, k:,Mem. Inst. Oswaldo Cruz, 1995, 90, 759-763. Harbome, J.B. Nat. Prod. Rep., 1986, 3, 323-344. Davin, L.B.; Bedgar, D.L.; Katayama, T; Lewis, N.G.; Phytochemistry, 1992, 31, 3869-3874. Katayama, K.; Davin, L. B.; Chu, A.; Lewis, N.G.; Phytochemistry, 1993, 33, 581-591. Ozawa, S.; Davin, L.B.; Lewis, N.G.; Phytochemistry, 1993, 32, 643-652. Knaggs, A.R.; Nat. Prod. Rep., 1999, 16, 525-560. Rahman, M.M.A.; Dewick, P.M.; Jackson, D.E.; Lucas, J.A.; Phytochemistry, 1990,29, 1841-1846. Broomhead, A.J.; Rahman, M.M.A.; Dewick, P.M.; Jackson, D.E.; Lucas, J. A.; Phytochemistry, 1991, 30, 1489-1492. Davin, L.B.; Wang, H.B.; Crowell, A.L.; Bedgar, D.L.; Martin, D. M.; Sarkanen, S.; Lewis, N.G.; Science, 1997, 275, 362-366. Fujita, M.; Gang, D.R.; Davin, L B . ; Lewis, N.G.; J. Biol. Chem., 1999, 274, 618-627. Slanina, J.; Taborska, E.; Lojkova, L.; Planta Med, 1997, 63, 277280. Lojkova, L.; Slanina, J.; Mikesova, M.; Taborska, E.; Vejrosta, J.; Phytochem. Anal., 1997, 8, 261-265. Choi, Y.H.; Kim, J.; Jeon, S.H.; Yoo, K.P.; Lee, H.K.; Chromatographia, 1998, 48, 695-699. Chandra, A.; Nair, M.G.; Planta Med, 1995, 61, 192-195. Suto, K.; Ito, Y.; Sagara, K.; Itokawa, H.; J. Chromatogr. A, 1997, 786, 366-370. Wolf, C ; Pirkle, W.H.; Welch, C.J.; Hochmuth, D.H.; Konig, W.A.; Chee, G.L.; Charlton, J.L.; J. Org. Chem., 1997, 62, 5208-5210. Ryu, S.N.; Ho, C.T.; Osawa, T.; J. Food Lipids, 1998, 5, 17-28.

276 [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60]

Cavin, A.; Potterat, O.; Wolfender, J.L.; Hostettmann, K.; Dyatmyko, W. J. Nat. Prod., 1998, 61, 1497-1501. Facino, R.M.; Carini, M.; Aldini, G.; De Angelis, L.; Rapid Commun. Mass Spectrom., 1997,11, 1303-1308. Bergeron, C; Marston, A.; Wolfender, J.L.; Mavi, S.; Rogers, C; Hostettmann, K.; Phytochem. Anal, 1997, 8, 32-36. He, X.G.; Lian, L. Z.; Lin, L. Z.; J. Chromatogr. A, 1997, 757, 8187. Bambagiotti Alberti, M.; Coran, S.A.; Ghiara, C; Giannellini, V.; RafFaelli, A.; Rapid. Commun. Mass Spectrom., 1994, 8, 595-598. Bambagiotti Alberti, M,; Coran, S.A.; Ghiara, C; Moreti, G.; RafFaelli, A.; Rapid. Commun. Mass Spectrom., 1994, 8, 929-932. Annapoorani, K.S.; Damodaran, C; Chandra Sekharan, P.; J. Chromatogr., 1984, 303, 296-305. Annapoorani, K.S.; Periakali, P.; Ilangovan, S.; Damodaran, C; Sekharan, P. C; J. Anal. Toxicol, 1984, 8, 182-186. Ragupathi, G.; Prabhasankar, P.; Chandra Sekharan, P.; Annapoorani, K.S.; Damodaran, C; Forensic Sci. Int., 1992, 56, 127-136. Schaufelberger, D.E.; J. Chromatogr., 1991, 538, 45-57. Lee, Y.W.; Voyksner, R.D.; Pack, T.W.; Cook, C.E.; Fang, Q.C.; Ito, Y.; Anal. Chem., 1990, 62, 244-248. Gnabre, J.N.; Ito, Y.; Ma, Y.; Huang, R.C.; J. Chromatogr. A, 1996, 719, 353-364. Ma, Y.; Qi, L.; Gnabre, J.N.; Huang, R.C.C.; Chou, F.E.; Ito, Y.; J. Liq. Chromatogr. Relat. TechnoL, 1998, 27, 171-181. Fotsis, T.; Heikkinen, R.; Adlercreutz, H.; Axelson, M.; Setchell, K. D.; Clin. Chim. Acta, 1982, 121, 361-371. Bannwart, C; Adlercreutz, H.; Wahala, K.; Brunow, G.; Hase, T.; Clin. Chim. Acta, 1989,180, 293-301. Adlercreutz, H.; Fotsis, T.; Watanabe, S.; Lampe, J.; Wahala, K.; Makela, T.; Hase, T.; Cancer Detect. Prev., 1994,18, 259-271. Mazur, W.; Fotsis, T.; Wahala, K.; Ojala, S.; Salakka, A.; Adlercreutz, H.;Anal. Biochem., 1996, 233, 169-180. Setchell, K.D.; Lawson, A.M.; McLaughlin, L.M.; Patel, S.; Kirk, D.N.; Axelson, M.; Biomed. Mass Spectrom., 1983,10, 227-235.

277

[61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80]

Adlercreutz, H.; Bannwart, C; Wahala, K.; Makela, T.; Brunow, G.; Hase, T.; Arosemena, P.J.; Kellis, J.T.,Jr.; Vickery, L.E.; J. Steroid. Biochem. Mol. Biol, 1993, 44, 147-153. Adlercreutz, H.; Fotsis, T.; Kurzer, M.S.; Wahala, K.; Makela, T.; Hase, T.-Anal. Biochem., 1995, 225, 101-108. Ono, H.; Matsuzaki, Y.; Wakui, Y.; Takeda, S.; Ikeya, Y.; Amagaya, S.; Maruno, M.; J. Chromatogr. B Biomed. Appl, 1995, 674, 293-297. Atkinson, D.A.; Hill, H. H.; Shultz, T.D.; J Chromatogr. Biomed. Appl., 1993, 128, 173-179. Gamache, P.H.; Acworth, I.N.; Proc. Soc. Exp. Biol Med., 1998, 217, 274-280. Adlercreutz, H.; Wang, G.J.; Lapcik, O.; Hampl, R.; Wahala, K.; Makela, T.; Lusa, K.; Talme, M.; Mikola, H.; Anal Biochem., 1998, 265,208-215. Yoo, K.; Porter, J.R.; J. Nat. Prod, 1993, 56, 715-721. Kominami, G.; Ueda, A; Sakai, K.; Misaki, A.; J. Chromatogr. B, 1997, 704, 243-250. Kurzer, M.S.; Xu, X.; Annu. Rev. Nutr., 1997,17, 353-381. Setchell, K.D.; Lawson, A.M.; Mitchell, F.L.; Adlercreutz, H.; Kirk, D.N.; Axelson, M.; Nature, 1980, 287, 740-742. Axelson, M.; Sjovall, J.; Gustafsson, B.E.; Setchell, K.D.; Nature, 1982, 298, 659-660. Adlercreutz, H.; Mousavi, Y.; Clark, J.; Hockerstedt, K.; Hamalainen, E.; Wahala, K.; Makela, T.; Hase, T.; J. Steroid Biochem. Mol Biol, 1992, 41, 331-337. Adlercreutz, H.; Scand. J. Clin. Lab. Invest. Suppl, 1990, 201, 3-23. Adlercreutz, U.,Ann. N. Y. Acad Sci., 1990, 595, 281-290. Adlercreutz, H.; Environ. Health Perspect., 1995, 103, Suppl. 7, 103-112. Adlercreutz, C.H.; Goldin, B.R.; Gorbach, S.L.; Hockerstedt, K.A.; Watanabe, S.; Hamalainen, E.K.; Markkanen, M.H.; Makela, T.H.; Wahala, K.T.; Adlercreutz, T.A.; J. Nutr., 1995, 125, 757S-770S. Adlercreutz, H.; Mazur, W.; Ann. Med, 1997, 29, 95-120. Potter, J.D.; Steinmetz, K.; lARC. Sci. Publ, 1996, 61-90. Knight, D.C.; Eden, J.A.; Obstet. Gynecol, 1996, 87, 897-904. Rickard, S.E.; Thompson, L.U.; ACS Symp. Ser., 1997, 662, 113293.

278

[81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99]

Mazur, W.; Adlercreutz, H.; Pure Appl. Chem., 1998, 70, 17591776. Tham, D.M.; Gardner, CD.; Haskell, W.L.; J. Clin. Endocrinol. MetaboL, 1998, 83, 2223-2235. Rickard, S.E.; Orcheson, L.J.; Seidl, M.M.; Luyengi, L.; Fong, H. H.S.; Thompson, L.U.; J. Nutr., 1996, 726, 2012-2019, Gansser, D.; Spiteller, G; Z Naturforsch. C, 1995, 50, 98-104. Brzezinski, A.; Adlercreutz, H.; Shaoul, R.; Rosier, A.; Shmueli, A.; Tanos, V.; Schenker, J. G.; Menopause, 1997, 4, 89-94. Schottner, M.; Gansser, D.; Spiteller, G.; Planta Med, 1997, 63, 529-532. Schottner, M.; Gansser, D.; Spiteller, G.; Z Naturforsch. C, 1997, 52, 834-843. Schottner, M.; Spiteller, G.; Gansser, D.; J. Nat. Prod, 1998, 67, 119-121. Wang, C; Makela, T.; Hase, T.; Adlercreutz, H.; Kurzer, M.S.; J. SteroidBiochem. Mol. Biol, 1994, 50, 205-212. Evans, B.A.; Griffiths, K.; Morton, M.S.; J. Endocrinol, 1995, 147, 295-320. Mousavi, Y.; Adlercreutz, H.; J. Steroid Biochem. Mol Biol, 1992, ^7,615-619. Orcheson, L.J.; Rickard, S.E.; Seidl, MM.; Thompson, L.U.; Cancer Lett., 1998, 125, 69-76. Martin, M.E.; Haourigui, M.; Pelissero, C; Benassayag, C; Nunez, E.A.; Life Sci., 1995, 58, 429-436. Kardinaal, A.F.; Morton, M.S.; Bruggemann Rotgans, I.E.; van Beresteijn, E.G.; Eur. J. Clin. Nutr., 1998, 52, 850-855. Tou, J.C.L.; Chen, J. M.; Thompson, L.U.; J. Nutr., 1998, 725, 1861-1868. Kinghorn, A.D.; Fong, H.H.S.; Farnsworth, N.R.; Mehta, R.G.; Moon, R.C.; Moriarty, R.M.; Pezzuto, J.M.; Curr. Org. Chem., 1998, 2, 597-612. Luyengi, L.; Suh, N.; Fong, H.H.S.; Pezzuto, J. M.; Kinghorn, A.D.; Phytochemistry, 1996, 43, 409-412. Pettit, G.R.; Schaufelberger, D.E.; J. Nat. Prod, 1988, 57, 11041112. Ma, W.W.; Kozlowski, J. F.; McLaughlin, J.L.; J. Nat. Prod, 1991, 54, 1153-1158.

279 [100] Pieters, L.; De Bruyne, T.; Claeys, M.; Vlietinck, A.; Calomme, M.; van den Berghe, D.; J. Nat. Prod., 1993, 56, 899-906. [101] Kuo, Y.H.; Kuo, L.M.Y.; Chen, C.F.; J. Org. Chem., 1997, 62, 3242-3245. [102] Wu, Y.C.; Chang, G.Y.; Ko, F.N.; Teng, C. M.; PlantaMed, 1995, 67, 146-149. [103] Arslanian, R.L.; Bailey, D.T.; Kent, M.C.; Richheimer, S.L.; Thomburg, K.R.; Timmons, D.W.; Zheng, Q.Y.; J. Nat. Prod., 1995, 58, 583-585. [104] Takasaki, M.; Konoshima, T.; Yasuda, I.; Hamano, T.; Tokuda, H.; Biol. Pharm. Bull, 1997, 20, 776-780. [105] Chen, I.S.; Chen, J.J.; Duh, C.Y.; Tsai, I.L.; Phytochemistry, 1997, 45, 991-996. [106] van Uden, W.; Homan, B.; Woerdenbag, H.J.; Pras, N.; MaUngre, T.M.; Wichers, H.J.; Harkes, M.; J. Nat. Prod, 1992, 55, 102-110. [107] Novelo, M.; Cruz, J.G.; Hernandez, L.; Pereda Miranda, R.; Chai, H.; Mar, W.; Pezzuto, J.M.; J. Nat. Prod, 1993, 56, 1728-1736. [108] Middel, O.; Expert Op. Ther. Patents, 1996, 6, 547-554. [109] Middei, O.; Woerdenbag, H.J.; van Uden, W.; van Oeveren, A.; Jansen, J.F.; Feringa, B.L.; Konings, A.W.; Pras, N.; Kellogg, R.M.; J. Med. Chem., 1995, 38, 2112-2118. [110] San Feliciano, A.; Gordaliza, M.; Miguel del Corral, J.M.; Castro, M.A.; Garcia Gravalos, M.D.; Ruiz Lazaro, P.; Planta Med, 1993, 59, 246-249. [HI] Gordaliza, M.; Castro, M.A.; Garcia Gravalos, M.D.; Ruiz, P.; Miguel del Corral, J.M.; San Feliciano, A.; Arch. Pharm. (Weinheim), 1994, 327, 175-179. [112] Gordaliza, M.; Miguel del Corral, J.M.; Castro, M.A.; Lopez Vazquez, M.L.; San Feliciano, A.; Garcia Gravalos, M.D.; Carpy, A.; Bioorg. Med. Chem., 1995, J, 1203-1210. [113] Gordaliza, M.; Castro, M.A; Miguel del Corral, J.M.; Lopez Vazquez, M.L.; Garcia, P.A.; San Feliciano, A; Garcia Gravalos, M. D.; Broughton, H.; Tetrahedron, 53, 15743-15760. [114] Gordaliza, M.; Faircloth, G.T.; Castro, M.A.; Miguel del Corral, J. M.; Lopez Vazquez, M.L.; San Feliciano, A.; J. Med. Chem., 1996, 39, 2865-2868.

280 [115] Miguel del Corral, J.M.; Gordaliza, M.; Castro, M.A.; Lopez Vazquez, ML.; Garcia Gravalos, M.D.; Broughton, H.B.; San Feliciano, A.; Tetrahedron, 1997, 53, 6555-6564. [116] Subrahmanyam, D.; Renuka, B.; Rao, C.V.L.; Sagar, P. S.; Deevi, D.S.; Babu, J. M.; Vyas, K.; Bioorg. Med. Chem. Lett., 1998, 8, 1391-1396. [117] Ikeda, R.; Nagao, T.; Okabe, H.; Nakano, Y.; Matsunaga, H.; Katano, M.; Mori, M.; Chem. Pharm. Bull, 1998, 46, 871-874. [118] Ikeda, R.; Nagao, T.; Okabe, H.; Nakano, Y.; Matsunaga, H.; Katano, M.; Mori, M.; Chem. Pharm. Bull., 1998, 46, 875-878. [119] Hirano, T.; Fukuoka, K.; Oka, K.; Naito, T.; Hosaka, K.; Mitsuhashi, H.; Matsumoto, Y. Cancer Invest., 1990, 8, 595-602. [120] Hirano, T.; Fukuoka, K.; Oka, K.; Matsumoto, Y.; Cancer Invest., 1991, 9, 145-150. [121] Tomioka, K.; Ishiguro, T.; Mizuguchi, H.; Komeshima, N.; Koga, K.; Tsukagoshi, S.; Tsuruo, T.; Tashiro, T.; Tanida, S.; Kishi, T.; J. Med Chem., 1991, 34, 54-57. [122] Wickramaratne, D.B.; Pengsuparp, T.; Mar, W.; Chai, H.B.; Chagwedera, T.E.; Beecher, C.W.; Farnsworth, N.R.; Kinghom, A.D.; Pezzuto, J.M.; Cordell, G.A.; J. Nat. Prod, 1993, 56, 20832090. [123] Wickramaratne, D.B.; Mar, W.; Chai, H.; Castillo, J.J.; Farnsworth, N.R.; Soejarto, D.D.; Cordell, G.A.; Pezzuto, J.M.; Kinghorn, A. D.; PlantaMed, 1995, 61, 80-81. [124] Somanabandhu, A.; Nitayangkura, S.; Mahidol, C; Ruchirawat, S.; Likhitwitayawuid, K.; Shieh, H. L.; Chai, H.; Pezzuto, J.M.; Cordell, G.A.; J. Nat. Prod, 1993, 56, 233-239. [125] Hirano, T.; Gotoh, M.; Oka, K.; Life ScL, 1994, 55, 1061-1069. [126] Moritani, S.; Nomura, M.; Takeda, Y.; Miyamoto, K.; Biol Pharm. 5M//., 1996,79, 1515-1517. [127] Umehara, K.; Nakamura, M.; Miyase, T.; Kuroyanagi, M.; Ueno, A.; Chem. Pharm. Bull., 1996, 44, 2300-2304. [128] Kato, T.; Hirose, M.; Takahashi, S.; Hasegawa, R.; Kohno, T.; Nishibe, S.; Kato, K.; Shirai, T.; Anticancer Res., 1998, 18, 10531057. [129] Chen, J.J.; Ishikawa, T.; Duh, C. Y.; Tsai, I.L.; Chen, I.S.; Planta Med, 1996, 62, 528-533.

281 [130] Medarde, M.; Pelaez Lamamie de Clairac, R.; Tome, F.; Lopez, J.L.; San Feliciano, A.; Gravalos, D.G.; Arch. Pharm. (Weinheim), 1995, 328, 403-407. [131] Medarde, M.; Pelaez Lamamie de Clairac, R.; Lopez, J.L.; Gravalos, D.G.; San Feliciano, A.; Arch. Pharm. (Weinheim), 1995, 328, 640644. [132] San Feliciano, A.; Medarde, M.; Pelaez Lamamie de Clairac, R.; Lopez, J.L.; Puebla, P.; Garcia Gravalos, M.D.; Ruiz Lazaro, P.; Garcia de Quesada, M. T.; Arch. Pharm. (Weinheim), 1993, 326, [133] Cui, B.L.; Chai, H.Y.; Santisuk, T.; Reutrakul, V.; Farnsworth, N. R.; Cordell, G.A.; Pezzuto, J.M.; Kinghom, A.D.; Tetrahedron, 1997, 53, 17625-17632. [134] Lee, S.K.; Cui, B.; Mehta, R.R.; Kinghom, A.D.; Pezzuto, J.M.; Chem. Biol. Interact, 1998, 7/5,215-228. [135] Takaba, K.; Hirose, M.; Yoshida, Y.; Kimura, J.; Ito, N.; Shirai, T.; Cancer Lett., 1997, 113, 39-46. [136] Njoku, C.J.; Hopp, D.C.; Alali, F.; Asuzu, I.U.; McLaughlin, J.L.; PlantaMed., 1997, 63, 580-581. [137] Park, S.; Lee, D.K.; Yang, C.H.; Cancer Lett, 1998,127, 23-28. [138] Hempstock, J.; Kavanagh, J.P.; George, N.J.R.; Brit J. Urol, 1998, 82, 560-563. [139] Suh, N.; Luyengi, L.; Fong, H.H.; Kinghom, A.D.; Pezzuto, J.M.; Anticancer Res., 1995,15, 233-239. [140] Tsai, LL.; Hsieh, C.F.; Duh, C.Y.; Chen, I S . ; Phytochemistry, 1996, 43, 1261-1263. [141] Tsai, LL.; Hsieh, C.F.; Duh, C.Y.; Phytochemistry, 1998, 48, 13711375. [142] El Gamal, A.A.; Takeya, K.; Itokawa, K ; Halim, A.F.; Amer, M.M.; Saad, H.E.A.; Phytochemistry, 1997, 45, 597-600. [143] Ohse, T.; Ohba, S.; Yamamoto, T.; Koyano, T.; Umezawa, K.; J. Nat Prod., 1996, 59, 650-652. [144] Landstrom, M.; Zhang, J.X.; Hallmans, G.; Aman, P.; Bergh, A.; Damber, J.E.; Mazur, W.; Wahala, K.; Adlercreutz, H.; Prostate, 1998, 36, 151-161. [145] Yan, L.; Yee, J.A.; Li, D.H.; McGuire, M.H.; Thompson, L.U.; Cancer Lett, 1998,124, 181-186.

282

146] Sung, M.K.; Lautens, M.; Thompson, L.U.; Anticancer Res., 1998, 18, 1405-1408. [147] Thompson, L.U.; Seidl, M.M.; Rickard, S.E.; Orcheson, L.J.; Fong, H.n.S; Nutr. Cancer, 1996, 26, 159-165. [148] Li, D.; Yee, J.A.; Thompson, L.U.; Yan, L.; Cancer Lett, 1999, 142, 91-96. [149] Thompson, L.U.; Rickard, S.E.; Orcheson, L.J.; Seidl, M.M.; Carcinogenesis, 1996,17, 1373-1376. 150] Jenab, M.; Thompson, L.U.; Carcinogenesis, 1996, 17, 1343-1348. [151] Wang, C.F.; Kurzer, M.S.; Nutr. Cancer, 1997, 28, 236-247. [152] Wang, C.Kurzer, M.S.; Nutr. Cancer, 1998, 31, 90-100. [153] SchultzeMosgau, M.H.; Dale, I.L.; Gant, T. W.; Chipman, J. K.; Kerr, D. J.; Gescher, A.; Eur. J. Cancer, 1998, 34, 1425-1431. 154] Wang, W.Q.; Liu, L.Q.; Higuchi, CM.; Chen, H.W; Biochem. Pharmacol, 1998,56, 189-195. [155] Vlietinck, A.J.; DeBruyne, T.; Apers, S.; Pieters, L.A.; PlantaMed., 1998, 64, 97-109. [156] Matthee, G.; Wright, A.D.; Konig, G.M.; Planta Med, 1999, 65, 493-506. [157] Chen, D.F.; Zhang, S.X.; Xie, L.; Xie, J.X.; Chen, K.; Kashiwada, Y.; Zhou, B.N.; Wang, P.; Cosentino, L.M.; Lee, K.H.; Bioorg. Med. Chem., 1997, 5, 1715-1723. [158] Chen, D.F.; Zhang, S.X.; Chen, K.; Zhou, B.N.; Wang, P.; Cosentino, L.M.; Lee, K.H.; J. Nat. Prod, 1996, 59, 1066-1068. [159] Rimando, A.M.; Pezzuto, J.M.; Farnsworth, N.R.; Santisuk, T.; Reutrakul, V.; Kawanishi, K.; J. Nat. Prod, 1994, 57, 896-904. [160] Valsaraj, R.; Pushpangadan, P.; Smitt, U.W.; Adsersen, A.; Christensen, S.B.; Sittie, A.; Nyman, U.; Nielsen, C; Olsen, C.E.; J. Nat. Prod, 1997, 60, 739-742. [161] Fujihashi, T.; Hara, H.; Sakata, T.; Mori, K.; Higuchi, H.; Tanaka, A.; Kaji, H.; Kaji, A; Antimicrob. Agents Chemother., 1995, 39, 2000-2007. 162] Lee, S.S.; Lin, M.T.; Liu, C.L.; Lin, Y.Y.; Liu, K.C.S.C; J. Nat. Prod., 1996,59, 1061-1065. [163] Chang, C.W.; Lin, M.T.; Lee, S.S.; Liu, K.C.; Hsu, F.L.; Lin, J.Y.; Antiviral Res., 1995, 27, 367-374.

283 [164] Liu, K.C.S.C; Lee, S.S.; Lin, M.T.; Chang, C.W.; Liu, C.L.; Lin, J.Y.; Hsu, F.L.; Ren, S.; Lien, E.i.\Med. Chem. Res., 1997, 7, 168179. [165] Hwu, J.R.; Tseng, W.N.; Gnabre, J.; Giza, P.; Huang, R.C.; J. Med. Chem., 1998, 41, 2994-3000. [166] Chen, H.S.; Teng, L.; Li, J.N.; Park, R.; Mold, D.E.; Gnabre, J.; Hwu, J.R.; Tseng, W.N.; Huang, R.C.C.; J. Med Chem., 1998, 41, 3001-3007. [167] Asano, J.; Chiba, K.; Tada, M.; Yoshii, T.; Phytochemistry, 1996, 42, 713-717. [168] Kernan, MR.; Sendl, A.; Chen, J.L.; Jolad, S.D.; Blanc, P.; Murphy, J.T.; Stoddart, C.A.; Nanakorn, W.; Balick, M.J.; Rozhon, E.J.; J. Nat. Prod, 1997, 60, 635-637. [169] Tezuka, Y.; Terazono, M.; Kusumoto, T.I.; Kawashima, Y.; Hatanaka, Y.; Kadota, S.; Hattori, M.; Namba, T.; Kikuchi, T.; Tanaka, K.; Supriyatna, S.;//e/v. Chim. Acta, 1999, 82, 408-417. [170] Hattori, M.; Hada, S.; Watahiki, A.; Ihara, H.; Shu, Y.Z.; Kakiuchi, N.; Mizuno, T.; Namba, T.; Chem. Pharm. Bull, 1986, 34, 38853893. [171] Namba, T.; Tsunezuka, M.; Hattori, M.; Planta Med, 1982, 44, 100-106. [172] El Feraly, F.S.; Cheatham, S.F.; Breedlove, R.L.; J. Nat. Prod, 1983, 46, 493-498. [173] Atta-ur-Rahman; Ashraf, M.; Choudhary, M.I.; Habiburrehman; Kazmi, M.H.; Phytochemistry, 1995, 40, 427-431. [174] Zacchino, S.; Rodriguez, G.; Pezzenati, G.; Orellana, G.; Enriz, R.; Sierra, M.G.; J. Nat. Prod, 1997, 60, 659-662. [175] Zacchino, S.; Rodriguez, G.; Santecchia, C; Pezzenati, G.; Giannini, F.; Enriz, R.; J. EthnopharmacoL, 1998, 62, 35-41. [176] Sartorelli, P.; Young, M.C.M.; Kato, M.J.; Phytochemistry, 1998, 47, 1003-1006. [177] Gonzalez, J.A.; Estevez Braun, A.; Estevez Reyes, R.; Bazzocchi, L L.; Moujir, L.; Jimenez, I.A.; Ravelo, AG.; Gonzalez, AG.; Experientia., 1995, 51, 35-39. [178] Sauvain, M.; Kunesch, N.; Poisson, J.; Gantier, J.C; Gayral, P.; Dedet, J.?.;Phytother. Res., 1996,10, 1-4.

284

[179] Oketch-Rabah, HA.; Dossaji, S.F.; Christensen, SB.; Frydenvang, K.; Lemmich, E.; Cornett, C; Olsen, C.E.; Chen, M.; Kharazmi, A.; Theander, T;J. Nat. Prod, 1997, 60, 1017-1022. [180] Lopes, N.P.; Chicaro, P.; Kato, M. J.; Albuquerque, S.; Yoshida, M.; PlantaMed, 1998, 6< 667-669. [181] Cabral, M.M.O.; Azambuja, P.; Gottlieb, O.R.; Garcia, E.S.; Pamsitol Res., 1999, 85, 184-187. [182] Russell, G.B.; Singh, P.; Fenemore, P.G.; Aust. J. Biol. Sci., 1976, 29, 99-103. [183] Ishibashi, F.; Taniguchi, E.; Phytochemistry, 1998, 49, 613-622. [184] Chauret, D.C.; Bernard, C.B.; Arnason, J. T.; Durst, T.; BCrishnamurty, H.G.; Sanchez Vindas, P.; Moreno, N.; Roman, L. S.; Poveda, L.; J. Nat. Prod., 1996, 59, 152-155. [185] Nitao, J.K.; Nair, M.G.; Thorogood, D.L.; Johnson, K.S.; Scriber, J.M.; Phytochemistry, 1991, 30, 2193-2195. [186] He, K.; Zeng, L.; Shi, G.; Zhao, G.X.; Kozlowski, J.F.; McLaughlin, J.L.; J. Nat. Prod, 1997, 60, 38-40. [187] Li, P.C; Poon, M.K.T.; Ko, K.M.; Am. J. Chin. Med, 1996, 24, 255-262. [188] Li, P.C; Mak, D.H.F.; Poon, M.K.T.; Ip, S.P.; Ko, KM.; Phytomedicine, 1996, 3, 217-221. [189] Chen, C.C; Huang, Y.L.; Chen, H.T.; Chen, Y.P.; Hsu, H.Y.; PlantaMed, 1988, 54, 438-440. [190] Matsunaga, K.; Shibuya, M.; Ohizumi, Y.; J. Nat. Prod, 1994, 57, 1734-1736. [191] Abe, M.; Morikawa, M.; Inoue, M.; Nakajima, A.; Tsuboi, M.; Naito, T.; Hosaka, K.; Mitsuhashi, H.; Arch. Int. Pharmacodyn. Ther., 1989, 30J, 40-50. [192] Abe, M.; Morikawa, M.; Inoue, M.; Tsuboi, M.; Res. Commun. Chem. Pathol Pharmacol, 1990, 69, 271-284. [193] Abe, M.; Morikawa, M.; Inoue, M.; Tsuboi, M.; Aoyagi, Y.; Ohta, A.; Gen. Pharmacol, 1991, 22, 663-668. [194] Matsumura, Y.; Kita, S.; Morimoto, S.; Akimoto, K.; Furuya, M.; Oka, N.; Tanaka, T.; Biol Pharm. Bull, 1995,18, 1016-1019. [195] Kita, S.; Matsumura, Y.; Morimoto, S.; Akimoto, K.; Furuya, M.; Oka, N.; Tanaka, T.; Biol Pharm. Bull, 1995,18, 1283-1285. [196] Matsumura, Y.; Kita, S.; Tanida, Y.; Taguchi, Y.; Morimoto, S.; Akimoto, K.; Tanaka, T.; Biol Pharm. Bull, 1998, 21, 469-473.

285 [197] Braquet, P.; Senn, N.; Fagoo, M.; Garay, R.; Robin, J.P.; Esanu, A.; Chabrier, E.; Godfrain, T.; C. R. Acad Sci. Ill, 1986, 302, 443-448. [198] Braquet, P.; Senn, N.; Robin, J.P.; Esanu, A.; Garay, R.P.; J. Hypertens. SuppL, 1986, 4, S161-S164. [199] Braquet, P.; Senn, N.; Robin, J.P.; Esanu, A.; Godfraind, T.; Garay, R.; Pharmacol. Res. Commun., 1986,18, 227-239. [200] Hirano, T.; Oka, K.; Naitoh, T.; Hosaka, K.; Mitsuhashi, H.; Res. Commun. Chem. Pathol. Pharmacol, 1989, 64, 227-240. [201] Fagoo, M.; Braquet, P.; Robin, J.P.; Esanu, A.; Godfraind, T.; Biochem. Biophys. Res. Commun., 1986,134, 1064-1070. [202] Hussain, R.A.; Dickey, J.K.; Rosser, M.P.; Matson, J.A.; Kozlowski, MR.; Brittain, R.J.; Webb, M.L.; Rose, P.M.; Fernandes, P.; J. Nat. Prod., 1995, 58, 1515-1520. [203] Nikaido, T.; Ohmoto, T.; Noguchi, H.; Kinoshita, T.; Saitoh, H.; Sankawa, U.;PlantaMed, 1981, 43, 18-23. [204] Iwasaki, T.; Kondo, K.; Kuroda, T.; Moritani, Y.; Yamagata, S.; Sugiura, M.; Kikkawa, H.; Kaminuma, O.; Ikezawa, K.; J. Med. Chem., 1996, 39, 2696-2704. [205] Ukita, T.; Nakamura, Y.; Kubo, A.; Yamamoto, Y.; Takahashi, M.; Kotera, J.; Ikeo, T; J.Med Chem., 1999, 42, 1293-1305. [206] Ukita, T.; Sugahara, M.; Terakawa, Y.; Kuroda, T.; Wada, K.; Nakata, A.; Ohmachi, Y.; Kikkawa, H.; Ikezawa, K.; Naito, K.; J. Med Chem., 1999, 42, 1088-1099. [207] Nikaido, T.; Ohmoto, T.; Kinoshita, T.; Sankawa, U.; Nishibe, S,; Hisada, S.; Chem. Pharm. Bull, 1981, 29, 3586-3592. [208] Shen, T.Y.; Hussaini, I. M.; MethodsEnzymoL, 1990, 757, 446-454. [209] Shen, T.Y.; Lipids, 1991, 26, 1154-1156. [210] Iwakami, S; Wu, J. B.; Ebizuka, Y.; Sankawa, U.; Chem. Pharm. Bull, 1992,40, 1196-1198. [211] Castro-Faria-Neto, H.C.; Martins, M. A.; Silva, P.M.; Bozza, P.T.; Cruz, H.N.; de Queiroz Paulo, M.; Kaplan, M.A.; Cordeiro, R.S.; J. Lipid Mediators, 1993, 7, 1-9. [212] Castro-Faria-Neto, H.C.; Bozza, P.T.; Cruz, H.N.; Silva, C.L.; Violante, FA.; Barbosa-Filho, J.M.; Thomas, G.; Martins, M A ; Tibirifa, E.V.; Noel, F.; Cordeiro, R.S.B.; Planta Med, 1995, 61, 101-105.

286 [213] Herbert, J.M.; Castro-Faria-Neto, H.C.; Barbosa-Filho, J.M.; Cordeiro, R.S.; Tibiriga, E.; J. Lipid. Medial Cell. Signal, 1997, 77, 1-14. [214] Zhang, S.X.; Chen, K.; Liu, X.J.; Zhang, D.C.; Tao Wiedmann, T. W.; Leu, S.L.; McPhail, AT.; Lee, K.H.; J. Nat. Prod, 1995, 58, 540-547. [215] Zeng, H.W.; Jiang, Y.Y.; Cai, D.G.; Bian, J.; Long, K.; Chen, Z.L.; PlantaMed, 1997, 63, 296-298. [216] Chen, C.C; Hsin, W.C; Ko, F.N.; Huang, Y.L.; Ou, J.C; Teng, CM.; J. Nat. Prod, 1996, 59, 1149-1150. [217] Tanabe, Y.; Seko, S.; Nishii, Y.; Yoshida, T.; Utsumi, N.; Suzukamo, G.,J. Chem. Sac. Perkin Trans. 1, 1996, 2157-2165. [218] Jung, K.Y.; Lee, LS.; Oh, S.R.; Kim, D.S.; Lee, H.K.; Phytomedicine, 1997, 4, 229-231. [219] Lee, I.S.; Jung, K.Y.; Oh, S.R.; Kim, D.S.; Kim, J.H.; Lee, J.J.; Lee, H. K.; Lee, S. H.; Kim, E. H.; Cheong, C; Arch. Pharmacol. Res., 1997, 20, 633-636. [220] Lee, LS.; Jung, K.Y.; Oh, S.R.; Park, S.H.; Ahn, K.S.; Lee, H.K.; Biol. Pharm. Bull, 1999, 22,265-267. [221] Jung, K.Y.; Kim, D.S.; Oh, S.R.; Park, S.H.; Lee, I. S.; Lee, J.J.; Shin, D.H.; Lee, H.K.; J. Nat. Prod, 1998, 61, 808-811. [222] Hirose, N.; Inoue, T.; Nishihara, K.; Sugano, M.; Akimoto, K.; Shimizu, S.; Yamada, H.; J. Lipid Res., 1991, 32, 629-638. [223] Ogawa, H.; Sasagawa, S.; Murakami, T.; Yoshizumi, H.; Clin. Exp. Pharmacol. Physiol. Suppl, 1995,1, S310-S312. [224] Fukuda, N.; Miyagi, C; Zhang, L.; Jayasooriya, A. P.; Sakono, M.; Yamamoto, K.; Ide, T.; Sugano, M.; J. Nutr. Sci. Vitaminol, 1998, 44, 715-722. [225] Kwon, B.M.; Jung, H.J.; Lim, J.H.; Kim, Y.S.; Kim, M.K.; Kim, Y.K.; Bok, S.H.; Bae, K.H.; Lee, L R.; PlantaMed, 1999, 65, 7476. [226] Iwasaki, T.; Kondo, K.; Nishitani, T.; Kuroda, T.; Hirakoso, K.; Ohtani, A.; Takashima, K.; Chem. Pharm. Bull, 1995, 43, 17011705. [227] Kuroda, T.; Kondo, K.; Iwasaki, T.; Ohtani, A.; Takashima, K.; Chem. Pharm. Bull, 1997, 45, 678-684. [228] Prasad, K. Atherosclerosis, 1997, 132, 69-76.

287 [229] Prasad, K.; Mantha, S.V.; Muir, A.D.; Westcott, N.D.; Atherosclerosis, 1998, 136, 367-375. [230] Prasad, K.; Circulation, 1999, 99, 1355-1362. [231] Potterat, O.; Curr. Org. Chem., 1997,1, 415-440. [232] Faure, M.; Lissi, E.; Torres, R.; Videla, L.A.; Phytochemistry, 1990, 29, 3773-3775. [233] Morikawa, M.; Fukuchi, K.; Inoue, M.; Tsuboi, M.; J. Pharm. Pharmacol., 1992, 44, 859-861. [234] Morikawa, M.; Abe, M.; Yamauchi, Y.; Inoue, M.; Tsuboi, M.; Biochem. Biophys. Res. Commun., 1990, 168, 194-199. [235] Nakasone, Y.; Takara, K.; Wada, K.; Tanaka, J.; Yogi, S.; Nakatani, N.;Biosci. BiotechnoL Biochem., 1996, 60, 1714-1716. [236] Lu, H.; Liu, G.T.; Chem. Biol Interact., 1991, 78,11-U. [237] Lu, H.; Liu, G.T.; PlantaMed, 1992, 58, 311-313. [238] Xue, J.Y.; Liu, G.T.; Wei, H.L.; Pan, Y.; Free Radical Biol Med, 1992,12, 127-135. [239] Malaisse, W.J.; Liu, G.T.; Med Sci. Res., 1998, 26, 225-226. [240] Li, L.N.; Pure Appl Chem., 1998, 70, 547. [241] Maeda, S.; Masuda, H.; Tokoroyama, T.; Chem. Pharm. Bull, 1994, 42, 2500-2505. [242] Maeda, S.; Masuda, H.; Toyokoroyama, T.; Chem. Pharm. Bull, 1994,^2,2506-2513. [243] Wang, J.P.; Raung, S.L.; Hsu, M.F.; Chen, C.C; Br. J. Pharmacol, 1994, 113, 945-953. [244] Kuriyama, K.; Murai, T.; J. Jpn. Soc. Biosci. Biotech. Agrochem., 1996, 70, 161-167. [245] Kang, M.H.; Naito, M.; Tsujihara, N.; Osawa, T.; J. Nutr., 1998, 128, 1018-1022. [246] Ashakumary, L.; Rouyer, I.; Takahashi, Y.; Ide, T.; Fukuda, N.; Aoyama, T.; Hashimoto, T.; Mizugaki, M.; Sugano, M.; Met. Clin. Exp., 1999,48, 1303-1313. [247] Fukuda, N.; Zhang, L.; Kodama, M.; Sakono, M.; Ide, T.; Yamamoto, K.; Sugano, M.; J. Nut. Sci. Vitaminol, 1999, 45, 437448. [248] Haraguchi, H.; Ishikawa, H.; Shirataki, N.; Fukuda, A.; J. Pharm. Pharmacol., 1997, 49, 209-212.

288 [249] Matsuda, H.; Ishikado, A.; Nishida, N.; Ninomiya, K.; Fujiwara, H.; Kobayashi, Y.; Yoshikawa, M.; Bioorg. Med. Chem. Lett, 1998, 8, 2939-2944. [250] Peng, H.L.; Chen, D.F.; Lan, H.X.; Zhang, X.M.; Gu, Z.; Jiang, MB.., Acta Pharmacol. Sin., 1996,17, 538-541. [251] Fukuyama, Y.; Nakahara, M.; Minami, H.; Kodama, M.; Chem. Pharm. Bull, 1996, 44, 1418-1420. [252] Jong, T.T.Chau, S.W.; Phytochemistry, 1998, 49, 2145-2148. [253] Hikino, H.; Kiso, Y.; Wagner, H.; Fiebig, M.; Planta Med, 1984, 50, 248-250. [254] Hikino, H.; Kiso, Y.; Taguchi, H.; Ikeya, Y.; Planta Med, 1984, 50, 213-218. [255] Maeda, S.; Takeda, S.; Miyamoto, Y.; Aburada, M.; Harada, M.; Jpn. J. Pharmacol., 1985, 38, 347-353. [256] Miyamoto, K.; Hiramatsu, K.; Ohtaki, Y.; Kanitani, M.; Nomura, M.; Aburada, M.,Biol. Pharm. Bull., 1995, 75, 1443-1445. [257] Ohkura, Y.; Mizoguchi, Y.; Morisawa, S.; Takeda, S.; Aburada, M.; Hosoya, E.; Jpn. J. Pharmacol., 1990, 52, 331-336. [258] Ohkura, Y.; Mizoguchi, Y.; Sakagami, Y.; Kobayashi, K.; Yamamoto, S.; Morisawa, S.; Takeda, S.; Aburada, M.; Jpn. J. Pharmacol., 1987, 44, 179-185. [259] Nagai, H.; Yakuo, I.; Aoki, M.; Teshima, K.; Ono, Y.; Sengoku, T.; Shimazawa, T.; Aburada, M.; Koda, A.; Planta Med, 1989, 55, 1317. [260] Kubo, S.; Ohkura, Y.; Mizoguchi, Y.; Matsui Yuasa, I.; Otani, S.; Morisawa, S.; Kinoshita, H.; Takeda, S.; Aburada, M.; Hosoya, E.; Planta Med, 1992, 58, 489-492. [261] Ohtaki, Y.; Nomura, M.; Hida, T.; Miyamoto, K.; Kanitani, M.; Aizawa, T.; Aburada, M.; Biol. Pharm. Bull., 1994, 77, 808-814. [262] Ohtaki, Y.; Hida, T.; Hiramatsu, K.; Kanitani, M.; Ohshima, T.; Nomura, M.; Wakita, H.; Aburada, M.; Miyamoto, K.; Anticancer Res., 1996,76,751-755. [263] Nomura, M.; Ohtaki, Y.; Hida, T.; Aizawa, T.; Wakita, H.; Miyamoto, K, Anticancer Res., 1994, 14, 1967-1971. [264] Nomura, M.; Nakachiyama, M.; Hida, T.; Ohtaki, Y.; Sudo, K.; Aizawa, T.; Aburada, M.; Miyamoto, K. I.; Cancer Lett, 1994, 76, 11-18.

289 [265] Yamada, S.; Murawaki, Y.; Kawasaki, H.; Biochem. Pharmacol, 1993, 46, 1081-1085. [266] Shiota, G.; Yamada, S.; Kawasaki, H.; Res. Commun. Mol. Pathol. Pharmacol, 1996, 94, 141. [267] Ko, K.M.; Mak, D.H.; Li, P.C; Poon, M.K.; Ip, S.P.; Jpn. J. Pharmacol, 1995, 69, 439-442. [268] Ko, K.M.; Ip, S.P.; Poon, M.K.; Wu, S.S.; Che, C.T.; Ng, K.H.; Kong, Y.C.-PlantaMed, 1995, 67, 134-137. [269] Ko, K.M.; Mak, D.H.F.; Li, P.C; Poon, M.K.T.; Ip, S.P.; Phytother. Res., 1996, JO, 450-452. [270] Li, J.X.; Shi, Q.; Xiong, Q.B.; Prasain, J.K.; Tezuka, Y.; Hareyama, T.; Wang, Z.T.; Tanaka, K.; Namba, T.; Kadota, S.; Planta Med, 1998, 6< 628-631. [271] Yang, X.W.; Miyashiro, H.; Hattori, M.; Namba, T.; Tezuka, Y.; Kikuchi, T.; Chen, D.F.; Xu, G.J.; Hori, T.; Extine, M.; Mizuno, H.; Chem. Pharm. Bull, 1992, 40, 1510-1516. [272] Fernandes, E.R.; Carvalho, F.D.; Remiao, F.G.; Bastos, M.L.; Pinto, M.M.; Gottlieb, O.R.; Pharm. Res., 1995,12, 1756-1760. [273] Shimizu, S.; Akimoto, K.; Shinmen, Y.; Kawashima, H.; Sugano, M.; Yamada, H., Lipids, 1991, 26, 512-516. [274] Shimizu, S.; Kawashima, H.; Akimoto, K.; Shinmen, Y.; Sugano, H.; Yamada, H.; Phytochemistry, 1992, 31, 757-760. [275] Umeda-Sawada, R.; Takahashi, N.; Igarashi, O.; Biosci. Biotechnol Biochem., 1995, 59, 2268-2273. [276] Umeda-Sawada, R.; Ogawa, M.; Okada, Y,; Igarashi, O.; Biosci. Biotechnol Biochem., 1996, 60, 2071-2072. [277] Griffiths, G.; Jones, H.E.; Eaton, C.L.; Shimizu, S.; Stobart, A.K.; Phytother. Res., 1998, 12, 417-421. [278] Umeda-Sawada, R.; Ogawa, M.; Igarashi, O.; Lipids, 1998, 33, 567572. [279] Akimoto, K.; Kitagawa, Y.; Akamatsu, T.; Hirose, N.; Sugano, M.; Shimizu, S.; Yamada, H.; Am. Nutr. Metab., 1993, 37, 218-224. [280] Yamashita, K.; Nohara, Y.; Katayama, K.; Namiki, M.; J. Nutr., 1992, 122, 2440-2446. [281] Yamashita, K.; lizuka, Y.; Imai, T.; Namiki, M.; Lipids, 1995, 30, 1019-1028. [282] Kamal Eldin, A.; Pettersson, D.; Appelqvist, LA.; Lipids, 1995, 30, 499-505.

290 [283] Kimura, M.; Suzuki, J.; Yamada, T.; Yoshikazi, M.; Kikuchi, T.; Kadota, S,; Matsuda, S.; PlantaMed., 1985, 57, 291-293. [284] Prieto, J.M.; Recio, M.C.; Giner, R.M.; Manez, S.; Massmanian, A.; Waterman, P.G.; Rios, J.L.; Z. Naturforsch. C, 1996, 57, 618-622. [285] Yasukawa, K.; Ikeya, Y.; Mitsuhashi, H.; Iwasaki, M.; Aburada, M.; Nakagawa, S.; Takeuchi, M.; Takido, M.; Oncology, 1992, 49, 6871. [286] Baba, A.; Kawamura, N.; Makino, H.; Ohta, Y.; Taketomi, S.; Sohda, T.; J. Med. Chem., 1996, 39, 5176-5182. [287] Baba, A.; Oda, T.; Taketomi, S.; Notoya, K.; Nishimura, A.; Makino, H.; Sohda, T.; Chem. Pharm. Bull, 1999, 47, 369-374. [288] Tsumga, T.; Ebizuka, Y.; Nakajima, J.; Chun, Y. T.; Noguchi, H.; litaka, Y.; Sankawa, U.; Chem. Pharm. Bull, 1991, 39, 3265-3271. [289] Nonaka, M.; Yamashita, K.; lizuka, Y.; Namiki, M.; Sugano, M.; Biosci. Biotechnol Biochem., 1997, 61, 836-839. [290] Serra, M.F.; Diaz, B. L.; Barreto, E.G.; Pereira, A.P.B.; Lima, M.C.R.; Barbosa-Filho, J.M.; Cordeiro, S.B.; Martins, M.A.; de Silva, F.M.K.; PlantaMed, 1997, 63, 207-212. [291] Therien, M.; Fitzsimmons, B.J.; Scheigetz, J.; Macdonald, D.; Choo, L.Y,; Guay, J.; Falgueyret, J.P.; Riendeau, D.; Bioorg. Med. Chem. Lett., 1993, 3, 2063-2066. [292] Ducharme, Y.; Brideau, C; Dube, D.; Chan, C.C; Falgueyret, J.P.; Gillard, J.W.; Guay, J.; Hutchinson, J.H.; McFarlane, C.S.; Riendeau, D.; Scheigetz, J.; Girard Y.; J. Med. Chem., 1994, 37, 512-518. [293] Chauret, N.; Li, C; Ducharme, Y.; Trimble, L.A.; Yergey, J.A.; Ramachandran, C; Nicoll Griffith, D.A.; Drug Metab. Dispos., 1995, 23, 65-71. [294] Delorme, D.; Ducharme, Y.; Brideau, C; Chan, C.C; Chauret, N.; Desmarais, S.; Dube, D.; Falgueyret, J.P.; Fortin, R.; Guay, J.; Hamel, P.; Jones, T.R.; Lepine, C; Li, C; McAuliffe, M.; McFarlane, C.S.; Nicoll Griffith, D.A.; Riendeau, D.; Yergey, J.A.; Girard, Y.; J. Med Chem., 1996, 39, 3951-3970. [295] Yu, S.M.; Ko, F.N.; Wu, T.S.; Lee, J.Y.; Teng, CM.; Eur. J. Pharmacol, 1994, 256, 85-91. [296] Yu, S.M.; Wu, T.S.; Teng, CM.; Br. J. Pharmacol, 1994, 777, 906912.

291 [297] Muller-Jakic, B.; Greger, H.; Vermes, B.; Bauer, R. In Flavonoids and Bioflavonoids, Antus, S.; Gabor, M.; Vetschera, K., Eds.; Akademia Kiado: Budapest, 1996, 149-156. [298] Chae, S.H.; Kim, P.S.; Cho, J.Y.; Park, J.S.; Lee, J.H.; Yoo, E.S.; Baik, K.U.; Lee, J.S.; Park, M.H.; Arch. Pharm. Res., 1998, 2/, 6769. [299] Cho, J.Y.; Park, J.; Yoo, E.S.; Yoshikawa, K.; Baik, K.U.; Lee, J.; Park, M.H. Arch. Pharm. Res., 1998, 27, 12-16. [300] Hirano, T.; Wakasugi, A.; Oohara, M.; Oka, K.; Sashida, Y.; Planta Med, 1991,57,331-334. [301] Okuyama, E.; Suzumura, K.; Yamazaki, M.; Chem. Pharm. Bull., 1995, 43, 2200-2204. [302] Cheng, H.T.; Chang, H.; Thorac. Cardiovasc. Surg., 1995, 43, 7376. [303] Pieters, L.; DeBruyne, T.; VanPoel, B.; Vingerhoets, R.; Totte, J.; vandenBerghe, D.; Vlietinck, A.; Phytomedicine, 1995, 2, 17-22. [304] Hall, A.v.; Parbtani, A.; Clark, W.F.; Spanner, E.; Keeney, M.; Chin Yee, I.; Philbrick, D.J.; Holub, B.J.; Am. J. Kidney Dis., 1993, 22, 326-332. [305] Hirano, T.; Homma, M.; Oka, K.; Naito, T.; Hosaka, K.; Mitsuhashi, H.,LifeSci., 1991, 49, 1871-1878. [306] Plante, G.E.; Prevost, C ; Chainey, A.; Braquet, P.; Sirois, P.; Am. J. Physiol., 1987, 253, R375-R378. [307] Watanabe, K.; Watanabe, H.; Goto, Y.; Yamaguchi, M.; Yamamoto, N.; Hagino, K.; Planta Med, 1983, 49, 103-108. [308] Yamazaki, M.; Hirota, K.; Chiba, K.; Mohri, T.; BioL Pharm. Bull, 1994, 77, 1604-1608. [309] Fukuyama, Y.; Hasegawa, T.; Toda, M.; Kodama, M.; Okazaki, H.; Chem. Pharm. Bull., 1992, 40, 252-254. [310] Bodesheim, U;.Holzl, J.; Pharmazie, 1997, 52, 386-391. [311] Madrigal Bujaidar, E.; Diaz Barriga, S.; Cassani, M.; Molina, D.; Ponce, G.;Mutat. Res., 1998, ^72, 139-144. [312] Madrigal Bujaidar, E.; Diaz Barriga, S; Cassani, M.; Marquez, P.; Revuelta, P.; Mutat. Res. Genet. Toxicol Eviron. Mutagen., 1998, 419, 163-168. [313] Miyazawa, M.; Okuno, Y.; Oshiro, K.; Kasahara, H.; Shimamura, H.; Nakamura, S.; Kameoka, H.; Biosci. Biotechnol Biochem., 1998, 62, 1425-1427.

292 [314] Kulling, S.E.; Jacobs, E.; Pfeiffer, E.; Metzler, M.; Mutat. Res., 199S,4J6, 115-124. [315] Kulling, S.E.; Jacobs, E.; Metzler, M.; N.-S. Arch. Pharmacol., 1998, 357^ 553

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

293

CROCUS SATIVUS'mOl.OGlCA\. ACTIVE CONSTITUENTS *M. LIAKOPOULOU-KYRIAKIDES^ AND D.A. KYRIAKIDIS^ ^Dept. Chem, Engineering, Section of Chemistry and ^Faculty of Chemistry, Aristotle University ofThessaloniki, Greece 54006 ABSTRACT: The recent advances in separation and characterization of the volatile and non-volatile components of the dried stigmas of Crocus sativus (saffron) are presented. The volatiles with a very strong odor are consistent of more than 34 components that are mainly terpenes, terpene alcohols and their esters. Non-volatiles include crocins 1,2,3 and 4 that are responsible for the red or reddish brown color of stigmas together with carotenes, crocetin, picrocrocin (a glycosidic precursor of safranal) the bitter substance and safranal the major organoleptic principle of the stigmas. The methodology and techniques developed for the analysis of saffron metabolites including various chromatographic and spectroscopic techniques (TLC, HPLC, GC-MS, LC-MS, NMR.) are described. An extended study of the biological effects of either extracts or specific constituents of saffron (crocetin, picrocrocin and safranal) or extracts from its roots, from experiments in vivo and in vitro are discussed. A growing body of evidence indicates that carotenoids possess anticarcinogenic, antimutagenic and immunomodulating effects. Other effects such as lowering of blood pressure, stimulation of respiration, sedative effects, inhibition of human platelet aggregation in vitro and the dietary effects are reported as well. In addition the use of its coloring agents (crocins and carotenes) in dying cotton and wool fabrics and/or other uses in industry as well as attempts for biotechnological production of saffron and its metabolites are mentioned.

INTRODUCTION Crocus sativus L (Family Iridaceae) a perennial herb, native to the Eastern Mediterranean region, is cultivated in Spain, France, Greece, Italy, India, Turkey, Iran, e.t.c. Saffron-the yellow-orange stigmas from a small purple Crocus sativus flower- is the world's most expensive spice. According to the Greek mythology a gorgeous mortal named Crocos fell hard for the nymph Smillax who rebuffed Crocos' overtures and then Crocos became a lovely purple flower, the well known Crocus sativus. Greek myths and poetry exhibit an extravagant admiration of the color

294 and perfume from saffron. Homer sings the "Saffron mom" whereas gods and goddesses, heroes and nymphs and vestals are clothed in robes of saffron hue. To the nations of Eastern Asia, saffron's yellow dye was the perfection of beauty and its penetrating odor a perfect ambrosia. From historical records, saffron yellow shoes formed part of the dress of the Persian Kings. Crocus sativus plants (schemes 1,2) require strict agroclimatic conditions for their growth, that influence the quality of the spice. Samples obtained from different geographical locations and from different processing methods show variations with respect to color, flavor and the bitter principle.

Scheme 1. Crocus flowers. CH3

CH3

OR2

all-trans CH2OH Q

OH HO HO

\. 'CH2

CH2OH Q

H0--^Y"^V-^^ \

^ OH X Y Crocetin (I) R„ R2=H, Crocins (II) R„ R2=X, Y

OH

295 Crocins (II) that are glycosyl esters of crocetin (I) are the major components of saffron. Crocins {trans and cis isomers) belong to the family of carotenoids that are water-soluble. Safranal (III), a monoterpene aldehyde and picrocrocin (IV) precursor of safranal are also main components.

CH3 Safranal (III) H3C

CH3

CH2OH Q

"''^^^^^W^''

-

^CH3

Picrocrocin (IV)

Volatile constituents Analysis of saffron volatile components has been reported by various investigators [1-7]. Volatiles with a very strong odor are consistent of more than 34 components, that are mainly terpenes, terpene alcohols and esters. The major organoleptic principle present (60-70% of the volatiles) is safranal (2,6,6-trimethylcyclohexa-l,3-dien-l-carboxaldeyde) and is formed by de-glucosylation of picrocrocin, the bitter compound. Zarghami and Heinz [2,3] have identified volatile compounds similar or different than safranal. The structure of picrocrocin has been established by Buchecker and Eugster [8]. Safranal has been determined by TLC, GC and spectrophotometry [4]. Thin layer chromatography plates impregnated with 0.25 mm thick silica gel G were used with n-butanol-acetic acidwater (4:1:1) solvent system. Crocins and crocetins were visible in white light whereas picrocrocin can be detected under UV lamp (254 nm) as a dark brown fluorescent spot. For determination of safranal [4], crude ethanolic extract was extracted with diethyl ether. Ether layers were pooled, air dried and the volume made up to 0.1 ml. TLC of the extract

296 developed in the solvent system hexane-ethyl acetate (9:1) gave a spot that became visible on spraying with 2,4-dinitrophenylhydrazine (DNPH) reagent. Safranal gives a UV spectrum with X^max 308 nm.

Scheme 2. Crocus sativus plant. 1, roots; 2, bulbs; 3, corm; 4, leaves; 5, close flower; 6 open flower.

Sujata et al [4] have also reported the use of SE-30 and Carbowax columns for gas chromatographic analysis of the saffron volatile constituents. Isothermal run with SE-30 column resolved safranal into a sharp single peak at a retention time of 3.6 min. Catwallander et al [9] and Tarantilis and Polissiou [7] have reported, from the analysis of saffron volatiles, a minor component 2-hydroxy4,4,6-trimethyl-2,5-cyclohexadien-l-one that is considered the most powerful aroma constituent of saffron followed by safranal. Monoterpene aldehydes, isophorone and isophorone-related compounds have been also found in saffron. Zarghami and Heinz [3] have identified six new isophorone-related compounds listed in Table 1 from ether extracts of saffron by IR, NMR, UV and MS. In the same table are listed the most common volatiles of safron. The structures of carotenoid derived volatiles found by various researchers are shown below (scheme 3). Quantitative determination of picrocrocin and crocetin by High Performance Thin Layer Chromatography (HPTLC) is reported by Corti et al [\Q]. In addition, Kanasawud and Crouzet [11,12] and Crouzet and

297 Kanasawud [13] have proposed a mechanism for the formation of volatile compounds of saffron by thermal degradation of carotenoids including acarotene (VI) and lycopene (VII) degradation.

Scheme 3. Crocus carotenoid-derived volatiles isolated from stigmas.

Non-volatile constituents Qualitative methods that are used for determination of the color and flavor profiles of saffron have been reviewed by Sampathu et al{\A]. Separation of crocetin glycosyl esters by HPLC is referred by Pfander and Rychener [15], Pfander and Wittwer [16,17], Pfister [18], Sujata et al [4], Tarantilis et al [19] and Solinas and Cicheli [20]. A pretreatment of crude ethanol extract of stigmata by gel filtration on a G-50 column prior to HPLC is reported by Pfander and Rychener [15]. This treatment is necessary to

298 remove carbohydrates, that disturb the separation of pigments. HPLC analysis was performed using I) Lichrosorb SI 60 column (Merck, Darmstadt) with eluent ethyl acetate-isopropanol-water (56:34:10) and II) Lichrosorb RP-18 (Merck) with methanol-water (60:40) and UV detector in both cases at 440nm. It was found that, the main pigment was crocin (40-45%) followed by the mixed ester (35%), the diglucosyl ester (10%) and the two monoesters (2% each). Reverse-phase systems were used for the separation of the diglycosyl esters (polar compounds) including digentiobiosyl, diglucosyl- and the mixed ester with Lichrosorb RP-18 column. Crocin, crocetins and picrocrocin can be separated according to Sujata et al [4] either on a gradient run from 20 to 80% (v/v) acetonitrile/water in 20min at a flow rate of 0.5 ml/min and detection at 308nm or isocratic run with 76% (v/v) using a Shimadzu 15cmx4.9 mm ID CLC-ODS column. Table 1. The most volatile components of saffron Compound 1. Picrocrocin 2. Safranal 3. Isophorone 4. Isophorone related compounds

Chemical name

Formula weight

4-(P-D-glycopyranosyloxy)-2,6,6-trimethyl-1 -cyclohexen-1 carboxaldehyde

C,6H2607

2,6,6-trimethylcyclohexa-1,3-dien-1 -carboxaldehyde

C.OHHO

3,5,5-trimethyl-2-cyclohexen-l-one

C9HJ40

3,5,5-trimethyl-4-hydroxy-l-cyclohexanon-2-ene

C9H,402

5.

»

3,5,5-trimethyl-l,4-cyclohexadione

C9H,402

6.

»

3,5,5-trimethyl-l,4-cyclohexadion-2-ene

C9H,202

7.

»

3,5,5-trimethyl-2-hydroxy-1,4-cycIohexadion-2-ene

C9H12O3

8.

»

2,6,6-trimethyl-4-hydroxy-1 -cyclohexene-1 -carboxaldehyde

C10H16O2

9.

»

2,4,4-trimethyl-3-formyl-6-hydroxy-2,5-cycIohexadien-1 -one

CloH,203

As it has been reported previously [21] direct spectrophotometric determination of the crude ethanol extract at A^max values corresponding to crocin, crocetin, picrocrocin and safranal respectively, does not give reproducible results. Nevertheless, use of crude plant extract directly for determination of the various components is referred by Tarantilis et al [19] using HPLC with UV-Visible, Photodiode-Array Detection (UV-Vis-DAD) and mass

299 spectrometry coupled in line. ES (electrospray) instead of TS (thermospray) interfaces combined with HPLC-UV-Vis-DAD can also be used for characterization of other unknown compounds as pointed out by the same authors. UV-Vis spectra of saffron components are characteristic; picrocrocin [4-(a-D-glycopyranosyloxy)-2,6,6-trimethyl-1 -cyclohexen-1 carboxaldehyde] exhibits a characteristic broad absorption band at 250nm Oxidative decarboxylation of picrocrocin gives the precursor isophorone (3,5,5-trimethyl-2-cyclohexen-l-one) that is also a minor component of saffron. Crocus derivatives show characteristic UV-Vis spectra of the carotenoid moiety in the molecule. Carotenoids absorb in the visible region with double peaks between 400 and 500nm, that vary with the position and number of cis double bonds, and a single peak between 320 and 340nm. These characteristic peaks are presented by the cis isomer. The UV-spectra of dlX-trans glycosidic carotenoids show two bands one at 256nm (glycosyl esters bonds of crocins) and the other one between 400 and 500nm with A^ax at 437nm. The above data show that UV-Vis spectra can be used for identification of cis and trans isomers of carotenoids. The cis isomer shows 3 absorption bands whereas two absorption bands correspond to trans isomers. A large number of carotenoid compounds have been isolated from saffron. Crocins and crocetin have been already mentioned above and are listed in Table 2. a-Carotene (V), p-carotene (VI), lycopene (VII), zeaxanthin (VIII), phytoene (IX), phytofluene (X) are the minor ones [22]. Straubinger et al [23] have reported the identification of four novel glycoconjugated carotenoid breakdown products of saffron that are the PD-glucosides of (4R)-4-hydroxy-3,5,5-trimethylcyclohex-2-enone, (4S)-4hydroxy-3,5,5-trimethylcyclohex-2-enone and (4S)-4-(hydroxymethyl)3,5,5-trimethylcyclohex-2-enone as well as the P-D-gentiobiosyl ester of 2-methyl-6-oxohepta-2,4-dienoicacid. It should be mentioned that, in saffron all crocin derivatives, except crocin-1, occur as pairs of cis-trans isomers [24,25]. Minor carotenoids such as isorhamnetin-4-0-a-L-rhamnapyranosyl(l-^2)-p-D-glucopyranoside and P-(p-hydroxyphenyl) ethanol-a-0-a-L-rhamnopyranosyl(l-->2)p-D-glucopyranoside have been identified by chemical and spectroscopic analysis. Preparative HPLC and TLC have been used by Castellar et al

300

[26] and Iborra et al [27] respectively, for the separation of saffron secondary metabolites.

a-Carotene (V) HsC^

P-Carotene (VI)

OH

"CHs

Zeaxanthin (VIII)

Responsible for the flavor of saffron, as it has been mentioned, is the bitter tasting glycoside picrocrocin, and its degradation product, safranal.

301 An oxidative cleavage of zeaxanthin has been proposed [28] for the formation of carotenoid degradation. Table 2. Crocetin and crocins of saffron Chemical name

Compound Crocetin Crocin-1

Formula weight

2,6,11,15-tetramethylhexadeca-2,4,6,8,10,12,14heptaenedioic acid

C20H24O4

Ri= P-D-glucosyl (Y),

C26H34O9

R2=H Crocin-2

Ri= p-D-gentiobiosyl (X),

C32H44O14

R2=H Crocin-2'

Ri=R2= P-D-glucosyl (Y)

C32H44O14

Crocin-3

Ri= P-D-gentiobiosyl (X),

C3SH54O19

R2= p-D-glucosyl (Y) Crocin-4

Ri=R2= P-D-gentiobiosyl (X)

C44H64O24

Crocin-5

Ri= Three P-D-glucosyl (Z), R2= P-D-gentiobiosyl (X)

C50H74O29

Extractions and pre-separations Various methods have been applied for the recovery of saffron volatiles including steam distillation (SD), micro-simultaneous steam distillation extraction (MSDE) and vacuum head space (VHS) analysis. Simultaneous distillation-extraction with diethyl ether-pentane is reported by Romer and Rennel [29] and head space analysis by Mookherjee et al [30] and Joulain et al [31]. As reported by Godefroot et al [32], Nickerson and Lickens [33] and Schmitt [34] in some cases MSDE method gives the volatiles in low concentrations whereas high boiling point components are obtained in high concentrations. The use of the three mentioned methods for isolation and identification of the aroma components of saffron has been reported [7]. Soxhlet extraction is mainly used for the preparation of saffron crude extract and isolation of coloring agents using methanol or ethanol in the dark and under nitrogen atmosphere [4,27]. Pre-separation of the various components is achieved by chromatographic techniques, including silica gel or gel filtration column chromatography [16], Multilayer Coil Coulter Current Chromatography (MLCCC) [10]. A pre-separation of methanolic extract of saffron by MLCCC method is reported by Straubinger et al

302

[23]. Fractions containing picrocrocin were re-chromatographed on the same column using CHCh/methanol (1.75:0.3) and the fraction corresponding to picrocrocin was further identified by ^"^C-NMR and MS. Of the various analytical methods described, the ISO recommended a thin-layer chromatographic technique [35] for qualitative analysis of saffron, whereas quantitative HPLC methods have been described for the analysis of commercial saffron products [36]. Other reports on chemical analysis of saffron constituents including combination of the reported methods, crystal structure analysis, Fourier Transform-Infra Red (FT-IR) and Raman analysis are worth mentioned [37-40]. Very interesting is also the review article by Rios et al[A\] on chemical analysis of saffron. Flavonoids and other constituents Flavonoids occur in plants as mono- and di-glycosides and derive from 2phenyl-y-benzopyrone. Their UV spectra depend on the number and substitution of the hydroxy groups in the phenol nucleus. Flavonoid derivatives have been also found in stigmata of Crocus sativus. Among them kaempherol diglycoside has been isolated and identified by mass spectra [38]. Four isolectins have been isolated from bulbs of Crocus sativus with approximately molecular weight 48 KDa as determined by gel filtration chromatography [42]. Rivoflavine and thiamin are also constituents of saffron [43]. Very recently, anthocyanins were identified in the flowers of Crocus sativus [44]. Pharmacology and Medicine Carotenoids have been studied extensively and their biological effects including antioxidant, antitumor and in general their immunomodulating effects are known [45-52]. It has been reported that extracts of saffron show various pharmacological activities including antiviral [50] and antitumor activity [53-56]. The effect of saffron on cell colony formation and cellular and intracellular DNA, RNA and protein synthesis in malignant and non-malignant cells has been studied by Abdullaev and Fenkel [57,58]. Abdullaev and Frenkel detected a dose-dependent decrease in colony formation of A549 lung adenocarcinoma, cervical

303

epitheliod carcinoma and HeLa cells using saffron extracts and isolated crocetin respectively. Inhibition of DNA and RNA synthesis in isolated nuclei and suppression of the activity of the purified RNA polymerase II is also reported [59-60] The inhibitory effect of Crocus sativus in chemical carcinogenesis in mice and the modulatory effects on cis platininduced toxicity in mice have been reported as v^ell [61,62]. The protective effect of crocetin against oxidative damage in rat primary hepatocytes has been further pointed out [63]. In vitro cytotoxic analysis of the main saffron compounds (crocins, crocetin, picrocrocin and safranal) using HeLa (human cervical epitheliod cancer) cells was carried out by Escribano et al [64]. It was found that in terms of LD50 values (dose inducing 50% cell growth inhibition on these cells) safranal has 0.8mM, crocin and picrocrocin 3mM respectively. Furthermore, Escribano et al [65] have also isolated from corms of Crocus sativus a glycoconjugate with cytotoxic activity as well. Experiments with carotenoids from Crocus sativus in HL60 cells were conducted by Tarantilis et al [66] where inhibition and induction of differentiation of these cells was studied. Furthermore, cytotoxicity experiments of dimethylcrocetin extracted from saffron were carried out on sarcoma-180, Ehrlich ascites carcinoma, P388 leukemia, Dalton's lymphoma ascites and primary cells from surgical excised tumor samples [53]. LD50 values were 9 mg/ml for Dalton's lymphoma and L1210 and a little higher for the other series with the maximum value in P388 cells. Dimethylcrocetin and crocin were also found to inhibit DNA and RNA synthesis [53]. These experiments were carried out using ^[H] thymidine or ^[H] uridine and evaluating DNA or RNA synthesis in the presence of saffron compounds. The protective effects of crocetin on bladder toxicity induced by cyclophosphamide has been reported by Nair et al [67]. The effect of saffron on vitamin A levels and the inhibition of growth of solid tumors in mice has been studied [68]. The reports of Lin and Wang [69] on the protection of crocin in the acute hepatic damage in rats induced by aflatoxin-Bl and dimethylnitrosamine and Wang et al [70,71] on the effects of crocetin on the hepatotoxicity and hepatic DNA binding of aflatoxins Bl in rats. We have recently shown (unpublished data) the antitumor activity in mice and the gonotoxic effect of a fraction isolated from methanol extract of saffron, that had been purified on a silica gel column using chloroform/methanol as eluent. P388 tumor cells were transplanted in

304

BDFi mice and the tumor static factor (T/C%, the percent of survival increase of mice) after the administration of the specific saffron extract was found to be 140 and 149 (blank 125).The same fraction w^as also tested against P388 cancer cells for its gonotoxic effect. The high SCE (sister chromatide exchange) values in combination with the significant decrease of PRI values are an indication of DNA damages caused by this fraction. SCEs values have been frequently used as a highly sensitive indicator of DNA damages and/or subsequent repair. The ability to excise and repair various types of damage to DNA seems to be a general property of living cells [72]. Lowering of blood pressure of anesthetized dogs and cats, stimulation of respiration, and reports for pulmonary oxygenation, cerebral oxygenation, alveolar increased oxygenation are among the other biological properties of saffron [73-75]. In a recent report, the reduction of hypervalent iron in myoglobin by crocin fiirther indicates the antioxidant action of carotenoids and saffron in general [76]. Saffron stimulates uterus and in ancient times it had been used for abortions. Crocus sativus bulbs are toxic to young animals and the stigmas in high doses are narcotic [77]. Hypoxia and in general the effect of saffron in cardiovascular diseases is reported by Grisola [78]. Inhibition of growth and induction of differentiation of promyelocytic Leukemia HL60 by the carotenoids of Crocus sativus has been studied by Tarantilis et al [19]. The IC50 values found, were 0,8 muM for dimethyl crocetin, 2 muM for crocetin and crocin respectively, and 5 muM induction of differentiation of HL60 cells. We have reported that bulbs of Crocus sativus contain a poly-histidine protein factor, which aggregates human platelets in vitro. The molecular weight of this factor as found by gel filtration chromatography on a Sephadex G-75 column and sodium dodecylsulfate (SDS) polyacrylamide slab gel electrophoresis is 42,000Da [79]. In addition, another low molecular weight protein factor isolated from the same crude extract was found to inhibit human platelet aggregation in vitro [80]. The effect of Crocus sativus on blood coagulation has been also examined by Nishio et al[Sll The antitumor activity of saffron was also evaluated against a variety of murine tumor models. Ascites tumors were induced by i.p. transplantation of one million cells in mice and appeared in 7-14 days depending on the type of tumor cells used [82]. Saffron was administered

305 orally (200ml/kg) for 9 days continuously, one day after tumor inoculum. It was found that saffron, at that dose delayed ascites tumor growth and increased the life span of the treated mice compared to untreated controls The role of saffron from dietary sources in chemoprevention and in general in modifying cancer has been nicely reviewed by Nair et al [82] and by Rios et a/ [41] where hypolipidemic and tissue oxygenation effect of active constituents of saffron are reported as well. Chemopreventive agents have been widely used in pharmacology in the treatment of various diseases including cancer [83]. The role of chemopreventers in human diet is also discussed by Stavric [84]. Alcohol extract from saffron has been also examined for learning and memory, in step through (ST) and step down (SD) tests, in normal and in leaming-and memory-impaired mice. The results showed that this extract ameliorates the impairment effects of ethanol on learning and memory processes and possesses a sedative effect [85-87]. In addition, recent experiments have shown that crocin and crocetin derivatives inhibit promotion of tumor in mice [88]. Other activities Dyeing of fabrics Recent reports have shown that saffron extracts were used for dyeing cotton and wool fibbers [89-92]. As it has been found, both crude methanol or aqueous extracts of saffron and the isolated crocin fraction give satisfactory results in terms of color fastness and color brightness. Crocin or saffron extracts together with other natural pigments such as berberine, azulen, curcumin, betanin, etc have been used as a hair dye [93]. In a US Patent [94] a scalp treatment composition is described for scalp and hair health and growth containing saffron, among the other plant extracts. Food coloring The use of natural pigments for food applications is gaining soil from day to day [95]. Curcumin, betaine, amarathine, anthocyanins and p-carotene are the most common and widely used pigments [95,96]. Saffron's coloring properties attributed mainly to water-soluble carotenoids are used for coloring of foods. The stability of these saffron pigments in aqueous

306 extracts and data on the condition of storage and applications as well as kinetic studies of caronenoid loss and changes under various conditions of water activity (aw) and temperature have been reported [97,98]. Changes in pigments (crocins) and volatiles of saffron {Crocus sativus L) during processing and storage have been reported by Raina et al [99]. According to these Authors, drying of stigmas between 35-50° C leads to highest percentage in pigments 15-17% and about 60% safranal.. A quality evaluation of saffron by sensory profile and gas chromatography has been presented by Narasimhan et al [100] where, by using a modified Steven's equation, the dose-response relationship for saffron flavor (safranal) can be estimated. Another interesting article on the various uses of saffron is that of Basker and Negbi [101]. Biosynthesis of C2o-carotenoids in Crocus sativus As Pfander and Schurtenberger have reported [28] the absence of C20hydrocarbon precursors of crocetin in saffi'on supports a degradation pathway for the biosynthesis of crocetin by the occurrence of picrocrocin and safranal. Two different pathways have been proposed for the biosynthesis of the C2o-aglycon of the main pigments a) oxidative degradation of a C40carotenoid such as zeaxanthin [102,28] or b) dimerization of two Ciocompounds such as geranylpyrophosphate followed by dehydrogenation and oxidation as reported for the formation of Cso-carotenoides from the famesylpyrophosphate and of C4o-carotenoids from geranyl geranylpyrophosphate (C20) [103]. The interesting article by Armstrong and Hearst [104] on the molecular biology of carotenoid pigment biosynthesis should be mentioned. Biotechnological production of Saffron and its metabolites Tissue culture of saffron including somatic embryogenesis and shoot regeneration has been first reported by George et al [105]. Induction of crocin, crocetin, picrocrocin and safranal synthesis in callus cultures of saffron-Crocw^* sativus L has been reported by Visvanath et al [106]. Callus cultures were obtained from floral buds on Murashige and Skoog's medium supplemented with 3% sucrose, 2,4-dichlorophenoxy acetic acid

307

(2 mg/1) and crocetin (0.5 mg/1). The experiments showed that red global callus (RGC) and red filamentous (RFC) callus cultures can produce crocin, crocetin, picrocrocin and safi-anal. The yield of safi-anal obtained fi*om RGC cultures was comparable to that obtained fi'om natural dried saffron. Picrocrocin content in both cultures was higher than that of stigmas and picrocrocin was less in both cultures. It was also shown in this study that C sativus callus tissues have the biosynthetic capability to produce the above reported metabolites that is considered an important step in developing a biotechnological process for saffron production. Enhanced crocetin glucosylation by encapsulation into maltosyl-Pcyclodextrin was reported by Cormier et al [107] as an attempt to overcome the difficulties of the previous method, where the use of DMSO inhibited the enzymatic glucosylation of crocetin catalyzed by cell free extracts of saffron. Furthermore, in vitro production of stigma-like cultures from stigma explants and synthesis of crocin, picrocrocin and safranal by saffron stigma-like structures proliferated in vitro and synthesis of aroma compounds by microbial transformation of isophorone with Aspergillus niger cultures have been reported [108-112]. General conclusions The biological effects and uses of saffron, extracts of saffron and its specific constituents show that the plant must achieve more attention from Scientists as a real "active plant". Biotechnological preparation of saffron and its active metabolites will bring more light to the above investigations and establish Crocus sativus as a "multipurpose" plant. In addition, the above data support the consideration that folklore is still a useful tool in predicting sources with various biological activities e.g. antitumor, antiviral activity, etc. Furthermore, future screening of plant extracts from folklore medicine might be more profitable, in terms of curing human infectious diseases.

308 ABBREVIATIONS TLC HPLC GC-MS LC DNPH TS ES DAD SD MSDE VHS MLCCC ISO FTIR T/C SCEs PRI aw RGC RFC DMSO

= Thin layer chromatography = High pressure liquid chromatography = Gas chromatography - Mass spectrometry = Liquid chromatography = 2,4 dinitrophenylhydrazine = Thermospray = Electrospray = Diode array detection = Steam distillation = Micro-simultaneous steam distillation extraction = Vacuum head space = Multilayer coil coulter current chromatography = International standardization organization = Fourier-transform infra red = Tumor static factor = Sister chromatide exchange values = Proliferation rate index = Water activity = Red global callus = Red filamentous callus = Dimethylsulfoxide

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8]

Rodel, W.; Petrzika, M. J. High Resolution Chromatogr, 1991, ]4, 771-774. Zarghami, N.S.; Heinz, D.E. Lebensm. Wiss. TechnoL, 1971, 4, 43-45. Zarghami, N.S.; Heinz, D.E. Phytochemistry, 1971, 10, 2755-2761. Sujata, V.; Ravinshankar, G.A.; Venkataraman, L.V. J. Chromatogr, 1992, 624, 497-502. Tarantilis, P.A.; Polissiou, M.; Manfait, M. J. Chromatogr, 1994, 664, 55-61. Alonso, G.L.; Salinas, M.R.; Esteban-Infantes, F.J.; Sanchez Femandes, M.A. J. Agric. FoodChem., 1996, 44, 185-188. Tarantilis, P. A; Polissiou, M. J. Agric. Food Chem., 1997, 45,459-462. Buchecker, R.; Eugster, C.H. Heh. Chim. Acta, 1973, 56, 1121-1124.

309

[9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34]

Catwallader, K.R.; Back, H.H.; Cai, M. ACS symposium series, N° 116 (spices); Amer. Chem. Soc. Pub. Washington DC, 1997, 66-79. Corti, C; Mazzei, E.; Ferri, S.; Franchi, G.G.; Dreassi, E. Phytoch. Analysis, 1996,7,201-203. Kanasawud, P.; Crouzet, J.C. J. Agric. Food Chem., 1990, 38, 237-243. Kanasawud, P.; Crouzet, J.C. J. Agric. Food Chem., 1990, 38, 1238-1242. Crouzet, J.C; Kanasawud, P. In Methods in Enzymology; Packer, L., Ed.; Academic Press: London, 1992, Vol. 213, Chapt. 7 Sampathu, M.; Shivashankar, S.; Lewis, Y.S. Crit. Rev. FoodSci. Nutr., 1984, 20, 123-157. Pfander, H.; Rychener, M. J. Chromatogr. 1982, 234, 443-447. Pfander, H.; Wittwer, F. Helv. Chim. Acta, 1975a, 58, 1608-1620. Pfander, H.; Wittwer, F. Helv. Chim. Acta, 1975a, 58, 2233-2236. Pfister, S.; Meyer, P.; Steck, A.; Pfander, H. J. Agric. Food Chem., 1996, 44, 2612-2615. Tarantilis, P. A.; Tsoupras G.; Polissiou, M. J. Chromatogr., 1995, 699, 107-118. Solinas, M.; Cichelli, A. Industrie Alimentari, 1988, 27, 634-640. Speranza, G.; Dada, G.; Manito, P.; Monti, D.; Grammatica, P. Gazz. Chim. ItaL, 1984,114, 189-192. Stahl, W.; Sundquist, A.R.; Hanusch, M.; Schwartz, W.; Sies, H. Clin. Chem., 1993,59,810-814. Straubinger, M.; Jezussek, M.; Waibel, R.; Winterhalter P. J. Agric. Food Chem., 1997,^5,1678-1681. Dhingra, V.; Seshadri, T.R. Indian J. Chemistry, 1975,13, 339-341. Song, C.Q.; Xu, R.S. Acta Chimica Sinica, 1991, 49, 917-920. Castellar, M.R.; Montijano, H.; Manjon, A.; Iborra, J.L. J. Chromatogr., 1993, 648, 187-190. Iborra, J.L.; Castellar, M.R.; Canovas, M.; Manjon, A. J. Food ScL, 1992, 57, 714-716. Pfander, H.; Schurtenberger, H. Phytochemistry, 1982, 21, 1039-1042. Romer, G.; Renner, E. Lebensmittel-Vnters. u-Forsch, 1974, 156, 329-331. Mookherjee, B.D.; Trenkle, R.W.; Wilson, R.A. J. Essent. Oil Res., 1987, 2, 8590. Joulain, D. In Progress in Essential oil Research; Brunke E.J., Ed., Walter de Gruyter& Co.: Berlin, 1986, 57-67. Godefroot, M.; Sandra, P.; Verzele, M. J. Chromatogr., 1981,203, 325-335. Nickerson, G.B.; Likens, S.T. J. Chromatogr., 1986, 21, 1-5. Schmitt, L. Perfum Flavor, 1984, 9, 149-154.

310

[35] [36] [37] [38] [39]

[40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62]

International standard, saffron specification, ISO 3632-1 and ISO 3632-2. 1993 (E), International Organization for standardization: Geneva, 1993. Lozano, P.; Castellar, M.R.; Simancas, M.J.; Iborra, J.L. J. Chromatogr. A, 1999, 830, 477-483. Corradi, C; Michelii, G. Boll Chim. Farm., 1979,118, 553-562. Oberdieck, R. Dtsck Lebensm. Rundsck, 1991, 87, 246-252. Van Calsteren, M.R; Bissonnette, M.C; Cormier, F.; Dufresne, C.; Takahito, I.; Yves LeBlank, J.C; Perreault, D.; Roewer, I. J. Agric. Food Chem., 1997, 45, 1055-1061. Tarantilis, P.A.; Polissiou, M.; Mentzafos, D.; Terzis, A.; Manfait, M. J. Chem. Crystalogr., 1994, 24, 739-742. Rios, J.L.; Ressio, M.L.; Giner, R.M. Phytother. Res., 1996, 10, 189-193. Oda, Y.; Tatsumi, Y. Biol. Pharm. Bulletin, 1993,16, 978-981. Bhat, J.V.; Brooker, R. Nature, 1953,172, 544-545. Norbaek,R.; Kondo, T. Phytochemistry, 1998, 47, 861-864. Barton, G.W. J. Nutr., 1989, 119, 109-111. Olson, J.A. J. Nutr., 1989,119, 94-95. Ziegler, R. J. Nutr. 1989,119, 116-122. Colburn N. In Carcinogenesis: A comprehensive survey; Colburn Tsl. Ed.; Raven Press: New York, 1980, 35-36. Schwartz, J.; Suda, D.; Light, G. Biochem. Biophys. Res. Commun., 1986, 136, 1130-1135. Jun-Kim, H. Int. J. Biochem,, 1993, 25, 911-915. Jyonouchi, H.; Hill, R.J.; Tomita, Y.; Good, R.A. Nutr. Cancer, 1991, 16, 93-105. Madan, C.L.; Karup, B.M.; Gupta, U.S. Econ. Bot., 1966, 20, 377-381 Nair S.C; Pannikar B.; Panikkar K.R. Cane. Lett., 1991, 57, 109-114. Nair, S.C; Salomi, M.J.; Varghese, CD.; Pannikar, B.; Pannikar, K.R. Biofactors, 1992, 4, 51-54. Morjani, H.; Tarantilis, P.; Polissiou, M.; Manfait, M. Anticancer Res., 1990, 10, 1398-1399. Tarantilis, P.A.; Polissiou, M.; Morjani, H.; Avot, A.; Bel Jebbar, A.; Manfait, M. Anticancer Res. 1992,12, 1889. Abdullaev, F.I.; Fenkel, G.D. Biofactors, 1992, 3, 201-204. Abdullaev, F.I.; Fenkel, G.D. Biofactors, 1992, 4, 43-45. Abdullaev, F.I. Toxicology Lett., 1994, 70, 243-251. Abdullaev, F.I. Biofactors, 1993, 4, 83-86. Salomi, M.J.; Nair, S.C; Panikkar, K.R. Nutr. Cancer, 1991,16, 67-72. Nair, S.C; Salomi, M.J.; Panikkar, B. J. Ethno-pharmacoL, 1991, 31, 75-83.

311

[63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [16] [11] [78] [79] [80] [81] [82] [83]

[84] [85] [86] [87]

Tseng, T.H.; Chu, C.Y.; Huang, J.M.; Shiow, S.J.; Wang, C.J. Cancer Lett., 1995, 97,61-67. Escribano, J.; Alonso, G.L; Coca-Prados, M. Cancer Lett., 1996,100, 23-30. Escribano, J.; Rios, I.; Fernandez, J.A. Biochim. Biophys. Acta, 1999, 1426, 2X1222. Tarantilis, P.A.; Morjani, H.; Polissiou M.; Manfait, M. Anticancer Res., 1994, 14, 1913-1918. Nair, S.C; Panikkar, K.R.; Parathod, R.K. Cancer Biother., 1993, 8, 339-343. Nair, S.C; Varghese CD.; Panikkar, K.R.; Kurumoor S.K. Int. J. Pharm., 1994, 32, 104-114. Lin, J.K.; Wang, C J. Carcinog., 1986, 7, 595-599. Wang, C.J.; Shiow, S.J.; Lin, J.K. Carcinog., 1991,12, 459-462. Wang, CJ.; Shiow, S.; Lin, J.K. Carcinog., 1991,12, 1807-1810. Gaudin, D.; Yieldin, K.L. Proc. Soc. Exp. Biol. Med, 1969,131, 1413-1416. Holloway, G.M.; Gainer, J.L. J. Appl. Physiol, 1988, 65, 683-686. DiLuccio, R.C; Gainer, J.L. Aviat. Space Envirom. Med., 1980, 57, 18-20. Seyde, W.C; McKeman, D.J.; Laueman, T.; Gainer, J.L.; Longnecker, D.E. J. Cereb. Blood Flow Metab., 1986, 6,103-107. Jorgensen, L.V.; Andersen, H.J.; Skibsted, L.H. Free Rad Res., 1997, 27,13-Sl. Chopra, R.N.; Badhwar, R.L. Indian J. Agri Sci., 1940,10, 40-42. Grisola, S. Lancet, 1997, 2, 41-42. Liakopoulou-Kyriakides, M.; Sinakos, Z.; Kyriakidis, D.A. Plant Science, 1985, 40, 117-120. Liakopoulou-Kyriakides M.; Skoubas, A.L Biochem. Intern., 1990, 22, 103-110. Nishio, T.; Okugawa, H.; Kato, A.; Hashimoto, Y.; Matsumoto, K.; Fujioka, A. Shoyaku gaku Zasshi, 1985,41,271-276. Nair, S.C; Kurumboor, S.K.; Hasegawa, J.H. Cancer Bioth., 1995, 10, 257-264. Birt, D.F.; Pelling, J.C; Nair, S.C; Lepley, D. In Genetics and Cancer Susceptibility: Implications for Risk Assessement, Waler, C; Groopman, J.; Slaga, T.J.; Klein-Szanto, A. Eds; Wiley-Liss, New York, 1996. Stavric, B. Clin. Biochem., 1994, 27, 319-332. Sujiura, M.; Shoyama, Y.; Saito, H.; Abe, K. J. Pharmacol Exper. Ther., 1994, 271, 703-707. Zhang, Y.X.; Shoyama, Y.; Sugiura, M.; Saito, H. Biol Pharmaceut. Bull, 1994, 77,217-221. Sugiura, M.; Shoyama, Y.; Saito, H.; Nishiyama, N. Proceedings of the Japan academy series B'Physical and biological sciences, 1995, 77, 319-324.

312 [88]

[89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112]

Konoshima, T.; Takasaki, M.; Tokuda, H.; Morimoto, S.; Tanaka, H.; Kawata, E.; Xuan, L.J.; Saito, H.; Sugiura, M.; Molnar, J.; Shoyama, Y. Phvtother. Res., 1998,72,400-404. Tsatsaroni, E.; Elefteriadis I., Journal ofSoc. Dyers Colourists, 1994, 110, 313315. Tsatsaroni, E.; Liakopoulou-Kyriakides, M.; Elefteriadis, I. Dyes and Pigments, 1998, J7, 307-315. Liakopoulou-Kyriakides, M.; Tsatsaroni, E.; Laderos, P.; Georgiadou, K. Dyes and Pigments, 1998, 36, 215-221. Takaora, A.; Kumico. M.; Mitsuto, K. Hyogo Kyoiku Daigaku, 1991, 11, 157166. Mizumaki, Katsumi, JPN kokoi Tokyo Koho., Chem. Abstr., 1990, 113, 217788280-82 Hua, W.Y.; Park, J.Y. U.S. Patent, 1992, 5.108.749. Knewstubb, C.J.; Henry, B.S. Food Technol Int., 1988, 179-186. Timperlake, C.F.; Henry, B.S. Endeavour, New Series, 1986,10, 31-36. Tsimidou, M.; Tsatsaroni, E. J. FoodSci., 1993, 58, 1073-1075. Tsimidou, M.; Biladeris, C. J. Agric. Food Chem., 1997, 45, 2890-2898. Raina, B.L.; Agarwal, S.G.; Bhafia, A.K.; Gaur, G.S. J. Sci. Food Agric, 1996, 71, 27-32. Narasimhan, S.; Chand, N.; Rajalakshmi, D. J. FoodQual, 1992, 75, 303-314. Basker, D.; Negbi, M. Econ. Bot., 1983, 37, 228-236. Kuhn, R.; Winterstein, A. Ber. Dtsch. Chem. Ges., 1934, 67, 344-357. Porter, J.W.; Lincoln, R.E. Arch. Biochem., 1950, 27, 3901-3903. Armstrong, G,A.; Hearst, J.E. FASEB, 1996, 2, 228-237. George, P.S.; Visvanath, S.; Ravishankar, G.A.; Venkataraman, L.V. Food BiotechnoL, 1992, 6, 217-223. Visvanath, S.; Ravishankar, G.A.; Venka Taraman, L.V. BiotechnoL Appl. Biochem., 1990, 72, 336-340. Cormiere, F.; Dutresne, C; Dorion, S. BiotechnoL Tech., 1995, 9, 553-556. Sarma, K.S.; Maesato, K.; Hara, T.; Sonoda, Y. J. Exp. Bot., 1990, 41, 745. Fakhrai, F.; Evans, P.K. J. Exp. Bot., 1990, 41, 47-52. Plessner, O.; Ziv, M.; Negbi, M. Plant Cell Tissue Organ Culture, 1990, 20, 8994. Himeno, H.; Sano, K. Agric. BioL Chem., 1987, 51, 2395-2400. Mikami, Y.; Fukunaga, Y.; Arita, M.; Obi, Y.; Kisaki, T. Agric. BioL Chem., 1981,^5,791-793.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

313

CHEMICAL COMPONENTS OF FRAXINUS ORNUS BARK - STRUCTURE AND BIOLOGICAL ACTIVITY IVANKA N. KOSTOVA AND TANYA lOSSIFOVA Institute of Organic Chemistry with Centre ofPhytochemistry, Bulgarian Academy of Sciences, Bg-1113 Sofia, Bulgaria ABSTRACT: This review describes the investigations carried out by the authors and coworkers on the chemical composition and biological activity of Fraxinus ornus. Its stem bark has been used in the traditional medicine for treatment of wounds, inflammation, arthritis and dysentery. Our results support the claims of the folk medicine by presenting scientific proofs for the antimicrobial, antiinflammatory, immuno-modulatory, skinregenerating, antioxidant, photo-dynamic damage prevention and antiviral properties of the bark extract and its components. The ethanolic extract of the stem bark and its main constituent esculin were found practically non-toxic. They inhibited the classical and alternative pathways of complement activation. The extract and esculin displayed antiinflammatory activity in both zymosan- and carrageenan-induced paw edema in mice. The extract exhibited a pronounced antioxidative activity and caused intense wound epithelization. The antimicrobial and photodynamic damage prevention properties of the extract and its fractions were dependable on their hydroxycoumarin composition. The isolated secondary metabolites belong to the groups of hydroxycoumarins, secoiridoids, phenylethanoids and lignans. Most of the coumarins and secoiridoids possess the ability to affect the complement activity. A clear correlation between the structure and the biological activity (antimicrobial, antioxidative and photodynamic damage prevention) of the studied hydroxycoumarins was observed. Isolation of new biologically active compounds and finding of new biological properties of already known bioactive substances has been achieved.

INTRODUCTION Throughout the ages plants have been the source of medicinal agents. Today a revival of thousand-years-old herbal remedies and a return to the ancient form of medicine is observed. Herbal remedies are common in Asia and Europe, particularly in Germany, France and Italy. More and more Americans are supplementing and replacing prescription medicines with various medicinal herbs. However, in many cases the claims of the folk medicine are still to be scientifically proved. Some claims may not be accurate. The herbal medicines should also meet the contemporary requirements for safety and effectiveness. The expansion of the market for herbs demands strict standards for ingredients and manufacturing. The standardization of the herbal preparations requires a detailed study of their chemical composition and finding of the active components. Herb remedies that

314 enjoy the greatest popularity are generally those that have been the most thoroughly investigated. Fraxinus ornus (L.) is a small tree belonging to the Oleaceae family widely found in Bulgaria. Its stem bark is used in the Bulgarian folk medicine for treatment of infected wounds, inflammation, arthritis and dysentery [1,2]. However, the claims of the traditional medicine were not scientifically confirmed, the toxicity of the extract was not examined and the chemical composition of the bark was not completely investigated. In order to explain and confirm the biological activities claimed by the traditional medicine, and to search for new biologically active compounds we studied the antimicrobial, antioxidative, immunomodulatory, antiinflammatory, skin-regenerating and antiviral properties of Fraxinus ornus bark extract and its components. In a parallel detailed phytochemical investigation of the extract we isolated and determined the structures of many hydroxycoumarins, secoiridoid glucosides, caffeoyl esters of phenylethanoid glycosides, lignans and other phenolic compounds. In this review we present the results from our phytochemical and biological investigations. PREVIOUS INVESTIGATIONS Previous investigations on bark, flowers and leaves of Fraxinus ornus have shown the presence of the hydroxycoumarins esculin (1), esculetin (2), fraxin (3 ), fraxetin (4), cichoriin (5) and of some phenolic acids [36]. The isolation of the flavonoids rutin, quercetin, quercetin-3-glucoside, quercetin-3-galactoside, quercetin-3,7-digalactoside and rhamnetin from leaves and flowers has also been reported [7-9]. GENERAL APPROACH The bark of Fraxinus ornus was subjected to systematic phytochemical investigation. The isolated compounds and some bark extracts were fiirther investigated for their biological activities. In some cases synthetic derivatives of the natural compounds were prepared and their biological properties studied.

315

CHEMICAL STUDIES Extraction Extraction of the dried and well ground plant material was carried out with hot ethanol. The extract was concentrated to a small volume and the deposited solid (a mixture of the coumarin glucosides 1 and 3) filtered. The mother liquor was concentrated to obtain the total ethanol extract 1 (TEl). In some cases the ethanolic extract was concentrated without removal of the deposited solid to give the total ethanol extract 2 (TE2). TEl and TE2 were further used in the biological studies and for isolation of pure components. Hydroxycoumarins A characteristic feature of Fraxinus species is the presence of simple hydroxycoumarins having OH or/and OMe groups only in the benzene ring. In Fraxinus ornus we found derivatives of 6,7-dihydroxy-, 6,7,8trihydroxy- and 5,6,7-trihydroxycoumarins. Isolation of Hydroxycoumarins

For isolation of hydroxycoumarins the TEl was subjected to liquid vacuum chromatography (LVC) with PE, CHCI3, EtOAc and MeOH to yield the corresponding fractions [10]. No coumarins were found in the PE fraction. The CHCI3 fraction was chromatographed over a silica gel column with a dichloroethane (DCE) - MeOH gradient. A TLC study of the DCE fractions on silica gel yielded the coumarins esculetin (2), fraxetin (4), scoparone (6), isoscopoletin (7), scopoletin (8), fraxidin (9) and fraxinol (10). Identification of all coumarins was achieved by UV, IR, ^H NMR and mass spectra, and direct comparison with authentic samples. NOE experiments confirmed the structures of 9 and 10. RP' HPLC Analysis of Hydroxycoumarins in Fraxinus ornus Bark Extract

The hydroxycoumarin content of plants varies significantly depending on the stage of the plant growth and development, and on the climatic conditions. Investigations of these variations are of scientific and practical interest. High pressure liquid chromatography (HPLC) is the most promising method for separation and detection of coumarins because it allows high resolution and a rapid and reproducible determination even of

316 trace compounds [11-13]. However, the very complex hydroxycoumarin composition of many plants requires further improvement and development of the HPLC procedure.

R'

R^

R^

R'

1

H

OGlc

OH

H

la

H

OGlc 4Ac

OAc

H

2

H

OH

OH

H

2a

H

OAc

OAc

H

3

H

OMe

OH

OGlc

4

H

OMe

OH

OH

4a

H

OMe

OAc

OAc

5

H

OH

OGlc

H

6

H

OMe

OMe

H

7

H

OH

OMe

H

8

H

OMe

OH

H

9

H

OMe

OMe

OH

OMe

OH

OMe

H

10 11

H

OGIc

OMe

H

12

H

OMe

OMe

OMe

13

H

H

H

H

Glc == glucose Ac = acetate

In this connection we attempted a reverse phase - high pressure liquid chromatography (RP-HPLC) determination of 11 naturally occurring hydroxycoumarins 1-8 and 11-13 [14]. Conditions were found for best resolution of the standard mixture - mobile phase H20-MeOH, detection at A.= 220nm, and appropriate gradient profile and flow rate. These conditions were used for the HPLC analysis of the total extract of Fraxinus ornus bark.

317 2

3

30

w. 40 Mm.

Fig. (1). HPLC profile of the ethanolic bark extract of F. ornus collected from region 1 (see Table 1): peaks are labelled with the corresponding compound numbers.

The HPLC profile of the extract showed a good resolution of the main constituents. Esculin (1), esculetin (2), and fraxin (3) are the major components of this species, while the others are present in smaller amounts, Fig (1). Coumarin (13) was not detected in the extract. In addition 7-methylesculin (11) and 6,7,8-trimethoxycoumarin (12 ) were detected. This was the first report of the occurrence of 11 and 12 in the Oleaceae family. In Bulgaria Fraxinus ornus bark is a major source for the industrial preparation of esculin (1), an antiinflammatory and vitamin-P-like agent. The selection of appropriate plant material, i.e. of higher esculin (1) and of lower esculetin (2), fraxin (3) and fraxetin (4) content is economically important. For this reason a quantitative determination of 1-4 in commercial samples of Fraxinus ornus bark from five different regions in Bulgaria was carried out and the results presented in Table 1. The table shows that the samples from regions 1-4 are characterized by higher esculin (more than 8%) and total coumarin content (more than 9%). These values are also higher than those reported for the Chinese species F. chinensis, F. bungeana and F. stylosa [12]. Region 5 exhibited lower

318 esculin (6.3%) and higher esculetin, fraxin and fraxetin content (total coumarins 7.8%). Table 1. Concentrations of Hydroxycoumarins 1-4 in Samples of F. ornus Bark from Different Regions of Bulgaria

Regions 1 2 3 4

1

5

Esculin (I) 8.06 8.48 8.83 8.53 6.27

Esculetin (2) 0.25 0.26 0.29 0.29 0.50

Concentration % w/w) Fraxin Fraxetin (4) (3) 0.79 0.05 1.02 0.04 0.05 1.03 0.94 0.05 0.07 1.25

1

Total

1

9.15 9.80 10.20 9.81

7.79

1

The proposed method is applicable to the analytical control of 1-4 for scientific and industrial purposes. This method has been applied for determination of the hydroxycoumarin composition of all extracts in our further biological studies. It could be successfully used for phytochemical investigation of many hydroxycoumarin-bearing plants belonging to different families. Secoiridoids and Phenylethanoids Isolation of Secoiridoids and Phenylethanoids

The general scheme which we have employed for isolation of pure secoiridoids and phenylethanoids is schematically described below. Solvent-solvent partitioning of the TEl with PE and EtOAc yielded the corresponding PE, EtOAc and MeOH-H20 extracts. The EtOAc extract was further subjected to LVC with DCE-MeOH gradient yielding residues Rl and R2 (DCE-MeOH, 10:1); R3, R4 and R5 (DCE-MeOH, 5:1), and R6 and R7 (DCE-MeOH, 3:1). From Rl ligstroside (14), omoside (15) and caffeic acid (16) were isolated by LVC and TLC. R3 was subjected to CC over silica gel and HPLC to give hydroxyomoside (17). R4 after repeated CC and RP-HPLC afforded secoiridoids 14, 17-20a,b and the lignan 21. RP-HPLC of R5 gave the coumarin glucosides esculin (1) and fraxin (3), and the caffeoyl esters of phenylethanoid glucosides 22 and 23. RP-HPLC of R6 and R7 yielded the caffeoyl esters of phenylethanoid glycosides 24- 27.

319 Secoiridoid Glucosides

Secoiridoid glucosides of oleoside (28) type frequently occur in Oleaceae family. Usually, in Fraxinus species they are present as mono- or diesters of p-hydroxyphenylethanol (29) derivatives. Our phytochemical studies on Fraxinus ornus resulted in the isolation of ligstroside (14), oleuropein (18), framoside (19) and of the new compounds ornoside (15), hydroxyornoside (17), hydroxyframoside A (20a) and hydroxyframoside B (20b) [15-17]. The FAB spectrum of ornoside (15) exhibited [M+H]^ at m/z 629. Its ^H and ^"^C NMR spectra (Table 2) revealed the typical signals of an oleoside (28) nucleus and suggested the presence of one 1,4-disubstituted benzene ring, one 1,2,4-trisubstituted benzene ring and two sets of OCH2CH2Ph moieties. The ^H NMR spectrum of its acetate 15a exhibited signals for only four alcoholic and one phenolic acetyl group in the trisubstituted benzene ring, and suggested no free PhCH2 CH2 OH groups. Alkaline hydrolysis of ornoside (15) afforded glucose and the new phenolic compound, named omosol, whose structure was elucidated as 30 on the basis of its spectral data and chemical behavior. These findings implied that in ornoside, the ornosol (30) moiety is linked to the oleoside (28) through the nonphenolic CH2CH2OCO ester bonds at C-7 and C-11. The position of attachment was unambiguously established by detailed NOE experiments on ornoside. The most characteristic NOEs are presented in Fig. (2). On irradiation of CH2-2", an enhancement of the H-3 singlet, the doublet for H-4" and H-8" and CH21" was observed. Irradiation of CHa- T" leads to enhancement of CH2-2*" and H-4'". This confirms the assignment of CH2-I", 2", 1'" and 2'" and proves, that the disubstituted (tyrosol) moiety is linked to C-11. Therefore, the structure 15 was assigned to ornoside.

0-Glc

Fig. (2). Most important NOEs observed for 15

320

14

14a

CH2CH2—(

. 7 R^OOC

11 , C00R2

8

\ OGIcR^

>—OH

-OAc

CH2CH2-

R^

R^

R^

Me

OH

H

Me

OAc

8" T

CH2CH2-¥ T-OH \ / 5" 4' ^ _ ^ ^ C ' H / H 2

15

OH

r~8" CH2CH2—/

\—OAc

15a

OAc Cn2Cn2

CH2CH2—f

17

^

>—OH ^

CHjCHz-

/

\

^O—
OH

-2"

x:

y-CHCHz

-OAc

17a

O—/

OAc

V-CHCH2

=/ i,

>AC

OH

18

19

Me

-OH

CH2CH2"

CH2CH2—{

)—OH

HO—(f

OH

\—CH2CH2

OH

OH

20a

20b

CH2CHJ

-OH

CH2CH2—(

\—OH

Hu

\

CH2CH2—\

/ -OH

C H 2C H 2

OH

HO^ HO—/

H Me

/

V-CH2CH2

OH

H Me

OH OH

H H

Me

OH

OY

321

Table 2. ^H and ^^C NMR Data of Insularoside (15) and Hydroxyornoside (17)inCD30D(ppm) 17

15 1 Positions 1 3 4 5 6a 6b 7 8 9 10 11

r

2' 3' 4* 5' 6'a 6'b I'a I'b 2"a 2'b 3" 4" 5' 6" 7" 8" r"a r"b 2'" 3.,, 4'" 5"' 6"' 7'"

1 ^Z 1

(J in HZ) 5.87 brs 7.55 s

1

6H

!1 3.80 dd (10.9, 3.6) 2.17 dd (15.1, 10.9) 2.32 dd (15.1, 3.6)

172.6s 124.9d 129.9s 13.7q 167.9s 100.9d 74.2d 77.9d 71.4d 78.3d 62.7t

6.06 qd (7.2, 0.8)

1.62 dd (7.2, 1.5)

4.79 d (7.8) 3.28 dd (9.1, 7.8) 3.401 (9.1) 3.2-3.4 obscured by MeOH 3.2-3.4 obscured by MeOH 3.65 dd (11.8, 5.7) 3.87 dd (11.8, 2.0) 4.49 ddd (11.2, 8.5, 3.1) 4.53 ddd (11.2, 5.9,4.4) 2.91ddd (14.5, 5.9, 3.1) 3.02 ddd (14.5, 8.5,4.4)

65.9t 35.8t

-

7.21 d (8.6) 6.94 d (8.6)

-

6.94 d (8.6) 7.21 d (8.6) 4.03 ddd (10.8, 6.5,4.3) 4.27 ddd (10.8, 6.0, 4.0) 2.761 (5.3) 2H

1

-

6.53 d (2.0)

6.85 d (8.1) 6.77 dd (8.1,2.0)

1 |

135.4s 131.5d 120.9d 157.4s 120.9d 131.5d 66.3t 34.7 132.2s 120.0d 147.0s 147.6s 117.7d 125.1d

1

-

-

1

1

|6H(/inHz) 5.85 brs 7.55 s

5c 95.1d 155.3d 109.7s 31.3d 40.8t

3.80 dd (10.9, 3.9) 2.13 dd (15.1, 10.9) 2.23 dd (15.1,3.9)

172.6s 125.2d 129.9s 13.6q 167.6s 100.9d 74.7d 77.9d 71.4d 78.3d 62.7t

6.05 brq(7.0)

1.61 d (7.00)

4.78 d (7.8) 3.29 dd (9.1, 7.8) 3.421 (9.1) 3.2-3.4 obscured by MeOH 3.2-3.4 obscured by MeOH 3.66 dd (11.9, 5.7) 3.88 dd (11.9, 1.5) 4.28 dd (10.8, 7.8) 4.56 dd (10.8, 3.8) 4.90 dd (7.8, 3.8)

71.6d

1

138.3s I29.2d 120.5d 158.5s 120.5d 129.2d 66.3t

7.36 d (8.7) 6.97 d (8.7)

-

'

1

6.86 d (8.2) 1 7.79 dd (8.2, 1.8)

1

68.7t

-

6.58 d (1.8)

1

95.2d 155.6d 109.5s 31.3d 40.8t

-

-

j 6.97 d (8.7) j 7.36 d (8.7) 4.04 ddd (10.7, 6.5,4.0) 14.27 m 2.77 m2H

Sc

|

347t 132.3s 120.3d 146.7s 147.7s 117.8d 125.0d

1

1

Our paper on isolation and structure elucidation of omoside was already in press, when the publications of Tanahashi et al. [18] and Shen et al. [19] appeared in Phytochemistry. Tanahashi and co-authors were the first to report the isolation and structure determination of a novel secoiridoid glucoside from F. insularis, named insularoside, and to assign structure 15 to this compound. Later, Shen et al. described the occurrence of insularoside in Fraxinus uhdei under the name uhdoside [19,20].

322

Therefore, almost simultaneously and independently the secoiridoid glucoside 15 has been isolated from three different Fraxinus species under three different names. Our approach in its structure elucidation provided some additional data and more chemical evidence about this unusual secoiridoid. Insularoside (uhdoside, ornoside) belongs to the group of the rare secoiridoid glucosides in which a flexible macrocyclic ring is formed between one phenolic compound and the oleoside aglucone. Most probably, in the preferred conformation protons H-6a and H-6b are influenced by the ring current of the aromatic rings, which would explain the observed highfield shift of the corresponding protons in 15 versus 14 without a significant change of the coupling constants. This fact and the observed NOE enhancement of H-5 and H-1 on irradiation of H-6a and H6b, respectively, indicate that the configuration at C-5 and C-1 in insularoside is the same as that described for oleoside derivatives, Fig. (2) [21]. A molecular formula of C32H36O14 was established for hydroxyornoside (17) by its ^H and ^^CNMR and FAB spectra ([M+Na]"^ at m/z 667). Its ^H NMR spectrum clearly indicated that it is an analog of insularoside (15). However, instead of the two protons at 2" position as in 15, only one proton was observed at 6 4.80. The ^"^C NMR spectrum (Table 2) of 17 was in full agreement with that of 15 for all carbons except for C- 1", 2", 4" and 8". The most striking difference was observed for C-2" - a doublet at 8 71.5 for 17 instead of the triplet at 6 35.8 observed for insularoside. Upon acetylation, hydroxyornoside provided a hexaacetate (17a) which showed six acetyl singlets in its ^H NMR spectrum instead of five for insularoside. It was suggested that 17 is a hydroxyderivative of 15 with an additional OH group located at position 2". Extensive NOE studies on 17a revealed the close relationship in space between the acetate group at C-2" and H-3, H-l", H-2", H-4", and H.8", Fig. (3).

H „f^. AcO

Fig. (3). Important NOEs observed for 17a

323

On alkaline hydrolysis, 17 afforded a new phenolic compound whose structure was unambiguously shown to be 2"- hydroxyomosol (31) by its spectral data. Acid hydrolysis of hydroxyomoside (17) using 16.0% methanolic H2SO4 resulted in a complex reaction mixture from which 32a,b could be isolated as a minor product and 33a,b as a major one. Under the same reaction conditions insularoside (15) afforded 32a,b and 34a,b, while ligstroside produced only 32a,b.

(CH30)2HC

R^ 32a,b

CH3

33a,b

CH3

CH3 HOCH2CH2-

I V

CHCHo

in HOCH2CH2-

34a,b

CH3

36a,b

CH3

OH 2" / V •CH2CH2

ACOCH2CH2-

•OAc

/"V

2"

CH2CH2

RtlOC ^ C00R2

OH

35a

R' =

CHjCH

CH3

35b

R' =

CHjCHj

CH3

r\.

OH

324

The ^H NMR (Table 3) and MS data of the less polar compound supported the proposed structure 32a9b and were in agreement with those described for the natural compounds 35a and 35b [22]. However, no CHO signal was visible and the presence of two CH protons and four OMe signals was observed. This suggested that compound 32a,b is a derivative of 35a and 35b in which the CHO group at position 1 has been acetalized. To account for the presence of two diastereomers we accept that under the applied reaction conditions a rearrangement of the secoiridoid nucleus takes place via the mechanism already proposed by Gariboldi et al. for the formation of 35a and 35b from oleuropein (18 ) in plants [22]. Evidently, under the conditions of acid hydrolysis this rearrangement is followed by acetalization of the CHO group. The stereochemistry at C-8 and C-9 in 32a,b could not be assigned with the help of NMR experiments because of the overlapping of ^H NMR signals. Table 3. *H NMR Data for Compounds 32a,b, 33a,b, 34a,b and 36a,b in CDCI3 (ppm); Coupling Constants (Hz) Are Given in Parentheses 1 Positions H-1 d H-3s H-5m Ha-6 dd Hb-6 dd H-8m H-9m H-lOd CH2-l"td CH2-2" td H-4", H-8" d H-5", H-7" d CH2-r"t CH2-2"'t H-4"' d H-7'" d H-8"' dd COOCH3 OCH3

1 AcO a.b

32a,b 4.41 (3.4) 4.29 (7.8) 7.58 7.54 3.30-3.10 2H 2.83(15.3,3.5)^ 2.39(16.0,4.4)^ 2.23(15.5,11.0)* 2.57(16.0,8.0)'' 4.20-4.10 2H 2.00-1.90 2H 1.43(7.0) 1.41 (7.0)

3.71 3H, 3.69 3H 3.68 3H, 3.67 3H 3.36 3H, 3.34 3H 3.31 6H

-

33a,b 4.40-4.20 2H 7.56 7.52 3.30-3.10 2H 2.68(16.0,3.5)^ 2.27(16.0,4.0)" 2.17(16.0,11.0)* 2.58(16.0,8.0)" 4.40-4.20 2H 2.00-1.90 2H 1.41(7.00) 1.39(7.00) 4.40-4.20 4H 4.40-4.20 2H 7.30 (8.5) 4H 6.98 (8.5) 4H 3.78 (6.5) 4H 2.75 (6.5) 4H 6.79 (1.8) 2H 6.99 (8.0) 2H 6.92(8.0, 1.8) 2H 3.68 3H, 3.66 3H

36a,b 34a,b "1:41 (3.4^ 4.39 (3.4) 4.22 (7.0) 4.26 (7.7) 7.55 7.54 7.50 7.50 3.30-3.10 2H 3.30-3.10 2H 2.76(16.0,3.4)* 2.66(16.0,3.5)* 2.34(16.0,4.4)" 2.29(16.0,4.2)" 2.17(16.0,11.0)* 2.23(16.0, 11.0)* 2.60(16.0,7.8)" 2.57(16.0,7.9)" 4.40-4.20 2H 4.20-4.10 2H 2.00-1.90 2H 2.00-1.90 2H 1.42(7.8) 1.41 (7.0) 1.39(6.5) 1.39(7.0) 4.301 (6.0) 4H 4.32 (7.0, 2.0) 4H 2.921 (6.0) 4H 2.93 (7.0, 2.0) 4H 7.17 (8.5) 4H 7.18 (7.0) 4H 6.92 (8.5) 4H 6.92 (7.0) 4H 4.22 (7.0) 4H 3.76 (6.5) 4H 2.85 (7.0) 4H 2.71 (6.5) 4H 6.82 (1.8) 2H 6.75 (1.9) 2H 7.07 (8.0) 2H 6.97 (7.0) 2H 6.96(8.0, 1.8) 2H 6.89(7.9. 1.8) 2H 3.67 3H, 3.66 3H 3.67 3H, 3.66 3H

3.35 3H, 3.33 3H 3.31 3H, 3.29 3H

3.35 3H, 3.33 3H 3.29 3H, 3.30 3H

3.34 3H, 3.32 3H 3.30 3H, 3.29 3H 2.16 6H, 1.99 6H

Signals with identical indexes belong to the same isomer; the other signals are not assigned.

1

|

325 The H NMR spectrum (Table 3) of the more polar compound 33a,b showed the absence of the sugar unit and revealed the protons of hydroxyomosol moiety and the rearranged secoiridoid nucleus. The doubling of all signals in the spectra suggested the existence of two diastereomers in nearly 1:1 ratio. The presence of two COOMe signals in addition to the four OMe singlets from the acetalized CHO group indicated that after hydrolysis of one of the ester linkages in hydroxyomoside a transesterification with MeOH has taken place. A similar information regarding the rearranged secoiridoid nucleus was forthcoming and from the ^H NMR spectrum of 34a,b, where the protons of ornosol moiety were also clearly visible. Acetylation of 34a,b afforded the acetate 36a,b, whose ^H NMR spectrum showed one aliphatic and one aromatic OAc signal (Table 3). The NOE experiments performed on 36a,b resulted in an enhancement of H-3, H-8", H-4", and CU2-T upon irradiation of CH2- 1" which suggested that the p-substituted phenethoxy unit is attached to C-11 of the secoiridoid nucleus. Therefore, under the applied conditions the acid methanolysis of 15 and 17 leads to opening of the macrocyclic ring at C-7. This is in accordance with the results of Tanahashi and coworkers [18] for the mild alkaline hydrolysis of insularoside (15) and the previous observations of LaLonde [23] that saturated esters undergo easier solvolysis compared to a,p - unsaturated ones. The new secoiridoid glucosides hydroxyframoside A (20a) and hydroxyframoside B (20b) were isolated as an unresolved mixture 20a,b in a ratio 2:1 as suggested by the ^H and ^^C NMR spectra. The molecular formula C32H38O14 was established for 20a and 20b based on the NMR spectra (Table 4) and the negative ESIMS, where only one peak at m/z 645 was found for the [M-H]" ion in both compounds. The ^H NMR spectrum of 20a,b suggested that each of the isomeric compounds had one 4-hydroxyphenethoxy and one 3,4-dihydroxy-phenethoxy unit. The exact position of their attachment to the oleoside nucleus of 20a and 20b was deduced on the basis of HMBC, HH-LR-COSY and NOESY spectra. The heteronuclear long range correlations from the methylene protons at 6 4.10 and 6 4.21 of isomer 20a to the carbon signal at 5 172.45 (assigned to C-7), and from the methylene protons at 8 4.09 and 4.22 of isomer 20b to the carbon signal at 6 172.5 (C-7) placed the corresponding CH2 groups at position 1'" in 20a and 20b, respectively. In the COSY spectrum of 20a,b the CH2-I'" protons of 20a and 20b showed coupling cross peaks to the protons resonances of the methylene groups at 5 2.77 for 20a and 6 2.82 for 20b, assigned to the respective CH2-2'" in both compounds. The presence of the cross peaks 52.77(CH22'")/56.67(lH, d, y=2.09Hz) and 52.77(CH2-2'")/6.55(lH, dd, J= 8.02 and 2.09 Hz) for 20a and 62.82 (CH2-2"')/67.05 (2H, d,J= 8.55Hz) for 20b in the HH-LR-COSY spectrum of 20a,b indicated the linkage of one 3,4-

326 dihydroxybenzene ring to CH2-2'" in 20a and one 4-hydroxybenzene ring to the same position in 20b. These data unambiguously proved the substitution at C-7 in the two compounds: one 3,4-dihydroxyphenethoxy unit in 20a and one 4-hydroxyphenethoxy unit in 20b. Table 4. ^H and ^^C NMR Data of Hydroxyframoside A (20a) and Hydroxyframoside B (20b) in CD3OD (ppm) 20a 1 Positions 1 3 4 5 6a 6b 7 8 9 10 11

r

2' 3' 4' 5' 6'a 6'b l"a l"b 2" 3" 4" 5" 6" 7" 8"

r"a

1

1

l'"b 9'" 3.,, 4"' 5'"

6 H (7 in Hz) 5.90 br s 7.47 s

3.94 dd (9.40, 4.45) 2.39 dd (14.22, 9.40) 2.64 dd (14.22,4.45)

6.08 qd (7.10, 1.00)

1.66 dd (7.10, 1.48)

4.80 d (7.82) 3.2-3.4 obscured by MeOH 3.411 (8.83) 3.2-3.4 obscured by MeOH 3.2-3.4 obscured by MeOH 3.68 dd (11.87, 5.50) 3.89 dd (11.87, 1.70) 4.27 dt (10.80, 6.74) 4.31 dt (10.80, 6.74) 2.87 t 2H (6.74)

7.07 d (8.60) 6.71 d (8.60)

6.71 d(8.60) 7.07 d (8.60) 4.10 dt (10.71, 7.07) 4.21 dt (10.71, 7.07) 2.77 t2H (7.07)

20b 8C 94.51 154.39 108.84 31.04 40.47

172.45 124.14 130.02 12.88 167.47 100.19 74.06 77.71 70.77 77.25 62.03

65.67

34.57 129.72" 130.25 115.54 156.34 115.54 130.25 66.17

-

6H(/inHz) 5.90 brs 7.48 s

3.94 dd (9.40,4.61) 2.39 dd (14.22, 9.40) 2.65 dd (14.22,4.61)

6.07 qd (7.05, 1.00)

1.64 dd (7.05, 1.48)

4.80 d (7.82) 3.2-3.4 obscured by MeOH 3.411 (8.83) 3.2-3.4 obscured by MeOH 3.2-3.4 obscured by MeOH 3.68 dd (11.87, 5.50) 3.89 dd (11.87, 1.70) 4.27 dt (10.80, 6.74) 4.31 dt (10.80, 6.74) 2.81 t2H (6.71)

6.68 d (2.02)

6.69 d (8.02) 6.56 dd (8.02, 2.02) 4.09 dt (10.70, 7.17) 4.22 dt (10.70, 7.17) 2.82 t2H (7.17)

34.70 129.52'^ 7.05 d (8.55) 6.67 d (2.09) 116.36 6.72 d (8.55) 145.53 6'" i 144.23 7'" 1 6.70 d (8.02) 115.75 6.72 d (8.55) 120.62 7.05 d (8.55) 6.55 dd (8.02,2.09) 8"' *' ^ Values with the same superscript are interchangeable.

dC 94.45 154.36 108.86 31.07 40.47 172.50 124.14 130.02 12.86 167.47 100.15 74.06 77.74 71.80 77.25 62.06 65.63 34.46 129.76'' 116.29 145.53 144.22 115.66 120.58 66.17 34.82' 129.33" 130.29 115.60 156.34 115.60

1

1 1

1

130.29 1

327

Furthermore, the HMBC, COSY and HH-LR-COSY spectra of 20a,b gave evidence for the attachment of one 4-hydroxyphenethoxy unit in 20a and one 3,4-dihydroxyphenethoxy unit in 20b to C-11 of the oleoside moieties. The proposed arrangements of 20a and 20b were further confirmed by the following cross peaks in the NOESY spectrum of 20a,b - for 20a: H3/H-2" and H-27H-4",8"(5 7.07, 2H, d,J= 8.60 Hz), and for 20b: H-3/H2" and H-27H-4"(8 6.68, IH, dd, J =2,02 Hz), and H-2 7H-8" (6 6.56, dd, J = 8.02 and 2.02 Hz), Fig. (4). Therefore, 20a and 20b are hydroxyderivatives of framoside (19) isolated from the same extract.

H OGIc

20a 20b

OH H

H OH

Fig. (4). Important NOEs observed for 20a and 20b

Tanahashi and co-authors [18] postulated that insularoside (15) is biosynthesized by oxidative C-0 coupling of the two p-hydroxyphenethyl moieties from framoside (19) found at that time in Fraxinus formosana [24]. The presence of 15, hydroxyframoside A (20a) and framoside (19) in Fraxinus ornus is in support of their postulate. However, the cooccurrence of 15 and omosol (30) in Fraxinus ornus allows to speculate that 30 could be also involved in the biosynthesis of insularoside (15). Caffeic Acid Esters of Phenylethanoid Glycosides

It is known that the Oleaceae family is a rich source of phenylethanoid glycosides (PhGs). Many cinnamic esters of PhGs of diverse structures have been isolated from the genera Syringa, Forsythia, Ligistrum, Jasminum and Osmanthus [25]. However, only the occurrence of verbascoside (24) and calceolarioside A and B in Fraxinus species have been reported [19,26,27].

328 Our investigations revealed the presence of calceolarioside B (23) in Fraxinus ornus leaves [28] and of six caffeic acid esters of PhGs in the polar part of the EtOH extract of the bark [29]. Five of them were identified as the known 2-(4-hydroxyphenyl)-ethyl-(6-0-caffeoyl)-p-Dglucopyranoside (22), calceolarioside B (23), verbascoside (24), isoacteoside (25) and lugrandoside (26). The occurrence of 22 and 26 in Oleaceae has not been reported so far. 0R5

r.R2 0R1

R^ 22 H 23 H 24 H 25 H 26 H 27 H 27a Ac Caff=Caffeoyl Ester Unit

H OH OH OH OH OH OAc

H H Rha Rha H Caff Caff 2 Ac

H H Caff H Caff H Ac

Caff Caff H Caff Glc Glc Glc 4Ac

The structure of the novel PhG-ester, named isolugrandoside, was deduced as 27 by concerted application of ID and 2D NMR methods and MS studies. A molecular formula of C29H36O16 was established on the basis of its negative HRFAB-MS ([M-H]" at m/z 639.1950) and the ^H and ^^C NMR spectra (Table 5). The ^H NMR spectrum revealed the signals typical of a /ra«5-caffeoyl ester unit, a 3,4-dihydroxyphenethyl unit and two P-glucose units. The detailed analysis of its ^H NMR, COSY, ID TOCSY, ^^C NMR, GHSQC and HMBC spectra fixed the position of the ester and glucosidic linkages to one central Pglucopyranosyl moiety (Glc-1). The HMBC correlation from H-1 to C-8" gave evidence for the attachment of the 3,4-dihydroxyphenethoxy moiety to C-1 of the central glucose unit, as usual in the PhGs [25]. The position of linkage of the terminal glucopyranosyl moiety (Glc-2) at C-6 of the central glucose unit was deduced from the ^*^C NMR spectrum where the resonances of C-6 and C-5 were observed at 5 69.9 and 76.9, respectively [25]. The HMBC correlation from the anomeric proton T" to C-6 indicated that the interglucosidic linkage is between C-T" of Glc-2 and C-6 of Glc-1. The position of the rraw^-caffeoyl unit at C-3 of the central glucose unit was confirmed by the HMBC correlation from H-3 to C-9'.

329 Table 5. ^H and ^^C NMR Data of Isolugrandoside (27) and Isolugrandoside Acetate (27a) 27" 1 Moieties

[ Positions

1 1 GIc-1

2 3 4 5 6a 6b

r

Caffeoyl

1 Phenethyl 1 Alcohol

2' 3' 4' 5' 6' ?• 8' 9' 1" 2" 3" 4" 5" 6" 7" 8"a 8"b j,„ 9'"

Glc-2

3'" 4'" 5'" 6'"a 6'"b AcO-Ph

1

6H(yinHz) 4.51 d (7.8) 3.44 t-like (9.0) 5.061 (9.0) 3.64 m 3.62 m 3.86 dd (11.3, 4.6) 4.18brd(11.3)

[

1

'

-

-

128.3 115.9 146.8 149.6 117.9 123.7 147.7 115.8 170.0 132.2 117.3 144.6 146.1 117.2 122.2 36.6 72.9

7.12 d (1.9)

6.85 d (8.2) 7.02 dd (8.2,1.9) 7.61 d (15.9) 6.37 d (15.9)

6.76 (2.0)

6.75 d (8.1) 6.64 dd (8.1, 2.0) 2.82 t, (7.1) 2H 3.79 dt (9.7, 7.1) 4.05 dt (9.7, 7.1) 4.43 d (7.8) 3.28 t-like (9.0) 3.42 t (9.0) 3.35m 3.30m 3.70 dd (11.6, 5.2) 3.89 brd (11.9, 1.9)

-

1 1

bC 104.5 73.6 79.7 69.8 76.9 69.9

7.22 d (8.5) 7.38 d (8.5, 2.0) 7.58 d (16.0) 6.28 d (16.0)

7.08 br s

j

104.9 75.2 78.1 71.7 78.0 62.8

AcO-Glc 1 ^ Measured in CD3OD (ppm). ^ Measured in CDCI3 (ppm).

1

-

1

7.34 d (2.0)

-

1

^^

6H(7inHz) 4.50 d (7.9) 5.08 t (9.0) 5.301 (9.0) 4.97 t (9.0) 3.72 m 3.64 m 3.89 brd (10.9)

1

1

1 1 1 1 1 1

7.06 d (8.0) 7.06 br d (8.0) 2.90 m 3.65 m 1 4.12 m 4.58 d (8.14) 5.00 m 5.211 (9.0) 5.08 m 3.67 m 4.14 brd (12.6) 4.28 dd (12.6,4.8) 2.32,2.31,2.29,2.28 2.10, 2.04, 2.02, 2.00, 1.98,1.91 1

Acid hydrolysis of 27 produced caffeic acid, 3,4-dihydroxyphenylethyl alcohol and D-glucose. The identification of D-glucose including its absolute configuration was conducted according to the procedure of Oshima et al. [30]. The proposed structure of 27 was in full agreement with the ^H NMR spectrum of its diacetate 27a. Our experimental data excluded the isomerization of 26 to 27 during the isolation and purification procedure and confirmed the natural occurrence of isolugrandoside.

330

Minor Phenolic Compounds The isolation of the minor components caffeic acid (16) and the lignan (+)-l-hydroxypinoresinol-4'-P-D-gIucoside (21) is described in the section "Isolation of secoiridoids and phenylethanoids". GC-MS analysis of the ethanolic extract revealed the presence of tyrosol (29) and of the new phenylethanoid ornosol (30) in trace amounts.

HO^

in

HO-"^ ^ ^ ^ ^ ^ ^ C O O H

16

_f

\—CH2CH2OH

29

ROCH2CH2—{

\-0—/

y-CH2CH20R

30 R = H 30aR = Ac

BIOLOGICAL STUDIES Antimicrobial Properties of Hydroxycoumarins, Phenylethanoids and Fraxinus ornus Bark Extracts

Secoiridoids,

The weak bacteriostatic effect of ethanolic extracts of Fraxinus ornus bark against Staphylococcus aureus has been reported [31]. The antimicrobial properties of the natural dihydroxycoumarins esculin (1), esculetin (2) and daphnetin have been also investigated using different test systems [32]. However, there was no detailed investigation on the antimicrobial activity of Fraxinus ornus extract and the compounds responsible for have not been identified. We studied the antimicrobial properties [33] of three Fraxinus ornus bark extracts and their main components 1-4 against Staphylococcus aureus, Candida species, Escherichia coli and Pseudomonas aeruginosa using the procedure of Heiss [34]. Evaluation of microbial growth on the contact surface was performed by measuring the Contact Growth Index (CGI). The crude extract (extract A) of the bark was separated into

331

petroleum ether (extract B), ethyl acetate (extract C) and water-methanol (extract D) applying the solvent-solvent partition method. Extract B was not further investigated because of its low solubility and absence of coumarins 1-4. The content of coumarins 1-4 in extracts A, C and D determined by HPLC is presented in Table 6 [14]. Table 6.

Coumarin Composition of Extracts A, C and D

Extracts 1 1

Extract A Extract C Extract D

Escuiin (1) 36.50 32.17 43.50

Content of the main coumarins in % w/w Fraxin (3) Esculetin(2) 5.00 2.33 1.83 5.33 7.00 0.17

Fraxetin (4) 0.33 0.80

0.12

| |

1

The propylene glycol solutions of all tested compounds and extracts were not active against E, coli and P. aeruginosa. Against Candida species fraxin (3) at concentration of 0.62% and fraxetin (4) at concentration of 0.84% exhibited some activity (CGI = 3). Against S. aureus, a 2.80 % solution of escuiin (1) showed lower activity (CGI=4) than 0.62% solution of fraxin (3 ) (CGI =3), while a 0.72% solution of esculetin (2) and a 0.84%) solution of fraxetin (4) revealed full inhibition (CGI=0). Extract C exhibited the highest antimicrobial activity among the tested extracts: full inhibition (CGI=0) of S. aureus and Candida species at a concentration of 8.80% and weak inhibition (CGI=3) of Candida species at a concentration of 2.10%). The antibacterial activity of the 0.2% ethanolic solutions of the studied compounds (for fraxin 0.1%) and bark extracts on S. aureus, compared with that of famesol, is presented in Table 7. The data allow a better comparison of the antibacterial power of the products. It follows from the results that the antimicrobial properties of the studied pure compounds depend on their structure and substitution: a. fraxetin (4), a 6-OMe-7,8-dihydroxycoumarin is a stronger inhibitor of S, aureus than esculetin (2), a 6,7- dihydroxycoumarin; b. glucosidation decreases the inhibitory power of both types of dihydroxycoumarins; c. the coumarins fraxin (3) and fraxetin (4) inhibit Candida species, while escuiin (1) and esculetin (2) do not. The activity of extracts A, C and D (Table 7) indicates a clear correlation with their hydroxycoumarin composition. The extract C enriched with the coumarins esculetin (2), fraxin (3) and fraxetin (4) shows the highest antimicrobial activity of all studied extracts. The inhibitory power of extracts C and D is comparable with that of famesol. The HPLC profiles of the tested extracts indicated the presence of some other bark constituents, different from 2-4, that could also contribute to their activity [14].

332

Table 7.

Sample Esculin(l) Esculetin(2) 1 Fraxin (3) 1 Fraxetin (4) 1 Extract A Extract C Extract D 1 Farnesol

Contact Growth Index (CGI) of 5. aureus Concentration (w/v%) in 96% EtOH 0.2 0.2 0.1 0.2 0.2 0.2 0.2 0.2

10^ 4 1 3 0 3 0 0 0

Suspension for inoculation (CFUy ml) 10^ 4 2 4 0 4 0 0 0

10* 4 3 4 0 4 0 1

2

1

This gave us a reason to study and compare [35] the antimicrobial properties of a series of Fraxinus ornus bark constituents - the hydroxycoumarins 1-4, 6-7, 11-12, and acetates 2a and 4a; the secoiridoids 14, 14a, 15, 15a, and the tyrosol derivatives 29 and 30 applying the method described by Bankova et al. [36]. In this case the antibacterial properties of the studied compounds were also dependable on their structure and substitution (Table 8). In the group of coumarins, both glucosides esculin (1) and fraxin (3) showed negligible activity (Minimum inhibitory concentration (MIC) > lOOO^g/ml). This in contrast to the higher potency of the corresponding aglucones esculetin (2) and fraxetin (^4) (MIC = 500 and 125 |^g/ml, respectively) and indicates, that glucoside bonding of one of the phenolic hydroxyls leads to a significant weakening of the inhibitory properties. As already established by the method of Heiss, fraxetin (4), a 6-OMe,7,8dihydroxycoumarin, is a more potent inhibitor of the growth of S. aureus than esculetin (2), a 6,7-dihydroxycoumarin derivative. Compound 2 inhibits both the gram-positive and gram-negative organisms, while 4 is an inhibitor only of the gram-positive. This suggests that the position of the phenolic hydroxyls is essential for the activity. Isoscopoletin (7), a 7- methyl derivative of esculetin is less inhibitory than esculetin itself, while scoparone (6), a 6,7-dimethyl derivative of esculetin, is completely inactive. The 7-0-methylesculin (11) and 6,7,8trimethoxycoumarin (12) are totally deprived of activity. Evidently, a methylation of the phenolic OH decreases the activity. The MIC of the acetates esculetin 2 Ac (2a) and fraxetin 2 Ac (4a ) were found to be equal to those of the parent compounds 2 and 4. Most probably, acetylation of the phenolic OH does not alter the antibacterial activity of the hydroxycoumarins. Compounds 4 and 4a are the most potent inhibitors of S. aureus among the tested Fraxinus ornus components.

333

Tabie 8. Antibacterial Activity of Hydroxycoumarins, Secoiridoids and Tyrosol Derivatives

1 Compound 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Inhibitory zone (d jmrnl) S. aureus E. coli

1 Hydroxycoumarins 16 18 Esculin (1) Esculetin (2) 17 19 16 17 Esculetin 2Ac (2a) 17 20 Fraxin (3) 21 Fraxetin(4) 0 20 0 Fraxetin 2AC (4a) 0 0 Scoparone(6) 14 0 Isoscopoletin (7) 0 0 Methylesculin(ll) 0 0 6,7,8-Trimethoxy-coumarin (12) Secoiridoids 28 Ligstroside (14) 26 26 0 Ligstroside 5 Ac (14a) 24 Insularoside (15) 16 23 0 Insularoside 5 Ac (15a) Tyrosol derivatives 6 0 Tyrosol (29) 24 0 Omosol(30) ' MIC was determined only for compounds with inhibitory zone more than

MIC

1

S. aureus

£. coli

> 1000 500 500 >1000 125 125 -

> 1000 500 500 > 1000

500 500 500 1000

500

1 1 1 1 1 1 500

1 500 16mm.

1

-

1 1 1 I

The resistance of some plants to attacks of insects and microbes has been attributed [37] to the presence of the bitter secoiridoid ligstroside (14), which predominantly occurs in Fraxinus species. In this investigation ligstroside (14), its acetate (14a), insularoside (15) and the tyrosol derivative omosol (30) exhibited equal inhibition (MIC = 500 jag/ml) on the growth of S. aureus. The natural secoiridoids 14 and 15 inhibited both the gram-positive and the gram-negative microorganisms (MIC = 500 |Lig/ml), while their respective acetates 14a and 15b, and tyrosol derivatives 29 and 30 inhibited only the gram-positive. Our preliminary screening of the secoiridoids 14, 14a, 15, 15a and omosol 3 Ac (30a), using the method of Romans et al. [38], showed clearly visible zones of inhibition of the growth of Cladosporium cucumerinum and suggested them to be fungitoxic. The antibacterial activity of the secoiridoids 14,15, 17-19, 20a, b, 28a and 38, and the caffeoyl esters of phenylethanoid glycosides 22 -27 has also been tested using the direct bioautographic TLC assay as published by Hamburger and Cordell [39]. Bacillus subtilis spp. and Pseudomonas fluorescens were the representatives of the gram-positive and gramnegative bacteria, respectively. The minimum inhibition amount (MIA) was determined. Cefotaxime was used as a positive control.

334

Most of the tested secoiridoids displayed an inhibition of the growth of B. subtilis and P, fluorescens (Table 9). Insularoside (15) and hydroxyornoside (17) are more active against B. subtilis. The mixture of hydroxyframoside A, B (20a, b) showed the best activity against Pseudomonas fluorescens. Table 9. Minimum Inhibition Amount (MIA) of Secoiridoids Against B, subtilis and P. fluorescens Compound Ligstroside (14) Insularoside (15) Hydroxyornoside (17) 1 Oleuropein (18) Framoside (19) Hydroxyframoside A,B (20a,b) 7,11 -Dimethyloleoside (28a) 10-Hydroxyligstroside (38) 1 Cefotaxime ^NA- not active at a concentration of 50 |ig/spot

B. subtilis [|ag/spot] 0.5 0.2 0.2 1.0 1.0 2.5 NA 40 0.01

P. fluorescens [|ag/spot]

10

1

10 20 10 10 2.5 40 NA

0,01

1

The caffeoyl esters 22-27 showed no activity against Pseudomonas fluorescens at a concentration of 50 |uig/spot. The MIA of 22-27 against Bacillus subtilis is the following: 22- 20; 23 - 10; 24 - 2.5; 25 - 2.5; 26 20; 27 - 50 |ig/spot [40]. Skin - regenerating Properties of Esculin and Fraxinus ornus Bark Extract We investigated the skin regenerating properties of esculin (1) and Fraxinus ornus bark extract on male white Wistar rats having standard oval wounds [41]. The rats were divided into four groups of 6 animals and treated as follows: I. (control) group, destined for spontaneous recovery; II. (control) group, treated with propylene glycol; III. (test) group, treated with extract. A solution (18.2%) of the total ethanol extract of the bark (TE2) in propylene glycol was applied. RPHPLC analysis of the extract revealed the following hydroxycoumarin composition: esculin (27.1%), esculetin (0.1%), fraxin (0.3%), fraxetin (0.4%) and minor components scoporone (6), isoscopoletin (7), scopoletin (9), methylesculin (11) and 6,7,8- trimethoxycoumarin (12). The presence of coumarins 10 and 11, and of compounds 14 - 27 was confirmed by TLC and HPLC.

335

IV. (test) group, treated with a 3.45% solution of esculin in propylene glycol. The percent of epithelization with respect to the beginning of the experiment (zero day) was calculated (Table 10). Table 10. 1

Group

i II III

1

IV

Epithelization (%) of Wounds with Respect to the Zero Day

V

3^*^ day

•^th

Day

day

lO*" day

10.0 9.4 39.8 38.6

30.7 31.1 55.8 50.4

43.9 45.3 84.9 67.8

89.9 88.8 96.9 87.4

day

i^

11

963

1

96.5 100.0

98.9

1

The III group of animals exhibited a more intense epithelization of the wounds in comparison with the control groups at every stage of the investigation. A weaker regenerating effect was found in the IV group of animals treated with esculin. The application of propylene glycol alone ( II group) did not result in an epithelizing effect. On the 7^^ day the biopsy of the I and II (control) groups established mostly a massive leucocytic necrotic swellings on the wound surface. A less pronounced leucocytic necrotic elevation was established in the III group, where definite zones of the wounds were almost completely filled with granular tissue. The amount of the riper collagen fibrils exceeded the amount of young fibrils. The epithelial regenerate consisted of ten and more cells. The basal prismatic layer, which was rich in mitoses, was separated from the overlying layers consisting of larger cells with round light nuclei. The III and IV groups of animals exhibited similar developments. These investigations have shown that the alcoholic extract and esculin obtained from Fraxinus ornus bark exercise moderate skin regenerating effects, no toxicity or local irritation being observed. The experimental results are in line with the antimicrobial properties of the extract and its constituents, and with the use of the bark in the traditional medicine for wound treatment and against inflammation. In addition, we studied the acute toxicity of the total ethanolic extract TE2 and its main component esculin (1) and found that applied p.o. to white mice and white Wistar rats in doses from 50 to 8000 mg/kg they were practically non-toxic. No lethality was observed up to 21 day and with the highest doses used. No significant changes were found both in the behavior and the reflexes of the animals. No pathological deviations from the physiological values were found in all hematological and clinical -chemical indices studied.

336 Antioxidative Action of the Ethanolic Extract Hydroxycoumarins from Fraxinus ornus Bark

and

some

The need to prepare lipids and lipid-containing products that are stable with respect to the effect of atmospheric oxygen has increased the interest in finding suitable natural sources of harmless antioxidants. The high biological activity of extracts from such sources is often due to the presence of phenolic compounds, many of them exercising a pronounced antioxidative effect on lipids [42,43]. In the literature there was no information about the antioxidative action of Fraxinus ornus extract and the data on hydroxycoumarins were scarce [44]. We examined the antioxidative action of the total ethanolic extract (TE2) of Fraxinus ornus bark as well as of its main hydroxycoumarin components esculin (1), esculetin (2), fraxin (3) and fraxetin (4) [45]. Quantitative reverse phase HPLC analysis [14] of the extract used in this study established the following hydroxycoumarin content: esculin 53.7%, esculetin 0.5%, fraxin 8.7% and fraxetin 0.3%. The investigations were performed at 100^ with kinetically pure triacylglycerols of lard and sunflower oil (TGL and TGSO), which represent models of two types of natural lipid unsaturation. The stabilization factor F as a measure of the effectivity and the oxidation rate ratio (ORR) as a measure of the strength of the tested antioxidants were estimated. The parameters F and ORR for TGL and TGSO oxidation in the presence of 0.05 and 0.10% ethanolic extracts oi Fraxinus ornus bark are presented in Table 11. The data indicate that the ethanolic extract of Fraxinus ornus bark has a pronounced antioxidative activity during the oxidation of both lipid substrates. This activity is commensurate with the inhibiting effect of the same concentration of butylated hydroxytoluene (BHT) and butylated hydroxyanisole (BHA) during TGL oxidation [46]. In order to elucidate the contribution of the main phenolic components [esculin (1), esculetin (2), fraxin (3) and fraxetin (4)] in the extract to its stabilizing action, we studied the autoxidation kinetics of the two natural lipid systems (TGL and TGSO) in the presence of different concentrations of 1-4. For TGL autoxidation in the presence of 0.04% esculin and 0.04% fraxin the following data for F and ORR were obtained: esculin, F = 2.5, ORR = 0.7; fraxin, F = 2.1, ORR = 1.0. In TGL addition of 0.02% esculin and 0.02% fraxin resulted in practically no stabilizing action. In TGSO these glucosides exhibited no antioxidative action. The kinetic parameters characterizing the inhibitor action of esculetin and fraxetin are also presented in Table IL The experimental data demonstrate that the fraxetin possesses a higher effectivity and greater strength than does the esculetin. Both substances are less effective

337

inhibitors in the lipid system with a higher oxidizability (TGSO). Both hydroxycoumarins are antioxidants of a relatively high effectivity and of a great strength because, under the same oxidation conditions for 9.1x 10' "•M B H T in TGL F = 4.8, ORR = 0.33; for 4.6 x 10"* M a-tocopherol F = 6.5, ORR = 0.13; for 1.0 x 10"^M ferulic acid F = 3.1, ORR = 0.59; for 1.3 x lO'^M 3,4-dihydroxybenzoic acid F = 13.4, ORR = 0.07 and for 1.1 x lO'^M caffeic acid F = 62.1. For 1.0 x 10"^ M ferulic acid in TGSO F = 2.6, ORR = 0.60, ; for 1.3 x lO'^M 3,4-dihydroxybenzoic acid F = 3.6, ORR = 0.60 and for 1.1 x 10"^M caffeic acid F = 33.6, ORR = 0.07 [46,47]. Table 11. Stabilization Factor F and Oxidation Rate Ratio ORR for the Inhibited Oxidation of TGL and TGSO at 100°C in the Presence of Ethanolic Extract, Esculetin (2) and Fraxetin (4) from Fraxinus omus bark 1 Antioxidant 1 Extract

Esculetin (2)

Fraxetin (4)

Inhibitor concentration M % 0.05 0.10 0.56x10-^ 0.01 0.02 1.12x10-^ 0.05 2.81x10-' 0.10 5.56x10-' 0.48x10-' 0.01 0.02 0.95x10"' 0.05 2.38x10-' 0.10 4.76x10-'

TGSO

TGL F 4.8 6.1 22.7 42.7 64.5 73.1 38.2 86.3 208.0 340.0

ORR 0.28 0.28 0.07 0.06 0.05 0.05 0.05 0.03 0.02 0.01

F 3.6 4.0 14.8 20.8 37.6 41.2 13.2 27.2 72.0 125.0

1 ORR 1 0.60 0.50 0.15 0.09 0.06 0.05 0.15 0.09 0.04

0.02 1

It is worth mentioning that when ORR is larger than 1, the oxidation proceeds faster in the presence of an inhibitor than in its absence. The lower the ORR, the stronger the inhibitor. Comparison of the F and ORR values for the glucosides with the values of the same parameters for the aglucones at almost the same molar concentrations (0.02% aglucones) shows that the glucoside bonding of one of the phenolic OH groups leads to a significant weakening of the inhibiting properties. The results obtained indicate blocking of the more active phenol group in fraxin, due to which the difference in antioxidative activities of glucosides esculin and fraxin is, in contrast to aglucones esculetin and fraxetin, negligible. Considering the antioxidative effects of coumarins 1-4 in TGL and TGSO and the content of these compounds in the extract, and eliminating the possibility of antagonism or synergism between them, it was calculated that the four compounds under consideration determined twothirds of the antioxidative effectivity of the extract in TGL and half of the effectivity in TGSO.

338 TLC analysis of the antioxidatively acting compounds [48] in the ethanoHc extract from Fraxinus ornus bark revealed the presence of additional antioxidative acting compounds. Among them calceolarioside B (23) demonstrated a significant activity. Most probably, caffeic acid (16) and the caffeoyl esters of phenylethanoid glycosides 22, 24-27 also contribute to the antioxidative properties of the extract. Photodynamic Damage Prevention by Extracts Hydroxycoumarins from Fraxinus ornus Bark

and

Some

The antioxidants or quenchers of free radicals are knovm to minimize skin photoaging and the protective action of sun screens correlates closely to their free radical scavenging activity [49,50]. This prompted us to use the prevention of photodynamic yeast cell damage to comparatively investigate the protective activity of the four hydroxycoumarins esculin, esculetin, fraxin and fraxetin, and a widely used sun screen paminobenzoic acid (FABA). We applied the same test to characterize the protective activity of four Fraxinus ornus bark preparations containing these coumarins in different concentrations and to examine the activity of caffeic acid (16), a minor constituent of these preparations [51]. Table 12. Preparation 1 Total extract 1 Fraction A 1 Fraction B 1 Fraction C

Protective Effect of the Fraxinus ornus Preparations Protectio n factor in% 9.8 2.3 54.6 6.1

Hydroxycoumarin composition in % EsculinCl) 40.5 0.0 28.4 45.3

Esculetin (2) 1.0 0.0 6.3 0.3

Fraxin (3) 6.9 0.0 2.5 9.0

Fraxetin (4) 0.2 0.0 1.6

|

0.2

1

The total extract of Fraxinus ornus bark v^as prepared and further subjected to solvent-solvent partition to obtain the fractions A, B, and C w^ith the hydroxycoumarin composition given in Table 12. The test for protective effect evaluation was performed using yeast cells (strain Kluyveromyces fragilis 129-1) according to the procedure described by Lazarova and Ignatova [52] and the protection factor determined. All of the tested pure compounds showed protective activity (Table 13). The protection achieved was higher for esculetin (2) and fraxetin (4) as compared to the corresponding glucosides esculin (1) and fraxin (3). This finding is in accordance with the data obtained by Bakalova et al [53] concerning the effect of the same coumarins on the lipid peroxidation of liver microsomes, as well as with our own results on their antioxidative action presented above.

339 The protective effects of the four Fraxinus ornus preparations depend on their hydroxycoumarin composition (Table 12). Fraction B, which is enriched in aglucones, exerts the highest protective activity. Fraction C, with lower aglucone and higher glucoside concentration, demonstrated a protective activity lower than that of the total extract. The negligible protective activity of fraction A correlates very well with its composition, i.e. no measurable concentration of the coumarins under investigation. Protective Effect of the Pure Compounds

Table 13. 1

Compound Esculin (1) Esculetin (2) Fraxin (3) Fraxetin (4) PABA

Caffeic acid (16)

Concentration in mg/1

Protection factor in %

20 50 20 50 20 50 20 50 2.5 5 25 2.5 5 25

38.7 26.3 97.6 99.8 23.1 72.8 94.7 98.2 47.2 943 92.6 49 1 91.0

98,2

1

Our results suggest that the caffeic acid (16) and the aglucones esculetin (2) and fraxetin (4) approximate the skin protective effect exerted by the conventional sun screen PABA. The Fraxinus ornus bark preparations enriched in the compounds 2 and 4 seem to be effective protectors too. Complement Inhibition and Antiinflammatory Activity of Hydroxycoumarins, Secoiridoids and Extracts from Fraxinus ornus Bark The use of Fraxinus ornus bark in the Bulgarian folk medicine for treatment of inflammation, arthritis and dysentery [1,2] suggests the presence of some active principles with anti-inflammatory activity. Our phytochemical investigations have shown that the bark contains hydroxycoumarins, secoiridoid glucosides, caffeoyl esters of phenylethanoid glycosides and other phenolic compounds [10,14-17,28]. RP-HPLC analysis of commercial samples of Fraxinus ornus bark revealed high esculin (1) content (6-9%) [14].

340

Since the complement system is highly involved in an inflammatory response [54,55] many substances exhibiting anticomplementary activity have proved to be effective antiinflammatory agents [56]. In this connection we studied and compared the effects of the ethanolic extract of Fraxinus ornus bark and its main component esculin (1) on some in vitro and in vivo reactions related to acute inflammatory processes [57]. Quantitative RP-HPLC analysis of the total extract used in this study showed the following hydroxycoumarin composition: esculin 40.0%, esculetin 2.4%, fraxin 7.8% and fraxetin 0.4% [14]. The inhibitory effects of TE2 and 1 on the classical pathway (CP) and the alternative pathway (AP) of the complement activation in mouse serum were estimated at final concentrations varying from 1 to 50)ag, Fig. (5). The TE2 caused a more pronounced reduction of CP hemolysis compared to esculin. In the AP assay they exhibited nearly equal dosedependent inhibition of complement - mediated lysis.. inhibition (%)

Concentration (|ig)

Inhibition (%)

Concentration (|ig)

Fig. (5). Inhibition of CP (A) and AP (B) complement activity in mouse serum by different concentrations of TE2 (o) and esculin (•)

341 The comparison between the effects obtained with TE2 and 1 in the hemolytic inhibitory assay indicates that the anticomplementory action of TE2 is not due only to esculin. The effect of esculin on CP activity was less pronounced than that of the total extract, although it represents 40% of the content of TE2 and in these experiments it was tested in equal concentrations with TE2. The full inhibition of AP activity was achieved at concentration of 50)ig for both esculin and TE2. This also suggested that, excepting esculin, some other extract constituents contribute to its anticomplementary action. The esculin concentration causing 50% inhibition of complement activity in vitro is about 10"^ M which appears to be similar to that established for chemically similar compounds. Carrageenan- and zymosan- induced paw edema were chosen as suitable models for evaluation of the antiinflammatory activities of TE2 and esculin (Table 14). The results showed that both TE2 and esculin significantly reduced formation of the zymosan-induced paw edema in mice. In the case of carrageenan-induced edema, only TE2 at a dose of 15 mg/kg significantly reduced the inflammation, while esculin was ineffective. This indicated that the extract contains active components with different mode of action. Evidently, TE2 and esculin possess the ability to influence complement activity in vitro and to suppress some complement-mediated reactions after in vivo application. Coumarin derivatives are known to possess antiinflammatory and antimetastatic properties. The mode of action of coumarins is mainly attributed to their direct action on cells participating in the inflammatory process [58]. The influence of coumarins on the complement system, which is involved in the different stages of inflammatory response, has not been thoroughly investigated. Table 14. Effect of Esculin (1) and Total Extract from F. ornus Bark on Zymosan- and Carrageenan- Induced Paw edema in Mice Carrageenan- induced Zymosan-induced oedema oedema | Paw volume" Paw volume' 1 Control 42.6 ±5.5 89.0 ±9.5 15 30.3 ±5.4* 15.7 ±1.3* Total extract 5 28.6 ±10.0* 21.0 ±2.0 15 38.0 ±6.2* 36.0 ±2.6 Esculin (1) 5 38.2 ±2.0 1 40.6 ±3.5* ® Difference (mg) between the weight of a zymosan or carrageenan treated paw and the concentrated saline treated paw. Significant from respective control: * P< 0.05. Test material

Dose (mg/kg)

For this reason we examined the in vitro effect of 12 hydroxycoumarins on classical and alternative complement activity in

342

normal human serum and the consumption of the key components CI and C3 [59]. The coumarins 1-4, as well as their acetylated and methylated derivatives la, 2a, 4a, 6-8,11 and 12 have been investigated at different concentrations. The effect of the substances at concentrations of 1.2x10"'^ and 5.0 xlO'^^M, as the most representative, are shown in Table 15. All the substances tested had a moderate or weak ability to affect at least one of the complement pathways. The effect was not strictly dose-dependent. Esculin 5Ac (la), esculetin 2Ac (2a) and 7-methylesculin (11) exhibited good inhibition on CP activity. Scoparone (6) strongly reduced AP activity in normal human serum (NHS). Scopoletin (8), esculin (1) and esculetin (2) enhanced complement mediated hemolysis. Some of the compounds exhibited combined effect - activated one of the pathways and inhibited the other. Table 15. Inhibition (-) or Activation (+) of CP and AP Activity by Some Hydroxycoumarins

Hydroxycoumarins 1 1

1 1 1 1 1 1 1

Esculin (1) Esculin 5Ac (la) Esculetin (2) Esculetin 2Ac (2a) Fraxin (3) Fraxetin (4) Fraxetin 2Ac (4a) Scoparone (6) Isoscopoletin (7) Scopoletin (8) Methylesculin(ll) Trimethoxycoumarin (12) NA = not active

CP + 28.2 -21.5 + 7.1 + 8.2 + 28.2 -8.2 -8.2 + 5.9 + 8.2 + 22.4 -28.2 -16.5

Concentration 1.2xlO-^M Activity (%) AP NA -28.2 -5.9 -28.2 NA NA -20.6 -45.9 NA NA -34.1 -25.9

CP + 27.1 -40.0 + 24.3 -32.9 NA -17.5 -28.6 -11.4 + 24.3 + 10.0 -51.4 + 13.6

Concentration 5.0xlO"*M Activity (%)

|

AP

1

+ 10.7 + 17.9 + 15.7 -28.6 + 15.0 -15.0 -12.1 -57.1 NA + 20.0 -8.6

-28.6

1

Seven hydroxycoumarins were further tested at a single concentration (5.0x10"^ M) for their ability to influence CI and C3 functional activities after preincubation with undiluted NHS. 7-Methylesculin (11) had a good effect on reducing total, CI, and C3 hemolysis via both pathways. Scoparone (6) strongly inhibited C3 alternative activity but in the case of the classical pathway only the total hemolysis was diminished without influence on CI and C3. Esculin (1) slightly increased C3 classical activity but caused exhaustion of alternative C3 activity. In subsequent experiments 1 and 8 altered the effect of other complement activators (heat aggregated IgG, suramin and zymosan) when applied with them simultaneously in vitro.

343

The experimental data presented above suggest that hydroxycoumarins may counteract with some of the complement proteins and thus inhibit their functional activity. Also, it is possible that the substances form complexes with the serum proteins which are able to activate complement system. It is difficult to make conclusions about the relationship between the structure of the coumarins and their action on complement mediated reactions. It might be concluded, however, that the methylated hydroxycoumarins 6 and 11 are the most potent inhibitors of AP and CP activities and deserve attention as a possible antiinflammatory agents. A series of pure secoiridoid glucosides isolated from different Fraxinus species was compared in vitro for anticomplement action as well as for their ability to prevent cobra venom-induced complement activation in normal human serum [60]. Table 16 shows that most of the secoiridoids possess the ability to suppress CP and AP activities. The most effective inhibitors of CP in guinea-pig serum (GPS) were ligstroside (IC50 33 |Lig/ml) and insularoside (IC50 62 |iig/ml). With regard to NHS the most pronounced decrease of CP was caused by 7,11-dimethyloleoside (28a) at a concentration of 250 |ag/mL Altemative pathway hemolysis was slightly altered, although the substances were used in a higher concentration (Img/ml) than in the CP assay (250 |ag/ml). Evidently, secoiridoid glucosides exhibit a greater effect on CP activity. It makes them interesting for further investigation as there is a need for selective inhibitors of the complement system for possible therapeutic use. Table 16. Effect of Some Secoiridoids and Fraxinus ornus Bark Extracts on CP and AP Activity in GPS and NHS 1 1

Product

GPS ICP5o(Mg/ml)*

?3HS Inhibition (%)

CP**

AP'

1

Extract TEl 302 ±10 1 Extract TE2 1438 ±20 12.7 15.5 1 Ligstroside (14) 33 ±4 38.2 18.2 1 Insularoside (15) 62 ± 8 28.2 1 Hydroxyornoside (17) 5.5 185 ±10 23.6 18.2 1 Oleuropein (18) 130 ±8 29.1 1 Framoside (19) 4.5 160 ±6 96.4 7,11 -Dimethyloleoside (28a) 8.2 180±4 1 10-Hydroxyligstroside (38) 34.5 >1000 4.5 1 ' The concentration giving 50 % inhibition of CP (ICP50). ^ Measured at concentration of 250 ^g/ml. Measured at concentracion of l^ig/ml 1

1

In our further experiments the secoiridoids 15, 17-19 expressed the ability to prevent CP and AP activation, caused by cobra venom [61]. This

344

suggested that among the active secoiridoid constituents may exit one with C3- convertase inhibitory property. Our investigations revealed the contribution of secoiridoid glucosides to the antiinflammatory action of the extract. Antiviral Activity of Some Hydroxycoumarin Derivatives Studies on the antiviral effects of 6,7-dihydroxy- and 6,7,8trihydroxycoumarin derivatives were very limited [62,63]. This prompted us to study the antiviral properties of the structurally related compounds 1, 2, 4, 6, 7,11,12 and acetates la, 2a, 4a [64]. Primary screening for antiviral activity was carried out using the agardiffusion plaque-inhibition method with cylinders [65]. The compounds were tested against one representative in each of the four taxonomic viral groups; namely, picoma-, orthomyxo-, paramyxo- and herpes viruses, which represent a few of the most important families of the human patogens. The viruses used were poliovirus 1 (PVl), influenza virus A (FPV), Newcastle disease virus (NDV) and pseudorabies virus (PsRV). Two of the ten hydroxycoumarin derivatives tested showed activity against NDV, namely esculetin (2) and its diacetate (2a), when applied in cylinders at a dose of 2.8 and 1.9 mM/0.1 ml, respectively. Their activity was significant, although inferior when compared to that of ribavarin (used as a reference paramyxovirus inhibitor) at a dose of 2.0 mM /O.l ml. The remaining compounds tested were without effect on the replication of the four viruses studied. Evidently, no definitive conclusion could be drawn regarding structure-activity correlations. It could be noticed, however, that methylation and glucosidation of esculetin (2) lead to a loss of activity. SUMMARY AND CONCLUSIONS The work presented in the preceding pages described the antimicrobial, antioxidative, antiinflammatory, immunomodulatory, skin regenerating, photodynamic damage prevention and antiviral properties of Fraxinus ornus bark extract and its constituents. The total ethanolic extract of the stem bark and its main constituent esculin (1) were found practically non-toxic. They inhibited the classical pathway and alternative pathway of complement activation. The total extract and 1 displayed antiinflammatory activity in both zymosan- and carrageenan-induced paw edema in mice. The extract exhibited a pronounced antioxidative activity and caused intense wound epithelization. The antimicrobial and photodynamic damage prevention properties of the extract and its fractions were dependable on their

345

hydroxycoumarin composition. These results to some extent explained the traditional use of the bark for treatment of wounds, inflammation, dysentery and arthritis. The isolated secondary metabolites belong to the groups of hydroxycoumarins, secoiridoid glucosides, phenylethanoids and lignans. Of them 6 are new compounds. Isolation of new biologically active compounds and finding of new biological properties of already known bioactive substances has been achieved. Most of the secoiridoids and the coumarins possess the ability to affect at least one of the complement pathways. A clear correlation between the structure and the biological activity (antimicrobial, antioxidative and photodynamic damage prevention) of the studied hydroxycoumarins was observed. Our investigations provide more data on the chemical composition and biological properties of Fraxinus ornus and support the claims of the traditional medicine by presenting scientific proofs for the biological activity of the bark extract and its components. It was concluded that the total ethanolic extract of the bark can be included in the group of the practically non-toxic substances and may be used in the therapeutic practice for treating wounds, bums etc., as well as in the perfumery and cosmetics [66]. ABBREVIATIONS Ac AP BHT BHA Caff CGI CP DCE FPV Glc GPS HPLC IR LVC MIA MIC MS NA NAS NDV

Acetate (COCH3) Alternative Pathway Butylated Hydroxytoluene Butylated Hydroxyanisole Caffeoyl Ester Unit Contact Growth Index Classical Pathway Dichloroethane Influenza Virus A Glucose Guinea-pig Serum High Pressure Liquid Chromatography Infra-red Liquid Vacuum Chromatography Minimum Inhibition Amount Minimum Inhibitory Concentration Mass Spectrometry Not Active Normal Human Serum Newcastle Disease Virus

346 NMR ORR PABA PhG PsRV PVl Rha RP TEl TE2 TGL TGSO TLC UV

= = = = = = = = = = = = = =

Nuclear Magnetic Resonance Oxidation Rate Ratio p-Amino Benzoic Acid Phenylethanoid Glykoside Pseudorabies Virus Poliovirus 1 Rhamnose Reverse Phase Total Ethanol Extract 1 Total Ethanol Extract 2 TriacylglycerolsofLard Triacylglycerols of Sun Flower Oil Thin Layer Chromatography Ultra-violet

ACKNOWLEDGEMENTS The phytochemical investigations described in this review were carried out at the Institute of Organic Chemistry, Bulgarian Academy of Sciences and partially in a collaboration with Prof. W. Kraus, Dr. B. Vogler and Mrs. L Klaiber from the University of Hohenheim, Stuttgart. We are grateful for their participation. The work carried out at the Institute of Organic Chemistry was made possible thanks to the skillful efforts of Dr. B. Mikhova, Dr. G. Stoev, Dr. E. Vassileva, Mr. N. Nykolov, and for the antioxidative action thanks are due to Dr. N. Yanishlieva and Dr. E. Marinova. The biological investigations are a result of an effective collaboration with our following co-workers from the Institute of Microbiology, Bulgarian Academy of Sciences: Prof A. Galabov, Dr. N. Ivanovska, Dr. A. Kuyumgiev, Dr. G. Lazarova, Dr. H. Neychev, Dr. Z. Stefanova and Mrs. A. Ignatova. We gratefully acknowledge their valuable contribution. We also acknowledge the collaboration with Dr. L. Chipilska from the National Centre of Hygiene and Medical Ecology and Dr. E. Klouchek and Dr. A. Popov from the Higher Medical Institute. Dr. I. Kostova thanks Mrs. A. Ivanova for the technical assistance in the preparation of the manuscript.

REFERENCES [1] [2]

Stoyanov, N. In Our Medicinal Plants (Bulg.); Ularova, K., Ed.; Nauka i izkustvo: Sofia, 1973; Vol./, pp. 321-322. Asenov, I.; Nikolov, S. In Pharmacognosy (Bulg.); Popova, M., Ed.; Medizina i Phizkultura: Sofia, 1988, p. 41.

347

[3] [4]

Dezelic, M.; Gujic-Vasic; J.; Ind. Chim. Beige, 1967, 32, 310-311. Ivanov, V.; Yaneva, A.; Savchev, P.; Tr. Nauchnoizsled. Khim-Pharm. Inst., 1972, 8, 147-148. [5] Grujic-Vasic, J.; Bosnic, T.; Arh. Farm, 1981, 31, 273-278. [6] a. Artemeva, M.V.; Karryev, M.O.; Nikonov, G.K.; Rastit. Resur, 1975, 11, 368371. b. Steinegger, E.; Steiger, K. F.; Pharm, Acta Helv, 1959, 34, 334-344. [7] Paris, R.; Stamboli, A.; Am.Pharm. Franc, 1960, 18, ^13-^11. [8] Harbome, J.B.; Green, P.C.; Bot. 1 Linn. Soc, 1980, 81, 155-167. [9] Shammas, G.; Philianos, S.; Verykokidou-Vitsaropoulou, E.; Ann. Pharm. Fr, 1990,48, 13-16. [10] Kostova, I.; Planta med, 1992, 58,484. [11] Villeneuve, F.; Abravanel, G.; Moutounet, M.; Alibert, G.; J. Chromatogr, 1982, 234, 131-140. [12] Guo, X.-S.; Zhang, Y.-Z.; Yaoxue Xuebao, 1983, 525-528. [13] Tamma, H.; Raman, V.; Miller, G.C., Everett, R.; J. Chromatogr, 1985, 322, 236239. [14] Nykolov, N.; lossifova, T.; Vassileva, E.; Kostova, I.; Stoev, G.; Phytochem. Analysis, 1993,4, 86-88. [15] lossifova, T.; Mikhova, B.; Kostova, I.; Phytochemistry, 1993,34, 1373-1376. [16] lossifova, T.; Mikhova, B.; Kostova, I.; Monats.f. Chemie, 1995, 126, 1257-1264. [17] lossifova, T.; Vogler, B.; Kostova, I.; Phytochemistry, 1998,49, 1329-1332. [18] Tanahashi, T.; Shimada, A.; Nagakura, N.; Inoue, K.; Kuwajima, H.; Takaishi, K; Chen, C.C.; Phytochemistry, 1993, 33, 397-400. [19] Shen, Y.-C.; Chen, C.-H.; Lee, K.-H.; Phytochemistry, 1993, 33, 1531-1533. [20] Shen, Y.-C; Chen, C.-H.; J. Nat. Prod, 1993, 56, 1905-1911. [21] Kuwajima, H.; Morita, M.; Takaishi, K.; Inoue, K.; Fujita, T.; He, Z.-D.; Yang, CK,\ Phytochemistry, 1992,31, 1277-1280. [22] Gariboldi, P.; Jommi, G.; Verotta, L.; Phytochemistry, 1986, 25, 865-869. [23] LaLonde, R.T.; Wong, Ch.; Tsai, A. I.-M.; J. Am. Chem. Soc, 1976, 98, 30073013. [24] Tanahashi, Y.; Watanabe, H.; Itoh, A.; Nagakura, N.; Inoue, K.; Ono, M.; Fujita, T.; Morita, M.; Chen, C.-C; Phytochemistry, 1993, 32, 133-136. [25] Jimenez, C; Riguera, R.; Nat. Prod Reports, 1994, 591-606. [26] Damtoft, S.; Franzyk, H.; Jensen, S.R.; Phytochemistry, 1992, 31,4197-4201. [27] Tanahashi, T., Shimada, A., Nagakura, N., Inoue, K., Kuwajima, H., Takaishi, K., Chen, C.-C, He, Z.-D., Yang, C.-R.; Chem. Parm. Bull, 1993, 41, 1649-1651. [28] Kostova, I.; Yossifova, T.; Vassileva, E.; Mikhova, B.; Planta med, 1993, 59, A705. [29] lossifova, T.; Vogler, B.; Klaiber, I.; Kostova, I.; Kraus, W.; Phytochemistry, 1999, 50,297-301. [30] Oshima, R.; Yamauchi, Y.; Kumanotani, J.; Carbohydrate Research, 1982, 107, 169-176, [31] Lambrev, J.; Yankov, N.; Bachvarova, T.; Adjarova, E.; Nauchni Trudove na Vissh Selskostopanski Institute 'V. Kolarov", Plovdiv, 1961, 9, 311-317.

348

[32] Jurd, L.; Gorse, J.; King, A.D.; Bayne, H.; Mihara, K.; Phytochemistry, 1971, 10, 2971-2974. [33] Kostova, I.N.; Nikolov, N.M.; Chipilska, L.N.; J. Ethnopharmacol, 1993, 39, 205208. [34] HQiss,F.; Aerosol report, 1975, 12,540-553. [35] lossifova, T.; Kujumgiev, A.; Ignatova, A.; Vassileva, E.; Kostova, I.; Pharmazie, 1994, 49, 298-299. [36] Bankova, V.; Kujumgiev, A.; Ignatova, A.; Dyulgerov, A.; Pureb, O.; Zanyansan, Z.; Popov, S.; Proc. Vint. Conf. Chem. Biotechnol Biol Active Nat. Prod, Varna, 1989,2,239-343. [37] Kubo, I.; Matsumoto, A.; Takase, I. J. Chem. Ecol, 1985, 11,251-263. [38] Romans, A.L.; Fuchs, A.; J. Chromatogr, 1970, 51, 327-329. [39] Hamburger, M.O.; Cordell, G.A.; J. Nat. Prod, 1987, 50, 19-22. [40] lossifova, T.; Vogler, B.; Klaiber, I.; Kostova, I.; Kraus, W.; 46^*^ Ann. Congress Soc. Med. Plant Res., Vienna 1998, Abstract and Poster 034. [41] Klouchek, E.; Kostova, I.N.; Popov, A.; Compt. Rend. Acad Bulg. Sci, 1994, 47, 125-128. [42] Houlihan, H.M.; Ho, C.-T.; Chang, S.S.; J. Am. Oil Chem. Soc, 1985, 62, 96-100. [43] Torel, J.; Cillard, J.; Cillard, P.; Phytochemistry, 1986,25, 383-385. [44] Dziedzic, S.Z.; Hudson, B.J.F.; Food Chem, 1984, 14,45-51. [45] Marinova, E.M.; Yanishlieva, N.Vl.; Kostova, I.N.; Food Chem, 1994, 51, 125132. [46] Yanishlieva, N.VL; Marinova, E.M.; Fat Sci. Technol, 1992, 94, 374-379. [47] Marinova, E.M.; Yanishlieva, N.; Fat Sci. Technol, 1992, 94, 428-432. [48] Marinova, E.M.; Yanishlieva, N.; Comm. Dept. Chem. Bulg. Acad. Sci, 1986, 19, 524-527. [49] Pathak, M.A.; Photochem. Photobiol, 1990,51,68-69 8. [50] Cesarini, J.P.; Msika, P. Photochem. Photobiol, 1988,47, 73-74 S. [51] Lazarova, G.; Kostova, I.; Neychev, H.; Fitoterapia, 1993, LXIV, 134-136. [52] Lazarova, G.; Ignatova, M.; Biotechn. BioE, 1991, 5, 37-42. [53] Bakalova, R.; Somleva, T.; Serebinova, E.; Ognianov, A.; Kagan, V.; Stoichev, Z.; Free Radicals and Biostabilizers, Sofia,19S7, Abstract from the Symposium, p.9. [54] Dias da Silva, W.; Lepow, I.H.; J. Exp. Med, 1967,125,921-946. [55] Frank, M. M.; Frues, L.F.; Immunology Today, 1991,12, 322-326. [56] Engleberger, W.; Hadding, U.; Etschenberg, E.; Graf, E.; Leyck, S.; Winkelman, J.; Pamham, M.J.; Intern. J. Immunopharmacol, 1988, 10, 729-737. [57] Stefanova, Z.; Neychev, H.; Ivanovska, N.; Kostova, I. J. Ethnopharmacol, 1995, 46, 101-106. [58] ?\\\QX,'^B.\Lymphology, 1980, 13, 109-119. [59] Ivanovska, N.; Yossifova, T.; Vassileva, E.; Kostova, 1.; Meth. Find. Exp. Clin. Pharmacol 1994, 16, 557-562. [60] Ivanovska, N.; lossifova, T.; Kostova, I.; Phytotherapy Research, 1996, 10, 555558. [61] Van Dijk, H.; Rademaker, P.M.; Klerx, J.P.A.M.; Willers, J.; J. Immun. Meth, 1985,85,233-243.

349

[62] Alakom, B.; Lacal, I.C.; Femandez-Souza, J.M.; Carrasco, L.; Antiviral Res, 1984, 4,231-244. [63] Tan, G.T.; Pezzuto, J.M.; Kinghom, D.A.; J. Nat Prod, 1991, 54, 143-154. [64] Galabov, A. S.; lossifova, T.; Vassileva, E.; Kostova, I.; Z Naturforsch, 1996, 51c, 558-562. [65] Rada, B.; Zavada, J.; Neoplasma, 1962, 9, 57-62. [66] Kostova, I.N.; Borissova, M.V.; Bulg. Patent # 51077/1991.

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

351

NATURAL COMPOUNDS FOR THE MANAGEMENT OF UNDESIRABLE FRESHWATER PHYTOPLANKTON BLOOMS KEVIN K. SCHRADER, AGNES M. RIMANDO, STEPHEN O. DUKE United States Department ofAgriculture, Agricultural Research Service, Natural Products Utilization Research Unit, P.O. Box 8048, University, MS 38677-8048, United States ofAmerica ABSTRACT: Significant losses occur every year in aquaculture, farm livestock, and waterfowl due to problems with toxin and "off-flavor" compound production from microalgae blooms. In addition, an overabundance of algae can clog streams and create flavor problems for municipal drinking water suppHes. At present, application of synthetic compounds to the affected aquatic ecosystem is one management method used to control and prevent the growth of noxious phytoplankton. Unfortunately, many of these synthetic compounds have limitations for their usefulness in controlling phytoplankton blooms including restricted use (by government), broad-spectrum toxicity towards non-target organisms, high toxicity to non-target organisms, and public perception of adverse health risks associated with their use. The discovery, characterization, and use of natural compounds for phytoplankton control would provide environmentally safe alternatives to chemical compounds. This chapter focuses on past and present research in natural algicides including isolation and characterization of bioactive lead compounds, toxic selectivity and mechanism of action of lead natural compounds (and materials), and efficacy in aquatic ecosystems in terms of impact on target phytoplankton blooms and non-target organisms. The economic cost of using natural algicides is also discussed.

INTRODUCTION Freshw^ater phytoplankton blooms commonly occur in reservoirs, lakes, canals, and ponds under eutrophic and other physicochemical conditions that are favorable for bloom formation. Among the different types of phytoplankton blooms that can occur in freshwater ecosystems, cyanobacterial (blue-green algal) blooms are usually the most undesirable for the following reasons: 1) certain species of cyanobacteria can produce toxins that kill aquatic and terrestrial animal life; 2) some species of cyanobacteria produce "off-flavor" compounds that can impart an undesirable taste to cultured fish; 3) cyanobacteria are a poor base for

352 aquatic food chains in aquaculture ponds [1]; 4) cyanobacteria are poor oxygenators of water in food-fish production ponds [1]; and 5) filamentous cyanobacterial blooms can block filters used in drinking water supply systems. In this chapter, we will focus on freshwater cyanobacterial blooms. Cyanobacterial toxins are increasingly associated with human and animal acute and chronic health problems [2]. Occurrences of toxic cyanobacterial blooms are worldwide [2,3], and toxin production has been confirmed in bloom-forming cyanobacteria such as members in the genera Anabaena, Aphanizomenon, Cylindrospermopsis, Lyngbya, Microcystis, Nodularia, Nostoc, Oscillatoria (Planktothrix), and Umezakia [2,3]. Of the dozens of cyanobacterial toxins identified, most are classified as cytotoxic, hepatotoxic, neurotoxic, or causing allergic or irritation-type reactions. Examples of some of the health-related incidents in humans associated with cyanobacterial toxins include the following: 1) severe hepatoenteritis in 149 persons exposed to Cylindrospermopsis raciborskii [4]; 2) skin irritations and gastroenteritis after bathing in freshwater containing cyanobacteria [5]; 3) rashes and fever after contact with a microcystin-containing bloom oi Oscillatoria agardhii [2]; 4) 2000 cases of gastroenteritis and 88 deaths linked to Microcystis and Anabaena bloom in a Brazilian reservoir [2]; 5) 60 deaths in haemodialysis patients in Brazil linked to microcystin from cyanobacteria [6]; and 6) irritation and bhstering due to a toxic bloom of Microcystis [7]. Water filtration through activated carbon in a filter bed at the end of the water treatment process is one method used by some municipal water systems to help remove cyanobacterial toxins [8]. The oxidants chlorine and ozone also help degrade some types of cyanobacterial toxins [8]. Management options that involve removal of cyanobacteria or prevention of their growth in water supply reservoirs and lakes include the use of artificial mixing to reduce water stratification which favors the growth of cyanobacteria [1,9], reduction of nutrient loading to the water body, and the use of algicides, although release of toxins due to lysis of cyanobacterial cells from an established bloom can result in a short term increase in dissolved toxins levels [8]. There are numerous reports attributing cyanobacterial blooms to poisoning and death incidents in wild animals and domestic livestock. Wild and domestic animals will usually avoid drinking water contaminated with cyanobacteria. However, fences that restrict the movements of farm animals and limit their access to cyanobacteria-free

353 drinking water has probably contributed to more cases of poisoning incidents involving farm and livestock animals than for wild animals [7]. The most prevalent reports of cyanobacterial toxicosis involving livestock have been in sheep [8] and in cattle [10]. Other cyanobacterial poisonings of domestic animals include lambs [11-13], horses [14], pigs [14], and dogs [14,15]. Waterfowl have also been poisoned lethally by ingestion of a filamentous species of cyanobacteria [16]. There are also other repeated cases of wildUfe poisonings after ingestion of water containing cyanobacteria [17]. In aquaculture ponds, fish kills due to toxic cyanobacterial blooms are infrequent, and there is little direct evidence of harmful consequences of toxin-producing cyanobacteria in aquaculture ponds [18]. At least one report [19] attributes large losses of channel catfish {Ictalurus punctatus L.) in a catfish production pond in the southeastern United States to the toxic cyanobacterium Aphanizomenon flos-aquae. Fish may avoid ingesting harmful quantities of toxic cyanobacteria to help reduce the adverse consequences posed by such blooms. In any case, prolonged exposure to toxin-producing cyanobacteria may stress fish and cause them to become more susceptible to adverse conditions, thereby reducing growth [18]. Numerous off-flavor compounds have been identified and implicated as the cause of noxious taste and odor episodes in drinking water supplies and in cultured fish. The compounds of microbial origin most often implicated in such episodes are geosmin (^ran5-l,10-dimethyl-^ra«5-9decalol) (lA) and 2-methyHsobomeol or MIB ((1-R-exo)-1,2,7,7tetramethyl bicyclo-[2,2,l]-heptan-2-ol) (IB).

H3C

B

Fig. (1). Geosmin (A) and 2-methylisobomeol (B)

CH3

354

Geosmin is often described as having an earthy odor, similar to the odor encountered from freshly plowed soil. MIB has a musty odor, similar to the odor from a damp cellar. Planktonic, filamentous species of cyanobacteria are most commonly associated with geosmin and MIB production in freshwater ecosystems. In water, the sensory threshold odor concentration (concentration recognized as earthy or muddy by 75% of judgements) for geosmin is 18-20 ng/L [20] and 40 ng/L for MIB [21]. In fish, sensory threshold concentrations differ from species to species, and the range for geosmin is 6-10 |Lig/kg [20,22] and for MIB is 0.1-0.7 |ig/kg [20,23]. Other off-flavor compounds of microbial origin have been isolated and identified from cultured fish and drinking water supplies. P-Cyclocitral has a tobacco-like odor and has been isolated from blooms of the cyanobacterium Microcystis spp. [24] and the flesh of farm-raised channel catfish [25]. P-Cyclocitral does not appear to be a large off-flavor problem in aquaculture [18]. Isopropylmercaptan, a compound with an oniony odor, has been isolated from cultures of Microcystis sp. [26], and 1-octen3-one, a mushroom-like odor-producing compound, has been isolated from cultures of an Anabaena sp. [27]. In addition, offensive sulftircontaining compounds are released into the water during cyanobacterial cell decomposition [26,28]. In catfish aquaculture, offensive odors can be rapidly absorbed in the flesh of the fish. The major uptake route for geosmin and MIB is across the gills [18], and these compounds are primarily stored in the adipose tissue [29]. Elimination of these odorous compounds is much slower than the uptake rate, and elimination rate decreases as water temperature decreases and adipose tissue content of the catfish increases [30]. The problems with off-flavor episodes in municipal drinking water systems (reservoirs) are frequent and can occur worldwide. Earthy and musty off-flavor incidences in water supply systems have been attributed to cyanobacteria in CaHfomia, USA [31-36], Florida, USA [37], Pennsylvania, USA [38], Austraha [39,40], Canada [41], Japan [42-47], and Norway [48]. In 1989, 63% of 388 water utilities surveyed in the United States reported earthy- or musty-odor episodes in their drinking water supply [49]. Such episodes usually result in consumer (customer) complaints to the local water utility board or company but do not represent a health hazard to humans. Re-occurrences of off-flavor episodes erode consumer confidence in the quality and safeness of public drinking water

355 supplies [50] and have contributed to an increase in the consumption of bottled water [51]. Water utilities usually rely on filtration of the water supply through powdered or granular activated carbon at the treatment plant to remove the undesirable earthy and musty odors [49-51]. However, during cyanobacterial bloom die-offs in the water supply ecosystem, lysis of the dead or dying cyanobacterial cells releases such large amounts of the off-flavor compounds that carbon filtration usually can not completely remove the earthy and musty odor compounds. Ozonation is another method used by some municipal water districts to reduce undesirable tastes and odors in drinking water. The addition of the oxidant ozone to water at alkaline pH levels will significantly reduce levels of geosmin and MIB [49,51]. Conditions that favor development of cyanobacterial blooms are high nutrient loading rates, diminished vertical mixing of the water column, and warm water temperatures [1]. Cyanobacteria are able to out-compete green algae and other types of phytoplankton for light and nutrients due to their ability to regulate their position in the water column by changing cell buoyancy. In addition, conditions in which nitrogen is limited relative to phosphorus favor growth of nitrogen-fixing species of cyanobacteria. Unfortunately, cyanobacteria are able to compete well for resources but have slow growth rates compared to most species of eukaryotic phytoplankton [55]. Slower growth rates of cyanobacteria translate into reduced aquaculture yields for types of fish or crustaceans that depend on in-pond primary production. Cyanobacteria are poorly utilized as a food source by herbivorous crustacean zooplankton, an early step in the aquatic food chain, thereby reducing transfer efficiency of nutrients from primary production to aquaculture crop [1]. Biomass-specific rates of net carbon fixation by cyanobacteria are lower than for eukaryotic phytoplankton, and the slower growth rates of cyanobacteria also translate to their poor oxygenation of water. The formation of surface scums of cyanobacteria followed by their massive die-offs can endanger fish crops due to acute dissolved oxygen depletion [56]. The worldwide economic impact of undesirable cyanobacterial blooms in municipal water systems is unknown. Increased costs can occur due to increased amounts of activated carbon used in water treatment facilities to counteract higher levels of off-flavor compounds in the water. In addition, filters and filtration equipment can become clogged with certain types of filamentous cyanobacteria and result in significant "down" time to correct these problems. Certain species of cyanobacteria (e.g., Lyngbya wollei)

356 can also form floating mats of filaments that clog waterways, ponds, and lakes, thereby impeding navigation and recreation. Although cyanobacterial-derived off-flavors can occur in many aquatic animals and in aquaculture throughout the world, the only economic impact studies of such off-flavor problems have been in channel catfish aquaculture in the southeastem United States. Off-flavor catfish are held until flavor quality improves, and these delays in harvest result in economic losses from inventory management problems and increased production costs [57]. Based on off-flavor problem added costs [57-60], catfish producers may have lost as much as U.S.$60 million in 1998. Approaches to the management of undesirable cyanobacterial blooms in freshwater ecosystems include managing around the bloom, treatment of the water or aquatic animal (depuration) to purge undesirable microbial metabolites, and prevention of the development of cyanobacterial blooms. Several prevention methods and strategies utilized in drinking water reservoirs and lakes are the following: 1) water shed protection including the control of point sources of nutrient inputs; 2) hypolimnetic aeration to destratify the water column; 3) control of inorganic chemical composition (e.g., alkalinity) that promotes cyanobacterial growth; and 4) the application of chemicals to kill or prevent the growth of cyanobacteria [52]. In aquaculture pond management, several approaches have been used to help prevent cyanobacterial growth. These approaches include the following: 1) the apphcation of alum (aluminum sulfate) to precipitate (remove) phosphorus from the water; 2) nutrient removal from pond water by macrophytic plants; 3) manipulating the ratio of total nitrogen to total phosphorus in the water; 4) water exchange to reduce nutrient availability; 5) biomanipulation using planktivorous fishes; and 6) application of synthetic algicides [18]. The use of algicides is an approach similar to that used for dealing with weeds in terrestrial agriculture, i.e., application of phytotoxic chemicals (herbicides). Many commercial aquaculturists favor this approach. Currently, only copper-based products are approved by the United States Environmental Protection Agency for widespread use as an algicide in food-fish production ponds and municipal drinking water reservoirs in the United States. These copper-based products, such as copper sulfate and chelated copper compounds, are commonly used in aquaculture and in municipal drinking water systems. Unfortunately, copper-based algicides do not provide a large degree of toxic selectivity towards noxious

357

cyanobacteria [61]. In addition, strains within species of cyanobacteria can vary considerably in their tolerance to copper [62,63]. Copper sulfate toxicity towards non-target organisms such as plants and fish is very susceptible to environmental variables. Environmental factors can decrease the toxicity of copper sulfate so that it is ineffective in controlling the growth of target cyanobacteria or increase the toxicity towards nontarget organisms [18]. Copper from copper sulfate will rapidly disappear from the water column after application and accumulate in the aquatic ecosystem sediments. Copper in chelated copper algicides does not rapidly precipitate out of solution and is claimed to be more effective than copper sulfate since chelated copper provides a higher concentration of total dissolved copper in the water for a longer period of time [64]. However, these benefits may be outweighed by the increased cost of using chelated copper products compared to copper sulfate. Although repeated applications of copper algicides at low dosages reduces the overall abundance of cyanobacteria and off-flavor problems in catfish production ponds [65], the broad spectrum toxicity of copper reduces water quality in terms of depressed dissolved oxygen levels and increases in ammonia concentration [66]. Such deterioration in water quality can endanger fish health and reduce aquaculture productivity. The unacceptable risks that synthetic herbicides pose to the environment and human health prevent their approval for use in drinking water supply systems and aquaculture ponds. The discovery of environmentally safe compounds that have a greater degree of toxic selectivity towards noxious species of cyanobacteria than copper algicides currently available would greatly benefit commercial aquaculturists and municipal drinking water suppliers and consumers. The use of natural products to selectively control cyanobacteria is desirable due to environmental safety issues and consumers' negative perceptions of the use of synthetic compounds (herbicides). Cyanobacterial Inhibitory Compounds and Sources Compounds reported to have inhibitory effects on cyanobacteria have been isolated from a diverse range of organisms including terrestrial plants, aquatic plants, fungi, bacteria, actinomycetes, protozoa, and even from some species of green algae and cyanobacteria. The structural types

358 of the secondary metabolites identified from these organisms are equally diverse, with a number being phenolic compounds. Barley straw has been shown to be effective in controlling general algal blooms [67-70]. Cyanobacteria inhibited by barley straw include Anabaena sp., Aphanizomenon sp., Microcystis sp., and Oscillatoria sp. [71]. Barley straw introduced into the water inhibited algal growth during its decomposition [72]. In a study using an artificial rumen reactor, decomposing barley straw was shown to contain lignin-derived phenolic compounds [73]. Derivatives of coniferyl, coumaryl, and sinapyl alcohols were identified by UV and mass spectrometry. Phenol, /7-cresol, and biphenyl-2-ol, compounds of known toxicity to algae, were present in close proximity to rotting straw at concentration ranges of 10-100, 10-100, and 1-10 |J,g/L, respectively [74]. In their studies. Barret and coworkers [75] also found a relatively high concentration of phenol, /?-cresol, and additionally, 4-ethyl-phenol released from the straw. The phenolics tannic acid, propyl gallate, and gallic acid were found to inhibit growth and pigment synthesis of Nostoc sp. strain MAC and Agmenellum quadruplicatum PR-6 [76]. In the same study, methyl gallate was found to also inhibit the growth of ^. quadruplicatum PR-6, but not of Nostoc sp. Further studies showed that these tannin compounds, except for gallic acid, inhibited chlorophyll a and c-phycocyanin synthesis in Nostoc sp. strain MAC, and, except for methyl gallate, also had an effect on electrolyte efflux [77]. Tannins have been reported to be the algicidal constituents of several Acacia species [78,79]. The algicidal activity was lost after acetylation and methylation of the plants' methanolic extracts indicating that the activity was due to the free phenolic hydroxyl groups present as galloyl residues [79]. Tannins from the extracts of the fhiits of Acacia nilotica, in a spray-dried form, controlled algal populations including Oscillatoria and Microcystis [78]. Leachate from the hardwood tree aspen (Populus tremuloides) inhibited algal growth [80] and was found to be rich in phenols. However, experimental data obtained do not support toxicity attributable only to the phenolic compounds. No further work has been reported on the algal toxicity of aspen wood leachate. Theoretically, aquatic plants, both higher plants and algae, might be expected to be a source of algicides since these organisms compete with cyanobacteria. There is some evidence to support this theory. Cattails {Typha latifolia) are rich in p-sitosterol, which has algicidal activity against Phormidium autumnale, Naviculla pelliculosa, Chlorella emersonii, Chlorella vulgaris, and Stichococcus bacillaris [81]. In a

359

bioassay-directed study of the components of water hyacinth {Eichhornia crassipes), two highly algicidal compounds identified as A^-phenyl-Pnaphthylamine and 7V-phenyl-a-naphthylamine were found [82]. Few such bioassay-directed studies directed toward algicide discovery have been made. Species of the watermilfoil genus Myriophyllum (Haloragaceae) have been reported to release phenohc cyanotoxins [83]. Three 1 '-0-caffeoyl-6'-0-galloyl-p-Dphenylpropanoid glucosides: glucopyranose (2), r-0-coumaroyl-6'-0-galloyl-P-D-glucopyranose (3), and r-0-sinapoyl-6'-0-galloyl-P-D-glucopyranose (4) were found to be toxic to the common freshwater cyanobacterium Synechococcus leopoliensis [84]. Phenylpropanoid 2 was found to undergo hydrolysis under experimental conditions, yielding gallic and caffeic acids. Observation of higher activity of the intact molecules of compounds 2, 3, and 4 compared to the individual phenolic acid hydrolytic products led the authors to believe that synergism was occurring between the phenolic acids. Synergism was confirmed with mixtures of gallic-p-coumaric and gallic-sinapic acids. Ellagic acid, (+)-catechin, gallic acid, and pyrogallic acid released by M spicatum were shown to synergistically inhibit the growth of Microcystis aeruginosa [85]. In a separate study, tellimagrandin II (eugeniin) (5) was found to be the main inhibitory compound in M. spicatum following a bioassay-directed fractionation [86]. Tellimagrandin II strongly inhibited Anabaena sp. PCC7120, Synechococcus sp. PCC 6911, Synechocystis sp. CB-3, and Trichormus var. P-9 in agar-diffusion assays. Investigation of M alterniflorum and M, verticilatum from northern Germany did not reveal the presence of tellimagrandin II, but several other hydrolyzable polyphenols were found. Proserpinaca palustris (another Haloragaceae plant) also inhibited Trichormus var. P-9. Hydrolysis of the crude extract of M. alterniflorum yielded gallic and ellagic acids, while tannase treatment of the crude extracts of M, verticilatum and of P. palustris yielded mainly gallic acid. In M. brasiliense, tellimagrandin II was also identified as the most toxic components to Microcystis aeruginosa and Anabaena flos-aquae, along with 1-desgalloyleugeniin (6) [87]. Additionally, a mixture of epicatechin-3-gallate and catechin-3-gallate, gallic acid, and the flavonoids quercetin, quercitrin, and avicularin were reported as anti-algal compounds from this aquatic plant.

360

R1 2 OH 3 H 4 OMe

HO

R2

H H OMe

OH

5 R = galloyi 6R = H

361 Fatty acids appear to also have activity against cyanobacteria. Oxygenated fatty acids resembling prostaglandin structures and with selective inhibition against cyanobacteria were isolated from the aquatic plant Eleocharis microcarpa [88]. The prominent components from the active chromatographic fraction were identified as C^^ trihydroxycyclopentyl (7) and C^g hydroxycyclopentenone (8, 9) fatty acids. These compounds were also found in other higher aquatic plants (Potamogeton sp., Najas sp.) as well as in the waters in which these plants grow. Long chain fatty acids were also found in the waters around decomposing barley straw [74]. Analysis of the waters by GC-MS showed the presence of 9-eicosene, octadecanoic acid, hexadecanoic acid, heptanoic acid, and hexanoic acid in high concentrations [74]. From the ether extract of the aquatic plant Typha latifolia, the compounds linolenic acid, 6-linolenic acid, and an unidentified CI8:2 fatty acid were isolated, along with three steroidal compounds; these compounds exhibited selective inhibition of T625 Synechococcus leopoliensis and T1444 Anabaena flos-aquae in vitro [89].

COOH

8 R = .CH=CHCH2CI^ 9 R = -(CH2)3CH3

Ricinoleate is in the same chemical class of compounds as those allelochemicals isolated from Eleocharis microcarpa and was determined

362 to be selectively toxic towards cyanobacteria [90]. Potassium ricinoleate, the active constituent of the commercial algicide Solricin 135®, was shown to inhibit the growth of cyanobacteria in laboratory tests [90]. However, when its effectiveness was evaluated in ponds, potassium ricinoleate did not inhibit the growth of cyanobacteria [91]. The authors could not explain the loss of activity. Phenylpropanoid-type compounds have been isolated as the algal inhibitory constituents from the aquatic plant Acorus gramineus [92] using a filter paper-dish bioassay. It was noted that l,2-dimethoxy-4-(£'-3'methyloxiranyl) benzene (10), l,2,4-trimethoxy-5-(Z-r-propenyl)benzene (11), l,2,4-trimethoxy-5-(£'-3'methyloxiranyl)benzene (12), the three most abundant phenylpropanes isolated, were either inhibitory or not in each of the cyanobacterial strains tested which included Anabaena flos-aquae, Nostoc commune, and Synechococcus leopoliensis.

MeO OMe R1

R2

10 H C - C H - C H o

H

O 11 HC = CH-~CH3

OMe

12 H C - C H - C H o

OMe

\ / O

^

A number of structurally unrelated compounds were tested in a general screen using microtiter plates to discover natural products with selective growth inhibition towards off-flavor-compound producing cyanobacteria [93]. In this study, ^raw5-ferulic acid was found to be the most selectively toxic to Oscillatoria cf chalybea among the phenolic compounds tested. However, it was not toxic to Anabaena sp. LP 691. Cinnamic acid, one of the compounds released from decomposing barley straw, was also selectively toxic to O. cf chalybea. Among the naturally-occurring

363 selectively toxic to O. cf. chalybea. Among the naturally-occurring quinones tested anthraquinone (13) was the most toxic, and also selectively toxic, to O. cf. chalybea. Anthraquinone, 2methylanthraquinone, and juglone had the highest differential activity to cyanobacteria compared to green algae tested. Artemisinin (14), the antimalarial sesquiterpenoid lactone peroxide from Artemisia annua, was shown in this screening study to be highly toxic with selectivity towards O. cf chalybea. The use of A. annua leaves and/or flowers in ways similar to the use of barley straw may be of great potential.

Some cyanobacteria produce secondary metabolites that inhibit the growth of other cyanobacteria. Cyanobacterin (15), bearing a y-ylidene-ybutyrolactone structure, was isolated from the freshwater cyanobacterium Scytonema hofmanni and was found to inhibit Synechococcus cultures and other microorganisms [94]. Cyanobacterin appears to act as an allelopathic substance by allowing the survival of the slow-growing Scytonema while among the more prolific-growing species. Studies with Synechococcus sp. and Euglena gracilis indicate cyanobacterin acts on the thylakoid membranes [95]. Hapalosiphon intricatus and Hapalosiphon fontinalis were reported to be inhibitory to Anabaena sp., but the extracellular substance(s) secreted was not identified [96]. Subsequent work with H. fontinalis led to the isolation of a tetracyclic indole alkaloid, hapalindole A (16), the compound responsible for antialgal acitivity [97]. The isothiocyanate form of hapalindole A (17) was also isolated, among other minor constituents.

364

OCHo 15

16 R = - N ^ C 17 R=:_N=C=S A structurally unique group of compounds (those with a long hydrocarbon chain bearing a conjugated enediyne and a heterocyclic ring at the end) was isolated from the fresh-water cyanobacterium Fischerella muscicola UTEX 1829 [98,99]. The compounds, named fischerellins A (18) and B (19), were also found in Fischerella ambigua [99,100]. Fischerellin A inhibited other cyanobacteria {Anabaena variabilis P9, Phormidium sp. UTEX 1540, Synechococcus sp. PCC 6911, Synechocystis sp. CB-3), as well as several chlorophytes, but not eubacteria [100]. Fischerellin B was also found to have algicidal properties but is produced in smaller quantities than fischerellin A [99]. Fischerellin A is a potent photosystem II (PSII) inhibitor [98] and affects the fluorescence transients and O^ evolution by the cyanobacterium Anabaena P9 and the green alga Chlamydomonas reinhardtii, as well as higher plants [101]. An algicidal compound from Oscillatoria late-virens was found to be toxic towards

365 Microcystis aeruginosa via a mechanism that appears to be an inactivation of PSII-mediated electron flow [102]. In an earUer study, the active fractions of an ether extract from O. late-virens were found to contain long-chain saturated fatty acids as major components [103].

18

Metabolites from heterotrophic bacteria have also been reported to inhibit cyanobacteria. Pseudomonas aeruginosa inhibited the growth of several test species, which included the cyanobacteria Anabaena sp., Phenazine Phormidium bohneri, and Oscillatoria agardhii [104]. pigments released by the bacterium mediated inhibition of algal growth. Oxychlororaphine (20) and 1-hydroxyphenazine (21) showed strong antialgal activity while pyocyanine completely inhibited algal growth.

O II C-NH2

^-^

N^ ^ ^ ^ 20

OH

^ ^

N^ ^ ^ \ ^ 21

366 Antibiotics have also been studied as useful compounds to selectively inhibit the growth of cyanobacteria. Matsuhashi et al. [105] found several antibiotics (e.g., bacitracin, D-cycloserine, novobiocin, ristocetin, penicillin G, and vancomycin) that inhibited the growth of Anabaena variabilis at concentrations ranging from 10-1000 |Lig/mL while several green algal strains tested were unaffected at the same concentrations. The semisynthetic P-lactam antibiotic amoxicillin showed strong inhibition of Microcystis aeruginosa as indicated by a decrease in chlorophyll concentration measured fluorometrically [106]. The use of antibiotics to selectively control the growth of cyanobacteria is not a favorable approach due to the broad-spectrum toxicity of certain antibiotics toward bacteria and the subsequent potential for the selection of antibiotic-resistant strains of catfish disease-producing bacteria. Yeast extract was reported to contain components that are toxic to cyanobacteria [107]. L-lysine and malonic acid completely killed Microcystis viridis at a concentration of 1 and 40 ppm, respectively. Lysine malonate was found to be more toxic than DL-lysine. A few other types of microorganisms have been reported to inhibit the growth of cyanobacteria; however, no work has been performed on the structural characterization of the responsible constituents. In the screening of 65 cyanobacterial strains for production of phages and/or antibiotics, seven N^-fixing strains of cyanobacteria {Anabaena doliolum, Fischerella muscicola, and five strains of Nostoc) were found to produce metabolites which caused clearing of growth of indicator cyanobacterial strains on sohd media [108]. The metabolite produced by Nostoc sp. 78-11 A-E appears to be a proteinaceous, low molecular weight compound. None of the other metabolites were structurally characterized. In another study, cultures (whole cell) of 198 cyanobacterial strains isolated from soil and freshwater samples were screened for inhibitory activity against green algal species [109]. Twenty isolates found to be active against green algae were also tested for activity against cyanobacteria. Anabaena doliolum was inhibited by Fischerella JAVA 94/20. Fischerella NEP 95/1 inhibited Anabaena sp.; the toxicity was lost with the inclusion of proteinase K in the media, suggesting that the active agent was a peptide. In the same study, three Fischerella strains (JAVA 94/10, NEP 95/1, LOM 95/17) were inhibitory to Synechocystis PCC 6803 and Nostoc sp. from the plant Macrozamia communis (Cycadaceae). Also, Calothrix WA 96/8 inhibited only A. doliolum, and three strains of Calothrix, 9 strains of Fischerella,

367 and Nostoc NSW 95/10 showed toxicity toward Anabaena circinalis. Microcystis aeruginosa, and Nodularia spumigena. These cyanotoxins were neither isolated nor characterized. Several other viral, fungal, bacterial, and protozoal agents were reported to inhibit cyanobacteria but no phytochemical studies on the inhibitory components from these organisms have been done [110]. The diversity of the classes of compounds reported to be toxic towards cyanobacteria reflects the potential for many sites and mechanisms of actions for growth inhibition (to be discussed in the next section). The diversity of the sources that has been reported so far substantiates the fact that natural products are a rich source of cyanobactericidal compounds. Modes of Action Knowing the mode of action of pesticides is valuable for several reasons. For example, this information can be useful in the design of better pesticides, in anticipating toxicological problems, and in predicting the evolution of resistance. Algae and cyanobacteria are photosynthetic and share with higher plants many of the biochemical target sites of synthetic herbicides and other phytotoxins. There is considerable literature on the mode of action of synthetic herbicides [e.g., 111-114] and, to a lesser degree, on that of phytotoxins produced by plant pathogens [e.g., 115,116]. Many herbicides are as toxic to algae as to higher plants [e.g., 61], and, because of their ease of manipulation, are used by some scientists in studies for indication of mode of action of herbicides in higher plants [e.g., 117]. Relatively few studies have been done to identify compounds that might be used as algicides, and even fewer studies have been done to determine the mechanism of action of those compounds with useful activity as algicides. However, just as algicide mode-of-action studies have been extrapolated to higher plants, the reverse can also be done. Two points should be kept in mind when studying mode or mechanisms of action of biocides. First, a compound that kills will eventually affect every physiological system in the target organism. Most of the literature on the mode of action of herbicides deals with these secondary effects. Proof of a primary molecular target site is often very difficult, especially when the target is one that has not been previously described. Usually, the first papers on a biocide describe secondary effects, and are then followed by more definitive studies. For example.

368 the first paper on the biological effects of cyanobacterin (15) described its effects on growth [94], to be followed by a study showing its effects on PSII [118]. A strategy for determination of molecular target sites of herbicides or algicides has been recently published [119]. A second important aspect of modes of action that one should be aware of is that a compound can have more than one molecular site of action. For example, fischerellin A (18) is both a PSII inhibitor and an antifungal agent [98]. Clearly, the molecular target as a fungicide is not PSII. If both sites are of equal value in killing the target organism, determination of the molecular sites of action can be challenging. There are only about fifteen well-defined molecular target sites for commercial herbicides and about twenty for natural phytotoxins (Table I). Most of these have been documented to be toxic to green algae or cyanobacteria. Furthermore, we are not aware of any studies showing that any of these compounds are not toxic to green algae or cyanobacteria. There has been no systematic study of the toxicity of these compounds to algae or cyanobacteria. The closest attempt at this has been the work of our laboratory [61,93]. There is relatively little overlap between the molecular target sites of commercial herbicides and those of natural compounds (Table I) [116,120,121]. We will describe what we know of the mechanism of action of a few selected compounds that are discussed in other parts of this chapter. The largest class of commercial herbicides are those that inhibit photosynthesis by inhibition of PSII by stopping electron transport [112]. This phenomenon is easily monitored by measuring increases in variable fluorescence or decreases in oxygen evolution. A vital component of PSII is a quinone-binding protein called D-1 in eucaryotic plants and L in cyanobacteria. A quinone (plastoquinone) is necessary as a redox compound to transport electrons. PSII inhibitors compete for the quinonebinding site on D-1. PSII inhibitors can be divided into two groups: those that depend more on binding the Ser264 amino acid of D-1 and those that bind the His2i5 moiety. The Ser264 inhibitors include triazines, substituted ureas, and carbamates, whereas the His2i5 family of inhibitors include benzoquinones, naphthoquinones, and acridones. There is apparently enough homology between D-1 and L for similar activity of all PSII inhibitors studied against electron flow in both eucaryotic plants and cyanobacteria. However, this cannot be said for all photosynthetic bacteria [101].

369

Table I. Modes of Action of Commercial Herbicides and Natural Phytotoxins (Adapted in part from references 116 and 120). Physiological site

Molecular site

Herbicide or natural product^

Amino acid synthesis aromatic amino acids branched chain amino acids

EPSP synthase acetolactate synthase

glutamine synthesis

glutamine synthetase

glutamate synthesis general amino acid synthesis ornithine synthesis methionine synthesis

aspartate amino transferase many transaminases ornithine carbamoyl transferase P-cystathionase

glyphosate imidazolinones, sulfonylureas, others many, including phoshinothricin, oxetin, and tabtoxinine gostatin gabaculin phaseolotoxin

Photosynthesis electron transport

D-1, quinone-binding

photophosphorylation

CF| ATPase

many, including triazines, substitued ureas, etc.; sorgoleone, cyanobacterin, fischerellin A ten toxin

electron transport diverters plastoquinone synthesis

photosystem 1 4-hydroxyphenylpyruvate dioxygenase

bipyridiliums {e.g., paraquat) isoxazoles, pyrazoles, and triketones; leptospermone

protoporphyrinogen oxidase ALA synthase causes accumulation phytoene desaturase

p-nitrodiphenyl ethers, oxadiazoles, etc. gabaculin ^-aminolevulinic acid many, including pyridazinones, etc. isoxazolidinones

Pigment synthesis porphyrins

carotenoids

a prenyl transferase

rhizobitoxin

Cell division mitotic disruptors

P-tubulin

dinitroanilines, phosphoric amides; vinca alkaloids, colchicine, etc.

Vitamin synthesis folate synthesis

dihydropteroate synthase

asulam

370

Table 1 - cont'd Lipid synthesis acetyl-Co A carboxylase acetyl-Co A transacylase 3-oxoacyl-ACP synthase ceramide synthase

Nucleic acid synthesis plastid nucleic acid synthesis

aryloxyphenoxypropanoates, cyclohexanediones thiolactomycin cerulenin AAL-toxin and analogues, australifungin

RNA polymerase adenylsuccinate synthase AMP deaminase

tagetitoxin hydantocidin carbocyclic coformycin

unknown at present

rose bengal and many others; cercosporin, hypericin, and many others

H"^-ATPase

syringomycin

Direct photodynamic action

Plasma membrane function

Cell wall synthesis cellulose synthesis? unknown site ^compounds in italics are natural products

dichlobenil isoxaben

Caution should be exercised in assuming that plastoquinone analogues will exert most of their phj^otoxicity through PSII inhibition because some are also excellent inhibitors of mitochondrial respiration [122]. In some cases, this may be because they bind to the electron transport protein that normally binds to ubiquinone, a redox compound involved in mitochondrial electron transport in a similar way that plastoquinone is involved in photosynthetic electron transport. Thus, some inhibit respiration by a very similar mechanism to that of PSII inhibitors. Some natural quinones, such as sorgoleone and fischerellin A (18) are apparently sufficiently close analogues of plastoquinone that they are excellent PSII inhibitors [123,124]. Juglone (22) is phytotoxic to algae and has been reported to inhibit both photosynthesis and respiration in higher plants [125,126], although exactly how it inhibits photosynthesis is not clear. An analog of juglone, anthraquinone (13), is highly phytotoxic to certain

371 species of cyanobacteria [93] and is a strong inhibitor of photosynthesis [127].

OH

O 22

One potential problem with the use of PSII inhibitors as algicides is that resistance may evolve relatively quickly. In higher plants, resistance has evolved to certain classes of PSII inhibitors, but not to others [128]. In all but a few cases, amino acid substitutions in D-1, resulting in reduced affinity for the inhibitor, account for the evolved resistance [129]. These alterations in D-1 reduce photosynthetic efficiency under some environmental conditions, but under strong selection pressure of the herbicide, herbicide resistance is favored. Another potential drawback of PSII inhibitors is that there is little likelihood of such an algicide being effective at very low doses because of the very large amount of target site in each cell. Compared to some of the "low dose" herbicides, all PSII herbicides are applied at relatively high doses. Artemisinin (14) and several of its analogues are quite toxic to both cyanobacteria and higher plants [93,130]. However, its mode of action is still a mystery. It affects a large number of physiological parameters [130], but no primary site of action has been estabhshed. Like another sesquiterpene lactone, dehydrozaluzannin C, it may act by binding SH groups of proteins [131]. Effects of artemisinin can be reversed with cysteine [132]. If this is the mechanism of action, resistance is unlikely to evolve by alterations in binding sites. Photodynamic compounds, that is those that generate singlet oxygen in the presence of light and molecular oxygen, are toxic to all living things. Singlet oxygen causes rapid membrane disruption by peroxidation of membrane lipids [112]. Hypericin (23), a photodynamic product of several species of Hypericum, is an effective biocide on higher plants [133] and cyanobacteria [134]. However, photodynamic compounds have

372

the inherent problem of being toxic to everything. There are both natural and synthetic compounds that cause the accumulation of toxic levels of Some porphyrins are particularly porphyrins in plants [135]. photodynamic, and normally do not accumulate in healthy cells.

The natural glutamate synthetase (GS) inhibitor, tabtoxinine, is toxic to the green alga Chlorella vulgaris, and this green alga was used in elucidating its mode of action [136]. There are a number of other highly active natural GS inhibitors [137] that should be examined for their effects on both green algae and cyanobacteria. There are other natural compounds that are known to be toxic to algae for which we do not have knowledge of a specific molecular target. These include many of the phenolic acids, such as ferulate, and hydrolyzable polyphenols such as tellimagrandin II (5). One mode-of-action theory for these compounds as algicides is that they non-competitively inhibit algal exoenzymes such as alkaline phosphatase [83,138]. In fact, these compounds do bind enzymes non-competitively, often inhibiting their activity. Furthermore, these compounds can be taken up by cells and inhibit intracellular enzymes. For example, tannic acid, gallate, and two analogues of gallate were found to lower levels of glutamate synthase and nitrate reductase activity extracted from Nostoc sp. [77]. However, whether this was a direct or indirect effect was not clear because no in vitro effect of the compounds was reported. If non-specific inhibition of enzymes is the mode of action of many simple phenohcs and hydrolysable

373

phenolics, they have no specific molecular site of action, although each compound might be expected to have a slightly different spectrum of enzymes that it preferentially inhibits. An advantage to such a mode of action is that evolution of resistance to such a compound at molecular target sites is highly unlikely. An alternative or complimentary theory for the mode of action of simple phenolic compounds is that they are converted to much more toxic quinones. Pillinger et al. [69] found that various phenolic decomposition products of barley straw were most toxic under conditions favorable for oxidation of the compounds to quinones, and that quinones were up to one thousand-fold more toxic to algae than the parent compounds. The most likely route to conversion to a quinone is enzymatic. Peroxidases and polyphenol oxidases can perform such a reaction, depending on the substrate. However, polyphenol oxidase cannot be detected in most green algae [139] and has not been reported in cyanobacteria. There are studies that indicate a particular mode of action as a natural algicide, without elucidating that it is a primary site. For example, Kida [140,141] reported that the Streptomyces purpeofuscus product 7-deoxyD-glycero-D-glucoheptose inhibited chlorophyll synthesis in the green alga Scenedesmus obliquus. Yet, only chlorophyll accumulation was measured. Phytotoxins with many other sites of action could cause a similar effect. Ideally, a biocide should selectively kill the target organism(s) at doses that have little or no effect on non-target organisms. In aquaculture, a compound that would selectively remove the cyanobacteria while leaving the more desirable green algae is much more desirable than non-selective algicides. A selectivity factor of greater than ten-fold is most advantageous because of uncertainties in efficacy due to environmental factors and human error. Selectivity can be based upon differences in target sites, in movement to the target sites, and/or in metabolic detoxification of the biocide or toxicants generated by the biocides (e.g., singlet oxygen). To our knowledge, we are the only laboratory that has searched for selective cyanobactericides [61,93], although others have noted differences in activity of specific compounds to different algae and algal groups [e.g., 118,142]. In our studies, we found several synthetic herbicides to have 100- to 1000-fold greater activity in causing complete inhibition of growth of the filamentous cyanobacterium Oscillatoria perornata (formerly referred to as Oscillatoria cf. chalybea) than the unicellular green alga Selenastrum capricornutum: diquat, paraquat.

374

diclofop, and bromoxynil [61]. Diquat and paraquat act by diverting electrons from photosystem I (PSI) to generate toxic levels of superoxide radical, resulting in similar damage as that caused by photodyamic compounds. Diclofop inhibits lipid synthesis via specific inhibition of acetyl CoA carboxylase. Bromoxynil is a PSII inhibitor. Only diquat and paraquat were effective on O. perornata at economical rates; however, these herbicides are perhaps the most acutely toxic of all herbicides to animals. We are unaware of any commercial herbicides that have the desired levels of selectivity and safety for use in aquaculture. However, our limited experience with natural products indicates that there may well be a natural product that has the optimal properties. Among a very limited number of natural products, we found anthraquinone, 2-methylanthraquinone, juglone, and sorgoleone to be two to three orders of magnitude more toxic to 0. perornata than to S. capricornutum [93]. These compounds are either known [128,143] or suspected PSII inhibitors. Only the first two of these compounds were active enough to be of interest as cyanobactericides. Whether there is a large difference in binding affinities of these compounds for D-1 versus L or some other mechanism of selectivity is unknown. A strategy can be derived for selective cyanobactericide discovery, based on physiological or biochemical differences between cyanobacteria and eucaryotic photosynthetic organisms. For example, a compound that would specifically inhibit synthesis of phycobilins (photosynthetic pigments peculiar to cyanobacteria) or phycobiliproteins (a phycobilinprotein complex of cyanobacteria) could be expected to be selective. This sort of biorational approach has not been attempted for discovery of algicides and has been unsuccessfiil in the discovery of herbicides. Those antibiotic cyanobactericides, such as bacitracin and penicillin G mentioned previously, probably act on cyanobacteria via a mode of action similar or identical to that by which they kill pathogenic bacteria. The selectivity of these cyanobactericides found by Matsuhashi et al. [105] supports the view that they are procaryote-specific. A variation of this approach is to target more general processes that might differ in the target organismfi-omother organisms. Cyanobacteria are more primitive photosynthetic organisms than eucaryotic plants, and thus, evolved when atmospheric oxygen was at much lower concentrations than today. Thus, they might be more sensitive to toxic oxygen species than more recent evolutionary arrivals. One could argue that the evolution of the chloroplast was in response to the toxic effects of oxygen. Our

375 finding that O. perornata was 1000-fold more susceptible to paraquat and diquat than the green alga S. capricornutum and the result of Brody et al. [134] that cyanobacteria are highly sensitive to hypericin supports this view. Thus, concentrating on natural compounds that might directly or indirectly cause toxic oxygen species to increase might be a viable strategy for the discovery of algicides and, perhaps even selective cyanobacterial algicides. Knowledge of the site of action can be valuable in designing a better algicide through structure-activity relationship (SAR) studies. This approach works quite well with synthetic compounds. However, if the biological activity of interest is the function of a natural compound in nature, its activity may have already been optimized by evolution. Examples are summarized by Duke et al. [144]. In the only case of a SAR study of a natural product algicide of which we are aware, Gleason et al. [145] could not produce a more active cyanobacterin (15) molecule than the parent molecule. There have been exceptions to the natural compound being the optimal pesticide, such as the pyrethroid insecticides. Modifications are often most needed to increase the environmental halflife or to improve uptake by the target organism. Our knowledge of the mechanisms of action of algicidal compounds is fragmentary. We still have much to do. Bioassay Methods for Discovering Selective Algicidal Compounds A variety of approaches have been used in the development of rapid bioassay systems to determine the toxicity of compounds towards algae and cyanobacteria. The variables that have been measured to determine algal response to a particular compound include rate of chlorophyll fluorescence and ^"^C-uptake to measure photosynthesis, measurement of growth inhibition zones on solid agar, growth rate, biomass, cell number at a given time, alkaline phosphatase activity, and motility. In most cases, the toxicity of the compound can be expressed as the reduction of the response relative to the control response as percentage of inhibition [146]. Several algal toxicity methods are discussed, with an emphasis on a rapid bioassay developed specifically for screening a large number of compounds to determine their usefulness as selective cyanobactericides [147].

376 An initial consideration of algal and cyanobacterial toxicity testing involves selecting a microcosm-type screen versus a screen using individual species. Microcosm-type screening involves the use of natural water containing many different species of phytoplankton in addition to a target species of algae or cyanobacteria. The argument against using single-species screens is that assumptions are made concerning the organism's responses at levels of biological organization above a single species [148]. Although microcosm-type screening may provide a more reliable indication of the target organism's response toward a toxicant while in its natural environment [148], it may limit the ability to rapidly screen compounds due to the difficulty in measuring responses of a separate individual target species amongst a mixture of phytoplankton. Therefore, microcosm-type screening should be reserved for later use, after lead algicidal compounds have been discovered using rapid, singlespecies bioassays and as an intermediate step in the progression from smaller bioassays to large full-scale lake or pond testing. Chlorophyll fluorescence, ^"^C-uptake, and growth rate measurements provide an indication of viable cells, whereas biomass and cell number measurements do not distinguish between the living and dead cells. In photosynthetic algae and cyanobacteria, the reaction centers referred to as photosystem I (PSI) and photosystem II (PSII) are special chlorophyll molecules used to dissipate excitation energy derived from the deexcitation of an excited pigment molecule that has absorbed the energy of a photon. Chlorophyll fluorescence occurs when the excited pigment returns to the ground state by the route of emitting a fluorescence photon, of a wavelength longer than that of the photon initially absorbed [149]. Most of the chlorophyll fluorescence from algae is associated with PSII [150], while in cyanobacteria, most of the chlorophyll fluorescence emanates from both PSII and phycobiliproteins [149], the major lightPhycobiliprotein harvesting pigments in cyanobacteria [151]. fluorescence overlaps with the spectrum of chlorophyll emission [149]. Chlorophyll fluorescence measurements and analyses for cyanobacteria must be modified from those used for algae due to the difference in the pattern of chlorophyll fluorescence signals between the two groups of microorganisms [149]. For certain compounds, i.e., those inhibiting electron flow in photosynthesis, fluorescence measurement can help evaluate the toxicity of a compound towards algae and cyanobacteria. Fluorescence measurement is more sensitive than measuring culture absorbency (optical

377

density). However, in vivo chlorophyll fluorescence measurement bioassays do not provide as broad a screening method of toxic compounds as with the use of growth inhibition assays. Only those compounds adversely affecting photosynthesis are best suited for such an assay. In addition, in vivo chlorophyll a fluorescence measurements are very complex and have been discouraged as a method for conducting screening bioassays [150]. Radiocarbon-uptake, such as ^"^C-assimilation, has been used for species of green algae, and this test essentially measures the rate of photosynthesis. In addition, ^"^C-uptake is mainly used with phytoplankton in natural water and rarely used with cultured algae as the test organisms [152]. The duration of the ^"^C-assimilation test is a few hours, and compounds that are phytotoxic towards algae adversely affect the rates of ^^C-uptake. Although radiocarbon-uptake testing is reproducible, this toxicity testing method is less sensitive than growth inhibition toxicity testing [153], and, therefore, it is not applicable for a thorough, rapid screening of compounds to determine their toxicity towards algae and cyanobacteria. For growth inhibition bioassays using hquid media, large-scale laboratory and batch-culture toxicity [Algal Growth Inhibition Toxicity Physical/chemical (AGIT)] tests have been widely used [154]. experimental parameters cause variability between batch culture tests, and these parameters need to be understood and controlled. The most important experimental parameters include light, temperature, carbon dioxide availability, pH, and the test organisms used [155]. Continuous light is recommended over a lightidark cycle because it is more practical and most algae grow well in continuous light [155], Temperature needs to be maintained near the optimum for the particular test species and near uniform among batch-culture flasks [155]. Algal laboratory cultures can become carbon limited very quickly and carbon dioxide availability can be maintained by adding bicarbonate to the medium [155]. Also, buffered media will reduce any effect of pH on the toxicity of the test compound. Different groups of algae (cyanobacteria, diatoms, green algae) and different species within the same group of algae should be used in the bioassay to adequately determine the toxicity of the test compound and its potential usefulness as a selective algicide. The most commonly used freshwater green alga species used in toxicity tests is Selenastrum capricornutum [155]. Of course, the other species included in the bioassay must include the target species, i.e., noxious species to be

378 selectively eliminated from the natural aquatic ecosystem. Algal and cyanobacterial bioassays should be run for relatively short periods (i.e., no more than four days) before the growth of the control cultures becomes adversely affected by growth-limiting factors such as nutrient depletion or the build-up of toxic cataboHc metabolites excreted from the cells [156]. For growth bioassays using sohd media, modification of the agar diffusion assay used by Flores and Wolk [108] has been shown to provide reliable results [100]. Gross et al. [100] spotted methanol and ethanol extracts on 1% agar plates, dried the extracts in sterile air, and then overlaid the plate with a suspension of indicator cells in 1% agar. After several days of incubation, areas (zones) of no growth were measured and quantification of the inhibitory effects of the test extracts was estimated by comparison with/to the serial dilutions and controls. The most widely measured response variables in growth inhibition toxicity tests are specific growth rate and biomass. Specific growth rate is determined by calculating the linear regression of logarithmic biomass data from an exponential growth curve [157]. Biomass is usually determined by measuring either cell biomass (dry weight or cell volume counting using a electronic particle counter) or optical density (absorbance) using a spectrophotometer [155]. In certain types of toxicity testing [e.g., determination of EC50 (50% effective concentration) values], monitoring growth rate is better than biomass measurements [156,157]. The measurement of alkaline phosphatase activity (APA) of target phytoplankton is a recently developed bioassay that has been used to determine the algicidal effects of polyphenols from Eurasian watermilfoil {Myriophyllum spicatum) [80]. Phytoplankton produce extracellular enzymes, such as alkaUne phosphatase, to provide additional sources of nutrients. Fluorescence spectrometry is used to measure APA, with methylumbeliferyl-phosphate used as substrate and mixed with the algal or cyanobacterial suspension and the suspected inhibitor. Rapid bioassays using the response of cyanobacterial motility as an indication of the toxicity of a compound have been used [158]. However, this type of bioassay is impractical when using non-motile species of cyanobacteria and algae such as Anabaena spp. and Selenastrum capricornutum, respectively. An alternative to batch algal bioassays is the use of cage-culture turbidostats (CCT) [154]. The CCT is computer controlled and will monitor algal growth responses to the addition of test compounds. Although easy and quick to set up, the CCT limits the number of

379 concentrations of a particular compound that can be tested during an experimental run, thereby reducing data essential for statistical analysis. Microbiotests (microtiter plate bioassays) offer a much more rapid, simpler, and low cost alternative to batch-culture toxicity testing of compounds [147,159-162]. Microtiter plate bioassays require the use of continuous culture (chemostat) systems to provide a source of cells (inoculum) in continuous, steady-state growth. The continuous culture system maintains a particular growth rate of the culture by the addition of fresh medium at a set rate [163]. The volume in the continuous culture vessel is kept constant by an overflow apparatus so that the rate of fresh medium entering the vessel is equal to the removal rate of medium and cells. Cells growing at a constant rate should adsorb test compounds more consistently, thereby reducing the potential influence of certain test compounds (e.g., lipophihc organics) on distorting test results [155]. In addition, minor changes in the growth of the test organism toward a particular compound are easier to observe and analyze. Schrader et al. [147] developed a rapid, inexpensive bioassay to screen a large number of natural compounds for selectivity as cyanobactericides. In this bioassay, continuous culture systems provided a source of cells (inoculum), and a representative green alga, Selenastrum capricornutum, and the MIB-producing cyanobacterium Oscillatoria perornata were used. O. perornata is a filamentous plankton that has been attributed as being the major cause of musty off-flavor in farm-raised channel catfish in west Mississippi [164]. This organism was previously identified as Oscillatoria cf chalybea [165]. It does not conform very well to the classic description of that species [166] and, therefore, a more appropriate designation of this organism is Oscillatoria perornata f. attenuata [167]. The bioassay utilizes 96-well cell culture plates (microtiter plates). Test compounds at various concentrations and pre-dissolved in either water, ethanol, or methanol are placed in wells of the culture plate, followed by the addition of either unialgal or cyanobacterial culture to the wells. Absorbance (650 nm) is measured at 24-h intervals for four to five days using a microplate reader, and growth of the treated cultures (wells) relative to untreated cultures is used to determine the relative toxicity of the test compounds to the two microorganisms. The supplementation of culture inoculum placed in the microplate wells with sodium bicarbonate as a means to provide an additional source of carbon dioxide was found to be unnecessary. Pediastrum simplex and Anabaena sp. LP-691 are other

380 species of green algae and cyanobacteria, respectively, that have been used as test organisms with this bioassay [93]. One of the problems encountered in the bioassay developed by Schrader et al. [147] was the limitation in the type of compound amenable for the screening procedure. Only compounds that are soluble in water, ethanol, or methanol could be screened since these solvents will not react with the polystyrene in the microplates. However, the substitution of a quartz glass microplate for a polystyrene microplate allows solvents such as pentane, hexane, and acetone to be used in the bioassay. Therefore, natural compounds and crude extracts that may only be soluble in one or several of these types of solvents can now be screened. The potential change in toxicity of a compound towards a particular organism should be considered in the progression from a 96-well microplate (300 \iL capacity per well) to a large aquatic ecosystem, such as a lake or pond. Scale-up bioassays from 96-well to six-well microplates discovered a reduction in the toxicity of the natural compound ferulate towards O. perornata [168]. Other scale-up studies from microplate to flask bioassays have also found a reduction in phytotoxicity of the test compound [161]. In addition to small-scale laboratory studies, dose-response experiments as a prelude to full-scale lake or pond testing should be performed in an intermediate-scale system using a much larger volume (liters) of culture or natural water containing the target microorganism(s). Any pond or lake testing should include close monitoring of the persistence of the test compounds in the aquatic environment to help determine their effectiveness as a selective algicide and any potential threat that they might pose to the environment. Efficacy Testing of Natural Algicides in Freshwater Ecosystems Lead cyanobactericidal compounds from laboratory screening must undergo efficacy testing in appropriate freshwater habitats to ultimately determine their usefulness as selective algicides. In catfish aquaculture, efficacy testing using whole ponds has its drawbacks. For example, a large amount of lead compound is required when testing in several ponds. The cost of synthesis or direct purchase from an appropriate supplier can be high. Also, there is a need to test in several ponds at the same time to obtain a large number of replicate observations, i.e., to determine if any statistically significant treatment effect occurs. The use of several ponds

381 to obtain such information introduces high pond-to-pond variabihty of phytoplankton community structure with inconsistent estabhshment of target cyanobacterial populations. In addition, high pond-to-pond variability of pond water chemistry could affect the stoichiometry of the test compound after application. An alternative to fiill-scale lake or pond (real or artificial) studies is the use of "limnocorrals." Limnocorrals are enclosures that isolate a volume of water for study and are also referred to as experimental enclosures or in situ mesocosms. Their use provides an intermediate step between laboratory-scale studies, which are often too artificial, and full-scale lake or pond studies. In addition, a much smaller amount of test compound is required, and the initial environmental conditions are the same within all of the enclosures. Limnocorrals can be made of translucent fiberglass, polyvinyl chloride, or some other flexible, inert material. These enclosures can be open-bottomed so that they will sink into the bottom sediment to help form a watertight seal to separate the intemal environment fi*om the outside. In larger-scale (lake) studies using limnocorrals, divers have been used to help ensure that a proper seal in the sediment is obtained [169]. Vertical mixing of the water column inside small-scale limnocorrals is necessary in order to help achieve environmental conditions more similar to those in the lake or pond outside of the enclosure. One method used to achieve water mixing/aeration inside limnocorrals is by pumping air through submerged airstones placed within the limnocorrals. Some of the earliest studies using limnocorrals have been to determine the effects of mercury on algae [170], to study phytoplankton growth [171,172], and to study the fates of insecticides in aquatic ecosystems [173]. Other studies using limnocorrals have assessed the short-term impacts of various chemicals and pesticides on aquatic ecosystems [174177]. However, limnocorrals are not suitable for long-term impact studies due to deterioration of environmental conditions within the enclosures over longer periods of time (e.g., weeks). In addition, the intensive labor demands required in setting up the limnocorrals often limit the number of units that can be used for a study. Dose-response studies can be performed using as few as eight enclosures [178]. Although limnocorrals do not provide an exact duplication of the lake or pond environment, their use does permit an adequate determination of the impact of a test compound on the phytoplankton community.

382 CONCLUSION Although previous research concerning the use of natural compounds as selective cyanobactericides is not very extensive, past and present research indicate that such an approach has merit and should continue to be investigated. The vast majority of natural compounds in nature remain unscreened as to their potential for use as environmentally-safe, selective compounds to help control noxious species of cyanobacteria in freshwater ecosystems. The discovery of such compounds that are cost-effective would be very beneficial to industries such as catfish aquaculture and municipal drinking water systems by reducing their dependence on synthetic compounds (herbicides) to control cyanobacterial blooms. The use of synthetic herbicides carries with it environmental safety concerns and a negative perception by the public. ABBREVIATIONS APA OCT EC50

= = =

GS MIB PSI PSII SAR

= = = = =

Alkaline Phosphatase Activity Cage-culture Turbidostat 50% Effective Concentration (toxicant concentration affecting a specific response in 50% of exposed test organisms) Glutamate Synthetase 2-Methyhsobomeol Photosystem I Photosystem II Structure-activity Relationship

ACKNOWLEDGEMENTS We have been fortunate to work with Dr. Craig Tucker who has contributed greatly to current knowledge about water quality and phytoplankton communities in catfish aquaculture ponds. We thank two reviewers, Drs. Casey Grimm and Mario Tellez, for their many helpful suggestions during preparation of this chapter.

383

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15] [16] [17] [ 18] [19] [20] [21] [22] [23] [24] [25] [26]

Paerl, H.W.; Tucker, C.S.; / World Aquacult Soc, 1995, 26, 109-131. Codd, G.A.; Bell, S.G. In Proceedings of the 1998 WaterTECH Meeting, Brisbane, Australia; AWWA; Artarmon: New South Wales, Australia, 1998; pp. 1-14. Codd, G.A.; Water Sci. TechnoL, 1995, 32, 149-156. Hawkins, P.R.; Runnegar, M.T.C.; Jackson, A.R.B.; Falconer, I.R;. Appl. Environ. Microbiol., 1985, 50, 1292-1295. Pilotto, L.S.; Burch, M.D.; Douglas, R.M.; Cameron, S.; Roach, G.J.; Cowie, C.T.; Beers, M.; Robinson, P.; Kirk, M.; Hardiman, S.; Moore, C; Attewell, R.G.; Austral. New Zealand J. Public Health, 1997,21, 562-566. Pouria, S.; de Andrade, A.; Barbosa, J.; Cavalcanti, R.L.; Barreto, V.T.S.; Ward, C.J.; Preiser, W.; Poon, G.K.; Neild, G.H.; Codd, G.A.; Lancet, 1998, 352, 2126. Falconer, I.R.; Environ. Toxicol, 1999,14, 5-12. Steffensen, D.; Burch, M.; Nicholson, B.; Drikas, M.; Baker, P.; Environ. Toxicol., 1999,14, 183-195. Paerl, H. W. In Growth and Reproductive Strategies of Freshwater Phytoplankton; Sandgren, CD., Ed. Cambridge University: Cambridge, England, 1988; pp. 261-315. Puschner, B.; Galey, F.D.; Johnson, B.; Dickie, C.W.; Vondy, M.; Francis, T.; Holstege, D.M.; J. Am. Vet. Med Assoc, 1998, 213, 1605-1607. McBarron, E.J.; May, V.; Austral. Vet. J., 1966, 42, 449-453. McBarron, EJ.; Walker, R.I.; Gardner, I.; Walker, K.H.; Austral. Vet. J., 1975, 51, 587-588. Main, D.C.; Berry, P.H.; Peet, R.L.; Robertson, J.P.; Austral. Vet. J., 1977, 53, 578-581. Francis, G.; Nature, 1878,18, 11-12. DeVries, S.E.; Galey, F.D.; Namikoshi, M.; Woo, J.C; J. Vet. Diagn. Invest., 1993, 5, 403-408. Beasley, V.R.; Cook, W.O.; Dahlem, A.M.; Hooser, S.B.; Lovell, R.A.; Valentine, W.M.; Clin. Toxicol., 1989, 5, 345-361. Carmichael, W.W.; J. Appl. Bacteriol, 1992, 72, 445-459. Boyd, C.E.; Tucker, C.S.; Pond Aquaculture Water Quality Management, Kluwer Academic: Boston, 1998. English, W.R.; Schwedler, T.E.; Dyck, L.A.; J. Appl. Aquacult, 1993, 3, 195209. Persson, P.-E.; Water Res., 1980,14, 1113-1118. Persson, P.-E.; Aqua Fennica, 1979, 9, 48-52. Yurkowski, M.; Tabachek, J.L.; J. Fish. Res. Board Can., 1974, 31, 1851-1858. Johnsen, P.B.; Kelly, C.A.; J. Sens. Stud, 1990, 4, 189-199. Juttner, F.; Water Sci. Technol, 1995, 31(11), 69-78. Martin, J.F.; Suffet, I.H.; Water Sci. Technol, 1992, 25(2), 73-79. Jenkins, D.; Medsker, L.L.; Thomas, J.F.; Environ. Sci Technol, 1967,1, 731735.

384 [27] [28] [29] [30] [31] [32] [33] [34] [35] [36[ [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58]

Mohren, S.; Juttner, F.; Water Set TechnoL, 1983,15(6/7), 221-228. Slater, G.P.; Blok, V.C; Water Sci TechnoL, 1983, J5(6/7), 181-190. Johnsen, P.B.; Lloyd, S.W.; Can. J. Fish. Aquat. Sci, 1992, 49, 2406-2411. Johnsen, P.B.; Lloyd, S.W.; Vinyard, B.T.; Dionigi, C.P.; J. World Aquacult. ^oc, 1996, 27, 15-20. Izaguirre, G.; Hwang, C.J.; Krasner, S.W.; McGuire, M.J.; Appl. Environ. Microbiol, 1982, 43, 708-714. Izaguirre, G.; Water Sci. TechnoL, 1992, 25(2), 217-223. Seligman, K.; Enos, A.K.; Lai, H.H.; Water ScL TechnoL, 1992, 25(2), 19-25. Izaguirre, G.; Taylor, W.D.; Water Sci TechnoL, 1995, 31(1]), 41-48. Izaguirre, G.; Taylor, W.D.; Water Res., 1998, 32, 1673-1677. Izaguirre, G.; Taylor, W.D.; Pasek, J.; Water Sci TechnoL, 1999, 40(6), 85-90. Rosen, B.H.; MacLeod, B.W.; Simpson, M.R.; Water Sci TechnoL, 1992, 25(2), 185-190. Burlingame, G.A.; Dann, R.M.; Brock, G.L.; J. Am. Water Works Assoc, 1986, 78(3), 56-61. Hayes, K.P.; Burch, M.D.; Water Res., 1989, 23, 115-121. Bowmer, K.H.; Padovan, A.; Oliver, R.L.; Korth, W.; Ganf, G.G.; Water Sci TechnoL, 1992, 25(2), 259-267. Anderson, B.C.; Quartermaine, L.-K.; J. Great Lakes Res., 1998, 24(4), 859867. Negoro, T.; Ando, M.; Ichikawa, N.; Water Sci TechnoL, 1988, 20(8/9), 117123. Hoson, T.; Water Sci TechnoL, 1992, 25(2), 177-184. Hosaka, M.; Murata, K.; likura, Y.; Oshimi, A.; Udagawa, T.; Water Sci TechnoL, 1995, 31(11), 29-34. Hoson, T.; Hayashi, S.; Hattori, K.; Water Sci TechnoL, 1995, 31(11), 139-144. Kajino, M.; Sakamoto, K.; Water Sci TechnoL, 1995, 31(11), 153-158. Sugiura, N.; Iwami, N.; Inamori, Y.; Nishimura, O.; Sudo, R.; Water Res., 1998, 32, 3549-3554. Berglind, L.; Johnsen, I.J.; Ormerod, K.; Skulberg, O.M.; Water Sci. TechnoL, 1983,15(6/7), 241-246. Suffet, I.H., Corado, A.; Chou, D.; Butterworth, S.; McGuire, M.J.; J. Am. Water Works Assoc, 1996, 88(4), 168-180. Jardine, C.G.; Gibson, N.; Hrudey, S.E.; Water Sci TechnoL, 1999, 40(6), 9198. McGuire, M.J.; Water Sci TechnoL, 1995, 31(11), 1-8. McGuire, M.L; Gaston, J.M.; Water Sci TechnoL, 1988, 20(8/9), 215-228. Chen, G.; Dussert, B.W.; Suffet, I.H.; Water Res., 1997, 31, 1155-1163. Kilham, P.; Hecky, R.E.; LimnoL Oceanogr., 1988, 33, 776-795. McGuire, M.J.; Water Sci TechnoL, 1999, 40(6), 153-163. Tucker, C.S.; Rev. Fish. Sci., 1996, 4(1), 1-55. Engle, C.R.; Pounds, G.L.; van der Ploeg, M.; J. World Aquacult. Soc, 1995, 26, 297-306. Kinnucan, H.; Sindelar, S.; Wineholt, D., Hatch, U.; South. J. Agric Econ., 1988,20,81-91.

385

[59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71 ] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82]

[83] [84]

Coats, W.A.; Dillard, J.G.; Waldrop, J.E.; Mississippi Agricultural and Forestry Experiment Station, Agricultural Economics Research Report No. 184,1989; Mississippi State University, Mississippi, USA. Tucker, C.S.; Rev. Fish. Set, 2000, 8(1), 45-88. Schrader, K.K.; de Regt, M.Q.; Tidwell, P.D.; Tucker, C.S.; Duke, S.O.; Aquaculture, 1998,163, 85-99. Twiss, M.R.; Welboum, P.M.; Schwartzel, E.; Can. J. Bot., 1993, 77, 333-338. Price, N.M.; Morel, F.M.M.; Arch. Hydrobiol, 1994, 42, 79-97. Masuda, K.; Boyd, C.E.; Aquaculture, 1993, 777, 287-302. Tucker, C.S.; Catfish J., 1998, Xni(3), 5. Boyd, C.E.; Hydrobiologia, 1973, 41, 357-390. Gibson, M.T.; Welch, I.M.; Barret, P.R.F.; Ridge, I.; J. AppliedPhycol, 1990, 2, 241-248. Welch, I.M.; Barrett, M.T.; Gibson, M.T.; Ridge, I.; J. Appl Phycol, 1990, 2, 231-239. Pillinger, J.M.; Cooper, J.A.; Ridge, I.; J. Chem. Ecol, 1994, 20, 1557-1569. Harriman, R.; Adamson, E.A.; Shelton, R.G.J.; Moffet, G.; Biocontrol Sci. Technoi, 1997, 7, 2S1-296. Ridge, I. and Barret, P.R.F.; Aspects ofAppl. Biol., 1992, 29, 457-562. Newman, J.R.; Barrett, P.R.F.; J. Aquat. Plant Manage., 1993, 31, 203-206. Kivaisi, A.K.; Op den Camp, H.J.M.; Lubberding, H.J.; Boon, J.J.; Vogels, G.D.; Appl. Microbiol Biotechnol, 1990, 33, 93-98. Everall, N.C.; Lees, D.R.; Water Res., 1996, 30, 269-276. Barret, P.R.F.; Cumow, J.C; Littlejohn, J.W.; Hydrobiologia, 1996, 340, 307311. Chung, K.T.; Zhao, G.; Stevens, S.E., Jr.; J. Aquat. Animal Health, 1995, 7, 341344. Zhao, G.; Watson, J.; Crowder, C; Stevens, S.E., Jr.; J. Appl. Phycol, 1998,10, 1-7. Ayoub, S.M.H.; J. Chem. Tech. Biotechnol, 1982, 32, 728-734. Ayoub, S.M.H; Yankov, L.K.; Fitoterapia, 1985, 56, 227-229. Taylor, B.R.; Goudey, J.S.; Carmichael, N.B.; Environ. Toxicol Chem., 1996, 15, 150-159. Aliotta, G.; Monaco, P.; Pinto, G.; Pollio, A.; Previtera, L.; J. Chem. Ecol, 1991,17,2123-113 A. Huang, S.; Wang, W.; Ma, K.; Zhou, Y.-Z.X.; Wu, H.; Sun, W.-H.; Yang, S.-Y.; Huang, H.; Huang, Z.-A.; Wu, H.; Yu, S.-W. In Recent Advances in Allelopathy. Vol. I. A Science for the Future. Macias, F.A.; Galindo, J.C.G.; Molinillo, J.M.G.; Cutler, H.G., Eds.; Univ. Cadiz Press, 1999; pp. 197-204. Gross, E.M. In Principles and Practices in Plant Ecology: Allelochemical Interactions', Dakshini, K.M.M.; Fay, C.L., Eds.; CRC Press: Boca Raton, Florida, 1999; pp. 179-199. Aliotta, G.; Molinaro, A.; Monaco, P.; Pinto, G.; Previtera, L.; Phytochemistry, 1992,57,109-111.

386 [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111]

Nakai, S.; Inoue, Y.; Hosomi, M.; Abstract from Second World Congress on Allelopathy. Critical Analysis and Future Prospects. Lakehead University, Canada, Aug. 8-13, 1999; p. 137. Gross, E.M.; Meyer, H,; Schilling, G.; Phytochemistry, 1996, 41, 133-138. Saito, K.; Matsumoto, M.; Sekine, T.; Murakoshi, I.; Morisaki, N.; Iwasaki, S.; J. Nat. Prod., 1989, 52, 1221-1226. Van Aller, R.T.; Pessoney, G.F.; Rogers, V.A.; Watkins, E.J.; Leggett, H.G.; ACSSymp. Ser., 1985, 268, 387-400. Aliotta, G.; Delia Greca, M.; Monaco, P.; Pinto, G.; Pollio, A.; Previtera, L.; / Chem. EcoL, 1990,16, 2637-2646. Van Aller, R.T.; Pessoney, G.F.; Aquacult Mag., 1982, 8, 18-22. Tucker, C.S.; Lloyd, S.W.; Aquaculture, 1987, 65, 141-148. Delia Greca, M.; Monaco, P.; Previtera, L.; Alliotta, G.; Pinto, G.; Pollio, A.; Phytochemistry, 1989, 28, 2319-2321. Schrader , K.K.; de Regt, M.Q.; Tidwell, P.R.; Tucker, C.S.; Duke, S.O.; Bull Environ. Contam. Toxicol., 1998, 60, 651-658. Mason, C.P.; Edwards, K.R.; Carlson, R.E.; Pignatello, J.; Gleason, F.K.; Wood, J.M.; Science, 1982, 215, 400-402. Pignatello, J.J.; Porwoll, J.; Carlson, R.E.; Xavier, A.; Gleason, F.K.; Wood, J.M.;/ Org. Chem., 1983, 48, 4035-4038. Moore, R.E.; Cheuk, C; Patterson, G.M.L.; J. Am. Chem. Soc, 1984, 106, 64566457. Smitka, T.A.; Bonjouklian, R.; Doolin, L.; Jones, N.D.; Deeter, J.B.; / Org. Chem., 1992, 57, %51-%6\. Hagmann, L.; Jlittner, F.; Tetrahedron Lett., 1996, 36, 6539-6542. Papke U.; Gross, E.M.; Francke, W.; Tetrahedron, 1997, 38, 379-382. Gross, E.M.; Wolk, C.P.; Juttner, F.; J. PhycoL, 1991, 27, 686-692. Srivastava, A.; Juttner, F.; Strasser, R.J.; Biochim. Biophys. Acta, 1998, 1364, 326-336. Bagchi, S.N.; Chauhan, V.S.; Marwah, J.B.; Curr. Microbiol., 1993, 26, 223228. Chauhan, V.S.; Marwah, J.B.; Bagchi, S.N.; New Phytol, 1992, 120, 251-257. Dakhama, A.; de la Noue, J.; Lavoie, M.C.; J. Applied PhycoL, 1993, 5, 297306. Matsuhashi, M.; Furuyama, M.; Maruo, B.; Agr. Biol. Chem., 1969, 33, 17581760. Holten Liitzhoft, H.-C; Halling-Sorensen, B.; Jorgensen, S.E.; Arch. Environ. Contam. Toxicol., 1999, 36, 1-6. Kaya, K.; Sano, T.; Phycologia, 1996, 35, 117-119. Flores, E.; Wolk, P.; Arch. Microbiol., 1986,145, 215. Schelegel, I.; Doan, N.T.; de Chazal, N.; Smith, G.D.; J. Appl. PhycoL, 1999, 10,471-419. Sigee, D.C.; Glenn, R.; Andrews, M.J.; Bellinger, E.G.; Butler, R.D.; Epton, H.A.S.; Hendry, R.D.; Hydrobiologia, 1999, 396, 161-172. Duke, S.O.; Environ. Health Perspect., 1990, 87, 263-271.

387

[112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136]

Devine, M.D.; Duke, S.O.; Fedtke, C; Physiology of Herbicide Action, Prentice Hall: Englewood Cliffs, New Jersey, 1993. Cobb, A.; Herbicides and Plant Physiology, Chapman and Hall: London, 1992. Roe, R.M.; Burton, J.D.; Kuhr, R.J., Eds.; Herbicide Activity: Toxicology, Biochemistry and Molecular Biology, lOS Press: Amsterdam, 1997. Gilchrist, D.G. In Toxins and Plant Pathogenesis; Daly, J.M.; Deverall, B.J., Eds.; Academic Press: New York, 1983; pp. 81-136. Duke, S.O.; Abbas, H.K.; Amagasa, T.; Tanaka, T. In Crop Protection Agents from Nature. Natural Products and Analogues; Copping, L.G., Ed.; Royal Chemical Society: Cambridge, 1996; pp. 82-113. Boger, P. In Target Assays for Modern Herbicides and Related Phytotoxic Compounds; Boger, P.; Sandmann, G., Eds.; Lewis Publishers: Boca Raton, Florida, 1993; pp. 83-91. Gleason, F.K.; Paulson, J.L.; Arch. Microbiol, 1984,138, 273-277. Dayan, F.E.; Romagni, J.G.; Duke, S.O.; J. Chem. EcoL, 2000, (In press). Duke, S.O.; Dayan, F.E.; Hernandez, A.; Duke, M.V.; Abbas, H.K.; Brighton Crop Protection Conference, Weeds, 1997, 1, 83-92. Dayan, F.E.; Romagni, J.G.; Tellez, M.R.; Rimando, A.M.; Duke, S.O.; Pestic. Outlook, 1999, JO, \S5ASS. Moreland, DE. In Weed Physiology, Vol. II Herbicide Physiology; Duke, S.O., Ed.; CRC Press: Boca Raton, Florida, 1985; pp. 37-61. Czamota, M.A.; Weston, L.A.; Dayan, F.E.; WeedSci. Soc. Amer. Abstr., 1998, 38, 52. Duke, S.O.; Dayan, F.E.; RIKENRev., 1999, 21, 9-10. Hejl, A.M.; Einhellig, F.A.; Rasmussen, J.A.; 7. Chem. EcoL, 1993, J9, 559568. Koeppe, D.E.; Physiol. Plant., 1972, 27, 89-94. Schrader, K.K.; Dayan, F.E.; Allen, S.N; de Regt, M.Q.; Tucker, C.S.; Paul, R.N.; Internal J. Plant Sci., 2000, (In press). De Prado, R.; Jorrin, J.; Garcia-Torres, L., Eds.; Weed and Crop Resistance to Herbicides, Kluwer: Amsterdam, 1997. Gronwald, J. In Herbicide Resistance in Plants; Powles, S.B.; Holtum, J.A.M., Eds.; CRC Press: Boca Raton, Florida, 1994; pp. 27-60. Dayan, F.E.; Hernandez, A.; Allen, S.N.; Moraes, R.M.; Vroman, J.A.; Avery, M.A.; Duke, S.O.; Phytochemistry, 1999, 50, 607-614. Galindo, J.C.G.; Hernandez, A.; Dayan, F.E.; Tellez, M.R.; Macias, F.A.; Paul, R.N.; Duke, S.O.; Phytochemistry, 1999, 52, 805-813. Duke, S.O.; Paul, R.N.; Lee, S.M.; Amer. Chem. Symp. Soc. Ser., 1988, 380, 318-334. Knox, J.P.; Dodge, A.D.; Plant Cell Environ., 1985, 8, 19-25. Brody, S.S.; Papageorgiou, G.; Alygizaki-Zorba, K.Z;. Naturforsch., 1997, 52c, 165-168. Duke, S.O.; Rebeiz, C.A., Eds.; Porphyric Pesticides, ACS Symp. Ser. 559, American Chemical Society: Washington, DC, 1994. Braun, A.; Phytopathology, 1955, 45, 659-664.

388 [137]

Lydon, J.; Duke, S.O. In Plant Amino Acids; Singh, B.K., Ed.; Marcel Dekker: New York, 1999; pp. 445-464. [138] Wetzel, R.G.; Verk Int. Ver. Theor. Angew. LimnoL, 1993, 25, 122-128. [139] Sherman, T.D.; Vaughn, K.C.; Duke, S.O.; Phytochemistry, 1991, 30, 24992506. [ 140] Kida, T. In Novel Microbial Products for Medicine and Agriculture; Demain, A.; Somkuti, G.; Hunter-Cevera, J.; Rossmoore, H., Eds.; Elsevier Science B.V.: Amsterdam, 1989; pp. 195-202. [141] Kida, T; Shibai, H.; Agric. Biol Chem., 1986, 50, 483-484. [142] Gleason, F.K.; Baxa, C.A.; FEMSMicrobiol. Lett., 1986, 33, 85-88. [143] Rimando, A.M.; Dayan, F.E.; Czamota, M.A.; Weston, L.A.; Duke, S.O.; / Nat. Prod, 199S, 61,921-930. [144] Duke, S.O.; Dayan, F.E.; Romagni, J.G.; Rimando, A.M.; Weed Res., 2000, 40, 99-111. [145] Gleason, F.K.; Thoma, W.J.; Carlson, J.L. In Progress in Photosynthesis Research; Biggens, J., Ed.; Martinus Nijhoff Publishers: Dordrecht, 1987; pp. 763-766. [146] Nyholm, N.; Sorensen, P.S.; Kusk, K.O.; Christensen, E.R.; Environ. Toxicol. Chem., 1992,11, 157-167. [147] Schrader, K.K.; de Regt, M.Q.; Tucker, C.S.; Duke, S.O.; Weed Technol., 1997, 11,161-11 A. [148] Cairns, Jr., J.; Hydrobiologia, 1983, 100, 47-57. [149] Campbell, D.; Hurry, V.; Clarke, A.K., Gustafsson, P.; Oquist, G.; Microbiol. Mol Biol. Rev., 1998, 62, 667-683. [150] van der Heever, J.A.; Grobbelaar, J.U.; Arch. Environ. Contam. Toxicol., 1998, 55,281-286. [151] Fay, P.; Microbiol. Rev., 1992, 56, 340-373. [152] van der Heever, J.A.; Grobbelaar, J.U.; J. Appl. Phycol, 1996, 8, 65-71. [153] Nyholm, N.; Damgaard, B.M.; Chemosphere, 1990, 21, 671-679. [154] Clarkson, N.; Leftley, J.W.; Meldrum, D.T., Watson, J.W.; Water Res., 1998, 32, 1162-1168. [155] Nyholm, N.; Kallqvist, T.; Environ. Toxicol. Chem., 1989, 8, 689-703. [156] Nyholm, N.; Water Res., 1985,19, 273-279. [157] Nyholm, N.; Arch. Environ. Contam. Toxicol, 1990, 19, 518-522. [158] Zullei, N.; Benecke, G.; Bull Environ. Contam. Toxicol, 1978, 20, 786-792. [159] Blaise, C; Legault, R,; Bermingham, N.; van Coillie, R.; Vasseur, P.; Tox. Assess., 19^6,1,26\'2%\. [160] Blaise, C; Environ. Toxicol Water Qual, 1991, 6, 145-155. [161] St. Laurent, D.; Blaise, C; MacQuarrie, P.; Scroggins, R.; Trottier, B.; Environ. Toxicol Water Qual, 1992, 7, 35-48. [162] Cerejeira, M.J.; Pereira, T.; Silva-Femandes, A.; Chemosphere, 1998, 37, 29492955. [163] Tempest, D.W. In Methods in Microbiology; Norris, J.R.; Ribbons, D.W., Eds.; Academic: London, 1970; Vol. 2, pp. 259-276. [164] van der Ploeg, M.; Tucker, C.S.; J. Appl Aquacult., 1993, 3, 121-140. [165] Martin, J.F.; Izaguirre, G.; Waterstrat, P.; Water Res., 1991, 25, 1447-1451.

389 [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178]

Gomont, M.; Ann. Sci. Nat Bot. VIL, 1892, 75, 263-368 [Reprinted in 1962 by Wheldon & Wesley, Ltd. and Hafner Publishing, New York]. Skuja, H.; Nova Acta Reg. Soc. Sci. Ups., Series 4,1949, J4(5), 1-188. Schrader, K.K.; Rimando, A.M.; Tucker, C.S.; Duke, S.O.; Pestic. Sci, 1999, 55, 726-732. Liber, K.; Solomon, K.R.; Carey, J.H.; Environ. Toxicol. Chem., 1997,16, 293305. Blinn, D.W.; Tompkins, T.; Zaleski, L.; J. Phycol, 1911,13, 58-61. Lund, J.W.G.; Verh. Int. Ver. Theor. Angew. Limnol, 1972, 18, 71-77. Lack, T.J.; Lund, J.W.G.; Freshwater Biol, 1974, 4, 399-415. Solomon, K.R.; Smith, K.; Guest, G.; Yoo, J.Y.; Kaushik, N.K.; Can. Tech. Rep. Fish. Aquat Sci., 1980, 975, 1-9. Stephenson, R.R.; Kane, D.F.; Arch. Environ. Contam. Toxicol, 1984, 13, 313326. Solomon, K.R.; Yoo, J.Y.; Lean, D.; Kaushik, N.K.; Day, K.E.; Stephenson, G.L.; J. Fish. Aquat. Sci, 1985, 42, 70-76. Solomon, K.R.; Bowhey, C.S.; Liber, K.; Stephenson, G.R.; J. Agric. Food Chem., 19SS, 36, 1314-1318. Solomon, K.R.; Stephenson, G.L.; Kaushik,; N.K. Environ. Toxicol Chem., 1989, 8, 659-669. Liber, K.; Kaushik, N.K.; Solomon, K.R.; Carey, J.H.; Environ. Toxicol Chem., 1992,77,61-77.

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

391

PHARMACOLOGICAL ACTIVITIES AND APPLICATIONS OF SALVIA SCLAREA AND SALVIA DESOLEANA ESSENTIAL OILS *ALESSANDRA T. PEANA, MARIO D.L. MORETTI Dipartimento di Scienze del Farmaco, Universita degli Studi di Sassari 1-07100 Sassari, Italy ABSTRACT: Pharmacological properties of S. sclarea and S. desoleana oils are discussed in relation to their chemical composition. After systemic administration, these oils had a depressant action on the CNS in mice and a hydrocholeretic effect in rats. Other studies demonstrated also a good anti-inflammatory activity in rats as well as a peripheral analgesic action in mice. These essential oils possess in vitro antimicrobial properties against some human pathogen strains and their activity is comparable to S. officinalis oil, well known for its antiseptic properties. In vitro studies carried out on mucoadhesive preparations showed the ability of the oil components to permeate the oral mucous. This could be of interest in the treatment of human inflammatory diseases of mucous tissues, frequently associated with microbial infections. These oils are also able to inhibit the growth of some phytopathogenic fungi and could therefore be useful in the agronomic field as an alternative to synthetic compounds, with a view to reducing environmental pollution. Some biological effects were correlated with the chemical composition and the kind of the formulations utilized in order to examine some possible applications of these oils in human medicine. All pharmacological activities seem to be attributable to the content of some oxygenated compounds, like alcohols (mainly linalool and alphaterpineol) and esters (linalyl and alpha-terpinyl acetate). Experimental observations point the hypothesis of a synergic action between the different components, even if the oils in toto were more active than their fractions or single components.

INTRODUCTION A remarkable variety of aromatic plants has long been used in folk medicine, cosmetics, and the food industry. Most of the properties of aromatic species can be attributed to their essential oil content, that is the volatile component of the secondary metabolism of the plant, whose biological function is still not completely understood. It accumulates in the secretory gland system and is released into the environment after various kinds of stimulation, such as a change in the weather (rain, wind, a rise in temperature), or the intervention of animal species like pollinator insects, etc.

392 Several studies analyzing essential oils and describing new techniques to optimize yield while retaining the characteristics of the oil during extraction have recently appeared in the literature [1-5]. When aromatic plants and essential oils are used as additives to food, besides giving a particular aroma and taste, they aid preservation since it is likely that they inhibit the proliferation of microorganisms affecting the organoleptic quality of the food. They are also used in phytotherapy to relieve symptoms and cure a variety of ailments, both acute and chronic. In alternative medicine, particularly aromatherapy, numerous preparations are based on aromatic plants and their essential oils. Many claims are made in non-specialist magazines and on television about the healing properties of the plants, but these are not supported by data from a sufficient number of specific studies. Only in a few cases, has research been carried out according to clear protocols, with results in line with the authors' objectives. Different research groups have studied various species of aromatic plants to assess their use in different sectors, with the further aim of creating a source of income for the local population by utilizing the most promising native species of plants [6-9]. Moreover, several studies have been initiated to discover why some species are used in folk medicine and examine the possibility of using the essential oils and their fractions in new, potentially profitable sectors. SALVIA SCLAREA L. General and Botanical description This plant (named also clary sage) is a member of the Labiatae (lipped flowers) or more correctly the Lamiaceae family, which gets its name from Lamium, one of the genera of this family. Lamiaceae (and Lamium) are thought to have been given the name because it is akin to Lamia, a man-devouring monster with the head and breast of a woman and the body of a serpent. According to Willis [10], there are about 700 members of the Salvia genus that can be found in tropical and temperate parts of the world. According to Briquet [11] the genus Salvia, which contains both Old world and New world species, is subdivided into subgenera and sections [8]. S. sclarea (2n=22) grows wild in countries around the Mediterranean (Southern Europe, North Africa and Western Asia). It is also wide-

393 spread from Iran to the Transcaucasican region [8]. In Italy it is found mainly in Sardinia and Sicily [12]. S. sclarea is a perennial herbaceous plant, very occasionally biennial, rhizomatous, with typical quadrangular stems and opposite leaves. It grows to a height of about 100 cm, although in very long photoperiods and dry summers it can reach more than 150 cm. Although the steams are erect, they are densely haired and much branched. The leaves are simple petiolate, broadly ovate-cordate and pubescent, possessing numerous glands. The inflorescence comprises verticillasters generally possessing 4-6 flowers with the bracts exceeding the 20-30 corolla which can be purple, lilac or white in color with the upper lip strongly falcate. The calyces are about 10 mm with spinase pubescent teeth which are glandular peltate. The 2-3 mm long nutlets are dark brown. The plant flowers between mid-May and August. It blooms from the second year and the flowers are protandrous. It grows in sunny dry coastal and submontane areas in stony or loose soil, usually calcareous, and along dry slopes, particularly in rocky places [12-14]. Commercial production of 5. sclarea crops has greatly increased around the world, mainly in some Soviet countries (Russia, the Crimean region of the Ukraine, Georgia and Uzbekistan), the USA, Bulgaria, France, China, Hungary and India [8]. The reason for this expansion is related to production of the essential oil, which is still used as a major ingredient in perfumes because of its good fixative properties and characteristic pleasant, intense smell of ambergris. The essential oil is also used in flavoring foodstuffs, particularly confectionery, giving a pineapple taste, besides being added to liqueurs and aromatic wines, not only for flavor but also for its eupeptic and digestive properties [15-16]. Phytochemical investigations From the economic and pharmaceutical point of view the main components of S. sclarea are the essential oil and a diterpene alcohol named sclareol. The former was first produced by Schimmel [17] and is usually obtained by steam distillation of fresh plant material. The latter is usually obtained by counter current extraction of spent steamed plant material using hexane as the solvent. After concentration by solvent evaporation a "partial concrete" is obtained and sclareol is selectively removed by a second counter current extraction with a polar solvent. In the USA, most of the S. sclarea is grown for the production of this non -volatile compo-

394 nent while the oil is produced as a by-product [18]. Sclareol is mainly utilized in the production of sclareolide, a diterpene lactone obtained by a two-stage oxidation process of sclareol [19] or by its fermentation with various soil-borne microorganisms [20-22]. Sclareolide and its precursor sclareol are of value as a tobacco flavor additive and fragrance fixative. The former is also the precursor of norambrienolide ether (know as "ambrox" or "ambroxan"), an extremely valuable amber-like odored compound which has a low threshold as well as fixative qualities [23]. Oil yield and composition Many workers have determined that the maximum oil content is obtained when all the flowers have reached full maturity because the calyces contain by far the most essential oil glands per unit area [24-28]. In contrast to some other oil rich Lamiaceae species, the leaves of the plant in question are fairly poor in oil. Table 1. Yield date w/v on a fresh weight basis Part of plant:

Yield %

stems

0.002 - 0.06

leaves

0.002-0.15

flowers spikes

0.11 -0.51

calyces

0.50-1.20

new buds 1 whole above ground plants

0.04 - 0.09 0.06 - 0.30

Like many aromatic plants, a diurnal fluctuation in oil content has been found in 5. sclarea with the maximum reached during the early morning. The most important contribution to an understanding of the composition of the oil came from the studies of Teisseire et al [29-31]. Further information on the oil composition can be found in the publications of Bulgarian researchers [32-33] and, more recently, Lawrence and Moretti [3435]. Comparative studies on the composition of the oil have come from the major commercial sources (USA, France, Italy and Russia); the main

395 differences between these oils are in the linalool, linalyl acetate and alpha terpinyl acetate contents. Table 2. Comparative chemical composition of commercial S. sclarea oil Compounds:

USA

France

Italy

Russia

alpha-pinene

0.1 -0.2

0.1 -0.3

trace - 0.9

0.2-0.3

myrcene

1.2-1.7

0.1 -0.2

0.1 -0.1.6

0.3-0.5

limonene

0.4-0.8

0.1-0.2

0.1 -0.9

0.1-0.2

(Z)- beta-ocimene

0.4-0.7

trace

trace

trace - 0.2

(E)- beta-ocimene

0.4-1.4

0.1-0.2

0.1-0.3

0.1-0.4

44.9-53.4

49.0-73.6

23.3-29.0

45.3-61.8

0.9-1.3

1.4- 1.6

0.2 - .03

1.1-1.8 10.4-19.3

linalyl acetate beta-caryophyllene linalool

20.3 - 28.6

9.0- 16.0

10.6- 14.6

alpha-terpineol

1.0-3.0

0.2-0.6

0 . 1 - 1.12

1.2- 2.5

neryl acetate

1.0- 1.7

0.2-0.3

1.2-2.0

0.4 - 0.6

geranyl acetate

1.9-3.2

0.3-0.5

1.2-3.2

0.8-1.2

nerol

0.6-1.1

trace - 0.1

0.9 - 2.7

0.3 - 0.5

geraniol

1.7- 3.3

0.1 - 0.3

2 . 5 - 3.1

0.6- 1.2

germacrene D

2.6-3.5

1.6-2.0

0.6- 1.4

0.7-2

caryophyllene oxide

0.2-0.3

0.3 - 0.5

trace

0.5

sclareol

0.2-0.4

0.1 -0.2

-

0.1-0.2

n.r.

n.r

30.7-38.5

n.r.

alpha-terpinyl acetate

Differences in linalool and linalyl acetate content were ascribed to the state of the plant material on distillation. In the USA and Italy, the plant material is freshly harvested, chopped and distilled within a few hours, whereas in France and the Russian countries, the harvested material is airdried prior to distillation. The excess moisture which is associated with processing of fresh plant material could lead to the major constituent, linalyl acetate, hydrolyzing to linalool, probably facilitated by an increased pH of the medium during distillation [8]. The presence of a large amount of alpha-terpinyl acetate in the essential oil extracted from 5. sclarea plants of Italian origin was reported [6]. As can be seen in the following table, the composition of an essential oil obtained from S. sclarea plants growing wild in a typical calcareous soil [35], contained high levels of methyl chavicol.

396

Table 3. Yield and composition of 5. sclarea oil obtained from Sardinian plants Compounds:

Cultivated

Wild

alpha-pinene

0.1

0.2

j alpha-thujene

n.d.

<0.1

camphene

n.d.

n.d.

beta-pinene

0.4

0.2

sabinene

0.2

0.2

alpha-phellandrene

0.6

n.d.

myrcene

0.1

l.O

limonene

1.0

0.2

1.8 cineole

1.5

3.2

(Z)- beta-ocimene

n.d.

0.8

gamma-teqjinene

0.9

<0.1

(E)- beta-ocimene

n.d.

0.1

para-cymene

0.1

<0.1

cis-linalool oxide (furanoid)

0.5

0.1

0.4

0.1

12.7

19.2

beta-caryophyllene

2.9

0.2

terpinen-4-ol

n.d.

0.3

alpha-terpinyl acetate

22.1

4.3

gamma-terpinyl acetate

n.d.

0.2 9.9

trans-linalool oxide (furanoid) ! linalyl acetate

linalool

2.6

alpha-terpineol

47.4

7.5

methyl chavicol

0.1

49.0

neryl acetate

2.1

0.4

geranyl acetate

1.3

0.2

nerol

0.2

<0.1

geraniol

0.6

<0.1

delta cadinol

n.d.

n.d.

methyl eugenol

n.d.

2.0

germacrene D

1.6

0.2

alpha-eudesmol

n.d.

0.2

1 beta-eudesmol

n.d.

03

1 Oil yield (ml/100 g)

1.9

1.5

j

397 This compound has never before been found in S. sclarea oil [6, 8, 14]. Further investigations must be carried out to determine whether the plant in question is a particular chemotype of S. sclarea rich in methyl chavicol, or a particular variety of the species. In relation to these data, Lawrence [9] pointed out that, unlike most Lamiaceae oils, to date no infraspecific chemical differences for S. sclarea oil have been found. The chemical composition of the S. sclarea oil was determined by various authors by GC, GC-MS and ^^C-NMR analyses [8, 34-37]. Studies to date have shown more than 120 compounds in oils from different countries. The main components are linalool and linalyl acetate representing a whole range of more than 50% of the oil in toto [34]. Numerous constituents of the oil are present in very low concentrations. The presence of abnormally high levels of epoxides, like trans and cis linalool epoxide and caryophyllene oxide, indicates that the oil has suffered some oxidation processes during distillation or storage [8]. As regards the content of monoterpene hydrocarbons, significant differences have been found, particularly for (Z)- and (E)- beta-ocimene which are present in much higher concentrations in the oil from Israeli [8]. Sciareol yield It has been determined [24-26] that the calyces are the plant part richest in sciareol and the epimer 13-episclareol. Table 4. Sciareol contents of the various plant parts Parts of plant:

Percentage of sciareol

roots

none

leaves

<0.1

flowering stem

0.4

corolla

0.2

bracts

0.4

immature calyces

0.4 - 0.7

partially mature calyces

0.7- 1.1

fully mature calyces

0.9- 1.7

fully mature tiller calyces

0.9- 1.7

398 The highest yield data of the oil and of the sclareol reveal that to maximize yield only the fully mature flower spikes should be harvested. Empirical uses In the early herbal literature, S. sclarea is often called "clear eye" or "eye bright" [8, 38-39], because the juice of the fresh plant was used to treat eye diseases. More recently, in the 19^^ century it was used to flavor wines and beers in Germany [40]. In fact, S. sclarea is still used in muscatel wine. Infusions or decoctions of the plant were and still are empirically used in baths to treat nervous polyarthritis and acute rheumatisms [8]. Moreover, aqueous extracts were used to treat various digestive disorders (eupeptic, aperitif, digestive, carminative, antispasmodic and antimeteoric). They also act as an antiperspirant, and anti-catarrhal substance. In the form of a throatwash S. sclarea is used as an antiseptic-disinfectant and lenitive-anti-inflammatory agent in oral cavity infections. The plant has also been empirically used to treat symptoms of some CNS disorders [41-42]. On account of its emmenagogic properties (as an estrogen-like substance), 5. sclarea was used in amenorrhea and in dysmenorrhea to relieve premenstrual pain [43]. Hypoglycemic and antiglycosuric properties were also attributed to S. sclarea. More recently, S. sclarea has also been used to produce a flavor system that would allow reduced dependence on tobacco, having many of the desirable flavor characteristics associated with Oriental tobacco, an integral part of the American blend cigarettes [8]. In Sardinian folk medicine 5. sclarea decoctions were also given as a "tonic-stimulant" in convalescence and in cases of asthenia, stress, neurasthenia and hysterical conditions. To treat a sore throat during a cold, the leaves were dipped in very hot olive oil, wrapped in pieces of paper and applied to the neck in a silk scarf, after the neck had been gently massaged. Pharmacology Activity on the central nervous system (CNS)

Various empirical uses of S. sclarea extracts suggest the likelihood that some components of the plant could interact with the complex structure of

399 the CNS. Recent studies carried out on experimental animals treated with S, sclarea essential oil showed the oil produces some interesting effects on the CNS that could justify the empirical use of S. sclarea in the treatment of the symptoms of some disorders like paroxysms, epilepsy, swooning, asthenia, stress, neurasthenia and hysterical conditions [42]. Atanossova-Shopova and Roussinov [41] as well as Peana [44] reported that the essential oil of S. sclarea has, on the whole, a depressant effect on the CNS related to the dosage. After systemic acute administration (50400 mg/kg), the oil produced in the experimental animals some changes in behavior, a reduction in spontaneous motor activity and potentiation of the narcotic effect of hexobarbital as well as of chloral hydrate and alcohol. The oil also had an anticonvulsive action upon electroconvulsions (maximal electroshock seizure) but had no effect upon pentamethylenetetrazole and strychnine convulsions. On the other hand, the oil was able to protect the experimental animals against lethal doses of the same convulsive agents. The effect on amphetamine- or caffeine- stimulated motor activity was not marked. Moreover, repeated systemic treatments with lower doses of the oil (50-100 mg/kg) were followed by death, preceded by a hypothermic effect and respiration impairment. These results show that the essential oil has a non-specific general inhibiting effect on the CNS. This non-specific action is expressed by the anticonvulsive effect, the influence on the spontaneous motor activity and the potentiation of narcotic drugs. Moreover, further evidence of its non-specific general depressant effect is that the essential oil itself has a narcotic effect when administered in doses close or equal to toxic ones (LD50 520 mg/kg in mice and 740 mg/kg in rats) [41]. More investigations on the action upon the CNS are necessary to better understand which components are responsible for the inhibiting effect and, eventually, which are responsible for a stimulating effect. In view of the reported data and the above considerations it is important to emphasize that the plant extracts should be utilized with the maximum care, particularly when the target is the CNS. Choleretic activity

S. sclarea oil, administered by parenteral route in rats to obtain a stronger reaction closer to the effect produced by continued oral treatment, had a weak effect on choleretic activity and was substantially hydrocholeretic

400

[45]. The effect on bile flux could explain the efficacy of S. sclarea used in empirical treatment of some digestive disorders [41-42]. Doses of the oil higher than 250 mg/kg greatly increased the narcotic effect of urethane, used as a general anesthetic in the test of choleretic activity, leading to a deeply depressant effect on the CNS. Anti-inflammatory and peripheral analgesic properties

Some empirical uses of S. sclarea as a cure for cold ailments of the respiratory apparatus as well as nervous polyarthritis and acute rheumatism [8] led to the hypothesis of an anti-inflammatory action of certain components, in particular the essential oil. With the aim of determining antiinflammatory and anti-nociceptive properties of S. sclarea essential oil, studies were conducted on experimental models of inflammation and pain [35, 46]. Table 5. Salvia sclarea oil: Anti-inflammatory activity (carrageenin edema) % inhibition

Mean edema volume (ml) ± s.e.

Compounds: Dose control

Basal

Ih

3h

5h

0.93 ± 0.02

1.04 ± 0 . 0 2

1.21 ± 0 . 0 7

1.23 ±0.08

Ih

3h

511

indomethacin

5

0.89 ± 0 . 0 3

0.93 ± 0.03

0.98 ± 0.05

0.96 ± 0.03

-64**

-68**

-77**

5". sclarea oil

250

0.92 ± 0.04

0.89 ± 0 . 0 4

0.99 ± 0 . 1 0

1.01 ±0.06

."73**

-75**

-70**

0.93 ± 0.02

1.10 ± 0 . 0 5

1.49 ± 0 . 0 4

1.41 ±0.04

ester fraction

140

0.95 ±0.01

0.98 ± 0.02

1.06 ±0.04

1.09 ±0.02

-82**

-80**

-71**

alcohol fraction

80

0.84 ± 0.03

0.92 ± 0.02

1.01 ±0.09

1.02 ±0.09

-53**

-70**

-58**

control

0.83 ± 0.03

1.15 ± 0 . 0 5

1.39 ±0.09

1.42 ±0.06

162.5

0.77 ± 0.02

0.96 ± 0.04

1.13 ±0.08

1.13±0.07

-41*

-36*

-39*

pineol mix.

32.9

0.77 ± 0.03

0.93 ± 0.05

1.04 ± 0 . 0 8

1.07 ±0.04

-50*

-52**

-49*

linalyl acetate

10.3

0.81 ± 0 . 0 4

0.99 ± 0.05

1.10 ± 0 . 0 8

1.12 ± 0 . 0 4

-44*

-48*

-47*

control methyl chavicol linalool and ter-

Tukey test compared with control values (*; p<0.05; **: p<0.01)

In carrageenin-induced hind paw edema [47] the systemic administration of the essential oil produced a reduction in the local inflammation process. This effect was in intensity and duration equivalent to therapeutic

401 doses of indomethacin, well known for its good anti-inflammatory analgesic properties. S. sclarea oil was also able to reduce histamine-induced hind edema [48]; the antiedematous effect increased progressively in time the administration of the edema-inducing agent and was greater chlorphenamine, used as antihistaminic reference drug.

and paw after than

Table 6. Salvia sclarea oil: Anti-inflammatory activity (histamine edema) % inhibition

Mean edema volume (ml) ± s.e. Compounds:

Basal

30'

60'

90'

120'

control

0.63±0.03 0.99±0.08 0.99±0.08 0.88±0.07 0.88±0.07

clorphenamine

0.69±0.02 1.05±0.06 0.99±0.05 0.96±0.04 0.96±0.05

control

0.70±0.06 1.03±0.11 1.03±0.08

S. sclarea oil

0.90±0.07 1.16±0.05

control

1.00±0.04

ester fraction

60'

90'

120'

0

-17

8

8

1.01±0.11 1.00±0.08 -21* -61** -61** -57**

1.03±0.01 1.02±0.02

1.03±0.02

1.30±0.03

1.25±0.05

1.20±0.04

0.91 ±0.02 1.21 ±0.03 1.13±0.03

1.09±0.02

1.07±0.03 -23** -27*

-28*

-20

1.08±0.04

1.04±0.04 -38** -33*

-48*

-55*

-35*

-37*

1.39±0.04

alcohol fraction 0.95±0.03 1.19±0.05 control

30'

1.15±0.06

0.70±0.06 1.03±0.11 1.03±0.08

1.01±0.11 1.00±0.08

1.41±0.05

1.28±0.04

1.28±0.04

1.27±0.04

0

pineol mix.

0.88±0.09 1.37±0.09

1.32±0.08

1.20±0.07

1.21±0.11

48*

33*

3

10

linalyl acetate

0.84±0.05

1.11 ±0.09 1.05±0.08

1.02±0.07

1.03±0.09

-18

-36*

-42*

-37

methyl chavicol 1.08±0.08

-39*

linalool and ter-

Tukey test compared with control values (*: p<0.05; **: p<0.01)

The above mentioned studies showed that these actions are mainly due to the oxygenated fraction of the oil containing alcohols (mainly linalool and alpha-terpineol), corresponding esters and methyl chavicol, this last extracted from samples of Sardinian wild 5. sclarea plants only. All these compounds, when administered separately, had a lesser effect than the essential oils in toto. These results suggest that the antiedematous effect is due to the synergistic action of the constituents of the essential oil. On the other hand, the peripheral analgesic action of the oil, evaluated in mice by inhibition of writhings induced by formic acid injection [49], was only moderate and weaker than the action of indomethacin. In this case also, the effect seems to be attributable to the oxygenated compounds of the oils.

402

The observations on the degree of remission of the inflammation, artificially induced with irritants and measured in vivo, indicate that the S. sclarea oil interferes to some extent with the agents involved in the inflammatory process (arachidonic acid derivatives, histamine, protease, quinine, radicals reactive to oxygen, etc.) [48, 50]. The results obtained so far in this field could justify the empirical use of S. sclarea in various inflammatory ailments. Antimicrobial activity towards bacteria and fungi

Some empirical uses of S, sclarea suggest that its essential oil, like many essential oils of the Lamiaceae family, has an inhibitory effect on microorganisms pathogenic to man. Ulubelen et al. [51] examined the antimicrobial activity of some diterpenes and sesquiterpenes found in an acetone extract of S. sclarea. They found that 2, 3-dehydrosalvipisone, sclareol, manool, 7-oxoroyleanone, spathulenol and caryophyllene oxide were active against Staphylococcus aureus; 2, 3- dehydrosalvipisone and manool were active against Candida albicans; and caryophyllene oxide was active against Proteus mirabilis. More recently, other authors [52] studied the antimicrobial activity of the essential oil in toto in comparison with the essential oil of S. officinalis, known to have antimicrobial properties [53-54], and chlorhexidine, a synthetic antimicrobial drug widely used in diseases of the oral cavity [55]. These experiments showed that S. sclarea oil had microbiostatic inhibitory activity against Staphylococcus aureus, Escherichia coli. Staphylococcus epidermidis and Candida albicans at concentrations of between 1.5 and 2.0 mg/ml. The minimal inhibitory concentrations (MIC) of S. officinalis oil, expressed as the lowest concentrations which completely inhibited bacterial growth in the microorganisms tested, are in agreement with the MIC reported in other studies in the literature (2 mg/ml). The inhibition of growth of Streptococcus salivarius. Streptococcus sanguis, Klebsiella pneumoniae, Citrobacter freundii, Enterobacter cloacae and Serratia marcescens required much higher concentrations of both oils (MIC more than 2 mg/ml). Of the alcoholic components of 5. sclarea oil, alpha-terpineol showed marked microbiostatic activity against two Gram positive (5'. aureus and 5. epidermidis) and two Gram negative bacteria {E. coli and Pseudomonas

403

aeruginosa) and C. albicans, a fungi used as strain of reference (MIC between 0.250 and 1 mg/ml), with candidicidal activity at doses >2 mg/ml. The inhibitory effect of linalool was weaker (MIC 1-2 mg/ml), with fungicidal action at a concentration of >2 mg/ml. In the range of concentrations used in the tests, the corresponding esters (linalyl acetate and alpha terpinyl acetate), the alpha- and beta-thujone mixture and camphor, which are the main components of S. officinalis oil, showed no significant effects. The inhibitory activity of S. sclarea oil on the growth of C. albicans increased progressively with contact time (time required to inhibit the growth of fungus) and was significantly higher than that of S. officinalis oil. On the other hand, the oil of 5. sclarea had practically no effect on S. aureus after 15 min exposure and S. officinalis oil produced only a slight degree of inhibition. Antimicrobial activity of 5. sclarea essential oil, although much weaker than chlorhexidine (MIC 0.006 - 0.25 mg/ml), could be of interest in some areas where 5. officinalis oil is employed, since its action is as good, if not better. The performance of the oil could be enhanced, not only with regard to its inhibitory effect on the growth of susceptible microorganisms, by technological and biopharmaceutical improvements to the formulations. If the vehicles of the oil were either natural or synthetic matrices able to adher to the skin or mucous, a longer inhibitory effect on the growth of microorganisms, responsible for local or systemic infections, would be obtained, besides peripheral anti-inflammatory and analgesic action, demonstrated in experimental animals after systemic administration. Pharmaceutical applications The above mentioned effects suggest the possible use of S. sclarea oil in the stomatological field to treat infective conditions which are often accompanied by inflammation. Since anti-inflammatory and analgesic effects require some drug absorption, some research [56-57] has been done recently to evaluate the capacity of the essential oil components to permeate the oral mucous in various mucoadhesive formulations of the oil (microemulsions, hydrogels and gelled microemulsions), useful in local treatments of stomatological ailments. The permeation properties of the

404

essential oil through porcine oral mucous were evaluated using modified Franz diffusion cells [58]. This mucous was chosen because its permeability is very similar to that of human oral tissue [59-61]. These experiments showed that 5. sclarea oil was able, on the whole, to permeate the oral mucous with significant statistical differences between the pure oil and its formulations. The best permeation profiles through the oral mucous were found with gelled microemulsions while the gels and the simple microemulsion matrices tested had negative effects on the absorption of the main oil components. A comparison of the Kp of the single components of the essential oil shows that beta-pinene, 1, 8-cineole, alphaterpineol and linalool have a higher Kp than the ester fraction of the oil. The former components are usually used as enhancers in percutaneous absorption [62-63] but they are also good permeants of the oral mucous. These authors pointed out that some components of the essential oil are able to permeate the porcine oral mucous to a good extent from suitable preparations of the oil. Monoterpene alcohols to which at least a part of the anti-inflammatory, analgesic and antimicrobial activity of the oil can be attributed, exhibited good permeation profiles, supporting the hypothesis that the essential oil could be useful in the treatment of conditions in which infective processes and inflammation are present together. Activity toward phytopathogenic fungi Considerable interest is given nowadays to the possibility of using natural products as parasiticides in agriculture to replace or enhance phytochemical products, because of the serious environmental problems caused by the often indiscriminate use of pesticides. Massive use of synthetic fungicides against parasites has led to the appearance of strains which are resistant to both the dicarboximide derivatives (iprodione, procymidone, vinclozolin [64-65] and the benzimidazoles (benomyl, carbendazim, thiophanates, etc.) [66-67] making it much more difficult to control the pathogens. Peana et al. [68] assessed the effectiveness in vitro of S. sclarea oil and its LC fractions, containing monoterpenic esters and alcohols, against a variety of phytopathogenic fungi which infect different species of host plants {Botrytis cinerea Pers., Fusarium solani (Mart.) Sacc, Phytophthora nicotianae var. parasitica (Dastur) Waterh., Rhizoctonia solani

405

Kuhn, Sclerotinia sclerotiorum (Lib.) de Bary and Sclerotium rolfsii Sacc). It was demonstrated that S. sclarea essential oil had a lethal effect on five of the six pathogens tested; the most susceptible microorganisms were B. cinerea and S. rolfsii (MIC between 1.3 and 1.6 mg/ml) while Fusarium solani was insensitive (MIC >5.4 mg/ml). The chromatographic fractions were less active compared with essential oil in toto. The most marked inhibitory action was observed with the alcohol-containing fractions. All in all, these observations indicate that further research should be done to evaluate both the synergic action of the components of the different fractions and the activity of the essential oil in toto, when applied directly to the plant species infected by the fungi, in suitable formulations. Other activities Lis Balchin et al [69] studied the action of S. sclarea oil on the rat isolated phrenic nerve diaphragm preparations and compared with activity on field- stimulated guinea-pig ileum preparations. The oil produced a contracture and inhibition on the skeletal muscle of the bi-phasic response to nerve stimulation, whilst only a contracture, with or without a decrease in response to field stimulation, was produced in smooth muscle. Biological activities of sclareol This diterpene derivative, extracted from S. sclarea, is naturally an epimeric mixture in a ratio of approximately 9:1, where the 13R-epimer is the major epimer in the mixture [8]. Malone et. al [16] screened the sclareol i.p. administration in rats for its potential pharmacological and toxicological activities using an observational open-field method, rating and measuring 64 parameters per animal for 7 days. The apparent effects appeared to be due to the physical nature of the material rather than to any intrinsic biological activity of the molecule. However, Georgieva et al. [70] demonstrated that sclareol glycol is able to influence cAMP levels in different brain areas of rats (anterior hypophyses, sections of brain cortex and cerebellum), suggesting that its action is due to a direct activation of the catalytic sub-unit of the enzyme adenylate cyclase.

406 Concluding remarks Wide range investigations of S. sclarea oil have demonstrated an enormous progress in the field. In general, S. sclarea oil has interesting biological activities which could be used not only in the treatment of human ailments but also in agriculture, particularly in the so-called biological cultivations which are being encouraged worldwide. In the last ten years, great advances have been made in studies on the production, biological activities and applications of S. sclarea oil. However, fundamental questions remain to be answered like the mechanism by which the oil acts on the CNS, and works as an anti-inflammatory and analgesic drug and as a hydrocholeretic and antimicrobial drug. Future studies will continue to be directed towards an understanding of the interactions of the oil with target biological structures. A challenge arises from the recent findings that the oil specifically interacts with the inflammatory process together with microbial growth. These fascinating topics are currently under further investigation because broader applications of suitable preparations of the oil are to be expected in the odontostomatological and gynecological fields. SALVIA DESOLEANA ATZEIET PICCI General and Botanical description Salvia desoleana is an endemic species which for many years was considered to be a variety of 5. sclarea, because of close morphological, systematic and phytochemical affinities [71]. This Salvia species is a perennial herbaceous plant, evergreen, rhizomatous, with multiple caules, flowering between mid-May and mid-July. Sometimes further efflorescence takes place between August and November, but this is limited to new axillary buds. It grows in sunny areas in different types of soil (calcareous, granitic and granitic-porphyritic) [42, 71-72]. Phytochemical investigation Like other aromatic plants, 5. desoleana releases its essential oil when submitted to a steam distillation process. Studies carried out up to now [6,

407

72] have described a production process based on fresh plant material collected within the last 24 hours, using a plant material-water ratio of 2 and a recycling system of aqueous phase. The distillation time was about 2 hours and the oil obtained was separated from the aqueous solution and then dried by treating with anhydrous Na2S04. Essential oil yield, determined in a Clevenger-type apparatus in conformity with Italian Pharmacopoeia standards [73] and relating to dry material, ranged from 0.5-2.5% depending on the part of the plant utilized, time of harvest and cultivation soil. GC and GC/MS analyses The chemical composition of S. desoleana oil was determined by GC and GC/MS analyses by different authors [6, 71-72, 74]. Generally, GC analyses were performed using fused silica capillary columns with different polarity like 5% diphenyl: 94% dimethyl: 1% vinylpolysiloxane bonded phase and Carbowax 20M bonded phase columns. Moretti et al. [72, 74] reported the relative retention indices of the oil constituents, calculated as described in the literature [75], and quantitative data obtained using ethylene glycol monobutyl ether as an internal standard. These authors also reported GC/MS data obtained from a GC apparatus directly coupled to a mass selective detector (MSD, 70 eV). Influence of environmental conditions on the composition of S. desoleana oil In 1999, Moretti et al. pointed out that the oil yield and composition depend on the parts of the plant utilized, time of harvesting and soil [72]. The data obtained in the same experimental conditions showed that the flowering leaves gave the highest essential oil yield while no statistically significant differences were found between non-flowering leaves (upper and lower parts). In relation to the composition, these authors reported that the ester component, between 27.3% and 61.1%, was characterized mainly by linalyl acetate and alpha-terpinyl acetate. Linalyl acetate was the highest component of the oil obtained from flowering leaves and upper part in toto but was in much lower concentrations in oil from non flowering leaves.

408 Table 7. Yield and composition of oils related to the parts of plants utilized Flowering leaves

Upper part in toto

Compounds:

Non-flowering leaves Upper layers

Lower layers

alpha-pinene

0.5

0.4

0.1

0.1

sabinene

0.9

0.9

0.1

0.2

1 beta-pinene

1.2

1.0

0.1

0.3

myrcene

0.5

0.2

0.1

0.1

alpha-phellandrene

2.0

2.3

0.2

0.5

para-cymene

0.3

0.3

0.1

0.1

limonene

1.2

1.6

0.5

0.8

1,8 cineole

8.2

6.4

1.0

1.2

gamma-terpi nene

0.9

1.5

0.9

0.7

cis-linalool oxide

1.0

1.3

0.3

0.4

trans-linalool oxide

0.3

0.5

0.2

0.3

linalool

5.9

6.7

0.9

2.1

camphor

0.1

0.2

0.3

0.4

terpinen-4-ol

0.1

0.2

0.7

-

alpha-terpineol

5.3

12.1

40.8

38.7

gamma-terpineol

0.2

-

-

-

methyl chavicol

0.1

0.2

0.2

0.1

nerol

0.4

0.4

0.1

0.2

linalyl acetate

37.0

29.0

6.1

10.4

1.1

1.0

0.3

0.5

alpha-terpinyl acetate

20.3

19.8

18.1

18.0

gamma-terpinyl acetate

1.6

-

-

-

neryl acetate

0.8

0.9

2.5

1.7

geraniol

geranyl acetate

1.4

1.7

0.6

1.1

beta-caryophyllene

0.5

0.9

2.7

2.4

germacrene D

0.3

0.8

7.6

1.3

caryophyllene oxide

0.1

-

-

Oil yield

2.1

1.4

0.7

1 0.7

409 The alpha-terpinyl acetate concentration was practically the same in all parts of the plant (18.0-20.3%). The alcohols, making up about 13.042.8%, were mainly linalool and alpha-terpineol, the former being more abundant in flowering leaves and in the upper part in toto while the latter was found in higher quantities in other samples (about 40%). The hydrocarbon component, between 6.5% and 12.4%, showed differences in both quality and quantity in correlation with the level of alphaphellandrene, prevalent in the oil obtained from the first two samples, and with the sesquiterpenic hydrocarbons (beta-caryophyllene and germacrene D) prevalent in remaining samples. The epoxide component was characterized by a prevalence of 1, 8 cineole in a concentration of between 6.4 and 8.2 % in samples containing flowering leaves but was significantly lower in the others (about 1%). The yield and composition of the oil obtained from aerial parts of plants, from the same experimental station and gathered at different times during the flowering period, showed no significant variations in either the main classes of constituents or the single components. The influence of soil and climate on the composition and characteristics of the oil were studied using S. desoleana plants from four different experimental stations in Sardinia (Decimoputzu, Laconi, Orosei, Sassari) where small plots had been planted, starting from seeds of common origin. The plant material was gathered in all stations at the beginning of June. No statistical differences were found in the oil yield. As regards composition, the ester fraction was most abundant in the plants from the Sassari station (62.4% of the oil in toto), while it was much lower in the samples from Orosei (38.2%). The main esters were linalyl acetate and alpha-terpinyl acetate in each case, the former more abundant in the oil from the samples grown in a calcareous area (42.5% of the oil in toto) while the latter was found in all the samples in concentrations of between 17.5 and 22.4%. The alcohol component was higher in samples collected in the Orosei station (38.9%) and much lower in the Sassari samples (9.4%). The most abundant alcohols were linalool and alpha-terpineol. The hydrocarbon content was between 6.6% (Laconi) and 10.8% (Sassari) in each oil in toto. The above data reveal that 5. desoleana is able to adapt itself to different soil types and climatic variations. It was also found that the oil obtained from plants grown in calcareous soil was particularly rich in esters while the alcohol component was higher in plants

410

grown in chalk-free soil, particularly in soil with a high nitrogen content and low level of assimilable phosphorus. The pH of soil appeared to have little effect; likewise the texture of the soil, whether sandy or clay. The data regarding the biomass showed that the flowers were richer in oil than the other parts of the plant. Table 8. Composition of 5. desoleana oil at different harvesting time Beginning of June

End of June

Compounds:

Middle of July

alpha-pinene

0.4

0.9

09

sabinene

0.9

1.3

1.5

beta-pi nene

1.0

1.6

1.9

myrcene

0.2

0.1

-

alpha-phellandrene

2.3

2.3

2.2

0.3

-

-

limonene

1.6

1.8

1.0

1 para-cymene 1,8-cineole

6.4

8.6

8.8

gamma-terpinene

1.5

0.2

0.1

cis-linalool oxide

0.5

0.5

0.4

trans-linalool oxide

1.3

1.4

1.0

linaiool

6.7

8.3

10.1

camphor

0.2

0.2

0.2

terpinen-4-ol

0.2

0.2

0.2

12.1

8.4

9.2

gamma-terpineol

-

0.1

-

methyl chavicol

0.2

0.1

0.2

alpha-terpineol

nerol

0.4

0.4

0.7

linalyl acetate

29.0

32.4

31.1

geraniol

1.0

1.1

1.7

alpha-terpinyl acetate

19.8

17.5

16.4

gamma-terpinyl acetate

-

-

-

neryl acetate

0.9

0.8

1.0

geranyl acetate

1.7

1.6

1.8

beta-caryophyllene

0.9

0.7

0.6

germacreneD

0.5

0.6

2.2

-

0.1

-

2.2

2.4

2.5

1 caryophyllene oxide 1 Oil yield

411

Table 9. Composition of 5. desoleana oil from different experimental stations Compounds:

Decimoputzu (17msl)

Laconi (583 msl)

Orosei (7 msl)

Sassari (78 msl)

alpha-pinene

0.4

0.6

0.5

0.8

sabinene

0.9

0.8

0.9

1.4

beta-pinene

1.0

1.0

1.2

1.8

myrcene

0.2

n.d.

1.9

0.2

alpha-phellandrene

2.3

1.5

n.d.

3.2

para-cymene

0.3

n.d.

n.d.

0.1

limonene

1.6

0.9

2.0

2.1

1,8-cineole

6.4

7.7

9.5

10.7

gamma-terpinene

1.5

0.1

0.9

0.2

cis-linalool oxide

1.3

0.9

1.0

1.5

trans-linalool oxide

0.5

0.3

0.5

0.6

linalool

6.7

9.5

24.8

5.3

camphor

0.2

n.d.

trace

0.1

terpinen-4-ol

0.2

0.2

0.3

n.d.

alpha-terpineol

12.1

16.5

8.6

2.7

gamma-terpineol

n.d.

n.d.

n.d.

n.d.

methyl chavicol

0.2

0.3

0.3

0.2

nerol

0.4

0.5

1.4

0.4

linalyl acetate

29.0

28.0

10.0

42.5

geraniol

1.0

1.4

3.8

1.0

alpha -terpinyl acetate

19.8

17.6

22.4

17.5

gamma -terpinyl acetate

n.d.

n.d.

n.d.

n.d.

neryl acetate

0.9

0.8

2.0

0.8

geranyl acetate

1.7

1.5

3.8

1.6

beta -caryophyllene

0.9

1.2

0.6

0.3

germacrene D

0.5

0.5

n.d.

0.6

caryophyllene oxide

n.d.

n.d.

n.d.

0.2

total hydrocarbons

9.6

6.6

8.0

10.8

total esters

51.4

47.9

38.2

62.4

total alcohols

20.4

28.1

38.9

9.4

oil yield

2.1

1.8

1.9

1.8

412 The oil obtained from the inflorescences had a higher concentration of esters and fewer sesquiterpenic hydrocarbons, making it suitable for use in the perfume industry. The biomass without inflorescences produced an oil whose main component was alcohol. This oil could be of interest to the pharmaceutical sector since some biological action can be attributed to it. Analysis of the composition of the oils showed that, whereas the alphaterpinyl acetate concentration was about the same in all the samples, the linalyl acetate content varied considerably. This appears to be correlated not only with the corresponding alcohol but also with the alpha-terpineol level. An increase in these two components produced a corresponding reduction in the amount of linalyl acetate present in the oil. The phenology of the species would allow two harvests of the flowers a year, one at the first flowering (mid-May to mid-July) and the other some time between August and November. In this way, the oil yield could be optimized and distillates of a particular quality could be obtained (i.e. with a high ester content). These characteristics could be further enhanced by growing the species on calcareous soil. Furthermore, an oil rich in alcohols, of value in the pharmaceutical industry, could be obtained from the leaves all year round. Chemical affinity between the essential oils ofS. desoleana and S. sclarea

In a previous study, Ghizzoni et al. [6] examined comparatively the chemical composition of the essential oils of S. sclarea and S. desoleana growing wild in Sardinia and pointed out that these oils had a very similar composition. As regards the cultivated species, the essential oils obtained from plants collected in the same station and the same time of harvest showed a similar composition from a qualitative point of view. The oil obtained from S. sclarea plants cultivated in the same environmental conditions contained a higher quantity of alcohols than S, desoleana oil. The main component of the alcoholic fraction was alpha terpineol, which represented about 47% of the S. sclarea oil in toto. The amount of esters was lower than in S, desoleana oil owing to a smaller percentage of linalyl acetate while the percentage of alpha-terpinyl acetate was almost the same (about 22% against 19%). Some further differences in composition were related to the quantities of epoxides (mainly 1,8 cineol), monoterpene hy-

413 drocarbons such as alpha phellandrene and limonene, and sesquiterpenes like beta-caryophyllene and germacrene D. Empirical uses In Sardinian folk medicine S. desoleana was used to treat a sore throat associated with local infection and a temperature, in the form of local applications of the leaf (sometimes after dipping in hot olive oil) inside cloths soaked in hot water. It was also drunk as a decoction of the leaf and leafy twigs as an antipyretic. Ground fresh leaves were also used to treat cutaneous pustules. Pharmacology Considering the morphological and systematic affinities of S. desoleana with S. sclarea, used more in folk medicine, these authors (Peana and Moretti) examined comparatively some pharmacological activities of the essential oils of these aromatic plants after their bioassay-guided fractionation and chemical and physical characterization. Some pharmaceutical preparations, suitable for topical applications, were also prepared and tested to evaluate the ability of the main oil components to permeate oral mucous membrane in an in vitro system. Activity on the central nervous system

Peana et al. [44], in a study on the CNS, showed that the S. desoleana oil produced in mice effects similar to that described above for S. sclarea oil. Some quantitative differences in the action on the CNS were related to the exploratory activity: in fact S. desoleana oil produced a greater reduction in exploratory activity than S. sclarea oil. Its effect was comparable to that of 5 mg/kg of diazepam, a drug known to have a marked sedative effect when administered in large doses. S. desoleana oil was also more active than S. sclarea oil in the test with pentamethylenetetrazole to assess its action as an anticonvulsant and anti-anxiety agent [50].

414 Choleretic activity

Peana and Moretti [45,74] demonstrated that S. desoleana oil and some of its components possess a choleretic activity in the rat. This oil and its components at doses corresponding to the quantity contained in the oil, when administered by parenteral route (i.p.) to obtain a stronger reaction closer to the effect produced by continued oral treatment, produced a significant increase in bile volume. Fig. (1).

1.5

m tfi ffi rfl r[l Control

S. desoleana e.o.

Esteric fraction

Alcoholic fraction

Linalool+alphaterpineol

Dbasal D Ih tl2h BSh • 4 h

Data are expressed as mean ± s.e. Statistical analyses were performed by ANOVA followed by Tukey test compared with basal values (*: p<0.05;**: p<0.01).

Fig. (1). S. desoleana oil: choleretic activity after ip treatment

The dry residue of bile collected after oil administration, was higher than that of the controls during the first two hours. Moreover, the dry bile residue in animals treated with LC fractions of the oil was related to the chemical composition of the fraction administered. Alcoholic components of the oil and a synthetic mixture of linalool and alpha-terpineol produced a dry bile residue lower than basal values, showing a typical hydrocholeretic action of these essential oil derivatives. These obtained findings confirm previous observations on terpenic alcohols described as agents able to increase the bile flux [76]. Results so far obtained from these authors suggest that the activity of S. desoleana oil on bile flux is substantially

415 hydrocholeretic. This activity was much higher than that of 5. sclarea oil, Fig. (2), characterized by a lower monoterpenic alcohol content.

r^ 0.5

g 0.0 Control Dbasal • ih E2h BSh

S. desoleana e.o.

S. sclarea e.o.

•4h

Data are expressed as mean ± s.e. .Statistical analyses were performed by ANOVA followed by Tukey test compared with basal values (*: p<0.05; **: p<0.01). Fig. (2). S. desoleana and S. sclarea oil: choleretic activity after ip treatment.

Anti-inflammatory and peripheral analgesic properties

The behavior of 5. desoleana essential oil towards a local inflammation process induced by carrageenin injection depends on its composition. In fact, studies carried out on oil samples characterized by different concentrations of alcohol and ester components, LC fractions of the oil and the single oil components showed that the composition is very important as regards the capacity of the oil to reduce the experimental edema, Fig. (3). The experimental data [77] showed that the anti-inflammatory effect is mainly due to the presence of oxygenated compounds, like alcohols (mainly linalool and terpineol) and corresponding esters, as previously reported in this review with regard to S. sclarea oil. All these compounds, when administered separately, had a lesser effect than the essential oil in toto. A higher concentration of alcohols led to an antiinflammatory effect which was more intense and had an earlier onset than that of ester-rich oils, Fig. (4). On the other hand, data so far indicate that the latter possess an intrinsic capacity to reduce local edema, attributable to their biological transformation by hydrolysis in the corresponding alcohols. This claim is sup-

416 ported by experiments with pure linalyl acetate, the main ester component of the oil which showed typical pro-drug behavior characterized by a delayed and prolonged action.

S. desoleana wild •Ih

BSh

Laconi

Decimoputzu

mSh

Bars represent the percent of inhibition (± s.e.) of edema with respect to the corresponding control group. (ANOVA followed by F tests for contrasts; groups were compared pooling together data from the three time points). *: p<0.05; **: p<0.01. Fig. (3). Antiinflammatory activity of S. desoleana essential oil (250mg/kg s.c.)

The antiedematous action of S. desoleana oil did not last as long as that of S. sclarea oil which has an higher ester concentration and continued unaltered for a length of time comparable to that of the reference drug (indomethacin). As regards the action against histamine-induced edema, S. desoleana oil showed a lower inhibitory effect than S. sclarea oil and chlorphenamine, used as reference drug [77]. This experimental model of edema, related to the presence of the autacoid histamine alone, does not allow an extensive evaluation of how the oils in question act, because of the short half-life of the edematous agent used. The writhing test, a model that permits an assessment of the peripheric antinociceptive effect of drugs by i.p. injection of formic acid, showed that S. desoleana oil possesses a remarkable analgesic effect comparable to the action of indomethacin, administered in the same experimental conditions. The action of 5. desoleana oil seems to be attributable to its oxygenated components and is more intense than that of S. sclarea oil [35,77].

417 100 1

S. desoleana e.o.

Ester fraction Alchol fraction

tragole

-25

Linalool and terpineol nnixture

Linalil acetate

D l h lll3h B S h Bars represent the percent of inhibition (± s.e.) of edema with respect to the corresponding control group. Statistical analysis was performed by ANOVA followed by Newman Keuls test; *: p<0.05; **: p<0.01; ***: p<0.005. F i g . (4). S. desoleana

In vitro antimicrobial

essential oil: antiinflammatory activity.

activity

against microorganisms

pathogenic

to man

Comparative studies on the antimicrobial activity of S. desoleana and 5. sclarea oils demonstrated that the first possessed in vitro only a weak inhibitory effect on the growth of the microorganisms tested [52]. Considering the composition of the oils utilized in these experiments, their different behavior against microorganisms could be explained by taking into account their chemical composition. In fact, the inhibitory action could be mainly due to the alcohol content while esters alone did not have any significant effect on microorganism growth. The samples of S. desoleana oil used in these experiments contained a lower concentration of active alcohols (20%) and a very high ester content (about 57%) compared with S. sclarea oil (about 51% and 38% respectively). The weak in vitro activity of S. desolana oil could be markedly increased in vivo by enzymatic hydrolysis of esters in the corresponding alcohols. This indicates that 5. desolana essential oil may be more effective in vivo as an antimicrobial agent.

418 Pharmaceutical application From the studies reviewed in this paper, it is evident that 5. desoleana essential oil, like S. sclarea oil, has a good anti-inflammatory and analgesic activity and a lower antimicrobial efficacy in vitro. All these effects are mainly linked to the content of alcohols. Moreover, all pharmacological action could be potentiated by in vivo transformation of the ester derivatives of the oil by enzymatic hydrolysis. Since an essential requisite for the metabolism of the drugs is absorption, our research group [78-80] evaluated the capacity of the S. desoleana oil components to permeate mucous membranes from various mucoadhesive formulations in view of their possible use in the stomatological and gynecological fields. The permeation capacity of the main oil components was markedly increased with the aid of diethylene glycol monoethyl ether (Transcutol®), a new enhancer agent useful in topical and mucoadhesive preparations [81-83]. All the formulations allowed a high permeability coefficient in comparison with the pure essential oil. In particular, the components with a terpenic structure (beta-pinene, cineole, alphaterpineol and linalool) had the highest capacity to pass through the porcine buccal mucous when compared to the other components (linalyl acetate and alpha-terpinyl acetate). Moreover, diethylene glycol monoethyl ether increased the amount of permeate in relation to its concentration in the pharmaceutical preparations studied. These experimental observations support the hypothesis that S. desoleana oil, like S. sclarea oil, if suitably formulated, could be useful in the treatment of conditions in which infective processes and inflammation are present. Effects against phytopathogenic fungi Peana and Moretti [68] reported that 5. desoleana essential oil had in vitro a lethal effect on some phytopathogenic fungi (Botrytis cinerea, Fusarium solani, Phytophthora nicotianae var. parasitica, Rhizoctonia solani, Sclerotinia sclerotiorum and Sclerotium rolfsii). The chromatographic fractions of the oil were less active compared with the essential oil in toto. The most marked inhibitory action was observed in the alcohol-containing fractions, and was not so evident as that obtained from the essential oil of S, sclarea. The activity of the essential oil in toto, also in this case, if ap-

419

plied directly to the plant species infected by the fungi, would need suitable formulations. Concluding remarks The current state of knowledge of the reviewed biological activities of S. desoleandi oil confirms the hypothesis that the behavior of the oil is very similar to that of S. sclarea oil. Although some reported data seem to be in disaccord, it is evident that the biological activities are closely related to the chemical composition of the essential oils, which is dependent on various factors such as time of harvesting, parts of the plants utilized, and environmental conditions, in particular soil type and climate. ABBREVIATIONS n.r. n.d. GC GC-MS ^^C-NMR CNS DL50

mix. s.e. MIC Kp Js

LC cAMP MSD i.p. s.c. e.o. ANOVA ®

= Not reported = Not detected = Gas-chromatography = Gas-chromatography Mass Spectometry = Nuclear Magnetic Resonance on ^^C = Central Nervous Sistem = Lethal dose 50 = mixture = Standard Error = Minimal Inhibitory Concentrations = Permeability Coefficient = Flux = Liquid Chromatography = Adenylate cyclase =Mass Selective Detector = Intraperitoneal = Subcutaneous = Essential oil = Analysis of Variance =Trade Mark

420

ACKNOWLEDGEMENTS The authors wish to thank Prof. A.D. Atzei (Dipartimento di Scienze del Farmaco, University of Sassari) for his professional aid on identification of plant material and also for his expert advice on uses in folk medicine.

REFERENCES [I] Lawrence, B.L.; Perfum. Flavor, 1996, 21, 57-68. [2] Kerrola, K.M.; Kallio, H.P.; J. Agric. Food Chem., 1994, 42, 2234-2245. [3] Tanker, M.; Tanker, N.; Sarer, E.; Atasu, E.; Senser, B.; Kurue, S.; Mericli, P. Proceedings of the IFEAT meeting. Antalya 1990, pp. 16-29,1996. [4] Punyarajun, S.; Nandhasri, P. J. Pharm. Sci. Mahidol Univ., 1981, 8, 65-70. [5] Singh, B.; Sood, R.P.; Singh, V.; 7, Essenl Oil Res., 1992, 4, 525-526. [6] Ghizzoni, C ; Picci, V.; Atzei, A.D.; Colombo, E.; Rivista Italiana EPPOS, 1991, 5, 25-34. [7] Lawrence, B.L.; Perfum. Flavor, 1986, 11, 73-80. [8] Lawrence, B.L.Production of clary sage oil and sclareol in North America. 4emes Rencontres Internationals Plantes aromatiques et medicinales. Nyons, 5-7 Dec. 1994, pp 41-58, CFPPA, Nyons 1995. [9] Lawrence, B.L. In Advances in Labiatae Science; R. M. Harley and T. Reynolds, Royal Botanic Gardens, Kew, 1992. [10] Willis, J.C. In A Dictionary of Flowering Plants and Ferns.; T^ Edn. revised by Airy Shaw, H.K., University Press, Cambridge, 1966. [II] Briquet, J. Labiatae In Die NatUrlichen Pflanzenfami lien; Eugler, A. and Prantl, K., Ed.; Englemann: Leipzig, 1897; Vol. 4 3a, pp. 183-375. [12]Pignati, S. Flora d'ltalia, Edagricole: Bologna, 1982; Vol. 2. [13]Chiappini, M.; Falchi, L ; Annali di botanica, 1966, 28, 1-21. [14] Picci, V; Clementi, P.; Boll. 1st. Bot. Univ. Sassari, 1975, 9, 293-299. [15]Fenaroli, G. Sostanze aromatiche naturali, Hoepli, U.: Milano, 1963; Vol. 1 [16]Malone, M.H.; McChesney, J.D.; Kouzi, S.A. ; Fitoterapia, 1991, 67,123-129. [17] Anon. Medicinal and Aromatic Plants in Germany in 1937. Deutsh Heilpflanze 4, 1938; Ann. Rept. Schimmel and Co., 1938; pp.127. [18]Leffingwell, J.C; Stallings, J.W.; Sellers, P.O.; Lloyd, R.A.; Kane, F.C. Vith International Essential Oil Congress. S. Francisco, 1974. [19]Schumaker, J.N.; Henley, W.M.; Bernasek, E.; Teague, C.E. Two-Stage Oxidation of Sclareol. U.S. Patent No.3,050,532, Aug. 21 1962. [20]Farbood, M.I.; Willis, B.J. Process for Producing Diol and Furan And Microorganisms Capable of Same. U.S. Patent No.4,798,799, Jan. 17 1989. [21]Farbood, M.I.; Morris, J.A.; Downey, A.E. Process for Producing Diol and Lactone And Microorganisms Capable of Same. U.S. Patent No.4,970,163, Nov. 13 1990. [22]Farbood, M.I.; Morris, J.A.; Downey, A.E. Process for Producing a Cyclic Ether. U.S. Patent No.5,155,029, Oct. 13 1992.

421 [23]Ohloff, G. In Scent and Fragrances; Springer-Verlag: Berlin, 1974; pp. 238. [24] Nevstrueva, R.I. In Gosud. Nikitskii Bot. Sad. Sbornik. Rabot: Moscow, 1955; pp. 39-51. [25] Janol, M.M.; Annales Chimie, 1932, 17, 5-127. [26]Guseva, K.A.; Rafanova, R.Ya.; Bulanova, A.V.; Virezub, S.I. Maslob. -Zhir. Prom., 1959, 25, 29-30. [27]Ivanov, N.N.; Grigor'eva, V.F.; Ermakov, A.I. Bull Appl Genet. Plant Breeding, 1929, 21, 321-324: Chem. Abs., 1932, 26, 5604. [28]Palamar, N.S.; Khotin, A.A.; Alekseeva, E.I. Gosud. Izd. SeVskhozyaistvennoi Lit., 1953, 45-69. [29]Teisseire, P.; Bernard, P.; Recherches, 1955, 5, 32-37. [30]Teisseire, ?.\ Recherches, 1960, 10, 10-21. [31]Corbier, B.; Teisseire, P. Recherches, 1974, 19, 275-278. [32]Tchorbadjiev, S.; Ivanov, D.; Marinov, V.; Stovanova, V.; Compt. Rend. Acad. Bulg. ScL, 1969, 22, 297-299. [33] Vlakhov, R.; Ivanov, I; Ivanov, D.; Parfum. Cosmet. Savons., 1970, 13, 600-602. [34]Lawrence, B.L.; Perfum. Flavor, 1990, 15, 69-71. [35]Moretti, M.D.L.; Peana, A.T.; Satta, M.; J. Essent. Oil Res., 1997, 9, 199-204. [36]Bernreuther, A.; Schreier, P.; Phytochem. Analysis, 1991, 2, 167-170. [37]Ravid, U.; Putievsky, E.; Kutzir, I.; Flav. Fragr. J., 1995, 10, 281-284.

[38] Leung, A.Y. In Encyclopedia of Common Natural Ingredients; Wiley Interscience, J.: Wiley and Sons, New York, 1980. [39] Grieve, M. Modern Herbal., Hafner Publ. Co. Darrin.Conn. 1970. [40] Guenther, E. The Essential oils', D. van Nostrand, Princeton, 1949. [41] Atanasova-Shopova, S.; Rusinov, K.; Izv. Inst. Fiziol. Bulg. Akad. NauL, 1970, 13, 89-95. [42] Marchioni, A.R.; Distefano, E.G. Le piante medicinali della Sardegna, Ed. Delia Torre: Cagliari (Italy), 1989. [43] Fackler, E.; Acta phytoterap., 1964, 2, 22. [44] Peana, A.; Satta, M. ARS Pharmaceutica, 1992, Tomo XXXIII N. 1, 458-461. [45] Peana, A.; Satta, M.; Moretti, M.D.L.; Orecchioni, M.; Acta Phytoterapeutica, 1995, 1, 12-16. [46]Peana, A.; Satta, M.; Pharmacol. Res., 1993, 27, 25-26. [47] Winter, C.A.; Risley, E.A.; Nuss, G.W.; Proc. Soc. Exp. Biol. Med., 1962, 111, 544547. [48] Pirisino, R. in Farmacologia e Tossicologia Sperimentale, Ed. Pitagora: Bologna, 1989; pp 191-192. [49] Eddy, N.B.; Leimbach, D. J. Pharmacol. Exp. Ther., 1953, 107, 385. [50] Goodman & Gilman's, The Pharmacological basis of therapeutics,9^ ed, Joel G. Hardman, Alfred Goodman & Gilman's, Lee E. Limbird 1996. [51]Ulubelen, A.; Topeu, G.; Eris, C.; Sonmez, U.; Kartal, M.; Kuruku, S.; Bozok Johansson, C.; Phytochemistry, 1994, 36, 971-974. [52] Peana, A.T.; Moretti, M.D.L.; Juliano, C.; Planta Med., 1999, 65, 752-754. [53]Biondi, D.; Cianci, P.; Geraci, .C.; Ruberto, G.; Piattelli, M.; Flavour and Fragrance Journal, 1993, 8, 331-337.

422 [54]Pellecuer, J.; Roussel, J.L.; Andarj, C; Privat, G.; Jacob, M.; Tomei, R. Riv. ItalianaEPPOS, 1979, 6, 10-11. [55]Goodson, J.M.; J. Clin. PeriodontoU 1996, 23, 268-272. [56]Ceschel, G.C.; Maffei, P.; Moretti, M.D.L.; Peana, A.T.; Demontis, S.; STP Pharma Sciences, 1998, 8, 103-106. [57]Ceschel, G.C.; Maffei, P.; Peana, A.T., Moretti, M.D.L. Atti 22^ Congresso Internazionale della Societa Farmaceutica del Mediterraneo Latino, Montpellier, 19-22 Settembre 1996. [58] Franz, T.J.; 7. Invest. Dermatol, 1975, 64, 190-196. [59] Friend, D.R. J. ofCont. ReL, 1992, 18, 235-248. [60] Squier, C.A.; Rooney, L.; J. Ultrastruct Res., 1976, 54, 286-290. [61]Squier, C.A.; Philip, W.W. In: Oral mucosal drug delivery, Rathbone, M.J., Ed.; Dekker, M. Inc.: New York, 1996 pp. 1-26. [62] Williams, A.C.; Barry, B.W.; Int. J. Pharm., 1991, 74, 157-168. [63]Obata, Y.; Takayama, K.; Machida, Y.; Nagai, T.; Drug des. discovery, 1991, 8, 137-144. [64]Maraite, H.; Gilles, G.; Meunier, S.; Weins, J.; Bal, E.; Parasitica, 1980, 36, 90101. [65]Nielsen, S.L.; Lundsgaard, J.; Vaxtskyddsrapporter, 1987, 48, 70-78. [66]Bollen, G.J.; Scholten, G.; Netherland Journal of Plant Pathology, 1971, 77, 83-90. [67] Pepin, H.S.; MacPherson, E.A.; Plant Disease Reporter, 1982, 66, 404-405. [68] Peana, A.T.; Moretti, M.D.L.; Acta Phytoterapeutica, 1996, 2, 25-29. [69]Lis Balchin, M.; Hart, S.; / Ethnopharmacol, 1997, 58, 183-187. [70]Georgieva, Zh.; Uzunov, P.; Eksp. Med. Morfol, 1989, 28, 1-7. [71] Atzei, A.D.; Picci, V.; Webbia, 1982, 36, 71-78. [72]Moretti, M.D.L.; Peana, A.T.; Sanna Passino, G.; Satta, M.; Tuberoso, C.I.G.; /. Essent. Oil Res., 1999, 11, 635-641. [73]F.U.I. X, Poligrafico dello Stato: Roma, 1998; pp. 156-158. [74] Peana, A.; Satta, M.; Moretti, M.D.L.; Orecchioni, M.; Planta Med., 1994, 60, 478479. [75] Van Den Dool, H.; Kratz, P.; J. Chromatogr. 1963, 11, 463-471. [76] Yamahara, J.; Kimura, H.; Kobayshi, N.; Yakugaku Zassh, 1983, 183, 979-985. [77] Peana, A.T.; Moretti, M.D.L.; Satta, M. VIII Convegno SIF. Arzana (Nu) 22-25 May 1996. [78]Ceschel, G.C.; Maffei, P.; Moretti, M.D.L.; Peana, A.T.; Demontis, S.; Farmacevtski Vestnik, 1997, 48, 240-241. [79]Ceschel, G.C; Maffei, P.; Moretti, M.D.L.; Peana, AT.; Demontis, S. 6^*^ Congresso Nazionale dei Docenti di Odontoiatria, Roma, 21-24 Aprile 1999. [80]Ceschel, G.C.; Maffei, P.; Moretti, M.D.L.; Demontis, S.; Peana, A.T. in print on International Journal of Pharmaceutics 1999. [81]De Vries, M.E.; Junginger, H.E.; Bodde, H.E.; Verhoef, J.C.; Ponec, M.; Craane, W.I.; Int. J Pharm, 1991, 49, 25-29. [82]Babar, A.; Solanki, U.D.; Cutie, A.J.; Plakogiannis, F.; Drug Dev. Ind. Pharm., 1990, 16,523-540.

423

[83]Krost, J.; Langer, R. In Hydrogels in Medicine and Pharmacy, CRC-Press Inc.: Boca Raton, Florida, 1987; Vol. 3, pp. 98.

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

425

SECONDARY METABOLITES WITH ANTINEMATODAL ACTIVITY EMILIO L. GHISALBERTI Department of Chemistry, University of Western Australia, Nedlands 6907 W,A, Australia ABSTRACT: Nematodes represent a serious threat to both plants and animals. New methods of control of these parasites are being sought since a number of soil-applied commercial nematocides are being phased out and resistance to anthelmintics is an increasing problem. Integrated pest management, transgenic plant resistance and biological control strategies are being investigated as methods of control. At the same time, the search for naturally occurring antinematodal compounds continues unabated. The discovery of the potent anthelmintic activity of the avermectins and milbemycins has stimulated interest in the identification of other natural products with high activity and selective mode of action towards nematodes. In this report, the wide range of metabolites shown to have antinematodal activity are reviewed. Aspects of their chemistry and activity are outlined and, wherever possible, the relationships between structure and bioactivity are emphasised.

INTRODUCTION Plant- and animal-parasitic nematodes cause considerable damage to various crops and contribute significantly to malnutrition and disease in domestic livestock and humans. Since phytoparasitic nematodes are typically soil-borne, they are a hidden threat to plants and their significance can be easily overlooked. World-wide, they are responsible for economic losses to agriculture of US$100 billion per year [1]. Great losses of domestic livestock are caused by the animal-parasitic species Haemonchus and Trychostrongylus, In humans, intestinal nematodes affect a large proportion of the world population; 25% harbour Ascaris lumbricoides, the large gut roundworm that causes malnutrition and obstructive bowel disease. The blood-sucking hook-worms, Ancylostoma diiodenale and Necator americanus, infect over 1 billion people [2,3]. The control of these organisms has become a matter of some concern. The most efficient synthetic nematocides have been withdrawn from agricultural use because of their toxicity to begnin soil organisms, their persistence in the environment and adverse effects on public health.

426 Increasing resistance towards anthelmintics has also become evident with various parasites. The need for a more ecologically rational pest management strategy has refocused attention on natural products. The discovery in 1979 of the potent anthelmintic and insecticidal activity of the avermectins and milbemycins (eg 1, 2), structurally related 16-membered macrocycles from Streptomyces species [4], has stimulated a search for other new and specific naturally occurring anthelmintics.

H,C

In the present review, an attempt is made to summarise the information available on natural products with antinematodal activity. Secondary metabolites from terrestrial plants and marine, fungal and bacterial sources are considered and aspects of their chemistry, structure and biological activity are reviewed. They are grouped according to their biosynthetic origin, often presumed, to facilitate comparison between structure and activity. Two main reviews on the topic of naturally occurring nematocides have been published in recent years. The first considered metabolites

427

from higher plants with activity against phytoparasitic nematodes [5], and the second described the nematocidal metabolites available from higher fungi [6]. Other reviews on the more general topics of antiparasitic agents from plants [7] and from marine animals [8] contain related information. Given the voluminous literature on the avermectins and milbemycins, these compounds will not be considered here except for comparison purposes. A number of pertinent reviews on these compounds are referenced in a general overview of the topic [4]. Recent results on their biosynthesis [9,10], activity [11-13], clinical [14,15] and veterinary pharmacology [16], mechanism of action [17,18], and environmental aspects [19] have been published. NEMATODES The study of nematodes has developed into two disciplines. Nematology deals with nematodes that are parasites of plants and invertebrates, and free-living nematodes in soil and water. The sequencing of the genome of the 'model' free-living nematode Caenorhabditis elegans, the first animal genome to be mapped, has attracted much interest to this field [20]. Helminthology is the study of nematodes and worms parasitic in vertebrates and mainly those of importance in human and veterinary medicine. Nematodes (also known as eelworms or roundworms) are millimitersized, non-segmented animals unrelated to true worms (Annelida). They are quite small, usually 100-1000mm in length, although some may be several millimetres. Nematodes are composed of an elongate cylindrical body usually tapering at both ends and the body is protected by an external, metabolically inert, flexible, multi-layered and transparent cuticle covering the outermost living layer, the hypodermis. They move with a serpentine motion and undulate through soil, water or viscous material and are found in marine, aquatic and terrestrial environments [21]. There are about 30000 described species of which '-^SOVo are marine, 15% are parasites of invertebrates or vertebrates, 10% are plant parasites and 25% are free-living. Phytoparasitic nematodes Nematodes are an important component of the soil microfauna. In one square metre of soil 1-20 million nematodes, contributing a biomass of

428 1.5-30 g, can be found [22,23]. Some are free-living saprophytes feeding upon bacteria, whereas others feed on the cytoplasm of hypae of different fungi, resulting in poor growth or death of the host. Free-living soil nematodes do not feed direcdy on living plant material and belong to the 'decomposer subsystem' where they are influential in nutrient cycling in the soil. Phytoparasitic nematodes are obligate feeders on plants, most attack the roots of their host, and can only obtain nutrients from the cytoplasm of living plant cells. When feeding, they use a protrusible structure, the stylet, to pierce the cell wall and, following injection of secretions, they ingest the cytoplasm. Migratory ectoparasites move along the root and feed more or less from specific root tissue. Migratory endoparasites are in closer contact with the plant but migrate freely between host and soil. Sedentary endoparasites spend most of their life in the roots of the host and live off induced feeding sites [23]. From a global perspective, the most damaging species are the root-knot nematodes (Meloidogyne spp) and the cyst nematodes {Heterodera and Globodera spp) [24, 25]. Some of the more commonly studied free-living and phytoparasitic nematodes are listed in Table 1 which also serves as a key for the abbreviated binomials of nematodes most often mentioned in this review. Vertebrate-parasitic nematodes Helminths constitute a major health problem in humans and domestic animals, especially in the tropical and sub-tropical areas of the world. They are classified in two phyla; the flatworms (Platyhelminthes), comprising the flukes (or trematodes) and the tapeworms (or cestode), and the roundworms or nematodes (Nemathelminthes). Most nematodes are not parasitic, but some infect vertebrates by entering the digestive tracts of animals where they develop into adults. Diseases caused by infection are dependent on the residence site of the adult nematode; the gastrointestinal tract or body tissues. All nematodes proceed through five developmental stages, LI to L5, in a series of moults. From the gastrointestinal tract, adults release eggs which pass to the external environment in the feces. Under the appropriate conditions, the eggs may hatch releasing the LI stage or, in some nematodes, the larvae may develop to the second or third stage before being released. The infective larvae spread into the surrounding herbage thus increasing the possibility of infecting other animals. The animal-

429 infecting nematodes frequently mentioned in this review and their abbreviations [26,27] are included in Table 1. Table 1.

Selected types of nematodes J_P_^P/AS!J nfativc species

1

Comments

|

1 C elegans

1 bacteriophagous nematode

|

1 P. redivivus

1 bacteriophagous nematode

|

Phytoparasitic Cyst forming

j 1 G. rostochiemis

1 Heterodera

potato cyst nematode

|

1 /y. glycines. H. schachtii 1 soybean cyst, beet cyst nematode

|

1 Endoparasitcs 1 Mehidof^yne

1 Tvlenchulus

I 1 /i/. incognita

root-knot nematode

|

M.javantca

root-knot nematode

|

P. penetrans

root lesion nematode

|

I 7! semi penetrans

1 citnjs nematode

|

i Z'. hamatus

\ pin nematode

|

1 Ectoparasites 1 Parafx'lenchus 1 Above ground feeders

I

1 Angifina

A. tritici

wlKat gall nematode

1 Aphelenchoides

A. besseyi

white-tip nematode

j

1 dursaphelenchus

B. xylophilus {lignicolus)

pine wood nematode

|

1 Ditylenchus

D. dipsaci

stem and bulb nematode

|

1 Animal parasitic nematodes 1 Ancylostoma 1 Ascaris j Coopeha

| A. duodenale \ A. lumbricoides C. curticei

\ hookworm

j

\ roundworms

1

j small intestine parasite guinea worm

1

1 Draawculus

\ D. medinensis

1 Haemonchus

\ //. contortus

j true-stomach nematode

| |

1 Nippostrongyit4s

j /V. brasiliensis

| rodent gut nematode

j

Ostertagia

j 0. ostertagi

Strongyloides

| S. stercolaris

1 threadworms

|

Toxocara

\ f. canf5

1 dog worm

|

Trichostrongylus

\ r. colubriformis

| small intestine parasite

|

1 Trichuris

jr. Irichuria

tRie-stomach nematode

| whipworms

|

J

430 NATURAL METHODS OF NEMATODE CONTROL The ability of some plants to exert nematocidal or nematostatic activity has been known for a long time. The marigold (Tagetes), recognised as an important medicinal plant since the 1st century A.D., has traditionally been used in India as a pest control in agriculture [28]. Marigolds are grown between beds of vegetables of the potato family and the direction of the marigold beds is changed every year. The resistance oiTagetes spp to Meloidogyne, the root-knot nematode that has up to 1700 plant hosts, was first investigated in 1938 [29,30]. M.Javanica failed to penetrate the roots and those that entered did not develop beyond the infective L2 form. Subsequent chemical studies resulted in the isolation of two nematocidal principles (3, 4) from the roots of Tagetes [31]. As another example, the inhibiting effects of root diffusates of some Cruciferae on the emergence of larvae of G. rostochiensis, the potato cyst nematode, has been ascribed to the production of isothiocyanates (5). Other examples of plants that contain antinematodal compounds have been summarised [32,33]. In the sequel, it will be obvious that many plants use defence compounds as a strategy against nematode infection. These metabolites represent a source of relatively simple compounds that could provide active principles and leads in the design of antinematodal agents. A recent compilation of plants with nematocidal activity contains 150 entries and, for most of these, the active agent(s) have yet to be identified [33].

3

CH.

\

N= C=S

4

5

//

In the past, treatment of diseases caused by nematodes was primarily concerned with those parasites that infected humans. Plants or plant juices were traditionally used as vermifuges. The juice of marigold flowers dropped into the ear was reputed to kill worms [28]. Some marine algal

431 species were used by early Greek, Turkish, Maoris and Japanese communities as vermifuges to combat intestinal parasites [34]. Ascaridole (6), the active principle in the oil pressed from the male fern {Dryopteris filixmas) and the oil of Chenopodium ambrosioides, has long been used in antiparasitic remedies [35]. In the early 1900s, ascaridole was one of the major anthelmintics used to treat ascarids and hookworms in humans and domestic animals. However, the oil is effective only at almost toxic doses [36]. In vitro studies have shown that ascaridole inhibited egg development of//, contortus, a parasite of small ruminants [37]. Modern integrated pest management tactics for the control of phytoparasitic nematodes are based on knowledge of the biology and ecology of the crops, nematodes and other interacting organisms. Crop sanitation programmes, use of nematode-resistant varieties, crop rotation, use of poor or non-host cover crops, fallowing and conservation tillage, soil amendments can all contribute to pest management [38]. The control of animal parasitic diseases is also a complex problem. At present, anthelmintic treatment and grazing management are used to control nematode parasites of livestock [39]. For humans, reduction of infections to levels below those which lead to significant morbidity requires a greater emphasis on health education, improved sanitation and water supplies on the one hand, and control of disease vectors, rational chemotherapy and vaccines, on the other [40,41]. Biological control of plant and animal parasitic nematodes has been investigated for some time and a number of approaches have been suggested [27,42]. One of these involves the use of fungi which are predatory on nematodes. The predator-prey relationship between a fungus and an animal was first discovered by Zopf in 1888 [21]. He observed that the fungus Arthrobotrys oligospora was able to capture and colonise motile nematodes with net-like bails of aerial hyphaes. Moreover, he showed that the fungus penetrated the wall of the nematodes and that the hyphaes grew within the body of the animal and consumed the contents. It was later found that the nets produce an adhesive which was largely responsible for holding the prey [21]. With nematophagous fungi, there is evidence for sophisticated interactions between prey and predator. The presence of nematodes stimulates the development of trapping organs and the fungus undergoes a switch from a saprophytic to predatory habit. The trapping organs may also exude attractants [43]. Studies on fungal biocontrol of nematode parasites of livestock has concentrated

432

mainly on Arthrobotrys oligospora. However, there are a number of other fungal predators which can survive passage through the gastrointestinal tract of ruminants [27, 39]. Moreover, a good case can be made for controlling the population at the free-living stage which occurs in the external environment [44]. In this situation, a number of predator organisms or their active metabolites could be used for control. For example, more than 100 species of fungi are known to have nematode destroying capabilities. Along these lines, two antinematodal agents have been commercialised. DiTera®, prepared from the lyophilised mycelium of the fungus Myrothecium spp., controls several phytoparasitic nematode species, eg Meloidogyne, Heterodera, Globodera, Pratylenchus, and does not affect begnin free-living nematodes [45,46]. The use of a strain of another fungus, Paecilomyces lilacinns, is being developed. This fungus is an egg parasite, infecting eggs at all stages of development, and seems to be most effective on eggs of root- and cyst-nematodes. The dried fungus (Paecil™)is used in a soil drench [47,48]. SCREENING FOR NEMATOCIDAL COMPOUNDS Various screening techniques using parasitic nematodes as the test organisms are available [49]. However, such studies are restricted by the often complex life cycle of the nematodes and by the difficulties involved in monitoring all stages of development. In contrast, free-living nematodes are more accessible to analysis under controlled conditions and, for example, Rhabditis sp and P, redivivus were selected as the test organisms for screening of a large number of Actinomycetes [50]. The much studied free-living nematode C. elegans was advanced as an alternative to the time-consuming in vivo assays with mice infected with N. dubius, the vertebrate parasitic nematode used in the assay that disclosed the avermectins, C elegans, which has a motor-neuronal system similar to that of N. dubiiis [51] and is sensitive towards peptides with ionophoric properties [52], has been favoured as the test organism in recent years [53]. A group of milbemycin analogues was discovered in 1987 using C elegans [49]. Compounds related to the milbemycins were detected by using K contortus obtained from infected sheep [54]. Several in vitro assays for anthelmintic activity have been described and an evaluation of the methods has been published [55]. Some assays rely on assessment of motility [56], although more objective assays are available [57]. The screening of

433

crude plant extracts can be complicated by the presence of tannins and polyphenols which are known to exhibit antinematodal effects [58, 59]. In these cases, pretreatment of the extracts with polyamide or polyvinylpolypyrrolidone has been recommended [60], The need to test the activity of compounds towards more than one organism [61] and, more specifically, to combine results from different tests to obtain a detailed picture of the effect of a potential nematocide, has been emphasised [62]. A new assay that provides rapid results and has high sensitivity has been reported. A good correlation between action mechanism and nematode shape was obtained with 15 nematocidal compounds [63]. In many of the studies collected in this review, assessment of antinematodal activity, understandably, was restricted to a single test organism. Much more information on the metabolism of parasites and the mode of action of nematocidal agents is being acquired, and more will result now that the C elegans genome has been mapped. Unique enzymes, transporters, receptor or cellular components in nematodes will become targets for evaluating potential drugs [49]. SECONDARY METABOLITES WITH ANTINEMATODAL ACTIVITY Fatty acids A number of simple fatty acids have been found to be toxic to nematodes, although very high concentrations (-1 mg/ml) were required [64-66]. Dicarboxylic acids produced a faster, more intense action; maleic acid was active, but fumaric acid was not [67]. One of the more unexpected findings was the observation that linoleic acid had nematocidal activity towards C. elegans, Linoleic acid was the only nematocidal agent found in the mycelial extract of several predacious Arthrobotrys species [68]. It showed significant activity (LD50 5-10 |Lig/ml), although it was much less active than ivermectin (LD50 0.1 |ig/ml). Linoleic acid production was stimulated in cultures containing nematodes and, when the concentration reached 25 |ig/ml, the nematodes were rendered immotile. Extracts of the roots of Iris japonica (Iridaceae) showed strong activity towards the rice white-tip nematode, A, besseyi, and this activity was shown to reside in the acidic portion of the extract [69]. The active prin-

434

ciples were identified as tetradecanoic, palmitic and linoleic acids. Of the several other acids tested, undecanoic acid, 11-undecylenic (LDgo-ioo 20 ppm), 2-undecylenic acid (LD60-8O 10 PPi^) ^^^ nonanoic (pelargonic) acid (LD60-8O 20 ppm) were the most potent. The activities found for a number of naturally occurring fatty acids are given in Table 2 [6,53,6871]. Whereas the methyl esters of oleic, linoleic and linolenic acid were inactive [6], methyl nonanoate (3.2 |LL1/1) reduced egg production of M. incognita [72]. Table 2.

Nematocidal activity of fatty acids Aphelencoides

Compound

1 Caenorhabditis elegans

besseyi LLD60-80(PP"^)

1 FIcxanoic acid

>200

Meloidogyne

Panagrellus

incognita

redivivus

1 LD50 (ppm)

|j-^50(ppm)

1

1

>ioo

ED95(ppm)

100

2500

>I00

156

1

1 Octanoic acid

-

1 Nonanoic acid

20

-

-

1 Oecanoicacid

20

50

>I00

156

1 Dodccanoic acid

20

25

>I00

156

1 Tetradecanoic acid

50

5

>100

inactive

|

1 Palmitic acid

200

25

>I00

inactive

|

1 Stearic acid

>200

50

>100

-

10

>100

25

100

1 9Z-Tctradecenoic acid 1 Oleic acid 1 9£-Octadccenoic acid 1 Linoleic acid 1 Linolenic acid

>100

10

>200 |

J

1

1 1 1 1 1 1

1

5

150

25

>I00

50

>100

5Z,8Z, IIZ, 14Z1 Eicosatetracnoic acid

|

1 Dccanedioic acid

{

1 10£"-Decenedioicacid

|

1 S-coriolicacid(7)

|

. . -

-

-

312 300

.

10

>100

50

>ioo

1

>ioo

1

10-Hydroxy.8£1 decenoic acid Cibaric acid (8)

| |

^ 1

100

1

-

1

435

A number of acids, alkyl amines and alcohols were tested for nematocidal activity against B. lignicolus. For the alkyl amines, the activity was correlated with the length of the alkyl chain and a maximum was reached with octadecylamine (LC50 2.1 ppm) [73]. In another study, a selection of amines and amides that disrupt hormone regulated processes and block sterol metabolism in insects were tested against the nematodes P. redivivus and M incognita. N, N-dimethyl amines with a CI 1-C14 chain were lethal at 5-10 ppm, with N-ethylpentadecamine the most active (<5 ppm). The N, N-dimethylamide of dodeca- and tridecanoic acid were active (LDioo 5-10 ppm) against P. rediviviis. These compounds (20 ppm) effectively controlled infection of tomato seedlings by M incognita [74]. In this context, it is interesting to note that N, N-dimethylocta-decylamine has been used as an antinematode [75]. Long chain amines and alkylamides showed only moderate activity against the dog round-worm (r. canis) L2 larvae [76]. Claims that decanol is a general nematocide which is not toxic to animals have been made in a patent [77]. Against B. lignicolus, decanol had weak activity (LD 18 16 ppm) [73]. Triacontanol (9) and triacontyl tetracosonate (10), isolated from Mucima aterrima, caused paralysis and death of M incognita juveniles. At 1% concentration, the compounds inhibited (100 and 50%) the escape of the adult insect from the pupa. Treatment of tomato plants with solutions of (9) and (10) eliminated the formation of galls on roots [78,79]. The triglyceride (11), from Argemone mexicana seeds, was active against M incognita [80]. OH CO2H

COsH

CH3

(CH2)28

CH2OH

CH3

(CH2)28

CHaO"

^{0Hz)22-

10

9 0\ Linoleoyr

Eicosa-9,12-dienoyl

11

-CH,

436 Polyynes The first polyacetylene with nematocidal activity was isolated from the roots of a Helenium hybrid. It was identified as the pentayne (12) and shown to have in vitro activity towards P. penetrans and other nematodes [31]. The 3'trans (13) diene and its S-cis isomer, obtained from the flowers of safflower {Carthamus tinctorius), were nematocidal (LD95; LDgo 10 |Ug/ml) to A. besseyi [81, 82]. The pentayne (12) and the epoxide (14) from the roots of the thistle {Cirsium japonicum) completely inhibited reproduction of B. xylophilus in the cotton ball bioassay at 16 and 250 |Lig/ball. The three acetylenes (15-17) were found to be active at 110 |Llg^all [83]. Roots of the daisy, Erigeron philadelphicus, contained (18) and its cis -isomer which showed activity (LD50 3 |ig/ml) against P. coffeae and another four natural compounds (17, 19, 20 and its cisisomer) were also active [84]. H3C

(C^=C)5

CHZIZCH2

12 DH CH3(CH2)2

r^

(C^C)2

y^^^^' CHI=CH2

14

(CzzzzQa—CH3

15 CH3

V

(C^C)3

CH3

16

N

- ^ ^ ^7

HfiO^^ 0

r - ^ 0

18

Several synthetic analogues containing only one triple bond were tested and demonstrated significant activity at levels <1 |Lig/ml [85]. Maximum activity was observed for compounds containing a ketone conjugated with a triple bond, aryl or an ester group. Falcarindiol (21), from the fresh roots of Angelica pubescens, was nematocidal to^. besseyi [69].

437

l,2Dithiins The thiarubrines are red 1, 2-dithiin polyynes that occur predominantly in resin canals in the cortex and periderm of certain species of the Asteraceae family, eg Ambrosia and Verbesina [86]. They are derived from acetate via the fatty acid biosynthetic pathway [87]. Since they are light sensitive compounds (^max 340-350, 480-490), subdued light is required for their isolation [86]. Visible light irradiation of solutions of thiarubrine A (22) leads to discharge of colour and the dithiin ring rearranges to the intermediates (23, 24) [88,89]. On standing, some of the original colour is reformed, but most of the intemiediates are converted to the thiophene (25) with loss of an atom of sulphur. The thiarubrines exhibit potent antimicrobial and antifungal activity in the dark, light-dependent antiviral activity, and appear to induce pulmonary oedema in mice which may restrict their use [86]. The more polar thiarubrines are toxic to fungi at lower concentrations than those that affect mammalian cells. Thiarubrine A (22) (5 ppm) killed C. elegans in the dark, but was more potent (0.03 ppm) after exposure to light [90]. Thiarubrine C (26) was active (LD50 12.4, 23.5 ppm) towards A/, incognita and P, penetrans Exposure of M mcog/i/Va juveniles to (26) (20 ppm)greatly reduced their ability to infect tomato plants [91]. Polythienyls In studies investigating the resistance of Tagetes to Meloidogyne species, the extract from the roots of Tagetes species was found to be more active than those obtained from other parts of the plants. From this fraction, an active compound was isolated and shown to be identical to a-terthienyl (3). Subsequent work led to the identification of a second compound (4). Both were toxic to H. rostochiensis (0.1 ppm; 100 ppm), D. dipsaci (5

438 ppm; 500 ppm) and A, tritici (0,5 ppm; 12,5 ppm) and (4) was active (3.13 ppm) towards P, redivivus [92,93]. However, marigolds are not active against other nematodes, some of which use the plant as the host.

(rF=c)2

HX"

(CF=C)2

(rF=C)2

HqO

C=

. _ / /

CKzzrcHg

CK=CH2

CHZZICHa

23

24

25

\^

A chance observation revealed that the in vitro activity of a-terthienyl (3) was not expressed in the dark. Subsequent studies showed that light (K --350 nm) was required for activation. Moreover, it was found that a-terthienyl (3) was a singlet oxygen sensitiser and that it inactivated enzymes such as glucose-6-dehydrogenase and acetylcholinesterase [94-96], The ability of a-terthienyl (3) to function as a sensitiser to convert triplet oxygen to the singlet state accounts for its enhanced activity in the presence of light. It has been proposed that, in a plant root system, stimulation of peroxidase by damage to the roots leads to a cascade of reactions producing compounds that control endoparasitic nematodes [97]. In fact, peroxidase activity in Tagetes roots increased markedly in plants infected with P. penetrans [97]. Interestingly, some herbivorous insects,

439 Manduca sexta and Pieris rapae, are sensitive to topically applied a-terthienyl (3) in the presence of light, whereas others such as Heliothis virescens are insensitive. a-Terthienyl (3) also showed potent activity (LD50 19 ppb) towards the light-exposed water-living larvae of the mosquito Aedes aegyptii [98]. A study of the the metabolic and environmental fate of a-terthienyl (3) has been carried out [99], Metabolism in rats resulted in the formation of the two metabolites (27, 28), and the dithienyl acid (28) could be detected as a metabolite in mosquito larvae (Scheme 1). Under conditions of simulated sunlight, a methanol solution of (3) yielded (28), a number of 2-substituted thiophenes, eg (29, 30), their methyl esters, and thiophene 2, 5-dicarboxylic acid.

CH>^, // V 29

OH

// W

°

30

^

'

28

Scheme 1. Products of metabolism and photolysis ofa-tcrthienyl (3)

Furans and tetrahydrofurans The furans (31, 32), isolated from the basidiomycete Irpex lacteus, were shown to be nematocidal (LD50 50 ppm) towards A. besseyi [100]. The pentyl furan (31) was also isolated from an ascomycete and shown to have comparable activity (LD50 75 |Lig/ml and 50 |ig/ml) towards C. elegans and M, incognita [101].

440

35

g

36

H

,CN

37

Nothenia anomala, a brown alga which occurs as an epiphyte on another brown alga Hormosira banksii, produces the /ran^-dihydroxytetrahydrofuran (33) and the corresponding cw-diastereomer (34) as major and minor metabolites. The Z^/j-epoxide (35), a known metabolite of A^. anomala, is almost certainly a biosynthetic precursor. In fact, acid catalysed conversion of (35) leads to a similar mixture of (33) and (34) [102]. The tranS'isomQX inhibited (LD50 1.8, 9.9 ppm) the development of eggs of T. colubriformis and K contortus to the infective free-living L3 stage. This level of activity is comparable to that of the commercially available nematocides levamisole (36) and closantel (37). A number of semisynthetic

441

and synthetic analogues were also tested to gain an insight into SAR. The following conclusions were reached: the most active stereoisomers were the CIO epimers of (34) and (35) (LD50 0.54, 0.41 ppm), the mono- and diketone-derivatives retained activity, but the corresponding acetates were inactive, and the terminal double bond is required for high activity [102]. The activity was not manifested in vivo, but this is not surprising since deUvery of such hydrophobic compounds would appear to present problems. The Annonaceous acetogenins have attracted considerable attention because of their antitumor and pesticidal activity. These compounds, found only in the plant family Annonaceae, are essentially derivatives of fatty acids [103]. Their bioactivity is expressed by depletion of ATP levels arising from inhibition of complex I in mitochondria and inhibition of NADH oxidase of plasma membranes of tumor cell. The consequence of ATP deprivation is apoptosis [103]. More than 350 acetogenins are known, all of them containing either C-32 or C-34 chains. A number have been described as toxic to mosquito larvae, corn borers, aphids, beetles and free-living nematodes. Since the acetogenins in a plant extract occur as complex mixtures containing up to 40 compounds, the development of resistance by pest organisms would be rendered more difficult, making such extracts attractive as preparations for pest control. A number of pure acetogenins have been tested for their antinematodal activity. Asimicin (38) was remarkably active (LDioo 0.1 ppm) towards C. elegans [104,105]. Annonin VI (C20 epimer of (39) was also reported to be active towards C elegans [106,107]. Other acetogenins showed strong activity against Molinema dessetae, a rodent filarial parasite. The compounds tested, and their activity after 1 and 7 days, were as follows; annonacin (39) (LD50 0.66, 0.08 |ig/ml), annonacin tetraacetate (10.2, 1.20), annonacinone (the 10-keto derivative of 39) (0.52, 0.28), corossolin (40) (5.12, 0.41), murisolin (41) (1.5, 0.30), cherimoHn (42) (0.67, 0.04), and otivarin (43) (6.66, 0.25). The activity of these compounds compares favourably with that exhibited by ivermectin (LD50 1.37, 0.27 |ig/ml) [108,109], So far, no studies on the biosynthesis of these compounds seem to have been carried out; perhaps a consequence of the difficulty of dealing with such complex mixtures. The acetogenins appear to be derived biogenetically from the polyketide pathway in which the tetrahydrofuran rings arise from epoxidation of isolated double bonds and participation of a y-hydroxyl [103].

442

OH

Annonacin 39

Cherimolin 42

OH

°" erythro trans threo

Otivarin 43

threo trans

Other fatty acid-derived metabolites Two compounds containing the a, P-unsaturated azoxy system were isolated from a Streptomyces sp. [110,111]. Compounds containing this functionality are widespread in nature and have beeen isolated from bacteria, fungi and higher plants [112]. Jietacin A (44) and B (45) were ten times more toxic (LDioo 0-25 |ig/ml) to B. xylophylus than avermectin [110,111]. Modification of the vinyl group in jietacin A (44) greatly reduced the activity, but the length of the alky I chain did not seem

443

to play a major role [113]. A number of synthetic compounds containing the azoxy functionality have been reported [112-115]. There is some evidence that for valanimycin the N-N bond is formed by the reaction of a hydroxylamine with an amine to generate an intermediate hydrazine [116]. However, in a fungal organism, the coupling of two hydroxylamino units seems to be preferred [112]. Sulfmemycin (46), a metabolite containing a rra«5-a,P-unsaturated thioamide S-oxide group from a Streptomyces sp., added (100 ppm) to the diet of gerbils reduced (52%) levels of T. colubriformis in 4 days [117].

<^^^N

< ^ ^ ;

44

45

46 Cardol (47), one of the phenolic lipids that characterise extracts of members of the plant family Anacardiaceae [118] showed pronounced antifilarial activity (LDioo 3.5 ppm) [119]. In rats, it was tolerated up to concentrations of 5 g/kg body weight. For activity, the phenolic hydroxyls and the alkyl side chain were necessary [119,120]. The related compound urushiol (48) showed toxicity to Ascaris suilla [121]. Immature lace bugs of the genera Stephanitis and Corythuca have secretory setae on their bodies, including antennae, which produce small droplets of colourless fluid. This secretion consists of a mixture of simple acetogenins. In attempts to define the role of these compounds in the biology of the lace bug, their bacterial, fungal and nematocidal activity was evaluated. Although the compounds showed minimal antimicrobial activity, four showed growth inhibition of larvae of Ascaris suum [122]. Compounds (49) and (50) (10 mM) decreased the motility of the larvae and, at the end of the assay (7 days), >80% of the larvae were immotile or

444

appeared abnormal in size. The effect of (51) and (52) was obvious only at the end of the test when >60% were found to be immotile. These effects were quantitatively similar to those induced by the anthelmintics levamisole (37) (1 mM) and parbendazole (53) (1 mM).

47

OH CoHii

'(CH2),4CH3

OH

48

49

Ci i H j

O

OH

C„H,,

51

50 C.H, -(CHj),

\\ OH

/

52

Polyketides Musacin C (54) cooccurs in Streptomyces griseoviridis with the cyneromycins, 14-membered ring lactones of heptaketide origin. Administered

445

to jirds, (54) (10 mg/kg) led to a 95% reduction of T. colubriformis, but the cyneromycins were not active [123].

The leaf tissue of mature plants of some eucalyptus species, eg E. grandis, produce amounts (up to 7.5 mg/g dry wt) of three plant growth regulators, the G-inhibitors (55-57), which are characterised by the presence of an epidioxy functionality. Since these compounds lacked optical activity, the possibility that they were artefacts was considered. In fact, the p-triketone (58) (Scheme 2) rapidly formed the epidioxide (57) on exposure to air. However, isolation of the inhibitors in the presence of ^^02 did not result in incorporation of isotopic oxygen. It has been suggested that (55-57) arise from the corresponding Mannich bases which are con-

446

verted to the endoperoxides in response to plant damage or to a biological stimulus. A probable biosynthetic scheme, proposed on the basis of in vitro transformations is illustrated for (57) in Scheme 2 [124].

NR1R2

Scheme 2. Proposed in vitro formation of the G-inhibitors The presence of the epidioxy moiety in the G-inhibitors suggested that they might exhibit ascaridole-like anthelmintic activity [125]. Moreover, Mannich bases are known to exhibit anthelmintic activity and it is interesting to note the similarity between the G-inhibitors and artemisinin (59), a highly effective antimalarial agent [126]. Parasites are sensitive to free radicals and have a limited capacity to cope with them since they often lack catalase activity and their glutathione metabolising enzymes have low activity [127]. Fresh leaves of E. grandis were fed for 7 days to goats in-fected with H. contortus. An autopsy showed that the treatment group harboured 92% less parasites than the control group. Of the synthetic analogues of G3 and the Mannich bases tested against H. contortus L3 larvae, the most active was (60) (0.08 mM). Although it had little activity in vivo to K contortus (mainly due to the difficulty of delivery), it reduced infection by N. brasiliensis in mice by 73%. An isolate of the ascomycete Lachnum papyraceum, collected in southem Germany, produced more than 30 metabolites. These include a number of metabolites of pentaketide origin, some of which have incorporated an isoprene unit [6,53]. The two major metabolites, lachnumol A (61) and

447

lachnumon (62) showed significant activity towards C elegans (LD50 5, 25 |ig/ml), but little against M. incognita [128,129]. By way of comparison, ivermectin showed LD50 0.1 and 2 |Lig/mI. In an attempt to produce brominated analogues of these compounds, CaBr2 was added to the culture medium. In this case, lachnumol Bl (63) and B2 (64) were obtained and were shown to be active (LD50 50, 25 |Lig/ml) towards C. elegans but not to M. incognita [130-132].

H,C0

H3C0

OH

CI

61

63R = H 64 R = Br H,C0' OH

Br

OH

CI

Scheme 3. Hypothetical pathway for the formation of the mycorrhizins

448

66 R = CI (1 ^g/ml) 67 R = H (2Sml) 69 R= Br(2Mg/ml)

^„ „ ^. ,. , „ ^8 R = CI(1ng/ml) 70 R = Br (5ng/ml)

73R = H (50^g/ml) 74R = CH3(50^g/ml)

71(2ng/ml)

75R = H (50ng/ml) 76R = CH3 (50ng/ml)

H

HO b

CI

77 (50 ^g/ml)

78 (50 ^ig/ml)

79 (100 ^g/ml)

Fig. (I), Activity (LD50) of mycorrhizin derivatives towards C eiegans

PhSH Et^N

CH3O

o

81 In media containing CaBr2, the two brominated analogues (69, 70) as well as (71) were obtained. These metabolites were strongly active to C eiegans (LD50 2-5 |ig/ml) but only weakly towards M incognita. Moderate activity was found for a number of other metabolites (72-79), Fig. (1). The predisposition of these compounds containing a cyclohexenedione system to react with nucleophiles was illustrated by the fact

449 that treatment of (80) with thiophenol and triethylamine gave compound (81) in good yields [135]. Compounds (66, 67, 69) also formed adducts with cysteine, indicating their potential to undergo (J-addition with bionucleophiles. This observation provides a rationale for the high cytotoxicity and strong inhibition of DNA, RNA and protein synthesis shown by these compounds [136]. Isolates of the nematode-trapping fungal species Arthrobotrys oligospora produce a cocktail of metabolites containing variations of a prenylated epoxydon structure. Oligosporon (82) was inactive towards C elegans [137], but (82) (LD50 25 |Lig/ml) and the 4', 5'-dihydroanalogue (LD50 50-100 |ig/ml) were toxic to the intestinal parasite K contortus. Variation in the stereochemistry of the allylic hydroxyl in the ring, or the oxidation level in the ring or side-chain, resulted in loss of activity [138]. 82 CH.OAc

O.X'

83

^0-^0

.a I

84 Cladobotryum rubrobrunnescens (Hyphomycetes), a mycophilic deuteromycete, produces cladobotrin (83) which showed weak activity against M incognita (LD50 100 l^g/ml) [139]. Cladobotrin had previously been isolated from C varium and had been shown to have antifungal activity to Ganoderma lucidum [140]. Aspyrone (84), a metabolite of Aspergillus melleiis, showed moderate activity (LD39 100 |Lig/l) towards P. penetrans. Although the related metabolites mellein, penicillic acid and dihydropenicillic acid had little effect on P. penetrans [141], patulin (85)

450

was active towards M incognita [6] and penicillic acid (86) showed weak activity towards Anguillula aceti [142].

90 R = H 91 R = OH Species of the genus Zanthoxylum (Rutaceae) are reputed to have anthelmintic activity and some are known to contain ascaridole (6). Z liebmannianum has been used in Mexico to treat intestinal parasites. Sanshool (87), a hexaketide-derived polyene isobutylamide isolated from this plant, was toxic to Ascaris suum (LC50 206 |ig/ml), but it also induced convulsions in mice at 55 mg/kg [143]. Eliamid (88), a metabolite oi Sporangium cellulosum, has been claimed to be nematocidal [144]. Griseulin, an aromatic nitro compound which showed in vitro activity (LDioo 1 ppm) towards a number of nematodes, including P. redivivus, was isolated from a Streptomyces sp. [145]. The structure initially assigned to this compound [146] was incorrect and was later revised to (89) on the basis of synthetic studies [147]. The azaphilones are a large group of fungal pigments isolated mainly from perfect and imperfect stages of Ascomycetes such as Aspergillus and Penicillium, On exposure to ammonia or primary amines, the azaphilones develop a red colour due to exchange of the pyran oxygen by nitrogen. The new azaphilones (90, 91), isolated from Bulgaria inquinans (Asco-

451 mycetes), showed activity (LD50 5, 10-25 |Lig/ml) towards C. elegans. They inhibited the binding of the dopamine antagonist 3H-SCH23390 to the Dl receptor in rat brain striatal homogenates, but not the binding of ^H-kainic acid to the glutamate receptor subtype [148]. Maduramicin (92), a polyether ionophore, has been found as a metabolite of Nocardia and Actinomadura species [149]. Its structure was determined by X-ray crystallographic studies and an investigation of its biosynthesis revealed that it incorporated 8 acetate, 7 propionate units with 4 methoxy methyls derived from methionine. It displayed antimalarial [150] and nematocidal activity (0.5-30 ppm) towards phyto- and animal-parasitic nematodes [151]. H3C0.,

H3C0 H3G,

n I •

0-^1

CH3

^o4~

CH3

92 The hexaketide-derived 10-membered ring lactone, lethaloxin (93), was active (LD50 25 |Lig/ml) towards C. elegans [6]. This compound is produced by Mycosphaerella lethalis, a pathogen of sweet clover [152]. Chlonostachydiol (94), a 14-membered ring macrodiolide from Chlonostachys cylindrospora, exhibited potent activity towards the abomasum nematode, K contorsus. Subcutaneous injection of (94) (2.5 mg/kg) in artificially infected lambs caused an 80-90% reduction of nematodes [153,154]. Helmindiol, a symmetric 16-membered macrodiolide isolated from an Alternaria alternata strain was shown to have anthelmintic properties. In vivo testing (2.5 mg/kg) with lambs infected with //. contorsus, resulted in reduction (50%) of nematode levels over 14 days. Surprisingly, both helmindiol and clonostachydiol (94) could be shown reproducibly to be active with one group of lambs but not a second group. It has been suggested that these differences might reflect variations in metabolism in the test groups [155]. Although the stereochemistry of

452 helmindiol was not fully assigned, it appears to be the same as that established for pyrenophorol (95) [156]. Brefeldin (96), an octaketide macrolide with antiviral, cytotoxic, cancerostatic and phytotoxic activity, was active (50 |Xg/ml) towards Anguillula aceti [142, 157]. Radicicol (monorden) (97), a macrolide antibiotic produced by Deuteromycetes, possesses cytotoxic, antiprotozoal and antineoplastic activity [158]. The diethyl ether derivative showed weak activity (LD50 0.2 mg/ml) against a soil nematode, but the cytotoxicity was considerably higher (LD50 3.1 |ig/ml) [159].

93 (relative stereochemistry)

453 Elaiophylin (98), a metabolite of several Streptomyces spp, has a 16membered unsaturated ring like the avermectins. It belongs to a small group of macrodiolides with C2-symmetry [160,161]. It is acid and base sensitive because of the facility of a retro-aldol cleavage at C9/C10. Interest in this compound arose from its promising antibacterial, antifungal, cytocidal and anticoccidial activities. Elaiophylin has also been used as a starting material for the syntheses of nematocidal compounds. A study involving 40 of its derivatives identified a number which were antinematodal. In particular, the 11,11'-dimethoxy derivative and some unsymmetric deglycosidation products showed activity (100 ppm) towards C elegans [160]. This compound has recently been found as a natural product and does not seem to be an artefact of elaiophylin [162]. The bafilomycins, a group of 16-membered macrocyclic lactones bearing a pendant side chain at the lactone terminus, have been isolated from several Streptomyces species [163]. They are members of the hygrolide family of macrolide antibiotics and are of interest as potent vacuolar ATPase inhibitors that possess a broad spectrum of over-lapping biological activity against bacteria, fungi, insects, cestodes and nematodes. A study of the biosynthesis of bafilomycin Al (99) revealed the incorporation of seven propionate units and four acetate residues, the isopropyl group in the tetrahydropyran ring presumably arising from valine. The unexpected oxygenation at C2 and C14 may suggest involvement of glycollate or glycerate, rather than acetate, at these positions. The absolute configuration of this group of compounds has been established [163] and recent synthetic efforts demonstrate the interest in the bafilomycin [164], Bafilomycin Al (99) showed comparable activity to that of the avermectins in stimulating GABA release from rat brain synaptosomes [165]. Bafilomycin CI (100) was noted as a weak inhibitor (LD50 32.5 |Lig/ml) of C. elegans motility [166] and bafilomycin Bl (101) (LD50 10 |Lig/ml) and leucanicidin (102) were also active (LD50 10 |ig/ml) to B. lignicolus [167,168]. In a more detailed study, the bafilomycins (99-102, 103) were found to paralyse LI larvae of H. contortus and to be lethal within 24 hr (LD50 from 0.23 |Lig/ml for (102) to 2.5 |ig/ml for (103)) [169]. Moreover, these compounds also inhibited the development of T colubriformis and O. circumcinta larvae. Although 700 -fold less potent than ivermectin, bafilomycin B2 (19-methoxy-Bl) exhibited similar potency to many avermectins and milbemycins as an inhibitor of L3

454

motility. Nematode strains resistant to the anthelmintics levamisole (36) and avermectin showed no cross-resistance to the bafilomycins, suggesting that the latter operate through a different mechanism.

102 R =

„a

'"•?

Faeriefungin, a polyol polyene macrolactone, was isolated from the mycelium of Streptomyces griseus var autrophiciis. It exhibited antifungal (MIC 3-12 |Lig/ml), antibacterial (MIC 16-32 jiig/ml), insecticidal {Aedes aegypti) and nematocidal activity {P, redivivus ; LC50 100 |i g/ml). Faeriefungin is a mixture of two homologues (104, 105) [170]. Three other compounds with more potent nematocidal activity (LDioo 0.1-1.0 )ig/ml) against ?. redivivus, C. elegans and H, glycines, were isolated from the same organism, but their structures have not been disclosed [171] The crude ethanol extract of two Amphimedon spp. inhibited (LD50 4.0 mg/ml) larval development of the free-living stages of //. contortus, Bioassay-guided fractionations yielded the amphilactams (106-109), a group of novel macrocyclic lactone/lactams. These metabolites affected the LI stage (LD99 7.5, 4.7, 8.5, 0.39 |ig/ml) of the nematode but had little or no effect on the eggs. The in vitro activity observed for (106),

455 (107) and (109) is comparable to that exhibited by the commercial anthelmintics levamisole and closantel.

103

Comparison of the structures of (106) and (108) suggests that the A2»3 geometry of the unique enamino lactam/lactone moiety plays no part in modulating the activity. However, for (107) and (109) a change from Eto Z- geometry results in 100-fold increase in activity [172],

H3C0

108 R = H 109 R = CH3

456

Terpenoids Monoterpenes The first monoterpene found to have anthelmintic activity was ascaridole (6) which, alone or as a component of plant extracts, has been used in the treatment of hookworm infection [37]. Ascaridole is also a potent in vitro inhibitor (LC 0.05 fiM) of the development of the malarial parasite Plasmodium falciparum [173]. Thymol also exhibits anthelmintic activity, but is too toxic to be considered for use in animals and humans [37]. Since many monoterpenes taken internally in high doses are known to cause problems (irritation, acute toxicity) [174], more attention has been paid to evaluating these compounds against phytoparasitic nematodes. Thymol exhibited strong antinematodal activity against the phytoparasitic M. arenaria, H. glycines, Paratrichodorus minor and free-living dorylaimid nematodes. When added to soil (25-250 ppm), population densities of the nematodes and disease incidence declined [175]. Other simple monoterpenes have been implicated as antinematodal agents, particularly when used as soil amendments: citronellal [176]; citral, menthol, a-terpineol [177]; geraniol [178] and limonene [179]. Menthol, geraniol, linalool and cineole, and a number of esters of menthol and geraniol, were toxic to the L2 stage of the phytoparasitic nematodes, Anguina tritici, M. Javanica and Tylenchulus semipenetrans [ 180,181 ]. 1, 2-Dihydroxymintlactone (110), the only monoterpene isolated from the wood-inhabiting fungus Cheimonophyllum candidissimus showed nematocidal activity (LD50 25 |Lig/ml) towards C elegans [182].

b

111

457 The anthelmintic activity of the seeds of Butea frondosa, which are reputed to be an effective treatment for roundworm infection in the Ayuverdic system of medicine, was attributed to palasonin (111) [183]. Palasonin bears a striking similarity to cantharidin (112), an insect derived vesicant which plays a role in the mating behaviour of pyrochroid beetles [184]. The biosynthetic origin of the two compounds is obscure, although they might best be considered as modified monoterpenes. The anthelmintic action of palasonin on Ascaridia galli possibly involves inhibition of glucose uptake and/or of the motor activity of the parasite [185]. Palasonin showed larvicidal activity towards the vertebrate parasite K contortus [186]. Recent interest in these compounds has been stimulated because of their activity as inhibitors of protein phosphatase 2A [187,188] and has resulted in two syntheses of palasonin [189,190]. Sesquiterpenes The macrocyclic sesquiterpene, a-humulene (113), is repellent to B, xylophilus [191]. The turmeric {Curcuma longa) component, (+)-ar-turmerone (114), was shown to be effective (25 |ig/ml) 2i%?imsX Anisakis larvae [192]. A group of bisabolanes (115-119), isolated from Cheimonophyllum candidissimus [193], were weakly antimicrobial but exhibited strong cytotoxic effects. Cheimonophyllon A (115) and D (116) (LD50 10 |ig/ml), B (117) and cheimonophyllal (118) (LD50 25 )ig/ml) showed moderate activity towards C elegans whereas C (119) was less active (LD50 50 Hg/ml). However, (115), (117 ) and (119) were not active (100 mg/ml) towards M incognita [193, 194]. Ivermectin, the standard used in this assay, showed LD50 0.1 |ig/ml. It has been suggested that the loss of biological activity of these compounds after incubation with S-cysteine implicates the a, P-unsaturated carbonyl group as the toxophore [194].

114

458

116R = 0H 117R = H

115

HO

119

118 OH

Fumagillin (120) was first isolated fifty years ago from Aspergillus fiimigatus and was later found to exhibit potent amoebicidal and anti microsporidial activity [195-197]. It is toxic to humans and is only used topically to resolve microsporidian-induced ocular infection [198]. 120 R = CO2M

V

CH30

121 R =

[1 CI

\ '

459 Recently, fumagillin (120) and its semi-synthetic derivative (121) have attracted considerable interest because of their oustanding ability to inhibit the proliferation of endothelial cells in vitro and tumor-induced angiogenesis in vivo [199]. Fumagillin (120) may act by irreversibly inhibiting MetAP-2, a cobalt-containing metalloprotease that removes methionine residues from the N-termini of proteins [200]. In fact, it has been shown that the ring epoxide is involved in the inhibition, whereas the side chain epoxide is dispensable [201]. A crystal structure of inhibited human MetAP-2 showed a covalent bond formed between the methyleneoxy carbon and His-231 in the active site of the enzyme [202]. Fumigallin has been reported to be moderately active (50 |ig/ml) as a vermifugal towards the free-living nematode Angiiillula aceti [142]. It would be interesting to test the more water soluble fumidil B (the bicyclohexyl ammonium salt) and the semisynthetic TNP-40 (121) against a wider range of nematodes. This surge of interest in the activity of fumagillin has resulted in a new synthesis of fumagillol (122) [203]. The biosynthetic pathway suggested by Birch [204] on the basis of preliminary studies still awaits confirmation and amplification. Studies on the biosynthesis of the related ovalicin (123) have provided additional evidence for the intermediacy of P-bergamotene (124) [205] in accordance with the original proposal for fumagillin (Scheme 4). 1

. ^ ^ HXTN

B

v\

xir

' T]

^ ^ ^ ^ / N ^ ^ ^

1

^

.-^^ - ^

124

1 1

_ H

1 1


1

1

11 _/_

-^^

y ii
^ Scheme 4. Hypotlietical derivation of parent skeleton of fumagillin

1

460 The cochlioquinones are a group of fungal metabolites from Cochliobolus miyabeanus that have their biosynthetic origin from the combination of famesol and a hexaketide moiety [206]. CochHoquinone A (125) exhibited moderate effects (70 |Lig/ml) on the motility of C. elegans. Interestingly, it was shown to compete for the specific binding site of ivermectin [207]. The 14-epidihydrocochlioquinone B (126) showed stronger activity (10 jag/ml) towards C. elegans but was ten times less active towards M. incognita [208]. For C elegans, the change in stereochemistry at C-14 was significant. The cochlioquinones are well known as phytotoxins [209] and (125) is an inhibitor of diacylglycerol kinase which regulates the activity of protein kinase C [210]. The resistance of plants to pathogens is mediated by a variety of mechanisms, either induced or constitutive. The induction of phytoalexins in plants in response to biotic and abiotic factors has been studied intensively. The strategy of engineering plant resistance by the introduction of appropriate genes coding for phytoalexin expression is an exciting prospect which is limited by the availability of the required genes [211]. In this context, it is interesting to note the antinematodal activity of two sesquiterpene phytoalexins. Rishitin (127), a sesquiterpene phytoalexin formed in potatoes, showed antagonistic activity towards Xiphimena diversicaiidatum [212]. The nematodes became agitated and were repelled by a point source of rishitin (20 |Lig), and a solution of 200 |Lig/ml inactivated the nematodes, killing them after 2 hr. At 100 |ag/ml, rishitin inhibited the movement of Ditylenchus destructor and levels of rishitin in potato tubers were correlated with resistance to the nematodes [213].

461

HO,,^

127

128

Solavetivone (128), a biosynthetic precursor of rishitin, became the dominant sesquiterpene (75-98% of total sesquiterpenes) in some potato clones, whereas in standard cultivars it represented less than 15% of the sesquiterpenes. The clones producing high levels of solavetivone have been found to carry the HI gene that confers resistance towards G. rostochiensis which attacks the roots of the potato plant [214]. Circumstantial evidence suggests that the gene(s) controlling solavetivone production are located on chromosome V close to the HI locus for resistance to the nematode. A number of terpenes containing an a, P-unsaturated 1, 4-dialdehyde group have been implicated in the defence mechanism of terrestrial and marine plants, fungi, molluscs, sponges and termites [215-217]. Many have been found to have repellent or antifeedant effects on potential predators, but only a few have been tested for antinematodal activity. Polygodial (129), a hot-tasting sesquiterpene, shows diverse activities including antifungal and insect antifedant activity. It was weakly active (IC50 0.07 |Lig/ml) against the infective L3 stage of T. colubriformis [60]. Fruiting bodies of the pungent Russulaceae species of fungi {Lactarius and Russula) contain the stearoyl ester of velutinal (130). On damage to the fruit, the ester is enzymatically transformed in seconds to the pungent dialdehydes isovelleral (131) and/or related sesquiterpenes, eg (132), depending on the species [218]. Isovalleral is weakly toxic (LD50 50 |Lig/ml) to C elegans. In the conversion of (130) to isovalleral, stearic acid is generated. It has been suggested that this might contribute to the nematocidal activity [219]. In fact, stearic acid is significantly active (LD50 50 |Lig/ml) towards C elegans [68]. The mild-tasting Lactarius mitissimus (Russulales) produces three sesquiterpenoids (133-135) with weak activity (100 |ig/ml) to C elegans [219]. In a number of the Agaricales, nematocidal activity increases in response to injury and appears to be associated with the production of linoleic, S-coriolic acid and other fatty acids with known nematocidal activity [219].

462 .

OCO(CH2),(,CH3

lX\ (

CHO

,,/-4

/ 129

130 PH

CHO CHO

132

133

OH

134 R = H 135R = OH

2

""CHO

136

137 R = H 138R = OH

139

The structurally related hydroxy butanolide, marasmic acid (136), found in Marasmius, Lachnella and Peniophora species, also showed weak activity to M incognita (LD60 100 |Llg/ml) but not to C. elegans [6]. The simplistic explanation that the bioactivity of the unsaturated aldehyde may, to a large extent, arise from the reaction of the 1, 4dialdehyde group with amino acids such as lysine on receptor molecules does not appear to hold. For example, polygodial (129) is ten times more reactive in vitro with lysine than warburganal which shows ten times the bioactivity of (129). However, in vivo, factors such as shape can overide the reactivity operating in vitro. Santonin (137), a potent anthelmintic, has been used to treat ascaridiasis and oxyuriasis, but several cases of fatal poisoning by santonin have been reported. Small doses may produce effects on vision, cause headache, nausea and vomiting, and higher doses, epileptiform convul-

463 sions [37]. Santonin (0.1%) does not kill Ascaris, but acts on the ganglion cells of the nematode to induce paralysis and the parasite is eliminated via the feces [220]. In a recently developed bioassay, santonin was found to have LD50 40 |Lig/ml [63]. Synergistic effects of santonin and a-kainic acid towards mice infected with Symphacia obvelata have been observed [221]. Importantly, 1, 2-dihydrosantonin, which lacks the a, p-unsaturated ketone in the A-ring, showed complete lack of activity [222]. Artemisin (138) is another sesquiterpene lactone reported to have anthelmintic properties [223]. Alantolactone (139) and related sesquiterpenes are active against M. incognita juveniles (L2 stage). Maximum toxicity was associated with an a-me thy lene-y-lactone moiety; modification of this group reduced the activity [224], Alantolactone (139) (0.05%) killed Ascaris in 16 hr [225]. Extracts of the sponge Dysidea herbacea showed potent activity towards the vertebrate parasitic nematode N. brasiliensis. Bioassay guided fractionation yielded furodysin sesquiterpenes (eg 140). Follow-up testing with mixed N. dubius and Hypselodory nana revealed that the sesquiterpenes were weakly active[226]. These compounds, as single enantiomers or mixtures of enantiomers, administered (200 ppm) for 18 days to mice infected with N. dubius reduced infection by 59% [227]. The terpenoid aldehyde fraction (50 |ig/ml) from Gossypium hirsutum inhibited the motility of M incognita, Gossypol (141) was one of the active components (125 |Lig/ml) [228]. The glandular terpenoid aldehydes constitute an important component in the cotton plant's defense to insects [229].

140 Sponges of the genus Axynyssa have been reported to produce a wide variety of sesquiterpene metabolites containing isonitriles, isothiocyanate and thiocyanates groups. Most of the interest in these compounds has concerned the biosynthetic origin of the carbon of the thiocyanate group.

464

It appears that inorganic thiocyanate and cyanide is utilised for isocyanide biosynthesis and that organic isothiocyanates can also be precursors [230,231]. NCS

SCN

142* 143* A^ * absolute stereochemistry unknown

144*

145*

Compounds of this type are produced by members of the order Axinellida, Halichondrida and Lithistida. Prescreens on samples of the first two orders showed that extracts had activity towards the parasitic stage of A^. brasiliensis, Bioassay-guided fractionation led to the isolation of a number of active sesquiterpenes (142-144) from Axinyssa fenestratus and a Thopsentia sp. (145). No specific details of the relative activity were given [232]. In view of the antinematodal activity of the isothiocyanates derived from the glucosinolates {vide infra), the activity of these sesquiterpenes is not surprising and suggests that tests on other members of this series could prove fruitful. The trichothecenes are a group of of sesquiterpenes that have attracted attention because of their involvement in toxicoses of farm animals and humans. They are produced by several fungi, eg Fusarium, Myrothecium, Trichothecium, that are parasitic on cereals. Trichothecin (100 |ig/ml) decreased the motility of Anguillula aceti [142]. Verrucarin A and T-2 toxin (2 ppm) reduced the viability of M, javanica, and T-2 toxin (0.2, 2 ppm) was toxic to M. hapla and Pratylenchus neglectus [233]. Given the toxicity of these compounds to plants and animals, they do not appear to be candidates as useful nematocides. Diterpenes Remarkably few diterpenes appear to have been tested for antinematodal activity. It would seem worthwhile to evaluate the activity of

465

the clerodane diterpenes, some of which show remarkable insect antifeedant properties [234]. Odoracin (146) and odoratrin (147), from the nemato-cidal extracts of the roots of Daphne odorata, showed potent toxicity (LD70 and LD96 1 ppm) towards A, besseyi. The diacetate and dibenzoate were inactive, but the desbenzoyl derivative of odoracin retained activity [235, 236].

146

147 Ri 148 149 150 151

R2

=CH2 NC CH3 CK3 NCS CH3 NC

A group of diterpenes obtained from a number of marine sponges are characterised by the presence isonitriie, isothiocyanate and/or formamido functionalities. Over forty examples have been recorded and are generically named the kalihinols. The absolute configuration of these diterpenes has recently been established [237]. They exhibit a broad spectrum of biological activities, including antimicrobial, antifungal, anthelmintic, cytotoxic and antifouling properties [238]. From the Fijian sponge Acanthella cavernosa, a number of kalihinols were isolated by bioassay-guided fractionation using the parasitic stage of iV. brasiliensis. Kalihinol Y(148) at 50 |i g/ml was reported to be extremely active and

466

(149-151) were active [239] at the same concentration. Two related diterpenes isolated from the sponge Thopsentia sp also showed activity in the same assay [232], The in vitro antimalarial activity of some of these compounds is noteworthy. For example, kalihinol A (149) was found to have a potent (EC50 1.2x10"^ M) and selective activity towards Plasmodium falciparum [238]. Triterpenes and sesterterpenes Oleanolic acid has been reported to be active towards C elegans [240] and ursolic acid to have anthelmintic activity [241]. The diol acid (152) was active (50 |Lig/ml) towards C. elegans [242]. The bark of a Melia cultivar (Meliaceae), used against infection by Ascaris nematodes, contains the limonoid 28-deacetyl sendanin (153) as the active principle. The anthelmintic effect is similar to that of santonin, but the compound is less toxic and, unlike santonin, does not need to be coadministered with a cathartic agent [243]. The leaf and kernel extracts and oil cake of Azadirachta indica (neem) showed nematocidal activity against a wide range of nematodes of agricultural importance [33]. One of the components, the potent biopesticide azadirachtin, had significant antinematodal activity towards M. incognita [244, 245] and B. xylophilus [246], Three quassinoids from the seeds of Hannoa (Quassia) undulata, applied (5.0 |ig/ml) as a mixture to tomato roots for 5 days, completely inhibited penetration by M. javanica [247] and reduced (1.0 ppm) reproduction [248]. The first step in the life cycle of nematodes towards parasitism is hatching of the eggs. In many phytoparasitic species, the hatch rate is enhanced by specific plant root exudates. This dependence is a possible target for control strategies. Glycinoeclepin A (154) is capable of initiating hatching in eggs of//, glycines at concentration as low as 10*^2 g/ml. It occurs in trace amounts and only 1.25 mg were obtained from 1058 kg of kidney bean roots [249], The structure was determined from spectroscopic data and X-ray diffraction analyses [250] and confirmed by three different total syntheses [251]. 12-Desoxyglycinoeclepin has been synthesised from the abundant cycloartane spirolactone, abietospiran in a sequence that chemically emulates the presumed biosynthetic route [252]. Two analogues of (154) which retain the side-chain were also isolated, but showed no activity to levels of lO"'^ g/ml [253]. Model compounds

467

containing an a, p-unsaturated-y-keto acid (eg 155) were prepared and found to inhibit egg hatching of//, glycines by 87% after 14 days [254]. The Soianaceous steroidal glycoalkaloids a-tomatine (156) and achaconine (157) stimulate hatching of Globodera spp eggs by a mechanism which involves membrane bilayer destabilisation resulting from the formation of glycoalkaloid/sterol complexes [260]. They also showed toxicity (ED50 50, 85 |ig/ml) to P. redivivus [261], The related solamargine showed in vivo activity against adults and microfilariae of Setaria cervL Oral administration of solamargine (100 mg/kg) to rats infected with S. cervi, reduced the blood count by more than 30% and after four treatments elimination of the parasite was achieved [262]. A number of triterpenoid saponins show antifilarial activity [255], although they exhibit low toxicity in vitro. Acaciaside A and B, saponins isolated from Acacia auriciiloformis, administered orally (100 mg/kg) to adult rats with implanted S. cervi decreased (<80%) blood microfilariae following the third phase of treatment [256]. The saponins showed weak activity towards M incognita [257]. Two cardenolides asperoside and strebloside, obtained from the stembark of Streblus asper, a plant used in Indian traditional medicine, showed antifilarial activity in vivo and in vitro towards macrofilaria [258]. These agents interfere with glutathione metabolism which ultimately results in the death of the parasites [259]. AcQ

COjH

HO^

'U

CO2H

153

C0,H

155

468

156 D-Xyl(P 1 —3)D-Glu( pi— 4)D-pGar

I

D-Glu(pi-^2) L-6-deoxyMann( a 1-^-4) L-6-deoxyMann( a 1 -^2)

N

D-pQu

157

The ophiobolins are metabolites of the pathogenic fungi Dreschlera spp and are potent phytotoxins. Ophiobolin C (158) and M (159) inliibited (LD50 5, 13 |LiM) the motility of C elegans, the epimers at C6 being an order of magnitude less potent (130 |iM). They are non-competitive inhibitors of ivermectin, binding to membranes obtained from C elegans [263]. Ophiobolin K (160), which contains an extra double bond in the side chain, was active (ED50 10 |Lig/ml), but its C6-epimer was not [264].

158 159 A^^ 160 A^^

Shikimate-derived compounds The shikimate pathway is well developed in plants, producing aromatic amino acids and a large range of secondary metabolites [265]. Several of these mediate activation of genes in root nodulation (isoflavanoids, pterocarpans) and many play a role as phytoalexins [266]. Resistance mechanisms in plants often involve metabolites from this pathway [267].

469 Pyrocathecol, isolated from the root exudate of Eragrostis curvula (Gramineae), was highly toxic to root-knot nematodes [268]. Salicylic acid and 4-hydroxybenzoic acid showed nematocidal activity towards C elegans and M incognita and were effective in drench applications [269]. A number of other simple aromatic compounds were toxic to, and suppressed egg-hatch of, M incognita. Of these, the more active were ^-cinnamic acid, 1, 2, 3-trihydroxybenzene and 2-hydroxynaphtoic acid [270]. From the culture broth of Irpex lacteiis, the methyl phenoxyproprionate (161) was isolated and found to be active (LD50 25 |Lig/ml) against A. besseyi [271]. A number of synthetic analogues were prepared and tested, but none showed greater activity [272]. Eugenol (162) showed toxicity to A, tritici, T. semipenetrans, M. incognita and H. cajani [181]. The five propenylphenols (163-167) from the leaves of Piper betle, at 200 |Lig/ml, killed C elegans. The high levels of these compounds in the leaves could constitute an effective nematocidal effect [273]. In another study, chavicol (163) was found to be active (300 |Lig/ml) towards C elegans, but the demethoxy analogue (164) was much less active (600 |ig/ml) [274]. Using microspectrofluorometry and image analysis, it was found that (163, 164) caused a significant increase in levels of Ca^"^ in the intestinal tract and destruction of the cell membrane, allowing leakage of cytosol from the intestinal tract into the pseudocolelomic cavity. The compounds induced a higher level of Ca^"^, but for a shorter period, than DOPA [275]. CO2CH3

162R=:OCH3 163 R = H

H3C0

164 R = H 165R = Ac

RO'

166 R = H 167R = Ac

H,C0'

OCH CO2CH3

H5C0"

"Y" OCH,

168 E

169 Z

170 0CH3

470

Acorus gramineus (sweet-flag) has long been used as an insecticide [276]. The hexane extracts of the plant showed in vitro nematocidal activity. The activity was shown to be due to a- and p-asarone (168, 169) [277], compounds known to have mammalian toxicity and carcinogenicity and not suitable for use as pesticides [278]. A mixture of the isomers was toxic (MLC 250 |Lig/ml) to T. canis (L2) [279]. Methyl ferulate (170) was moderately active (LC46 100 \x, g/ml) towards B. xylophilus [280]. The naphthalene derivatives (171, 172), from Daldinia concentrica (Ascomycetes), were significantly active towards C elegans (10, 25 |LLg/ml) and also showed antimicrobial and cytotoxic activity [6]. Since the black fruiting bodies of D. concentrica are highly melanized, it is likely that these compounds arise from melanin biosynthesis. Some naphthoquinones have been found to inhibit the motility and survival of H. contortus (LI). Plumbagin ( 173) was the most potent (ED 1 |Lig/ml) followed by 1, 4-naphthoquinone, juglone (174) and 1, 2-naphthoquinone [281].

171 R = H 172R = CH3

y ^ ] f i„ y

173R = CH3 174 R = H

176 R = CH(CH3)2 177 R = CH(CH3)CH2CH3 178

471

^^^::^

180

A number of insect growth regulators are biologically active in parasitic and free-living nematodes. Certain steroids inhibit growth and reproduction, sesquiterpene insect juvenile hormones and their analogues disrupt development of a wide variety of nematodes, and non-terpenoid juvenile hormone mimics inhibit egg hatch. The insect antijuvenile hormone precocene II (175) was active against C remanei, but its role as a nematode antihormone has not been demonstrated [282,283]. A number of precocene analogues containing a 5-methyl and different substituents at the 7-oxygen have been synthesised and found to be active (ED50 0.7812.3 mM) towards C elegans [284,285]. The precocene analogues (176, 177), isolated from Rhyncholacis penicillata, showed strong insecticidal activity (<2ppm) and moderate activity towards C elegans [286]. Cannabiorcichromenic acid (178) and its 8-chloro analogue have been isolated from Cylindrocarpon olidum, a fungus found on M. incognita, A mixture (4:1) of the two was active (LD50 20 |ig/ml) against Heterorhabditis nematodes [287]. The dihydroquinone (179) from the brown alga Desmarestia menziesii, showed in vitro activity (EC50 ~70 |lg/ml) towards//, brasiliensis [288]. ^-Chalcone and 115 of its derivatives exhibited high activity against the zooparasitic nematodes Syphacia obvelata, N, dubius, T. mystax, T. canis and A. caninum [289]. The greatest increase in activity was obtained by replacing the double bond by a triple bond to give diphenylpropynone. It has been suggested that the activity was mediated by the conjugated carbonyl and that this may act as an uncoupler of mitochondrial oxidative phosphorylation [290], Chalcone showed nematocidal activity (LC50 33 \\M) and inhibition of hatching of L2 juveniles of Globodera spp. (IC50 7 liiM) [291]. f'-cinnamaldehyde (LC50 24 ppm) and 4-phenyl-3-butyn-2one (2 ppm) were also active. B. xylophilus causes serious damage to the Japanese red pine, Pinus densiflora, and to the black pine, P, thunbergii. The L4 stage of the nema-

472

tode is carried by the sawyer beetle from dead to live trees where it changes into the propagative form. Some pines, eg P. massoniana, are resistant to the nematode. The heartwood of this species contained pinosylvin monomethyl ether (180) which was toxic (LCioo 10 Rg/ml) to the nematode, an effect that was not due to inhibition of acetylcholinesterase [280]. The stilbenes, 4-hydroxy-3'-methoxy-bibenzyl (130 |Lig/ml), Z-4Hydroxy-3'-methoxystilbene (LD50 60 |Lig/ml) and the £-isomer (LD50 70 |Lig/ml) rendered T, coliibriformis immotile [60]. Phytoalexins can be elicited in plants by a variety of agents, including viruses, microorganisms and nematodes [292, 293]. When resistant plants are infected by nematodes, phytoalexins with antinematodal activity can be produced [294]. Some coumestans have been implicated in the resistance of plants to nematodes. For example, when the roots of the resistant lima bean (Phaseolus lunatus) were inoculated with Pratylensus schbneri, coumestrol (181) accumulated at the site of nematode attack. In vitro, coumestrol (5 |Lig/ml) inhibited the motility of the nematode [294]. Correlative evidence for a functional role of related compounds in resistance towards nematodes has been obtained [295-297]. In particular, nematode attack on the roots elicits the transcription of genes encoding several enzymes of the shikimate pathway that leads to phytoalexin production. Induction of enzyme activity results from transcriptional activation of the genes leading to increased levels of translatable mRNA [298].

181

183

0

184 R = H 185R = OH

473

186 0CH3 OCH,

The glyceollins, eg (182), are isoflavone phytoalexins that accumulate in soybean roots under challenge from M. incognita [299,300]. Levels of glyceollins were found to be related to resistance to the cyst nematode H. glycines [301]. Glyceollin I-III inhibited the motility of M incognita (22, 22, 12 |ig/ml) but had no effect on M Javanica, The inhibitory effect is reversible and the nematodes recovered on removal from the test solution. These compounds did not inhibit acetylcholinesterase activity [302]. Equol (183) was weakly toxic (LDioo 250 ppm) towards C. elegans and P, redidivus [303]. The related medicarpin (184) and its 4-hydroxy derivative (185), from the dried roots of the Ethiopian medicinal plant Taverniera abyssinica, showed toxicity (LD50 25 |i,g/ml) to C. elegans and inhibited oxygen consumption of axenically grown C elegans [304]. In comparison, the insecticide rotenone (186) (LD90 1 |ig/ml) and ivermectin (LD90 5 }ig/ml) were significantly more active. Medicarpin (184) inhibited the motility of P. penetrans in vitro [305]. H3O

HX(

H3C0. HXO

187

OH 188 R = H 180 R = OH

190

Pycnanthus angolensis, a plant used in Cameroon as a vermifuge, produced dihydro-guaiaretic acid (187) which showed low activity (LC50

474

>100 ppm) in the brine shrimp lethality assay, but significant activity (LC50 10.1 ppm) towards C elegans [306]. The lignans (188) and (189) (50 ppm) inhibited cyst hatching of G. rostochiensis and G. pallida [307]. (-)-Nor-trachelogenin (190) and (+)-pinoresinol (191), from Pinus massoniana, showed weak activity (LD67 and LD46 100 |Lig/ml) towards the pine wood nematode B. xylophihis [280, 308].

191

The plant family Zingiberaceae includes the well known spices, turmeric, cardamon (Elettaria cardamomum) and ginger {Zingiber officinale), which have been used for centuries in traditional Chinese, Japanese and Indian medicine. The curcuminoids are specific to several genera of this family and are the colouring substances of turmeric. Their chemistry and pharmacology have been reviewed recently [309]. The methanolic extract of the roots of C comosa was found to significantly inhibit the motility of C elegans. Bioassay-guided fractionation of the hexane soluble portion of the extract provided four active diarylheptanoids (192-195) that showed significant activity (EC95 9, 9, 0.7, 1 |Lig/ml respectively) to C. elegans in motility assays [310]. Compound (194) was racemic and (195) was assigned the R-configuration. Curcumin I-III (196198) were isolated from the nematocidally active extract of Curcuma longa. When tested individually in an assay using T. canis, the compounds were ineffective, but a mixture of the compounds showed activity (0.1 |ig/ml) [311]. Roots of the banana plant Musa acuminata infected with the nematode Radopholus similis produced the phenalenone (199) [312]. Interestingly, it seems likely that diarylheptanoids, also produced in response to fungal challenge, are possible precursors of phenalenones [313]. A group of diarylnonanones, the malabaricones A-C (200-202), showed nematocidal activity (MLC 5, 25, 6-10 |LiM) against T. canis [314]. Twenty analogues were synthesised, but none showed greater activity than the natural compounds. Zingiber officinale is eaten along with raw fish and used in traditional Chinese medicine. An extract from this plant effectively destroyed Anisakis larvae in vitro. The active components were identified as [6]-

475 shogaol (203) and [6]-gingerol (204). In a solution containing [6]-shogaol (62.5 |ig/ml), more than 90% of larvae lost spontaneous movement within 4 h and were destroyed completely within 16 h. Microscopic examinations showed destruction of the digestive tract and disturbances of the cuticle [315]. [6]-Gingerol showed reduced activity (250 |ig/ml) towards Anisakis larvae, whereas the drug pyrantel pamoate was inactive at 1 mg/ml. Synergistic effects between [6]-gingerol and a small amount of [6]-shogaol were observed.

Rl

" - ^

^-^^

192 193 194 195

^OH

R2

H, OAc 0 H,OH 0

196 197 198

H H H OH

Rl OCH3 OCH3 H

R2 OCH3 H H

(Cllj),

199

200Ri = R 2 = H 201 Ri = OH; R2 =H 202 Rl = R2 = OH

203

H,0 (CHj)4CH3

476

H,CO, (CH2)4CH3

204

Metabolites derived from amino acids 7?-2-Methylene cyclopropaneacetic acid (MCPA) (205) is a metabolite of hypoglycin A (206), the amino acid responsible for Jamaican vomiting sickness [316]. MCPA has been shown to be a strong inhibitor of flavincontaining Acyl CoA dehydrogenases. The inhibitory activity of MCPA to galling induced by Meloidogyne incognita was determined. MCPA (0.5 ppm) inhibited galling on fox-tail millet {Setaria italica) grown in sand by 90% over a period of 14 days [317]. A number of derivatives of MCPA showed similar activity. Interestingly, higher concentrations of MCPA were required to achieve the same effect if the plants were grown in soil, presumably due to the instability of MCPA. It is not known if MCPA acts as a fatty acid oxidation inhibitor in nematodes.

.CO.H

206

205

(Hof);

COjH

208

(HjQjN

209

The seeds of Quisqualis indica are commonly used in China to treat ascariasis. The active component is quisqualic acid (207). In clinical trials, potassium quisqualate (125 mg) showed anthelmintic activity equi-

477

valent to that of santonin [318]. Quisqualic acid (0.1% w/v) in vitro caused cessation of movement of Ascaris siium, but was not nematocidal [319]. Quisqualate binds to glutamate receptors in mammalian brain tissue and increases the flux of monovalent cations [320]. Since nematodes contain specific glutamate binding sites [321,322], this suggests that the activity of quisqualate may be mediated by a glutamate receptor interaction. Both the roots and edible portions of Asparagus officinalis contain asparagusic acid (208), a member of the relatively rare group of 1, 2dithiolanes which includes the coenzyme a-(+)-lipoic acid. Asparagusic acid inhibited (LD80-99 55 ppm) the growth of M hapla, P. ciirvitus and P. penetrans. It also inhibited (50 |Ug/ml) hatching of G. rostochiensis and H. glycines even in the presence of hatching stimulants [323]. The biosynthesis of asparagusic acid has been investigated (Scheme 5) [324]. The results showed that it is derived from isobutyric acid via methacrylic acid, 2-methyl-3-mercaptopropionic acid and S-(2-carboxy-n-propyl)cysteine, this last also being found in asparagus. The conversion of isobutyric acid to methacrylic acid involves the oxidation of the l-pro-S methyl group of isobutyrate. In contrast, in the biosynthesis of lipoic acid, which is derived from octanoic acid, sulphur is introduced without loss of hydrogen on the carbon adjacent to the sites of sulphur introduction [324].

pro-R H,C

HjC, -CO2H

H,C ^

-CO,II

CH3 pro-S

Asparagusic acid (208)

Scheme 5. Biosynthesis of asparagusic acid

478

Members of the Laminariales and other brown algae produce the amino acid laminine (209) which has general anthelmintic activity [325]. Laminine, y-amino-butyric acid betaine and glycinebetaine reduced the number of M javanica females infecting Arabidopsis thaliana [326] and L2 juveniles infecting tomato plants [327]. The kainoid amino acids are a group of non-proteinogenic pyrrolidine dicarboxylic acids which have come under close scrutiny in recent years because of their anthelmintic, insecticidal and neuroexcitatory properties [328]. The parent member, (-)-a-kainic acid (210), and its C4-epimer, (+)-allokainic acid, were first isolated from the marine red alga Digenea simplex which has been used in Japan against intestinal worms for over a thousand years. It also occurs in a related algae, Centrocerus clavulatum, in the Corsican moss, Alsidium helminthocorton and the plant Caloglossa liprieurii [329]. a-Kainic acid is a broad spectrum anthelmintic which is ten times more potent than the sesquiterpene santonin (137) without the complicating side effects [330], a-Kainic acid kills adult Ascaris worms by causing a neuromuscular block with no apparent adverse side effects on humans at doses of 5-10 mg [331]. It also showed synergistic activity with santonin in deparasitising mice infected with Symphacia obvelata [332,333]. The c/^-stereochemistry at C3 and C4 appears to be crucial to its biological activity.

479

Domoic acid (211) was first isolated from the red algae Chondria armata, also known for its anthelmintic and insecticidal activity. It has since been found in two other Rhodophytes, C. baileyana and Alsidium corallimim, and in two diatoms [328]. Several stereoisomers of domoic acid are known; isodomoic acids G and H are two recently isolated diastereoisomers from Chondria armata [334]. Oral administration of domoic acid (20 mg) to children effectively expelled ascaris and pinworm, without any observable side effects. Domoic acid has been identified as the toxin responsible in paralytic shellfish poisons. It accumulates in mussel through its food source, the phytoplankton Nitzschia pungens, which can produce more than 1% of its dry weight of domoic acid. Both kainic and domoic acids exert their toxic effects through interaction with the glutamate receptors in the nervous system and have found use in neurophysiological studies. Their activity in both invertebrate and vertebrates leads to specific neuronal degeneration in the brain. Little is known of the biosynthetic pathways to these compounds, although some labelling studies using [l-'^C]- and [1, 2-^^C]-acetate have been carried out with domoic acid. Evidence was obtained for a novel condensation of a geranyl unit with an activated glutamic acid derivative arising from the citric acid cycle [335]. Presumably, similar alkylation with isopentenyl diphosphate would provide kainic acid. In contrast, the acromelic acids A and B, produced by the poisonous mushroom Clitocybe acromelalga, [336] arise from a modified tyrosine unit. Both exhibit greater neuroexcitatory activity (glutamate agonists) than domoic acid and are highly toxic (7 and 8 mg/kg) to mice [328]. Glucosinolates are amino acid derived plant metabolites that contain a sulphate and a thioglucose moiety (eg 212). They are produced by members of the Brassicaceae and at least 10 other families of dicotyledous angiosperms. Their biosynthesis from amino acids involves a multifunctional P450 enzyme which converts the amino acids to the aldoxime [337,338]. Glucosinolates have been detected in all organs of plants that produce them and are located within the vacuole of the cell. They are considered to have little biological activity themselves but, on hydrolysis, they produce compounds which contribute to plant defence. The hydrolysis is catalysed by endogenous p-thioglucosidases (myrosinases) which are activated on tissue damage, in the case of nematodes, when the stylet perforates the cell [339], The hydrolysis of glucosinolates initially involves cleavage of the thioglucoside linkage yielding D-glucose and an

480

unstable thiohydroximate-O-sulphonate that spontaneously rearranges, resulting in the production of sulphate and a range of possible reaction products, thiocyanate, isothiocyanate or nitrile depending on such factors as substrate, pH or the availability of ferrous ions (Scheme 6) [340, 341].

o.so

212

0,80

215

214

+ NH3

, ^ '

217

216

y OH HO' HO

OH HO

H

NItrlle

s^

^R

-..-.^^^^^^

Thiocyanate S = C = N—R Isothiocyanate

/

°'»"^'

^ i r "N^°

^ Scheme 6. Products generated from glucosinolates

S Oxazolidine-2-thJone

481

The antinematodal effect of Brassica species has been recognized for a long time and soil amendments incorporating these plants has been recommended from time to time [342]. Tissues of plants containing glucosinolates reduce the number of soil-borne parasitic nematodes, including Meloidogyne spp., K schachth, and the root-lesion nematode P, penetrans. Glucosinolates are not active by themselves, they become active in the presence of myrosinase. Thus, glunapin (212), dehydroerucin (213), glucotropolin (214) and sinigrin (215) (0.5%), with added myrosinase, caused 93-100% mortality of L2 juveniles of H. schachtii [343]. Isothiocyanates display a range of activities towards a wide variety of soil organisms including fungi, viruses, insects and nematodes. They are potent electrophiles which react, inter alia, with free amino groups of amino acids and proteins (cf Edman degradation) [339], Allyl isothiocyanate (216) (50 |ig/ml) inhibited hatching of eggs of G. rostochiensis [344]. It was also toxic (LD50 40 ppm) to C. elegans and was nearly three orders of magnitude more potent than the corresponding glucosinolate (sinigrin) [345, 346]. It is interesting to note that methylisothiocyanate was used as a soil fumigant, and 4-isothiocyanato-N-(4-nitrophenyl)benzeneamine (amoscanate) and 1, 4- diisothiocyanatobenzene (bitoscanate) as anthelmintics against Ancylostoma and Necator infections [75]. Of 150 natural and simple synthetic aryl and arylalkyl isothiocyanate tested against the free-living nematode Turbatrix aceti, a number with 4substitution were shown to be active [347,348]. Carbon disulfide (CS2) is a volatile liquid with bacteriostatic, fungicidal and insecticidal properties. Its nematocidal activity was first noted in 1932 [349]. Many plants of the subfamily Mimosoideae produce CS2 when their roots are injured, and this might serve to control soil pathogens. CS2 is known to be produced when soils are amended with sulfur containing amino acids or isothiocyanate [350], Glucosinolates are apparently absent in mimosoid plants, but the seeds contain S-alkyl cysteine derivatives in the amino acid pool, of which the most notable is I-djenkolic acid (217). Enzymes that cleave these amino acids are present in parts of the plant, including the roots, and the products presumably are ammonia, pyruvate and methylene dithiol. Addition of (217) or Z-cysteine to injured roots of Mimosa pudica led to increased levels of CS2 and carbonyl sulfide. These plants could be useful as rotational crops or intercrops; CS2 acting as a soil fumigant to control nematode levels [350].

482

The peptaibols are a family of peptides that contain a high proportion of the unusual amino acid a-aminoisobutyric acid, a C-terminal alcohol and an acylated N-terminal group. They occur as microheterogeneous mixtures in which the components differ only in the type of amino acid at one or more positions, a reflection of the non-ribosomal biosynthetic origin and have been isolated from a large number of fungi including Trichoderma and Penicillium [351]. They exhibit a wide range of bioactivities, eg bactericidal, fungicidal and insecticidal, because of their ability to interact with phospholipid bilayers, increasing the permeability of liposomes and forming transmembrane channels in lipid bilayer membranes [352]. Antiamoebin I (218) has been reported to possess anthelmintic activity [6] and claims that the chrysospermins (eg 219) show anthelmintic, antiinflammatory, immunosuppressive, nematocidal, and antitumor activity have been made in a patent [353]. Ac-Phe-Aib-Aib-Aib-lva-Gly-Leu-Aib-Aib-Hyp-Gln-lva-Hyp-Aib-Pro-Phe-OH

218 Ac-Phe-Aib-Ser-Aib-Aib-Leu-Gln-Gly-Aib-Aib-Ala-Ala-Aib-Pro-Aib-Aib-Aib-Trp-OH

219 Cyclic peptides The enniatins are a group of cyclodepsipeptides produced by Fusarium spp [206]. They are produced by a non-ribosomal mechanism and their biosynthesis is regulated by the amino acids of the metabolic pool [354]. Enniatin A (220) (5 |ig/ml) showed nematocidal activity towards N, brasiliensis, Trichinella spiralia and Heterakis spumosa, and other enniatins were active at 1-100 |ig/ml [355]. Enniatin B (221) (100 |ig/ml) was toxic towards Anguillula aceti [142] and M javanica (LD43 20 ppm) [233]. A novel depsipeptide PF1022A, isolated from a Mycelia sterilia, represents a promising new class of potent anthelmintics [356]. Compound (222) consists of eight residues, four N-methyl-L-leucines, two D-lactates and two 3-phenyl-D-lactates, elaborating a floppy 24-membered ring. In motility assays, it showed potent activity towards Heterakis spumosa (10^ '7 g/ml) [361] and H.contortiis (0.1 |LLM) [357]. It appears however that it acts as a neurotoxin by paralysing the nematode [358] and its activity is

483

not related to the ability to act as an ionophore, a property it shares with other cyciodepsipeptides [359].

fC

N

N

O

220 R = sBu 221 R = iPr

222

An indication of the broad range of antinematodal activity in vivo can be gained from the results given in Table 3. It is not, however, without its limitations. Whilst it has useful activity in clearing intestinal parasites in ruminants, it is less efficient against nematodes in tissues [360, 361]. Moreover, it is less effective when delivered by routes other than oral [358, 360]. However, it holds some promise as a nematocide and total syntheses of (222) [362, 363] and its enantiomer [363] have been described. The enantiomer had no significant in vivo activity [363] and poor in vitro activity (100 |Lig/ml) towards Trichinella spiralis and A^. brasiliensis [364]. In attempts to understand better its mode of action, SAR studies involving systematic replacement of the leucine residues by N-alkylated amino acids [365], and construction of analogues which restrict the conformations available have been undertaken [362,366]. Omphalotin A (223), a metabolite of the basidiomycete Omphalotus olearius, is a modified cyclododecapeptide containing 5 valine, 3 glycine, 3 isoleucine and one tryptophan, with methylation on 9 a-nitrogens [367]. The structures of omphalotins A and more highly oxygenated congeners, omphalotins B-D, have been elucidated [367,368] and the determination of the L-configuration for the amino acids of omphalotin A has established its absolute configuration [368].

484 Antinematodal Activity of PF1022A (222)

Table 3. 1 Host

Nematode Angiostrongylus

mice

\ costaricensis

Time

% clearance

Ref.

2.5-40 mg/kg (oral)

5 days

dose dependent

356

0.625 mg/ks (ip)

5 davs

complete

[Dose (method)

>91

357

95

358

3 days

none

360

complete

360

1 Ascahdia ^alli

chicken

2 mg/kg

1 H. contortus

jird

-2.75 mg/kg (oral)

4hr

10 mg/kg (oral)

Hymenolepis nana mice 1 A^. brasiliensis

rats

2.5 mg/kg (oral)

3 days

1 O. ostertagi

jird

-8.25 mg/kg (oral)

4hr

2.5 mg/kg (oral)

3 days

complete

5 mg/kg (ip)

3 days

| no effect

Thchinella spiralis mice

T. colubriformis

223

jird

1 -2.75 mg/kg (oral)

4hr

95

1

95

358 361

358 1

224 R = CH3 225 R = H

The omphalotins are lipophilic cyclopeptides that can exist in different conformations in different solvents [368]. Omphalotin A is highly toxic (LD90 0.76 |xM) to M. incognita, but 50 times less potent towards C.

485 elegans. In comparison, ivermectin exhibited LD90 4.6 and 0.46 |LIM towards the two nematodes [369]. In glasshouse tests, omphalotin A (2.510 |j.g/l) completely protected cucumbers and lettuce from M incognita and, since it is not phytotoxic and is weakly cytotoxic, it has potential as a nematocide [370]. Bursaphelocide A and B (224, 225) are two nematocidal metabolites from an imperfect fungus of the Mycelia sterilia [371]. Both compounds (100 |Lig) almost completely inhibited the growth of 5. xylophilus in the cotton ball assay [372,373]. The cyclic depsipeptide jaspamide (226), also known as jasplakinolide, was originally isolated from marine sponges of the genus Jaspis and, later, from a number of other sponge genera [374]. Jaspamide possesses a remarkable range of biological activities, including antifungal [375] and antiproliferative activities [376], It showed insecticidal activity (LC50 4 ppm) against the tobacco bud worm Heliothis viscerens [377], selective antimicrobial properties [375,378], potent nematocidal activity (ED50 <1 ppm) against A^. brasiliensis [375], cytotoxic [374] and ichthytoxic activity [379]. A group of novel ketide amino acid derivatives known as the bengamides cooccurr with jaspamide in Jaspis spp. [380-383]. Bengamide A and B (227, 228) showed activity towards A^. brasliensis [383,384]. Compared to levamisole as a standard (0.5 |ig/ml), which showed reduction of casts (100%), viability (99%) and motility (95%), bengamide A and B achieved the same results at 50 and 3 \iglm\ [383].

226

486

227 R = H 228 R = CH3

(CH2),2CH3

229

The soil-borne fungus Tolypocladium niveus has long been known to produce antifungal metabolites with a narrow spectrum of activity. These metabolites, now collectively known as the cyclosporins, are a group of more than 25 cyclic undecapeptides that became better known because of their potent immunosuppressive effects and low toxicity [385,386]. They have also shown promising results in the treatment of several disorders such as rheumatoid arthritis and myasthenia gravis [387]. Cyclosporin A (229) exhibited nematocidal activity against M incognita, although it was 100 times less potent than omphalotin [6]. It showed in vivo activity towards the intestinal nematode Strongyloides ratti. Subcutaneous injection of cyclosporin A (30 mg/kg) resulted in elimination of the parasite [388]. In dogs, cyclosporin A was active towards S, stercolaris, a nema-

487 tode capable of massive, often fatal, hyper-infection. It also conferred protection against the rodent filarial Dipetalolema viteae [389]. Alkaloids The extract from di Xestospongia sp. provided xestoaminol A (230) w^hich, at 50 |ig/ml, showed 100% reduction of viability and motility towards the L4 stage ofN, brasiliensis. By comparison, levamisole produced the same effects at 5 |ig/ml [390]. The compound probably arises from condensation of a fatty acid with alanine. The pyrrolidine alkaloid (231) occurs in the tropical legumes of the Lonchocarpus and Derris genera. It inhibited (IC41 25 |Lig/ml) cyst hatch of G. pallida and immobilised juveniles of G. rostochiensis. It reduced root galling by Meloidogyne spp. when used as a root drench, as a seed dressing or applied as a foliar spray, showing that it is a phloem-mobile compound [391]. An extract of the dried unripe fruits of the Chinese medicinal plant Evodia rutaecarpa was fractionated using O. circumcinta L4 larvae in a bioassay. The ant-helmintic principle was identified as the anthranilic acid-derived quinolone atanine (232) which immobilised larvae and adults of C elegans (IC50 10 M-g/rnl) and L4 larvae of (9. circumcinta (IC50 50 fig/ml) [392]. Monocrotaline (10 |Lig/ml), an ornithine-derived pyrroUizidine alkaloid, inibited movement of M. mcogmVa juveniles [393]. The anthelmintic activity of the quinolizidine alkaloids has been reviewed [394]. It is well recognised that quinolizidine alkaloids deter or repel insects and non-insect herbivores [394]. These compounds can interfere with protein biosynthesis and some bind to acetylcholine receptors with high affinity. (-)-N-methylcytisine (233) and (-)-anagyrine (234) were isolated from the roots of Sophora Jlavescens, a plant used in traditional Chinese medicine as an anthelmintic. N-methyl cytisine was twice as active as anagyrine but only half as active as (-)cytisine and nicotine [395,396] The alkaloids (3-6 )Lig) inhibited reproduction of B. xylophilus in the cotton balls assay.

(CH2)5-

230

488

o 233

234

The bengazoles and homologues have been isolated from Jaspis spp [397] and are representative members of a small group of bisoxazole metabolites [398]. Bengazole A (235) shows moderate activity (LD50 50 )Lig/ml) towards N. braziliensis. On storage, the compound decomposed to generate an alditol, probably by a pathway that involves photochemical addition of oxygen to the oxazole B ring [397]. The compound formed was devoid of antinematodal activity. Following the assignment of the absolute configuration by chiroptical studies [399], a result that appears to have been overlooked [381], stereoselective syntheses of bengazole A and 10-epibengazole A have been published recently [398]. Chelerythrine (236), a benzophenanthridine alkaloid, and the related sanguinarine and bocconine, isolated from the poppy Bocconia cordata, were toxic (50-100 |ig/ml) to the free-living nematodes Rhabditis and Panagrolaimus [400]. A crude extract obtained from a culture of a soil actinomycete showed activity toward C. elegans. The major active component isolated from this mixture was dioxapyrrolomycin (237) [401], a compound reminiscent of the antifungal metabolites, pyoluteorin and pyrrolnitrin, produced by some Pseudomonads from tryptophan. Recent progress in the biosynthesis of these compounds has been reviewed [402]. Dioxapyrrolomycin (237) was found to be active against H. contortus in the jird model (>91% clearance at 0.33 Jig/jird) and in lambs at 12.5 mg/kg (99.9% clearance). The indolyl oxazoles (238, 239), from the lipophilic extract of a Streptomyces sp., have been claimed to be active against C elegans [403]. Red algae of the genus Chondria produce the chondriamides A, B (240, 241) [404], and the lO'-l 1' Z-isomer of (240) [405]. Chondriamide A and its isomer showed in vitro activity (ECgo 0.26, 0.09 mM) towards N, brasiliensis, Chondriamide B was too unstable for biological testing, but its trimethyl derivative showed ECgo 0.30 mM. The magnesium salt (242) of a macrocyclic tetramic acid lactam, isolated from a sponge of the genus Geodia, showed significant toxicity (LD99 1 |ig/ml) towards H. contortus [406].

489

y=^N

OH

OH

HjCO'

238 R = CI 239 R = H

240 R = H 241 R = OH

490

Mg^-

242 Paraherquamide A (243) is an indole alkaloid first isolated from Penicillium paraherquei. Its structure was determined by X-ray diffraction analysis [407]. The potent nematocidal activity associated with this compound was discovered with its reisolation from P, charlesii together with five other related paraherquamides (C-G) [408]. It exhibited anthelmintic activity against immature T. colubriformis in a rodent model system and was highly effective in a single oral dose in sheep infected with T, colubriformis, H. contortus, 0. circumcinta and C curticei, Paraherquamide showed LD50 0.033, 0.058 and 2.7 |ag/ml in inhibiting the motility of L3 larvae of O. circumcincta, T. colubriformis and H. contortus [409]. In vivo, paraherquamide was still effective against ivermectin- and thiabenzadole-resistant nematodes, which suggests a different mode of action [410]. A series of naturally occurring analogues demonstrated strong activity when tested in an immotility assay using C. elegans [411]. More than a hundred structural analogues of paraherquimade have been tested, but none have proven to be as active either in the motility assay or binding to a membrane preparation of C elegans, [412]. Unfortunately, the relatively low yield of the paraherquamides recoverable from fungal cultures have slowed the development of these promising agents. Marcfortine A (244), also a potent anthelmintic, is a fungal metabolite oi Penicillium roqueforti [413], a variety of which is used for cheese starter cultures. All the three recently recognised varieties of this fungus, now named P. roqueforti, P. carneum and P. paneum, produce the marc-

491 fortines [414]. Marcfortine A is structurally related to paraherquamide containing a five-membered, instead of a 6-membered G-ring, and a gemhydroxyl and a methyl at CI4.

244

245 O

CHj

Chemical conversions of marcfortine to paraherquamides have been achieved [415]. Utilising various soil-derived microorganisms, individual hydroxylation at carbon atoms 5, 10, 12, 14, 15, 16 and 27 has been realized [416] but no improvement on the activity of (244) was observed. A study of the biosynthesis of (244) has shown that it is derived from methionine, tryptophan, lysine and two isoprene units, the latter two being derived from acetic acid. The pipecolic acid moiety arises from lysine via a-ketoglutarate [417]. Bioassay-guided fractionation (//. contortus L3 larvae) of the culture extract of a strain oi Aspergillus led to the isolation three paraherquamide derivatives and two members of a novel class of compounds which have been named aspergillimides [418]. They differ from the paraherquamides in the absence of the 7-membered dioxygenated ring and the phenyl ring to which this is fused. The anthelmintic activities of the aspergillimides were tested on gerbils infected with T. colubriformis. Aspergillimide A (245) achieved 98% reduction in faecal egg count at 20 mg/kg. Paraherquamide A, in comparison, showed 99% reduction at 4 mg/kg.

492 The plakidines are novel pentacyclic aromatic alkaloids originally isolated from a Vanuatan red sponge of the genus Plakortis [419,420]. The isolation of plakinidines from an ascidian [421,422] suggests that a symbiotic microorganism might be the source of these compounds, but the origin of the pyrroloacridine skeleton remains to be determined. The plakinidines are purple and red compounds whose colour varies according to pH (purple in chloroform and dark green in acidic solution) [419]. Plakinidine A (246) and B (247) exhibited in vitro activity towards N, brasiliensis at 50 |Lig/ml [423].

248 249

246 R = H 247 R = Chfe

NH,

NH2OC

X. CH20H

N(CH3)2

250

NHCH,

251

493 Other compounds Garlic (Allium sativum) is reputed to have a number of desirable biological activities and some of its many metabolites have been tested to this end [424, 425]. Garlic and its extracts have been used as pesticides and antinematodals, although the plant can be a host to some nematodes, eg Ditylenchus dipsaci. Limited information on the nematocidal activity of individual components is available. Diallyl disulfide has been reported to be nematocidal [426]. AUicin (248) inhibited (LD90 5 ppm, LD50 0.5 ppm) egg hatching of M incognita and was lethal to juveniles (LDgy-ioo 2.5-5 ppm) [427]. 2-Aminoquinoline, from the mushroom Leucopaxillus albissimus, was moderately active against N, brasiliensis (L4) and, at 50 |ig/ml, reduced motility (50%), viability (74%) and cast formation (52%) [428]. In the fresh mushroom, the concentration of this compound (2 |Lig/g) is 40 times higher. Phenoxazone (249) was toxic (LD50 50 |ig/ml) to M incognita [6]. Some members of the orthosomycins, a group of compounds known for their antibiotic activity, showed activity against intestinal nematodes [429]. The nucleoside peptide aspiculamycin (250), isolated from Streptomyces toyocaensis, at 50 mg/kg, completely eliminated Aspiculahs tetraptera and Syphacia obvelata (pinworm) from the intestinal tract of infected mice [430]. The related compound (251) has been claimed to show nematocidal activity towards B, lignicolus [431]. Anthelmycin (252) (hikizimycin) has been found to be effective against pinworms, roundworms and whipworms [432].

vun^

M^

N^

^ ^ ^

252

^o

494

CONCLUDING REMARKS The number of secondary metabolites found to have antinematodal activity is remarkable and ranges from the structurally simple, non-specific ascaridole to the more complex and more specific avermectins, paraherquamides and depsipeptides. A distinction can also be made between those metabolites that have evolved in plants as defence compounds, the polyynes, 1, 2-dithiins, glucosinolates and phytoalexins and many others w^hose activity, at this stage, appears incidental. The in vitro activity determined for a number of compounds can be misleading. Satisfactory delivery of the test compound is often not achieved because of solubility problems. A number of cases in which total extracts display greater activity than can be attributed to the individual components have been noted. While this is often rationalised by assuming the presence of a highly active, unisolated principle, there is some evidence of synergistic effects. Thus, for example, simple fatty acids can show activity alone or in a synergistic manner with other metabolites. The modest activity observed for some plant metabolites might belie the fact that phytoparasitic nematodes, in situ, could be exposed to high local concentration of the metabolite(s). Studies aimed at gaining an insight into the the spectrum of action of potential antinematodal compounds have indicated that results from several tests need to be evaluated since different parasitic species can react differently to the same treatment. It has been suggested that separation of nematodes into two groups, free-living and parasitic, is not helpful and that more importance be placed on how much species have in common [62]. For example, phylogenetic studies indicate that C elegans is likely to be an excellent model for the human hookworms Ancylostoma and Necator [2]. The C elegans genome has revealed many nematodespecific systems that are potential targets for nematode control [433]. With predatory fungi, studies on the interaction of attractants produced by the fungus with the chemosensory receptors of nematodes [434] could provide significant leads. Apart from studies on avermectins, paraherquamide and some cyclodepsipeptides, few attempts to delineate the mode of antinematodal action of the metabolites contained in this review have been undertaken. Currently, there are three classes of broad spectrum anthelmintics in use: the benzimidazoles, the levamisoles and the avermectins. However,

495 evidence for resistance to these is becoming apparent and it is encouraging to note that, for example, the paraherquamides [408], PF1022A [358] and the bafilodides [169] seem to act by mechanisms which differ from those presently known. The analysis of the genes in drug-resistant mutants of C. elegans should provide a better insight into the underlying mechanisms.

496 REFERENCES [1]

[2] [3] [4] [5]

[6] f7]

[8]

[9] [10] [II] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23]

[24]

[25] [26]

[27] [28] [29] [30] [311 [32] [33] [34] [35] [36] [37] [38]

Atkinson, H.J.; Lilley, C.J.; Urwin, P.E.; McPherson, M.J. In The Physiology and Biochemistry of Free-living and Plant-parasitic Nematodes; Perry, R.N.; Wright, D.J., Eds.; CABI International: Wallingford; 1998; pp. 3 8 M l 3 . Blaxter, M. Science, 1998, 282,2041. The World Health Report, World Health Organisation, Geneva, 1998. Komis, G.I, In Agrochemicals from Natural Products; Godfrey, C.R.A., Ed.; Marcel Dekker, Inc: New York, 199S; pp. 215-255. Chitwood, D.J. In Pest Control with Enhanced Environmental Safety: Duke, S.O.; Menn, J.J.; Piimmer, J.R„ Eds.; American Chemical Society: Washington DC, 1993; pp. 300-315. Anke, H.; Sterner, O. Curr. Org Chem., 1997, I, 361. Boris, R.P.; Scaeffer, J.M. In Phytochemical Resources for Medicine and Agriculture', Nigg, H.N.; Seigler, D., Eds.; Plenum Press: New York, 1992; pp. 117-158. Crews, P.; Hunter, L.M. In Marine Biotechnology: Pharmaceutical and Bioactive Natural Products; Attaway, D.H.; Zaborsky, O.R., Eds.; Plenum Publishing Corporation: New York, 1993; pp. 343-349. Ikeda, H.; Omura, S. Chem. Rev., 1997, 97, 2591. Ikeda, H.; Omura, S. J. Antibiot., 1995, 54, 1995. Lanusse, C.E.; Prichard, R.K. Vet. Parasitoi, 1993,49, 123. Sangster, N.C.; Gill, J. Parasitology Today, 1999, 15, 141. Thompson, D.P.; Klein, R.D.; Geary, T.G. Parasitology, 1996, 113 (Suppl S), S217 DeSilva, N.; Guyatt, H.; Bundy, D. Drugs, 1997, 53, 769. Ottesen, E.A.; Campbell, W.C. J. Antimicrob. Chemother., 1994, 34, 195. McKellar, O.A.; Benchaoui, H.A. J. Vet. Pharmacol. Therap., 1996, 19, 331. Martin, R. /. Pharmacol. Therap., 1993, 58, 13. Cleland, T.A. Molec. Neurobioi, 1996, \3, 97. Wratten, S.D.; Forbes, A.B. Ann. Appl. Biol., 1996, 128,329. Wilson, R.K. Trends in Genetics, 1999, 15,51. Barron, G.L. The Nematode-Destroying Fungi, Canadian Biological Publications Ltd: Guelph, Ontario, 1977. Wallwork, J.A. Ecology of Soil Animals, McGraw-Hill Publishing Co. Ltd.: London, 1980. Hussey, R.S.; Grundler, P.M. W. In The Physiology and Biochemistry of Freeliving and Plant-parasitic Nematodes', Perry, R.N.; Wright, D.J., Eds.; CABI International: Wallingford, 1998; pp. 213-243. Opperman, C.H.; Conkling, M.A. In Plant and Nematode Interactions-, Barker, K.R.; Pederson, G.A.; Windham, G.L., Eds.; American Society of Agronomy: Madison, Wisconsin, 1998; pp. 239-250. Riffle, J.W. Phytopathology, 1967, 57, 541. Gibson, D.I. In Microbiology and Microbial Infections: Parasitology: Cox, F.E.G.; Kreier, J.P.; Wakelin, D., Eds.; vol, 5; Oxford University Press: New York, 1998; pp. 453-477. Jansson, H.-B., Tunlid, A.; Nortbring-Hertz, B. In Fungal Biotechnology: Anke, T., Ed; Chapman and Hall: Weinheim, 1997; pp. 38-50. Rice, E.L. Pest Control with Nature's Chemicals; University of Oklahoma Press: Norman, 1983. Tyler, J. Plant Disease Reporter, 1938, 109 (Suppl.), 133. Steiner, G. Proc. Bioi Soc. Washington, 1941, 54, 31. Gommers, F.J. Helmintoi Abstr., 1981, 50B, 9. Gommers, F.J. Nematicidal Principles in Compositae, Dissertation, Agricultural University: Wageningen, The Netherlands, 1973. Prakash, A; Rao, J. Botanical Pesticides in Agriculture, CRC Lewis Publishers: Boca Raton, 1997. Scheuer, PJ. In Biomedical Importance of Marine Organisms; Fautin, D.G., Ed.; Memoirs of the Califomian Academy of Science: San Francisco, 1988, p 37. Kliks, M.M. Social Sc. Med., 1985,21, 879. Burger, A. http://pubs.acs.org/books/burgerO I .htm (16/10/1997) del Castillo, J. In Chemical Zoology; Florkin, M.; Scheer, B.T., Eds.; vol. 3, Academic Press: New York, 1969; pp. 521-554. Duncan, L.; Noling, J. W. In Plant and Nematode Interactions; Barker, K.R.; Pederson, G.A.; Windham, G.L., Eds.; American Society of Agronomy: Madison, Wisconsin, 1998; pp. 251-287.

497

[39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75]

Waller, P.J.; Larsen, M. Int. J. Parasit., 1993, 23, 549. Albonico, M.; Crompton, D.W.T.; Savioli, L. Adv. Parasitol. , 1998, 42, 278. Molyneux, D.H. In Microbiology and Microbial Infections, vol. 5.; Cox, F.E.G.; Kreier, J.P.; Wakelin, D. Eds.; Arnold, Hodder Headline Group, London, 1998, pp 85-112. Kerry, B.R.; Jaffee, B.A. In The Mycota IV: Environmental and Microbial Relationships; Wicklow, D.T.; Soderstrom, B.E., Eds; Springer-Verlag: Berlin, 1997; pp. 203-218. Dix, N.J.; Webster, J. Fungal Ecology, Chapman and Hall: London, 1995; pp. 284-301. Waller, P.A.; Faedo, M. Int. J. Parasit., 1996, 26, 915. Devidas, P.; Crovetti, A.J. Eur. Patent EP 363, 888, 1990. Chem. Abstr., 1990, 113,2069. http://www.ars.usda.gov/is/np/mba/jan97/ditera.htm Holland, R.J.; Williams, K.L. Microbiology Australia, 1998, 19, 16. Holland, RJ.; Williams, K.L.; Khan, A. Nematology, 1999, 1,131. Otoguro, K.; Tanaka, H. In The Search for Bioactive Compounds from Microorganisms; Omura, S. Ed.; Springer-Verlag: New York, 1992; pp. 63-78. Mori, R. J. Antibiot. Ser. A., 1961, 14, 286. Simpkin, K.G.; Coles, G.C. J. Chem. Tech. Biotechnol., 1981, 31, 66. Hino, T.; Kanda, M.; Kumazawa, S.; Yoshikawa, N.; Mikawa, T. European Patent EP 488, 224, 1992. Chem. Abstr., 1992, 1117, 232208. Anke, H.; Stadler, M.; Mayer, A.; Sterner, O. Can. J. Bot., 1995, 73 (Suppl.), S392. Hood, J.D.; Banks, R.M.; Brewer, M.D.; Fish, J.P.; Manger, B.R.; Poulton, M.E. J. Antibiot., 1989,42, 1593. Johannsen, M.V. Vet. Res. Commun., 1989. 13,455. Jurgens, T.M.; Frazier, E.G.; SchaeffTer, J.M.; Jones, T.E.; Zink, D.L.; Borris, R.P.; Nanakom, W.; Beck, H.T.; Balick, M.J. J. Nat. Prod., 1994, 57,230. Douch, P.G.C.; Morum, P.E. Int. J. Parasit., 1994, 24, 321. Taylor, C.E.; Murant, A.F. Nematologica, 1966, 12,488. Hewlett, T.E.; Hewlett, E.M.; Dickson, D.W. J. NematoL, 1997, 29, 737. Lorimer, S.D.; Perry, N.B.; Foster, L M.; Burgess, E.J.; Douch, P.G.C.; Hamilton, M.C.; Donaghy, M.J.; McGregor, R.A. J. Agric. Food Chem., 1996, 44, 2842. Mayer, A.; Anke, H.; Sterner, O. Mededelingen Faculteit Landbouwkundige en Toegepaste Biologische Wetenschappen Universiteit Gent 61(3A), 1996, 839. Bemt, U.; Junkersdorf, B.; Londerhausen, M.; Harder, A.; Schierenberg, E. Fundament. Appl. Nematol., 1998, 21, 251. Yanagida, J.; Matsuhashi, R.; Watanabe, I.; Satou, T.; Koike, K.; Nikaido, T.; Akao, S. Chem. Pharm. Bull., 1998, 46, 1261. Chitwood, D.J. In Pest Control with Enhanced Environmental Safety; Duke, S.O.; Menn, J.J.; Plimmer, J.R.; Eds.; American Chemical Society: Washington, D.C., 1993; pp. 300-315. Sayre, R.M.; Patrick, Z.A.; Thorpe, H.A. Nematologica, 1965, 11,263. Dijan, C; Pijarowski, L.; Ponchet, M., Arpin, N.; Favre Bonvin, J. Nematologica, 1991, 3 7, 101. Djian, C; Pijarowski, L. Pest. Biochem. Physiol., 1996, 56, 12. Stadler, M.; Anke, H.; Sterner, O. Arch. Microbiol., 1993, 160, 401. Munakata, K. In Natural Products for Innovative Pest management. Current Themes in Tropical Science; Whitehead, D.L.; Bowers, W., Eds.; Pergamon Press: Oxford, 1983; Vol. 2, pp. 299-309, Stadler, M.; Mayer, A.; Anke, H.; Sterner, O. Planta Med, 1994,60, 128. Kwok, O.C.H.; Plattner, R.; Weisleder, D.; Wicklow, D.T. J. Chem. Ecoi, 1992, 18, 127. Davis, E.L.; Meyers, D.M.; Dullum, C.J.; Feitelson, J.S. J. Nematol, 1997, 29, 677. Nagase, A.; Kuwahara, Y.; Tominaga, Y.; Sugawara, R. Agr. Biol Chem., 1982, 46, 167. Feldmesser, J.; Thompson, M.J.; Robbins, W.E.; Sponaugle, R.P. Experientia, 1976,32,466. Islip, P.J. In Burger's Medicinal Chemistry and Drug Discovery: Therapeutic Agents; Wolff, M.E., Ed.; John Wiley & Sons: New York, 1997; Vol. 4, 5th. edition, pp. 365-413.

498

[76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [U] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99]

[100] [101] [102] [103] [104] [105]

[106] [107] [108] [109] [110] [Ill]

Kiuchi, F.; Nishizawa, S.; Kawanishi, H.; Kinoshita, S.; Ohsima, H.; Uchitani, A.; Sekino, N.; Ishida, M.; Kondo, K.; Tsuda, Y. Chem. Pharm. Bull., 1992, 40, 3234. Melle-Bezons, French Patent 1, 599, 014, 1970. Chem. Abstr., 1971, 74, 86811. Nogueria, M.A.; de Oliveira, J.S.; Ferraz, S. Phytochemistry, 1996, 42, 997. Nogueria, M.A.; de Oliveira, J.S.; Ferraz, S.; dos Santos, M.A. Nematol Medit., 1996,24,249. Saleh, M.A.; Rahman, F.H.A.; Ibrahim, N.A.; Taha, N.M. J. Chem. Ecol., 1987, 13, 1361. Kogiso, S.; Wada, K.; Munakata, K. Agr Biol. Chem., 1976,40, 2085. Kogiso, S.; Wada, K.; Munakata, K. Tetrahedron Letters, 1976, 109. Kawazu, K.; Nishii, K.; Nakajima, S. Agr. Biol. Chem., 1980,44, 903. Kimura, Y.; Mori, M.; Suzuki, A.; Kobayashi, A. Agr. Biol. Chem., 1981,45, 2915. Mori, M.; Heon, S.-B.; Kimura, Y.; Suzuki, A. Agr. Biol. Chem., 1982, 46, 309. Ellis, S.M.; Balza, F.; Constabel, P.; Hudson, J.B.; Towers, G.H.N. In LightActivated Pest Control; Heitz, J.R.; Downum, K.R., Eds.; American Chemical Society: Washington, D.C., 1995; pp. 164-178. Gomez-Barrios, M.L.; Parodi, J.F.; Vargas, D.; Quijano, L.; Hjortso, M.A.; Flores, H.E.; Fischer, N.H. Phytochemistry, 1992, 31,2703. Block, E.; Page, J.E.; Toscano, J.; Wang, C.-X.; Zhang, X.; DeOrazio, R.; Guo, C; Platz, M.S.; Sheridan, R.S.; Towers, G.H.N. J. Am. Chem. Soc, 1996, 118, 4719. Towers, G.H.N.; Page, J.E.; Hudson, J.B. Current Org Chem., 1997, 1, 395. Towers, G.H.N.; Abramowski, Z.; Finlayson, A.J.; Zucconi, A. Planta Med., 1989,51,225. de Viala, S.S.; Brodie, B.B.; Rodriguez, E.; Gibson, D.M. J. Nematol., 1998, 30, 192. Uhienbroek, J.H.; Bijioo, J.D. Rec. Trav. Chim., 1958, 77, 1004. Uhienbroek, J.H.; Bijioo, J.D. Rec. Trav. Chim., 1959, 78, 382. Bakker, J.; Gommers, F.J.; Nieuwenhuis, I.; Wynberg, H. J. Biol. Chem., 1979, 254,1841 Gommers, F.J., Bakker, J., Smits, L. Nematologica, 1980,26, 369. Kagan, J. Prog. Chem. Nat. Prod, 1991, 56, 87. Gommers, F.J.; Bakker, J, In Chemistry and Biology of Naturally-Occurring Acetylenes and Related Compounds Bioactive Molecules', Lam, J.; Breteler, H.; Arnason, T; Hansen, L., Eds.; Elsevier: New York, Vol. 7, pp. 61. Addor, R. W. In Agrochemicals from Natural Products; Godfrey, C.R.A., Ed.; Marcel Dekker, Inc.: New York, 1995; pp. 1-62. Arnason, J.T.; Durst, T.; Kobaisy, M.; Maries, R.J.; Szenasy, E.; Guillet, G.; Kacew, S.; Hasspieler, B.; Downe, A.E.R. in Light-Activated Pest Control; Heitz, J.R., Downum, K,R„ Eds.; ACS Symposium Series 616; American Chemical Society: Washington, D.C., 1995: pp. 145-151. Hayashi, M.; Wada, K.; Munakata, K. Agr. Biol. Chem., 1981, 45, 1527. Mayer, A.; Koepcke, B.; Anke, H.; Sterner, O. Phytochemistry, 1996, 43, 375. Capon, R.J.; Barrow, R.A.; Rochfort, S.; Jobling, M.; Skeene, C; Lacey, E.; Gill, J.H.; Friedel, T.; Wadsworth, D. Tetrahedron, 1998, 54, 2227. Aiali, F.Q.; Liu, X.-X.; McLaughlin, J.L. J. Nat. Prod, 1999, 62, 504. Mikolajczak, K.L.; McLaughlin, J.L.; Rupprecht, J.K. U. S. patent No. 4, 721, 717,1988. Chem. Abstr., 1987, 106, 63044. McLauglin, J.L.; Zeng, L.; Oberlies, N.H.; Alfonso, D.; Johnson, H.A.; Cummings, B.A. In Phytochemicals for Pest Control; Hedin, P.A.; Hollingworth, R.M.; Miyamoto, J.; Thompsom, D.G.; Eds., American Chemical Society: Washington, D.C., 1997; pp. 117-133. Rupprecht, J.K.; Hui, Y.-H.; McLaughlin, J.L. J. Nat. Prod, 1990, 53, 237. Johnson, A.H.; Gordon, J.; McLaughlin, J.L. In Progress in New Crops; Janick, J., Ed.; ASHS Press: Arlington, VA, 1996; pp. 609-614. Cave, A.; Figaddre, B.; Laurens.; Cortes, D. Progr Chem. Nat. Prod, 1997, 70, 81. Bories, C; Loiseau, P.; Cortes, D.; Myint, S.H.; Hocquemiller, R.; Gayral, P.; Cave, A.; Laurens, A. Planta Med, 1991, 57,434. Omura, S.; Otoguro, T.; Imamura, N.; Kuga, H.; Takahashi, Y.; Masuma, R.; Tanaka, Y.; Tanaka, H. J. Antibiot., 1987,40,623. Imamura, N.; Kuga,H.; Otoguro, K.; Tanaka, H.; Omura, S. J. Antibiot., 1989, 42, 156.

499

[112] Ye, Y.; Aulinger, K.; Arnold, N.; Spahl, W.; Steiglich, W. Tetrahedron Letters, 1997,38,8013. [113] Tsuzuki, K.; Yan, F.-S.; Otoguro, K.; Omura, S. J. Antibiot., 1991, 44, 774. [114] Giencke, W.; Ort, O.; Stark, H. Liebigs Ann. Chem., 1989, 671. [115] Nakata, M.; Kawazoe, S.; Tamai, T.; Tatsuta, K.; Ishiwata, H.; Takahashi, Y.; Okuno, Y.; Deushi, T. Tetahedron Letters, 1993, 34, 6095. [116] Parry, R.J.; Li, W. 1 Chem. Soc. Chem. Commun., 1994, 995. [117] Lee, T.M.; Siegel, M.M.; Morton, CO.; Goodman, J.J.; Testa, R.T.; Borders, D.B. J. Antibiot., 1995, 48, 284. [118] Tyman, J.H.P. Chem. Soc. Rev., 1979, 8, 499. [119] Suresh, M.; Raj, R.K. Current ScL, 1990, 59, 477. [120] Suresh, M.; Raj, R.K. Ind. J. Pharm. ScL, 1990, 52, 271. [121] Sa Ack Hong Seoul! Med, 1961, 2, 113. Chem. Abstr., 1962, 57, 6563h. [122] Neal, J.W.; Oliver, J.E.; Fetterer, R.H. Ann. Entomol. Soc. Am., 1995, 88,496. [123] Burkhardt, K.; Fiedler, H.-P.; Grabley, S.; Thiericke, R.; Zeek, A. J. Antibiot., 1996,49,432. [124] Ghisalberti, E.L. Phytochemistry, 1996,41, 7. [125] Bennet-Jenkins, E.; Bryant, C. Int. J. ParasitoL, 1996, 26, 937. [126] van Agtmael, M.A.; Eggelte, T.A.; van Boxtel, C.J. Trends Plant Sci., 1999, 20, 199. [127] Callahan, H.L.; Crouch, R.K.; James, E.R. Parasitology Today, 1988, 4, 218. [128] Stadler, M.; Anke, H.; Arendholz, W.-R.; Hansske, F.; Anders, U.; Sterner, O.; Bergquist, K.-E. J. Antibiot., 1993, 46, 961. [129] Stadler, M.; Anke, H.; Bergquist, K.-E.; Sterner, O. J. Antibiot., 1993,46, 968. [130] Stadler, M.; Anke, H.; Sterner, 0. J. Antibiot., 1995,48, 149. [131] Stadler, M.; Anke, H.; Sterner, O. J. Antibiot., 1995, 48, 154. [132] Stadler, M.; Anke, H.; Sterner, 0. J. Antibiot., 1995,48, 158. [133] Stadler, M.; Anke, H.; Sterner, 0. J. Antibiot., 1995, 48, 261. [134] Stadler, M.; Anke, H.; Sterner, 0. J. Antibiot., 1995, 48, 267. [135] Trofast, J.; Wickberg, B. Tetrahedron, 1911, 33, 875. [136] Stadler, M.; Anke, H.; Arendohlz, W.-R.; Hansske, F.; Anders, U.; Sterner, 0.; Bergquist, K.-E. J. Antibiot., 1993, 46, 961. [137] Stadler, M.; Sterner, 0.; Anke, H. Z Naturforsch., 1993,48C, 843. [138] Anderson, M.G.; Jarman, T.B.; Rickards, R.W. J. Antibiot., 1995,48, 391. [139] Wagner, C; Anke, H.; Sterner, O. J. Nat. Prod, 1998, 61, 501. [140] Tezuka, Y.; Huang, Q.; Kikuchi, T.; Nishi, A.; Tubaki, K. Chem. Pharm. Bull, 1994,42,2612. [141] Kimura, Y.; Nakahara, S.; Fujioka, S. Biosc. Biotech. Biochem., 1996, 60, 1375. [142] Bacikova, D.; Betina, V.; Nemec, P. Nature, 1965, 206, I37I. [143] Navarette, A.; Hong, E. Planta Med, 1996, 62, 250 [144] Reichenbach, H.; Hoefle, G.; Garth, K. Ger. Ojfen. DE 19,718,843; 1998. Chem. Abstr., 1999, 130,3107. [145] Nair, M.G.; Chandra, A.; Thorogod, D.L. J. Antibiot., 1993, 46. 1762. [146] Nair, M.G.; Chandra, A.; Thorogod, D.L.; Davis, R.M.G. Pest. ScL, 1995,43, 361. [147] Ishibashi, Y.; Nishiyama, S.; Yamamura, Y. Chem. Lett., 1994, 1747. [148] Stadler, M.; Anke, H.; Dekermendjian, K.; Reiss, R.; Sterner, O.; Witt, R. Nat. Prod Lett., 1995,1,1. [ 149] Wadsworth, D.J. In Agrochemicals from Natural Products', Godfrey, C.R.A., Ed.;Marcel Dekker: New York, 1995; pp.257-284. [150] Oronski, A.L. US Patent A,A96,5^9; 1985. Chem. Abstr., 1985, 102, 125582. [151] Wood, l.B. US Patent ^,5\Q,U^', 1985. Chem. Abstr., 1985, 103, 36174. [152] Amone, A.; Assante, G.; Montorsi, M.; Nasini, G.; Ragg, E. Gazz. Chim. ItaL, 1993, 123,71. [ 153] Grabley, S.; Hammann, P.; Thiericke, R.; Wink, J.; Philipps, S.; Zeeck, A. J. Antibiot., 1993, 46, 343. [154] Zeeck, A.; Philipps, S.; Kind, R.; Grabley, S.; Hammann, P.; Thiericke, R.; Wink, J. Ger. Offen. DE 4,225,284, 1994. Chem Abstr., 1994, 120, 321513. [155] Kind, R.; Zeeck, A.; Grabley, S.; Thiericke, R.; Zerlin, M. J. Nat. Prod, 1996, 59, 539. [156] Machinaga, N.; Kibayashi, C. Tetrahedron Letters, 1993, 34, 841. [ 157] Betina, V. Folia Microbiol., 1992, 37,3. [158] Soga, S.; Neckers, L.M.; Schulte, T.W.; Shiotsu, Y.; Akasaka. K.; Narumi, H.; Agatsuma, T.; Ikuina, Y.; Murakata, C; Tamaoki, T.; Akinaga, S. Cancer Res., 1999, 59, 1239.

500

[158] Tanaka, Y.; Shiomi, K.; Kamei, K.; Sugohhagino, M.; Enomoto, Y.; Fang, F.; Yamaguchi, Y.; Masuma, R.; Zhang, C. G.; Zhang, X. W.; Omura, S. J. AntibioL, 199S,5\, 153. [159] Calton, G.C. US patent 955, 705, 1978. Chem. Abstr., 1978, 94, P121499. [160] Hammann, P.; Kretzschmar, G.; Seibert, G. J. Antibiot., 1990, 43, 1431. [161] Ziegler, F.E.; Tung, J.S. J. Org. Chem., 1991, 56, 6530. [162] Ritzau, M.; Heinze, S.; Fleck, W.F.; Dahse, H.M.; Grafe, V.J. Nat. Prod, 1998, 61, 1337. [163] O'Shea, M.G.; Rickards, R.W.; Tothschild, J.M.; Lacey, E. J. Antibiot., 1997, 50, 1073. [164] Scheldt, K.A.; Tasaka, A.; Bannister, T.D.; Wendt, M.D.; Roush, W.R. Angew. Chem. Int. £J., 1999, 38, 1652. [165] Hensens, O.D.; Monaghan, R.L.; Huang, L.; Albers-SchOnberg, G. J. Am. Chem. Soc, 1983, 105,3672. [166] Huang, L.; Albers-Schonberg, G.; Monaghan, R.L.; Jakubas, K.; Pong, S.S.; Hensens, O.D.; Burg, R.W.; Ostund, D.A.; Conroy, J.; Stapley, E.O. J. Antibiot., 1984, 37, 970. [167] Otoguro, K.; Liu, Z.-X.; Fukuda, K.; Li, Y.; Iwai, Y.; Tanaka, H.; Omura, S. J. Antibiot.,\9S8,4\,573. [168] Otoguro, K.; Nakagawa, A.; Omura, S. J. Antibiot., 1988, 41, 250. [169] Lacey, E.; Gill, J.H.; Power, M.L.; Rickards, R.W.; O'Shea, M.G.; Rothschild, J.M. Int. J. Parasit., 1995, 25, 349. [170] Nair, M.G.; Putnam, A.R.; Mishra,S.K.; Taft, W.H.; Keller, J.E.; Miller, J.R.; Zhu, P.-P.; Meinhart, J.D.; Lynn, D.G.J. Nat. Prod., 1989, 52, 797. [171} Nair, M.G. In Natural and Engineered Pest Management Agents, Hedin, P.A.; Menn, J.J.; Hollingworth, R.M. Eds.; ACS Symposium Series 551; American Chemical Society; Washington, DC, 1994; pp. 145-161. [172] Ovenden, S.P.B.; Capon, R.J.; Lacey, E.; Gill, J.H.; Friedei, T.; Wadsworth, D. J. Org Chem., 1999,64, 1140. [173] Pollack, Y.; Segal, R.; Golenser, J. Parasitology Res., 1990, 76, 570. [174] Bruneton, J. Pharmacognosy, Phytochemistry, Medicinal Plants, Lavoisier Publishing: Paris, 1995. [175] Soler-Seratosa, A.; Kokalis-Burelle, N.; Rodriguez-Kabana, R.; Weaver, C.F.; Kng, P.S. Nematropica, 1996, 26, 57. [176] Ishikawa, M.; Shuto, Y.; Watanabe, H. Agric. Biol. Chem., 1986, 50, 1863. [177] Bauske, E.M.; Rodriguez-Kabana, R.; Estaun, V.; Kloepper, J.W.; Robertson, D.G.; Weaver, C.F.; King, P.S. Nematropica, 1994, 24, 143. [178] Osman, A.A.; Viglierchio, D.R. Rev. NematoL, 1988, 11, 93. [179] Viglierchio, D.R.; Wu, F.F. Nematropica, 1989, 19, 75. [180] Malik, M.S.; Sangwan, N.K.; Dhindsa, K.S.; Verma, K.K.; Bhatti, D.S. Pesticides, 1987,21,30. [181] Sangwan, N.K.; Verma, B.S.; Verma, K.K.; Dhinda, K.S. Pest. Sci, 1990, 28, 331. [182] Stark, A.; Anke, T.; Mocek, U.; Steglich, W.; Kirfel, A.; Will, G.Z. Z. Naturforsch., 1988,43c, 177. [183] Bochis, R.J.; Fisher, M.H. Tetrahedron Letters, 1968, 1971. [184] Schmid, H.; Snatzke, G.; Snatzke, F.; Nagarajan, K. N.; Peter, M. G. Helv. Chim. Acta, 1974,57,32. [185] Kumar, D.; Mishra, S.K.; Tandan, S.K.; Tripathi, H.C. Indian J. Pharmacol., 1995,27, 161. [186] Sathianesan, V.; Kumar, N.G.; Reghunandanan, V.R. Kerala J. Vet. ScL, 1984, 15, 159. [187] Yue-Ming, L.; Mackintosh, C; Casida, J.E. Biochem. Pharmacol., 1993,46, 1435. [188] Enz, A.; Zenke, G.; Pombo-Vilar, E. Bioorg. Med. Chem. Lett., 1997, 7, 2512. [189] Rydberg, D B.; Meinwald, J. Tetrahedron Letters, 1996, 37, 1129. [190] Dauben, W.G.; Lam, J.Y.L.; Guo, Z.R. J. Org Chem., 1996, 61,4816. [191] Suga, T.; Ohta, S.; Munesada, K.; Ide, N.; Kurokawa, M,; Shimizu, M.; Ohta, E. Phytochemistry, 1993, 33, 1395. [192] Suzuki, J.; Murata, I.; Enokida, R.; Yasuda, I. Tokyo-toritsu Eisei Kenkyusho Kenkyu Nempo, 1994,45, 35. Chem. Abstr., 1995, 123,246130. [193] Stadler, M.; Anke, H.; Sterner, O. Tetrahedron, 1994, 50, 12649 [194] Stadler, M.; Anke, H.; Sterner, O. J. Antibiot., 1994,47, 1284. [195] McCowen, M.C.; Callender, M.E.; Lawlis, J.F. Science, 1951, 112,202. [196] Katznelson, H.; Jamieson, C.A. Science, 1952,115, 70. [197] Killough, J.H.; Magill, G.B.; Smith, R.C. Science, 1952, 115, 71

501

[198] Curry, A. In Microbiology and Microbial Infections; Cox, F.E.G.; Kreier, J.P.; Wakelin, D. Eds., vol. 5; Arnold, Hodder Headline Group, London, 1998, pp 411-430. [199] Ingber, D.; Fujita, T.; Kishimoto, S.; Sudo, K.; Kanamaru, T.; Brem, H.; Folkman, J. Nature, 1990, 348, 555. [200] Yoshida, T.; Kaneko, Y.; Tsukamoto, A.; Han, K.; Ichinose, M.; Kimura, S. Cancer Res., 1998,58,3751. [201] Griffith, E.C.; Su, Z.; Niwayama, S.; Ramsay, C.A.; Chang, Y.-C; Liu, J.O. Proc. Natl. Acad. Sci. USA, 1998, 95, 15183. [202] Liu, S.; Widom, J.; Kemp, C.W.; Crews, CM.; Clardy, J. Science, 1998,282, 1324. [203] Vosburg, D.A.; Weiler, S.; Sorensen, E. J. Ange\v. Chem. Int. Ed., 1999, 38, 971. [204] Birch, A.J.; Hussain, S.F. J. Chem. Soc. (C), 1969, 1473. [205] Cordeil, G.A. Chem. Rev., 1976, 76,425. [206] Turner, W.B.; Aldridge, D.C. Fungal Metabolites II, Academic Press: London. 1983. [207] Schaeffer, J.M.; Frazier, E.G.; Bergstrom, A.R.; Williamson, J.M.; Liesch, J.M.; Goetz, M.A. J. Antibiot., 1990, 43, 1179. [208] Lorenzen, K.; Anke, T.; Anders, U.; Hindermayr, H.; Hansske, F. Z Naturforsch., 1994, 49C, 312. [209] Lim, C. H.; Miyagawa, N.; Akamatsu, N.; Nakagawa, Y.; Ueno, T. J. Pest. Sci., 1998,23,281. [210] Machida, T.; Higashi, K.; Ogawara, H. J. Antibiot., 1995,48, 1076. [211] Hain, R. In Pesticide Chemistry and Dioscience, Brooks, G.T.; Roberts, T.R. (Eds.); Royal Society of Chemistry, Cambridge, 1999; pp. 190-203. [212] Alphcy, T.J.W.; Robertson, W.M.; Lyon, G.D. Rev. Nematoi, 1988, 11, 399. [213] Zinovieva, S.V.; Chalova, L.I. Helminthologia, 1987, 24, 303. [214] Desjardins, A.E.; McCormick, S.P.; Plaisted, R.L.; Brodie, B.B. J. Agric. Food Chem., 1997,45,2322. [215] Jonassohn, M. Sesqiiiterpenoid unsaturated aldehydes. Ph. D. Thesis, Lund University, 1996; pp. 1-83. [216] Rogers, S.D.; Paul, V.J. Mar. Ecoi: Prog Ser., 1991, 77, 221. [217] Powell, G.; Hardie, J.; Pickett, J.A. Physiol. Entomoi, 1995, 20, 141. [218] Sterner, 0.; Bergman, R.; Kihiberg, J.; Wickberg, B.J. Nat. Prod, 1985,48, 279. [219] Stadler, M.; Sterner, O. Phytochemistry, 1998, 49, 1013. [220] Heinrich, M.; Robles, M.; West, J.A.; Ortiz de Montellano, B.; Rodriguez, E. Ann. Rev. Pharmacol. Toxicol., 1998, 38, 539. [221] Cavier, R.; Notteghem, M.J. Therapie, 1960, 15, 1138. [222] Badrina, F.A.; Zaghloul, A.M.; Maatooq, G.T.; El Sharkawy, S.H. Int. J. Pharmacogn., 1997, 35, 375. [223] Van Sumere, C.F., Lea, P.J. The biochemistry ofplant phenolics. Annual Proceedings of the Phytochemical Society of Europe, Clarendon: Oxford, 1985. [224] Mahajan, R.; Singh, P.; Bajaj, K.L.; Kalsi, P.S. Nematologica, 1986, 32, 119. [225] Ozeki, S.; Kotake, M.; Hayashi, K. Pro. Imp. Acad, 1936, 12,233. Chem. Abstr., 1937,31, 1878. [226] Horton, P.; Inman, W.D.; Crews, P. J. Nat. Prod, 1990, 53, 143. [227] Crews, P.; Matthews, T. R.; Horton, P. US Patent 4, 943, 589; 1990. Chem. Abstr., 1991, 114,35981. [228] Veech, J.A. J. Nematoi, 1979, 11, 240. [229] Stipanovic, R.D.; Bell, A.A.; Benedict, C.R. In Biologically Active Natural Products: Agrochemicals; Cutler, H.G.; Cutler, S.J., Eds.; CRC Press: Boca Raton, 1999; pp. 211-220. [230] Simpson, J.; Garson, M.J. Tetrahedron Letters, 1999,40, 3909. [231] Simpson, J.; Garson, M. J. Tetrahedron Letters, 1998, 39, 5819. [232] Alvi, K.A.; Tenenbaum, L.; Crews, P. J. Nat. Prod, 1991, 54, 71. [233] Ciancio, A. Fundam. Appl. Nematoi, 1995, 18, 451. [234] Merritt, A.T.; Ley, S.V. Nat. Prod Rep., 1992, 9, 243. [235] Munakata, K. In Advances in Pesticide Science, GeissbUhler, H., Ed.; 4th International Congress of Pesticide Chemistry, Pergamon Press: Oxford, 1978, pp. 295-302. [236] Kogiso, S.; Wada, K.; Munakata, K. Agric. Biol. Chem., 1976,40,2119. [237] Shimomura, M.; Miyaoka, H.; Yamada, Y.Tetrahedron Letters, 1999,40, 8015. [238] Miyaoka, H.; Shimomura, M.; Kimura, H.; Yamada, Y.; Kim, H.-S.; Wataya, Y. Tetrahedron, 1998, 54, 13467. [239] Omar, S.; Albert, C; Fanni, T,; Crews. P. J. Org Chem., 1988,53, 5971.

502

[240] Conrad, J.; Vogler, B.; Klaiber, Roos, G.; Walter, U.; Kraus, W. Phytochemistry, 1998,48,647. [241] Njoku, C.J.; Zeng, L.; Asuzu, I.U.; Oberlies, N.H.; McLaughlin, J.L. Int. J. Pharmacol., 1997,35, 134. [242] Fayez, M.B.E.; Negm, S.A.R.; Sharaf, A. Planta Med, 1963, 11, 439. [243] Shu, e x . ; Liang, X.T. Acta Chim. Sinica, 1980, 38, 196. [244] Akhtar, M.; Mahmood, I. In Tests ofAgrochemicals and cultivars. Association of Applied Biologists: Wellesboume, 1996. [245] Siddiqui, M.A.; Alam, M.M. Ann. Appl. Biol., 1988, 20, 1212. [246] Tanaka, M.; Aoki, M.; Aoki, H. Jpn Kokai Tokkyo Koho JP 09 67, 221,1977. Chem. Abstr., 1997, 126, 273675. [247] Prot, ].-C.', Komprobst, J.-M. C. R. Acad Sci. Ser. Ill, 1983, 296, 555. [248] Prot, J.-C. Komprobst, J.-M. Rev. NematoL, 1985, 8, 383. [249] Masamune, T.; Anetai, M.; Takasugi, M.; Katsui, N. Nature, 1982, 297,495. [250] Fukusawa, A.; Matsue, H.; Ikura, M.; Masamune, R. J. Chem. Soc. Chem. Comm., 1985, 222 (Correction: J. Chem. Soc. Chem. Comm., 1985, 748). [251] Corey, E.J.; Houpis, I.N. J. Am. Chem. Soc, 1990, 112, 8997. [252] Corey, E.J.; Hong, B. J. Am. Chem. Soc, 1994, 116, 3149. [253] Fukusawa, A.; Matsue, H.; Ikura, M.; Masamune, R. Tetrahedron Letters, 1985, 5539. [254] Kraus, G.A.; Vander Louw, S.J.; Tylka, G.L.; Soh, D.H. J. Agric Food Chem., 1996,44, 1548. [255] Emerson, R.W.; Crandall, B.C. W^O96,41,528; 1966. Chem. Abstr., 1997, 126, 128284. [256] Ghosh, M.; Sinhababu, S.P.; Sukul, N.C.; Mahato, S.B. Ind J. Exp. Biol., 1993, 31,604. [257] Roy, D.; Sinhababu, S.P.; Sakul, N.C.; Mahato, S.B. Indian J. NematoL, 1993, 23, 152. [258] Chatterjee, R.K.; Fatma, N.; Murthy, P.K.; Sinha, P.; Kulshrestha, D.K.; Dhawan, B.N. Drug Dev. Res., 1992, 26, 67. [259] Singh, S.N.; Raina, D.; Chatterjee, R.K.; Srivastava, A.K. Helminthologia, 1998, 35, 173. [260] Allen, E.H.; Feldmesser, I. J. Phytopathol., 1970, 60, 1013. [261] Allen, E.H.; Feldmesser, J. y./Ve/wa^o/., 1971,3,58. [262] Ghosh, M.; Sinhababu, S.P.; Sukul, N.C.; Sahu, N.P.; Mahato, S.B. Int. J. Pharmacogn., 1994,32, 184. [263] Tsipouras, A.; Adefarati, A.A.; Tkacz, J.S.; Frazier, E.G.; Rohrer, S.P.; Birzin, E.; Rosegay, A.; Zink, D.L.; Goetz, M.A. Biorg Med. Chem., 1996, 4, 531. [264] Singh, S.B.; Smith, J.L.; Sabnis, G.S.; Dombrowski, A.W.; Schaeffer, J.M.; Goetz, M.A.; Bills, G.F. Tetrahedron, 1991, 47, 6931. [265] Luckner, M. Secondary Metabolism in Microorganisms, Plants, and Animals, Springer-Verlag:Berlin, 1990. [266] Barz, W.; Welle, R. In Phenolic Metabolism in Plants, Recent Advances in Phytochemistry, Stafford, H.A.; Ibrahim, R.K., Eds.: Plenum Press: New York, 1992, vol. 26, pp. 139-164. [267] Dixon, R.A.; Harrison, M.J.; Paiva, N.L. Physiol. Plant., 1995, 93, 385. [268] Scheffer, F.; Kickhut, R.; Visser, J.H. Z. Pjlanzenernaehr. Dueng Bodenk., 1962, 98, 114. Chem. Abstr., 1963, 58, 1279. [269] Maheshwari, D.K.; Anwar, M.J. Phytopathol., 1990, 129, 159. [270] Mahajan, R.; Singh, P., Bajaj, K.L. Rev. NematoL, 1985, 8, 161 [271] Hayashi, M.; Wada, K.; Munakata, K. Agr. BioL Chem., 1981,45, 1527. [272] Hayashi, M.; Wada, K.; Munakata, K. Agr. BioL Chem., 1983, 47, 2653. [273] Evans, P.H.; Bowers, W.S.; Funk, E.J. J. Agric Food Chem., 1984, 32, 1254. [274] Yoshizawa, Y.; Kawaii, S.; Kanauchi, M.; Chida, M.; Mizutani, J. BioscL Biotech. Biochem., 1993, 57, 1572. [275] Kawaii, S.; Yoshizawa, Y.; Mizutani, J. BioscL Biotech. Biochem., 1994, 58, 9982. [276] Mox\Qy,T.]. Economic Botany, 1994,48,397. [277] Perrett, S.; Whitfield, P. J. Phytotherapy Research, 1995, 9,405. [278] Morales-Ramirez, P.; Madrigal-Bujaidar, E.; Mercader-Martinez, J.; Casasni, M.; Gonzalez, G.; Chamorro-Cevallos, G.; Salazar-Jacobo, M. Mutation Research, 1992, 279, 269. [279] Ali, M.A. Bangladesh J. ScL Ind Res., 1994,29, 87. [280] Suga, T.; Ohta, S.; Munesada, K.; Ide, N.; Kurokawa, M.; Shimizu, M.; Ohta, E. Phytochemistry, 1993, 33, 1395. [281] Fetterer, R.H.; Fleming, M.W. Comp. Biochem. Physiol.-C, 1991, 100, 539.

503

[282] Binder, B.F.; DeMilo, A.B.; Kochansky, J.P.; Chitwood, D. J. Agric Food C/7ew.,1992,40, 1475. [283] Fodor, A.; Deak, P.; Kiss, I. Gen. Comp. Endocrmoi, 1982, 46, 99. [284] Fodor, A.; Timar, T.; Kiss, I.; Hosztafi, S.; Varga, E.; Soos, J.; Sebok, P. Gen. Comp. Endocrinol., 1989, 74, 18 [285] Fodor, A.; Timar, T. Gen. Comp. Endocrinol., 1989, 74, 32. [286] Burkhardt, G.; Becker, H.; Grubert, M.; Thomas, J.; Eicher, T. Phytochemistry, 1994,37, 1593. [287] Quaghebeur, K.; Coosemans, J.; Toppet, S.; CompemoUe, F. Phvtochemistry, 1994,37,159. [288] Davyt, D.; Enz, W.; Manta, E.; Navarro, G.; Norte, M. Nat. Prod. Lett., 1997, 9, 305. [289] Laliberte, R.; Campbell, D.; Brauderlein, F. Can. J. Pharm. Sci., 1967, 37. [290] Gonzalez, J.A.; Estevez-Braun, A. J. Agric. FoodChem., 1998, 46, 1163. [291] Gonzalez, J.A.; Estevez-Braun, A. Pest. Biochem. Physiol., 1997, 58, 193. [292] O'Neill. M.J. Parasitology Today, 1986, 2, 358. [293] Veech, J.A. J. NematoL, 1982, 14, 2. [294] Rich, J.R.; Keen, N.T.; Thomason, I.J. Physiol. Plant. Pathol, 1911, 10, 105. [295] Plowright, R.A.; Grayer, R.J.; Gill, J.R.; Rahman, M.L.; Harbome, J.B. Nematologica, 1996, 42, 564. [296] Cook, R.; Tiller, S.A.; Mizen, K.A.; Edwards, R. J. Plant Pathol., 1995, 146, 348. [297] Edwards, R.; Mizen, T.; Cook, R. Nematologica, 1999, 41,51. [298] Kuc, J. Ann. Rev. Phytopathol., 1995, 33, 275. [299] Kaplan, D.T.; Keen, N.T.; Thomason, I.J. J. Nematoi, 1978, 10,291. [300] Kaplan, D.T.; Keen, N.T. J. Nematoi, 1977, 9, 274. [301] Huang, J.S.; Barker, K.R. Plant Physiol., 1991, 96, 1302. [302] Kaplan, D.T.; Keen, N.T.; Thomason, IJ. Physiol. Plant. Pathol., 1980, 16, 319. [303] Chang, Y.-C; Nair, M.G.; Nitiss, J.L. J. Nat. Prod, 1995, 58, 1901. [304] Stadler, M.; Dagne, E.; Anke, H. Planta Med, 1994,60, 550. [305] Baldridge, G.D.; O'Neill, N.R.; Samac, D.A. Plant MoL BioL, 1998, 38, 999. [306] Njoku, C.J.; Hopp, D.C.; Aiala, F.; Asuzu, I.U.; McLaughlin, J.L. Planta Med., 1997,63,580. [307] Gonzalez, J.A.; Est^vez-Braun, A.; Est^vez-Reyes, R.; Ravelo, A.G. J. Chem. £co/., 1994, 20, 517. [308] Suga, T. Ger. DE 4, 207, 301; 1993. Chem. Abstr., 1993, 119, 175897. [309] ROedi, P.; Juch, M. Curr. Org Chem., 1999, 3, 623. [310] Jurgens, T.M.; Frazier, E.G.; Schaeffer, J.M.; Jones, T.E.; Zink, D.L.; Borris, R.P.; Nanakom, W.; Beck, H.T.; Balick, M.J. J. Nat. Prod, 1994, 57, 230. [311] Kiuchi, F.; Goto, Y.; Sugimoto, N.; Akao, N.; Kondo, K.; Tsuda, Y. Chem. Pharm. Bull., 1993,41, 1640. [312] Kamo, T.; Kato, N.; Hirai, N.; Tsuda, M.; Fujioka, D.; Ohigashi, H. Phytochemistry, 1998, 49, 1617. [313] Roth, G.N.; Chandra, A.; Nair, M.G. J. Nat. Prod, 1998, 61, 542. [314] Hosoi, S.; Kiuchi, F.; Nakamura, N.; Imasho, M.; All, M.A.; Sasaki, Y.; Tanaka, E.; Tsumamoto, Y.; Kondo,, K.; Tsuda, Y. Chem. Pharm Bull., 1999,47,37. [315] Goto, C; Kasuya, S.; Koga, K.; Ohtomo, H.; Kagei, N. Parasitology Research, 1990, 76, 653. [316] Lai, M.; Liu, M.; Liu, H. J. Am. Chem. Soc, 1991, 113, 7388. [317] Pechacek, J.T.; Bargar, T.M.; Sabol, M.R. Bioorg Med Chem. Lett., 1997, 7, 2665. [318] Borris, R.P.; Schaeffer, J.M. In Phytochemical Resources for Medicine and Agriculture, Nigg, H.N.; Seigler, D., Eds; Plenum Press:New York, 1992, pp. 117-158. [319] Ishizaki, T.; Kato, K.; Kumada, M.; Takemoto, T.; Nakajima, T.; Takagi, N.; Koike, K. Jap. J. ParasitoL, 1973,22,181. [320] Krogsgaard-Larsen, P.; Honore, T. Trends Pharm. Sci., 1983, 31,33. [321] Rohrer, S.P.; Evans, D.V.; Bergstrom, A.R. Comp. Biochem. Physiol, 1990, 95C, 223. [322] Scaeffer, J.M.; White, T.; Bergstrom, A.R.; Wilson, K.E.; Turner, M.J. Pest. Biochem. Physiol, 1990, 36, 220. [323] Takasugi, M.; Yachida, Y.; Anetai, M.; Masamune, T.; Kegawawa, K. Chem. Lett., 1915, 43. [324] Parry, R. J.; Mizusawa. A. E.; Chiu, I. C; Naidu, M. V.; Ricciardone, M. J. Am. Chem. Soc, 1985, 107,2512. [325] Blunden, G.; Gordon. S.M. Pharm. Int., 1986,7,162.

504

[326] Wu, Y.; Jenkins, T.; Blunden, G.; Von Mende, N.; Hankins, S.D. J. Appl. Phycoi, 1998, 10,91. [327] Wu, Y.; Jenkins, T.; Blunden, G.; Whapham, C; Hankins, S.D. Fund. Appl. NematoL, 1997,20,99. [328] Parsons, A.F. Tetrahedron, 1996, 57,4149 [329] Liu, J.S. Yaoxue Tongboa, 1979, 14, 256. [330] Takemoto, T.; Daigo, K. Archiv. Pharmazie, 1958,203. 627. [331] Dixon, G. K. In Crop Protection Agents from Nature, Copping, L. G., Ed.; Royal Society of Chemistry .'Cambridge, 1996; pp. 114-216. [332] Cavier, R.; Notteghem, M.J. Therapie, 1960, 15, 1138. [333] Cavier, R.; Notteghem, M.J. Therapie, 1960, 15, 1142. [334] Zaman, L.; Arakawa, O.; Shimozu, A.; Onoue, Y.; Nishio, S.; Shida, Y.; Noguchi, T. Toxicon, 1997, 35, 205. [335] Douglas, D.J.; Ramsey, U.P.; Walter, J.A.; Wright, J.L.C. J. Chem. Soc. Chem. Commun., 1992,714. [336] Konno, K.; Hashimoto, K.; Ohfune, Y.; Shirahama, H.; Matsumoto, T. J. Am. Chem. Soc, 1988, 110,4807. [337] Herbert, R.B. Nat. Prod Rep., 1999, 16, 199. [338] Fleming, F.F. Nat. Prod Rep., 1999, 16, 597. [339] Vaughn, S.V. In Biologically Active Natural Products: Agrochemicals; Cutler, H.GT; Cutler, S.J., Eds.; CRC Press: Boca Raton, 1999; pp. 81-91. [340] Halkier, B.A.; Du, L. Trends in Plant Science, 1997, 11, 425. [341] Chew, F.S. In Biologically Active Natural Products: Potential Use in Agriculture; Cutler, H. G., Ed.; ACS Symposium Series 380. American Chemical Society: Washington, D.C., 1988; pp. 155-181. [342] Halbrendt, J.M. J. Nematoi, 1996, 28, 8. [343] Lazzeri, L.; Tacconi, R.; Palmeri, S. J. Agr. Food Chem., 1993,41, 825. [344] Ellenby, C. Ann. Appl. Biol., 1951, 38, 859. [345] Donkin, S.G.; Eiteman, M.A.; Williams, P.L. J. Nematoi, 1995, 27, 258. [346] Potter, M.J.; Davies, K.; Rathjen, A.J. J. Chem. Ecoi, 1998, 24, 67. [347] Augustin, J.; Drobnica, L; Nemec, P.; Antos, K.; Kalamar, J.; Kristian, P. J. Antibiot., 1968,21,497. [348] Augustin, J.; Drobnica, L; Nemec, P.; Antos, K.; Kalamar, J.; Kristian, P.; Martvon, A.; Uher, M. J. Antibiot., 1968, 21, 504 [349] Piluk, J.; Hartel, P.G.; Haines, B.L. Plant Soil, 1998, 200, 27. [350] Kelly, D.P.; Wood, A.P.; Jordan, S.L.; Padden, A.N.; Gorlenko, V.M.; Dubinina, G.A. Biochem. Soc. Trans., 1994, 22, 1011. [351] BrUckner, H.; Reinecke, C; Kripp, T.; Kieb, M. Proceedings of the Fourth International Mycological Congress, Regensburg, Germany, 1990, p. 224. [352] Samson, M.S.P. Quart. Rev. Biophys., 1993, 26, 365. [353] Metzger, J.; Schlegel, B.; Fleck, W.F.; Domberger, K.; Ihn, W.; Schade, W. Can. Pat. Appl. CA 2,115,966, 1994. Chem. Abstr. 1995, 122, 104448. [354] von DOhren, H.; Keller, U.; Vater, J.; Zocher, R. Chem. Rev., 1997, 97, 2675 [355] Pleiss, U.; Turberg, A.; Harder, A.; Londerhausen, M.; Jeschke, P.; Boheim, G. J. Label. Comp. Radiopharm., 1996, 38, 651. [356] Sasaki, T.; Takagi, M.; Yaguchi, T.; Miyadoh, S.; Okada, T.; Koyama, M. J. Antibiot., 1992,45,692. [357] Terada, M. Jap. J. Parasitol., 1992,41, 108. [358] Conder, G.A.; Johnson, S.S.; Nowakowski, D.S.; Blake, I.E.; Dutton, F.E.; Nelson, S.J.; Thomas, E.M.; Davis, J.P.; Thompson, D.P. J. Antibiot., 1995,48, 820. [359] Gessner, G.; Meder, S.; Rink, T.; Boheim, G.; Harder, A.; Jeschke, P.; Scherkenbeck, J.; Londerhausen, M. Pest. Sci., 1996,48, 399. [360] Wang, M.; Watanabe, N.; Shomura, T.; Ohtomo, H. Jap. J Parasitol., 1995,44, 306. [361] Wang, M.; Watanabe, N.; Shomura, T.; Ohtomo, H. Jap. J Parasitol., 1994,43, 346. [362] Dutton, F.E.; Lee, B.H. Tetrahedron Letters, 1998, 39, 5313. [363] Ohyama, M.; linuma, K.; Isogai, A.; Suzuki, A. Biosci. Biotech. Biochem., 1994, 58, 1193. [364] Martin, R.J.; Harder, A.; Londershausen, M.; Jeschke, P. Pest. Sci., 1996,48, 343. [365] Scherkenbeck, J.; Harder, A.; Plant, A.; Dyker, H. Bioorg. Med. Chem. Lett., 1998,8,1035. [366] Scherkenbeck, J.; Dyker, H.; Gondol, D.; Harder, A.; Plant, A. Reichel, F. Pest. 5c/., 1999, 55, 457.

505

[367] Sterner, O.; Etzel, W.; Mayer, A.; Anke, H. Nat. Prod Lett., 1997, 10, 33. [368] Buchel, E.; Martini, U.; Mayer, A.; Anke, H.; Sterner, O. Tetrahedron, 1998, 54, 5345. [369] Buchel, E.; Mayer, A.; Martini, U.; Anke, H.; Sterner, O. Pest. ScL, 1998, 54, 309. [370] Mayer, A.; Anke, H.; Sterner, O. Nat. Prod Lett., 1997, 10, 25. [371] Mayer, A.; Hoster, K. M.; Sterner, O.; Anke, H. Pest. Sci., 1999, 55, 27. [372] Kawazu, K.; Murakami, T.; Ono, Y.; Kanzaki, H.; Kobayashi, A.; Mikawa, T.; Yoshikawa, N. Biosci. Biotech. Biochem., 1993, 57, 98. [373] Kawazu, K.; Kobayashi, A.; Yoshikawa, N.; Jpn. Kokai Tokkyo Koho JP 03, 251, 597, 1991. Chem. Abstr, 1992, 116, 172456. [374] Zampella, A.; Giannini, C; Debitus, C; Roussakis, C; D'Auria, M.V. J. Nat. Prod, 1999, 62, 332. [375] Crews, P.; Manes, L.V.; Boehler, M. Tetrahedron Letters, 1986, 27, 2797. [376] Sasse, F.; Kunze, B.; Gronewold, T.M.A.; Reichenbach, H. J. Nat. Cancer Inst., 1998,90, 1559. [377] Zabriskie, T.E.; Klocke, J.A.: Ireland, CM.; Marcus, A.H.; Molinski, T.F.; Faulkner, D.J.; Xu, C; Clardy, J.C. J. Am. Chem. Soc, 1986, 108, 3123. [378] Scott, V.R.; Boehme, R.; Matthews, T.R. Antimicrob. Agents Chemother., 1988, 32, 1154. [379] Braekman, J.C; Dalzoe, D.; Moussiaux, B.; Riccio, R.J. Nat. Prod, 1987, 50, 994. [380] Adamczeski, M.; Quiflo^, E.; Crews, P. J. Am. Chem. Soc, 1989, 11, 647. [381] D'Auria. M.V.; Giannini, C; Minale, L.; Zampella, A.; Debitus, C; Frostin, M. J. Nat. Prod, 1997,60,814. |382] Fernandez, R.; Dherbomez, M.; Letoumeux, Y.; Nabil, M.; Verbist, J.F.; Biard, J.F. J. Nat. Prod, 1999, 62, 678. [383] Adamczeski, M.; Quiflo^, E.; Crews, P. J. Org. Chem., 1990, 55, 242. [384] Quiiloi, E.; Adamczeski, M.; Crews, P. J. Org Chem., 1986, 51,4494. [385] Kleinkauf, H.; von DOrhen, H. In Fungal Biotechnology; Anke, T., Ed.; Chapman and Hall: Weinheim, 1997; pp. 147-161. [386] Schreier, M.H. In Fungal Biotechnology; Anke, T., Ed.; Chapman and Hall: Weinheim, 1997; pp. 137-146. [387] Borel, J.F. Pharmacol. Rev., 1989, 41, 259. [388] Schad, G.A. J. Inf. Dis., 1986, 153, [389] Bout, D.; Haque, A.; Capron, A. Trans. Roy. Soc. Trop. Med. Hyg., 1984, 78, 670. [390] Jimenez, C; Crews, P. J. Nat. Prod, 1990, 53, 978. [391] Birch, A.N.E.; Robertson, W.M.; Geoghegan, I.E.; McGavin, W.J.; Alphey, T.J.W.; Phillips, M.S.; Fellows, L.E. Watson, A.A.; Simmonds, M.S.J.; Porter, E.A. Nematologica, 1993, 39, 521. [392] Perrett, S.; Whitfield, P.J. Planta Med, 1995, 61, 276. [393] Fassuliotis, G.; Skucas, G.P. J. Nematoi, 1969, 1, 287. [394] Wink, M. In Phytochemistry and Agriculture; van Beek, T.A.; Breteler, H., Eds.; Oxford Science Publishers:Oxford, 1993; pp. 171-213. [395] Matsuda, K.; Yamada, K.; Kimura, M.; Hamada, M. J. Agr. Food Chem., 1991, 39, 189. [396] Matsuda, K.; Kimura, M.; Komai, K.; Hamada, M. Agr. Biol. Chem., 1989, 53, 2287. [397] Adamczeski, M.; Quirto^, E.; Crews, P. J. Am. Soc. Chem., 1988, 110, 1598. [398] Mulder, R.J.; Shafer, CM.; Molinski, T.F. J. Org Chem., 1999, 64, 4995. [399] Searle, P.A.; Richter, R.K.; Molinski, T.F. J. Org. Chem., 1996,61,4073. [400] Onda, M.; Abe, K.; Yonezawa, K.; Esumi, N,; Suzuki, T. Chem. Pharm. Bull, 1970, 18,435. [401] Conder, G.A.; Zielinski, R.J.; Johnson, S.S.; Kuo, M.-S.T.; Cox, D.L.; Marshall, V.P.; Haber, CL.; DiRoma, P.J.; Nelson, S.J.; Conklin, R.D.; Lee, B.L.; Geary, T.G.; Rothwell, J.T.; Sangster, N.C J. Antibiot., 1992,45, 977. [402] Knaggs, A.R. Nat. Prod Rep., 1999, 16, 525. [403] Watabe, H.; Amano, S.; Yoshida, J.; Takase, Y.; Miyadoh, S.; Sasaki, T.; Hatsu, M.; Takeuchi, Y.; Kodoma, Y. Meiji Seika Kenkyu Nenpo, 1988, 55. Chem. Abstr., 1989, 110, 169904. [404] Palermo, J.; Blanch Flower, P.; Seldes, A. Tetrahedron Letters, 1992,41, 3097. [405] Davyt, D.; Entz, W.; Fernandez, R.; Mariezcurrena, R.; Mombru, A.W.; Saldafia, J.; Domfnguez, L.; Coll, J.; Manta, E. J. Nat. Prod, 1998, 61, 1560. [406] Capon, R.J.; Skeene, C; Lacey, E.; Gill, J,H.; Wadsworth, D.; Friedel, T.; J. Nat. Prod. 1999, 62, 1256.

506

[407] Yamazaki, M.; Okuyama, E.; Kobayashi, M.; Inoue, H. Tetrahedron Letters, 1981,22,135. [408] Ondeyka, J.G.; Goegelman, R.T.; Schaeffer, J.M.; Keleman, L.; Zitano, L. J. Antibiot.,\990,43, 1375. [409] Gill, J.H.; Lacey, E. Int. J. Parasitoi, 1993, 23, 375. [410] Ostlind, D.A.; Mickle, G.W.; Ewanciew, D.V.; Andriuli, F.J.; Campbell, W.C. Res. Fe/. 5c/., 1990,48,261. [411] Shoop, W.L.; Egerton, J.R.; Eary, C.H.; Suhayda, D. J. ParasitoL, 1990, 76, 349. [412] Schaeffer, J.M.; Blizzard, T.A.; Ondeyka, J.; Goegelman, R.; Sinclair, P.; Mrozik, H. Biochem. Pharmacol., 1992,43, 679. [413] Polonsky, J.; Merrien, M.A.; Prange, T.; Pascard, C.; Moreau, S.J. Chem. Soc, Chem. Cowm., 1980,601. [414] Boysen, M.; Skouboe, P.; Frisvad, J.; Rossen, L. Microbiology, 1996, 142, 541. [415] Lee, B.H.; Clothier, M.F.; Pickering, D.A. J. Org. Chem., 1997, 62, 7836. [416] Lee, B.H.; Komis, G.I.; Cialdella, J.I.; Clothier, M.F.; Marshall, V.P.; Martin, D.G.; McNally, P.L.; Mizsak, S.A.; Whaley, H.A.; Wiley, V.H. J. Nat. Prod., 1997,60, 1139. [417] Kuo, M.S.; Wiley, V.H.; Cialdella, J.I.; Yurek, D.A.; Whaley, H.A.; Marshall, V.P. J. Antibiot., 1996, 49, 1006. [418] Banks, R.M.; Blanchflower, S.E.; Everett, J.R.; Manger, B.R.; Reading, C. J. Antibiot., 1997,50,840. [419] Inman, W.D.; O'Neill-Johnson, M.; Crews, P. J. Am. Chem. Soc, 1990, 112, 1. [420] West, R.R.; Mayne, C.L.; Ireland, CM.; Brinen, L.D.; Clardy, J. Tetrahedron Letters, 1990,31,3271. [421 ] Smith, C.J.; Venables, D.A.; Hopmann, C; Salomon, C.E.; Jompa, J.; Tahir, A.; Faulkner, D.J.; Ireland, CM. J. Nat. Prod, 1997, 60, 1048. [422] Ford, P.W.; Davidson, B.S.J. Nat. Prod, 1997,60, 1051. [423] Crews, P.; Matthews, T.; Inman, W.D.t/^ 4,959,370, 1990. Chem. Abstr., 1991, 114,55776 [424] Auger, J.; Lecomte, C; Thibout, E. Acta Bot. Gallica, 1993, 140, 157. [425] Block, E. Angew. Chem. Int. Ed, 1992, 31, 1135. [426] Nath, A.; Sharma, N.K.; Bhardwaj, S.; Thappa, CN. Nematologia, 1982, 28, 253. [427] Gupta, R.; Sharma, N.K. Int. J. Pest Manag, 1993, 39, 390. [428] Pfister, J.R. J. Nat. Prod, 1988, 51, 969 [429] Wright, D.E. Tetrahedron, 1979, 35, 1207. [430] Haneishi, T.; Arai, M.; Kitano, N.; Yamamoto, S. J. Antibiot., 1974. 27, 339. [431 ] Kawazu, K.; Kobayashi, A. Jpn. Kokai Tokkyo Koho JP 02 72175, 1990. Chem. Abstr., 1990, 113,96163. [432] Hamill, R.L.; Hoehn, M.M. J. Antibiot., 1964, 17, 100. [433] Bargmann. CI. Science, 1998, 282, 2028. [434] Mombaerts, P. Science, 1999,286, 707.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

507

RESVERATROL OLIGOMERS: STRUCTURE, CHEMISTRY, AND BIOLOGICAL ACTIVITY ROBERT H. CICHEWICZ ^ AND SAMIR A. KOUZI ^ ^ Department of Horticulture, Michigan State University, East Lansing, Michigan 48824, ^ School ofPharmacy, University of the Pacific, Stockton, California 95211, USA ABSTRACT: Both plants and fungi have been reported to be capable of polymerizing the natural bioactive trihydroxystilbene resveratrol into a host of complex oligomers. In recent years, these resveratrol oligomers have received the attention of natural products researchers worldwide as scientists have come to appreciate their pharmacological potential. During this time, the number of known resveratrol oligomers has swelled with scores of new compounds reported. In this chapter, the structure, chemistry, and biological activity of the resveratrol oligomers will be reviewed. This review consists of a brief assessment of previous resveratrol research followed by a thorough examination of the resveratrol ohgomers including their roles as phytoalexins, detoxification products of the fungal metabolism of resveratrol, and potential medicinal agents for humans. A detailed analysis of the methods used for the isolation and structure elucidation of these compounds is provided.

INTRODUCTION Stilbenes represent a unique class of biologically active natural products produced primarily by plants. The stilbene nucleus is based on a 14carbon skeleton composed of two phenyl rings joined by an ethylene bridge. One of the most well known and widely distributed of the stilbenes is the compound resveratrol (3,4',5-trihydroxystilbene) (1) (the structures of all compounds noted in this review can be found in Appendix (1)). Since its original description by Michio Takaoka [1] as a constituent of the roots of the white hellebore (Veratrum grandiflorum O. Loes.), 1 has been identified as a component of numerous plant species, several of which are utilized by humans for food and medicinal purposes [2]. To date, a variety of biological activities have been ascribed to 1 including cancer chemopreventive and cardiovascular modulating effects [2].

508

The biosynthesis of the stilbenoids, including 1, has been previously reviewed. Briefly, the synthesis of 1 is dependent upon a single key enzyme known as stilbene synthase or resveratrol synthase as part of a mixed phenylpropanoid-polyketide pathway [2,3,4,5,6] (Fig. (1)). Stilbene synthase catalyzes the formation of 1 through the condensation of one /?-coumaroyl CoA and three malonyl CoA molecules, both of which are ubiquitous intermediary plant metabolites. Stilbene synthase is not expressed constitutively in plants. Instead, resveratrol synthase is expressed in response to stress-factors such as injury to the plant. Several factors are capable of initiating the synthesis of resveratrol in plants by facilitating the up-regulation of stilbene synthase. Some of these initiating factors include ultraviolet irradiation, chemical stimuli, mechanical injury, and microbial attack [2]. Under these conditions, 1 is believed to function as a phytoalexin-like stress metabolite. For example, when Vitis spp. are exposed to cultures of the fungal grapevine pathogen Botrytis cinerea, the plants respond immediately with a localized up-regulation of stilbene synthase and the subsequent accumulation of 1 at the site of inoculation [2]. Studies have demonstrated that 1 plays a vital defensive role in plants. It has been shown that 1 possesses modest in vitro antimicrobial properties [2,7]. In addition, several oligomeric metabolites of 1 are known to exhibit greater antimicrobial activity [8]. Further evidence for the important defensive role of 1 in plants comes from studies in which transgenic barley, wheat, and tobacco plants containing a stilbene synthase gene exhibit enhanced resistance to fungal infection [9,10]. In addition to the phytoalexin activity of 1, a multitude of studies have examined the potential health effects of the consumption of 1 by humans. Research has revealed a number of intriguing health benefits associated with the consumption of 1. Two of the most promising of these potential phytoceutical applications are the cardioprotective and cancer chemopreventive activities of 1.

509

OH NH2 phenylalanine

phenylalanine ammonia lyase resveratrol

^ ^ NH3 1

'

0

ii

u

^-'V^^^^OH /-^

4CO2

cinnamic acid stilbene synthase (resveratrol synthase) cinnamate-4-hydroxyl ase + 3 HOOC-CH2CO-SC0A 1

^

0 II

ff'V^^'^^ "OH 1

0 CoA ligase + CoASH

^

^-^^^

—^

^^''^>v. ^^^"^^^

ff^^^^^^

SCoA

« o ^

p-coumaric acid

/>-coumaroyl-CoA

Fig. (1). Bio synthesis of resveratrol (1)

Cardioprotective Properties of Resveratrol Oxidative processes are believed to play a significant role in the development of vascular tissue damage and promotion of cardiovascular disease. Several studies [11,12,13,14] have now demonstrated that 1 possesses potent in vitro antioxidant effects; however, its in vivo activity is questionable [15]. Once oxidative tissue damage has occurred, blood platelet accumulation at the site of injury can lead to localized blood vessel occlusion or facilitate the development of plaques that may breakfree resulting in the formation of blockages in the peripheral vasculature. Therefore, drug therapies targeted at inhibiting platelet aggregation have been employed to help reduce the chance of clot formation in high-risk patients. Research now indicates that 1 may play a role in stemming this

510

risk as an inhibitor of the blood clotting process [16]. Furthermore, 1 has been shown [17,18,19] to be a potent in vitro inhibitor of platelet aggregation induced by arachidonic acid, ADP, and collagen; however, its in vivo activity appears to be more modest [16]. Vascular tissue damage and subsequent platelet clotting processes are also influenced to a great extent by leukocyte activity. Some leukocytes are responsible for the release of a host of pro-inflammatory substances that lead to additional vascular tissue damage and promote further platelet aggregation. In vitro tests [20] employing isolated leukocytes have demonstrated that 1 inhibits both the expression and subsequent release of several pro-inflammatory substances associated with the development of vascular tissue damage. In addition, studies have shown that 1 interferes with blood vessel endothelial cell adhesion molecule expression [21] and other associated tissue factors [22], thus providing further protection from leukocyte induced vascular injury. Furthermore, recent evidence now indicates that 1 possesses estrogenic properties that may convey additional cardioprotective effects [23,24,25]. In light of these cardioprotective properties exhibited by 1, it is conceivable that 1 may help to reduce the risk of developing cardiovascular disease in humans. Cancer Chemopreventive Properties of Resveratrol Research has also shown that 1 possesses potent cancer chemopreventive activity as exhibited by the inhibition of the initiation, promotion, and progression of cancer cell growth [26,27]. Several in vitro studies have demonstrated that 1 prevents the initiation of tumor formation through antimutagenic effects [28], inhibition of the expression of the phase I enzyme CYPlAl that is responsible for procarcinogen activation [29], and induction of the carcinogen-detoxifying phase II enzyme quinone reductase [26]. Further research has demonstrated that 1 inhibits the promotion of cancer cell growth through the inhibition of the enzymes responsible for the production of cancer promoting agents. For example, 1 appears to interfere with prostaglandin synthesis by directly inhibiting cyclooxygenase I and II activity and by further suppressing COX II expression [26,30]. In addition, 1 blocks the progression of cancer cell growth as illustrated in in vitro studies employing human estrogen dependent and independent epithelial breast carcinoma [31] and promyelocytic leukemia cell lines [26]. In these studies, 1 was noted to

511 alter cancer cell growth by preventing exponential growth, reducing cell viability, and promoting cellular differentiation to a nonproliferative phenotype. In related in vivo studies employing animal models, 1 has also been shovm to be a potent inhibitor of tumor development. In studies using both mice [26] and rats [32], 1 selectively inhibited tumor growth while failing to produce any noted systemic toxicity. Additional studies attempting to further elucidate the mechanisms behind the cancer chemopreventive and inhibitory effects of resveratrol have noted several possible mechanisms of action. For example, the cancer inhibitory effects of 1 may be mediated in part through the inhibition of protein tyrosine kinase and protein kinase C signal transduction pathways [33,34]. Further studies lending support to the anticancer activity of 1 demonstrate that it can inhibit ribonucleotide reductase activity [35], cleave DNA through the formation of an oxidative copper complex [36], and induce apoptosis in growing cancer cells [37,38]. Considering the promising pharmacological properties of 1, it is conceivable that 1, or one of its analogs, could serve as an important cancer chemopreventive agent for humans. In light of the important role that 1 plays in plants and its potential pharmacological applications for humans, significant interest has developed to identify new analogs of 1 that may possess interesting biological activity. One group of derivatives of 1 that have received significant attention in recent years is the resveratrol oligomers. To date, a host of resveratrol oligomers have been described. These compounds have demonstrated an array of promising pharmacological properties. In addition, new research is providing valuable insight into the in vivo roles that these compounds play in plants.

THE RESVERATROL OLIGOMERS The resveratrol oligomers are composed of a diverse assemblage of polyphenolic compounds derived from the trihydroxystilbene monomer resveratrol (1). Previously, in a review by Sotheeswaran and Pasupathy [7], the authors proposed a simple dichotomous classification scheme for categorizing these compounds based on a single structural feature. Under

512

this system, 'group A' oligomers consisted of those compounds that contain one or more five-membered oxygenated heterocyclic ring systems as typified by a dihydrobenzofuran moiety. In contrast, all compounds that lacked this moiety were assigned to 'group B'. While this categorization method does provide a simple means for gross structural classification, it is readily apparent that this system lacks any biogenic foundation. Therefore, due to this organizational scheme's limited applicability, it will not be utilized in this review. Instead, we have chosen to present the resveratrol oligomers based on their biogenic origin. Resveratrol oligomers have been isolated from a relatively small assemblage of plant families. More recently, resveratrol oligomers have also been identified as fungal detoxification products of resveratrol metabolism. In Table (1), a summary of all the reported occurrences of resveratrol oligomers based on their biogenic origins is presented. This information was compiled following a thorough review of the available literature. Several computerized databases were employed for this purpose including AGRICOLA®, AGRIS®, Biological Abstracts®, Chemical Abstracts®, Current Contents®, MEDLINE®, and TOXLINE®. In Table (1), the names of the resveratrol oligomers and their sources are provided followed by a description of the oligomer type (i.e. dimer, trimer, tetramer, or pentamer), molecular formula, and reported optical rotation. In almost all cases, the resveratrol oligomers have been isolated from their natural sources in an optically active form. There are several cases reported in which both the (+) and (-) enantiomeric forms of a resveratrol oligomer have been isolated from different plant families. In some instances, researchers have been able to assign the absolute stereochemistry of each enantiomer; however, there are still several resveratrol oligomers whose absolute configuration remains unknown. In order to avoid confusion regarding the numbering of the resveratrol oligomers in this text, we have adopted a simple classification scheme in

Table 1.

Sources of Resveratrol Oligorners Source

iPlACEAE beniculirm vulgure Miller

Part Examined

fruit

Compound

foeniculoside I (2a)

fruit

foeniculoside I1 (3a)

fruit

foeniculoside I11 (4a)

fruit

foeniculoside IV (5a)

fruit

miyabenol C (6a)

fruit

cis-miyabenol C (7a)

NA'

(+)-a-viniferin (8a)

NA

kobophenol A (9)

NA

pallidol (10)

roots and rhizomes roots and rhizomes

.c-viniferin (1 1 )

Oligomer Type (Molecular Formula) trimer + 1 Glc (C4xH420I 4) trimer + 1 Glc (C48H42014) trimer + 2 Glc (C54HS2019) trimer + 3 Glc (C60H62024) trimer (C42H3209) trimer (C42H3209)

Optical Rotation (Concentration, Solvent)

+46.8" (c 0.9, MeOH)

+89.1" (c 0.9, MeOH) +4 1.9" (c 1.0, MeOH)

+40.7" (c 0.8, MeOH)

+22 I .4" (c 1.1, MeOH)

+98.5" (c 1.0, MeOH)

YPERACEAE arex ciliuto-marginuta Nakai

arex fedia Nees ex Wight

var. miyabei (Franchet) T. Koyama

miyabenol A (12b)

trimer (C42H3009) tetramer (C56H440 13) dimer (C2XH2206) dimer (C28HZ206) tetramer (cS6H42012)

NA NA NA

NA -91" (c 0.18, MeOH)

Table 1.

Sources of Resveratrol Oligorners (Continued) Part Examined

Compound

Oligomer Type (Molecular Formula)

roots and rhizomes roots and rhizomes NA

miyabenol B (13b)

(+)-a-viniferin (82)

NA

kobophenol A (9)

NA

pallidol (10)

roots

(+)-a-viniferin (8a)

tetramer (C56H400I 2 ) trimer (C42H3209) trimer (C42H3009) tetramer (CS6H440 13) dimer (C28H22°6) trimer (C42H3009) dimer (C28H2206) tetramer (C56H44013) trimer (C42H3209) trimer (C42H3009) tetramer (CS6H44013) dimer (C28H2206)

miyabenol C (6a)

roots and rhizomes

(-)-&+inifwin ( l l b )

roots and rhizomes roots and rhizomes NA

kobophenol A (9a)

(+)-a-viniferin (8a)

NA

kobophenol A (9)

NA

pallidol (10)

miyabenol C (6a)

Optical Rotation (Concentration, Solvent) -126" (c 0.23, MeOH) +69" (c 0.07, MeOH)

NA NA NA +S I .4" (c 1 0 , EtOH)

-46" (c 0.5, MeOH) +227" (c 0 17, MeOH)

+206" (c 0.070. MeOH)

NA NA NA

Table 1.

Sources of Resveratrol Oligomers (Continued) Source

Part Examined

Compound

Oligomer Type (Molecular Formula)

Optical Rotation (Concentration, Solvent)

NA

(+)-a-viniferin (8a)

NA

NA

kobophenol A (9)

NA

pallidol (10)

roots and rhizomes roots and rhizomes

(-)-c-viniferin ( I l b )

roots and rhizomes roots and rhizomes roots and rhizomes

kobophenol B (16b)

trimer (C42H3009) tetramer (CS6H44013) dimer (C28H2206) dimer (c281.12206) tetramer (C56H42012) tetramer (C56H40O 12) tetramer (cS6H42012) trimer (C42H3209)

heartwood

(-)-hopeaphenol(14b)

alanocarpus zeylanrcus Trimen

bark

balanocarpol(17b)

opea brevipetiolarrs (Thwaites) P.S. Ashton

bark

copalliferol A ( H a )

hrex multifolia Ohwi

'arexpumila Thunb

NPTEROCARPACEAE alanocarpus heimii King

hopeaphenol(14b)

miyabenol A (12) miyabenol C ( 6 )

tetramer (C56H42012) dimer (C28H2207) trimer (C42H3209)

NA NA

(c

-27.9" 0.843, MeOH)

-370" (EtOH) - 199" (c 0.17, MeOH) NA

NA

-407" (EtOH) -17" (c 0.005, MeOH) +115.6" (MeOH)

Table 1.

Sources of Resveratrol Oligomers (Continued) Source

Hopea cordfolio Trimen

Part Examined

Compound

bark

copalliferol A (18a)

Oligomer Type (Molecular Formula)

Optical Rotation (Concentration, Solvent)

trimer

+115.6" (MeOH) -5.6" (MeOH) -17" (c 0.005, MeOH)

(C42H3209)

bark

stemonoporol (19a)

trimer (C42H3209)

Hopeajucunda Thwaites

bark

balanocarpol (17b)

dimer (C28H2207)

Vopea malibato Foxworthy

leaves

balanocarpol (17b)

dimer (C28H2207)

leaves leaves

dibalanocarpol (ZOb) maliabatol A (21b)

Yopea odorala Roxb.

heartwood

malibatol B (22b) (-)-hopcapheno1 (14b)

tetramer

-227" (c 0.73, MeOH)

dimer

bark bark

distichol (23b) E-viniferin (11)

-38" ( c 0.37, MeOH)

dimer

-77"

(C28H2009)

(c 0.06, MeOH)

tetramer

(C4ZH3209)

-407" (EtOH) -44" (MeOH)

dimer

NA

(CS6H42012)

ihorea disticha (Thwaites) P.S. Ashton

0.52, MeOH)

(CShH42014)

(C28H2207)

leaves

-18" (c

trimer

(C28H2206)

;horea robusra A. DC

heartwood

hopeaphenol(14b)

tetramer

-402.9"

(C56H42012)

;horea lalura Roxb.

heartwood

hopeaphenol(14b)

tetramer (C~6H42012)

-402.9"

Table 1.

Sources of Resveratrol Oligomers (Continued)

I Source

Part Examined

Compound

Oligomer Type (Molecular Formula) trimer (C4ZH3209) trimer (C42H3209) trimer (C42H3209) trimer (C4ZH3209) tetramer (CS6H12012) tetramer (CS6H42012) trimer (C42H3209) trimer (C42H3209) trimer (C42H3209) trimer (C42H3209) tetramer

I Shorea stipularis Thwaites

bark

copalliferol A (18a)

Istemonoporus affinis

bark

copalliferol A (18a)

bark

stemonoporol (19b)

bark

canaliculatol (24b)

bark

vaticaftinol (25b)

bark

vaticaffmol (25b)

bark

copalliferol A (18a)

I

Thwaites

Stemonoporus cordfolius (Thwaites) Alston Stemonoporus elegans (Thwaites) Alston

Stemonoporus kannelryensis A.J.G.H. Kostermans

Stemonoporus lancijolius (Thwaites) P.S. Ashton Stemonoporus oblongifolius Pierre

bark

stemonoporol (19b)

bark

copalliferol A (18a)

bark

stemonoporol (19b)

bark

vaticaftinol(25b)

bark

copalliferol A (18a)

(C56H4201

2)

trimer (C42H3209)

Optical Rotation (Concentration, Solvent)

+ I 15.6" (MeOH) + I 15" (MeOH) -5.58" (MeOl-I) -25.5" (c 0.003, MeOH) -22" (MeOH) -22.5" (MeOH) +115"

(MeOH) -5.58" (MeOH) + I 15" (MeOH) -5.58"

(MeOH) -22.5" (MeOH) + I 15" (MeOH)

Sources of Resveratrol Oligomers (Continued)

Table 1.

I'atica af/inis Thwiates

Vatica diospyroides Symington

Valeria copallifera (Retz.) Alston

Valeria copallifera (Retz.) Alston

FABACEAE Caragana chamlagu Lam

Part Examined

Compound

Oligomer Type (Molecular Formula)

Optical Rotation (Concentration, Solvent)

bark

stemonoporol(19h)

bark

6-viniferin (11)

-5.58" (MeOH) NA

bark

vaticaffinol(25b)

trimer (C42H3209) dirner (C28fi2206) tetramer (C56H42012) tetramer (C56H4.2012) tetramer (C56H42012) trimer (C42H3209) trirner (C42H3209) trimer (C42H3209) trinier (C42H3209)

stem

vatdiospyroidol (26b)

stem

vaticaphenol A (27h)

bark

copalliferol A (18)

bark

copalliferol B (28h)

bark

copalliferol A (18a)

bark

stemonoporol(19b)

roots

(+)-a-viniferin (8a)

roots

caraganaphenol A (29a)

-22.5' (MeOH) -67" ( c 0.1, MeOH) -29" (c 0.1, MeOH) NA -186" (pyridine)

+ I 15.6" (MeOH) -5.6" (MeOH)

trimer (C42H3009) tetramer

+188.5"

(C56H42013)

(c 0.785, MeOH)

+50.7" (EtOH)

Table 1.

Sources o f Resveratrol Oligomers (Continued) Part Examined

Compound

roots

kobophenol A (9a)

roots

(+)-a-viniferin (8a)

roots

kobophenol A (9a)

roots

miyabenol C (6)

roots

davidiol D (30b)

roots

(-)-E -vini ferin (1 1b)

roots

hopeaphenol(14b)

roots roots

roots roots roots

leachianol A (31b) leachianol B (32a) leachianol C (33b) leachianol D (34b) leachianol E (3Sb)

Oligomer Type (Molecular Formula) tetramer (c56H44°13) trimer (C42H3009) tetramer (C5~3~44~13) trimer (c42113209) pentamer (C70H53015) dimer (C2SH2Z0G) tetramer (C56H42012) trimer (C4ZH3209) trimer (C42H3209) tetramer (C56H44012) trimer (C42H3209) trimer (C4ZH3209)

Optical Rotation (Concentration, Solvent) +199.4" +52.9" (c 0.52, EtOH)

+250.6" (c 0.39, MeOH) NA

-135" (MeOH) -52.4" (c 0 057, MeOH) -312.0" (c 0.13, MeOH)

-159.8" (c 0.25, MeOH)

+147.4" (c 0.25, MeOH)

-30.3" (c 0.10, MeOH)

-18.7" (c 0. I 1, MeOH)

-84.8" (c 0.12, MeOH)

Reference

Part

Compound

Examined

roots

leachianol F (36)

roots

leachianol G (37)

roots

pallidol (lob)

roots

davidiol B (38)

roots

(-)-E-viniferin (1 1 b)

roots

stenophylla A (39b)

roots roots

1Gnerum hainanense C.Y. Cheng

Gnetum ieyboldi Tul.

wl

Sources of Resveratrol Oligomers (Continued)

Table 1.

lianas

stenophylla B (40b) stenophylla C (41b)

(-)-z-viniferin ( l l b )

N 0

Oligomer Type

Optical Rotation

(Moleeular Formula)

(Concentration, Solvent)

dimer (C28H2407) dimer (C28H2407) dimer (C28H2206) trimer (c42h4O 13) dimer (C28H2206) tetramer (C&tr1012) trimer (C42H3209) tetramcr (C56H42012)

NA NA

-36.3" (c 0.13, MeOH) NA NA

-345" (c 0.7, MeOH) (c

-2" 0.1, MeOH)

(c

-66" 0. I , MeOH) -38.6"

(c 0.12, MeOH)

lianas

resveratrol trans dehydrodimer (42)

NA

bark

gnetin A (43)

NA

rable 1.

Sources of Resveratrol Oligomers (Continued) Source

Part Examined

Compound

bark

gnetin B (44)

Oligomer Type (Molecular Formula)

Optical Rotation (Cnncentration, Solvent)

dimer

NA

(C28H2406)

bark

gnetin C (45)

dinier

NA

(C28H2206)

bark

gnetin E (46)

trimer

NA

(C42H320d

hefumparvfolium (Warb.) Cheng

lianas

hetum venosum

seeds

c-viniferin (11)

dimer

NA

(C2RH2206)

gnetin C (45)

Spruce ex Benth.

dimer

NA

(C28H2206)

seeds

gnetin E (46)

trimer

NA

(C42H3209)

Vehvitschia mirabilis

wood

gnetin F (47)

J.D. Hook

dimer

NA

(C28H2406)

wood

gnetin G (48)

dimer

NA

(C28H2207)

wood

gnetin H (49)

trimer

NA

(c4203209)

wood

IAEMODORACEAE nigoranfhospreissii Endl.

gnetin I (SO)

trimer

NA

(c4203209)

root cell culture

anigopreissin A (51)

dimer (C28H2006)

NA

Sources of Resveratrol Oligomers (Continued)

Table 1.

VI

N

N

I

Source

Musa cavendishii Lambert ex Paxt. PAEONIACEAE Paeonra suffruticosa Andrews

VITACEAE Ampelopsis brevipedunculala Maxim. ex Trautv.

Part

Compound

Optical Rotation (Concentration, Solvent)

Examined

rhizome

anigopreissin A (51)

NA

seeds

suffruticosol A (52)

NA

seeds

suffruticosol B (53)

NA

seeds

suffruticosol C (54)

NA

roots

ampelopsin A (SSa)

+167" (c 2.12, MeOH)

roots

ampelopsin B (56a)

+ 123" (c 0.93, MeOH)

roots

ampelopsin C (S7a)

+24" (c 1.04, MeOH)

roots

ampelopsin D (58)

roots

/ram -ampelopsin E (59)

NA 0" (c 2.06, MeOH)

roots

cis -ampelopsin E (60)

0"

(c 1.OO,MeOH)

Reference

Sources of Resveratrol Oligomers (Continued)

Table 1. Source

Part

Compound

Examined

roots roots

ampelopsin F (61a) ampelopsin G (62a)

Oligomer Type

Optical Rotation

(Molecular Formula)

(Concentration, Solvent)

dimer

+14.0"

(C28H2206)

( c I .98, MeOH)

trimer (C42H3209)

roots

ampelopsin H (63a)

tetramer (CS6H320

roots

miyabenol C (6)

12)

trimer

+32.0" ( c 0.23, MeOti)

+ I050 (c 0.83, MeOH) NA

(C42F13209)

roots

pallidol (10)

dimer

NA

(C28H2206)

'issuspallida Planch. 'issus quadrangularis L

stemwood stems

pallidol (10) pallidol (10)

dimer

0"

(C2nH2206)

(MeOH)

dimer

NA

(C28H2206)

stems

parthenocissin A (64)

dimer

NA

(C28H2206)

stems yphosfemmacrotalarioides (Planch.) Descoings

roots

quadrangularin A (65b) cyphostemmine A (66)

dimer

-2"

(C2nH2206)

(MeOH)

dimer

NA

(C2aH2206)

roots

cyphostemmine B (67)

dimer

NA

(C28H2206)

roots

trans- z-viniferin (11)

dimer (C28H2206)

NA

Sources of Resveratrol Oligomers (Continued)

Table 1.

~

Source

Part

Compound

Examined

roots

cis -&-viniferin (68)

Oligomer Type

O p z a l Rotation

(Molecular Formula)

(Concentration, Solvent)

dimer

NA

(C28H2206)

roots

gnetin C (45)

dimer

NA

(C28H2206)

roots

gnetin E (46)

trimer

NA

(c42133209)

roots

pallidol (10)

dimer

NA

(c28f42206)

roots

parthenocissin A (64)

dimcr

NA

(C2RH2206)

'arthenocissusquinquefolia Planch.

stems stems

parthenocissin A (64b) parthenocissin B (69b)

dimer

-25"

(C28H2206)

(c 0.13, MeOH)

trimer (C42H3209)

arthenocissus tricuspidata (Sieb. & Zucc.) Planch.

stemwood stemwood

isoampelopsin F (70b) (+)-&-viniferin ( I l a )

(c

-39" 0.13, MeOH)

dimer

-56.7"

(C28H2204

(c 0.23, MeOH)

dimer

NA

(C2RH2206)

sternwood

(-)-pallidol (lob)

dimer

NA

(C28H2206)

stemwood

tricuspidatol A (71)

dimer

NA

(C28132407)

itrs

betulrfolia Diels. & Gilg.

stems

arnpelopsin A (55)

dimer (C2RH2Z07)

NA

Sources of Resveratrol Oligomers (Continued)

Table 1. ~

Source

Part Examined

Compound

Oligomer Type (Molecular Formula)

Optical Rotation (Concentration, Solvent)

stems

ampelopsin C (57)

NA

stems

betulifol A (72a)

trimer (C42H3209) dimer (C28H2006) dimer (C28H240x) dimer (cZ8i tetrarner (C56H42012) tetramer (C56H42012) tetramer (C56H4201 2) dimer (C282207) trimer (C42H3209) dimer (C28H2206) dimer (C28H22Od dimer (C28H2206)

stems

betulifol B (73a)

stems

(+)-s-viniferin ( I l a )

stems

heyneanol A (74)

stems

hopeaphenol (14)

stems V i m corgneliae Pulliat ex Planch.

vitisin A ( 7 ~ ) ~

NA

(+)-ampelopsin A (55a)

NA

ampelopsin C (57a)

NA

ampelopsin F (61a)

leaves

(+)-s-viniferin (Ila)

NA

(+)-E-viniferin (1l a )

+28.9" (c 0.28, MeOH)

+8 1.6" (c 0.19, MeOH) NA NA NA NA

+167" (c 2.0, MeOH) +24" (c 1.0, MeOH) +I40 (c 2.0, MeOH) NA

+38" (c I .83, MeOH)

Sources of Resveratrol Oligomers (Continued)

Table 1. Part

Compound

Examined

stems roots roots stems stems bark bark

E-viniferin diol (78a) (+)-trans-vitisin A (76a)

(+)-cis -vitisin A (77a) (-)-trans-vitisin B (79b) (-)-cis-vitisin B (80b)

(+)+itisin D (82a) (+)-vitisin E (83a)

stems

(+)-E-viniferin ( l l a )

stems

flexuosol A (84b)

stems

gnetin A (43)

stems

hopeaphenol(l4)

stems

vitisin A ( 7 ~ ) ~

Oligomer Type

Optical Rotation

(Molecular Formula)

(Concentration, Solvent)

dimer (C28H2408) tetramer (CS6H42OI2) tetramer (C56H42012) t et r am e r (CS6H42012) tetramer (C56H420I 2) tetramer (CS6~~42012) trimer (C42H3209) dimer (C28H2206) tetramer (C56H42013) dimer (C28H2206) tetramer (CS&WI~) tetramer (CS6H420 12)

+136.0" (c O19,MeOH)

+195.1' (c 1.1, MeOH)

+ 184.0" (c 0.5, MeOH)

-90.0" (c 2.28, MeOH)

-4 I .9" (c 0.72, MeOH)

+222.0" (c 0.20, MeOH)

+94.5" (c 0.20, MeOH) NA

-99.6" (c 0.15, MeOH)

NA NA

NA

Table 1.

Sources of Resveratrol Oligomers (Continued) -Source

'rtrs heyneana Roem. & Schult.

Part Examined

Compound

stems

ampelopsin A (55)

Oligomer Type (Molecular Formula)

Optical Rotation (Concentration, Solvent)

dimer

NA

(C28H22O7)

stems

ampelopsin C (57)

trimer

NA

(C42H320Y)

stems stems ills

sp

roots

(+)-&-viniferin ( I l a ) heyneanol A (74b) E-viniferin (11)

dimer (C~Hzz0d tetramer

(c 0.40,

(C56I142O12)

(c 0.31, MeOH)

dimer

NA

+39" MeOH) -53"

(C28H2206)

roots

r-viniferin (8Sb)

tetramer (C56H42012)

roots

r-2-viniferin (86a)

tetramer (C56H42012)

itis vinifera L

leaves

a-viniferin (8)

trimer

-86.4" (c 1.1, MeOH)

+132.9" (c I . I , MeOH)

NA

(C42H3009)

UV-irradiated leaves leaves

(-)-a-viniferin (8b)

trimer

NA

(C42H3U09)

&-viniferin (11)

dimer

NA

(C2gH2Z06)

stalks

c-viniferin (11)

dimer

NA

(C28H2206)

leaves

resveratrol trans dehydrodimer (42)

dimer (C28H2206)

NA

Table 1.

Sources of Resveratrol Oligomers (Continued) Part Examined

Compound

roots

ampelopsin A (55)

Oligomer Type (Molecular Formula) dimer

Optical Rotation (Concentration, Solvent) NA

(C28H2207)

roots

hopeaphenol(l4)

tetramer

NA

(C56H42012)

roots wood

(+)-ampelopsin A (55a) (+)-E-viniferin (1 l a )

dimer

+167.0"

(C28H2207)

(c 2. I, MeOH)

dimer

+49.1" (c 1.9, MeOH)

(c28t12206)

wood

(+)-hopeaphenol(14a)

tetramer (C56H420

wood

(-)-isohopeaphenol (15b)

12)

tetramer (C56H42012)

wood

(-)-viniferal (87b)

from a trimer (C3SH2608)

wood

viniferifuran (88)

dimer

+374.8" (c 0.78, MeOH)

- 1 14.5" (c 0.44, MeOH)

- 1 3 1.7" (c 1.6, MeOH)

NA

(C28H2U06)

wood wood

(+)-vitisifuran A (89a) (-)-vitisifuran B (90b)

tetramer

+236.1"

( C 5 6 H 4 0 0 12)

(c 0.44, MeOH)

tetramer (C56H40012)

wood

(-)-from-vitisin B (79b)

tetramer (CS6H42012)

wood

(+)-vitisin C (81a)

-133.7" (c 0.12, MeOH)

-90" (c 2.3, MeOH)

tetramer

i239.9"

(C56H42012)

(c 0.5, MeOH)

Sources of Resveratrol Oligomers (Continued) I

Source

Part

Compound

Examined

MICROBIAL CULTURES AND ENZYMES Botrylis cinerea ATCC 1 1542

Bottytyliscinerea isolate Horseradish peroxidase

growing incubations growing incubations growing incubations growing incubations growing incubations growing incubations culture medium filtrate NA NA NA

a

not applicable or data not available

leachianol F (36) pallidol (10) restrytisol A (91b) restrytisol B (92b) restrytisol C (93b) resveratrol trans dehydrodimer (42b) resveratrol trans dehydrodimer (42) resveratrol trans dehydrodimer (42) (-)-trans-vitisin B (79b) (+)-vitisin C (81a)

Oligomer Type (Molecular Formula)

Optical Rotation

dimer (C28H2407) dimer (C28H2206) dimer (C2UH2407) dimer (CZRH2407) dimer (C2RHZZ06) dimer (C2UH2206) dimer (CZSH2206) dimer (C28H2206) tetramer (Cj6H42012) tetramer (C56H42012)

NA

(Concentration, Solvent)

NA -2 20" (c I .8, acetonc)

-0.42" (c

4.7, acetone) -0.92"

(c 1. I , acetone)

-1.15" (c 7.3, acetone)

NA NA -90" (c 2.3, MeOH)

t239.9" (c 0.5. MeOH)

Reference

530

which the (+) enantiomer has been assigned a number followed by the letter 'a', while the (-) enantiomer has been assigned the letter 'b'. Furthermore, for cases in which the stereochemistry of the reported oligomer was not defined, no letter designation was given. For example, Ito and Niwa [103] reported isolating (H-)-6'-viniferin from Vitis vinifera, Kurihara and colleagues [44] obtained (-)-6'-viniferin from Carex kobomugi, and Lin and colleagues [76] acquired an unspecified form of ^•-viniferin from Gnetum parvifoUum. Under our classification scheme, sviniferin has been assigned the number 11, (+)-6'-viniferin the number 11a, and (-)-£•-viniferin the number lib. The structures of these and all the other resveratrol oligomers can be found in Appendix (1). It should be noted that 1 and its glucoside conjugates, piceid (94) and resveratroloside (95), have been isolated from a large number of plant families; however, only a portion of these have been examined as a source of oligomeric stilbenes. Table (2) provides a brief listing of all the plant families known to produce 1, glucoside conjugates of 1, and resveratrol oligomers. This information was compiled based on the distribution of these compounds as reported by Gorham [3] and those references cited in Table (1). Future examination of many of the plant families presented in Table (2) will likely reveal many new sources of resveratrol oligomers. Biosynthesis of Resveratrol Oligomers Biosynthetic processes leading to the formation of the resveratrol oligomers have been hypothesized by several authors [3,4,44,51,52,57,62,84,89,103]; however, conclusive evidence supporting these conjectures is still lacking. In general, the plant derived resveratrol oligomers are believed to be the products of a successive series of oxidative couplings of resveratrol radicals (Fig. (2)). Recent efforts to identify the enzymes responsible for the biotransformation of 1 in Vitis spp. have uncovered at least two stilbene-metabolizing peroxidases as likely catalysts. These peroxidases have been shown to carry out reactions similar to that of horseradish peroxidase that has been documented for its role in the formation of structurally analogous neolignans [78,104, 109]. It has been shown that grapes contain several peroxidases that vary both temporally and spatially with regards to their in vivo expression in grape tissues indicating that these enzymes may

Table 2.

Distribution of Resveratrol and Its Derivatives by Plant Family Resveratroloside (94)

532

OH

Fig. (2). Depiction of a proposed biosynthetic scheme for the dimerization of resveratrol (1) to (+)-£--viniferin (11a)

have highly specific functional roles such as the polymerization of 1 [110]. Morales and colleagues [111] have identified a possible candidate peroxidase isozyme that is expressed constitutively in grape vacuoles. This isozyme has demonstrated a high affinity for 1 in an acidic medium in which it readily oxidizes 1. It is likely that additional peroxidases also play a role in the formation of defensive resveratrol oligomers such as those produced by plants during periods of microbial infection [112,113]. Based on the proposed free-radical mechanism for the formation of the resveratrol oligomers (Fig. (2)), it is likely that accessory proteins are involved in the biosynthesis of these compounds. It has come to light in recent years that auxiliary or dirigent proteins play an important role in the free-radical coupling of monolignols in lignin biosynthesis [114,115]. These proteins are believed to capture the oxidized free-radical substrate providing a scaffold upon which radical coupling can occur yielding an optically active product. Due to the optical activity generally exhibited by the resveratrol oligomers (Table (1)), it is a distinct possibility that a form of dirigent protein may be involved in the biosynthesis of these compounds. It is readily apparent that a great deal of research is needed to elucidate the biosynthetic pathways of the resveratrol oligomers. It is interesting to note that in vitro incubations of 1 with horseradish peroxidase [108] as well as with growing B. cinerea cultures [106] and B. cinerea culture medium filtrates [107] have yielded a unique resveratrol oligomer known as the resveratrol rraw^'-dehydrodimer (42). This dimer shares a similar structural design with the more common plant-derived resveratrol dimer, 6"-viniferin (11). It has been proposed that the B. cinerea mediated transformation of 1 to 42 may be facilitated by other stilbene oxidizing enzymes such as copper-containing laccases [106,107].

533 In another study, Pezet [116] provided evidence that a putative laccase isozyme from a B. cinerea isolate was capable of transforming resveratrol into 11 under in vitro conditions. ISOLATION OF RESVERATROL OLIGOMERS Resveratrol oligomers have been obtained from a wide variety of plants (Table (1)). These compounds have been isolated primarily from the roots, wood, and bark; however, they have occasionally been extracted from the seeds, fruits, and leaves of some species. In addition, a resveratrol oligomer has been obtained from a plant cell culture. Other resveratrol oligomers have also been obtained following the incubation of 1 with fungal cultures and isolated enzyme systems. Extraction Only a few solvents have been used to successfully extract resveratrol oligomers from both fresh and dried plant materials. Most isolation schemes utilize a single solvent extraction with methanol [43,86,92,99,103], ethanol [77,84,89,97], or acetone [51,62,65,93,96,105] at room temperature. Many resveratrol oligomers contain an ethylene bridge in a trans configuration; however, this double bond has been shown to be highly susceptible to ultraviolet induced isomerization [93,107]. In an effort to avoid isomerization of the double bond, several research groups have reported incorporating measures to shield the plant materials from light during the extraction process [54,59,72,98,99]. This has helped in some cases to prevent the formation of contaminating cis isomer artifacts. Partitioning Following extraction, the crude plant extracts are often subjected to one or more liquid-liquid phase partitioning steps. Ethyl acetate and water are most frequently utilized with the resveratrol oligomers readily accumulating in the organic phase [45,71,84,86,92,96,106]. In other cases, more elaborate partitioning schemes have been used such as

534

defatting with hexanes and/or chloroform to help remove unwanted highly non-polar substances before proceeding with ethyl acetate [39,103]. Still others have reported success by directly partitioning their crude extracts between a non-polar solvent such as hexanes and aqueous methanol [80]. Isolation A host of methods has been used to facilitate the isolation of single chemical entities from the complex polyphenolic mixtures obtained following their initial extraction and partitioning. The most common procedure is a simple two-step process beginning with silica gel column chromatography to yield a crude resveratrol oligomer mixture. This is followed by preparative thin-layer chromatography (PTLC) to separate the individual oligomers from each other. Often times the same or a similar solvent system is employed for both steps. The most common solvent system that has been employed successfully to isolate resveratrol oligomers is a chloroform-methanol solution [46,62,87,89, 96,103,104,106]. Other useful solvent systems for siHca gel chromatography include hexane-acetone [72], chloroform-benzenemethanol [41,43,44,45], and toluene-ethyl acetate-acetic acid [106]. Occasionally, some resveratrol oligomers have proven difficult to isolate using the previously outlined methods. In these cases, reversephase Cg and Ci8 bonded silica gel materials were invaluable for isolation [42,84,92,98,102,106,107]. This stationary phase material has been widely used in both preparative HPLC and PTLC. The most commonly employed solvent systems in these applications are aqueous methanol and aqueous acetonitrile solutions. Other less frequently used stationary phases include both Sephadex LH-20 and a small assortment of resins [39,43,55,65,66,71,98,99,106]. Another less commonly used, but promising technique that has been employed in the isolation of cis and trans isomers of resveratrol oligomers is recycled HPLC. Using an aqueous methanol solvent system, Ito and colleagues [93] were able to separate both the cis and trans isomers of the tetrameric resveratrol oligomer (+)-vitisin A (77a and 76a, respectively).

535

STRUCTURE ELUCIDATION OF RESVERATROL OLIGOMERS Structure elucidation studies of resveratrol oligomers have benefited greatly from the development of NMR spectroscopy. However, other spectroscopic methods such UV, IR, and MS still remain important tools in this research. UV Spectroscopy The extended conjugated stilbene moiety of 1 has been examined using UV spectroscopy. This chromophore typically presents a UV absorption spectrum with a X^max of approximately 310-320 nm. However, it has been noted that for resveratrol oligomers in which the ethylene bridge has been reduced, such as that seen in (-)-a-viniferin (8b) [100], kobophenol A (9) [43], and tricuspidatol A (71) [89], the A^max is shifted to a lower wavelength of approximately 280-290 nm. Most resveratrol oligomers present relatively similar UV absorption spectra; however, Mattivie and Reniero [117] have examined the spectra of an assortment of resveratrol oligomers and noted several key UV spectral features that they in turn have applied as a tool for predicting some basic structural features of these compounds. Using these methods, Mattivie and Reniero have reported that it is possible to determine the presence and stereochemistry of an ethylene bridge and predict the number of resveratrol monomers incorporated in an oligomer. IR Spectroscopy Infrared absorption spectroscopy has been used to aid in determining the structures of resveratrol oligomers. Resveratrol oligomers present several characteristic peaks such as a broad 0-H stretch and C=C stretching and bending vibrations. However, some of the most useful information obtained from the IR spectra of resveratrol oligomers result from some of the less commonly seen structural features in these compounds. The most informative of these is the presence or absence of a carbonyl stretch as was noted for caraganaphenol A (29a) [65], kobophenol B (16b) [45], gnetin A (43) [74], and leachianol C (33b) [70].

536 Mass Spectrometry Mass spectrometry has proven to be an important tool for quickly and reliably identifying the number of monomeric units of 1 that are incorporated into an oligomer. A variety of ionization techniques have been used on resveratrol oligomers with widely varying results. Standard electron impact techniques have occasionally resulted in failure to observe a molecular ion [100]; however, the use of other 'soft' ionization methods has helped circumvent this problem. Some of the more commonly employed techniques that have repeatedly been used with a high degree of success include electrospray [106], field desorption [47,65], and fast atom bombardment [39,42,53,67,68,92,99,103] ionization. In addition the utility of mass spectrometry for identifying the molecular weight of resveratrol oligomers, this technique has been used in a few limited applications for direct structural analysis. This type of application of mass spectrometry can be seen in a portion of the work regarding the structure elucidation of selected resveratrol oligomers [45,52,59,63]. More recently, Senda and colleagues [40] have used a combination of electron impact and secondary ionization mass spectrometry to investigate the fragmentation of three resveratrol oligomers. It is interesting to note that in some cases a greatly enhanced molecular ion has been observed using negative ion analysis [70,86,98,99,106]. *H and ^^C NMR Spectroscopy The development of modern NMR techniques has had a significant impact on structure elucidation studies of resveratrol oligomers. Highresolution ^H NMR spectra of resveratrol oligomers present several key features that aid in the structure elucidation of these compounds. For example, the presence of para-substituted phenolic moieties are readily discernible as a pair of AA'XX' doublets, each integrating to two protons, between 6.4-7.6 ppm with an approximate coupling constant of J=8 Hz. The 1,3,5-trisubstituted resorcinol moiety also presents some readily recognizable features including a doublet integrating to two protons at

537

approximately 6.2 ppm (J=2 Hz) and a triplet integrating to one proton around 6.0 ppm (J=2 Hz). Most resveratrol oligomers also possess two or more aliphatic proton spins between 3.0-5.5 ppm depending on the degree of unsaturation. A great deal of important information can also be gleaned from the ^"^C NMR spectra of the resveratrol oligomers. These compounds contain several characteristic spins that aid in the rapid determination of the presence or absence of certain key structural features of the resveratrol oligomers. For example, the presence of a signal between 95-103 ppm is a key indicator of a dihydrobenzofuran system, which is a common structural moiety observed in approximately 80% of the resveratrol oligomers. Most other spins appear at a lower field between 105-160 ppm. The high-field portion of the NMR spectra of resveratrol oligomers (approximately 35-95 ppm) contain, on average, about 10% of the ^^C spins, but these are exceptionally useful for rapidly identifying (or excluding) many of the structural moieties incorporated in the resveratrol oligomers. The ^H and ^"^C NMR data for the majority of the resveratrol dimers are presented in Tables (3) and (4), respectively. The data presented in these tables represent the typical range of chemical shift values and coupling constants that are encountered with all known resveratrol oligomers. Since most resveratrol oligomers can be considered as a dimer to which additional resveratrol monomers are fused (i.e. the trimer, leachianol D (34b), is the dimer pallidol fused with a third resveratrol moiety), these tables can be used to quickly help identify which structural moieties may be present in a new resveratrol oligomer. Other ID and 2D NMR Techniques Several ID and 2D NMR techniques have become standard tools for structure elucidation studies of the resveratrol oligomers. The most commonly employed of these include the one- bond ' ^ C - ' H H E T C O R and HMQC, as well as, the two- and three-bond '^C-*H FLOCK [106], COLOC [67,72,88,106], and HMBC [53,71,78] pulse sequences. Typical ^"Vc-H coupling constant values used for these experiments range between J=8-12 Hz. Other commonly used ' H - ' H correlation methods include standard COSY [64] and NOESY [67,72,78] pulse sequences; however, the most widely employed experiment concerning proton-proton

Table 3.

H N M R Data of Selected Resveratrol Dimers (6, multiplicity, J in Hz)

1

Reference

ampelopsin D (58)

ampelopsin F (61a)

(acetone-d,)

(acetone-d,)

1821

1831

anigopreissin A (51) (acetone-d,)

1791

Position 2

6.90 d (8.3)

6.99 d (8.5)

6.97 d (7.5)

7.46 d (8.8)

3

6.65 d (8.3)

6.62 d (8.5)

6 63 d (7 5)

6.80 d (8.8)

5

6.65 d (8.3)

6.62 d (8.5)

6.63 d (7.5)

6.80 d (8.8)

6

6.90 d (8.3)

6.99 d (8.5)

6.97 d (7.5)

7.46 d (8.8)

7 8

5.45 d (5.0)

4.15 brs 4.02 brs

4.07 brs

5.42 brs

3.23 brs

5.97 d (2.0)

10

6.49 d (2.2)

12

6.16 d (2.3)

5.99 t (2.0)

5.94 d (2.5)

6.41 t (2.2)

14

6.62 d (2.3)

5.97 d (2.0)

6.39 d (2.5)

6.49 d (2.2)

2‘

7.12 d (8.3)

7.05 d (8.5)

6.66 d (7.5)

7.45 d (8.5)

3’

6.78 d (8.3)

6.52 d (8.5)

6.44 d (7.5)

6.85 d (8.5)

5,

6.78 d (8.3)

6.52 d (8.5)

6.44 d (7.5)

6.85 d (8.5)

6’

7.12 d (8.3)

7.05 d (8.5)

6.66 d (7.5)

7.45 d (8.5)

7’

5.77 d ( I 1.7)

6.91 brs

3.52 brs

7.13 d (16.3)

8’

4.17d(11 7)

4.01 brs

7.06 d (16.3)

10’

7.25 d (0.8)

12’

6.43 d (2.3)

6.17 d (2.0)

6.03 d (2.5)

14’

6.24 d (2.3)

6.68 d (2.0)

6.32 d (2.5)

6.87 d (0.8)

1

Table 3. Compound (Solvent)

H N M R Data of Selected Resveratrol Dimers (6, multiplicity, J in Hz) (Continued)

balanocarpol (17b)’ (chloroform-d ,)

Reference

betulifol A (72a)

betulifol B (73a)

(pyridine-d,)

(acetone-d6)

cyphostemmin A (66) (acetone-d6)

1901

1901

i861

Position 2

7.59 d (8.7)

7.73 d (8.3)

7. I5 d (8.4)

7.29 d (9)

3

7.02 d (8.7)

7.25 d (8.3)

6.83 d (8.4)

6.91 d (9)

5

7.02 d (8.7)

7.25 d (8.3)

6.83 d (8.4)

6.91 d (9)

6

7.59 d (8.7)

7.73 d (8.3)

7.15 d (8.4)

7.29 d (9)

7

5.79 d (9.9)

5.79 d (10.3)

5.35 d (5.1)

8

5.28 d (9.9)

4.81 d(10.3)

4.40 d (5.1)

10

6.22 d (2.4)

6.76 brs

6.04 d (2.1)

6.32 d (2 5)

12

6.28 d (2.4)

6.83 brs

6.04 d (2.1)

6.28 d (2.5)

a 4.13 bd (15) b 4.02 bd (15)

6.04 d (2.1)

14 2’

6.77 d (8.7)

7.73 d (8.3)

6.93 d (8.4)

7.08 d (9)

3’

6.57 d (8.7)

7.25 d (8.3)

6.65 d (8.4)

6.79 d (9)

5’

6.57 d (8.7)

7.25 d (8.3)

6.65 d (8.4)

6.79 d (9)

6’

6.77 d (8.7)

7.73 d (8.3)

6.93 d (8.4)

7.08 d (9)

7‘

4.88 d (1.8)

5.79 d (10.3)

4.93 d (8.7)

8’

5.40 d (1.8)

4.81 d (10.3)

4.58 d (8.7)

10’

6.03 d (2.2)

6.76 brs

6.67 d (2.0)

12’

6.21 d (2.2)

6.83 brs

6.37 d (2.0)

14’ a

pentarnethyl ether derivative

6.49 d (2.5) 6.3 I t (2.5) 6.49 d (2.5)

Table 3.

1

H NMR Data of Selected Resveratrol Dimers (6, multiplicity, J in Hz) (Continued)

(Solvent)

cyphostemmin B (67) (acetone-d6)

Reference

1861

Compound

(+)-E-viniferin ( I l a ) (methanol-d4)

11051

E-viniferin diol (78a) (acetone-d,)

gnetin A (43) (acetone-d6)

1921

1751

Position 2

7.06 d (9)

7.13d(8.8)

6.77 d (8.6)

7.56 d (8.5)

3

6.80 d (9)

6.76 d (8.8)

6.62 d (8.6)

6.88 d (8.5) 6.88 d (8.5)

5

6.80 d (9)

6.76 d (8.8)

6.62 d (8.6)

6

7.06 d (9)

7.13 d (8.8)

6.77 d (8.6)

7.56 d (8.5)

7

4.28 brs

5.36 d (6.6)

4.35 d (4.0)

7.28 d (16.0)

4.51 d (4.0)

7. I5 d (16.0)

8

4.41 brs

4.34 d (6.6)

10

6.45 d (2.5)

6.15 d (2.2)

12

6.33 t (2.5)

6. I7 t (2.2)

6.34 d (2.0)

3.68 ddd (7.0, 1.5, 1.0)

14

6.45 d (2.5)

6.15 d (2.2)

6.68 d (2.0)

3.87 dd (1.5, 1.0)

2‘

7.35 d (9)

7.03 d (8.8)

7.17 d (8.5)

7.09 d (8.5)

3,

6.81 d (9)

6.64 d (8.8)

6.83 d (8.5)

6.74 d (8.5)

6.26 t (1.0)

5’

6.81 d (9)

6.64 d (8.8)

6.83 d (8.5)

6.74 d (8.5)

6‘

7.35 d (9)

7.03 d (8.8)

7.17d(8.5)

7.09 d (8.5)

7,

7.19 brs

6.56 d (16. I )

5.37 d (4.0)

3.77 t (7.0)

6.81 d (16.1)

4.40 d (4.0)

3.43 d (7.0)

6.18 d (2.5)

6.19 d (2.0)

8’ 10’

6.92 d (2.5)

6.62 d (2.2)

12’

6.41 d (2.5)

6.24 d (2.2)

14’

6.32 t (2.5)

6.26 t (2.0)

6. I8 d (2.5)

6. I9 d (2.0)

rable 3.

1

H NMR Data of Selected Resveratrol Dimers (6, multiplicity, J in Hz) (Continued)

Compound

gnetin B (44)

gnetin C (45)

gnetin F (47)

gnetin G (48)

(Solvent)

(acetone-d,)

(acetone-d6)

(acetone-d,)

(pyridine-d,)

Reference

1751

1751

1781

1781 7.48 d (8.5)

Position 2

7.38 d (8.5)

7.45 d (8.5)

6.82 d (8.5)

3

6.82 d (8.5)

6.85 d (8.5)

6.66 d (8.5)

7.00 d (8.5)

5

6.82 d (8 5)

6.85 d (8.5)

6.66 d (8.5)

7.00 d (8.5)

6

7.38 d (8 5)

7.45 d (8.5)

6.82 d (8.5)

7.48 d (8.5)

7

6.91 d (16.5)

7.11 d(16.5)

2.4-2.6

6.41 d (5.0)

8

6.76 d (16.5)

6.97 d (16.5)

2.4-2.6

6.16 d (5.0)

10

6.19s

6.60 brd (1.0) 6.26 d (2.0)

6.78 d (2 0)

12

3.15 d(14.0) 3.83 d (14.0)

14

3.02 s

6.70 brd (1 .O)

6.30 d (2.0)

7 45 d (2.0)

2.

6.92 d (8.5)

7.22 d (8.5)

7.14 d (8.5)

7.35 d (8.5)

3‘

6.65 d (8.5)

6.84 d (8.5)

6.89 d (8.5)

6.97 d (8.5)

5’

6.65 d (8.5)

6.84 d (8.5)

6.89 d (8.5)

6.97 d (8.5)

6’

6.92 d (8.5)

7.22 d (8.5)

7.14 d (8.5)

7.35 d (8.5)

7‘

3.15 d (5.5)

5.39 d (4.5)

5.34 d (6.0)

6.02d(11.5)

8’

3.02 d (5.5)

4.39 d (4.5)

4.19 d (6.0)

4.95 d (1 1.5)

6.20 d (2.0)

10’

6.14 d (2.0)

6.17 d (2.0)

12’

6.12 t (2.0)

6.24 I (2.0)

6.28 I (2.0)

7.09 d (2.5)

14’

6.14 d (2.0)

6.17 d(2.0)

6.20 d (2.0)

6.70 d (2.5)

Table 3. Compound

I

H N M R Data of Selected Resveratrol Dimers (6, multiplicity, J in Hz) (Continued)

leachinol G (37) (acetone-d,)

malibatol A (2Ib) (acetone-d,)

parthenocissin A (64b)

(Solvent)

(acetone-d6)

restrytisol A (91b) (acetone-d6)

Reference

1711

1531

1871

11061

2

7.07 d (8.0)

7.45 dd (8.5,2.5)

6.99 d (8.8)

7.17 d (8.4)

3

6.72 d (8.0)

6.80 d (8.8)

6.64 d (8.4)

5

6.72 d (8.0)

6.80 dd (8.5, 2.5) 6.80 dd (8.5, 2.5)

6.80 d (8.8)

6.64 d (8.4)

6

7.07 d (8.0)

7.45 dd (8.5,2.5)

6.99 d (8.8)

7.17 d (8.4)

7

4.48 dd (8.0, 4.0)

4.26 d (1.8)

5.82 d (4.4)

Position

8

3.40 dd (8.0,4.0)

3.45 brs

3.84 dd (4.4, 6.2)

10

5.75 d (2.0)

6.51 d (2.5)

6.19 bn

6.1 I d (2.3)

12

6.22 d (2.0)

6.30 d (2.5)

6.19 brs

5.95 t (2.2)

6.19 brs

6. I I d (2.3)

6.87 d (8.0)

7.02 dd (9.0, 2.5)

7.20 d (8.3)

7.28 d (8.4)

14 2’ 3’

6.7 I d (8.0)

6.33 dd (9.0, 2.5)

6.72 d (8.3)

6.77 d (8.8)

5’

6.71 d (8.0)

6.33 dd (9.0, 2.5)

6.72 d (8.3)

6.77 d (8.8)

6’

6.87 d (8.0)

7.02 dd (9.0, 2.5)

7.20 d (8.3)

7.28 d (8.4)

7’

4.26 d (4.0)

5.46 dd (2.5, 1.0)

6.31 brs

5.51 d (10.3)

8’

3.48 t (4.0)

5.28 ddd (2.5, 1.0, 1.0)

10’

6.14 d (2.0)

7.01 dd (2.0, 1.0)

12’

6.16 t (2.0)

6.57 dd (2.0, 1.0)

14’

6. I4 d (2.0)

3.95 dd (6.4, 10.4) 6.08 d (2.2) 6.26 d (1.5)

6.05 t (2.2)

6.52d(1.5)

6.08 d (2.2)

Table 3.

I

H NMR Data of Selected Resveratrol Dimers (6, multiplicity, J in Hz) (Continued)

Compound

restrytisol B (92b)

restrytisol C (93b)

tricuspidatol A (71)

viniferifuran (88)

(Solvent)

(acetone-d6)

(acetone-d6)

(acetone-d6)

(methanol-d,)

Reference

11061

11061

1891

11051

Position

2

7.09 d (8.4)

6.34 d (8.0)

7.21 d (8.5)

7.42 d (8.8)

3

6.62 d (8.4)

6.41 d (8.0)

6.87 d (8.5)

6.69 d (8.8)

5

6.62 d (8.4)

6.41 d (8.0)

6.69 d (8.8)

6

7.09 d (8.4)

6.34 d (8.0)

6.87 d (8.5) 7.21 d (8.5)

7

5.49 d (8.8)

7.42 d (8.8)

5.25 dd (5.0, 1.5)

8

3.97 t (9.2)

6.84 brs

3.50 dd (5.0, 1.5)

10

6.06 d (2.2)

6.00 brs

6.25 brs

12

5.96 t (2.0)

6.28 d (2.2)

6.25 brs

6.47 t (2.2)

14

6.06 d (2.2)

6.25 brs

6.40 d (2.2)

2’

7.31 d (8.4)

7.26 d (8. I )

7.21 d (8.5)

6.98 d (8.8)

3’

6.82 d (8.4)

6.84 d (8.0)

6.87 d (8.5)

6.65 d (8.8) 6 65 d (8.8)

6.40 d (2.2)

5’

6.82 d (8.4)

6.84 d (8.0)

6.87 d (8.5)

6’

7.31 d (8.4)

7.26 d (8.1)

7.21 d (8.5)

6.98 d (8.8)

7,

5.00 d (9.5)

4.79 d (4.0)

5.25 dd (5.0, 1.5)

6.94 d (16.3)

8‘

3.41 t (9.5)

4.62 d (4.4)

3.50 dd (5.0, 1.5)

6.85 d (16.3)

10‘

6.26 d (2.2)

6.50 d (2.2)

6 25 brs

6.99 d (2.0)

12’

6.18 t (2.0)

6.28 t (2.2)

6.25 brs

6.80 d (2.0)

14’

6.26 d (2.2)

6.50 d (2.2)

6.25 brs

VI P

W

544

X NMR Data of Selected Resveratrol Dimers (5)

Table 4. Compound (Solvent) Reference Position 1 2 3 4 5 6 7 8 9 10 11 12 13 14

|

r 2 3 4 5 6 7 8 9

loll

12 13

1

1£

J

ampelopsin D (58)

ampelopsin F (61a)

anigopreissin A (51)

{sLCttont-d 6)

(acetone-fiffi)

f821

(acetone-flffi) f83]

137.7 129.2 116.4 157.0 116.4 129.2 59.0 59.9 149.7 106.8 159.6 101.7 159.6 106.8 130.0 131.4 116.7 157.6 116.7 131.4 123.0 143.4 147.9 124.2 156.4 98.8 160.0

138.7 130.3 116.0 156.4 116.0 130.3 47.5 58.5 147.7 128.2 153.4 102.3 158.8 104.6 135.7 129.6 115.9 156.4 115.9 129.6 50.8 50.1 147.9 113.8 158.1 102.3 157.3

123.0 128.6 116.1 158.4 116.1 128.6 150.6 116.3 136.5 109.7 159.5 102.9 159.5 109.7 130.0 128.9 116.4 158.1 116.4 128.9 128.7 126.8 136.4 101.4 156.5 118.6 152.4

104JI

J

106_J

™

J

107_4

545

Table 4.

"C NMR Data of Selected Resveratrol Dimers (6) (Continued)

Compound

betulifol A (72a)

betulifol B (73a)

(+)-e-viniferin (11a)

(Solvent)

(pyridine-£?5)

(acetone-c?6)

(methanol 4)

Reference

[90]

1901

1105] 133.9

Position

1

1

130.5

134.2

2

130.3

128.7

128.2

3

116.7

116.5

116.3

4

159.7

159.1

1584

5

116.7

116.5

116.3

6

130.3

128.7

128.2

7

92.8

93.7

94.8

8 9

48.3 136.8

56.9 148.0

58.3 147.4

10

104.5

106.9

107.5

11

159.7

160.0

160.1

12

97.3

102.0

102.2

13 14

160.8

160.0

160.1

121.8

106.9

107.5

r

130.5

131.2

130.4

2

130.3

127.5

128.8

3

116.7

115.4

116.4

4

159.7

158.0

158.5

5

116.7

115.0

116.4

6

130.3

127.5

128.8

7

92.8

74.5

123.7

8

48.3

75.3

130.4 136.9

9

136.8

134.6

10'

104.5

108.2

104.4

ir 12

159.7 97.3

159.1 96.7

159.8 96.9

13

160.8

mj

161.9 n92

162.9

1£

120J

546

Table 4.

1

"C NMR Data of Selected Resveratrol Dimers (5) (Continued)

Compound

gnetin A (43)

gnetin C (45)

(Solvent)

(acetone-flffi)

(acetone-i/g)

(acetone-fifs)

^75]

[75]

1781

127.8

133.0

132.2

130.1

127.9

130.2

115.8 161.8 115.8

115.5

115.6 155.9

130.1

127.9

115.6 130.2

138.8

126.0

36.7

122.7

128.4

35.7

156.9

140.4

140.7

126.7

98.7

120.0

195.5

162.1

161.7

72.8

114.3

95.4

203.6

154.3

159.1

Reference

|

1

gnetin F (47)

Position

1 2 3 4 5 6 7 8 9 10 11 12 13 14

56.4

107.3

109.2

r

129.3

129.2

133.5

2 3 4 5 6 7 8 9 10'

130.3

127.0

116.4

115.7

127.8 116.0

159.3 116.4

156.9 115.7

157.8 116.0

130.3

127.0

127.8

54.2 51.7

92.8 55.0

93.8 56.8

146.9

145.3 106.2 158.4

147.1

105.5 159.7 102.1

101.4

101.9

159.7

158.4

ir 12 13

1

156.9 115.5

]£

J

105_5

J

1062

107.0 159.7 159.7

J

107^0

1

547

Table 4.

"C NMR Data of Selected Resveratrol Dimers (6) (Continued)

Compound (Solvent) Reference Position 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7 8 9 10' 11 12' 13

1

1£

1

gnetin G (48)

leachianol G (37)

(acetone-^/g) [78]

(acetone-rffi) [71]

129.8 127.8 114.7 157.1 114.7 127.8 69.8 42.5 138.8 116.9 158.6 95.9 157.6 104.2 131.2 128.6 114.7 155.9 114.7 128.6 87.1 48.2 141.7 117.5 154.7 100.3 157.6

135.6 129.4 115.2 157.2 115.2 129.4 77.3 62.5 147.2 105.6 158.5 102.4 154.8 123.1 137.9 129.2 115.6 157.1 115.6 129.2 55.9 59.0 151.3 106.2 159.2 101.0 159.2

124.7 130.9 116.4 159.1 116.4 130.9 151.2 117.3 135.8 109.7 156.7 102.2 157.5 121.3 133.4 130.6 114.7 155.2 114.7 130.6 48.8 74.8 139.7 109.9 155.4 95.9 156.2

1^2

n9A

1094

1

malibatol A (21b) (methanol-£f4) i

153]

1

548

Table 4.

C NMR Data of Selected Resveratrol Dimers (5) (Continued)

Compound (Solvent) Reference Position 1 2 3 4 5 6

|

"7

1

parthenocissin A (64b) (acetone-c^g)

restrytisol A (91b) (acetone-f/g)

187]

[1061

[106]

137.4 128.9 115.8 156.6 115.8 128.9 54.9 64.4 145.6 106.9 159.3 101.5 159.3 106.9 129.9 130.6 115.9 157.2 115.9 130.6 125.2 149.9 142.9 128.0 155.4 104.1 158.5

131.1 127.3 114.4 155.6 114.4 127.3 84.2 57.6 139.8 109.4 157.5 100.6 157.5 109.4 134.5 127.5 115.0 156.8 115.0 127.5 82.3 60.4 139.8 107.7 157.8 100.9 157.8

1317 128.1 114.2 155.9 114 2 128.1 83.6 59.2 142.6 108.0 157.7 100.5 157.7 108.0 131.5 128.1 115.0 157.0 115.0 128.1 86.6 59.8 142.5 106.8 158.5 101.2 158.5

8 9 10 11 12 13 14

r 2 3 4 5' 6 7 8' 9' 10 11' 12 13

1

MJ

J

mA

1

1077

|

restrytisol B (92b) (acetone-^g)

J

1_06^8

1

549 Table 4.

"C NMR Data of Selected Resveratrol Dimers (5) (Continued)

Compound (Solvent) Reference Position 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7 8 9'

loll 12 13

1

1£

1

restrytisol C (93b) (acetone-^/g)

tricuspidatol A (71) (acetone-fi^g)

[106]

189]

130.0 130.4 113.9 155.8 113.9 130.4 138.7 123.6 148.9 105.2 156.3 101.4 150.9 124.7 134.5 129.7 114.9 155.6 114.9 129.7 49.8 54.0 145.3 105.2 158.7 101.2 158.7 105^2

133.7 128.3 115.6 157.3 115.6 128.3 87.9 63.3 145.4 107.4 158.9 102.9 158.9 107.4 133.7 128.3 115.6 157.3 115.6 128.3 87.9 63.6 145.4 107.4 158.9 102.9 158.9 107_4

viniferifuran (88) (methanol-£^4) !

11051 123.8 128.4 116.2 158.4 116.2 1284 150.6 117.4 138.7 110.2 160.6 103.1 160.6 110.2 130.7 128.8 116.3 158.2 116.3 128.8 123.3 129.3 133.3 107.4 156.5 97.3 156.4

1

122^5

550

correlations is the difference NOE pulse sequence [65,72,87,106]. This experimental technique has proven to be extremely useful for assigning the relative stereochemistry of many resveratrol oligomers since coupling constants are notoriously poor predictors of stereochemistry in many of the five-, six-, seven-, and eight-membered ring systems encountered in these compounds [87,106]. Most NMR experiments with resveratrol oligomers are performed in deuterated acetone. This solvent has proven itself ideal for this class of compounds due to the high degree of solubility of the resveratrol oligomers and the lack of overlap between the solvent and analyte peaks. The second most commonly employed solvent is deuterated methanol, with deuterated dimethyl sulfoxide and pyridine used only occasionally. In some cases the NMR spectra of some 0-methyl and 0-acetyl resveratrol oligomer derivatives have been collected in deuterated chloroform [51,57,93,103]. Despite the exceptional analytical power afforded by the combination of these previously outlined techniques, discrepancies still remain regarding the structures and trivial names of some resveratrol oligomers (for example, see [2], [71,78], [82,86,118], [85], and [92,97,98]). Further comparative studies are needed to critically evaluate these incongruities. X-ray Crystallography In most cases, resveratrol oligomers have proven recalcitrant to crystallization preventing the widespread application of X-ray crystallographic techniques. Most resveratrol oligomers reportedly precipitate as amorphous off-white to reddish powders. This is likely the reason why there has only been one study published to date reporting the X-ray crystal structure of an 0-methyl, dibromo-derivative of hopeaphenol (14b) [50]. Circular Dichroism Due to the resistance of most resveratrol oligomers to forming suitable crystals for X-ray analysis, alternative measures have been sought in order to define the absolute stereochemistry of these compounds. As a result, circular dichroism (CD) studies have been applied to a large

551 number of resveratrol oligomers yielding important clues regarding the absolute configuration of many of these compounds. The first reported application of CD analysis to determine the absolute configuration of a resveratrol oligomer was published by Lins and colleagues [78]. In this study, the CD spectra of two neolignans, one possessing a 7-S, 8-5 absolute stereochemistry (96) and the other a 7-7?, 8-i? absolute stereochemistry (97), were compared with the CD spectra of gnetins F-I (47-50). In each case, the gnetins exhibited a negative Cotton effect at approximately 300 nm which was comparable to the negative Cotton effect exhibited by neolignan 96 at 280 nm. Therefore, the authors concluded that gnetins F-I each possessed an absolute stereochemistry of 7'S, S-S as shown for structures 47-50. In the years subsequent to this study, a number of research groups have utilized the absolute configuration assignments proposed for 47-50 in order to determine the absolute stereochemistry of a number of resveratrol oligomers. For example, Kurihara and colleagues [46] observed that a 7',8'-dihydroderivative of (-)-6'-viniferin (lib) obtained from Carex pumila exhibited identical EIMS and NMR spectral data to that obtained for gnetin F (47). However, the CD spectrum of this compound possessed a completely opposite pattern as compared to 47 indicating that the 7',8'-dihydroderivative of l i b was the enantiomer of 47. Based on these data, the authors were able to infer that the l i b found in the Cyperaceae possessed a 7-i?, 8-7? absolute configuration while the enantiomer, (+)-6'-viniferin (11a), obtained from the Vitaceae shared the same absolute configuration (7-iS', 8-iS) of the dihydrobenzofuran moiety as that observed in the Gnetaceae. More recently, Ito and colleagues [93] were able to independently assign the absolute stereochemistry of another resveratrol oligomer using spectroscopic methods. A sample of (-f)-ampelopsin A (55a) from Ampelopsis brevipedunculata var. hancei was converted by chemical means to yield a pentamethyl ketone derivative. Difference NOE spectroscopy was then used to assign the relative stereochemistry and conformation of the molecule in solution. CD spectroscopy of the derivatized (+)-ampelopsin A gave a positive Cotton effect at 357 nm; thereby, the authors were able to apply the 'octant rule' and deduce the absolute stereochemistry of (+)-ampelopsin A as shown for 55a.

552

Chemical Methods of Structural Analysis There have been no reports of the total synthesis of any resveratrol oHgomer. However, numerous chemical derivatization methods have been utilized to help elucidate the structures of these compounds. Two of the most commonly employed methods include simple methylation and acetylation procedures [81,92,103,105]. Methylation is generally performed in acetone with dimethyl sulfate and potassium carbonate under reflux for 12 h at approximately 80 °C. Acetylation is commonly carried out in pyridine with acetic anhydride for 24 h at room temperature. The 0-methyl and 0-acetyl derivatives of the resveratrol oligomers have been utilized to help deduce the number and type of hydroxyl groups present due to the resistance of aliphatic hydroxyl groups to methylation while remaining susceptible to acetylation [81]. In addition, the methylated and acetylated derivatives of the resveratrol oligomers provide a new source of non-exchangeable protons as targets for conformational NOE studies and also serve to protect labile hydroxyl groups when further chemical derivatization methods are employed. Some of the other chemical methods used to help deduce the structures of resveratrol oligomers include ozonolysis [92,93,105], hydrogenation [91], and enzyme-mediated catalysis [39,103]. In addition, ring-closure procedures have been performed on both r-viniferin (85b) [99] and miyabenol A (12) [119] to yield r-2-viniferin (86a) and miyabenol B (13), respectively, as a means of confirming their assigned structures and illustrating their biogenic origins. BIOLOGICAL OLIGOMERS

ACTIVITIES

OF

THE

RESVERATROL

Resveratrol oligomers are known to exhibit a wide-array of biological activities. Current evidence indicates that many resveratrol oligomers play a variety of important roles in plants ranging from inhibition of the spread of microbial infection to prevention of insect predation. It has also come to light in recent years that a number of resveratrol oligomers possess a host of promising pharmacological properties. Some of these resveratrol oligomers may ultimately serve as lead compounds in the development of new drug entities.

553 The Role of Resveratrol Oligomers in Plants In some plant species, the production of stilbene monomers and oligomers has been shown to be associated with a variety of stress factors [2,112,120,121,122,123,124]. The stilbene 1 is one of the most commonly encountered of these stress metabolites. While 1 has been shown to exhibit rather modest antimicrobial properties, some of its oligomeric derivatives have demonstrated more potent phytoalexin activity [2,8,125,126]. The biochemical events that unfold in plants following microbial infection and the subsequent production of defensive stilbenoids have been previously reviewed [2]. These events will be briefly outlined here with further consideration for the potential deleterious effects that the resveratrol oligomers may have on the host plant (Fig. (3)). Upon infection by an invading microorganism, plants respond by up-regulating the enzyme, stilbene synthase, resulting in a rapid and localized accumulation of 1 surrounding the area of infection. The phytoalexin 1 is then metabolized by the plant yielding a variety of stilbene monomers and oligomers. These compounds possess varying degrees of antimicrobial activity thereby preventing the spread of the microbial invader. For example, several resveratrol oligomers such as a-viniferin (8) [8], canaliculatol (24b) [57], ^-viniferin (11) [8], and vaticaffinol (25b) [58] have been shown to inhibit the growth of the fungal plant pathogen Cladospohum cladosporioides. In addition, 8 and 11 have demonstrated antifungal activity against other plant pathogens such as Botrytis cinerea, Plasmopara viticola, and Piricularia oryzae [8]. Some microbial pathogens can circumvent the defensive response of plants by biotransforming the antimicrobial stilbenoids in a multi-step oxidative detoxification process [106]. Research has shown that the pathogenicity of B. cinerea strains is positively correlated with these fungi's production of blue-copper oxidases known as stilbene oxidases or laccases [127,128]. These enzymes are polyphenol oxidases capable of catalyzing the oxidation and polymerization of numerous phenolic substrates [129,130,131,132]. It has been shown that 1 is readily transformed in the presence of B. cinerea culture medium filtrates that contain laccases [107]. Recently, six resveratrol dimers (restrytisols A-C

stilbene synthase transcription and translation

'\' @

resveratrol synthesis #

cellular fragments

invading fungal pathogen

inhibition

photosynthesis

resveratrol oligomers translocation to non-photosynthetic tissues Fig. (3). Biosynthesis of resveratrol oligomers in plants

555 (91b-93b), resveratrol trans-dehydrodimer (42b), leachianol F (36), and pallidol (10)) were reported as the products of the oxidative detoxification of resveratrol by growing B. cinerea cuhures [106]. Additional evidence suggests that these metabolites of 1 may be sequestered in the conidia vacuoles and subjected to further metabolic transformations by B. cinerea [133]. Further studies have indicated that 1 and its oligomers may be capable of inhibiting insect predation by serving as ecdysteroid antagonists [80]. In vitro studies have shown that cz^-resveratrol (98) and suffruticosols AC (52-54) may interact with insect ecdysteroid receptors in an antagonistic fashion, thereby preventing proper growth and development of the insect to occur. It is also important to consider any possible deleterious effects that resveratrol and its metabolites may have on plants. Production of defensive stilbenes in plants results in resource competition between parallel biosynthetic pathways. For example, one group of metabolites that may be severely impacted by this competition is the chalcones and their metabolites that play a variety of important roles in plants. These compounds share a pool of basic precursors with the stilbenoids [134,135]. Alterations in stilbenoid production could have far reaching effects on the health of the plant. The severity of this problem has been highlighted by Fischer and colleagues [136] who reported that overexpression of a stilbene synthase gene in transgenic tobacco and petunia plants resulted male in sterility. Stilbenes can cause a variety of other physiological problems for plants. Stilbenes alter ion transport by inhibiting ATPase activity [137,138,139]. Other stilbenes, including 1, have been shown to be rather effective in vitro inhibitors of photosynthesis [140,141]. As a result, the production of defensive stilbenes could have severe consequences on normal metabolic activities in plants. Accordingly, it is imperative that plants maintain guarded regulation over the synthesis of 1. Therefore, following the synthesis of 1 and other stilbenoids, it is important that these compounds are sequestered in storage vesicles in order to reduce their negative impact upon basic physiological functions in the plant [140]. As a feasible alternative, plants may transform 1 into oligomers as a means of detoxification (Fig. (3)). By forming oxidized resveratrol oligomers, the levels of the 'free' monomer are reduced, thereby sparing the plant from the deleterious effects of 1 on normal metabolic processes.

556 The resveratrol oligomers could then be stored by the plant in nonphotosynthetic tissues where they could act as defensive tannins [77] in the stems and roots. Alternatively, the resveratrol oligomers may be incorporated as part of the protective lignified tissues that are typically produced as a barrier at the site of infection [142]. Pharmacological Properties of Resveratrol Oligomers The diversity of biological properties associated with polyphenolic compounds is unquestionable [135,142,143,144]. These compounds have demonstrated great potential as pharmacological agents for treating a plethora of diseases. This includes the resveratrol oligomers that have been shown in recent years to exhibit an array of promising biological properties (Table (5)). Several resveratrol oligomers have been tested for antimicrobial properties. Many have shown moderate inhibitory activity against bacterial pathogens such as Escherichia coli and Staphylococcus spp. [41,43^44,45,52,54,59,62]. Other resveratrol oligomers have been evaluated for anti-HIV activity. In vitro testing of two resveratrol oligomers, balanocarpol (17b) and dibalanocarpol (20b), showed that both possessed moderate inhibitory activity against the HIV virus with EC50 values of 20 //g/mL and 46 //g/mL, respectively [53]. Several other resveratrol oligomers have been tested; however, these compounds did not inhibit HIV-viral replication [106]. Free-radical generation and inflammatory processes are associated with the development of a host of diseases such as arthritis, cancer, cardiovascular disease, and immune disorders. Several resveratrol oligomers have been screened for their ability to inhibit selected steps associated with the development of these diseases. A number of resveratrol oligomers have been shown to exhibit antioxidant, antihyaluronidase, anti-inflammatory, and cyclooxygenase inhibitory activities. For example, Ono and colleagues [145] reported that several resveratrol trimers including miyabenol C (6a), c/^-miyabenol C (7a), and foeniculosides I-IV (2a-5a) exhibited potent antioxidant effects, while two of these oligomers, miyabenol C (6a) and cz5'-miyabenol C (7a), also

Table 5. ~

~~

Pharmacological Properties of Resveratrol Oligomers ~

Pharmacological

Compound

Property

Reference

Pharmacological

malibatol B (22b)

copalliferol B (28b) (-)-E-viniferin ( l l a )

cytotoxic

dehydrodimer (42b)

distichol (23b)

pallidol (10) vatdiospyroidol(26b)

kobophenol B (16b)

cyclooxygenase I

miyabenol A (12b)

inhibition

-

stemonoporol(19a)

(+)-c-viniferin (I l a ) hepatoprotective

balanocarpol(17b)

ampelopsin E (59) cis -ampelopsin

E (60)

ampelopsin C (57) hepatotoxic

cis /trans -vitisin A

(+)-a-viniferin @a)

tumor necrosis

(-)-/runs -vitisin B (79b)

foeniculoside I (2a) foeniculoside II (3a)

factor inhibition prostaglandin H,

(+)-vitisin C (81a) (+)-a-vinifcrin (8a)

foeniculoside 111 (4a)

synthase inhibition

foeniculoside IV (5a)

protein kinase C

(+)-a-viniferin (8a)

[66,146]

inhibition

kobophenol A (9)

166,1461

miyabenol C ( 6 )

[66,146]

miyabenol C (6a) cis -mivabenol C (7a)

antioxidant

resveratrol-tram

dehydrodimer (42b)

dibalanocarpol(20b)

anti-inflammatory

restrytisol C (93b)

miyabenol C (6a) vaticafinol (25b)

antihyalurouidase

resveratrol-trans -

c-viniferin (11) kobophenol A (9a)

anti-HIV

Reference

malibatol A (21b)

copaiifero~A (18a)

antibacterial

1

Compound

Property

miyabenol C (6a) cis -miyabenol C (7a)

mixture (77a/76a)

558 inhibited hyaiuronidase activity with IC50 values of 135 and 60 juM, respectively. Another resveratrol trimer, (+)-a-viniferin (8a) has been reported to possess anti-inflammatory properties [64]. In a mouse carragenininduced paw edema model, ;7.o. injections of 10 mg/kg of 8a reduced paw edema by approximately 26% after five hours. One of the important groups of enzymes associated with inflammatory processes is cyclooxygenases. It has been shown that 8a can also inhibit the prostaglandin H2 synthase activity of cyclooxygenase with an IC50 value of 7 juM [42]. More recently, other resveratrol oligomers such as resveratrol trans-dohydrodirnQv (42b) and restrytisol C (93b) have also demonstrated moderate cyclooxygenase I inhibitory activity with IC50 values of 26 and 47 //M, respectively [106]. Some resveratrol oligomers have demonstrated toxicity toward mammalian cells. It has been reported that malibatols A (21b) and B (22b) (IC50 13 and 21 jug/mL, respectively) [53], as well as, pallidol (10) and resveratrol trans-dohydrodirnQv (42b) (IC50 32 and 49 juM, respectively) [106] exhibited moderate cytotoxicity toward human lymphoblastoid cells. The resveratrol tetramer, vatdiospyroidol (26b), has shown an increased degree of cytotoxicity against several cancer cell lines including oral epidermal carcinoma (ED50 1.0 jug/mL), colon cancer (ED50 1.9 jug/mL), and breast cancer (ED50 3.8 jug/mL) [61]. Other resveratrol oligomers have been evaluated for their hepatotoxic properties [91]. Mice treated with 30 mg/kg of ampelopsin C (57a) or a mixture of CIS' and rra^^-vitisin A (77a and 76a, respectively) expressed elevated levels of plasma alanine transaminase relative to control animals while the same dose of 11a exhibited hepatoprotective effects in mice treated with carbon tetrachloride. Two other resveratrol oligomers, transampelopsin E (59) and c/^-ampelopsin E (60), were also found to possess hepatoprotective properties based on in vitro studies utilizing primary cultured rat hepatocytes treated with carbon tetrachloride [82]. In the search for new anticancer agents, many compounds have been evaluated for their abilities to interrupt normal intracellular signal transduction pathways such as those mediated by protein kinase C. Xu and colleagues [146] reported that three resveratrol oligomers, (+ysviniferin (11a), miyabenol C (6), and kobophenol A (9), inhibited the activity of rat protein kinase C with IC50 values of 62.5, 27.5, and 52.0 juM, respectively. In addition, Ito and Niwa [103] reported that both

559 (-)-rra^5-vitisin B (79b) and (+)-vitisin C (90a) were able to inhibit the activity of tumor necrosis factor; however, the full experimental details of these tests were not described. In light of the many promising biological properties associated with these compounds, it is conceivable that many new pharmacological applications of the resveratrol oligomers await discovery. CONCLUSIONS The resveratrol oligomers are a unique class of biologically active polyphenolic compounds. Compelling evidence clearly indicates that the resveratrol oligomers are involved in a number of important physiological processes in plants. The most important of these appears to be the phytoalexin role of the resveratrol oligomer; however, studies suggest that this benefit may come at a price to the plant. Additional research is needed to further define the roles that the resveratrol oligomers play in plants and to discern how these compounds impact the physiological status of a plant. Moreover, the enzymes and cofactors involved in the formation of the resveratrol oligomers are of particular interest due to their value as tools for providing a greater understanding of the stereochemical precision involved in the formation of natural products by enzymes that utilize free-radical chemistry. Due to the wide array of promising pharmacological properties exhibited by resveratrol oligomers, these compounds may serve as leads for the development of new drugs. The resveratrol oligomers show intriguing potential for serving as agents that can inhibit inflammatory processes. Further research is needed to screen additional oligomers in order to deduce meaningful structure-activity relationships for this class of compounds. In addition, many plants that produce 1 have not yet been examined for the presence of resveratrol oligomers. It is conceivable that these and other plant species may contain many new resveratrol oligomers. The future of resveratrol oligomer research appears bright as scientists continue to discover new oligomeric entities and to further define the pharmacological potential of these compounds.

560

Appendix 1.

Structures of Resveratrol Oligomers HO

(

V-OR,

2a 3a 4a 5a l2i

foeniculoside I foeniculoside II foeniculoside III foeniculoside IV c/>miyabenol C

Rl D-Glc H D-Glc D-Glc H

R2 H D-Glc D-Glc D-Glc H

OH

1 /m^T5-resveratrol

8 or-viniferin 8a (+)-a-viniferin 8b (-)-cif-viniferin

^-r^< 10 pallidol

9 kobophenol A ^OH

m-r\ H-7 H-8 11 f-viniferin 11a (+)-//'<3rt5-f-viniferin llb(-)-/r<3«5-£'-viniferin

a /?

P a

12b miyabenol A

R3 H H H D-Glc H

561

Appendix 1.

Structures of Resveratrol Oligomers (Continued)

OH HO

13b miyabenol B

14a (+)-hopeaphenol 15b (-)-isohopeapheno!

14 hopeaphenol 14b (-)-hopeaphenol

HO

17b balanocarpol

HO

18a copal liferol A

H-7,H-7" p a

562

Appendix 1.

Structures of Resveratrol Oligomers (Continued)

19a stemonoporol

20b dibalanocarpol OH

21bmalibatol A

22b malibatol B HO

H-7 23b distichol a 24b canaiiculatol p

OH

25b vaticaffmol

563

Appendix 1.

Structures of Resveratrol Oligomers (Continued)

Ci 26b vatdiospyroidol OH HO

rf

"^

0"

29a caraganaphenol A OH

28b copalliferol B OH

OH ^ ^

32a leachianol B

OH

564

Appendix 1.

Structures of Resveratrol Oligomers (Continued)

„o-^

^OH ^ ^

// X\

OH

33b leachianol C HO

HO

"OH

^^

OH

^ff

^OH ^ ^

36 leachianol F 37 leachianol G

35b leachianol E

OH

40b stenophyllol B

41bstenophyllol B

OH

565

Appendix 1.

Structures of Resveratrol Oligomers (Continued)

42 resveratrol /ra/i^-dehydrodimer 43 gnetin A

45 gnetin C

o" V^OH 46 gnetin E

48 gnetin G

49 gnetin H

566

Appendix 1.

Structures of Resveratrol Oligomers (Continued) OH

.OH

HO

51 anigopreissin A 50 gnetin I

52 suffruticosol A 53 suffruticosol B

H-7 p a

54 suffruticosol C (R=4-hydroxyphenyl)

OH

55a (+)-ampelopsin A OH 56a (+)-ampelopsin B H

HO

58 ampelopsin D

59 /r^wj-ampelopsin E trans 60 c/5-ampelopsin E cis

567

Appendix 1.

Structures of Resveratrol Oligomers (Continued)

61a ampelopsin F

62a ampelopsin G (R=4-hydroxyphenyl)

OH

63a ampelopsin H

64b parthenocissin A

66 cyphostemmin A

OH

67 cyphostemmin B

68 c/\s-£'-viniferine

568

Appendix 1. HO

Structures of Resveratrol Oligomers (Continued) HO.

HO

OH

OH

70b isoampelopsin F

OH

69b parthenocissin B

72a betulifol A

74b heyneanol A

OH

73a betulifol B

75 vitisin A not defined 76a {+)-trans-\iiis'\n A trans ll2k (+)-cw-vitisin A cis

HO

569

Appendix 1.

Structures of Resveratrol Oligomers (Continued)

78a f-viniferin diol

«.,0°"

79b {-ytrans-\iiisin B 80b (-)-c;5-vitisin B 81a (+)-vitisin C

A'» trans cis trans

82a (+)-vitisin D 83a (+)-vitisin E

85b r-viniferin

H-8'" a a fS

570

Appendix 1.

HO'

^ ^ ^ HU

Structures of Resveratrol Oligomers (Continued)

--

OH

OH

86a r-2-viniferin

OH

88 viniferifuran 89a (+)-vitisfuran A

OH

OH

91brestrytisol A

571 Appendix 1.

Structures of Resveratrol Oligomers (Continued)

HO^J^

OH

93b restrytisol C

92b restrytisol B

OH 94 piceid

iX^ ^°-

95 resveratrol OS ide

H,CO. 0CH3

96 7-5', 8-^neolignan

OH

OH

98 c/5-resveratrol

HiCO

97 7-^, S-R neolignan

572

References Takaoka, M. J. Faculty Sci. Hokkaido Imperial U., 1940, 3,1-16. Soleas, G.J.; Diamandis, E.P.; Goldberg, D.M. Clin. Biochem., 1997,50,91-113. [3] Gorham, J. The Biochemistry of the Stilbenoids, Chapman and Hall: New York, 1995. [4] Kindl, H. In Biosynthesis and Biodegradation of Wood Components; T. Higuchi, Ed.; Academic Press: New York, 1985, pp. 349-377. [5] Rupprich, N.; Kindl, H. Hoppe-Seyler's Z. Physiol. Chem., 1978, 359, 165-172. [6] Schoppner, A.; Kindl, H. J. Biol. Chem., 1984,259, 6806-6811. [7] Sotheeswaran, S.; Pasupathy, V. Phytochemistry, 1993, 32, 10831092. [8] Langcake, P.; Pryce, R.J. Experientia, 1977, 33, 151-152. [9] Hain, R.; Reif, H.-J.; Krause, E.; Langebartels, R.; Kindl, H.; Vomam, B.; Wiese, W.; Schmeizer, E.; Schreier, P.H.; Stocker, R.H.; Stenzel, K. Nature, 1993, 361, 153-156. [10] Leckband, G.; Lorz, H. Theor. Appl. Genet, 1998, 96, 1004-1012. [11] Belguendouz, L.; Fremont, L.; Linard, A. Biochem. Pharmacol., 1997,55,1347-1355. [12] Frankel, E.N.; Waterhouse, A.L.; Kinsella, J.E. Lancet, 1993, 341, 1103-1104. [13] Miller, N.J.; Rice-Evans, C.A. Clin. Chem., 1995, 41, 1789. [14] Vinson, J.A.; Jamg, J.; Dabbagh, Y.A.; Serry, M.M.; Cai, S. J. Agric. Food Chem. 1995, 43, 2798-2799. [15] Turrens, J.F.; Lariccia, J.; Nair, M.G. Free Radical Res., 1997, 27, 557-562. [16] Pace-Asciak, C.R.; Rounova, O.; Hahn, S.E.; Diamandis, E.P.; Goldberg, D.M. Clin. ChimActa, 1996, 246,163-182. [17] Bertelli, A.A.E.; Giovarmini, L.; Giannessi, D.; Migliori, M.; Bernini, W.; Fregoni, M.; Bertelli, A. Int J. Tiss. Reac, 1995,17, 1-3. [18] Chung, M.-L; Teng, C.-M.; Cheng, K.-L.; Ko, F.-N.; Lin, C.-N. Planta Med, 1992, 58, llA-lie. [19] Pace-Asciak, C.R.; Hahn, S.; Diamandis, E.P.; Soleas, G.; Goldberg, D.M. Clin. Chim. Acta, 1995, 235, 207-219.

[1] [2]

573 [20] [21] [22] [23] [24] [25] [26]

[27] [28] [29] [30] [31] [32] [33] [34] [35]

Ferrero, M.E.; Bertelli, A.A.E.; Fulgenzi, A.; Pellegatta, P.; Corsi, M.M.; Bonfrate, M.; Ferrara, F.; DeCaterina, R.; Giovannini, L.; Gertelli, A. Am. J. Clin. Nutr., 1998, 68, 1208-1214. Ferrero, M.E.; Corsi, M.; Giovannini, L.; Bonfrate, C; DeCaterina, R.; Tillement, J.P.; Bertelli, A.A.E.; Bertelli, A. Int. J. Tiss. Reac.,1991,19, 121-124. Pendurthi, U.R.; Williams, J.T.; Rao, L.V.M. Artioscler. Thromb. Vase. Biol, 1999,19, 419-426. Ashby, J.; Pennie, H.; Brooks, A.N.; Lefevre, P.A.; Beresford, N.; Sumpter, J.P. J. Appl. Toxicol, 1998,19, 39-45. Gehm, B.D.; McAndrewa, J. M.; Chien, P.-Y.; Jameson, J.L. Proc. Natl Acad. Set U.S.A., 1997, 94, 14138-14143. Kopp, P. J. Endocrinol, 1998,138, 619-620. Jang, M.; Cai, L.; Udeani, G.O.; Slowing, K.V.; Thomas, C.F.; Beecher, C.W.W.; Fong, H.H.S.; Farnsworth, N.R.; Kinghom, A.D.; Mehta, F.G.; Moon, R.C.; Pezzuto, J.M. Science, 1997, 275, 218-220. Jang M.; Pezzuto, J.M. Cancer Lett., 1998,134, 81-89. Uenobe, F.; Nakamura, S.-I.; Miyazawa, M. Mutat. Res., 1997, 373, 197-200. Ciolino, H.P.; Daschner, P.J.; Yeh, G.C. Cancer Res., 1998, 58, 5707-5712. Subbaramaiah, K.; Chung, W.J.; Michaluart, P.; Telang, N.; Tanabe, T.; Inoue, H.; Jang, M.; Pezzuto, J.M.; Darmenberg, A.J. J. Biol Chem., 1998, 273,21875-24882. Mgbonyebi, O.P.; Russo, J.; Russo, I.H. Int. J. Oncol, 1998, 12, 865-869. Carbo, N.; Costelli, P.; Baccino, F.M.; Lopez-Soriano, F.J.; Argiles, J.M. Biochem. Biophys. Res. Commun., 1999, 254, 739743. Jayatilake, G.S.; Jayasuriya, H.; Lee, E.-S.; Koonchanol, N.M.; Geahlen, R.L.; Ashendel, C; McLaughlin, J.L.; Chang, C.-J. J. Nat. Prod, 1993, 56,1805-1810. Chang, C.-J.; Ashendel, C.L.; Geahlen, R.L.; McLaughlin, J.L.; Waters, D.J. In Vivo, 1996,10, 185-190. Fontcave, M.; Lepoivre, M.; EUeingand, E.; Gerez, C; Guittet, 0. FEBSLett., 1998, 421, 277-279.

574

[36]

Fukuhara, K.; Miyata, N. Biorg. Med. Chem. Lett., 1998, 8, 31873192. [37] Clement, M.-V.; Hirpara, J.L.; Chawdhury, S.-H.; Pervaiz, S.; 5/ooJ, 1998,92,996-1002. [38] Ragione, F.D.; Cucciolla, V.; Borriello, A.; Pietra, V.D.; Racioppi, L.; Soldati, G.; Manna, C; Galletti, P.; Zappia, V. Biochem. Biophys. Res. Commun., 1998, 250, 53-58. [39] Ono, M.; Ito, Y.; Kinjo, J.; Yahara, S.; Nohara, T.; Niiho, Y. Chem. Pharm. Bull, 1995, 43, 868-871. [40] Senda, N.; Kubota, Y.; Hoshino, T.; Nozaki, H.; Hayashi, H.; Nakayama, M. J. Mass Spectrom. Soc. Jpn., 1995, 43, 45-51. [41] Suzuki, K.; Shimizu, T.; Kawabata, J.; Mizutani, J. Agric. Biol. C/zem., 1987, 57,1003-1008. [42] Lee, S.-H.; Shin, N.-H.; Kang, S.-H.; Park, J.S.; Chung, S.R.; Min, K.R.; Kim, Y. Planta Med., 1998, 62, 204-207. [43] Kawabata, J.; Ichikawa, S.; Kurihara, H.; Mizutani, J. Tetrahedron Ze«., 1989, 30, 3785-3788. [44] Kurihara, H.; Kawabata, J.; Ichikawa, S.; Mishima, M.; Mizutani, J. Phytochemistry, 1991, 30, 649-653. [45] Kawabata, J.; Mishima, M.; Kurihara, H.; Mizutani, J. Phytochemistry, 1991, 30, 645-647. [46] Kurihara, H.; Kawabata, J.; Ichikawa, S.; Mizutani, J. Agric. Biol. C/zem., 1990, 54, 1097-1099. [47] Kawabata, J.; Fukushi, E.; Hara, M.; Mizutani, J. Magn. Reson. Chem., 1992, 30,6-10. [48] Coggon, P.; Janes, N.F.; King, F.E.; King, T.J.; Molyneux, R.J.; Morgan, J.W.W.; Sellars, K. J. Chem. Soc, Chem. Commun., 1965, 406-409. [49] Coggon, P.; King, T.J.; Wallwork, S.C. J. Chem. Soc, Chem. Commun. 1966,439-440. [50] Coggon, P.; McPhail, A.T.; Wallwork, S.C. J. Chem. Soc (B), Physical Organic Chem., 1970, 884-897. [51] Diyasena, M.N.C.; Sotheeswaran, S.; Surendrakumar, S. J. Chem. Soc, Perkin Trans. 1,1985, 1807-1809. [52] Sotheeswaran, S.; Sultanbawa, M.U.S.; Surendrakamar, S.; Bladon, P. J. Chem. Soc, Perkin Trans. 1,1983, 699-702. [53] Dai, J.-R.; Hallock, Y.F.; Cardellina II, J.H.; Boyd, M.R. J. Nat. Prod, 199S, 61,351-353.

575 [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70]

Sultanbawa, M.U.S.; Surendrakumar, S.; Bladon, P. Phytochemistry, 1987, 26, 799-801. Madhav, R.; Sechadri, T.R.; Subramanian, G.B.U. Phytochemistry, 1967, 6, 1155-1156. Samaraweera, U.; Sotheeswaran, S.; Sultanbawa, M.U.S. Phytochemistry, 1982, 21, 2585-2587. Bokel, M.; Diyasena, M.N.C.; Gunatilaka, A.A.L.; Kraus, W.; Sotheeswaran, S. Phytochemistry, 1988, 27, 377-380. Sotheeswaran, S.; Diyasena, M.N.C.; Gunatilaka, A.A.L.; Bokel, M.; Kraus, W. Phytochemistry, 1987, 26, 1505-1507. Sotheeswaran, S.; Sultanbawa, M.U.S.; Surendrakamar, S.; Balasubramaniam, S.; Bladon, P. J. Chem. Soc, Perkin Trans. I, 1985, 159-162. Sultanbawa, M.U.S.; Surendrakumar, S.; Wazeer, M.I.M. J. Chem. Soc, Chem. Commun., 1981, 1204-1206. Seo, E.-K. Ph.D. Thesis, University of Illinois at Chicago, Chicago, IL 1997. Gunawardena, Y.A.G.P.; Sotheeswaran, S.; Sultanbawa, M.U.S.; Surendrakumar, S.; Bladon, P. Phytochemistry, 1986, 25, 14981500. Sultanbawa, M.U.S.; Surendrakumar, S.; Bladon, P. J. Chem. Soc, Chem. Commun., 1980, 619-620. Kitanaka, S.; Ikezawa, T.; Yasukawa, K.; Yamanouchi, S.; Takido, M.; Sung,.; Kim, H.K,; I.H. Chem. Pharm. Bull, 1990, 38, 432-435. Kitanaka, S.; Takido, M.; Mizoue, K.; Kondo, H.; Nakaike, S. Chem. Pharm. Bull, 1996, 44, 565-567. Kulanthaivel, P.; Janzen, W.P.; Ballas, L.M.; Jiang, J.B.; Hu, CQ.; Darges, J.W.; Seldin, J.C; Cofield, D.J.; Adams, L.M. Planta Med., 1995, 61, 41-44. Ohyama, M.; Ichise, M.; Tanaka, T.; linuma, M.; Burandt, Jr., C.L. Tetrahedron Lett, 1996, 37, 5155-5158. linuma, M.; Ohyama, M.; Tanaka, T.; Lang, F.A. Phytochemistry, 1994,37,1157-1159. Ohyama, M.; Tanaka, T.; linuma, M.; Goto, K. Chem. Pharm. Bull, 1994. 42,2111-2120. Ohyama, M.; Tanka, T.; linuma, M. Tetrahedron Lett., 1994, 35, 7817-7820.

576 [71]

Ohyama, M.; Tanaka, T.; linuma, M. Phytochemistry, 1995, 55, 733-740. [72] Ohyama, M.; Tanaka, T.; linuma, M.; Burandt, Jr., C.L. Chem. Pharm. Bull, 1998, 46, 663-668. [73] Huang, K.-S.; Wang, Y.-H.; Li, R.-L.; Lin, M. J. Nat. Prod., 2000, 63, 86-89. [74] Lins, A.P.; Gottlieb, O.R.; Gottlieb, H.E. Anales Asoc. Quim. Argentina, 1982, 70, 257-262. [75] Lins, A.P.; Ribiero, M.N. DeS.; Gottlieb, O.R.; Gottlieb, H.E. J. Nat. Prod., 1982, 45, 754-761. [76] Lin, M.; Li, J.B.; Li, S.Z.; Yu, D.Q.; Liang, X.T. Phytochemistry, 1992, 31, 633-638. [77] Boralle, N.; Gottlieb, H.E.; Gottlieb, O.R.; Kubitzki, K.; Lopes, L.M.X.; Yoshida, M.; Young, M.C.M. Phytochemistry, 1993, 34, 1403-1407. [78] Lins, A.P.; Yoshida, M.; Gottlieb, O.R.; Gottlieb, H.E.; Kubitzki, K. Bull. Soc. Chim. Belg., 1986, 95, 737-747. [79] Holscher D.; Schneider, B. Phytochemistry, 1996, 43, 471-473. [80] Sarker, S.D.; Whiting, P.; Kinan, L.; Sik, V.; Rees, H.H. Tetrahedron, 1999, 55, 513-524. [81] Oshima, Y.; Ueno, Y.; Hikino, H.; Yang, L.-L.; Yen, K.-Y. Tetrahedron, 1990, 46, 5121-5126. [82] Oshima, Y.; Ueno, Y.; Phytochemistry, 1993, 33, 179-182. [83] Oshima, Y.; Ueno, Y.; Hisamichi, K.; Takeshita, M. Tetrahedron, 1993, ^P, 5801-5804. [84] Khan, M.A.; Nabi, S.G.; Prakash, S.; Zaman, A. Phytochemistry, 1986, 25, 1945-1948. [85] Adesanya, S.A.; Nia, R.; Martin, M.-T.; Boukamcha, N; Montagnac, A.; Pais, S. J. Nat. Prod, 1999, 62, 1694-1695. [86] Ducrot, P.-H.; Kollmann, A.; Bala, A.E.; Majira, A.; Kerhoas, L.; Delorme, R.; Einhom, J. Tetrahedron Lett., 1998, 39, 9655-9658. [87] Tanaka, T.; linuma, M.; Murata, H. Phytochemistry, 1998, 48, 1045-1049. [88] Tanaka, T.; Ohyama, M.; Morimoto, K.; Asai, F.; linuma, M. Phytochemistry, 1998, 48,1241-1243. [89] Lins, A.P. ; Felicio, J.D'A.; Braggio, M.M.; Roque, L.C. Phytochemistry, 1991, 30, 3144-3146. [90] Li, W.; Li, B.; Chen, Y. Phytochemistry, 1998, 49, 1393-1394.

577

[91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110]

Oshima, Y.; Namao, K.; Kamijou, A.; Matsuoka, S.; Nakano, M. Experientia, 1995, 51, 63-66. Oshima, Y.; Kamijou, A.; Ohizumi, Y.; Niwa, M.; Ito, J.; Hisamichi, K.; Takeshita, M. Tetrahedron, 1995, 57, 1197911986. Ito, J.; Gobaru, K.; Shimamura, T.; Niwa, M.; Takaya, Y.; Oshima, Y. Tetrahedron, 1998, 54, 6651-6660. Oshima, Y.; Kamijou, A.; Moritani, H.; Namao, K.-I.; Ohizumi, Y. J. Org. Chem., 1993, 58, 850-853. Shinoda, K.; Takaya, Y.; Ohta, T.; Niwa, M.; Hisamichi, K.; Takeshita, M.; Oshima, Y. Heterocycles, 1997, 46, 169-172. Li, W.-W.; Li, B.-G.; Chen, Y.-Z. J. Nat. Prod, 1998, 61, 646647. Li, W.-W.; Ding, L.-S.; Li, B.-G.; Chen, Y.-Z. Phytochemistry, 1996,^2,1163-1165. Mattivi, F.; Reniero, F. Bull. Liais. Groupe Polyph, 1992, 16, 116-119. Korhammer, S.; Reniero, F.; Mattivi, F. Phytochemistry, 1995, 38, 1501-1504. Pryce, R.J.; Langcake, P. Phytochemistry, 1977,16,1452-1454. Ourtoule, J.-C; Bourhis, M.; Vercauteren, J.; Theodore, N. Tetrahedron Lett., 1996, 27,4697-4700. Reniero, F.; Rudolph, M.; Angioni, A.; Bernreuther, A.; Cabras, P.; Mattivi, F. Vitis, 1996, 35, 125-127. Ito J.; Niwa, M.; Tetrahedron, 1996, 52, 9991-9998. Ito, J.; Niwa, M.; Oshima, Y. Heterocycles, 1997, 45, 1809-1883. Ito, J.; Takaya, Y.; Oshima, Y.; Niwa, M. Tetrahedron, 1999, 55, 2529-2544. Cichewicz, R.H.; Kouzi, S.A.; Hamann, M.T. J. Nat. Prod, 2000, 63, 29-33. Breuil, A.-C; Adrian, M.; Pirio, N.; Meunier, P.; Bessis, R.; Jeandet, P. Tetrahedron Lett, 1998, 39, 537-540. Langcake, P.; Pryce, R.J. J. Chem. Soc, Chem. Commun., 1917, 208-210. Yeo, H.; Lee, J.H.; Kim, J. Arch. Pharm. Res., 1999, 22, 306-308. Barcelo, A.R.; Morales, M.; Pedreno, M.A. In Lignin and Lignan Biosynthesis; N.G. Lewis, S. Sarkanen, Ed.; American Chemical Society: Washington, D.C., 1998, pp. 84-95.

578

[111] Morales, M.; Alcantara, J.; Ros Barcelo, A. Am. J. Enol. Vitic, 1997, 48, 33-38. [112] Hoos, G.; Blaich, R. Vitis, 1988, 27,1-12. [113] Calderon, A.A.; Zapata, J.M.; Barcelo, A.R. Physiol. Mol Plant Pathol, 1994, 44, 289-299. [114] Davin, L.B.; Wang, H.-B.; Cowell, A.L.; Bedger, D.L.; Martin, D.M.; Sarkanen, S.; Lewis, N.G. Science, 1997, 275, 362-366. [115] Gang, D.A.; Costa, M.A.; Fujita, M.; Dinkova-Kostova, A. T.; Wang, H.-B.; Burlat, V.; Martin, W.; Sarkanen, S.; Davin, L.B.; Lewis, N.G. Chem. Biol, 1999, 6, 143-151. [116] Pezet, R. FEMSMicrobiol Lett., 1998,167, 203-208. [117] Mattivi, F.; Reniero, F. Polyph. Commun., 1996,1, 125-126. [118] Adesanya, S.A.; Nia, R.; Martin, M.-T.; Boukamcha, N.; Montagnac, A.; Pais, M. J. Nat. Prod., 1999, 62, 1694-1695. [119] Kawabata, J.; Mishima, M.; Kurihara, H.; Mizutani, J. Phytochemistry, 1995, 40, 1507-1510. [120] Jeandet, P.; Sbaghi, M.; Bessis, R. J. Wine Res., 1992, 3, 47-57. [121] Ingham, J.L. Phytochemistry, 1976, 75, 1791 -1793. [122] Adrian, M.; Jeandet, P.; Bessis, R; and Joubert, J.M. J. Agric. Food Chem., 1996, 44, 1979-1981. [123] Dercks W.; Creasy, L.L. Physiol Mol Plant Pathol, 1989, 34, 189-202. [124] Langcake, P.; Pryce, R.J. Phytochemistry, 1977,16, 1193-1196. [125] Jeandet, P.; Bessis, R.; Sbaghi, M.; meunier, P. J. Phytopathol, 1995,143, 135-139. [126] Langcake, P. Physiol Plant Pathol, 1981,18, 213-226. [127] Pezet, R.; Pont, V.; Hoang-Van, K. Physiol Mol Plant Pathol, 1991, 39, 441-450. [128] Sbaghi, M.; Jeandet, P.; Bessis, R.; Leroux, P. Plant Pathol, 1996, 45, 139-144. [129] Thurston, C.F. Microbiology, 1994,140,19-26. [130] Ducros, V.; Brzozowski, A.M.; Wilson, K.S.; Brown, S.H.; Ostergaard, P.; Schneider, P.; Yaver, D.S.; Pedersen, A.H.; Davies, G.J. Nat. Struc. Biol, 1998, 5, 310-316. [131] Mayer, A.M. Phytochemistry, 1987, 26,11-20. [132] Mayer, A.M.; Harel, E. Phytochemistry, 1979,18, 193-215. [133] Adrian, M.; Rajaei, H.; Jeandet, P.; Veneau, J.; Bessis, R. Phytopathology, 1998, 88, 472-476.

579 [134] Dixon, R.A. Curr. Opin. BiotechnoL, 1999,10, 192-197. [135] Mann, J.; Davidson, R.S.; Hobbs, J.B.; Banthorpe, D.V.; Harbome, J.B. Natural Products: Their Chemistry and Biological Significance, Longman: Essex, 1994. [136] Fischer, R.; Budde, I.; Hain, R. Plant J., 1997,11, 489-498. [137] Balke, N.E.; Hodges, T.K. Plant Physiol, 1979, 63, 42-47. [138] Balke, N.E.; Hodges, T.K. Plant Physiol, 1979, 63, 48-52. [139] Balke, N.E.; Hodges, T.K. Plant Physiol, 1979, 63, 53-56. [140] Gorham, J.; Coughlan, S.J. Phytochemistry, 1980,19, 2059-2064. [141] Rich, P.R.; Madgwick, S.A.; Brown, S.; Von Jagow, G.; U. Brandt, Biochemical Society Transactions, 1991,19, 263S. [ 142] Haslam, E. J. Nat. Prod, 1996, 59, 205-215. [143] Frankel, E.N. In Antioxidant Food Supplements in Human Health, L. Packer, M. Hiramatsu, T. Yoshikawa, Ed.; Academic Press: New York, 1999, pp. 385-392. [144] Okuda, T. In Antioxidant Food Supplements in Human Health, L. Packer, M. Hiramatsu, T. Yoshikawa, Ed.; Academic Press: New York, 1999, pp. 393-410. [145] Ono, M.; Masuoka, C; Ito, Y.; Niho, Y.; Kinjo, J.; Nohara, T. FoodSci. Technol. Int. Tokyo, 1997, 3, 53-55. [146] Xu, G.; Zhang, L.P.; Chen, L.F.; Hu, C.Q. Acta Pharm. Sinica, 1994,29,818-822.

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

581

BIOACTIVE METABOLITES FROM PHYTOPATHOGENIC BACTERIA AND PLANTS ANTONIO EVIDENTE* AND ANDREA MOTTAt *Dipartimento di Scienze Chimico-Agrarie, Universitd di Napoli 'Tederico IF, Via Universitd 100, 80055 Portici, Italy; fistituto per la Chimica di Molecole di Interesse Biologico (Istituto Nazionale di Chimica dei Sistemi Biologici) del CNR, Via Toiano 6, 80072 Arco Felice, Italy ABSTRACT.We report on the isolation, the chemical and biological characterisation of several plant growth regulators, and some phytotoxins produced by bacteria pathogens for plants with agrarian and ornamental interest. We also describe studies on the structure-activity relationships and on the mode of action of the metabolites by resorting to stereoselective synthesis and the chemical derivatisation of some phytohormones. The relationships of some bacterial metabolites with plant pathogenesis is also described. Furthermore, the isolation, chemical and biological characterisation of phenathridine alkaloids isolated from some Amaryllidaceae plants is illustrated. The NMR properties, microbiological transformation, structure-activity correlation and mode of action of lycorine, the main Amaryllidaceae alkaloid, is also reported.

INTRODUCTION Toxins produced by phytopathogenic fungi and bacteria are involved in pathogenesis that seriously damages agricuhural, forest and ornamental plants. Several studies have been carried out on the role of bioactive microbial metabolites in the pathogenic process and on their use against specific diseases. These studies have favoured isolation of phytohormones, phytoalexins, antibiotics, fungicides, herbicides and elicitors which are potentially important for practical applications. Microbial bioactive metabolites belong to several classes of natural products of low molecular weight (terpenes, macrolides, butenolides, cytokinins, auxins, etc.) as well as high molecular weight (proteins, glycoproteins and more recently polysaccharides) [1-6]. One of the potential applications is the control of weed diffusion, a serious problem

582 for crops and pastures, as well as for forest resources and ornamental plants. This is usually achieved by using large amounts of agrochemicals which cause high environmental pollution and serious problems to human and animal health. By contrast, biological agents (mainly insects, bacterial and fungal pathogens, and more recently phytotoxins) are compatible with the environment, showing high specificity, and represent a long term solution [4,6-8]. Moreover, some plant metabolites, as the phenanthridine alkaloid extracts from Amaryllidaceae species, are particularly important for their chemotaxonomic relevance, pharmacological applications, the role in modulating some important physiological plant mechanisms, and their biological effects. Some reviews concerning these aspects have been pubhshed [9-16]. We will report on the structure determination of plant growth regulators, phytotoxins and bacteriocins produced by some phytopathogenic bacteria. This is a fundamental step for their biological characterisation, derivatisation and the development of synthetic and analytical methods for their possible practical application. Furthermore, the isolation, chemical and biological characterisation of the cell wall and exopolysaccharides of these phytopathogenic bacteria are described_in order to reveal their probable role in the pathogenic process and chemotaxonomic importance. The results of similar studies carried out on the bioactive phenantridine alkaloids extracted from higher plants will be described. Bacterial metabolites During the last few decades, research on phytopathogenic bacteria belonging to different genera {Pseudomonas, Xanthomonas and Burkholderia) has rapidly progressed. Close collaboration among pathologists, physiologists and chemists has provided new insights into determinants of pathogeneticity and virulence. Progress has been made in elucidating chemical structures and synthesis as well as modes of action of phytototoxins, phytohormones, and structural metabolites from several phytopathogenic Pseudomonas species and related pathogens, establishing novel strategies for disease control and biotechnological apphcations [17].

583 Pseudomonay cytokinins

Several plant diseases are characterised by growth abnormalities of the infected tissues. Such effects are often determined by an alteration of the physiological hormone balance in the plant [18]. Among the phytopathogenic bacteria, Pseudomonas syringae subsp. savastanoi (Smith) Young, Dye et Wilkie, and Pseudomonas amygdali Psallidas and Panagopulos seemed microorganisms for which the development of the symptoms and the final appearance of the disease on host plants suggested the possible involvement of phytohormones in the pathogenic process. P. syringae subsp. savastanoi (=P. savastanoi) is the causal agent of olive knot, one of the most widespread diseases of olive and other Oleaceae (oleander, privet and ash). It is characterised by the formation of galls on young stems and, less frequently, on the other plant organs. Preliminary investigations evidenced accumulations of cytokinins in the culture filtrates of the bacterium [19, 20]. From the basic organic extract we isolated the well-known trans-ZQ2iim (^Z) and its 9-P-D-riboside (^ ZR) together with two new cytokinins, called I'-methylzeatin (I'MeZ) and l"-methylzeatin riboside (l"MeZR) [1, 2, 3 and 4, Fig. (1) and Table 1]. The two new cytokinins were characterised, using essentially spectroscopic methods [^H and ^^C NMR (Nuclear Magnetic Resonance) and EI and FAB MS (Electron Ionization and Fast Atom Bombardment Mass Spectrometry)], as 6-(4-hydroxy-1,3-dimethylbut-/ra«5-2enylamino)purine and its 9-P-D-riboside [21, 22]. In particular, the two new cytokinins (3 and 4) differed fi*om ^Z and ^ZR in the presence of a methyl group on the carbon of the hydroxyisoprenyl side chain attaching the N-10 of the adenine moiety. The structure of T'-MeZR was confirmed by the preparation of the corresponding tetracetyl derivative and in particular by its EI MS as well as the FAB MS data of the natural metabolite. Furthermore, acid hydrolysis of 4 showed the formation of pD-ribose [22]. The cytokinin activity of the new phytohormones was tested determining the stimulation of biosynthesis of chlorophyll in the cucumber cotyledon bioassay [23] and expressed as mg zeatin equivalent/1 [21, 22].

584 CH, CH2R2

CH, CH2OH

1 Ri=H R2=0H

3 R=H

2 Ri=p-D-ribofuranosyl R2=0H

4 R=p-D-ribofuranosyl

5 R,=R2=H 6 Ri=p-D-2-deoxyribofuranosyl

R2=0H

CH2OH CH2OH

7 R=H

9 R=H

8 R=p-D-ribofuranosyl

10 R=p-D-2-deoxyribofuranosyl

Fig. (1). Cytokinins (1-8) from Pseudomonas

spp.

However, the absolute stereochemistry of the chiral carbon bearing the new methyl group in the isoprenyl side chain of the I'MeZ and its 9-p-Driboside (r'MeZR) remained to be determined.

585

Table 1. Phytohormone

Plant Growth Regulators from Phytopathogenic Bacteria Molecular

Bacterium

lUPAC name

formula Ac-8-IAA-Lys

P. s. savastanoi 1 C,8H23N305

6-[(2£)-4-Hydroxy-3-methylbut-2-enylamino]

2'-Deoxyzeatin riboside

N-a -Acetyl-N-8-(indole-3-acetyl)-L-lysine

P. amygdali

C15H21N5O4

P. s. papulans

C9H9NO6

-9-(2-deoxy-(3-D-ribofuranosyl) purine

2',5'-Dihydroxy-3'nitrophenylacetic

2-(2,5-Dihydroxy-3-nitrophenyl)acetic acid methyl ester

acid methyl ester P. amygdali Dihydrozeatin

CioHisNsO

P. s. savastanoi

6-(4-Hydroxy-3 -methylbutylamino)

Dihydrozeatin riboside

P. s. savastanoi

CisHzsNsOs

P. s. papulans

CioHnNO^

-9-(P-D-ribofuranosyl)purine (5)-2-Hydroxy-3-(3-hydroxy-4-nitrophenyl)

1 o-Hydroxynitropapuline

6-(4-Hydroxy-3 -methylbutylamino)purine

1 Indolacetic acid

P. amygdali

I (lAA)

P. s. savastanoi

C,oH9N02

Indole-3-acetic acid

s-IAA-Lys

P. s. savastanoi

C,6H2lN303

N-8-(Indole-3-acetyl)-L-lysine

Indolaldehyde

P. s. savastanoi

C10H9NO

Indole-3-aldehyde

P. amygdali

CioHnNs

6-(3-Methylbutylamino)purine

CnHnN02

Indole-3-acetic acid methyl ester

CuHisNsO

6-[(2£:)-4-Hydroxy-l ,3-dimethylbut-2-enyl-

1 Isopentenyladenine

1

propanoic acid methyl ester

P. amygdali Methyl ester of IAA

I'-Methylzeatin

P. s. savastanoi P. s. savastanoi

aminojpurine

1 r'-Methylzeatin

6-[(2£)-4-Hydroxy-l ,3-dimethylbut-2-enyl P. s. savastanoi

C16H23N5O5

Papuline

P. s. papulans

C10H12O3

Papulinone

P. s. papulans

C13H14O5

riboside

-amino]-9-(P-D-ribofuranosyl)purine (5)-2-Hydroxy-3-phenylpropanoic acid methyl 1 ester 4-(l-Hydroxy-2-phenylethyl)-4-carbomethoxy-|

P. amygdali NZeatin

|

ethan-2-one 6-[(2£)-4-Hydroxy-3-methylbut-2-enylamino] 1

P. s savastanoi

CIOHBNSO

P. s. savastanoi

Cl5H2,N505

purine 6-[(2£)-4-Hydroxy-3-methylbut-2-enylamino] 1

/-Zeatin riboside

-9-((3-D-ribofuranosyl)purine

586 An enantioselective short and convenient sequence, starting from the D or L-alanine and adopting a key a-aminoaldehyde-olefination step, allowed the synthesis in a good yield of {I'R)- and {1'S)'VMQZ and of their corresponding 9-p-D-ribosides {V'R)- and (7"i)-l"MeZR, to estabHsh the identity of the diastereomer (77?)-rMeZ and (7";?)-l"MeZR with the two natural cytokinins (3 and 4) and to confirm their structures [24]. The availability of synthetic samples of these two cytokinins (3 and 4) allowed us to revise the l^C NMR chemical shift assignments of the two methyls of the side chain as well as those of C-2 and C-3 of the ribosyl moiety of 4. The first attributions of these carbons [21, 22] should be reversed as obtained from the comparison of the two cytokinins with the diastereomers {rS)'VMtZ and (7"*S)-r'MeZR and the cz.s'-isomers of 3 and 4 [25]. The latter, prepared to compare their biological activity with that of the naturally occurring geometrical isomers, were obtained by adopting a synthetic technology similar to that used to prepare the two natural cytokinins and their diastereomers [26]. P. amygdali is the causal agent of hyperplastic bacterial canker in almond. The most characteristic symptoms of the disease, found in Greece [27], Turkey [28] and Afghanistan [29], is the formation of perennial cankers on trunks, branches, twigs and shoots. Li liquid culture the microorganisms produced phytohormones. From the basic organic extract four cytokinins were isolated and three of them identified (using spectroscopic methods) as the well-known t-Z (1) [30], its corresponding dihydroderivative (diHZ), and the isopentenyladenine (iP) [7 and 5, Fig. (1) and Table 1] [31]. The remaining one proved to be a new cytokinin on the basis of spectroscopic data (essentially ^H and ^^C NMR and FAB MS), and was identified as a 6-(4-hydroxy-3-methylbut-^ra«5'-2enylamino)-9-(2-deoxy-p-D-ribofuranosyl)purine [6, Fig. (1) and Table 1], and therefore called 2'-deoxyzeatin riboside (2'deOZR) [31]. The latter differed from ^ZR (2) only in the sugar moiety, 6 being a 2deoxyriboside. The biological activity of 2'deOZR was characterised using the cucumber cotyledon bioassay [23]. The structure assigned to 6 was also confirmed by its total enantioselective synthesis. It was realised with a short and convenient sequence similar to that used to prepare I'MeZ and T'MeZR and their diastereomeric and geometric isomers with the final condensation of the 6-chloro-9-(2-deoxy-p-D-ribofuranosyl) purine with the (£)-4-amino-2-methylbut-2-en-l-ol [32]. A similar synthetic methodology adopting the a-aminoaldeyde-olefmation as a key

587 step, was used to synthesise the cz.s'-zeatin and its 9-(2-deoxy-P-Dribofuranosyl) derivative [9 and 10, Fig. (1)] [33]. When assayed in the stimulation of chlorophyll biosynthesis in etiolated cucumber cotyledons, the 2'deOZR and its c/^-isomer respectively showed a marked and a total loss of activity in comparison with t-ZR that proved to be the most active cytokinin [33]. These findings indicate that both the transstereochemistry of the side chain double bond, and an unaltered ribosyl residue, when present, should be two structural features important to stimulate the biosynthesis of chlorophyll in cucumber cotyledons. The interesting biological activities of cytokinins have stimulated a great deal of research on their structure-activity relationship. From the total synthesis of I'MeZ, its 9-P-D-riboside and 2'deOZR, it was possible to obtain a sufficient amount of these cytokinins to prepare their derivatives and to analyse their ability to stimulate the synthesis of chlorophyll in etiolated cucumber cotyledons. Chemical and enzymatic reactions allowed us to obtain derivatives of the three cytokinins (3, 4 and 6) modified on the isoprenyl side chain attached to the purine moiety, and on the sugar residue, when present. Li contrast to 2'deOZR, which showed very low cytokinin activity (see above) [33], I'MeZ and T'MeZR displayed a higher stimulating potency of chlorophyll synthesis when compared to ^Z and ^ZR, respectively. With regard to the synthetic analogues, any modification of the carbon skeleton of the side chain influenced the activity. In particular, the reduction of the trans-douhlQ bond in 3 and 4 produces two diastereomeric dihydro derivatives, which respectively showed a reduction and a complete loss of activity, probably due to the stereochemistry of the new side chain chiral carbon (C-3). Deoxygenation of the side chain in the two dihydro derivatives of I'MeZ and 1 "MeZR gave rise to two inactive compounds. The acylation of the side chain hydroxy group (at C-4) as well as the acetylation of the hydroxy groups of the ribose moiety, did not significantly affect the biological activity [34]. Moreover, a rapid diagnostic method for analysing the three new cytokinins (I'MeZ, its 9-p-D-riboside and 2'deOZR) was developed to detect them in complex samples. The natural cytokinins and some of their above-mentioned derivatives and analogues, were characterised by fast atom bombardment tandem mass spectrometry (FAB MS MS). The protonated molecular ions of the examined cytokinins could be fingerprinted fi-om breakdown patterns of their gaseous unimolecular

588 dissociations, thus providing means for their identification by the desorption ionisation method [35]. Pseudomonas auxins

The acid organic culture extract of an oleander P. savastanoi strain contained auxins that should have, together with cytokinins, a role in the pathogenic process induced by the bacterium on the host plant. A product of indole-3-acetic acid (lAA) metabolism was isolated together with lAA [11, Fig. (2) and Table 1]. By chemical and spectral determination (essentially ^H and ^^C NMR and FAB MS), this auxin was characterised as N-a-acetyl-N-8-(indole-3-acetyl)-L-lysine (Ac-8-IAALys) [14, Fig. (2) and Table 1]. It is structurally related to the conjugate N-s-(indole-3-acetyl)-L-lysine (s-IAA-Lys) [13, Fig. (2) and Table 1] previously isolated as metabolite from another strain of the same bacterium [36]. Li particular, by diazomethan esterification Ac-s-IAALys was converted into the corresponding methyl ester; by strong acid hydrolysis it yielded lysine whose stereochemistry proved to be L by an enzymatic determination carried out using the method described by Nakatani et al [37]. As expected, s-IAA-Lys was converted into 14 by mild acetylation [36]. When bioassayed, the conjugate Ac-s-IAA-Lys caused chlorosis in oleander and bean leaves, stimulated wheat hypocotyl elongation and induced hypertrophy in potato tubers, proving 60% less effective than lAA [36]. To further investigate the biological activity of these auxins, present at very low levels in the culture filtrate of the bacteria, the s-IAA-Lys and its corresponding a-acetyl derivative were synthesised by a key conjugative step of lAA with the N-a-CBZ-L-lysine /?-nitrophenylester. The same strategy was used to obtain the unnatural conjugated N-a-(indole-3-acetyl)-L-lysine (a-IAA-Lys) and its N-s-acetyl derivative (Ac-a-IAA-Lys) [15 and 16, Fig. (2)] [38]. Comparative bioassay for auxin activity on wheat coleoptiles proved that s-IAA-Lys and Ac-s-IAA-Lys significantly stimulated their growth, although their activity was less than lAA. No activity was shown by a-IAA-Lys and Aca-IAA-Lys [38]. Furthermore, from the acid crude culture extract of an olive strain of the same bacterium another product of the lAA metabolic pathway was isolated. It was identified by spectroscopic methods as the indole-3-

589 aldehyde [17, Fig. (2) and Table 1]. This represents the first isolation of 17 from culture filtrate of a bacterium whose hormonal production (lAA and/or cytokinins) has a decisive effect on its ability to be pathogen [39].The main auxins found in the acid organic extract of P. amygdali culture were lAA [30] and the corresponding methyl ester [12, Fig. (2) and Table 1], which had the same auxin activity as the free acid. Identification of 12, made essentially by spectroscopic methods and by comparison with the product obtained by lAA methylation with diazomethane, represents the first report of its occurrence as microbial metabolite [40]. NH(CH2)4- CH-COOH

10

CH2COOR

11 R=H

CH2CO

13 R=H

12 R=CH3

14 R=Ac

NH-CH-(CH2)4-NHR CH2CO

COOH

15 R=H 16 R=Ac

Fig. (2). Auxins (11-14,17) from Pseudomonas spp.

CHO

17

NHR

590 Mode of action of auxins and cytokinins

A TLC- and HPLC-based method was developed to quantitatively detect the cytokinins and auxins produced in the acid and alkaline extracts of culture filtrates from different P. amygdali strains. The lAA and cytokinin (/-Z and ^ZR) quantification was performed by an HPLC method developed using a C-18 reverse phase column and a linear gradient of MeCN-HiO (7:3, v/v) for the analysis of lAA, and 100 ^iM TEAB (triethylammonium bicarbonate adjusted at pH=7 with CO2) and MeCN-H20 (7:3, v/v) in steps 1 and 2 for the analysis of cytokinins. The presence and level of phytohormones produced by some strains were analysed with respect to their virulence on almond plants [41]. The same analytical methods were used to estimate the production of cytokinins and/or auxins by three wild strains of P. savastanoi and three phytohormone-deficient mutants. The pathogeneticity on olive and oleander plants of three wild-type strains of P. savastanoi (two from olive and one from oleander) was compared to those of three phytohormone-deficient mutants of oleander strains: laa'^/cytokinin", laa" /cytokinins^ and Iaa"/cytokinins'. Mutants not producing lAA only induced necrosis of the inoculated tissues or swellings on the stems attributed to cytokinin production accompanied by necrosis. By contrast, Iaa"^/cytokinins' mutant induced attenuated symptoms on stems and knots similar to those obtained with the parent strain on oleander leaves. Olive strains induced necrosis of oleander leaves and were virulent and avirulent, on olive and oleander stems, respectively. The wild oleander strain and its three mutants were all able to multiply in oleander leaves at similar rates, reaching the same final populations. By contrast, the two olive strains multiply poorly, reaching populations ca, lO^-fold lower [42]. These results confirm that expression of lAA-genes alone is sufficient to initiate the development of knots on oleander, while cytokinins are necessary for the frill expression of the disease symptoms (determining knot size). This finding also indicates that plant tissues (stems and leaves) react differently to various strains of the bacterium, and suggests that, besides phytohormones, other pathogenic factors could be involved in this host-pathogen interaction. The necrotic reaction of oleander leaves heavily inoculated with olive strains was interpreted as a possible form of hypersensitivity reaction [42].

591 Also ash strains, avirulent on both olive and oleander plants, were observed to produce in culture phytohormones belonging to auxin and cytokinin groups. Typical strains produced only small amounts of auxins, whereas relatively large amounts of the above phytohormones accumulated in the culture of an atypical strain. No cytokinins were isolated from cultures of typical strains. The metabolites, characterised by chromatographic (TLC and HPLC) and spectroscopic methods, were identified as the lAA and its methyl ester (the auxins), and the r-ZR, r'MeZR, diHZ and its 9-p-D-riboside [diHZR, 8, Fig. (1) and Table 1], which were the cytokinins found only in the atypical strain cultures. The diHZR was isolated for the first time as a phytohormone of this bacterium [43]. Isolates of P. savastanoi from ash were examined for their ability to produce phytohormones in culture and for pathogenicity in comparison with isolates from olive and oleander. The production level of the above auxins and cytokinins was evaluated using the TLC and HPLC methods cited above and correlated to their different virulence on the host plant. [44]. Nineteen isolates produced low levels of L\A and its methyl ester, but no cytokinin. In contrast, the remaining atypical isolate accumulates high levels of auxins and cytokins in culture, comparable to those of olive and oleander strains. Oleander isolates infected all three hosts, whereas those from olive caused symptoms on olive and ash only. All cultures were able to multiply in host plant tissues, but the growth rates and final population densities were correlated to the plant species inoculated and the host origin of the isolates. These results were consistent with those obtained by genetic investigation and pathogeneticity assays. In fact, 19 out of the 20 isolates caused disease only on ash, but the atypical strain caused knots on both ash and olive. Phytohormone production shown by ash, olive and oleander strains of P. savastanoi was in agreement with the type of symptoms: canker accompanied by wart-like excrescence on ash, and knots on olive and oleander. Furthermore, the pathogenic features of these isolates and, in particular, their growth patterns in the different host tissues, support previous evidence on the existence of three distinct pathovars in P. savastanoi [44].

592 Pseudomonas syringae/7v. papuhns plant growth regulators

Pseudomonas syringae pv. papulans is the causal agent of blister spots of apple {Mains pumila) and pear {Pyrus communis), a serious disease occurring in several areas of North America and Europe [45-47]. The disease is mainly characterised by dark brown blisters on fruit and tiny cankers on branches, but also midvein necrosis and distortion of leaves occasionally occur. Various P. syringae pathovars produce phytotoxins that in some cases are known to play an important role in the disease process. P. s, pv. papulans also produces phytotoxic metabolites in culture in vitro. Three weak phytotoxins, called papuline, ohydroxypapuline and papulinone [19, 20 and 21, Fig. (3) and Table 1] were isolated and characterised by spectroscopic (iH and l^C NMR and EI MS) and chemical methods. Papuline (19) proved to be the methyl ester of the p-phenyllactic acid with a iS-stereochemistry at C-2. This was also confirmed by identity of its chemical and physical properties with those of the methyl ester obtained by methylation with diazomethane of the (iS)-2-hydroxy-3-phenylpropanoic acid [18, Fig. (3)]. This structure was further confirmed on converting 19 into the corresponding 2-0acetyl derivative by standard reaction. o-Hydroxypapuline (20) was characterised as the meta-hydroxy-para-mXro derivative of papuline, and confirmed by the consistent spectral data of its diacetyl derivative prepared in standard conditions. Finally, the papulinone (21) proved to be a new P-lactone characterised as 4-( 1 -hydroxy-2-phenylethyl)-4-carbomethoxyethan-2one; this structure was confirmed by conversion of 21 into the corresponding y-butyrolactone [22, Fig. (3)] by acetylation carried out in standard conditions. The mechanism to explain this unexpected reaction, whose driving force is the basic properties of pyridine used as solvent and the high steric strain present in 21, was also hypothesised [48]. Compounds 19, 20 and 21 were indistinguishable by bioassay on bean or apple leaves. When 2-6 mg/ml of purified metabolites taken up in solution in citrate buffer (pH 6.0) were applied to both leaves, similar areas of necrosis resulted. At concentrations of less than 2 and 1 mg/ml only slight chlorosis and no toxic effects were observed [48]. Further investigation of the same organic extract allowed isolation of a new trisubstituted derivative of phenylacetic acid methyl ester. It was characterised as the methyl ester of 2',5'-dihydroxy-3'-nitrophenylacetic acid [23, Fig. (3) and Table 1] by spectral studies of the metabolite, and

593 the corresponding 2',5'-0,0-diacetyl derivative [49]. When assayed on pre-germinated tomato seedlings, 23 brought about a sHght reduction in shoot growth at lO""^ M and a sHght stimulation of rootless growth at 10"^ M. These findings confirmed the plant growth regulator activity of phenylacetic acid and its derivatives [50].

OH 3

2l

OH I CH2-CH-COOCH3

1

CH2-CH-COOR

HO,

K ^

O9N

20

18 R=H 19 R=CH3

6

H

OH COOCH.

OCOCH3

5 I

.CH2-CH4

3

1

" ^ ^

o

2

c

22

21 2

1

H2COOCH3

23

Fig. (3). Plant growth regulators (19-21, 23)fromPseudomonas syringae pv. paulans.

594

Moreover, papuline, o-hydroxynitropapuline and papulinone are structurally related to P-phenyllactic acid, which is well known as a plant growth regulator substance isolated from fungi and bacteria [51, 52]. Low molecular weight carboxylic acids, including the above cited phenylacetic acid, with phytotoxic activity against plant tissue, have been reported to accumulate in the culture of several bacteria and fungi [50, 53, 54]. However, not all the substances toxic to plant tissue can be considered phytotoxins. In fact, considering the weak phytotoxic activity observed, the isolated carboxylic acids probably play a minor role during the pathogenesis of microorganisms on their hosts. Therefore, at present, it cannot be excluded that the compounds 19, 20 and 21 isolated from P. s. pv. papulans have other detectable activity i.e. plant growth promoting or inhibiting activity, which is in some way related to symptom induction by the bacterium, just like certain carboxylic acids [55] or P-phenyllactic acid itself and the phenylacetic acid and its derivatives (including 23). Pseudomonas caryophylli toxins

Pseudomonas caryophylli, recently reclassified as Burkholderia caryophylli [56], is the causal agent of bacterial wilt of carnation (Dianthus caryophyllus L.). The vascular disease, first described by Jones in 1941 [57], causes yellowing and wilting of leaves and stems; in the later stage of development these symptoms are accompanied by basal stem rot. The severe economic losses for floriculture and specialised nurseries have prompted studies to clarify the mechanism of this pathogenesis. Holtzman and Thomas [58] reproduced the symptoms of bacterial wilt in carnation cuttings placed in culture filtrates of the pathogen. They correlated the wilting with the presence of thermostable and high molecular weight compounds in culture filtrates. Later, Kusumi et al [59] isolated from liquid culture of the pathogen three antibiotic substances active against P. syringae pv. phaseolicola. These lipophilic antibiotics, called caryonencins A, B, C, were characterised as extremely unstable poljmnsaturated hydroxy acids, hideed, they were characterised as an equilibrium mixture of stereoisomeric 7,9-diene-6-hydroxy1,13,15,17-tetrayneheptadecanoic acids. More recently, Lavermicocca et al. [60] working with a culture filtrate of a B. caryophylli strain (NCPPB2151) obtained from the National Collection of Plant Pathogenic Bacteria of Harpenden (United Kingdom), showed that this

595 microorganism produces stable hydrophilic metabolites with antibiotic and phytotoxic activity, which are certainly different from caryonencins. Extraction with «-butanol and dialysis experiments gave indications about the chemical nature of the bioactive metabolites. Purification of the culture filtrates by gel-filtration chromatography on a Bio-Gel P2 column gave a group of fractions that inhibited the growth of P. s. pv. phaseolicola and caused severe phytotoxic symptoms on host and nonhost test plants. The results of partial purification and preliminary chemical characterisation of bioactive metabolites produced by B. caryophylli indicated that these substances are polar molecules with a molecular weight of less than 1800 Da [60]. Comparing these structural features with the structures of the different phytotoxins and antibiotics, all characterised as lipodepsipeptides and produced by several phytopathogenic Pseudomonas species [61], it is possible to hypothesise that B. caryophylli produces bioactive metabolites belonging to the same chemical group. Research is now in progress to verify this hypothesis by purifying such metabolites for their chemical and biological characterisation. Pseudomonas ciccaronei bacteriocins

Strains of many bacterial species can produce proteinaceous antimicrobial substances called bacteriocins. Bactericidal specificity, restricted to species closely related to their producer, and chemical composition distinguish bacteriocins from other classic antibiotics [62-64]. Investigations of bacteriocin production by plant pathogens have found applications in epidemiological studies and in the typing of bacterial species [65-69]. Bacteriocins from Pseudomonas syringae pv. syringae were isolated and characterised with the aim of using them as biological control agents [70, 71]. Pseudomonas syringae pv. ciccaronei (strain NCPPB2355), isolated from carob tree leaf spots [72], was found to produce bacteriocin inhibition against P. savastanoi strains, the causal agent of olive knot disease (see above). Treatments with mytomycin C did not substantially increase the bacteriocin titre in culture. Purification of bacteriocin obtained by ammonium sulphate precipitation of culture supernatant fluid, membrane ultrafiltration, gel fihration and preparative PAGE, led to the isolation of a high molecular weight proteinaceous substance. The bacteriocin analysed by SDS-Page revealed three protein

596

bands with molecular weights of 76, 63 and 45 kDa, respectively. The bacteriocin was insensitive to heat and proteolytic enzymes, resistant to non-polar organic solvents and was active between pH 5.0-7.0. PlasmidDNA analysis of P. s. pv. ciccaronei revealed the presence of 18 plasmids; bacteriocin-negative variants could not be obtained by cure experiments [73]. Xanthomonas campestris pv. vitians phytotoxins

Bacterial leaf spots is a disease affecting lettuce plants caused by Xanthomonas campestris pv. vitians, recently reclassified as X. hortorum pv. vitians [74]. This disease does not normally produce severe losses [75], but outbreaks occur during periods of heavy rainfall [76]. The leaves of infected plants show translucent, water-soaked brown lesions that become dark with age. These types of symptoms, and the fact that the bacterium is generally not isolated from the chlorotic halos surrounding the dead tissues, suggest the involvement of toxic compounds in the disease developments. Two phytotoxic metabolites were isolated from the culture filtrates of X. h. pv. vitians, which were identified as 3-methylthiopropanoic and ^ra«5-3-methylthiopropenoic acid, using chemical and spectroscopic methods [77]. The two acids induced chlorosis on lettuce, while their infiltration on cabbage leaves did not produce any symptoms. The two metabolites showed a significant activity against the lettuce protoplast and low activity against those of cabbage [77]. The two acids are also produced by X. c pv. campestris, armoraciae, raphani, carotae and orizae [78, 79]. They probably accumulated as a result of biotransformation of methionine present in the medium [78-80]. These results indicate that presumably the abiUty to produce the two compounds is in common with many pathovars of X. campestris. The 3-nitropropanoic acid, structurally close to 3methylthiopropanoic acid (the two compounds differ only in the terminal group bonded at the C-3 of the propanoic acid chain), was recently isolated from the plant pathogenic fimgus Melanconis thelebola [81], and used as a bioherbicide against forest and spontaneous broadleafs such as Alnus rubra [82]. This suggests that the toxic activity of these acids is mostly associated with the aliphatic chain and not with the substituent group at C-3.

597 Polysaccharides

The involvement of polysaccharidic or glycopolysaccharidic substances in bacterial and fungal diseases has often been reported since Hodgson et al [83-86] but their effective action as phytotoxins is still to be clarified [85, 86]. Comparable damage is often produced by polysaccharides from nonpathogenic species to the test plant [87]. However, in several phytopathogenic bacterial species such as Agrobacterium, Clavibacter, Erwinia, Pseudomonas and Xanthomonas, production of polysaccharidic or glycopolysaccharidic substances proved to be associated with water soaking of leaves and/or wilting symptoms [85]. A role in plant colonisation by the bacterium was also observed [85]. Recently, extracellular polysaccharides from Xanthomonas campestris pv. vesicatoria were proved to induce phytotoxic effects (chlorosis, necrosis, electrolyte leakage) on the homologous host [88]. Not enough is known about the mechanism of action of these macromolecules. They commonly appear to act by interfering with water movement in plant tissues due to mechanical plugging of the vessels which leads to wilt symptoms. The phenomenon could well be related to molecule size and viscosity rather than to their structure [84], though some results on host specificity [89] and on viscosity interference [90] would suggest a possible different behaviour in some cases. Lipopolysaccharides

Lipopolysaccharides (LPS) are complex molecules of the cell wall Gram negative bacteria: they contain a polysaccharide (hydrophilic) tail and a lipid (hydrophobic) head, which is anchored in the outer membrane. The molecule consists of three segments: lipid A, core polysaccharide, and the antigenic 0-chain. The core polysaccharide is fiirther subdivided into an "inner" core and an "outer" core. The three segments differ in their composition, biosynthesis and biological fimction. Although this model of LPS was originally proposed for Salmonella spp., the basic features appear vaUd for most, if not all. Gram-negative bacteria [91]. The isolation and structure determination of phytopathogenic bacteria LPS, and in particular that of the corresponding 0-chain, have assumed great importance for the probable role of this molecule in the hostpathogen specificity. Furthermore, knowledge of the chemical structure

598

of LPS, together with the cultural behaviour and the biochemical and physiological properties, helps the taxonomic classification of the microorganisms [90]. Pseudomonas savastanoi LPS

P. savastanoi strain isolated from infected oleander plants (see above), proved to be virulent on both oleander and olive plants, while olive strain was shown to be virulent only on the host plant. The molecular basis of this host-pathogen specificity is poorly understood. Exopolysaccharides (EPS) and cell-wall LPS of phytopathogenic bacteria have been correlated with pathogenesis: its chemical structure and tridimensional conformation may play a role in the early infection process and in the determination of plant-bacteria interaction specificity [92]. The LPS of three strains of P. savastanoi were isolated and the structure of their Ochain portion were determined by chemical and NMR spectral analysis. For all the strains the 0-chain consists of a tetrasaccharide repeating unit of three a-L-Rha/? and one terminal non-reducing a-D-Fuc-/?3NAc. Two rhamnosyl residues are 3-linked and the third one 2,3-linked as reported [93]. ->3)-a-L-Rha/7-(l->2)- a-L-Rha;7-(l->3)- a-L-Rhaj^-(l-^

3

t

1 a-D-Fuc/73NAc

It is interesting to note the presence in this polysaccharide of 3-Nacetyl-3,6-dideoxygalactose (D-Fuc-/>3NAc), which is an uncommon monosaccharide identified by NMR spectroscopy and GC-MS analysis of its methylalditol acetate [92]. LPS from an olive strain, avirulent on oleander, attenuated the response of oleander plants inoculated with a homologous strain [93]. A conformational study of the 0-chain LPS has been recently carried out using a new NMR-molecular approach. The preliminary results ^

Amodeo, P.; Jourdan, F.; Strazzullo, G.; Forestier, C; Gorvel, J.P.; Evidente, A.; Motta, A Abstracts 10th European Carbohydrate Symposium, Galway, Ireland, July 11-16, 1999, p. 298.

599 suggest that the molecule assumes a helical structure, whose compactness is under investigation. The structural identity of the LPS O-chain of three different P. savastanoi strains and that contemporaneously reported for P. syringae pv. tomato [94], which is a pathogen of tomato, has prompted comparison with LPS of P. amygdali and P. ciccaronei (two other phytopathogenic bacteria: see above). The O-chain of the latter two bacteria consists of a tetrasaccharide repeating unit similarly to the three strains of P. savastanoi and the strain of P. s. pv. tomato [94]. LPS lipid residue favoured chemotaxonomic considerations classifying P. amygdali and P. s. pv. tomato in a chemotype IE, as found otherwise for the two other bacteria examined {P. savastanoi and P. ciccaronei) [95]. Burkholderia caryophylli LPS

Burkholderia caryophylli is the causal agent of bacterial wilt of carnation, which produces in vitro phytotoxic and antibiotic metabolites (see above), whose structure determination and role in the disease is now in progress. The lipopolysaccharides obtained from the cell-walls by phenol-water extraction were purified by gel-filtration and acid hydrolysed. The structure determination of the main product isolated from the hydrolysate was performed essentially by spectral (ID and 2D ^H and ^^C NMR, and FAB MS techniques) and chemical methods. This proved to be a novel 4-branched sugar called caryophyllose [24, Fig. (4)]. Its isolation was difficult due to its lability in acid conditions, which are normally used to separate the O-chain from LPS lipid. In fact, only by using very soft hydrolysing conditions were the caryophyllose and some of its oligosaccharides isolated. This 12-carbon sugar (24) contains a 3,6dideoxyhexopyranose ring, with a xylo configuration, branched at C-4 with a 1,3,4,5-tetrahydroxyhexyl side chain [96].The yersiniose A and B are the only two 4-C-branched sugars known until now [97, 98]. Both were isolated from bacteria and are 3,6-dideoxy-4-C-(l-hydroxyethyl)-Dxy/o-hexoses and differ in the configuration of the side chain. In addition, in a bacterial lipooligo-saccharide, a new branched 3,6-dideoxy sugar, which bears a y-chain at C-4, has been later identified [99]. Caryophyllose can be converted into a mixture of three glycosides by treatment with HCl/MeOH. The mixture consisted of a/p form of the methyl glycosides of 24, its isomeric methyl glycoside [25, Fig. (4)],

600

which was generated by the closure of the pyranosihc ring at C-T, and the intramolecular glycoside 26 [Fig. (4)]. OH H,C

OH

H.C OH

OH

OH 24

25

OHy^a HO^^H > H3C Al H O

1 \ ^ 3 '

OH 26

27

CHO -OH CH, 'CH9OR HO i^OR ^T H "

OH T 4'CH20H °VI

-'CH2

H

' CH2OR

^1 28

2'CH2 i'CH20H

-H

RO-

29

30

R=p-BrC6H4CO

R=/?-BrC6H4CO

^^\J^ HO-

-OH -H

CHo H-OH HH-

-OH -OH CH3 31

Fig. (4). Caryophyllose (24), its isomeric methyl and intramolecular glycosides (25 and 26) and some of its derivatives and degradation products (27-30).

The structure of these three glycosides was determined using ^H and ^^C NMR techniques while the relative configurations of carbinol carbons were deduced by measuring Nuclear Overhauser Effect (NOE) [100].The absolute configuration of caryophyllose was independently elucidated by

601 Mosher's [101] and Exciton Chiral Coupling [102] methods. Mosher's method is based on the analysis of the difference between the chemical shifts of the protons adjacent to the carbinolic centre measured in the corresponding (S)- and (jR)-MTPA (a-methoxy-a-trifluoropheylacetate) esters. Since this procedure cannot be applied when more hydroxy groups are closely located (the phenyl rings of the Mosher ester can interfere with each other [100]), this procedure was applied to the 4,r:3*,5'di-(9isopropyliden derivative [27, Fig. (4)] of 24 methyl glycoside, whose esterifiable hydroxyl groups at 2 and 4' positions are sufficiently distant. The reaction was performed on the a-isomer since it was available in larger amounts. On the basis of Mosher's argument concerning the chemical shifts of the pertinent protons of (5)-MTPA-27 and (i?)-MTPA27, C-4' was assigned the R configuration. Such a stereochemistry was confirmed by using the absolute method of Exciton Chiral Coupling applied to the hydrolysis products of the derivative 28 [Fig. (4)], obtained by the Smith degradation of the LPS fraction, which consists in the oxidation of the latter with NaI04, followed by NaBH4 reduction. The two products of 28 hydrolysis, by treatment with /?-bromobenzoyl chloride, were converted into the triester 29 [Fig. (4)] and r,3',4'butanetriol triester 30 [Fig. (4)], respectively. These two derivatives were subjected to CD analysis. On the basis of the above results, structure 24 [Fig. (4)] can be defined as 3,6-dideoxy-4-C-(D-(3/^rc>-l,3,4,5tetrahydroxy)-D-xy/(9-hexopyranose, which is the ring form assumed in the B. caryophylli LPS by the novel branched sugar 3,6,10-trideoxy-4-C[(i?)-l-hydroxyethyl]-D-^rv/r(9-D-gw/o-decose 31 [Fig. (4)] which was called caryophyllose [100]. This unusual monosaccharide is the repeating unit of the 0-chain of the main LPS oi B. caryophylli, called caryophyllan [Fig. (5)]. It proved to be a linear homopolysaccharide built up of (1^7) a-caryophyllose, as deduced from GC-MS analysis of the products of hydrolysis, reduction and acetylation of the methylated LPS.

CH, i Vr^-Ar-V-vCH, . /^_:r^^-7Vv-vcH:

re^

flipide A!—[CORE

'V:;V^cH3 -^^^Y-V^cH,) „

Fig. (5). The caryophyllan 0-chain.

602

In addition, the product [28, Fig. (4)] obtained as described above by the oxidation/reduction of the LPS represents further evidence of the interglycosidic bond as it remained unchanged during degradation. Finally, the a-configuration of this bond was deduced from the coupling constant ^Ji^2 of 3.4 Hz measured for the doublet assigned to the anomeric proton of caryophyllose in the ^H NMR spectrum of the LPS fraction and by the nature of the products of its degradation [103]. A further investigation of the same lipopolysaccharide fraction and the isolation of the oligosaccharides generated by its hydrolysis allowed further definition of the configuration of the caryophyllan interglycosidic linkages. By partial hydrolysis of LPS, a mixture of two disaccharides (one a- and the other P-linked) was obtained in a 3:2 ratio. This nonsignificant ratio could be a consequence of the different hydrolysis rate of the a- and P-glycosidic linkage. When the native LPS was degraded with NaI04, a reaction that did not hydrolyse the glycosidic bond, two glycosides generated from a P- and an a-linkage were obtained in a 1:8 ratio. The two glycosides and the two disaccharides [Fig. (6)] obtained by caryophyllan were completely characterised using one- and twodimensional NMR experiments. From these findings the 11% occurrence of P-linkages in the caryophyllan structure has been evidenced [104]. OH

H3C-

H.C

Fig. (6). The two glycosides obtained by caryophyllan degradation.

Later, the LPS fraction of B. caryophylli was observed to contain a minor component, whose 0-chain was a homopolysaccharide made of (l->7) linked P-caryose, another new unusual monosaccharide characterised as below described as the 4,8-cyclo-3,9-dideoxy-L-^r);rroD-/(io-nonose [32, Fig. (7)]. When the ^H and ^^C NMR spectra of the hydrolysed LPS fraction were recorded, the presence of a further anomeric signal, not attributable to the completely hydrolysed caryophyllan, was observed. The 0-chain structure of the new LPS, called caryan, was determined by applying the

603 same techniques used for caryophyllan. In particular, the p-configuration for the interglycosidic hnkage between the different caryose units was deduced from the value of 170 Hz measured for the Vc,H of the anomeric carbons. This bond proved to be of (l->7) type by analysis of the NMR, GC and FAB MS spectra of the product obtained by the total permethylation of the polysaccharide fraction containing caryose [103].

32

Fig. (7). Caryose (32).

The latter fraction was subjected to methanolysis and the product [33, Fig. (8)] analysed using one- and two-dimensional NMR techniques. The caryose proved to possess a uncommon new carbocyclic structure with a fiirane ring spirofused at C-4. The relative configuration of all carbinolic hydroxy groups was deduced from the evidence of several NOEexperiments. When the methylglycoside 33 was subjected to acid hydrolysis a mixture of three compounds was obtained. They proved to be the two anomeric furanose forms [34a and 34b, Fig. (8)] of the hydrolysed 33 and the pyranosidic isomer in its a-configuration [35, Fig. (8)]. On the basis of these results the structure 32 [Fig. (7)] is defined for this novel monosaccharide [105]. The absolute configuration of caryose was determined applying the Exciton Chiral CoupUng to 36 [Fig. (8)], which is the di-O-pbromobenzoyl derivative, obtained as the main product of the isopropylidenation of the monosaccharide equilibrium mixture (34a, 34b and 35) followed by esterification with /^-bromobenzoyl chloride. The results obtained have suggested for caryose the absolute configuration

604 depicted for 32 in Fig. (7). Caryose was named, in accordance with the systematic nomenclature used for carbohydrate, as the 4,8-cyclo-3,9dideoxy-L-erytrO'D-ido-nonosQ, derived from the formal cleavage of the C-4/C-8 bond by the addition of two hydrogens [105].

HOy^W.
|2 H

HO O H " 33

36

34a

Ri=OH, R2=H

34b

Ri=H, R2=0H

35

R=/7-BrC6H4CO

Fig. (8). The spirofused furanosidic and pyranosidic forms (34a and 34b and 35) of caryose and some of its derivatives (33 and 36).

The structural analysis of both 0-specific polysaccharides was performed directly on the LPS fraction, as caryophyllan was immediately degraded by mild acid treatment. Thus, the 0-antigenic polysaccharides were never isolated. Moreover, initial attempts to separate both LPS fractions using gel-permeation chromatography failed. However, using chromatographic fractionation on a deoxycholate column in the conditions described by Peterson and McGroarty [106], a separation of LPS components was possible. Besides the expected caryophyllan and caryan and a fraction containing a polj^fructose, the fourth components proved be to a novel polysaccharide that had been coextracted with LPS.

605 Its structure was determined using one- and two-dimensional NMR spectroscopy and is as follow [107]. ->6)-a-D-GlcH 1 -> 1 )-p-D-Fru/-(2-^

Fructose occurs rarely in the bacterial polysaccharide. It has been found in the LPS of several Vibrio species [108] and in the K4 [109] and Kll [110] capsular antigens of Escherichia coli. In all cases, fructose appears as a terminal residue. In several plant species, fructans consisting of P-D-fructofuranosyl units are present as important storage polymers. The structure of the polysaccharide isolated from B. caryophylli is different from those of the above polysaccharides; however, as levan can be isolated as bacterial exopolysaccharides, this polysaccharide may be a side-product of levan biosynthesis. Burkholderia (Pseudomonas) cepacia LPS

Burkholderia {Pseudomonas) cepacia, a non-fermentative, Gramnegative, motile ubiquitous bacterium, was first reported as a phytopathogenic bacterium by Burkholder [111], Subsequently, it has been reported as a biological control agent of plant-disease [112-115] and as an opportunistic human pathogen [116] isolated from cystic fibrosis patients and contaminated medical devices and chemical solution employed in hospital practices [117, 118]. To search for B. cepacia strains that may safely be used as biocontrol agents, bacterium strains recovered from soil, water and clinical environments were previously distinguished according to their cultural, biochemical, physiological and molecular differences. Other approaches relied on LPS components of the bacterial cells [119]. In particular, these studies were carried out on the LPS extract from B. cepacia strain PVFi-5A, isolated from roots of healthy tomato plants and currently used as a biocontrol agent for bacterial and fimgal diseases of tomato plants. On the basis of chemical degradation methods and one- and two-dimensional ^H and ^^C NMR experiments, the structure established for the 0-deacetylated repeating unit of the 0-chain of the main B. cepacia strain PVFi-5A LPS is as follows [120]. ->4)-P-D-Gal/7NAc-( 1 ->3)-a-D-Gal/7-( 1 ->6)-a-D-Glc;7NAc-( 1 -^

606 Xanthomonas sp. LPS

Xanthomonas hortorum pv. vitians, which is the causal agent of leafspots of lettuce (see above), synthesises the 3-methylthiopropanoic and the corresponding dehydroderivative as phytotoxic metabolites. The LPS produced by this microorganism were extracted, and the lipid A and the polysaccharide moieties prepared according standard procedure [121]. Both the lipid A and the 0-chain LPS were characterised using chemical and spectroscopic methods. Main sugar components are rhamnose and 3N-acetyl-3,6-dideoxy-galactose that are thought to build up the Ospecific polysaccharide. Other sugars are mannose, glucose, 6deoxygalactose and galacturonic acid, which should be core region constituents, and glucosamine, which provided the carbohydrate backbone of lipid A. The LPS contains several phosphate groups. The main fatty acids in the lipid A are C10:0, 3-OH-C10:0 and 3-OH-C12:0, The last is the only amide-linked fatty acid. Small amounts of C8:0 and CI 1:0 fatty acids were determined [122]. Strawberry is attacked by several pathogens including X. fragariae, the agent of strawberry angular spot disease, causing serious damage to this crop [123]. Recently a new bacterial disease of strawberry has been reported [124]. The observed symptoms, which occur mainly on leaves, differ both from those caused by X. fragariae and by plant pathogenic fungi, causal agent of leaf disease on strawberry. Preliminary investigations of the characterisation of the bacterium, consistently isolated from lesions, showed that on the basis of the pathogenic and biochemical features it may be ascribed to X. campestris. Since it has been suggested [124] that X. gragariae and X. campestris may occur in combination on the same plant/leaf, it seemed of interest to gain further information on this pathogenic bacterium. Therefore, combining both chemical and spectroscopic data, the structure of the LPS 0-chain of this strain oiX. campestris was determined and is as follows [125]: P-D-Xyl/7 1

i 2 ^2)-a-L-Rhap-( 1 -^3)-a-L-Rhap-( 1 ^2)-a-L-Rhap-( 1 ->3)-a-L-Rha^-( 1 ->3)-a-L-Rhap-( 1 -^ 4

t 1 p-D-Xyl/7

607 Exopolysaccharides

The idea that exopolysaccharides (EPS) produced by several plant pathogenic bacteria could be involved in the pathogenic processes and or/the saprophytic or epiphytic phases of the life cycle appears to be generally accepted. Copious EPS production is often associated with increased virulence [89]. It is conceivable that the EPSs prevent bacterial elicitors of host-defence responses from reaching the plant and may inhibit deleterious adherence during infection, thus "maintaining (in both cases) a compatible interaction" [89].

P. savastanoi EPS

As described above, the structure of the 0-chains of P. savastanoi LPS, isolated from olive and oleander infected plant tissues, proved to be identical. Furthermore, the fact that the same LPS 0-chain occurs in P. s. pv. tomato, which is pathogenic for tomato, suggests that the host specificity of these strains was not correlated with the chemical structure of their LPS 0-chains. Therefore, an investigation was also carried out to characterise the EPS produced by the two strains of P. savastanoi, one virulent on olive plants but avirulent on oleander, and the other virulent on both plants. Their EPS fractions were found to be similar when the two strains were grown on the same culture medium. Depending on the medium employed, the EPS fractions consisted of alginates, of glucans and lipopolysaccharides, or of heteropolysaccharides, the latter consisting of fiicose, galactose, N-acetyl-2-deoxy-2-aminoglucose, and N-acetyl-2deoxy-2-aminogalactose. Some EPS fractions were tested on watersoaking induction and on the inhibition of knot-formation on bean and oleander plants, respectively. The EPSs produced by P. savastanoi strain from olive did not interfere with knot induction on oleander by homologous strain [126]. A further investigation was carried out on the EPSs from oleander strain grown on the King's B medium [127]. As proved by preliminary results, P. savastanoi in these conditions produces a complex mixture of polysaccharides that were partially purified by gel-filtration chromatography. The main fractions were characterised by chemical and spectroscopic methods. The monosaccharide components were fucose, galactose, N-acetyl-2-deoxy-2-aminogalactose and N-acetyl-2-deoxy-2-

608 aminoglucose in the ratio 1.0:2.0:1.2:2.2. Some oligosaccharides of the acid hydrolysed mixture were isolated and identified. Although the high complexity of these EPS fraction precluded its complete arrangement, the elucidated structural features suggested a very interesting polysaccharide structure. Lideed, the type of sugar components and the presence of some pecuHar substructures, namely 2,3-P-D-Gal/7 and 3,6-P-D-Galj!?, mimic the structural situation found in human-blood groups [128-130] or human neuroblastoma cell oHgosaccharides[131]2. Amaryllidaceae alkaloids Plant species belonging to the Amaryllidaceae family are widely distributed in several countries in the world. They are also cultivated as ornamental plants for their beautiful flowers and for the production of volatile oil. The study of Amaryllidaceae alkaloids began in 1877 with the isolation of lycorine [37, Fig. (9) and Table 2] from Narcissus pseudonarcissus [9], and the interest around this group of naturally occurring compounds has increased in time because of their effective antitumoral and antiviral activities. The hundreds of new alkaloids isolated from different parts and in different vegetative phases of ca. 150 species belonging to 36 genera can be grouped into 12 distinct ring-types [14]. The advances made in the isolation and chemical and biological characterisation of such alkaloids have been extensively reviewed [9-16]. Lycorine is a pyrrolo[Je]phenathridine ring-type alkaloid extracted from different Amarylidaceae genera, whose structure was firstly determined by Nagakawa et al in 1956 [132]. However, considering the increasing interest generated by lycorine in the last few decades and in particular the ability of 37 to inhibit ascorbic acid (AA) synthesis in vivo [133], we further investigated the chemical and biological aspects of this interesting alkaloid.

^Barone, G.; Corsaro, M.M.; Evidente, A.; Parrilli, M.; Surico, G.; Carbohydr Polymer, 1999 submitted.

609

37R=H

39 Ri=OAc, R2=R3=OH, R4=OCH3

38 R=CH3

40 Ri=R2=R4=OH, R3=OCH3 41 Ri=R2=0H, R3=R4=OCH3

4V---OCH3

42

43 R=OAc 44 R=OH

Fig. (9). Lycorine (37) and minor alkaloids (38, 39, 42, 43 and 44) extracted from bulbs of Stembergia lutea Ker Gawl.

Lycorine type alkaloids

The effect of lycorine on several important physiological processes (inhibition of growth and cell division in higher plants, algae and yeasts [134]; inhibition at low concentration (10"^ M) of AA biosynthesis [133]; prevention of cyanide-insensitive respiration [135]) has made this substance a valuable tool for studying a number of biological effects. Thus, investigation of how lycorine acts of in vivo and in vitro are actively pursued [136].

610 TABLE 2.

Alkaloids from Amaryllidaceae Species

1 Alkaloid

Plant

Molecular

lUPAC name

formula 2,4,5,7,1 lb,l lc-Hexahydro-l//-[l,3]-dioxolo Deacetyllutessine

S. lutea

C17H19NO5

[4,5-j]pyrrolo[3,2,l-rfe]phenanthridine4-methoxy-l,2-diol

1 9-0-Demethylhomolycorine

Lycorenan-7-one, 9-hydroxy-lO-methoxyN. tazetta

C,7Hi9N04

1 -methyl 3,4,1 Ob, 11 c-Tetrahydro-6//-[ 1,3]-dioxolo-

11-Hydroxyvittatine

S. lutea

C,6Hi7N04

Hippamine

S. lutea

c,,H,,m,

5,1 Ob-ethanophenanthridine-3,11 -dihydroxy 2,4,5,7,1 lb,l lc-Hexahydro-l//-[l,3]-dioxolo [4,5-y]pyrrolo[3,2,l -^ejphenanthridine2-methoxy-l-ol 2,4,5,7,1 lb,l lc-Hexahydro-l//-[l,3]-dioxolo

Lutessine

S. lutea

C,9H2lN06

[4,5-y]pyiTolo[3,2,l -fi^ejphenanthridine1 -acetoxy-4-methoxy-2-ol 2,4,5,7,1 lb,l lc-Hexahydro-l//-[l,3]-dioxolo

Lycorine

S. lutea

Cl6H,7N04

[4,5-7]pyrrolo[3,2,l -^e]phenanthridine-l ,2-diol 3,4,4a,5-Tetrahydro-[l,3]-Dioxolo[3,2,l-Je]

Narciclasine

S. lutea

Narciclasine-4-O-p -D-glucopyranoside

P. maritimum C20H23NO12

CMHBNOV

phenanthridine-6(2//)-one-2,3,4,7-tetrahydroxy 3,4,4a,5-Tetrahydro-[l,3]-Dioxolo[3,2,l-^e] phenanthridine-6(2//)-one-4-0-3-D-glucopyranoside 2,3,7-trihydroxy Lycorenan-7-one, 4,12-dihydro-5-hydroxy-

Nobilisitine A

C. nobilis

C,7H,9N05

1

1 -methyl-9,10-[methylenebis(oxy)] Lycorenan-7-one, 4,12-dihydro-5-(3-hydroxy- 1

Nobilisitine B

C. nobilis

C21H25N07

1 -oxobutoxy)-1 -methyl-9,10-[methylene bis(oxy)] 2,4,5,7,1 lb,l lc-Hexahydro-l//-pyiTolo[3,2,l

Stembergine

S. lutea

C,8H2,N05

-c^e]phenanthridine-1 -Acetoxy-10-methoxy2,9-diol 4,5-Dihydro-2-hydroxy[ 1,3]-dioxolo[4,5-y]

Ungeremine

P. maritimum C , 6 H „ N 0 3

pyrrolo[3,2,l-^e]phenanthridinium hydroxide inner salt 4,5-Dihydro-2,10-dihydroxy-9-methoxy[4,5-y ]

Zefbetaine

P. maritimum C,6H,3N03

pyrrolo[3,2,l-^e]phenanthridinium hydroxide inner salt

611

New procedures for the extraction and rapid analysis oflycorine

The above studies have required large amounts of pure alkaloid (37). Therefore a new purification method was developed to extract lycorine from bulbs of Sternbergia lutea Ker Gawl, the wild Amaryllidaceae collected during the withering period near Bari, Italy, and used as starting material. The extraction of alkaloids is carried out with dilute (1%) sulphuric acid, and then lycorine was precipitated with concentrated (12 N) NaOH and crystallised with ethanol as white prisms from the crude cake. This method was more straightforward and convenient than those previously described for extracting lycorine as a free base with different organic solvents. Furthermore, the new method allowed a higher yield to be obtained (1-1.4%) compared with those obtained from other procedures [137]. Furthermore, a new method was developed for lycorine analysis (37) by HPLC. The method is based on the appUcation of a sulphate salt of the alkaloid to a C-18 reverse phase column and subsequent elution with a mixture (47:53 v/v) of acetonitrile and diluted (0.01 M) ammonium carbonate at a flow rate of 2 ml/min. The method was more rapid and accurate with respect to those based on multiple-step purification of the alkaloid and proved to be suitable for the analysis of 37 in crude extracts, as the acid extracts of bulbs and leaves oiS. lutea. The results of this latter investigation supported the hypothesis that lycorine is biosynthesised in the leaves and then translocated in the bulbs of the plant. The same method was used to appreciate the yield of lycorine in the different extraction procedures. Finally, the HPLC method allowed us to estimate a lycorine yield of 85% obtained using the described extraction process, while the missing part was found in the mother liquor of alkaline precipitation [138]. Structure-activity relationships

Studies on structure-activity relationships in this family of alkaloids were feasible as a large number of lycorine derivatives had been prepared from 37 and several naturally occurring Amarylllidaceae alkaloids were available. In an initial experiment, twenty-three lycorine derivatives and naturally occurring alkaloids, structurally related to 37, were tested for their ability to inhibit AA biosynthesis in potato tubers. The following relationships between the structure modification and activity were

612 observed: i) cleavage of the acetalic bonds on the dioxole ring had no effect on activity; ii) derivatives with a methoxy group on C-8 (A-ring) were inactive; iii) oxidation of NCH2-7 to an amide group (B-ring) caused loss of activity; iv) modification of the C/D ring junction had no effect on activity when the D ring assumed a P-configuration, whereas a great decrease in activity was observed when the ring assumed an aconfiguration; v) selective or complete acetylation of hydroxy groups of the C ring, epimerisation or oxidation of the hydroxy group at C-2 led to a loss of activity; vi) a compound with a double bond located in the 1,2position showed almost identical activity to lycorine; vii) stereoselective hydrogenation of the double bond of the C ring induced a considerable increase in activity; viii) protonation of the nitrogen atom had no effect on activity. These results showed that the lycorine structural features needed to preserve the inhibitory activity of the AA biosynthesis are intact A and B rings, a (J-configuration of the D ring when the C/D ring junction changes and the presence of a nucleophilic site between C-1 and C-2 of the C ring [139]. Subsequently, several synthetic lycorine derivatives, five other Amaryllidaceae alkaloids and narciclasine [45, Fig. (10) and Table 2], which is a strongly antimitotic [140] neutral metabolite extracted from various AmaryUidaceae species [141] but structurally related to 37, were assayed for their ability to inhibit the AA biosynthesis. The highest potency observed was displayed by narciclasine followed by compounds with an aromatised C-ring. Derivatives modified at C-1 and/or C-2 of Cring were inactive, while the compound with a double bond between these position is a weak inhibitor. Also lutessine [43, Fig. (9) and Table 2] and its deacetyl derivative [44, Fig. (9) and Table 2], isolated as minor alkaloids from S. lutea, having a methoxy group bonded at C-4 of the Dring, appeared completely inactive. These results confirm that the presence of an appropriate substituted C ring is a necessary requirement for optimal 'response-triggering' contact between the lycorine derivatives and the specific receptor. Functional groups jutting out from the a-side of the molecule do not allow a good fit with the binding sites [142]. Studies of the effect of lycorine on the AA biosynthesis are still in progress. In vivo conversion of L-galactono-y-lactone to ascorbate is being carried out in all of the different nine plants investigated.

613

45

46

Fig. (10). Narciclasine and its 4-0-p-D-glucopyranoside (45 and 46) extracted from bulbs and leaves of Pancratium maritimum L.

The enzyme catalysing this reaction, i.e. L-galactono-y-lactone dehydrogenase, which appears to be localised in the mitochondrial membrane, is strongly inhibited by lycorine: 5 yM lycorine almost completely inhibits the activity of the enzyme. The alkaloid does not affect the activity of the ascorbate free radical reductase, dehydrogenase reductase or ascorbate peroxidase. These results confirm that lycorine is a specific inhibitor of ascorbate biosynthesis in plants, and consequently, that it can be a usefiil tool to improve our understanding of the ascorbate metaboHsm in ascorbate synthesising organisms [143]. NMR studies of lycorine and of some its

derivatives

The basis of this investigation was the attribution of all the ^H and ^^C systems of 37 and a-dihydrolycorine, which is the only derivative showing an activity slightly higher than that of lycorine. The spectra were recorded in CD3OD-CD3COOD (3:1, v/v) at 40°C, the only conditions in which a good resolution could be obtained. Application of NMDR (Nuclear Magnetic Double Resonance) and ^H NOE-Difference experiments and the information derived from the spectra of l-(9-acetyl-, 2-(9-acetyl-, l,2-(9,0-diacetyl-, a-dihydro-lycorine derivatives and the lactame of the latter, allowed complete assignment of the ^H NMR spectrum of 37 [144]. The PND (Proton Noise Decoupling), SFORD (Single Frequency Off-Resonance Decoupling) and SFD (Selective

614 Frequency Decoupling) ^^C NMR spectra, recorded in the same conditions, allowed chemical shift identification of the protonated carbons, while the quaternary aromatic carbons were identified through 'NOE-Time Dependence' measurements. These assignments were further supported by the l^C NMR data of the a-dihydrolycorine and the corresponding lactame, which were consistent with those reported in literature [145] for other structurally related naturally occurring compounds [144]. Lycorine related alkaloids from Sternbergia lutea Ker Gawl and other Amaryllidaceae

Together with lycorine five other alkaloids, well-kown metabolites of this plant family, were found in the bulbs of S, lutea Ker Gawl as tazettine, galanthine, galanthamine, hippeastrine and emanthidine [146]. Moreover, from the lycorine mother liquors another alkaloid, present in very low concentrations, was isolated and identified by spectroscopic analysis as hippamine [38, Fig. (9) and Table 2] [147], which is rarely found as a naturally occurring compound. It was previously isolated, at a level comparable to that found in S. lutea [147], only in an unidentified Hippeastrum species [148]. Further examination of the acid extract of the bulbs of the same plant {S. lutea) allowed isolation of another metabolite named stembergine [39, Fig. (9) and Table 2]. To isolate this alkaloid the extraction process described above, was slightly modified to avoid the total deacetylation of stembergine in the strong alkaline conditions used to precipitate lycorine. Stembergine, whose stmcture was determined by spectroscopic techniques (essentially ^H and ^^C NMR and EMS), proved to be the 1O-acetyl-isopseudolycorine. It differed from pseudolycorine [40, Fig. (9)], a lycorine-type alkaloid first isolated fi-om Lycoris radiata [149], in the presence of the acetyl group at C-1 and in the reversed position of the methoxy and hydroxy groups at C-9 and C-10 of the A ring. This stmcture was confirmed by iH NOE-difference NMR spectra and chemical work. In fact, the presence of a phenolic group in 39, which had been suggested by its typical reaction with a FeCls solution, was confirmed by converting stembergine into the corresponding 9-0-(2,4dinitro)phenyl derivative by treatment with an ethanolic solution of 1fluoro-2,4-dinitrobenzene according to the Sanger procedure [150]. Furthermore, the physical and spectroscopic properties of the

615 deacetylstembergine were different from those of pseudolycorine, and those of the diacetyl derivative of 39 differed from those of triacetylpseudolycorine. Finally, by reaction with diazomethane both stembergine and pseudolycorine were converted into methylpsedolycorine [41, Fig. (9)] as 39 was deactylated for the KOH traces present into the diazomethane solution [151]. From the same organic extract a new alkaloid and the 11hydroxyvittatine [42, Fig. (9) and Table 2] were isolated. The latter, belonging to a crynine-type alkaloids [11, 14], was isolated for the first time from S. lutea [152], but was already known as a metabolite of Pancratium maritimum and Rodophilia bifida [11], two other Amaryllidaceae species. The new alkaloid, characterised using spectroscopic methods (essentially ^H and ^^C NMR, compared to those 1-0-acetyllycorine, and EI MS), proved to be the l-O-acetyl-4methoxylycorine [43, Fig. (9)], which represents the first lycorine-type alkaloid having a substituent group on the D ring. The a-configuration of the methoxy group at C-4 was deduced by accurate inspection of the longrange constants measured for the coupling of both H-1 and H-2 with H-4, and the allylic and vicinal coupling of the latter with H-3 and the two protons of H2C-5, respectively. This structure was confirmed by hydrolysing 43 into the corresponding deacetyl derivative [44, Fig. (9)], which was also isolated as a naturally occurring compound from the organic extract of the same plant, and by converting it into the 2-0-acetyl derivative by standard acetylation [153]. The isolation of new alkaloids from S. lutea have prompted similar studies on the alkaloids of other species of Amaryllidaceae. Two oxyphenantridinium alkaloids, correlated with lycorine, and a new nonbasic metabolite were isolated from bulbs and leaves of Pancratium maritimum L., a wild species collected on the northern coast of Egypt, together with 14 already known alkaloids found in other Amaryllidaceae species [154-157]. The new metabolite proved to be a glucosiloxy phenol [46, Fig. (10) and Table 2]. Structural investigation showed a close correlation with narciclasine [45, Fig. (10)], which is the non-basic metabolite structurally related to 37, with the interesting biological activities cited above [140, 141]. The need to have a reasonable amount of narciclasine to correlate its spectroscopic and biological properties to those of lycorine and other Amaryllidaceae alkaloids, required a new method for its extraction from bulbs of S. lutea in crystalline form and satisfactory yield [158]. Furthermore, this was the occasion to confirm 44

616 as a metabolite of S. lutea, which is a result previously reported only on the basis of chromatographic evidence [159]. An accurate NMR investigation was carried out on the crystalline narciclasine which allowed complete assignment of ^H and ^^C NMR spectra in terms of chemical shifts and coupling constants [158]. These findings favoured the characterisation of the new glucosiloxy phenol (46) correlated to 45 and extracted from P. maritimum. The latter, characterised using the same spectroscopic techniques, was defined as the 4-0-P-glucopyranoside of narciclasine [46, Fig. (10)]. Its structure was confirmed by preparing the corresponding heptacetyl derivative and acid hydrolysing it into narciclasine and D-glucose. Moreover, the p-configuration of the glycosidic bond was deduced from typical axial-axial value of the coupling constant (/r,2'=7.9 Hz) measured for the anomeric proton [160]. When assayed for antitumoral and mycotoxic activities on potato tubers and brine shrimp larvae {A. salina L.), respectively, 45 showed a significant activity comparable to that of narciclasine [160]. Subsequently, from the same extract we isolated two betaine-type alkaloids. Spectroscopic investigation, carried out using essentially ^H and ^^C NMR and FABMS, allowed their identification as ungeremine and zefbetaine [47 and 48, Fig. (11) and Table 2], already known as metabolites of other Amaryllidaceae with cytotoxic and antibiotic activities [161, 162, 163]. Ungeremine was earlier obtained from Ungernia minor [164], Crinum Americanum [165], C asiaticum [166] and Zephyranthesflava[161]. From this latter zefbetaine was previously isolated the but its structure was wrongly reported as /5o-zeft)etaine [49, Fig. (11)] [161]. Li fact, the physical and spectroscopical data described for the natural metabolite isolated from Z. flava [161] were very similar to those measured for 48 and for the product obtained by Se02 oxidation of pseudolycorine [40, Fig. (9)], by using the same conditions to oxidise lycorine to ungeremine [167]. By contrast^ the same data differed from those of its isomer, named z5o-zeft)etaine [49, Fig. (11)], which is a nonnaturally occurring alkaloid and was prepared from stembergine [39, Fig. (9)] by applying the same reaction [168]. These resuhs fiirther confirmed the structure of stembergine and allowed assignment of the correct structure to zefbetaine [168].

617

47

48 Ri=CH3, R2=H 49 Ri=H, R2=CH3

,

^

50

Fig. (11). Ungeremine and zefbetaine (47 and 48) extracted from bulbs and leaves of Pancratium maritimum L. and the products (47 and 50) of lycorine microbial degradation.

Microbial degradation of lycorine

The ability of lycorine to inhibit the AA biosynthesis in higher plants has stimulated some considerations on its environmental impact. Therefore, investigations were carried out on the effect on lycorine on microorganism growth. Of several strains isolated from soil only a Pseudomonas sp. strain (ITM 311), from the S. lutea rhizosphere, was able to actively degrade lycorine, which was used as a carbon source in a minimal medium growth. Within 24-48 hours, lycorine was transformed into 3 phenanthridinium derivatives. The transformation consisted in the oxidation of lycorine with the aromatisation of the C ring, and the conversion of the methylene group H2C(7)-N into an azomethyne group. Two of these metabolites were identified by spectroscopic investigation (iH and ^^C NMR and FAB MS) as the ungeremine, the oxyphenanthridinium alkaloid isolated from P. maritimum (see above) and the anhydrolycorinium chloride [50, Fig. (11)]. They were identical to the product obtained by oxidation of lycorine, respectively with Se20 [167], and POCI3 [169]. The third degradation product of lycorine was

618 probably the diyhroderivative of ungeremine. All attempts to isolate it failed because its well-known instability [167]. In fact, this metabolite easily reverts to ungeremine [163]. Compared to lycorine and some of its derivatives with an aromatised C ring and narciclasine, when assayed on Corynebacterium fescians, ungeremine and anhydrolycorinium chloride (47 and 50) showed higher antibiotic activity. It seems to be associated with the aromatisation of the C ring but it is unaffected by the oxidation of the H2C(7)-N group [163]. The ability of lycorine to affect the growth of some yeasts was subsequently evaluated. Lycorine (37) stimulated budding in Cryptococcus terreus, increasing the dry weight and the number of cells. By contrast^ C dimennae was not affected by lycorine, but its growth was stimulated by the filtered medium lycorine treated cultures of C. terreus [170, 171]. Identical effects were observed on the growth of C magnus. Further investigations suggested that lycorine was degraded into two metabolites, one of which was probably the anhydrolycorinium chloride [50, Fig. (11)]. Finally, lycorine inhibited the growth of rho"^, mit" and rho" strains of Saccharomyces cerevisiae, while rho° strains (devoid of mitochondrial DNA) were resistant to high concentrations of the alkaloid. Total protein synthesis and mitochondrial protein synthesis are only slightly inhibited by lycorine. It had, however, an inhibitory effect on DNA and RNA [172]. Lycorenine-type alkaloids

Investigation of the phenanthridine alkaloids was extended to other Amaryllidaceae widely grown in Egypt. In particular, we studied the alkaloids containing the [2]benzopyrano [3,4g]indole ring system. Narcissus tazetta alkaloids

Narcissus tazetta L. is also cultivated as an ornamental plant and for the production of volatile oil. From this species a number of alkaloids were isolated [10-12, 14] which showed great interest because of their effective antitumoral [173] and antiviral activities, particularly against choriomeningitis virus [174]. Lycorine, homolycorine and tazettine were

619 all isolated from the wild plant [175]. A further alkaloid was isolated from the ethanolic extract of the bulbs of the same plant. It proved to be, by spectroscopic investigation (essentially ^H and ^^C NMR and EIMS) the 9-0-demethylhomolycorine [51, Fig. (12) and Table 2], whose structure was confirmed by preparing the corresponding monoacetyl derivative[176].

H3C0

4

r

2'

3' 4"

.vvQCOCHoCHCHa H OH

Fig. (12). 9-O-demethylhomolycorine (51) and the nobilisitine A and B (52 and 53) extracted from bulbs of Narcissus tazetta L. and Clivia nobilis Regel, respectively.

Clivia species alkaloids

Clivia miniata and C noblis Regel are cultivated in Egypt as ornamental plants for their beautifiil flowers. Clivia species have been the major source of alkaloids represented by the 3a,4-dihydrolactone[2] benzopyrano[3,4-g] indole ring system and containing four chiral centres at the ring junction positions (3a, 5a, 1 lb and 1 Ic). The further chiral position at C-5 is due to the presence of an oxygen substitute. This class of alkaloids is represented by clivonine which was previously isolated from C. miniata [177-179], while from C. nobilis three alkaloids were isolated

620 and identified as lycorine, clivatine and nobilisine [180], the latter being a stereochemical variant of masanane ring system [180]. Further investigation of the alkaloids synthesised by C nobilis allowed isolation of two new masanane-type lactone alkaloids, called nobilisitine A and B [52 and 53, Fig. (12) and Table 2], whose structure was determined by spectroscopic techniques (essentially 1 and 2 D iR and ^^C NMR and EI MS). Both compounds were observed to contain the same [2]benzopyran[3,4-g]indole ring system and therefore belong to the same subgroup of Amaryllidaceae alkaloids. Li particular, nobilisitine A and B proved to be the 5p-hydroxy-3a,llc-e/?z-masan-7-one and the 3hydroxybutanoyl ester of clivonine (5a-hydroxy-5a-q^/-masan-7-one), which is the alkaloid previously isolated from C miniata and belongs to the same alkaloid subgroup. In particular, the cis B/C and C/D ring fusions and the P-configuration assumed by the hydroxy group at C-5 in 52, were deduced by accurate examination of the constants measured for the coupling between H-1 lb with both H-5a and H-1 Ic, and of H-5a with H-5 and the protons of H2C-4. These findings and the inspection of a Drieding model of 52 were in agreement with the diagnostic ^H chemical shift values recorded for the N-methyl group. This is very similar to those observed in other Amaryllidaceae alkaloids containing the 3a,4dihydrolactone [2]benzopyrano[3,4-g]indole ring system and belonging to the same stereochemical class II [180, 181]. On the basis of the same considerations a trans- and a c/5-j unction for the B/C and C/D ring fusion and an a-configuration of the 3-hydroxybutanoyl ester residue can be deduced for nobilisitine B, which was included in the stereochemical class in [180, 181]. NobiHsitine A, whose structure was also confirmed by preparing the corresponding 5-0-acetyl derivative, represents the first naturally occurring alkaloid of masanane series showing the epimerisation both C-3a and C-llc. As shown by the comparison of the physical (melting point) and ^H NMR data, nobilisitine B is an unusual 3hydroxybutanoyl ester of clivonine and epimer at C-3' of clivatine. The latter is an alkaloid of the same subgroup previously isolated from C miniata Kegel [177] and C. nobilis [180], and had been already described by Dopke [182] as an ester of clivonine and the (i-hydroxybutyric acid [183].

621

ABBREVIATIONS AA Ac-a-lAA-Lys Ac-8-IAA-Lys CD 13c NMR 2'deOZR diHZ diHZR IDNMR

= =

2DNMR EI MS EPS FAB MS FAB MS MS Fru/ Fuc/?3NAc Gal/? GaljoNAc Glcp GlcpNAc iHNMR HPLC lAA a-IAA-Lys s-IAA-Lys iP LPS rMeZ r'MeZR MTPA NCPPB

=

= =

Ascorbic Acid N-8-Acetyl-N-a-(indole-3-acetyl)-L-lysine N-a-Acetyl-N-s-(indole-3-acetyl)-L-lysine Circular Dichroism Carbon-13 Nuclear Magnetic Resonance 2'-Deoxyzeatin Riboside Dihydrozeatin Dihydrozeatin Riboside Monodimensional Nuclear Magnetic Resonance Bidimensional Nuclear Magnetic Resonance Electron Ionization Mass Spectrometry Exopolysaccharides Fast Atom Bombardment Mass Spectrometry Fast Atom Bombardment Tandem Mass Spectrometry Fruttofuranose N-Acetyl-3-deoxy-3-aminofucopyranose Galactopyranose N-Acetyl-2-deoxy-2-aminogalactopyranose Glucopyranose N-Acetyl-2-deoxy-2-aminoglucopyranose Hydrogen-1 Nuclear Magnetic Resonance High Performance Liquid Chromatography Indole-3-Acetic Acid N-a-(indole-3 -acetyl)-L-lysine N-s-(indole-3-acetyl)-L-lysine Isopentenyladenine Lipopolysaccharides I'-Metylzeatin r'Methylzeatin Riboside a-Methoxy-a-trifluorophenylacetate National Collection Plant Pathogenic Bacteria

622 NMDR NMR NOE PND Rha/> SFD SFORD

= = = = = = =

TEAB TLC Xylp ^Z t-ZR

= = = = =

Nuclear Magnetic Double Resonance Nuclear Magnetic Resonance Nuclear Overhauser Effect Proton Noise Decoupling Rhamnopyranose Selective Frequency Decoupling Single Frequency Off-Resonance Decoupling Triethylammomium Bicarbonate Thin Layer Chromatography Xylopyranose trans-Zeatin trans-ZQ2itin Riboside

ACKNOWLEDGEMENTS The work was supported by grants from the Italian Ministry of University and Scientific and Technological Research (MURST). Contribution (No. 187DISCA).

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

Strobel, G.A.; Ann. Rev. Biochem., 1982, 51, 309-333. Phytotoxins and Plant Pathogenesis; NATO ASI Series, Series H, Vol 27, Springer-Verlag, Berlin, 1989. Klement, Z.; Proceedings of 7th International Conference on Plant Pathogenic Bacteria; Hungary, June 11-16,1989. Ballio, A.; Graniti, A.; Experientia, 1991, 47, 751-684. Evidente, A.; In Recent Research Development in Phytochemistry; Pandalai, S.G., Ed.; Research Signpost: Trivandrum, India, 1997, pp. 255-292. Evidente, A.; Motta, A. In Bioactive Natural Products: Isolation, Structure Elucidation, and Biological Properties; Tringali, C, Ed.; Taylor & Francis Ltd.: London, 2000, Chapter 14, in press. Strobel, G.A.; Sugawara, F.; Clardy, J.; ACSSymp. Ser., 1987, 330, 516-523. Delfosse, E.S.; Proceedings of the VII International Symposium on Biological Control of Weed; Rome, Italy, March 6-11,1988. Cook, J.W.; Loudon, J.D. In The Alkaloids; Manske, R.H.F.; Holmes, H.L. Eds.; Academic Press: New York, 1952, Vol. II, pp. 331-352. Wildman, W.C. In The Alkaloids; Manske, R.H.F. Ed.; Academic Press: NewYork, 1960, Vol. VI, pp. 289-413.

623 [11] [12] [13] [14] [15] [16]

[17]

[18] [19] [20]

[21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35]

Wildman, W.C. In The Alkaloids, Manske, R.H.F. Ed.; Academic Press: New York, 1968, Vol. XI, pp. 307-405. Fuganti, C. In The Alkaloids; Manske, R.H.F., Ed.; Academic Press: NewYork, 1975, Vol. XV, pp. 83-164. Suffhes, M.; Cordell, G.A. In The Alkaloids; Manske, R.H.F., Ed.; Academic Press: NewYork, 1975, Vol. XV, pp. 198-212, 338-340. Ghosal, S.; Saini, K.S.; Razdan, S.; Phytochemistry, 1985, 24, 2141-2156. Martin, S.F. In The Alkaloids; Brossi, A., Ed.; Academic Press: New York, 1987, Vol. XXX, pp. 251-376. Bastida, J.; Viladomat, F.; Codina, C ; In Studies in Natural Products Chemistry; Atta-Ur-Rahaman, Ed.; Elsevier Science B.V.: Amsterdam, 1998, Vol 20, Part F, pp. 323-405. Rudolph, K.; Burr, T.J.; Mansfield, J.W.; Stead, D.; Vivian, A.; Von Kietzell, J. In Pseudomonas Syringae Pathovars and Related Pathogens; Kluwer Academic Publishers: Dordrecht, 1997. Hegelson, J.P. In Plant Disease; Horsfall, J.G.; Cowling, E.B, Eds; Academic Press: New York, 1978, Vol. 3. Surico, G.; Sparapano, L.; Lerario, D.; Durbin, D.R.; lacobellis, N.S.; Experientia, 1975, 31, 929-930. Surico, G.; Sparapano, L.; Lerario, D.; Durbin, D.R.; lacobellis, N.S.; Proceedings of Fourth Congress of the Mediterranean Phytopathological Union; Zara, Jugoslavia, 1976, pp 449-503. Evidente, A.; Surico, G.; lacobellis, N.S.; Randazzo, G.; Phytochemistry, 1986, 25, 525-526. Evidente, A.; Surico, G.; lacobeUis, N.S.; Randazzo, G.; Phytochemistry, 1985, 24, 1499-1502. Fletcher, R. A.; Kallidumbil, V.; Steele, P.; Plant Physiol, 1982, 69, 675-677. Itaya, T.; Fujii, T.; Evidente, A.; Randazzo, G.; Surico, G.; lacobellis, N.S.; Tetrahedron Lett., 1986, 27, 6349-6352. Ohba, M.; Fujii, T.; Evidente, A.; Surico, G.; lacobellis, N.S.; Heterocycles, 1990, 31, 599-602. Fujii, T.; Ohba, M.; Sakari, M.; Heterocycles, 1988, 27, 2077-2088. Psallidas, P.G.; Panagopoulos, C.G.; Annls Inst. Phytopath. Benaki, 1975, 11, 94-108. Gundogdu, M.; Kaya, S.J.; J. Turk. Phytopath., 1976, 5, 97-98. Ercolani, G.L.; Ghaffar, A.; Fao Plant Prot. Bull, 1985, 33, 37-39. lacobellis, N.S.; Evidente, A.; Surico, G.; Experientia, 1988, 44, 70-72. Evidente, A.; lacobellis, N.S.; Vellone, R.; Sisto, A.; Surico, G.; Phytochemistry, 1989, 28, 2603-2607. De Napoli, L.; Evidente, A.; Piccialli, G.; Santacroce, C.; Vellone, R.; Phytochemistry, 1990, 29, 701-703. Evidente, A.; Piccialli, G.; Sisto, A.; Ohba, M.; Honda, K.; Fujii, T.; Chem. Pharm. Bull, 1992, 40, 1937-1939. Evidente, A.; Fujii, T.; lacobellis, N.S.; Riva, S.; Sisto, A.; Surico, G.; Phytochemistry, 1991, 30, 3505-3510. Greco, F.; Sindona, G.; Uccella, N.; Evidente, A.; Org. Mass. Spectrom., 1992, 27, 750-755.

624

[36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60]

Evidente, A.; Surico, G.; lacobellis, N.S.; Randazzo, G.; Phytochemistry, 1986, 25, 125-128. Nakatani, Y.; Fujioka, M.; Higashino, K.; Anafyt. Biochem., 1972, 49, 225-231. De Napoli, L.; Evidente, A.; Piccialli, G.; Santacroce, C; lacobellis, N.S.; Sisto, A.; Phytochemistry, 1993, 33, 13-16. A. Evidente; Surico, G.; J. Nat. Prod., 1986, 49, 938-939. Evidente, A.; lacobellis, N.S.; Sisto, A.; Experientia, 1993, 49, 182-183. lacobellis, N.S.; Evidente, A.; Surico, G.; Sisto, A.; Gammaldi, R.; J. Phytopathol, 1990,129, 177-189. lacobellis, N.S.; Sisto, A.; Surico, G.; Evidente, A.; Di Maio, E.; J. Phytopathol, 1994,140, 238-248. Evidente, A.; Di Maio, E.; Caponero, A.; lacobellis, N.S. Experienita, 1995, 51, 990-993. lacobellis, N.S.; Caponero, A.; Evidente, A.; Plant Pathol, 1998, 47, 73-83. Dhavanthari, B.N.; Can. Plant Dis. Surv., 1969, 49, 36-37. Burr, T.J; Hurwitz, B.; Plant Dis. Rep., 1979, 63, 157-160. Bazzi, C; Calzolari, A.; Phytopathol medit, 1983, 22,19-21. Evidente, A.; lacobellis, N.S.; Scopa, A.; Surico, G.; Phytochemistry, 1990, 29, 1491-1497. Evidente, A.; lacobellis, N.S.; Impellizzeri, A.; Surico, G.; Phytochemistry, 1992,57,4105-4107. lacobellis, N.S.; De Vay, J.E.; Physiol Mol Plant Pathol, 1987, 30, 421-432. Tamura, S.; Chang, C.F.; Agr Biol Chem., 1965, 29, 1061-1062. Kimura, Y.; Tamura, S.; Agr. Biol Chem., 1973, 37, 2925. Takeda, Y.; Fujita, T.; Shingu, T.; Ogimi, C; Chem. Pharm. Bull, 1987, 35, 2569-2573. Noda, T.; Sato, Z.; Kobayashi, H.; Iwazaki, S.; Okuda, S.; Ann. Phytopath. Soc. Jpn., 9S0, 46, 663-666. Perreaux, D.; Maraite, H.; Meyer, J.A.; Physiol Plant Pathol, 1982, 20, 313319. Stead, D.E.; Int. J. System. Bacteriol, 1992, 42, 281-295. Jones, L.K.; Phytopathology, 1941, 31, 199. Holtzman, O. V.; Thomas Jr., W.D.; Phytopathology, 1953, 43, 587. Kusumi, T.; Ohtani, I.; Nishiyama, K.; Tetrahedron Lett, 1987, 28, 3981-3984. Lavermicocca, P.; lacobellis, N.S.; Di Maio, E.; Evidente, A.; Capasso, R.;

Petria,\99^,4,\l\-n0. [61] [62] [63] [64] [65] [66] [67]

Bender, C.L.; Alarcon-Chaidez, F.; Gross, D.C.; Microbiol. Mol Biol Rev., 1999, 63, 266-292. Tagg, J.R.; Dajani, A.S.; Wannamaker, L.W.; Bacteriol Rev., 1976, 40, 722756. Vidaver, A.K.; Plant Dis., 1983, 67, 471-475. Gross, D.C.; Vidaver, A.K. In Methods in Phytobacteriology, Klement, Z.; Rudolph, K; Sands, D.C. Eds; Akademiao Kiado: Budapest, 1990, pp. 245-249. Echandi, E.; Phytopathology, 1976, 66, 430-432. Vidaver, A.K.; Buckner, S.; Can. J. Microbiol, 1978, 24, 14-18. Xu, G.-W.; Gonzales, C.F;. Phytopathology, 1991, 81, 628-631.

625

[68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89]

[90] [91]

[92] [93] [94] [95]

[96]

lacobellis, N.S.; Contesini, A.M.; Surico, G.; Phytopathol medit, 1995, 34, 15-22. Frey, P.; Smith, J.J.; Albar, L.; Appl Environ. Microbiol, 1996, 62, 473-479. Haag, W.L.; Vidaver, A.K.; Antimicrob. Agents Chemoth., 191A, 6, 74-83. Smidth, M.L.; Vidaver, A.K.; Can. J. Microbiol, 1986, 32, 231-236. Ercolani, G.L.; Caldarola, M.; Phytopathol medit., 1972,11,1\-12>. Lavermicocca, P.; Lonigro, S.L.; Evidente, A.; Andolfi A.; J. Appl Microbiol, 1999, 86, 257-265. Vauterin, L.; Hoste, B.; Kersters, K.; Swings, J.; Int. J. System. Bacteriol, 1995, 45, 472-489. Patterson, C.L.; Grogan, R.G.; Campbell, R.N.; Plant Dis., 1986, 70, 982-985. Pennisi, A.M.; Pane, A.; Inf.tore Fitopatol, 1990, 4, 56-58. Scala, F.; Evidente, A.; Coppola, L.; Capasso, R.; Lorito, M; Zoina, A.; / Phytopathol, 1996,144, 325-329. Noda, T.; Sato, Z.; Kobayashi, H.; Iwasaki, S.; Okuda, S.; Ann. Phytopathol Soc. Jpn., 1980,46, 663-666. Robenson, D.J.; Cook, D.R.; Physiol Mol Plant Pathol, 1985, 26, 219-230. Ev^bank, E.; Maraite, H. J.; Gen. Microbiol, 1990,138, 1185-1189. Evidente, A.; Capretti, P.; Giordano, F.; Surico, G.; Experientia, 1992, 48, 1169-1172. Doworth, C.E.; Macey, D.E.; Sieber, T.N.; Woods, T.A.D.; Can. J. Plant Pathol, 1996,18,315^324. Hogdson, R.; Peterson, W. H.; Riker, A. J.; Phytopathology, 1949, 39, 47-62. Harbome J.B.; In Plant and Fungal Toxins; Keelr, R.F.; Tu Anthony, T., Eds.; Mercel Dekker: New York, 1983, pp. 743-782. Denny, T.P.; Ann. Rev. Phytopathol, 1995, 33, 173-197. Van Alfen, N K.; Ann. Rev. Phytopathol, 1989, 27, 553-550. Woodward, J.R.; Keane, P.J.; Stone, B.A.; Physiol Plant Pathol, 1980, 18, 438-454. Walkes, CM.; 0 ' Garro, L.W.; Physiol Mol Plant Pathol, 1996, 48, 91-104. Rudolph, K.W.E.; Gross, M.; Neugebauer, M.; Hokawat, S.; Zachowski, A.; Widra, K.; Klement, Z. In Phytotoxins and Plant Pathogenesis; Graniti, A.; Durbin, R.D.; Balho, A., Eds.; Springer Verlag: Berlin, 1989, pp. 177-218. McWain, P.; Gregory, G. F.; Phytochemistry, 1972,11, 2609-2612. Chatterjee, A.K.; Vidaver, A.K. In Genetics of Pathogenicity Factors: Applications to Phytopatogenic Bacteria; Igram, D.S; Williams, P.H., Eds.; Academic Press: London, 1986, Vol. 4, pp. 94-114. Halverson, L.J.; Stacey, G.; J. Microbiol Rev., 1986, 50, 193-225. Adinolfi, M.; Corsaro, M.M.; Lanzetta, R.; Marciano, C.E.; Parrilli, M.; Evidente, A.; Surico, G.; Can. J. Chem., 1994, 72, 1839-1843. Knirel, Y.A.; Shashkov, A.S.; Paramanov, N.A.; Zdorovenko, G.M.; Solyvanic, L.P.; Yakovleva, L.M.; Carbohydr. Res., 1993, 243, 199-204. Corsaro, M.M.; De Castro, C ; Evidente, A.; Lanzetta, A.; Lavermicocca, P.; Molinaro, A.; Sisto, A.; ParrilH, M.; Surico, G.; Phytochemistry, 1997, 46, 289292. Adinolfi, M.; Corsaro, M.M.; De Castro, C ; Lanzetta, R.; Parrilli, M.; Evidente, A.; Lavermicocca, P.; Carbohydr. Res., 1995, 267, 307-311.

626

[97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126]

Lindberg, B.; Adv. Carhohydr. Chem. Biochem., 1987, 48, 279-318. Gorshkova, R.P.; Zubkov, V.A.; Isakov, V.V.; Ovodov, Y.S.; Carbohydr. Res., 1984,72(5,308-312. Gilleron, M.; Vercauterem, J.; Puzo, G.; Biochemistry, 1994, 33, 1930-1937. Adinolfi, M.; Corsaro, M.M.; De Castro, C; Evidente, A.; Lanzetta, R.; Mangoni, L.; Parrilli, M.; Carbohydr. Res., 1995, 274, 223-232 (and references therein cited). Dale, J.A.; Mosher, H.S.; J. Am. Chem. Soc, 1973, 95, 512-519. Circular Dichroic Spectroscopy - Exiton Coupling in Organic Chemistry, University Press: Oxford, 1983. Adinolfi, M.; Corsaro, M.M.; De Castro, C; Evidente, A.; Lanzetta, R. Lavermicocca, P.; ParrilH, M Carbohydr Res., 1996, 284, 119-133. De Castro, C; Evidente, A.; Lanzetta, R.; Lavermicocca, P.; Manzo, E. Molinaro, A.; ParrilH, M.; Carbohydr Res., 1998, 307, 167-172. Adinolfi, M.; Corsaro, M.M.; De Castro, C; Evidente, A.; Lanzetta, R. Lavermicocca, P.; Parrilli, M.; Carbohydr Res., 1996, 284, 111-118. Peterson, A.A.; McGroarty, E.J.; J. Bacteriol, 1985,162, 738-745. Molinaro, A.; De Castro, C; Evidente, A.; Hoist, O.; ParrilH, M.; Eur J. Biochem., 1999, 259, 887-891. Knirel, Y.A.; Wildman, G.; Senchenkova, S.N.; Jansson, P.E.; Weintraub, A Eur. J. Biochem., 1997, 247, 402-410 (and references therein cited). Rodriguez, M.; Jann, B.; Jann, K.; Eur J. Biochem., 1988, 777, 117-124. Rodriguez, M.; Jann, B.; Jann, K.; Carbohydr Res., 1990,196, 101-109. Burkholder, W.H.; Phytopathology, 1950, 40, 115-117. Kawamoto, S.O.; Lorbeer, J.W.; Plant Dis. Rep., 1976, 60, 189-191. Janisiewicz, W.J.; Roitman, J.; Phytopathology, 1988, 78, 1697-1700. Hebbar, K.P.; Aktinson, D.; Tucker, W.; Dart, P.J.; Soil Biol. Chem., 1992, 24, 1009-1020. Cartwright, D.K.; Benson, D.M.; Plant Dis., 1995, 79, 301-308. Holmes, B.; J. Appl. Bacteriol, 1986, 61, 299-314. Goldman, D.A.; Klinger, J.D.; J. Pediatr., 1986,108, 806-812. Gilardi, G.L.; Lab. Managment, 1983, 21, 29-32. Beynon, L.M.; Cox, A.D.; Taylor, C.J.; Wilkinson, S.G.; Perry, M.B Carbohydr Res., 1995, 272, 231-239. Cimino, P.; Corsaro, M.M.; De Castro, C; Evidente, A; Marciano, C.E ParrilH, M; Sfalanga, A.; Surico, G.; Carbohydr Res., 1998, 307, 333-341. Barton-Willis, P.A.; Wang, M.C.; Holliday, M.J.; Long, M.R.; Keen, N.T Physiol. Plant Pathol, 1984, 25, 387-398. Molinaro, A.; Lanzetta, R.; Evidente, A.; Parrilli, M.; Hoist, O.; FEMS Microbiol Lett., 1999 181, 49-53. Maas, J.L.; Pooler, M.R.; Gallette, G.L.; Adv. Strawb. Res., 1995,14, 18. Scortichini, M.; Riv. Fruttic, 1996, (5, 51-53. Molinaro, A.; Evidente, A.; Fiore, S.; lacobellis, S.N.; Lanzetta, R.; Parrilli, M.; Carbohydr Res., 2000, 325, 222-229. Cimino, P.; Corsaro, M.M.; Evidente, A.; Lanzetta, R.; Molinaro, A; ParriUi, M.; Sfalanga, A.; Guastani, M.; Surico, G.; Phytopathol medit, 1998, 37, 1321.

627

[127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158]

King, E.O.; Ward, M.K.; Raney, D.E.; J. Lab. Clinic Med., 1954, 44, 301-307. Hounsell, E.F.; Feeney, J.; Scudder, P.; Biochem. J., 1988, 250, 9-13. Scudder, P.; Lawson, A.M.; Honsell, E.F.; Carruthes, R.A.; Childs, R.A.; Feizi, T.; Eur. J. Biochem., 1987,168, 585-593. Dua, V.K.; Narasinga Rao, B.N.; Wu, S.S.; Dube, V.E.; Bush, C.A.; / Biol. Chem., 1986, 261, \599-\60S. Santer, U.V.; Glick, M.C.; van Halbeek, H.; Vliegenthart, J.F.G.; Carbohydr. Res., 1983,120, 197-213. Nakagawa, Y.; Uyeo, S.; Yayima, H.; Chem. Ind, 1956, 1238-1239 (and references therein cited). Arrigoni, O.; Arrigoni Liso, R.; Calabrese, G.; Nature, 1975, 256, 513-514. De Leo, P.; Dalessandro , G.; De Santis, A.; Arrigoni, O.; Plant Cell Physiol., 1973,14, 487-496. Arrigoni, O.; Arrigoni Liso, R.; Calabrese, G.; Science, 1976,194, 332-333. Arrigoni, O.; J. Bioen. Biomem., 1994,26, 407-419. Evidente, A.; lasiello, I.; Randazzo, G.; Chem. Ind, 1984, 348-349 (and references therein cited) Evidente, A.; lasiello, L; Randazzo, G.; J. Cromatogr., 1983, 281, 362-366. Evidente, A.; Cicala, M.R.; Randazzo, G.; Riccio, R.; Calabrese, G.; Liso, R.; Arrigoni, O.; Phytochemistry, 1983, 22, 2193-2196. Ceriotti, G.; Nature, 1967,11, 595-596. Piozzi, F.; Fuganti, C; Mondelli, R.; Ceriotti, G.; Tetrahedron, 1968, 24, 11191131. Evidente, A.; Arrigoni, O.; Liso, R.; Calabrese, G.; Randazzo G. Phytochemistry, 1986, 25, 2739-2743. Arrigoni, O.; De Gara, L.; PacioUa, C; Evidente, A.; De Pinto, M.C.; Liso, R. J. Plant. Physiol, 1997,150, 362-364. Evidente, A.; Cicala, MR.; Giudicianni, I.; Randazzo, G.; Riccio, R. Phytochemistry, 1983, 22, 581-584. Fonseca, S.F.; Ruveda, E.A.; McChesney, J.D.; Phytochemistry, 1980, 19, 1527-1530. Phokas, G.K.; Farmakeftikon Deltion, 1971, 7, 9-12. Evidente, A.; lasiello, I.; Randazzo, G.; J. Nat. Prod., 1984, 47, 1061-1062. Dopke, W.; Arch. Pharm., 1962, 295, 920-924. Kondo, H.; Tomimura, K.; Ishiwata, S.; J. Pharm. Soc. Jpn., 1932, 52, 51-54. Sanger, F.; Biochem. J., 1945, 39, 507-515. Evidente, A.; lasiello, L; Randazzo, G.; J. Nat. Prod, 1984, 47, 1003-1008. Evidente, A J. Nat. Prod., 1986, 49, 168-169. Evidente, A.; J. Nat. Prod, 1986, 49, 90-94. Ahmed, Z.; Rizk, A.; Hamouda, F., Lloydia, 1964, 27, 115-134. El Tohamy, D.; Master Degree Thesis. Faculty of Pharmacy, University of Assiut, Egypt, 1988. Ali, A.A.; Mesbah, M.K.; Mohamed, M.H.; Bull. Pharm. Sci. Assiut University, 1984,7,351-362. Gaber, M.; Master Degree Thesis. Faculty of Pharmacy, University of Alexandria, Egypt, 1990. Evidente, A.; Planta Med, 1991, 57, 293-298.

628

[159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183]

Piozzi, F.; Marino, M.L.; Fuganti, C; Di Martino, A.; Phytochemistry, 1969, 8, 1745-1748. Abou-Donia, A.H.; De Giulio, A.; Evidente; A.; Gaber, M.; Habib, A.A.; Lanzetta, R.; El Din, S.; Phytochemistry, 1991, 30, 3445-3448. Ghosal, S.; Singh, S.K.; Srivastava, R.S.; Phytochemistry, 1986, 25, 19751978. Ghosal, S.; Singh, S.K.; Kumar, Y.; Unnikrishnan, S.; Chattparday, S.; Planta Med, 19SS, 54, 114-116. Evidente, A.; Randazzo, G.; Surico, G.; Lavermicocca, P.; Arrigoni, O.; J. Nat. Prod, 1985, 48, 564-570. Normatov, M.; Abduazimov, Kh.; Yunusov, S.Yu.; Uzbeksk. Khim. Zh., 1965, 9, 25-30. Ali, A.A.; El Sayed, H.M.; Abdallah, O.M.; Steglich, W.; Phytochemistry, 1986,25,2399-2401. Ghosal, S.; Kumar, Y.; Singh, S.K.; Kumar, A.; J. Chem. Res. Synop., 1986, 112-113. Fales, H.M.; Wamhoff, E.W.; Wildman, W.C; J. Am. Chem. Soc, 1955, 77, 5885-5890. Abou-Donia, A.H.; Habib, A.A.; El Din, A.S.; Evidente, A.; Gaber, M.; Scopa, A.; Phytochemistry, 1992, 31, 2139-2141. Takeda, K.; Kotera, K.; Mizukami, S.; J. Am. Chem. Soc, 1958, 80, 25622567. Garuccio, I.; Onofri, S.; Evidente, A.; Arrigoni, O.; Boll Soc. It. Biol. Sper., 1990, 66, 329-334. Garuccio, I.; Onofri, S.; Evidente, A.; Arrigoni, O.; Can. J. Bot, 1991, 69, 107111. Del Giudice, L.; Massardo, D.R.; Manna, F.; Evidente, A.; Randazzo, G.; Wolf, K.; Curr. Gen., 1984, 8, 493-498. Te-Cheng, Wu; Shuiting Chen; Tao-Yuan Wang; Pin Hsu; Yao Hsueh Hsueh Pao, 1965,12, 31-35; Chem. Abstr., 1965, 62, 11039g. Ramanatthan, S.; Furusawa, E.; Kroposki, M.; Furusawa, S.; Cutting, W.; Chemotherapy, 1968, 73, 121-127. Abou-Donia, A.H.; Darwish, F.A.; Ghazy, N.M.; Alex. J. Pharm. Sci., 1989, 3, 122-123. Evidente, A.; Lanzetta, R.; Abou-Donia, A.H.; Masauda, E.A.; Kassem, F.F.; Harraz, F.M.; Arch. Pharm., 1994, 327, 595-596. Kobayashi, S.; Ishikawa, H.; Sasakawa, E.; Kihara, M.; Shingu, T.; Kato, A.; Chem. Pharm. Bull., 1980, 28, 1827-1831. leven, M.; Vlietinck, J.; Vanden Berghe, D.A.; Totte, J.; Dommisse, R.; Esmans, E.; Alderweireldt, F.; J. Nat. Prod., 1982, 45, 564-573. Ah, A.A.; Ross, S. A.; El Moghazy, A.; El Moghazy, S.A.; J. Nat. Prod., 1983, 46, 350-352. Jeffs, P.W.; Mueller, L.; Abou-Donia, A.H.; Seif El-Din, A.A.; Campau, D.; J. Nat. Prod, 1988, 51, 549-554. Hawksworth, W.A.; Jeffs, P.W.; Tidd, B.K.; Toube, T.P.; J. Chem. Soc, 1965, 1991-2001. Dopke, W.; Heterocycles, 1977, 6, 551-5 Evidente, A.; Abou-Donia, A.H.; Darwish, F.A.; Masouda, E.A.; Kassem, F.F.; Hammoda, H.A.M.; Motta, A.; Phytochemistry, 1999, 51, 1151-1155.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

629

RUBIA TINCTORUML. GOVERDINA C.H. DERKSEN and TERIS A. VAN BEEK Laboratory of Organic Chemistry, Phytochemical Section, Wageningen University, Dreijenplein 8, 6703 HB Wageningen, The Netherlands ABSTRACT: The roots of Rubia tinctorum L. (madder) are the source of a natural dye. In this review for the first time all the different information on Rubia tinctorum available in the literature is summarised. The dye components are anthraquinones which probably contribute to the resistance of the plant against fungi in the soil. Madder roots have been used to dye textiles in many parts of the world since ancient times and an overview of the historical development, cultivation, harvesting and dyeing techniques of madder is given. The anthraquinone alizarin, the hydrolysis product of ruberythric acid, is supposed to be the main dye component of Rubia tinctorum. The chemical synthesis of alizarin and the biosynthesis of the anthraquinones in Rubia tinctorum are described. As far as the purification, structure elucidation and structures of isolated compounds are concerned, the review confines itself to the anthraquinones of madder. Finally the pharmacology and medicinal uses of madder and pure anthraquinones are discussed. This review supplements and updates earlier partial reviews on madder or anthraquinones by Schweppe, Thomson and Wijnsma.

HISTORY, TAXONOMY AND CULTIVATION The roots of Rubia tinctorum L. (madder) are the source of a natural dye and they have been used to dye textiles in many parts of the world since ancient times [1]. The dye components are anthraquinones [1] with alizarin, the hydrolysis product of ruberythric acid, being the main dye component of Rubia tinctorum. The anthraquinones probably contribute to the resistance of the plant against fungi in the soil [2]. In this review an overview of all the different information on Rubia tinctorum available in literature is given. Such a review article is not yet available. A number of good reviews deals with different aspects of Rubia tinctorum. For example: the history, cultivation and use of natural dye sources including Rubia tinctorum was described thoroughly by Schweppe [3]. Together with other references his book is used as a source for an overview of the historical development of Rubia tinctorum. Also the use of madder as a dye source is discussed in detail. In the excellent book by Thomson (1971) [1] all the naturally occurring anthraquinones and their spectral data are described. In 1986 Wijnsma and Verpoorte [4] wrote a

630 review article on the anthraquinones in Rubiaceae. Their review supplemented the review by Thomson with 50 newly identified antliraquinones from Rubiaceae. It also updated the spectral data of the anthraquinones. As far as the procedures and techniques used for purification and structure elucidation of Rubia compounds we confine ourselves to the anthraquinones. The chemical synthesis of alizarin and the biosynthesis of the anthraquinones in Rubia tinctorum are described. Leistner is one of the founders of the research on the biosynthesis of plant quinones including those of Rubia tinctorum [5]. This review also deals with the pharmacology and medicinal and other uses of Rubia tinctorum.

Fig. (1). Rubia tinctorum L. (madder). Reproduced from [7]

631 Name, Botany and Occurrence of Rubia tinctorum The Roman writer Plinius already used the name Rubia for the plant, because of the red colour of the roots. Tinctorum is derived from the Latin word for dyeing [3]. The genus Rubia belongs to the tribe Rubieae. The genera Galium and Asperula belong also to the Rubieae. The tribe is part of the subfamily Rubioideae. Other genera within the Rubioideae are Coffea, Cephaelis and Morinda. As well as the subfamily Cinchonoideae, the subfamily Rubioideae belongs to the Rubiaceae. The Rubiaceae are part of the order of the Gentianales [6]. The stalks of the madder plant (Fig. (1)) are so weak that they often lie along the ground, preventing the plant from rising to its maximum height of 60-100 cm. With little spines at the edges and midrib the plant can cling to other plants. The flower-shoots spring from the joints in pairs, the loose spikes of yellow, starry flowers blooming only in the second or third year, in June. The thick fleshy stalks that compose the perennial are about 0.6 cm thick. From the rhizome roots run under the surface of the soil for some distance, sending up shoots. The root is covered with a small blackish bark, the husk. Beneath this layer the root consists of a red layer, the intermediate layer. The imiermost layer of the root consists of large vascular bundle cells, the pith. These cells are surrounded by a small concentration of red dye. The highest concentration of red dye is found in the intermediate layer, less in the pith and almost nothing in the husk [8-

Fig. (2). Microscopic picture of a cross section of madder root [11]

632 11]. It is not clear how the red dye is stored in the intermediate layer. It is probably stored in compartments outside the cells, which are possibly divided by a pectin layer [11]. In Fig. (2) a microscopic picture of a crosssection of madder root is depicted. Rubia tinctorum is native in Southern and Southeast Europe, in the Mediterranean area, Asia Minor and in the Caucasus. Nowadays the plant also grows in China and Japan, up to the Malaysian Archipelago, in the Western part of North America, in Mexico and South America. In earlier days madder was cultivated in Central and Western Europe. Nowadays most of the plants are found in the wild [3]. In The Netherlands madder is again cultivated for its roots in the province of Groningen since about ten years. History Madder roots have been used for dyeing textiles in many parts of the world since ancient times. One of the eldest textiles dyed with madder is a belt found in Tutankhamun's grave and dates back to 1350 BC. During the times of the Pharaohs in Egypt neither alum nor another mordant was used during dyeing. In the surroundings of the Dead Sea a wool pouch dyed with madder from 135 AD was found. In China some archaeological grave treasures revealed the use of madder for dyeing fabrics. The ancient Greeks used madder for dyeing and the main red dye in the Roman Empire was obtained from madder. Around the time of the birth of Christ the quality of the dyed product improved by using a mordant. The dyeing knowledge got lost due to the migration of many nations after the decline of the Roman Empire. Only in the Byzantine Empire and in the East the dyeing technique was still practised. During the unrest between 600 and 900 AD a lot of dyers roamed off to Italy. From Italy they wandered to France, the Rhineland and as far as England. During the Middle Ages Charlemagne stimulated the cultivation of madder. He inserted the plant on the list 'Capitulare de villis'. On this list all the plants occurred which were allowed to be cultivated on land owned by the state (France, Belgium, Holland and Italy). The Flemish were very experienced in the dyeing of textiles. They used just three plants as dye sources, Reseda luteola ('Wau', yellow), Rubia tinctorum (red) and Isatis tinctoria ('Waid', blue). With these tliree plants they could create almost any colour. The textiles that were dyed during the Middle Ages were mainly derived from

633 animal sources, like wool and linen. The dyeing of plant based textiles was not developed during the Middle Ages. Dyeing of cotton and other plant based materials was still the secret of the East. During the Middle Ages the technique of dyeing was mainly based on oral tradition and only a few written manuscripts have been found. From the seventeenth century more books and articles on dyeing were published. In 1678 the Dutch East India Company (VOC) introduced the cotton printing trade in Europe. Especially in the deltas of the Dutch province Zeeland, the cultivation of madder was highly developed. In the 17th and 18th century the Turkish red dye technique was introduced in Europe. In 1747 GreeUTurkish dyers established in France (Rouen) a Turkish ruled red textile industry. Here the madder culture was the most famous of the 18th century. In the 19th century the madder cultivation also developed in the Alsace. From 1600-1900 there was a heavy trade in madder dyes throughout Europe. In 1868 Graebe and Liebermann discovered how to prepare alizarin synthetically. At the end of the 19th century the madder culture rapidly declined, due to the cheaper production of synthetic alizarin [3]. About one century later madder is again cultivated on a commercial scale in The Netherlands. Cultivation, Harvesting and Processing of Madder There is evidence that madder was already cultivated in The Netherlands in the 12th century. The madder cultivated in the clay soil of the province of Zeeland was long considered the finest madder in Europe [7,12,13]. The cultivation of madder started at the end of April or beginning of May. The young madder shoots were planted in well-manured soil. New plants were kept free of weeds and the soil around the roots was kept loose. In the autumn after the foliage had died the plant was covered with five to ten cm of earth. In this way the roots were protected against frost. In the following year the roots produced shoots. In the third year the weight of the roots increased with 70%. At the end of the third year in September and October the roots were harvested. The harvest was a heavy and time-consuming job. The farmer needed a lot of experience to dig out the entire root intact. The remaining soil was removed by shaking. The roots were left in the open air for drying during a couple of days [7,12,13]. The classical preparation of the roots was divided in two main processes: drying and stamping. After four to five days the madder was

634 placed inside the madder drying house. The roots that had to be dried were first laid on the bars of the lowest floor. After a couple of days the roots were moved higher and higher in the drying house. The drying house was heated by an oven at ground level. After drying the roots were trashed to loosen the earth and the husk of the roots. If the husk was not removed the product was sold as 'onberoofde' (unrobbed). The trash that was left after the threshing was called the 'Mullen'. This product had the lowest dyeing quality. The (robbed) roots were placed in a stamp trough fitted with iron plates where they were ground to powder [13]. The stamped product was sieved through coarser and finer sieves, which gave products of different quality. The thoughest part of the root remained and was stamped again and sieved. This process was repeated a couple of times. In this way first the outside of the root was collected and finally the inner part of the root. Von Wiesner wrote that the part of the root that was sieved first possessed the highest dyeing power. In contrast with Von Wiesner, De Kanter wrote that the inner part, i.e. the last part sieved had the highest dyeing quality [10,14]. In former days two main qualities were disfinguished called the 'Gemene' and the 'Krap'. The 'Krap' had a higher dyeing quality than the 'Gemene'. The madder product 'Two and One' consisted of two parts 'Krap' and one part 'Gemene'. The madder product 'One and One' consisted of one part 'Gemene' and one part 'Krap' [7,12-14], According to Van DijkVan der Peijl the outer part of the 'robbed' root was called 'Gemene' and the inner part 'Krap' [7,12,13]. According to microscopic analysis of some cross-sections of madder root the concentration of red colour is not the highest in the centre of the root. The highest concentration of red dye was found in the layer between the husk and pith of the root [8,10,11]. In the 18th century France became an important competitor in madder cultivation, due to improved production methods. An innovation was the treatment of dried and ground roots with sulphuric acid. The woody roots were destroyed and charred, while alizarin itself remained unaffected. The residue after drying and powdering was called 'Garancin'. 'Garancin' possessed a colouring capacity, which was four to six times higher than that of the original madder. 'Garancin' was favoured over the madder products from Zeeland and was a commercial success [7,8,10,12,13]. Another madder product was 'Madder Flower'. The ground madder was macerated in a solution of diluted sulphuric acid, washed, pressed, dried and ground. In this way sugars were removed and glycosides hydrolysed possibly through partial fermentation (some yeast was added). An enzyme

635 which occurs naturally in madder (Erythrozym) may have played a role in the hydrolysis as well. 'Madder Flower' had almost twice the dyeing capacity of the original madder. When 'Madder Flower' was further extracted with methanol and precipitated with water the product was called'Azale' [8,10]. SYNTHESIS AND BIOSYNTHESIS Synthesis of Alizarin In 1868 Graebe and Liebermann showed that alizarin could be reduced to anthracene by the then novel procedure of zinc distillation. At that time the exact structure of alizarin was still unknown. They suggested that the structure of alizarin was 1,2-dihydroxy anthraquinone [15]. In 1868 they verified this hypothesis by a total synthesis of alizarin. In the knowledge, that two of the halogen atoms in chloranil could be replaced by hydroxyl groups on treatment with alkali, they prepared a dibromoanthraquinone. They assumed that it would behave in the same way and indeed on fusion with NaOH it yielded a compound in all respects identical with the natural alizarin obtained from madder [16]. This was the first total synthesis of a natural pigment. Owing to the cost of bromine, this route was however not economically viable. Almost at the same moment Graebe, Liebermann and Caro in Germany and Perkin in England found a similar, alternative reaction pathway for the synthesis of alizarin. Graebe, Liebermann and Caro patented this reaction. They introduced the sulphonation of

O

o •

o

Fig. (3). Synthesis of alizarin introduced by Graebe, Liebermann and Caro [1]

OH

636 antliraquinone to 2-antliraquinone sulphonic acid and the subsequent conversion to alizarin (Fig. (3)) [16]. Nucleophilic substitution by a hydroxyl ion displaces a hydride ion. Alizarin plants were subsequently built in Germany and England [17-19]. The procedure developed by Koch (B.A.S.F.) is based on the sulphonation of the anthraquinone with oleum followed by alkali fusion under addition of an oxidising agent (Fig. (4)) [20]. o

oleum I ^^ 0

NaOH+NaNO, '

180°C

20%, 130°C 0

0

Fig. (4). Synthesis of alizarin based on sulphonation with oleum and alkali fusion [20]

For the synthesis of highly pure alizarin, one has to start with pure 2antliraquinone sulphonic acid during the alkali smelt. This can be obtained by choosing the reaction conditions and oleum concentration in such a way that only 40% anthraquinone will be sulphonated. Of the then formed anthraquinone sulphonic acids about 85% is the desired isomer. The remaining sulphonic acids consist of traces 1-anthraquinone sulphonic acid and about 15% 2,6-and 2,7-disulphonic acids. This result can be achieved with 20%) oleum, a reaction temperature of 130°C, a reaction time of 4 hours and a weight ratio SO3 to anthraquinone of 1:1. This method gives the highest possible yield of 2-sulphonic acid. After the reaction, the solution is diluted with 80%) sulphuric acid, until the remaining oleum is converted into 90%) sulphuric acid. Under these conditions the unreacted anthraquinone precipitates after which it is filtered, washed until neutral and dried [16,20-22]. In the next step the filtrate is diluted with 40%) sulphuric acid and NaCl is added till the sodium salt of 2-anthraquinone sulphonic acid precipitates. The other sulphonic acids remain in solution. The salt is converted with excess base and sodium nitrate into the sodium salt of alizarin in an autoclave at a temperature of 180°C. This reaction takes 7 hours with a yield of approximately 95%o [20]. After dilution with water, the alizarin is separated from the reaction solution by acidifying and filtration. Finally the residue is washed till it is free from acids and then mixed with 20%o water to form a paste. As such it is used for dyeing. The product was sold

637 as Alizarinred BB 20% (B.A.S.F.) or ditto B2 20% (ICI) [20]. Alizarin is not sold anymore by these companies. If the sulphonation step is prolonged all the anthraquinones are almost exclusively converted into 2,6- and 2,7 anthraquinone disulphonic acids. After dilution of this reaction mixture to 40%) sulphuric acid, saturating the solution with salts can precipitate the compounds. After fusion with base a mixture of 1,2,6- en 1,2,7-trihydroxyanthraquinones is obtained that leads to a more yellowish red on fabrics that have been impregnated with aluminium mordant [20]. Biosynthesis Leistner in Germany and Burnett in Great-Britain have performed much work on the biosynthesis of anthraquinones [5,23-28]. Two different pathways for the biosynthesis of anthraquinones in plants exist: either the polyketide pathway or the shikimate pathway [5]. Anthraquinones synthesised according to the former pathway are substituted in both ring A and C. Examples are emodin and chrysophanol. These polyketide-derived antliraquinones are found in Rliamnaceae, Polygonaceae and Leguminosae ^COOH

^%^

a-ketoglutaric aicd

OC/r o 0-siiccinylbenzoic acid

iso-chorismic acid

1,4-dihydroxy2-naphtoic acid

dimethylallyi pyrophosphate

1,4-dihydroxy-3-isopentenyl2-naphtoic acid

o o-succinylbenzoic-CoA

anthraquinone

Fig. (5). Biosynthesis of anthraquinones by the shikimate pathway based on the scheme proposed by Wijnsma and Verpoorte [4]

638 antliraquinones are found in Rhamnaceae, Polygonaceae and Leguminosae [5,23,28,29]. Anthraquinones formed by the latter pathway (Fig. (5)) are characterised by a substitution in one ring only. These anthraquinones like alizarin are found in Rubiaceae (e.g. Rubia tinctorum) [1,4,5,23,24,27,30]. Ring A and one of the carbonyl groups of the anthraquinones are derived from shikimic acid [26]. The other carbonyl group and the remainder of ring B are derived from a-ketogluturate that originates in turn from glutamate [24]. The Acetyl coenzyme ester of o-succinylbenzoic acid is cyclized to 1,4-dihydroxy-2-naphtoic acid [4]. Symmetrical 1,4naphtoquinone is not an intermediate in the biosynthesis as has been shown in labelling studies [5,24,25,27,30,31]. The intermediate 1,4dihydroxy-3-isopentenyl-2-naphtoic acid is formed by incorporation of dimethylallyl pyrophosphate at the unsubstituted carbon of the quinone ring [32]. Finally cyclisation of the prenyl with the C-1 carbon will lead to ring C of the anthraquinone. If the isopentenyl unit cyclizes on the OH group at C-10 this will lead to a pyran, for example mollugin [5,25,27,30,31]. The biosynthesis of these anthraquinones parallels those of the menoquinones in bacteria and naphtoquinones of plants for example juglone, vitamin K and lawsone. These compounds are also derived from shikimic (or chorismic) and a-ketoglutaric acids via o-succinylbenzoic acid [5,24]. 1,4-Dihydroxy-2-naphtoic acid is the branching point in the biosynthesis of menoquinones, naphtoquinones and anthraquinones [4]. Recently it was shown by Eichinger et al. in an elegant biosynthetic study that the isopentenyl pyrophosphate unit, forming the C-ring in the Rubia anthraquinones was not derived from [33]. Until a few years ago it was thought that isopentenyl pyrophosphate was always derived from mevalonate. However it has since been shown that two different pathways exist: (1) the mevalonate pathway, and (2) the deoxyxylulose pathway (operational in plastids). In the latter pathway pyruvate and glyceraldehyde 3-phosphate - via 1-deoxy-D-xylulose 5-phosphate - are the precursors of terpenoids. Cell cultures of Rubia tinctorum were fed with [1-^"^C]- or [U-^"^C6]-glucose. Amino acids were obtained by hydrolysis of biomass, and their '"^C labelling patterns were used to reconstruct the labelling patterns of acetyl CoA, pyruvate, phosphoenol pyruvate, erythrose 4-phosphate, and a-ketoglutarate by retrobiosynthetic analysis. The labelling patterns were used to predict the labelling patterns of lucidin primeveroside via three different hypothetical pathways

639 (polyketide, mevalonate and deoxyxylulose). The observed labelling pattern was in excellent agreement with the pattern predicted on the basis of the precursors o-succinylbenzoate and dimethylallyl pyrophosphate derived via the deoxyxylulose pathway [33]. Anthraquinones in Rubia tinctorum are mostly present as glycosides. Glycosylated secondary products differ from the free aglycones in two properties: they show increased water solubility and decreased chemical reactivity. Because of the better water solubility the glycosides are stored in the plant vacuole and the glycosides are less reactive toward cell compounds. This is probably the reason why glycosylated compounds, rather than free aglycones are accumulated [34]. Research on the biosynthesis of anthraquinones is mostly focused on the aglycone. Little research has been carried out on the modification reactions on the aglycone part and on the glycosylation and storage of anthraquinone glycosides in Rubia tinctorum. In Morinda the anthraquinone glycosides are stored in the vacuole. The enzymatic mechanism for hydrolysis of glycosides and glycosylation of anthraquinones is present or rapidly induced in cells that do not produce these compounds themselves [35]. Cell cultures Cell cultures have proven particularly valuable in studies to determine the biosynthetic pathway of antliraquinones. The biosynthesis of anthraquinones can be influenced by exogenous physiological and environmental factors. The composition of the medium, in particular the plant growth regulators present, has been found to exert major effects on anthraquinone synthesis in a number of species [36-45]. A number of articles deal with the optimisation of media for a maximum production of anthraquinones in in vitro cultures of Rubia tinctorum. The standard anthraquinone content in plant cells grown in vitro is somewhat lower than in intact plants. The anthraquinone composition in such in vitro cells can also differ from that in the roots and rhizomes of intact plants [46]. Schulte et al compared anthraquinone production in Rubia tinctorum plant roots with the production in suspension cultured cells [37]. They found for the roots a content of 110-|amol/g dry weight, and for the suspension culture a content of 158 |imol/g. However due to the faster growth rate of the suspension culture the productivity in the later system is higher. The effects of the phytohormones lAA, NAA, 2,4-D and kinetin, and of the

640 sucrose concentration on the growth of the roots and anthraquinone production in Rubia tinctorum hairy root cultures was studied by Sato et al [47]. Highest anthraquinone production was found in a culture grown in liquid Murashige and Skoog (MS) medium supplemented with 5 |LIM lAA and 3% sucrose, while kinetin had no effect. Higher concentrations of sucrose (6-18%) inhibited growth in the presence of phytohormones (5 |LiM lAA or 5 |^M NAA). In a phytohormone free medium, 12%-sucrose resulted in maximum growth and anthraquinone production [47]. Highest levels of lucidin, alizarin and purpurin were found in a culture growing on a phytohormone-free MS-medium supplemented with 18% sucrose, but growth was poor on this medium [47,48]. Anthraquinones were formed in Rubia tinctorum cell cultures in contrast to Morinda cell cultures when the phytohormone 2,4-D was used [35,40,45,47,49]. Kino-Oka et al found that the growth of the cells was highest in a culture using fructose as a carbon source. If both glucose and fructose were available as carbon source, the cell culture preferred the uptake of glucose. Culture growth was much slower on glucose than fructose. The best nitrogen source for the cell culture was NOs", N H / inhibited the growth [50]. Hairy roots of Rubia tinctorum can be induced by infection with Agrobacterium rhizogenes strains. Such cultures retain the capacity for the biosynthesis of antliraquinones. A team of Japanese scientists used A. rhizogenes strain 15834, and obtained a madder root culture with a low growth index (G.I. = final fresh weight/initial fresh weight = 7.5) on hormone free medium. The addition of lAA to the nutrient medium stimulated the growth of roots, but the authors failed to increase the G.I. above a value of 19 within four weeks [46,47]. Another group of Japanese scientists carried out a genetic transformation with the A. rhizogenes strain ATCC 13332. The derived hairy root culture had a G.I. of about 40 over a four-week period [46,50]. In this research summation of HPLC peak areas of ruberythric acid and alizarin was interpreted as the amount of anthraquinones present. As reference compound they purchased authentic ruberythric acid from Carl Roth Co. It is not clear whether they took into account that this (commercial) ruberythric acid contains only 20% of anthraquinones [51]. The growth of the transgenic root culture obtained withy4. rhizogenes strain r-1601 on MS-medium or Gamborg B-5 medium was similar to that obtained by Kino-Oka et al. The anthraquinone content of hairy roots grown on Gamborg B-5 medium was twice as high as that of cells grown on MS medium [46]. Transformation oiR. tinctorum

641 cotyledons with the agropine type A. rhizogenes strain LBA 9402 resulted in a root line, which produced a different type of anthraquinone structure. Nordamnacanthal was the major product for this root line in modified B5 liquid medium [48]. EXTRACTION, ISOLATION AND PURIFICATION OF ANTHRAQUINONES ¥ROM RUBIA TINCTORUM A lot of articles deal with the isolation, separation and purification of anthraquinones from Rubia tinctorum. Purification of these anthraquinones may be necessary for the study of their biosynthesis, the analysis of the dyeing properties or for screening of mutagenic activity. An enumeration of the techniques used by the different groups for Rubia tinctorum follows below. Extraction Many articles deal with the extraction of anthraquinones from the roots, the above ground parts or cell cultures of Rubia tinctorum. The isolation procedure depends on whether the free aglycones or the glycosides are desired. The anthraquinones can be isolated by sequential extraction with solvents of increasing polarity [52-56]. The different extract solutions can be further purified by a liquid-liquid partitioning step. As a first extraction step a non-polar solvent can be used such as ether, benzene, chloroform, dichloromethane or ethyl acetate [52,57-62]. The anthraquinone glycosides, however, should be extracted by using water, ethanol, methanol or water-ethanol mixtures [46,47,52,58,62-67]. The extraction can be performed at different temperatures. During the extraction of anthraquinones from plant material with hot methanol or ethanol artefacts can be formed. These anthraquinones show the presence of 2methoxymethyl or 2-ethoxymethyl groups respectively caused by the partial conversion of lucidin to the corresponding co-ethyl ether. This reaction is highly temperature-dependent [68]. Thus for the extraction of anthraquinones the use of hot methanol or ethanol should be avoided [4]. Also basic solutions are used for the extraction of anthraquinones or the fractionation of apolar plant extracts, e.g. sodium carbonate, sodium bicarbonate or sodium hydroxide [53,54,62,65,69]. Most of the anthraquinones in madder are phenols. In general phenols dissolve well in

642 basic solutions [62]. Masawaki et al (1996) used this property for the simultaneous extraction of both aglycones and glycosides. When an aqueous KOH solution (50 mmol/nr^) was used as extraction solvent, both ruberythric acid and lucidin primeveroside were extracted. When chloroform was used as extraction solvent, the anthraquinone primeverosides were not extracted and only a small amount of alizarin and lucidin was extracted. Selective extraction from dried pulverised madder roots of the anthraquinones alizarin and lucidin into chloroform and their primeverosides into the aqueous phase was achieved using a chloroformwater two-phase extraction with pH adjustment to pH=5 [70]. In many cases the glycosides in the extract are subsequently hydrolysed in aqueous (2-5%) H2SO4 or HCl solution at 80-100°C [47,64,66,71]. In this process the main glycosides of madder, ruberythric acid and lucidin primeveroside, are converted to their aglycones alizarin and lucidin with the simultaneous release of glucose and xylose. During the hydrolysis most of the aglycones precipitate [62]. Direct hydrolysis of madder root extract gives a black precipitate. An aqueous extract of madder root also contains asperuloside, an iridoid glycoside. On warming with dilute acids i.e. the conditions of the hydrolysis, asperuloside gives first a green colour and then a tarry black precipitate [64,72]. Partial hydrolysis of anthraquinone glycosides can be achieved with an aqueous solution of NaOH [73]. Another possibility is the hydrolysis of the glycosides with hydrolases. Masawaki et al claimed that ruberythric acid, in a two phase chloroform-water solution, could be selectively and completely converted to alizarin within 6 hours by enzymatic hydrolysis with B-glucosidase at 50°C at pH=5. In contrast only 60% of ruberythric acid was converted to alizarin in a one-phase aqueous solution after 6 hours. This could be explained by the fact that alizarin exerts an inhibitory effect on the enzymatic hydrolysis of ruberythric acid in the solution. In the two-phase extraction the enzymatic hydrolysis of ruberythric acid could proceed effectively because the inhibitory effect of alizarin was reduced by the extraction of alizarin into the chloroform layer [70]. In this experiment lucidin primeveroside was not enzymatically hydrolysed by B-glucosidase [70]. After extraction of cell-cultures containing anthraquinones often the amount of anthraquinones is calculated as alizarin. In most cases (pooled) extracts were adjusted to equal volume mostly with methanol or an 80% ethanol solution and the quantification was carried out at 434 nm and

643 compared to a calibration curve of alizarin at 434 nm [37,40,46,48] or at 550 nm after adding KOH to the solution [48]. The antliraquinones can be further purified by precipitation or by chromatography. Precipitation can be achieved with reagents as lead acetate [74]. Precipitation was very common in older days. Nowadays components are mainly purified by chromatography. Column Chromatography Crude extracts of anthraquinones can be further fractionated by column chromatography. A lot of different column materials have been used for the purification of anthraquinones of Rubia tinctorum. Eluents used for Table 1. Examples of Separation of Anthraquinones by Low-Pressure Column Chromatography 1 Product name

Material

Solvent

1 Silica gel

Si02

methanol

[52]

acetone

[53]

chloroform

[53]

petroleum ether 40-60

[53]

benzene

[53]

chloroform-methanol

[55]

hexane-ethyl acetate

[55]

hexane

[60]

chloroform-methanol

[75]

0.5% aqueous carbonate methanol

[65]

Sephadex G-25

crossiinked dextran

Sephadex LH20

crosslinked hydroxypropyl ated dextran

Amberlite XAD-2 Dowex 50 (H")

non-ionic polymeric absorbent strongly acidic cationexchange resin

1 Reference

sodium

[52]

methanol

[66]

chloroform-methanol

[60]

1 water-methanol | 1 methanol 1

[73] [66]

644 normal phase column chromatography usually consist of a series of solvents of increasing polarity. Different authors used different column materials for the fractionation of anthraquinones of Rubia tinctorum, see Table (1). Pre-purification and concentration of anthraquinones occurring in plant extracts can be effected by solid-phase extraction (SPE) using C8 material. The crude ethanol extract was diluted tenfold with water and passed dropwise through pre-activated SPE cartridges. These cartridges were then washed with water followed by methanol-water (30-70). After drying the cartridge with air the anthraquinone fraction was eluted from the column with methanol-water (80:20) [63,66]. Paper Chromatography and Thin Layer Chromatography Paper chromatography has been frequently used in the past. Shibata et al (1950) [76] applied paper chromatography for the separation and identification of anthraquinone pigments. Petrol ether (bp. 45-70°C) saturated with 97% methyl alcohol was used as solvent. After development the paper strip was sprayed with 0.5% magnesium acetate in methanol and heated at 90°C for five minutes. Distinct orange-red, purple or violet coloured spots, depending on the position of the hydroxyl groups in the anthraquinone nucleus, were produced with this reagent [76]. nButanol-acetic acid-water (4:1:5) was used as eluent for the separation of both the glycosides and aglycones [77]. Nowadays thin layer chromatography (TLC) has replaced paper chromatography. TLC is now frequently used for the planar separation of anthraquinones, the most frequently used adsorbent being silica [4]. Almost every research group has developed their own solvent system and, in Table (2) some examples of eluents are given for the separation of anthraquinones in Rubia tinctorum extracts. Preparative TLC has been applied for the purification of anthraquinones from Rubia tinctorum [78]. Detection of anthraquinones on TLC plates is very simple due to their yellow-orange colours. Some anthraquinones show a colour when observed in UV254 light and most of them show colours when observed in UV366 light. A change in colour is also observed when hydroxyantliraquinones are sprayed with a solution of KOH or NaOH in methanol (5%) w/v). This detection method for hydroxy anthraquinones is called the Borntrager reaction. The colour changes from yellow-orange to red-purple

645

Table 2. Eluents Used for the Separation of Anthraquinones on Silica Gel TLC 1 Eluent

1 Reference

1 n-butanol saturated with 6M ammonia-methanol 4:1

[64]

1 ethyl formate-formic acid-toluene 7:2:5

[64]

1 ethyl formate-tormic acid-toluene 4:1:5

[64]

1 toluene-pyridine-acetic acid 10:1:1

[64]

1 benzene-ethyl acetate 1:1

[64]

1 n-butanol-ethanol-water 4:1:5

[64]

1 n-butanol-pyridine-water-benzene 5:3:3:1

[64]

1 n-butanol-pyridine-water 6 4:3

[64]

1 methanol-10 vol-% aqueous acetic acid 6:4

[71]

1 methanol-10 vol-% aqueous acetic acid 7:3

[71]

1 methanol-10 vol-% aqueous acetic acid 8:2

[71]

1 chloroform-hexane-ethyl acetate-acetic acid 40:40:15:5

[71]

1 toluene-acetic acid 9:1

[71]

1 n-butanol-ethanol-ammonia 6:2:3

[65]

1 benzene-ethyl formate-formic acid 40:24:1

[65]

1 chloroform-ethyl acetate 6:1

[65]

ethyl acetate-methyl ethyl ketone-formic acid-water 5:2:0.2:1

[65]

1 toluene-methanol 9:1

[48]

1 chloroform-methanol-25% ammonia 85:14:1

[48]

1 benzene-ethanol 8:2

[66]

1 benzene-ethyl acetate-methanol 40:30:5

[61,79]

1 chloroform-benzene-ethyl acetate-acetic acid 40:40:15:5

[61]

1 methanol-10 vol-% aqueous acetic acid 8:2

1[58]

1 chloroform-benzene-ethyl acetate-ethanol 8:8:3:1

1[60]

646 1 benzene-acetone 9:1

l60]

1 chloroform-methanol 9:1

[55,60]

1 hexane-ethyl acetate 9:1

[55]

hexane-ethyl acetate 8:2

[55]

1 chloroform-methanol 8:2

[55]

chloroform-methanol 7:3

[55]

due to the ionisation of the OH group. The mesomerism is enlarged and results in a shift of the absorption maximum to higher wavelength [62,80]. Another identification test is the exposure of the spots to NH3 vapour; a colour change of the anthraquinone spots is observed [4,61]. Spots can also be revealed by spraying with 0.5% magnesium acetate in methanol and heating at 90°C for five minutes. Hydroxyanthraquinones react with this reagent if they have at least one hydroxyl group in the 1 -position. Compounds, which contain two hydroxyl groups in the 1,2-position, give a violet colour, those with two in the 1,3-position an orange-red or pinlc colour and those with two in the 1,4-position give a purple colour [76]. Gas Liquid Chromatography (GLC) For GLC studies of hydroxyanthraquinones more volatile derivatives have to be prepared, because the antliraquinones as such are not sufficiently volatile. Several derivatives, namely methyl ethers, trimethylsilyl (TMS) ethers and trifluoroacyl derivatives have been separated on SE-30, OV-17, and UC-W98 phases [81,82]. All anthraquinone derivatives other than reductively silylated ones gave excessive tailing when cliromatographed [81]. Reductive silylation was therefore considered the method of choice for derivatization of hydroxyanthraquinones [81]. The influence of the number and position of substituents in the anthraquinone nucleus on GLC behaviour is clear. An increase in retention time was observed with an increase in the number of hydroxyl groups. The GC elution order largely follows the molecular weights. Within the group of dihydroxyanthraquinones, 1,4-dihydroxyanthraquinones had shorter retention times than 1,2- or 1,3-dihydroxyantliraquinones. Hydroxyanthraquinones have longer retention times than the corresponding methylanthraquinones. The

647 retention time decreased when a hydroxyl group was replaced by a methoxyl group [81-83]. High Pressure Liquid Chromatography (HPLC) In spite of the excellent separation of anthraquinones by GLC, one preferably uses high pressure liquid chromatography (HPLC) for the separation of these compounds. For an HPLC separation the anthraquinones do not have to be silylated, which saves time and prevents possible losses during the derivatisation. Several screening methods for anthraquinones in Rubia tinctorum, based on HPLC have been described in the literature [57,58,63,66,71]. Cig reversed-phase materials are mostly used with an occassional mentioning of silica gel. Table 3. tinctorum

HPLC Systems Used for the Separation of Anthraquinones of Rubia

Eiuent

Anthraquinone type

0.1% aq. acetic acid-acetonitrile gradient

glycosides + agly cones 1 [57]

RP-C18Novapak 1 column Shin-Pack CLC-ODS

methanol-10% aq. acetic acid isocratic, internal standard: trichlorophenol methanol-5% aq. acetic acid 7:3 isocratic

aglycones

1 [71]

aglycones

[52]

Shin-Pack CLC-ODS

methanol-5% aq. acetic acid 6:4 isocratic

glycosides

[52]

aglycones

[63]

glycosides

[47]

aglycones

[47]

alizarin

[46]

methanol-O.r/oH^POj

glycosides + aglycones

[66]

glycosides + aglycones

[66]

glycosides + aglycones

[66]

RPCsLiChroCART

isocratic methanol- 5% aq. acetic acid 7:3 acetonitriIe-0.02 M ammonium acetate buffer pH 4 15:85 gradient methanol-10%0 aq. acetic acid isocratic

aglycones

[58]

Alltima end-capped Cis

water-acetonitrile gradient

glycosides + aglycones

[51]

1 HPLC column KONTRONRP18

isocratic elution methanol-5% aq. acetic 1 Hypersil 5 reversed 1 phase material acid 7:3 TSKODS 120 T column methanol-10% aq. acetic acid 3:7 gradient TSKODS 120 T column methanol-10% aq. acetic acid 3:2 gradient acetonitrile- 4%) aq. acetic acid gradient 1 Armsfer-C8 column Nucleosil-508 1 ODS Hypersil column. Superpac PEP S C2/C,K |RP

Reference 1

648 In general, the HPLC methods focus on the aglycones or even only on alizarin. In most cases the madder extract is first hydrolysed and the total quantity of alizarin is determined after HPLC separation [58,71]. Two HPLC methods for the simultaneous analysis of both glycosides and aglycones failed to give a baseline separation of the two main glycosides, ruberythric acid and lucidin primeveroside [57,66]. Recently a HPLC method was published for the quantitative detection of the most common glycosides and aglycones in madder root. In this HPLC method the glycosides ruberythric acid and lucidin primeveroside were baseline separated [51]. In all the articles the antliraquinones of Rubia tinctorum were (quantitatively) measured through determination of their ultraviolet (UV) absorbance at «254 nm [51,57,58,63,66,71,84], 280 nm [46,47,66] or at visible wavelengths more specific for anthraquinones such as 430 nm [50], 480 nm [58] or 500 nm [58]. The eluents used for the HPLC separation of the glycosides generally consisted of mixtures of water and acetonitrile or water and methanol [4]. Most research groups added some acid to suppress tailing of the antliraquinone peaks [51]. In Table (3) the HPLC systems used for the separation of anthraquinones of Rubia tinctorum are described. Liquid-Liquid Chromatography Techniques Liquid-liquid partition cliromatography has been applied successfully for the separation of anthraquinone glycosides and aglycones [30,78,85,86]. A liquid-liquid separation technique called droplet counter current chromatography (DCCC) has been developed in 1970 by Tanimura et al. [87]. The liquid separation is based on the partitioning of the different compounds in the sample between many tiny mobile phase droplets, which move through the stationary phase. DCCC was used for the separation of anthraquinone glycosides or aglycones [85,86]. The solvent system chloroform-methanol-water 5:5:3 was used for the separation of the two main anthraquinone glycosides of Rubia tinctorum [51]. During a DCCC experiment the gravity is the driving force for the movement of the mobile phase through the stationary phase. In the newer technique of centrifugal partition chromatography (CPC), centrifugal forces drive the process. Separation by CPC is much faster than by DCCC. Hermans-Lokkerbol et al. optimised the solvent system for the separation

649 of both the anthraquinones aglycones and glycosides in a Rubia tinctorum extract with CPC. The solvent systems n-hexane-ethyl acetate-methanolwater 9:1:5:5 and chloroform-methanol-water-acetic acid 5:6:4:0.05 were used and offer different selectivity. The first system was used for the separation of anthraquinone aglycones and the second system for the separation of aglycones and glycosides [78]. Capillary Electrophoresis (CE) Recently the separation of 10 antliraquinone aglycones and two glycosides from Rheum by capillary electrophoresis was described [88] and compared with an HPLC separation. Two of the investigated aglycones also occur in madder: alizarin and purpurin. Because all of the anthraquinones can be charged by means of complexation with a borate buffer, capillary zone electrophoresis (CZE) was chosen as separation mode. The separations were carried out with a 90 cm x 75 |im fused silica capillary. The detection window was located at 80 cm. Dectection occurred at 260 nm. The voltage was 23 kV and the temperature 20°C. The injection was pressure controlled (1.2 sec at 200 mbar). The total run time was 39 min (compared to 63 min for gradient RP-HPLC for the same set of test substances). Various borate concentrations were tested and retention times increased with increasing concentration. Best resolution was obtained at a 30 mM borate concentration. Next the pH was optimised. Increasing the pH led to longer retention times. Two values emerged as optimal: 10.56 and 11.12. Owing to the shorter analysis time and lower current, pH = 10.56 was selected. Both methanol and acetonitrile were tested as organic modifier. Acetonitrile gave better results and improved both the retention time and the resolution. The detection limit varied from 1.76-4.56 |Lig/ml and was roughly ten times higher than the corresponding HPLC detection limit. The linear range was 0.6-108 ^ig/ml and 1.3-156 |ig/ml for alizarin and purpurin respectively. The intra-day and inter-day reproducibility ranged between 1 -2% for both alizarin and purpurin. For HPLC these values were below 1% [88]. One can conclude that CZE has some merit for the fast analysis of anthraquinones although further research is necessary to see if it can be successfully employed for the quantitation of madder extractives.

650 STRUCTURE ELUCIDATION The basic antliraquinone structure is depicted in Fig. (6). In the Hterature two different methods are used for the numbering of the carbon atoms of the parent molecule (Fig. (6)). In this review the method that numbers the carbon atoms from 1 till 14 is chosen. The anthraquinones found in Rubia tinctorum differ in the nature of the substituents and the substitution pattern. These substituents are only found on carbon atoms 1, 2, 3 and 4. Hydroxyl and methoxyl groups are frequently encountered as substituent [1,4].

o o

,^'N2

13

J1

J4

^

10

' ^

\ < ^

2

II -^s.ga-s-ga-^i^. 5

M

4

0 Fig. (6). Parent antliraquinone structure and the two different numbering systems

Nowadays research on the structure of compounds relies on the interpretation of spectral information whereas older work was mainly dependent on the preparation of derivatives and on degradative experiments. The parent compound (9,10-anthraquinone) could be identified by zinc dust distillation or zinc dust fusion. The position of the hydroxyl groups could be established by the use of selective reagents. Hydroxyl groups at the 2-position are readily methylated with diazomethane. Chelated 1-hydroxyl groups are resistant to most chemicals but are affected by methyl iodide/silver oxide/chloroform or methyl sulphate/potassium carbonate/acetone mixtures [1]. In the following a brief description is given of the spectral techniques used for the identification of anthraquinones in Rubia tinctorum.

651 Ultraviolet and Visible Light Spectroscopy Ultraviolet and visible light (UV-VIS) analysis is the most used identification teclinique for anthraquinones. The UV-VIS absorption spectrum of an anthraquinone is a combination of the absorptions arising from partial acetophenone- and benzoquinone chromophores (Fig. (7)) [1,20,62]. O O

%:X Fig. (7). Benzoquinone and acetophenone chromophores [20]

The intense absorption within the range 220-230 nm is due to the local excitation (L.E.) bonds of the benzenoid rings of the acetophenone cliromophore. The absorption between 252-258 nm is caused by the electron transfer (E.T.) bond, which in case of the ortho- en meta substituted acetophenone chromophore is found between 240-270 nm. The benzenoid ring and one ketofunction are responsible for this E.T. bond. The absorptions between 265-280 nm and 285-290 nm are due to the E.T. bond of the benzoquinone chromophore. The absorption between 430-437

A JSJ«6I

n»

--

'1

131

—

—V-

—

j

\" \ 2.M.W-.

---

4jan;

"v^

I^

oJ \

\l

-

7^

»32l««

V '

*

Vj

/'

/' ^,±X D

Fig. (8). UV-VIS spectra of alizarin dissolved in MeOH and dissolved in MeOH/KOH

K

\

652 nm is caused by the L.E. bonds of the C=0 of the quinone cliromophore [62]. The UV-VIS absorption spectrum of ahzarin in methanol and in methanol + KOH from 200 to 650 nm is depicted in Fig. (8). Infrared Spectroscopy Anthraquinones are poorly soluble in common IR solvents and in practice most spectra are measured in potassium bromide discs [1]. The Infrared (IR) spectrum of 9,10-anthraquinone shows, due to the symmetrical character of this compound, only one carbonyl absorption at 1678 cm ^ The IR absorption of the carbonyl group is changed by substitution in either ring A or C. The carbonyl frequency is raised by -M substituents and by steric strain, both of which are relatively rare in natural anthraquinones. The frequency of the carbonylgroup is lowered by a +1 or +M group in ring A or C. The commonest substituents are alkyl, hydroxy 1 and alkoxyl. A 1-hydroxy 1 group lowers the frequency of the adjacent carbonyl group much more than the frequency of the other carbonyl group, due the hydrogen bonding between the 1-hydroxy 1 and the adjacent carbonyl group. Therefore the spectrum shows two carbonyl bonds.

'Pi ^

5;'I i

2500 2000 Wavenumber(cm-I)

Fig. (9). IR spectrum of alizarin in KBr

653 Antliraquinones, having no 1-hydroxyIgroup show only one carbonyl peak and its position is a Httle shifted to lower wave numbers by substitutions at C-2 or C-3. The absorption bond of the hydroxy stretch frequency occurs at 3320-3380 cm~^ [62]. The IR spectrum of alizarin in KBr is depicted in Fig. (9). Nuclear Magnetic Resonance Spectra ^H-NMR

^H-NMR spectroscopy is an indispensable tool for structure elucidation of anthraquinones. The ^H-NMR spectrum allows deduction of the nature and the number of substituents because all the substituents will give rise to signals with a characteristic chemical shift [4]. The 1 & 4, and 2 & 3 protons in the unsubstituted ring C of natural anthraquinones give multiplets [1]. The substitution pattern of ring A can be deduced from the splitting patterns observed in the region between 7.0 and 8.5 ppm. The usual coupling constants for aromatic protons are 7-8 Hz for ortho couplings and ca. 1.5 Hz for meta couplings. Also from the chemical shifts observed for the aromatic protons information about their position can be deduced. For a proton at C-1 or C-4 a signal above 7.5 ppm will be seen whereas a proton at C-2 or C-3 will give rise to a signal below 7.5 ppm. Because the two carbonyl groups in the molecule exert a relatively strong deshielding effect on substituents in the 1-position this can be of help in determining the location of substituents. A 1-positioned methoxyl group will give rise to a signal at ca. 4.05 ppm while a methoxyl group in the 2 or 3 position will give a signal at ca. 3.9 ppm. Hydroxyl groups at the 2- or 3- position are not always readily observable. The deshielding effect of hydrogen bonding with the carbonyl groups helps in discriminating between a free hydroxyl group and a hydroxyl group in the 1-position. The signal due to a 1-positioned hydroxyl group will be found between 12 and 13 ppm; the signal due to a free hydroxyl group - if it is seen at all - will be found at ca. 10 ppm [4]. Different solvents (CDCI3, DMS0d6, CD3OD) can be used for NMR analysis. Depending on the solvent the values will usually vary about 0.1-0.2 ppm. The best solvent for the NMR analysis of anthraquinones, in particular the glycosides of Rubia tinctorum, is DMS0d6. The 200 MHz proton NMR spectrum of alizarin in DMS0d6 is depicted in Fig. (10).

654

u 8.2

B.l

7.9

Fig. (10). ^H-NMR spectrum (200 MHz) of alizarin in DMS0-d6 ^^C-NMR

The spectra can be obtained by using ^H noise and noise off-resonance decoupling techniques. The overlapping of certain resonances can be resolved by off-resonance decoupling and by comparison with derivatives [89]. For 1-hydroxyanthraquinones the resonance of C-9 is shifted 5 ppm downfield relative to the carbonyl signal of the parent anthraquinone, as a result of strong intramolecular hydrogen bonding between the hydroxyl group and the carbonyl oxygen. The slight shielding of C-10 (1 ppm) is in part the consequence of the bond dipole moment created by the intramolecular hydrogen bond. For a 2-hydroxyanthraquinone, the slight shielding effect (1.5-ppm) of the C-2 on the C-10 may result from transfer of electron density by cross conjugation between C-2 and C-10. Transfer of electron density by cross conjugation also explains the slight shielding effect of the I-OCH3 on C-9 (0.7 ppm) and of the 2-OCH3 (3 ppm) on C10 [90]. In poly cyclic systems some quaternary carbons are difficult to observe, because of the very low signal intensities (see also Fig. (11)) due to the high value of the relaxation times and the absence of one-bond

655 interactions. To detect the frequencies of all the quarternary carbons, 10 mg of chromium tris(acetylacetonate) (Cr(AcAc)3) was added as a relaxation agent to the solution [89]. The 50 MHz carbon NMR spectrum of alizarin (DMS0d6) is depicted in Fig. (11).

W^wWl^fr^t^^^i^ifri^^^

^IMH/W^*^ vJwM

Fig. (11). '^C-NMR (50 MHz) spectrum of alizarin in DMS0-d6

Mass Spectrometry For aglycones the molecular ion almost invariably forms the base peak. Mass spectra of the 9,10-anthraquinone itself shows successive elimination of two molecules of carbon monoxide with strong peaks at m/z 180 (M-CO) and 152 (M-2 CO) (and strong double charged ions at m/z 90 and 76) which correspond to the molecular ions of fluorenone and biphenylene respectively ((Fig. (12)).

^^^^ X .

X^

Fig. (12). Fluorenone and biphenylene

r.^^^%,

^ ^ ^

Vx-'

x ^

656 Otherwise there is very little fragmentation. The spectra of substituted anthraquinones follow the same pattern with corresponding peaks appropriate to the substituents and their position. 2Hydroxyanthraquinone shows more intense M-CO and M-OH peaks than does 1-hydroxy-anthraquinone. Both hydroxyanthraquinone spectra have a peak at m/z 140 corresponding to the loss of three molecules of carbon monoxide, the third arising from the phenolic group Dihydroxyanthraquinones behave similarly and a peak at M - 4 CO+ appears. A I-OCH3 group gives rise to [M-OH]"^ and [M-H20]^ peaks which are not observed in 2-OCH3 derivatives [1]. The mass spectrum of alizarin is depicted in Fig. (13). I E+ 07 7.55

212 51 63 ,!

, !•

77 ,

•!--

92 105 .!

.. !•

23^1

128 138 155 I

.li

L,

. !. ,

Fig. (13). Mass spectrum of alizarin

SECONDARY METABOLITES OCCURRING IN RUBIA TINCTORUM The most important components of Rubia tinctorum are the anthraquinones. The basic anthraquinone structure is depicted in Fig. (6). The anthraquinones found in Rubia tinctorum differ in the nature of the substituents and the substitution pattern. Due to the biosynthetic route of anthraquinones {vide supra) in Rubiaceae these substituents are only found on carbon atoms 1, 2, 3 and 4. A hydroxylgroup is frequently encountered as substituent. Alizarin is the most well loiown anthraquinone of madder. In 1826 alizarin was first isolated from Rubia tinctorum by Colin and Robiquet [91]. In 1868 Graebe and Liebermann deduced the structure of alizarin by zinc dust destination. They verified their hypothesis by synthesis of alizarin. In 1869 they patented an improved chemical synthesis of alizarin.

657 After the first isolation of alizarin a lot of other anthraquinones were isolated from Rubia tinctorum for example purpurin, munjistin, rubiadin, pseudopurpurin, nordamnacanthal, lucidin, xanthopurpurin and anthragallol. Anthraquinones are primarily present as glycosides in Rubia tinctorum. Alizarin is mainly present as the glycoside ruberythric acid. Ruberythric acid was first isolated in a crystalline form by Rochleder in 1851 [92]. Ruberythric acid consists of the aglycone alizarin and a disaccharide. Many years later the sugar moiety was identified as primeverose, a disaccharide of xylose and glucose [93]. Lucidin is another anthraquinone in madder that is mainly present in its glycosidic form. The sugar moiety is also primeverose. In the 19^'^ century and early 20^^^ century literature the occurrence of additional glycosides was reported. However in more recent literature these findings have not been confirmed. In recent literature only ruberythric acid and lucidin primeveroside occur. It is not clear if these additional glycosides are really present in Rubia tinctorum or if these were confused with ruberythric acid and lucidin primeveroside. According to Von Wiesner in 1927 only ruberythric acid and rubiadin glucoside were isolated in pure form by respectively Rochleder in 1851 and Schunk and Marchlewski in 1893 [10,92,94]. Schweppe enumerate in his review the glycosides: galiosin (pseudopurpurin 1-B-primeveroside), rubiadin 3-Bprimeveroside, rubiadin glucoside and rubianin (a C-glucoside) [3,8,10,95]. In Ullman's Encyclopedic [96] rubiadin 3-B-D-glucoside and galiosin (pseudopurpurin B-primeveroside) are reported as glycosides. Recently three new glycosides were found in Rubia tinctorum: lucidin glucoside, 2-hydroxymethylanthraquinone 3-glucoside and 3,8-dihydroxy2-hydroxymethylanthraquinone [55]. Because the proposed biosynthetic route for anthraquinones in Rubia tinctorum does not allow substitution in the C-ring the isolation of the latter compound needs to be reconfirmed. Some anthraquinones isolated from Rubia tinctorum are believed to be artefacts for example the anthraquinones which show the presence of a 2methoxymethyl or 2-ethoxymethyl group. These anthraquinones have been formed during the extraction of lucidin with boiling methanol or ethanol [4,68,97]. According to Schweppe the anthraquinones purpurin (1,2,4-trihydroxyanthraquinone) and purpuroxanthin (1,3-dihydroxyanthaquinone) are formed from respectively pseudopurpurin (3-carboxy1,2,4-trihydroxyanthraquinone) and munjistin (2-carboxy-l,3-dihydroxyanthraquinone) during drying of the roots [3]. Some anthraquinones were only isolated once from Rubia tinctorum. It is thus doubtful whether these

658 anthraquinones are really present in madder, especially if there are few spectral data available as for example with quinizarin-2-carboxylic acid. All the antliraquinones which have been isolated so far from Rubia tinctorum are listed in alphabetical order below. Structures and synonyms are given. References referring to isolations and spectral data are included as well. Of the few non-anthraquinone type of compounds isolated from Rubia tinctorum only the name, compound class and a reference referring to their isolation are given.

Parent anthraquinone structure Alphabetical Listing of All Anthraquinones Isolated from Rubia: Alizarin (1,2-Dihydroxyanthraquinone) Structure: Ri = R2 = OH, R3 = R4 = H Isolation: [25-27,47,52,53,55,58,61,63,64,66,70,71,77,78,98,99] UV: [3,54,55,73] IR: [1,54,55,73] MS: [54,71], MW = 200, C,4H804 ^H-NMR: [54,55,73] *^C-NMR: [55,73,89] Alizarin dimethylether (1,2-Dimethoxyanthraquinone) Structure: Ri = R2 = OMe, R3 = R4 = H Isolation: [53] UV: [1] IR: [100,101] MS: [101], MW = 268, Ci6H,204 'H-NMR:

[101]

Alizarin 1-metliylether (2-Hydroxy-l-methoxyanthraquinone) Structure: Ri = OMe, R2 = OH, R3 = R4 = H Isolation: [53,78]

659 UV: [1,54] IR: [1,54] MS: [54], MW = 254, C15H10O4 ' H - N M R : [54,102] Alizarin 2-methylether (l-Hydroxy-2-methoxyanthraquinone) Structure: Ri = OH, R2 = OMe, R3 = R4 = H Isolation: [52,53,66] UV: [1,52,103] IR: [1,103] MS: [52,103], MW = 254, C15H10O4 'H-NMR:

[103]

'^C-NMR: [89] Anthragallol (1,2,3-Trihydroxyanthraquinone) Structure: Ri = R2 = R3 - OH, R4 = H Isolation: [53,66] UV: [1,104] IR: [1,104] Anthragallol 2,3-dimethylether (1 -Hydroxy-2,3-methoxyanthraquinone) Structure: Ri = OH, R2 = R3 = OMe, R4 = H Isolation: [53] UV: [1] IR: [1,105] 'H-NMR:

[105]

Anthragallol 3-methylether (1,2-Dihydroxy-3-methoxyanthraquinone) Structure: Ri = R2 = OH, R3 = OMe, R4 = H Isolation: [53] Christofin (1,4-Dihydroxy-2-ethoxymethylanthraquinone) Structure: R, = R4 = OH, R2 = CH20Et, R3 = H Isolation: [61] IR: [61] MS:[61],MW = 298,Ci7Hi405 'H-NMR:

[61]

2-Hydroxyanthraquinone Structure: Ri = R3 = R4 = H, R2 = OH

660 Isolation: [53,66] UV: [1,106] IR [107,108] MS: [109], MW = 224, CnHgOa 'H-NMR:

[108]

'^C-NMR: [90] l-Hydroxy-2-hydroxymethylanthraquinone Structure: Ri = OH, R2 = CH2OH, R3 = R4 = H Isolation: [54,110] UV: [54] IR: [54] MS: [54], MW = 254, C15H10O4 ' H - N M R : [54,102] 1 -Hydroxy-2-methylanthraquinone Structure: Ri = OH, R2 - Me, R3 = R4 = H Isolation: [52,53,64] UV: [73,111,112] IR: [73,112] MS: [73,112], MW = 238, C15H10O3 ' H - N M R : [73,112] '^C-NMR:[113] 2-Hydroxymethylanthraquinone-3-0-B-D-glucoside Structure: Ri = R4 = H, R2 = CH2OH, R3 = O-B-D-glucoside Isolation: [55] UV: [55] IR: [55] MS: [55], MW = 416, C21H20O9 'H-NMR:

[55]

"C-NMR: [55] 2-Hydroxymethyl-8-hydroxyanthraquinone-3-0-B-D-glucoside Structure: Ri = R4 = H, R2 = CH2OH, R3 = O-B-D-glucoside, at"C8: OH Isolation: [55] UV: [55] IR: [55] MS: [55], MW = 432, C21H20O10 'H-NMR:

[55]

661 2-Hydroxymethylquinizarin (1,4-Dihydroxy-2-hydroxymethylanthraquinone) Structure: Ri = R4 = OH, R2 = CH2OH, R3 = H Isolation: [3,110] MS:[110],MW = 270, CsHioOs 7-Hydroxytectoquinone (2-Hydroxy-7-methylanthraquinone) Structure: Ri = R3 = R4 = H, R2 = OH, at C7: -Me Isolation: [52,59] UV: [59] IR: [59] MS: [59], MW = 238, Ci5H,o03 'H-NMR:

[59]

'^C-NMR: [59] Lucidin (1,3-Dihydroxy-2-hydroxymethylanthraquinone) Structure: Ri = R3 = OH, R2 = CH2OH, R4 = H Isolation: [47,52,53,64-66,70,78] UV: [30,65,85,97] IR: [30,65,85,97] MS: [65,97], MW = 270, CijHioOs ' H - N M R : [30,65,85] '^C-NMR: [30,31,73] Lucidin co-ethylether (l,3-Dihydroxy-2-ethoxymethylanthraquinone) Structure: R, = R3 = OH, R2 = CH20Et, R4 = H Isolation: [52,55] UV: [55,73,97] IR: [55,97] MS: [55,73,97], MW =298, CIVHMO., ' H - N M R : [55,73,97,114] '^C-NMR: [55,73] Lucidin glucoside (1 -Hydroxy-2-hydroxymethylanthraquinone-3-0-B-Dglucoside) Structure: R| = OH, R2 = CH2OH, R3 = 0-B-D-glucoside, R4 = H Isolation: [55] UV: [55,64] IR: [55,64]

662 MS: [55], MW = 432, CiiHsoOio 'H-NMR:

[55]

'^C-NMR: [55] Lucidinco-methylether(l,3-Dihydroxy-2-methoxymethylanthraquinone) Structure: R, = R3 = OH, R2 = CH20Me, R4 = H Isolation: [52] UV: [52,54,114] IR: [52,54,114] MS: [52,54,114], MW = 284, CieHnOj ' H - N M R : [52,54,102,114] '^C-NMR: [102] Lucidin primeveroside (1 -Hydroxy-2-hydroxymethylanthraquinone-3 -OB-D-primeveroside) Structure: Ri = OH, R2 = CH2OH, R3 = O-B-D-primeveroside, R4 = H Isolation: [52,55,64,66,70] UV: [55,73,85] IR: [55,73,85] MS: [55,73], MW = 564, C26H28O14 'H-NMR: [55,73,85] '^C-NMR: [30,55,73] 2-Methoxyanthraquinone Structure: R, = R3 = R4 = H, R2 = OMe Isolation: [53] UV: [1] IR: [100] MS: [109], MW = 238, C15H10O3 'H-NMR: [100] '^C-NMR: [90] l-Methoxy-2-methylanthraquinone Structure: Ri = OMe, R2 = Me, R3 = R, = H Isolation: [53] UV:[115] IR:[115] 'H-NMR: [116]

663 Munjistin (2-Carboxy-1,3-dihydroxyanthraquinone) Structure: Rj = R3 =0H, R2 = COOH, R4 = H Isolation: [53,77,117] UV: [1] IR: [1,118] MS: [119], MW = 284, dsHgOf, 'H-NMR:[119]

'^C-NMR:[119] Munjistin methylester (2-Carboxymethyl-l ,3-dihydroxyanthraquinone) Structure: Ri = R3 = OH, R2 = COOMe, R, = H Isolation: [52,60] UV: [60,120] IR: [60,120] MS: [60,120], MW = 298, CicHioOft ' H - N M R : [60,120] Nordamnacanthal (1,3-Dihydroxy-2-formylanthraquinone) Structure: R, = R3 = OH, R2 = CHO, R4 = H Isolation: [48,52,60,78] UV: [112] IR:[112] MS: [48,112], MW = 268, Cj.sHsOs 'H-NMR: [48,97,112] Pseudopurpurin (2-Carboxy-1,3,4-trihydroxyanthraquinone) Structure: Ri = R3 = R, = OH, R2 = COOH Isolation: [27,53,64,77] UV: [1,121] IR: [1,121] Purpurin (1,2,4-Trihydroxyanthraquinone) Structure: Ri = R2 = R4 = OH, R3 = H Isolation: [27,47,53,61,64,77,110] UV: [1] MS: [122], MW = 256, CnHgOj 'H-NMR: [123], 5(DMSOd6) 6.62 (IH, s, H-3), 7.86-7.90 (2H, m, H-6, H-7), 8.10-8.15 (2H, m, H-5, H-8), 11.62 (IH, br s, OH), 13.07 (IH, s, OH), 13.34 (IH, s, OH) [Derksen, unpublished]

664 •^C-NMR: 6(DMSOd6) 110.7, 114.7, 117.3, 131.5, 137.4, 138.3, 139.1, 140.0, 154.4, 162.1, 165.5, 188.2, 191.6 [Derksen, unpublished] Quinizarin (1,4-Dihydroxyanthraquinone) Structure: Rj = R4 = OH, R2 = R3 = H Isolation: [3,61,110] UV: [1,124] IR: [125] MS: [109], MW = 240, C,4H804 *H-NMR: [126] *^C-NMR: [90,127] Quinizarin-2-carboxylic acid (2-Carboxy-l ,4-dihydroxyanthraquinone) Structure: Ri = R4 = OH, R2 = COOH, R3 = H Isolation: [110] UV: [128] Ruberythric acid (l-Hydroxyanthraquinone-2-O-B-D-primeveroside) Structure: Ri = OH, R2 = 0-6-D-primeveroside, R3 = R4 = H Isolation: [52,64,66,70,92] UV: [73,121] IR: [73,121] MS:[73],MW = 534,C25H260,3 ' H - N M R : [73]

'^C-NMR: [73] Rubiadin (1,3-Dihydroxy-2-methylanthraquinone) Structure: Ri = R3 = OH, R2 = Me, R, = H Isolation: [27,52,53,77] UV: [54,85,97,103] IR: [52,54,85,103] MS: [52,54,103], MW = 254, C,5H,o04 ' H - N M R : [52,54,85,103] Rubianin (1 -Hydroxyanthraquinone-2-C-6-D-glucoside) Structure: Rj = OH, R2 = -C-B-D-glucoside, R3 = R4 = H Isolation: [129] UV: [129] '^C-NMR: [73]

665 Tectoquinone (2-Methylanthraquinone) Structure: Ri = R3 = R4 = H, R2 = Me Isolation: [52,60] UV: [4] IR: [4] MS: [4], MW = 222, CisHioOz ' H - N M R : [130,131] '^C-NMR:[131] Xanthopurpurin (1,3-Dihydroxyanthraquinone) Structure: Ri = R3 = OH, R2 = R4 = H Isolation: [52,53,64,77] UV: [1,52] IR: [1,52] MS: [52,132], MW = 240, C14H8O4 ' H - N M R : [30,52,132] '^C-NMR: [52,90] Xanthopurpurin dimethylether (1,3-Dimethoxyanthraquinone) Structure: Ri = R3 = OMe, R2 = R4 = H Isolation: [53] UV: [1] IR: [1] ' H - N M R : [30]

'^C-NMR: [90] Xanthopurpurin 3-methyIether (l-Hydroxy-3-methoxyanthraquinone) Structure: Ri = OH, R2 = R4 = H, R3 = OMe Isolation: [53] UV: [133] IR: [133,134] 'H-NMR: [134] '^C-NMR: [89] Non-anthraquinone compounds Carboxylic acids Citric acid [135,136]

666 Coumarins Scopoletin [52] Cinnamic acid derivatives Hexadecyl ferulate [52] Octadecyl ferulate [52] Eicosyl ferulate [52] Flavonoids Hyperoside [137] Rutin [137] Iridoids Asperuloside [55,138] Asperulosidic acid [138] Monotropein [138] Naphtoquinones Lapachol methylether [52] Miscellaneous Mollugin [52] DYEING WITH MADDER Compounds Responsible for the Dyeing Effect The roots of Rubia tinctorum have been used for dyeing textiles in many parts of the world since ancient times. Madder was widely cultivated in Western Europe for the dye industry until the beginning of the twentieth century. Rubia tinctorum contains useful anthraquinone mordant dyes. Dried roots of madder contain the hydroxy anthraquinones alizarin, pseudopurpurin, rubiadin, purpurin, purpuroxanthin and some minor anthraquinones. Anthraquinone derivatives are good mordant dyes if they satisfy the following conditions: 1. The anthraquinone has a hydroxyl group at C-1 or C-4 position next to one of the carbonyl groups.

667 2. The hydrogen bonding between the carbonyl and the hydroxy 1 at CI or C4 must be weakened by a substituent at C-3 with a -I or -M effect, or a substituent at C-2 with a +1 or +M effect [20]. Thus anthraquinones in madder with only one free phenohc group are of no dyeing importance [3]. The next paragraph describes the dyeing with madder. At the end of the nineteenth century alizarin could be produced synthetically. As the colouring capacity of alizarin was very similar to that of dried madder roots, the cultivation of madder quickly came to an end and only synthetic alizarin was used for dyeing textile [3]. Theory and Practice of Dyeing with Madder A lot of different formulas for dyeing with madder have been described in the literature. The recipes can be divided in two main classes, according to the origin of the material to be dyed, the number of process steps and necessary chemicals. 1. The Alizarin red procedure, for dyeing animal derived fibres such as wool 2. The Turkish red procedure, for dyeing plant derived fibres such as cotton [3] In a lot of modern handbooks on dyeing, the difference between these procedures is not taken into account [13]. But actually it is as complicated to dye vegetable fibres like cotton, flax or hemp with madder derived dyestuff as it is uncomplicated to dye an animal fibre with madder dyestuff. In contrast with cotton, wool fibre can be dyed red that is hearty and has a good chroma with a much easier dyeing recipe using only alum as mordant. For madder dyeing on a plant-derived fibre like cotton yarns a much more complicated recipe has to be used with a large number of dyeing steps and mordant components [3,13,20]. This must be due to the different composition of both fibres. Cotton (a vegetable fibre) consists of 94% cellulose [3], which is composed of B-glucopyranosyl residues joined by l--> 4 linl<:ages. Wool (an animal fibre) consists of the protein a-keratin [139]. In the Alizarin red procedure the main steps are pre-treatment, mordanting, dyeing and washing of the textile. Between the mordanting and the dyeing step the wool does not have to be dried. [3]. In the Turkish red procedure the following steps can be distinguished:

668 1. Pre-treatment 2. Oiling 3. Treatment with tannic acid 4. Treatment with mordant 5. Fixation of the mordant 6. Dyeing 7. Washing [20] 1. The pre-treatment; The yarn or cloth is cleaned from fatty and waxy components by treatment of the textile with soda or sodium hydroxide solution. Often the textile is also bleached, for example with chlorine [20]. 2. Oiling; The fibre is repeatedly treated with Tournant oil (rancid olive oil). The oil has to be fixed so the textile is dried between the various oiling steps. In 1872 this procedure was optimised by the introduction of a new oil and a continuous drying process. The new oil known as Turkish red oil consisted of castor oil made soluble in water by sulphonation with sulphuric acid [13,20]. Before the oiling step the cloth was steamed under an overpressure of 0.5-2 atm for a couple of hours [20]. There is no widely accepted explanation available why an oiling step improves the colour of the dyed textile [13,20]. 3. Treatment with tannic acid; This treatment is supposed to have some advantage as fixing agent for the alum mordant, which is applied in the next step [20]. 4. Treatment with Mordant; The mordant is essential for the dyeing process of the cloth or yarn in both the Turkish red and alizarin procedure. The mordant is necessary as fixative for the dye. Before 1750 alum was the main mordant used. After 1750 other mordants like basic aluminium sulphate and soda or chalk, aluminium acetate and later on aluminium sulphoacetate were used [20]. 5. Fixation of the mordant; The fixation is mostly achieved by drying of the textile at low temperature. Kiel showed that the formation of A1(0H)3 out of aluminium acetate mordant and sulphonate mordant is essential for the fixation. He demonstrated that the deposition of A1(0H)3 in the fibres was maximal if the pH lies between 4 and 9 during the fixation. At pH < 4 or pH > 9 the A1(0H)3 dissolves again. Kiel found that fixation of the textile at 40°C for 30 min or 1 min at 70°C in 3 ml sodium silicate solution (or 10 g chalk/L) are the best conditions [20].

669 6. Dyeing; The stained cloth is dyed with madder or synthetic ahzarin. A colour complex is formed between the dye (alizarin), the mordant (aluminium) and the calcium. The water bath that contains the textile and the dye is slowly heated. Finally the cloth is briefly boiled in the dye bath [20]. 7. Washing; The excess dye and other used compounds are removed from the cloth. The cloth is rinsed thoroughly in flowing water, or the cloth is washed with soap [20]. Over the years a lot of dyeing procedures have been proposed to speed up the dyeing process. Most of these procedures combine two or more steps of the Turkish red procedure. Up to now all these procedures give a qualitatively less dyed product. A lot of other mordants have also been used in the past to change the dyeing result. Most of these mordants do not increase the quality of the dyed product but merely gave a different coloured product. For example the use of the alternative mordant FeS04 -7 (-)

O(-) HO—^/^---HoO

Fig. (14). Structure of the dye complex suggested by Kiel [20]

Ca""^' 2H2O

670 H2O results in a darker and blacker colour. Copper gives a warmer and deeper colour. If the textile is placed in a tin bath after drying, the colour turns more yellow. It is assumed that the Sn(II) ions displace some of the calcium ions. This process is named "aviveren" in Dutch [20]. In the past a lot of authors have proposed a structure for the coloured complex that will be formed during the dyeing process. It was already known before 1868 that besides dye and mordant also calcium is essential for the complex. The pKa values of the alizarin hydroxyl groups at position 1 and 2 are respectively 12.0 and 8.2 making the 2-hydroxyl group the more acidic one. During the formation of the dye complex a calcium ion reacts with the 2-hydroxyl while an aluminium ion forms a complex between the 1-hydroxyl and the carbonyl group. The ratio between both alizarin and aluminium and alizarin and calcium is 2:1 [20]. After analysing IR spectra of alizarin-metal ion complexes Kiel suggested the structure depicted in Fig. (14). The model showed a complex of A P ^ chelated in the li-ketol (C-1 or C4) position of the alizarin molecule. Thirty-five years later Soubayrol et al

Azi

0\

4Na'

Fig. (15). Structure for the dye complex suggested by Soubayrol et al. [140]. For an explanation of the dashed lines see Fig. (16) and main text.

671 showed with ^'^Al NMR in the soHd state that the complex is a binuclear co-ordination complex (Fig. (15)) of the structure proposed by Kiel [140]. The binuclear species consists of a tetra anion skeleton formed around two Al^^ ions linked by two hydroxyl bridges [140]. Depending on the nature of the cation and the drying conditions the complex is surrounded by a variable number of water molecules, [Al2(|Li-OH)2Na2(Ci 41^504)4 (H20)4](H20)2(Na)2 and [Al2(^-OH)2Ca2(Ci4H604)4(H20)4](H20)2 respectively. The four-alizarinate ligands are probably bound with their eight free oxygen atoms to the four H2O molecules as shown in Fig. (15).

Fig. (16). Three-dimensional structure of one half of the binuclear dye complex suggested by Soubayrol e^ Of/. [140]

Alizarin molecules Azl and Az3 partially overlap each other with two water molecules in front of the complex. Similarly molecules Az2 and Az4 overlap each other with the two remaining water molecules at the rear of the complex. The alizarin entities, which overlap each other with two benzene rings form a sandwich structure inside of which two of the Na^ or two Ca^^ cations are probably confined (Fig. (16)). The ionic radii of Na"^ and Ca^"^ fit nicely into a benzene hexagon of 2.4 A diameter, which therefore supports the formation of (C6H6)2Na sandwich subunits. Deformation of the four 0-Al bonds in the core is probably responsible for the decrease in the apparent co-ordination number of Al in the closed complex, (a lot of strain for including the bulk of the ligands) leading to an atypical chemical shift for ^^Al near 24 ppm. If instead K"^ or Ba^"^ are used as ions the complexes [Al2(|i-OH)2(Ci4H604)4(H20)]

672 (H20)5(K)4 and [Al2(^-OH)2(Ci4H604)4(H20)3] (H20)3(Ba)2 are formed. These are called binuclear open structure complexes. The bulkier K"^ or Ba^"^ cations are not able to enter the alizarin sandwich structure and rather than having Al-O-Al bridges that are too strained, the structures remain open and the water molecules are no longer necessary. In this case, the co97

ordination number six gives standard Al chemical shifts near zero and probably standard 0-Al bond lengths near 1.86 A as in [(Al2)^-OH)2]'^^ core complexes [140]. In the case of Ba^"^, the stable complex is a trihydrate, with probably two alizarinates connected by two water molecules, the other two ligands remaining free [140]. Inspection of molecular models shows that the two water molecules of each dibenzene sandwich may be replaced by a cellobiose entity representing a unit of the cellulose fibre. This possibility provides a good explanation of the strong fixation of this dye on cotton fabric [140]. BIOLOGICAL ACTIVITY AND NON-DYEING APPLICATIONS Biological and Pharmacological Activities of Rubia Extracts Antioxidant activity of alizarin in vitro and in vivo

The antioxidant activity of alizarin was established in four different assays: (1) suppression of light emission in the /^-iodophenol enhanced chemiluminescent assay, (2) scavenging of superoxide anion (02*~) in a hypoxanthine-xanthine oxidase system, (3) protection of rat liver microsomes from lipid peroxidation by ADP/iron(II) ions, and (4) protection of bromobenzene-intoxicated mice from liver injury in vivo [141]. Alizarin was compared with Trolox (water soluble vitamin E), the flavonoid baicalin and green tea proanthocyanidins. In assay (1) the activity of alizarin was 76% of that of Trolox. In assay (2) the inhibition of 02'~-induced chemiluminescence was 40%, 32%), 23%) and 14%) for Trolox, alizarin, green tea polyphenols and baicalin respectively. Alizarin was not significantly active in the lipid peroxidation assay but after baicalin the most active compound in the in vivo assay. This shows again the difficulty in the evaluation of antioxidant activity and the differences between in vitro and in vivo assays [141].

673 Antimicrobial activity

Rath et al. have studied the antifungal activity of ten anthraquinones occurring in Morinda lucida [142]. Several of these anthraquinones also occur in madder. As bioassay during the fractionation they used the activity against Cladosporium cucumerinum and Candida albicans in TLC bioautography. Of the compounds tested only 3-hydroxy-2-formylanthraquinone, nordamnacanthal, damnacanthal and alizarin 1-methyl ether showed significant activity against both fungi. Of these four compounds the latter one was the most active with detection limits at 0.5 and 1 |ig against Cladosporium cucumerinum and Candida albicans respectively. Alizarin 1-methyl ether was the only active compound in agar dilution assays when tested against the human pathogenic fungi C albicans, Aspergillus fumigatus and Trichophyton mentagrophytes. The minimum inhibitory concentrations were 10, 100 and 50 |Lig/ml respectively [142]. The antibacterial activity of aqueous (glycosides) and ether (aglycones) extracts of madder was assessed against Salmonella typhi, S. paratyphi A, S. paratyphi B, S. paratyphi C, S. enteritidis, S. typhi-murium. Shigella flexneri 2a, Sh sonnei, Sh. largei-sachsii, Sh. boidii, Sh. ambigua, Pseudomonas aeruginosa, Proteus vulgaris, Escherichia coli O55, Staphylococcus aureus. Streptococcus haemolyticus. Neisseria gonorrhoeae and Mycobacterium tuberculosis var. hominis [143]. The diffusion and dilution methods were used. The aqueous extract at 2% showed only activity against Sh largei-sachsii. The ether extract had a broader and stronger activity. It was active against Sh. largei-sachsii, Staphylococcus aureus and Streptococcus haemolyticus. Also the antifungal activity was determined. Both madder powder and the total aglycones showed activity against Candida albicans, Geotrichum candidum, Geotrichum louberi, Rhodotorula rubra, Rhinoclaviella sp. and Saccharomyces cervisiae in an inliibition zone assay. At 0.4 mg inhibition zones against all species varied from 11-25 mm. An aqueous extract as well as the aglycones in sunflower oil were not active against any fungus except the Rhinoclaviella species [143]. Activity against flagellates

Chagas's disease caused by the flagellate Trypanosoma cruzi is currently becoming more and more of a problem due to its wider occurrence in

674 Latin America and the lacic of good drugs. With the aim of developing new safe and efficient chemotherapy, 45 compounds were tested. Of the natural compounds tested, purpurin was the most effective one with an ID50 of 358 ± 35 |LiM. It was active against bloodstream forms also at 4°C and will be further tested for its usefulness for chemoprophylaxis in donated blood [144]. Antilithiasis activity

In rats oral intake of fresh madder root (10% of the food) decreased stone formation in bladder and kidney induced by 3% calcium carbonate [145]. In rabbits, oral intake of madder root extract (150 mg/kg) caused decreased calcium oxalate crystallization in the kidney. An increased death rate was observed in feeding experiments of rats. Furthermore, feeding experiments with rabbits showed hepatotoxic effects. Genotoxic effects were observed in bacterial in mammalian cell test systems [145]. Medicinal Uses of Rubia Extracts Phytopharmaceutical Uses

Extracts of madder root {Rubia tinctorum) contain some compounds of pharmacological interest. Crude extracts have been used for the treatment of bladder and kidney stones, especially those consisting of calcium oxalate and calcium phosphate in the urinary tract. In vitro experiments showed that ruberythric acid prevented the formation of calcium phosphate and calcium oxalate [146]. Extracts of madder roots have been used as ingredients of phytopharmaceuticals [4,39,46,57,61,63,64, 66,68,71,146]. Rubia Teep® tablets from Madaus (Koln, Germany) are an example of a madder phytopharmaceutical. With regard to the effect against stones no clinical trials exist [145]. Thus in view of the potential toxicity a negative advice about the medicinal use of madder was released by the German Commission E. Madaus withdrew Rubia Teep® from the market in 1990 due to the possible mutagenic properties of lucidin {vide infra).

675 Use as Diagnostic Tool

Alizarin can be used to stain calcium deposits in soft tissues. Dermatopathologists use it to detect dermal calcium in disorders such as pseudoxanthoma elasticum and calcinosis cutis [147]. It harmlessly stains also living tissues. Bones of animals that ate madder turned pinlc or red. The active compound responsible for the staining is supposed to be pseudopurpurin [147]. Alizarin is also used diagnostically as a marker for the study of bone growth. Alizarin and radioactive calcium are deposited similarly in growing bones, including those of the skull. The results that can be obtained by the use of alizarin or radioactive calcium are fully complementary. The hazard of the radioactive calcium to both the investigator and the patient does not arise from the use of alizarin [1,148]. Mutagenic and Carcinogenic Activity of Rubia tinctorum Extracts Because of the application of Rubia tinctorum extracts in phytopharmaceuticals and as food colourants studies on the safety of these products have been carried out. Brown and Dietrich were the first to report lucidin to be mutagenic [79]. Later different authors confirmed that this anthraquinone shows mutagenicity in several strains of Salmonella typhimurium [57,65,79,84,149-152]. The mutagenicity of lucidin was also tested in a battery of genotoxicity assays: mutagenicity in bacterial cells [65] and mammalian cells [65,68,79,84,153], induction of DNA repair in primary rat hepatocytes [68] and in vivo transformation of C3H/M2-mouse fibroblasts [68,79,154,155]. Lucidin induced unscheduled DNA synthesis in primary rat hepatocytes and transformed C3H/M2 mouse fibroblasts [57,79,84,153]. The genotoxic effect of lucidin is a matter of some concern in products containing Rubia extracts. For example, Rubia Teep® tablets contain small amounts of lucidin and lucidin primeveroside [66] and products with madder root extract are used as a food colourant. The glycoside lucidin primeveroside (a compound of madder root) is metabolised in rats to the genotoxic lucidin [57,84,150,153] and rubiadin [156,157]. In a thorough study Kawasaki et al (1992) elaborated that the mutagenicity of madder is not exclusively due to lucidin. Dried roots of

676 madder were extracted with different solvents and these extracts were fractionated by chromatography. Twenty compounds were isolated from the roots of Rubia tinctorum and these compounds were tested for their mutagenicity in Salmonella typhimurium strain TA 100 and/or TA98. 1Hydroxy-2-methylanthraquinone, lucidin-co-methylether, rubiadin, xanthopurpurin, 7-hydroxy-2-metliyl-antliraquinone, lucidin, lucidin-coethylether, lucidin primeveroside and the non-antliraquinone compound mollugin showed mutagenicity [52,84]. Mollugin is a direct mutagen. It is suggested that the phenolic hydroxyl group of mollugin reacts with oxygen and forms a phenoxyl radical and superoxide anion. Oxygen radicals react with guanine at the C-8 position [52]. Kawasaki et al further studied the mutagenicity of 25 anthraquinones to determine the structure-mutagenicity relationship. Lucidin and the alkoxy derivatives (lucidin-co-methylether and lucidin-co-ethylether) as such showed mutagenicity. Rubiadin that can be regarded as lucidin reduced at the hydroxymethyl group, showed mutagenicity only after metabolic activation. They concluded that 1,3-dihydroxyanthraquinones possessing a methyl or hydroxymethylgroup on carbon-2 show mutagenicity. An oxygenated state of the benzylic carbon-2 is essential for direct mutagenicity [52]. Lucidin-3-O-primeveroside showed mutagenicity towards Salmonella typhimurium TAIOO in the absence of metabolic activation [158]. When the glycoside was treated with hesperidinase during the preincubation period, it became more active, hi that case it was active both in the presence and absence of the S9 metabolic activation mix. This pointed in the direction of lucidin as the direct and indirect mutagen. When the preincubation period was prolonged, higher mutagenicity was found confirming the hypothesis about lucidin as the responsible compound [158]. The data confirmed earlier findings about the mutagenicity of lucidin. These mutagenic studies showed that lucidin can be metabolised to a reactive compound which forms covalent adducts with DNA and possibly other macromolecules. It was reported that lucidin forms ethers and esters upon heating with alcohol or acids. This supports the reactive character of lucidin [57,97,153]. Kawasaki et al. (1994) proved that lucidin forms adducts with the nucleic acids adenine and guanine under physiological conditions [159]. These adducts were identified as condensed reactants at the benzylic position of lucidin with a nitrogen atom of a purine base. This

677 indicated the formation of an exomethylenic compound as an electrophilic intermediate. Poginsky et al (1991) also suggested the formation of an electrophilic intermediate (Fig. (17)) that could react with DNA [57]. 0

OH CH2OH

sulphotransferase P

OH .CHsOSOaH

Fig. (17). Suggested formation of electrophilic intermediates out of lucidin that could form adducts with DNA [57,159]

To elucidate the possible carcinogenicity of madder roots, three groups of rats received either a normal diet or a diet supplemented with 1% or 10% madder for more than two years [160]. After this period all surviving animals were sacrificed and their organs studied. Weight gain and morbidity were not different among the three groups. Non-neoplastic lesions related to the treatment were evident in the liver and kidneys of both sexes. Moreover, dose-dependent increases in benign and malignant tumour formation were observed in the liver and kidneys of treated animals. ^^P-Post-labelling analysis showed an increase in the overall level of DNA adducts observed in the liver, kidney and colon of rats treated with 10% madder root in the diet for two weeks. HPLC analysis of 32p-labelled DNA adducts revealed a peak co-migrating with an adduct obtained after in vitro treatment of deoxyguanosine-3'-phosphate with lucidin. These observations implied that the long-term medicinal use of madder by humans is associated with the risk of formation of malignant tumours [160]. Use of Rubia Extracts as Food Colourants Another application of madder extract is its use as food colourant. Natural food colourants are used rather than synthetic ones, because of a consumer preference for natural products. It is widely believed that natural

678 colourants are generally safer than synthetic ones. Madder root extract has been used as a food colourant in confections, boiled fish and soft drinks in Japan due to their colours with distinctive heat and light resistant properties. Madder root anthraquinones turn purple after reacting with proteins in foods [66,71,154,161]. ABBREVIATIONS 2,4-D AD ATCC B.A.S.F. CPC DCCC DMSO DNA E.T. G.I. GLC HPLC lAA I CI IR L.E. MeOH MHz MS MS-medium NAA NMR 0V-17 SE-30 Si02 SPE TLC TMS UC-W98 UV

= = = = = = = = = = = = = = = = = = = = = = = = = = = = = =

(2,4-dichlorophenoxy)acetic acid anno domini fungal strain code chemical company in Germany centrifugal partition chromatography droplet counter cliromatography dimethyl sulphoxide deoxyribonucleic acid electron transfer growth index gas liquid chromatography high pressure liquid chromatography indoleacetic acid chemical company in England infrared local excitation methanol megahertz mass spectrometry Murashige and Skoog medium 1-naphthaleneaceticacid nuclear magnetic resonance type of stationary phase for gas chromatography type of stationary phase for gas chromatography silica gel solid phase extraction thin layer chromatography trimethylsilyl typeof stationary phase for gas chromatography ultraviolet

679 VIS VOC

= visible = Dutch East India Company

ACKNOWLEDGEMENTS The authors thank Prof. R. Verpoorte of the division of Pharmacognosy, State University of Leiden, The Netherlands for his comments on the biosynthetical and cell culture sections and Prof. JE. de Groot of the Laboratory of Organic Chemistry of Wageningen University, The Netherlands for his comments on the nomenclature of compounds.

REFERENCES [I] [2]

[3] [4]

[5] [6] [7] [8] [9] [10] [II]

[12] [13] [14]

Thomson, R. H. Naturally occurring quinones, second ed.; Academic Press: Londen, 1971. van Genderen, H.; Schoonhoven, L. M. Chemisch-ecologische flora van Nederland en Belgie. Een inleiding over aard en ecologische betekenis van secundaire plantestoffen; Stichting Uitgeverij van de Koninklijke Nederlandse Natuurhistorische Vereniging Utrecht: Utrecht, 1996. Schweppe, H. Handbuch der Naturfarbstoffe\ Ecomed: Landsberg, 1993. Wijnsma, R.; Verpoorte, R. In Fortschritte der Chemie Organischer Naturstoffe, Progress in the Chemistry of Organic Natural Products; Hill, R. A.; I<j*ebs, H. C , Ed.; Springer-Verlag: Wien, 1986; Vol. 49; pp. 79-141. Leistner, E. In Biochemistiy of Plants; Conn, E. E., Ed.; Academic Press: San Francisco, 1981; Vol. 7; pp. 403-423. Schripsema, J.; Foliola, 1989,2, 1-8. van Dijk-van der PeijI, A. Meekrap vroeger en nu\ Verhage & Zoon: Middelburg, 1980. Tschirch, A. Handbuch der Pharmakognosie, zweite Abteilung; Verlag Chr. Herm.: Tauchritz, 1925. Weeda, E. J.; Westra, R.; Westra, C ; Westra, T. Nederlandse oecologische flora, wildeplanten en hun relaties; IVN: Hilversum/Haarlem, 1988; Vol. 3. Von Wiesner, J. Krapp; Wilhelm Engelmann: Leipzig, 1927; Vol. 1. van Laarhoven, H. P. M.; Boot, M.; Bakhuizen, R. TNO Rapport V97.254 titled: "Rapportage betreffende microscopisch onderzoek aan worteldoorsneden van meekrap"; prepared for cooperation Rubia: Zeist, 1997; pp. 21. Zuurdeeg, J. P. Ode aan de meekrap: Sint-Annaland, 1995. Sandberg, G. The red dyes: cochineal, madder & murex purple; Lark Books: Asheville, 1997. de Kanter, J. De meekrapteler en bereider; of volledige beschrijving van het planten, telen, reeden, en bereiden der meekrap.\ A. Blusse en Zoon: Dordrecht, 1802.

680 [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41 ] [42] [43] [44] [45] [46] [47]

Graebe, C; Liebermann, A.; Dtsch. Chem. Ges., 1868, 49, 279-281. Fieser, L. F.; J. Chem. Educ, 1930, 7, 2609-2633. Peel, R. A.; The Dyer & Text He Printer, 1956, 775, 1013-1015. Peel, R. A.; The Dyer & Textile Printer, 1956, 775, 851-853. Peel, R. A.; 77?^Dyer &Textile Printer, 1956, 775, 931-933. Kiel, E. G. Thesis: Metaalcomplexen van alizarinerood (1,2-dihydroxyanthrachinon), Technische Hogeschool Delft, 1961. Lauer, K.; J. Prakt. Chemie, 1931, 130, 185-254. Lauer, K.; J. Prakt. Chemie, 1932, 7J5, 164-192. Leistner, E.; Phytochemistry, 1973, 72, 1669-1674. Leistner, E.; Phytochemistry, 1973, 72, 337-345. Leistner, E.; Zenk, M. H.; Tetrahedron Lett., 1971, 20, 1677-1681. Leistner, E.; Zenk, M. H.; Tetrahedron Lett., 1967, 5, 475-476. Burnett, A. R.; Thomson, R. H.; J. Chem. Soc. Chem. Commun., 1967, 1125-1126. Leistner, E.; Phytochemistry, 1971,10, 3015-3020. Fairbairn, J. W.; Muhtadi, F. J.; Phytochemistry, 1972, 77, 215-219. Inoue, K.; Shiobara, Y.; Nayeshiro, H.; Inouye, H.; Wilson, G.; Zenk, M. H.; Phytochemistry, 1984, 23, 307-311. Inoue, K.; Shiobara, Y.; Nayeshiro, H.; Inouye, H.; Wilson, G.; Zenk, M. H.; J. Chem. Soc. Chem. Commun., 1979, 957-959. Ramos-Valdivia, A. C; van der Heijden, R.; Verpoorte, R.; Phytochemistry, 1998, 4^,961-969. Eichinger, D.; Bacher, A.; Zenk, M. H.; Eisenreich, W.; J. Am. Chem. Soc, 1999, 727, 7469-7475. Hosel, W. In The biochemistry of plants. A comprehensive treatise; Conn, E. E., Ed.; Academic press: Davis, 1981; Vol. 7; pp. 725-753. van der Plas, L. H. W.; Hagendoorn, M. J. M.; Jamar, D. C. L.; J. Plant Physiol., 1998,752,235-241. Robins, R. J.; Payne, J.; Rliodes, M. J. C; Phytochemistry, 1986, 25, 2327-2334. Schulte, U.; El-Shagi, H.; Zenk, M. H.; Plant Cell Rep., 1984, 3, 51-54. Harkes, P. A. A.; Krijbolder, L.; Libbenga, K. R.; Wijnsma, R.; Nsengiyaremge, T.; Verpoorte, R.; Plant Cell Tissue Organ Culture, 1985, 4, 199-214. Suzuki, H.; Matsumoto, T.; Mikami, Y.; Agric. Biol. Chem., 1984, 48, 603-610. Zenk, M. H.; El-Shagi, H.; Schulte, U.; Planta Med SuppL, 1975, 79-101. Leistner, E.; Bauch, H. J.; In Plantaardige geneesmiddelen in de gezondheidszorg; Labadie, R. P., Ed.; Scheltema & Holkema: Utrecht, 1980; pp. 18-25. Zenk, M. H.; Schulte, U.; El-Shagi, H.; Naturwissenschaften, 1984, 71, 266. El-Shagi, H.; Schulte, U.; Zenk, M. H.; Naturwissenschaften, 1984, 71, 267. Hagendoorn, M. J. M.; Wagner, A. M.; Segers, G.; van der Plas, L. H. W.; Oostdam, A.; van Walraven, H. S.; Plant Physiol., 1994,106, 723-730. van der Plas, L. H. W.; Eijkelboom, C; Hagendoorn, M. J. M.; Plant Cell Tissue Organ Culture, 1995, ¥5, 111-116. Kuzovkina, I. N.; Mantrova, O. V.; Al'terman, I. E.; Yakimov, S. A.; Russ. J. Plant Physiol., 1996,45,291-298. Sato, K.; Yamazaki, T.; Okuyama, E.; Yoshihira, K.; Shimomura, K.; Phytochemistry, 1991,30, 1507-1509.

681

[48]

van der Heijden, R.; Verpoorte, R.; Hoekstra, S. S.; Hoge, J. H. C ; Plant Physiol. Biochem., 1994, 52, 399-404. [49] Hagendoorn, M. J. M.; van der Plas, L. H. W.; Segers, G. J.; Plant Cell Tissue Organ Culture, 1994, 38, 227-234. [50] Kino-Oka, M.; Mine, K.; Taya, M.; Tone, S.; Ichi, T.; J. Ferment. Bioeng., 1994, 77, 103-106. [51] Derksen, G. C. H.; van Beek, T. A.; de. Groot, IE.\ Capelle, A.; J. Chromatogr. A, 1998,57(5,277-281. [52] Kawasaki, Y.; Goda, Y.; Yoshihira, K.; Chem. Pharm. Bull., 1992, 40, 1504-1509. [53] Burnett, A. R.; Thomson, R. H.; J. Chem. Soc, 1968, 2437-2441. [54] Kuiper, J.; Labadie, R. P.; Planta Med, 1981, 42, 390-399. [55] El-Emary, N. A.; Backheet, E. Y.; Phytochemistjy, 1998, 49, 277-279. [56] Inouye, H.; Takeda, Y.; Nishimura, H.; Kanomi, A.; Okuda, T.; Puff, C ; Phytochemistry, 1988, 27, 2591-2598. [57] Poginsky, B.; Westendorf, J.; Blomeke, B.; Marquardt, H.; Hewer, A.; Grover, P. L.; Phillips, D. H.; Carcinogenesis, 1991, 72, 1265-1271. [58] Angelini, I. C. a. P. G.; Pistelli, L.; Belloni, P.; Bertoli, A.; Panconesi, S.; Industr. Crops Prod, 1997, 6, 303-311. [59] Kawasaki, Y.; Goda, Y.; Yoshihira, K.; Noguchi, H.; Shoyakugaku Zasshi, 1990, 44, 95-97. [60] Kawasaki, Y.; Goda, Y.; Yoshihira, K.; Shoyakugaku Zasshi, 1988, 42, 166-167. [61] Berg, W.; Hesse, A.; Kraft, R.; Hermann, M.; Pharmazie, 1974, 29, 478-482. [62] Labadie, R. P. Thesis: Onderzoek van farmaceutisch interessante anthraceenderivaten, Rijksuniversiteit Leiden, 1971. [63] Toth, Z. A.; Raatikainen, O.; Naaranlathi, T.; Auriola, S.; J. Chromatogr., 1993, 630, 423-428. [64] Murti, V. V. S.; Seshadri, T. R.; Sivakumaran, S.; Indian J. Chem., 1970, 8, 779782. [65] Yasui, Y.; Takeda, N.; Mutation Res., 1983, 727, 185-190. [66] Kjizsan, K.; Szokan, G.; Toth, Z. A.; Hollosy, F.; M.Laszlo; Khlafulla, A.; J. Liq. Chromatogr. Rel. Technol, 1996, 19, 2295-2314. [67] Denee, R.; Huizing, H. J.; J. Nat. Prod, 1981, 44, 257-260. [68] Poginsky, B.; Westendorf, O. J.; Kraus, L.; Marquardt, H.; Pharm. Weekblad Sci. Ed, 1987, P, 242. [69] Wijnsma, R.; Go, J. T. K. A.; Harkes, P. A. A.; Verpoorte, R.; Baerheim Svendsen, A.; Phytochemistry, 1986, 25, 1123-1126. [70] Masawaki, T.; Taya, M.; Tone, S.; J. Ferment. Bioeng, 1996, 81, 567-569. [71] Lodhi, A. H.; Sant'Ana, A. E. G.; Charlwood, B. V.; Phytochem. Anal, 1994, 5, 261-265. [72] Briggs, L. H.; Nicholls, G. A.; J. Chem. Soc, 1954, 3940-3942. [73] Itokawa, H.; Mihara, K.; Takeya, K.; Chem. Pharm. Bull., 1983, 31, 2353-2358. [74] Hauptmann, H.; Lacerda Nazario, L.; J. Am. Chem. Soc, 1950, 72, 1492-1495. [75] Onal, A.; Turkey. J. Chem., 1996, 20, 204-213. [76] Shibata, S.; Takito, M.; Tanaka, O.; J. Am. Chem. Soc, 1950, 72, 2789-2790. [77] Formanek, I.; Racz, G.; Farmacia, 1973, 27, 201-206.

682 [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96]

[97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [Ill]

Hermans-Lokkerbol, A. C. J.; van der Heijden, R.; Verpoorte, R.; J. Liq. Chromatogr., 1993,16, 1433-1451. Westendorf, J.; Marquardt, H.; Poginsky, B.; Dominiak, M.; Schmidt, J.; UdiV(\udiX(M,U.\ Mutation Res., 1990,240, 1-12. Auterhoff, H.; Boehme, K.; Archiv Pharmazie, 1968, 301, 793-799. Hendriksen, L. M.; KJ0sen, H.; J. Chromatogr., 1983, 258, 252-257. van Eijk, G. W.; Roeymans, H. J.; J. Chromatogr., 1976,124, 66-68. Furuya, T.; Shibata, S.; lizuka, H.; J. Chromatogr., 1966, 21, 116-118. Westendorf, J.; Poginsky, B.; Marquardt, H.; Groth, G.; Marquardt, H.; Cell Biol. Toxicol, 1988, 4, 225-239. Inoue, K.; Nayeshiro, H.; Inouye, H.; Zenk, M.; Phytochemistry, 1981, 20, 16931700. Hostettmann, K.; Planta Med., 1980, 29, 1-18. Tanimura, T.; Pisano, J. J.; Ito, Y.; Bowman, R. L.; Science, 1970, 169, 54-56. Weng, W.-C; Sheu, S.-J.; J. High Resolut Chromatogr, 2000, 23, 143-148. Berger, Y.; Castonguay, A.; Brassard, P.; Org Magn. Reson., 1980,14, 103-108. Berger, Y.; Castonguay, A.; Org Magn. Reson., 1978, //, 375-377. Colin; Robiquet; Ann. Chim. Phys., 1827, 34, 225-253. Rochleder; Liebigs. Ann. Chem., 1851, 78, 246-248. Richter, D.; J. Chem. Soc, 1936, 1701-1703. Schunck, E.; Marchlewski, L.; J. Chem. Soc, 1894, 969, 1137. Wehmer, C. Die Pflanzenstoffe botanisch-systematisch bearbeitet; Jena Verlag, 1911; Vol. 239. Bartholome, E.; Biekert, E.; Hellmann, H.; Ley, H.; Weigert, W. M. In Erdol und Erdgas bis Formazanfarbstoffe; Buchholz-Meisenheimer, H.; Moegling, D.; Pfefferkorn, R., Ed., Verlag Chemie GmbH: Weinheim, 1976; Vol. 11, pp. 112114. Leistner, E.; Planta Med Suppl, 1975, 214-223. Ellis, B. E.; Natural Prod Rep., 1988, 5, 581-607. Leistner, E.; Zenk, M. H.; Tetrahedron Lett., 1968, 7, 861-863. Dodsworth, D. J.; Calcagno, M.-P.; Ehrmann, E. U.; Devadas, B.; Sammes, P. G.; J. Chem. Soc. PerkinI, 1981, 2120-2124. Almeida, W. P.; Costa, P. R. R.; Syn. Comm., 1996, 26, 4507-4518. Banthorpe, D. V.; White, J. J.; Phytochemistry, 1995, 38, 107-111. Dosseh, C; Tessier, A. M.; Delaveau, P.; Planta Med, 1981, 43, 360-366. Roberge, G.; Brassard, P.; Synthesis, 1981, 381-383. Roberge, G.; Brassard, P.; J. Org Chem., 1981, 46, 4161-4166. Mihai, G. G.; Tarassoff, P. G.; Filipescu, N.; J. Chem. Soc. Perkin I, 1975, 26, 1374-1376. Morley, J. A.; Woolsey, N. F.; J. Org Chem., 1992, 57, 6487-6495. Pozzo, J.-L.; Clavier, G. M.; Colomes, M.; Bouas-Laurent, H.; Tetrahedron, 1997, 53, 6377-6390. Beynon, J. H.; Williams, A. E.; Appl. Spectrosc, 1960, 14, 156-160. Berg, W.; Hesse, A.; Herrmann, M.; Kraft, R,; Pharmazie, 1975, 30, 330-334. Inoue, K.; Ueda, S.; Nayeshiro, H.; Inouye, H.; Phytochemistry, 1983, 22, 131741.

683 [112] Tessier, A. M.; Delaveau, P.; Champion, B.; PlantaMed., 1981, 41, 337-343. [113] Hauser, F. M.; Takeuchi, C ; Yin, H.; Corlett, S. A.; J. Org. Chem., 1994, 59, 258259. [114] Chang, P.; Lee, K.-H.; Phytochemistiy, 1984, 23, 1733-1736. [115] Roberts, J. L.; Rutledge, P. S.; Trebilcock, M. J.; Austr J. Chem., 1977, 30, 15531560. [116] Sharghi, H.; Tamaddon, F.; Tetrahedron, 1996, 52, 13623-12640. [117] Schunck, E.; Roemer, H.; J. Chem. Soc., 1877, 3J, 666-66S. [118] Hirose, Y.; Kusuda, J.; Nonomura, S.; Fukui, H.; Chem. Pharm. Bull, 1968, 16, 1377-1379. [119] Itokawa, H.; Qiao, Y.; Takeya, K.; Phytochemistiy, 1989, 28, 3465-3468. [120] Demagos, G. P.; Baltus, W.; Hofle, G.; Z Naturforsch, 1981, 36B, 1180-1184. [121] Suzuki, H.; Matsumoto, T.; Obi, Y. In Proc. 5th Int. Cong. Plant Tissue Cell Culture-, A. Fujiwara Ed.; Tokyo, 1982, pp.285-286. [122] Wijnsma, R.; Verpoorte, R.; Mulder-Krieger, T.; Baerheim Svendsen, A.; Phytochemistry 1984,23, 2307-2311. [123] Bhide, N. S.; Rao, A. V. R.; Indian J. Chem., 1969, 7, 996-1000. [124] El Ezaby, M. S.; Salem, T. M.; Zewail, A. H.; Issa, R.; J. Chem. Soc. (B), 1970, 1293-1296. [125] Shafer, K. H.; Griffiths, P. R.; Shu-Qin, W.; Anal. Chem., 1986, 58, 2708-2714. [126] Cambie, R. C ; Dunlop, M. G.; Noall, W. I.; Rutledge, P. S.; Woodgate, P. D.; Austr. J. Chem., 1981, 34, 1079-1084. [127] Arnone, A.; Fronza, G.; Mondelli, R.; Pyrek, J. S.; J. Magn. Reson., 1977, 28, 6979. [128] Metwally, S. A. M.; J. Appl. Chem. Biotechnol, 1975, 25, 161-168. [129] Vmdymdiihdin, A.; Dyes Pigments, 1985,^,27-30. [130] Shi, S.; Katz, T. J.; Yang, B. V.; Liu, L.; J. Org Chem., 1995, 60, 1285-1297. [131] Danielsen, K.; Acta Chem. Scand, 1996, 50, 954-957. [132] Castonguay, A.; Brassard, P.; Can. J. Chem., 1977, 55, 1324-1333. [133] Savard, J.; Brassard, P.; Tetrahedron, 1984, 40, 3455-3464. [134] Iwao, M.; Kuraishi, T.; Bull. Chem. Soc. Jpn., 1987, 60, 4051-4060. [135] Rochleder; Liebigs. Ann. Chem., 1851, 80, 321-330. [136] Willigk, E.; Liebigs. Ann. Chem., 1852, 82, 339-348. [137] Borisov, M. I.; Lisovaya, L. I.; Pristupa, N. N.; Chem. Nat. Prod, 1970, 6, 370. [138] Bianco, A.; Guiso, M.; lavarone, C ; Passacantilli, P.; Trogolo, C ; Gaz. Chim. Itai, 191HJ08, 13-16. [139] Belitz, H.-D.; Grosch, W. Food Chemistry; Springer-Verlag: Berlin, 1987. [140] Soubayrol, P.; Dana, G.; Man, P. P.; Magn. Reson. Chem., 1996, 34, 638-645. [141] Zhang, J.; Shen, X.; J. Nutr. Environ. Med, 1997, 7, 79-89. [142] Rath, G.; Ndonzoa, M.; Hostettmann, K.; Int. J. Pharmacog, 1995, 33, 107-114. [143] Formanek, I.; Racz, G.; Pharmazie, 1975, 30, 617-618. [144] Castro, S. L.; Pinto, M. C. F. R.; Pinto, A. V.; Microbios, 1994, 78, 83-90. [145] Blumenthal, M.; Busse, W. R.; Goldberg, A.; Gruenwald, J.; Hall, T.; Riggins, C. W.; Rister, R. S. The Complete German Commission E Monographs; American Botanical Council: Austin, Texas, 1998. [146] Schilcher, H.; Dtsch. Apoth. Ztg, 1984,124, 2429-2436.

684

[147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161]

Norton, S. A.; J. Am. Acad. Dermatol., 1998, 39, 484-485. Dixon,A. D.;Hoyte,D. A.N.;/^/7a/. Record, 1963,745, 101-113. Brown, J. P.; Mutation Res., 1980, 75, 243-277. Brown, J. P.; Dietrich, P. S.; Mutation Res., 1979, 66, 9-24. Brown, J. P.; Brown, R. J.; Mutation Res., 1976, 40, 203-224. Tikkanen, L.; Matsushima, T.; Natori, S.; Mutation Res., 1983,116, 297-304. Poginsky, B. Thesis: Vorkommen und Genotoxizitdt von Anthracen-Derivaten in Rubia tinctorum L, University of Hamburg 1989. Ino, N.; Tanaka, T.; Okumura, A.; Morishita, Y.; Makita, H.; Kato, Y.; Nakamura, M.; Mori, H.; Toxicol. Industr. Health, 1995,11, 449-458. Wolfle, D.; Schmutte, C; Westendorf, J.; Marquardt, H.; Cancer Res., 1990, 50, 6540-6544. Blomeke, B.; Poginsky, B.; Dominiak, M.; Marquardt, H.; Westendorf, J.; Marquardt, H.; Planta Med, 1990, 56, 509. Blomeke, B.; Poginsky, B.; Schmutte, C; Marquardt, H.; Westendorf, J.; Mutation Res., 1992,25(5,263-272. Chulasiri, M.; Matsushima, T.; Yoshihira, K.; Phytotherapy Res., 1995, 9, 421424. Kawasaki, Y.; Goda, Y.; Noguchi, H.; Yamada, T.; Chem. Pharm. Bull, 1994, 42, 1971-1973. Westendorf, J.; Pfau, W.; Schulte, A.; Carcinogenesis, 1998,19, 2163-2168. Kawasaki, Y.; Goda, Y.; Maitani, T.; Yoshihira, K.; Takeda, M.; Shokuhin Eiseigaku Zasshi (J. Food Hyg. Soc. Japan), 1992, 33, 563-568.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

685

THE CHEMISTRY AND TOXICOLOGY OF BIOACTIVE COMPOUNDS IN BRACKEN FERN (PTERIDIUM SSF), WITH SPECIAL REFERENCE TO CHEMICAL ECOLOGY AND CARCINOGENESIS MIGUEL. E. ALONSO-AMELOT Grupo de Quimica Ecologica, Departamento de Quimica, Universidad de Los Andes, Merida, Venezuela. [email protected]

ABSTRACT: The Pteridium taxonomic complex grouped around the general denomination of "Bracken Fern" is a factory of obnoxious secondary metabolites, which in combination with a series of remarkable biological traits, contribute to its status as one of the five worst weeds in the world. Bracken excludes competing vegetation effectively, fends off the great majority of herbivorous insects, poisons numerous farm animal species, and resists fire and repeated cuttings. Evidence is also accumulating that it may affect people, not only economically but also by eliciting the onset of gastric cancer by direct ingestion of boiled croziers or consumption of animal products exposed to feed on bracken fronds. The present chapter reviews the xenobiotic chemistry of bracken by structural types, including all those compounds derived from the shikimic acid and mevalonic acid biosynthetic routes, and describes recent developments in the chemical behavior, toxicology, and chemical ecology of the most notorious of these compounds. Particular reference is made to the mechanisms that bring about the various cancer scenarios that the ingestion of this plant may cause, and the molecular processes involved.

INTRODUCTION Bracken fern {Pteridium ssp) is both an evolutionary w^onder and an agricultural scourge. The characters that endow this plant with its tremendous ecological success [1] are the same that powerfully influence managed ecosystems for forestry [2-4], conservation [5-7], agriculture [8] and farm animal production [9]. Some of these characters are primitive traits preserved over the ages, and others are idiosyncratic of modem plants [10]. Among the outstanding biological features, there is the extraordinary ability of bracken to colonize new territory through dispersion of enormous numbers of microscopic spores [11-13], which may amount up to 10 g per frond in the arachnoideum genotype [14]. Carried by wind, these spores travel long distances [15-16]. Bracken also takes over new land by the rapid expansion of its profuse rhizome system [17-18]. The

686 rate of growth of rhizomes and biomass above the ground, especially in the spring months in temperate climates [19-20] and during most of the rainy season in tropical regions [18,21], is frequently greater than that of potentially competing plants. On land cleared by slashing and/or fire, not only spores germinate abundantly [22] but if the ground previously harbored bracken rhizomes that grev^ unnoticed under the existing vegetation with all but a modest aerial representation of dispersed fern blades, rapidly growing fronds are usually the first to emerge from the bare ground surface. Numerous croziers appear during the first weeks after fire or clearing, achieving high densities, Fig. (1-A). The blades swiftly cover the area with its expanding pinnae, thus retarding or inhibiting the germination and growth of other heliophile, early succession plants for critically long periods of time. Areas of periodical wildfires are especially prone to bracken growth. Pteridium perpetuation results from the survival of fire-resistant rhizomes, located deep underground sometimes, in combination with the denudation of the ground surface from competing vegetation. In the neotropics, it has been recorded that the time from frond emergence to a fully expanded blade may be less than 45 days. Fig. (1-B) [18]. In such a short time pinnae spread out to a structure of considerable dimensions, measuring more than a meter long and occasionally up to 2.5 meters. The rapid growth of individual sporophytes and their longevity make Pteridium ssp one of the largest plants loiovm [23,24]. In addition, if not flattened by snow or severe wind storms, the wilted fi-onds will remain standing and providing mechanical support for new frond growth. The buildup of litter is such that the undercroft shade at ground level becomes equivalent to the darkness of the tropical rain forest [14]. (A)

687 (B) 100 -r

u w

u

80 1 60 1

cd

<<^ 40 o

1

£

CT

^

20 1 0

days of emergence

Fig. (1). Growth of pteridium aquilinum var caudatum fronds in a tropical mountain environment at 1850 m asl, from day of emergence above ground following fire. (A): variation in crozier density as a function of time. (B): elongation of blade central rachis. Accumulation of biomass may reach 5 g/d per frond each day at the maximum growth period. Data from AlonsoAmelot and Rodulfo, 1996, [18].

Shading alone contributes significantly to the inhibition of seed germination of heliophile plants. Paradoxically, it is this massive accumulation of litter, and not the competition of other plants, that in time induces bracken degeneration. It becomes suffocated by its own success, and opens the way for the establishment of other plant species that model the structure of later successional stages [25,26]. Besides these outstanding biological attributes, bracken is also gifted with a complex array of secondary chemicals. Most of them play a particular role in helping this plant to invade new territory, become firmly encroached, amass effective defenses against predators and competitors, and turn it into an often dominant element of the ecosystem. This is accomplished by the cost-saving production of these chemicals at specific times of the growing season [27], and a shrewd evolution-driven molecular structure design to render them strongly biologically active. Over one hundred compounds have been isolated and structurally characterized [28] from the various bracken genotypes that comprise the Pteridium taxonomic complex [10, 29,30]. The combination of the biological strategies and the timely production of bioactive chemicals [31,32], in addition to a genome of considerable plasticity residing in 104 chromosomes [30], give support to the spectacular ecological success of

688 bracken. Indeed, Pteridium is deemed as one of the five most important weeds in the entire world [33] It is no wonder that bracken fern has attracted the attention of scores of research groups from an array of disciplines over the past one hundred years since it was discovered that cattle was harmed by its ingestion [34,35]. Considerable efforts and resources have been invested in trying to explore and understand the intimate way of action of its many bioactive components [36-42]. Several reviews have appeared over the years on various aspects of bracken toxicity [27,28,36,43-50]. Interest around this plant has been spurred by recent discoveries relating bracken invasion with certain human health issues to be discussed here. The present chapter also covers the most relevant and recent aspects of the chemistry and mode of action of those bracken compounds that constitute the most biologically active fraction. Bracken biological impact, an overview The ingestion of bracken causes several deleterious effects in vertebrates, which are grouped under the general denomination oi bracken poisoning. An aggessive secondary chemistry is deeply involved. The clinical symptoms vary widely according to the animal's species. For example, sheep are blinded by progressive retinal degradation and often develop severe mouth lesions, whereas cattle show high fever, difficulty in breathing, excessive salivation, congestion of the mucous membrane, general weakness, increased coagulation time and, after prolonged exposure at relatively low dosages of bracken fi-onds, usually over 600 days, they may develop nasal, rectal and urinary bleeding (hematuria). In horses, donkeys and mules, the symptons are difficulty in breathing, unsteady gait, lack of coordination, drowsiness, dilated pupils, constipation, weak and irregular pulse, nervousness, twitching of muscles, backward inflection of the neck, yellowing of the eye sclera and extreme emaciation. Pigs on their part may develop heart insufficiency and lack of appetite. All groups of animals may diefiromthe condition if not treated conveniently. Generally, bovines carmot be spared once the haematuria sets in. Also, embryos may be impaired by their mother's intake of bracken. Some teratogenic effects have been observed in mice fetuses of bracken-fed mothers. Retarded ossification of the stemebrae and growth suppression along with maternal weight loss are the chief effects [51]. At the physiological level, various harmfiil processes have been identified such as hypoplasia of the bone marrow, thrombocytopenia, leucopenia particularly affecting polymorphonuclear leukocytes which promotes immunosupression, sharp decrease in the intestinal iron uptake and malfimctioning of liver and kidneys. These effects have been related to the deadly acute poisoning condition. Animals that ingest sublethal amounts of this fern may develop severe internal hemorrhages. Of

689 additional importance in sheep is cerebrocortical necrosis, retinal atrophy and severe vitamin Bl depression that also distress monogastric animals [27,28,36,43-47,49,50]. Polygastric or ruminant mammals synthesize their own vitamin Bl and are exempt from bracken-derived thiamine deficiency. Bracken and animal cancer In spite of all these severe consequences, the bracken-related health contingency of greatest concern in cattle is the hemorrhaging of the urinary bladder, known as bovine enzootic or vesical haematuria (BEH) [52]. This desease has been reported in several countries including central and southern Europe [53], Turkey [54], New Zealand [55], Australia [56], Korea, Japan, China [57], India [58-60], the new world: Venezuela [61], Colombia [62,63], Costa Rica [64], Brazil [65], and North Western United States [66], and isolated mid-ocean islands such as Hawaii, San Miguel in the Azores and Madeira [67,68]. Prolonged exposure to feeding bracken, usually in excess of several hundred days, leads to debilitating blood-letting through urine. In some animals, blood clots blocking the urethra and causing urinary retention have been reported [69]. This blood is poured into the urinary liquor by eventual rupture of highly vascularized vesicles of the inner wall of the bladder. Post mortem histological examination of these lesions reveal capillary ectasia, angiomatous cavity formation and vascularized proliferation involving the epithelium and the mesenchyme. Ultimately, malignant carcinogenic infiltration of large portions of the bladder wall take place, thus ruining irreversibly the animal's health. These bladder tumors comprise transitional epithelium carcinomas, papillary carcinomas, hemangiomas, hemangiosarcomas, fibromas and polyps. Furthermore, the weakened animal may die from various infections with invasion of the bloodstream by alimentary bacteria that overcome the debilitated immunologic response leading to septicemia [69,70]. That bracken is indeed responsible for the formation of these cancerous tumors was well established not only by the geographical overlap between bracken infestation in pasture land and the incidence of BEH in various countries [54,65,67,71], but also by inducing BEH in healthy cows with a diet containing Pteridium fronds [72,73]. Thus, Pteridium is the only plant known to cause naturally cancerous lesions in animals [75]. A sizable body of evidence has shovm that bracken carcinogenesis occurs also in other experimental animals such as rat, mouse, hamster, guinea pig, guppy fish (Lebistes reticulatus), Egyptian toad {Bufo regularis) and Japanese quail {Coturnix coturnix japonicd) [43,44,74-78]. Cancer types and affected organs depend on the species. For example, chiefly sarcomas, adenomas, adenocarcinomas of the ileum and in some instances carcinomas of the urinary bladder [76] affect the

690 rat whereas mice tend to develop lung adenomas and lymphatic leukemia. Other tumors reportedly develop in other organs in these rodents, leading to mammary cancer and the formation of hyperplastic nodules in the liver [79,80]. Cats also appear deeply affected in the liver after being fed with bracken artificial diets for just a few days [71]. All the test animals died from the condition shortly thereafter. However, the majority of these experiments indicate that it is the exposure to small, sublethal amounts of bracken feed for long periods that will cause the highest incidence of cancer. Bracken carcinogenicity may be spurred by ethiological cofactors, thus affecting other organs in the animal. This is the case of neoplasias found in specific sites of the upper alimentary tract of ruminants. Cattle have been observed to develop squamous carcinomas in the lateral dorsum of the tongue, the soft palate and oropharynx, the esophagus, the cardiax and specific places of the rumen [81]. Sheep are affected likewise [82]. Papillomas were found in 96% of the bovines which showed squamous carcinomas. It was significant that the papillomas found in a population (7,746 individuals) of non-cancerous cattle were located at exactly the same sites where squamous carcinomas occurred. The finding of DNA fragments from bovine papilloma virus type 2 (BPV-2) in 46% of the bracken-fed bovines where urinary bladder carcinomas had been detected naturally and 69% of those in which the tumors had been induced by BPV-4 inoculation [69], confirmed earlier correlations between BEH and papilloma virus [83,84]. The synergistic mechanism that emerges from Campo et al. experiments consists of natural PBV infection from mother to calf at birth and by contactfi'omnursing. This virus may become latent to be reactivated later in life by immunodeficiency caused inter alia by bracken. For the papillomatosis to progress into malignant neoplasias, the participation of powerfiil mutagen-carcinogens of bracken is apparently required. The carcinogenicity of Pteridium varies according to plant part and age, which correlates well with their corresponding chemical composition. In an early study performed when the actual carcinogens of bracken were still unknown, Hirono and coworkers [85] examined the tumor forming potential of inbred ACI rats. The rodents were given oral doses of bracken parts for two months that were inserted in rat basal diet pellets. Curled tips of corziers, crozier stalks, young fronds, rhizomes and rhizome starch of P. aquilinum probably of the Japanese variety latiusculum were employed. Carcinogenesis was induced more significantly by the curled tips and were the only bracken part also producing emaciation and pneumonia. Other plant sections were decreasingly active in the order rhizome and frond, whereas no activity could be observed in the rhizome starch. In all cancer-positive cases, rats developed multiple intestinal tumors whereas malignancies in the urinary bladder were recorded in a few animals only. These observations confirmed those of Evans who concluded that the concentration of the toxic principles causing "bracken poisoning" in cattle were five times that

691 found in the fronds [86]. As we shall describe later, the carcinogenic potential does not run in parallel with the concentration of known carcinogens detected in all bracken parts. Bracken and human cancer Bracken has coexisted with man since ancient times and has been utilized in several applications. For example, IVth century peoples of present-day Northumberland, England, used bracken fronds as bedding which also harbored insect pests. Excavations in Vindolanda, a site of Roman occupation, revealed as many as a quarter of a million pupae of the biting and blood sucking stable fly Stomoxys calcitrans, mingled with litter in just 30 m^ of a bracken floor-covering inside a human dwelling [87]. Although it has never been shown that skin contact with dried fronds leads to disease other than skin allergies, breathing the large number of spores and frond litter dust expected to be in such enclosures and the intimate contact with this bedding may not have been auspicious of good lung and skin health for Vindolanders. A few experiments have shown that compounds in orally administered spores caused neoplasias in mice [88] or formed DNA adducts with potential to disrupt cell function [89]. However, specific bracken carcinogens are still to be chemically characterized in spores [90]. The thickness of the bracken deposits in these archeological sites also indicates that harvesting of this abundant resource was a major affair of that community and it is likely that parts of bracken such as rhizomes were consumed as a source of starch. Excavations at much older deposits in Tasmania (6,500 years BP) [91] and Australia (1,500 y BP) [92] have unearthed evidence of rhizome processing for food. In 1799 Alexander von Humboldt described the preparation of grated rhizomes of Pteris aquilina, the former denomination of Pteridium aquilinum in the Canary Islands. This material toasted with barley flour to make gofio, a traditional cake there that made its way to parts of South America in colonial times [93]. Humboldt and other later researchers have valued bracken rhizomes chiefly as famine foods and as such may be still in use in various primitive cultures around the world [94]. In spite of Evans' observation [86] that rhizomes are responsible for bovine bracken poisoning and that it contains a higher concentration of thiaminase, there are no confirmed reports showing that bracken underground parts prepared for culinary purposes may be the cause of any illness in people, cancer least of all. It is likely that toxic components are washed away during preparation or remain in very small concentration. Young bracken fronds, on the other hand, are appreciated as an exquisite vegetable salad component in Japan. Over 13,000 metric tons of croziers -or Warabi- are imported annually in addition to the local production, specially of the prefectures of central Honshu island. Japan, with a 37.9 age-adjusted death rate per 100.000 people in males and 17.2

692 for females due to stomach cancer, fares as the fourth highest in the world after Korea, Costa Rica and Russia, and the consumption of croziers in a large scale may have something to do with this. Human gastric cancers, however, may result fi-om multifactorial dietary components and it is not completely certain that bracken intake may be involved. This is illustrated by the fact that Chile, the only country of the Americas with neither P. aquilinum in its territory nor any direct human intake of the plant, is the fifth highest in stomach cancer-derived death rates at 34.8 for males and 15.6 for females, followed not far by Ecuador (26.8 and 19.1 for males and females, respectively) where bracken infests large tracts of its densely populated sierras. Obviously, dietary factors other than bracken are likely to be involved. But compelling evidence is nonetheless accruing. A study embracing a small sample of 98 Japanese people of both sexes who were afflicted with cancer of the upper alimentary tract [95,96], unveiled that the habit of eating Warabi daily elevated by 2.68 the risk of developing cancer of the oesophagus and 1.53 for those eating bracken croziers only occasionally, as compared with a control of 480 age-matched people. It is interesting to note that the higher risk occurred among people who also smoked or took other key foods such a certain kind of local hot tea named chaguyu. Hirono and coworkers [97] examined the carcinogenic potential of Warabi that was processed in the Japanese manner by immersing the fresh croziers in boiling water containing wood ash, sodium carbonate, or plain boiling water to remove bitterness, and then seasoned with various sauces and served. Rats given a daily diet for several months of woodash water-boiled young fronds developed only 30% or less ileal tumors as compared with those who received unprocessed bracken croziers. The latter group also grew bladder papillomas and carcinomas. Although processing of bracken reduced its carcinogenic potential, it was not completely eliminated and still represented a risk not to be neglected. More recently, an epidemiological study of a Brazilian community in Ouro Preto, who consumes bracken croziers of the arachnoideum variety by first boiling them three times for a total of about 30 min, revealed that the risk of developing oesophageal and stomach cancer was 3.40 and 3.45, respectively, relative to a control population who did not eat bracken preparations [98-101]. The combined overall age-sex adjusted rates corrected for consumers of tobacco and smokers was 5.47-fold for those who ate Pteridium croziers. Cytogenetic analyses of peripheral blood lymphocytes of habitual consumers for more than 10 years displayed a significant increase in chromosomal aberrations [102]. Water from boiling crozier washes was shown to be highly tumorigenic [103]. Curiously, people consuming diets low in animal protein and rich in starch, which is a typical food pattern among the poor in rural areas of Ouro Preto and elsewhere in Latin America, were at a lower risk of acquiring gastric cancer from crozier consumption, as indicated by laboratory rats models [104].

693

A more subtle contact with bracken obnoxious chemicals has been proposed to explain the inordinately high rates of human upper alimentary tract cancers in other parts of the world where bracken grows profiisely but is not consumed as food. Epidemiological surveys revealed that bracken exposure during childhood was positively correlated with a more than double risk of acquiring stomach cancer in rural regions of Wales [105,106]. A more dramatic situation was found in Costa Rica where the gastric cancer rates are among the highest in the world. By comparing the health condition of inhabitants of the lowlands where bracken is less abundant and those of the central highlands where bracken can be seen growing profusely in pasture land, Villalobos-Salazar and co-workers [64,71] found that the standardized incidence rates per 100.000 people were 19.40 in the low regions vs a staggering 53.02 in the uplands. The oesophageal cancer rates were 1.28 and 3.82, respectively. At the same time, there was a high incidence of bracken-derived BEH characterized by hemangiomas, papillomas and transitional cell carcinomas in the epithelium of the urinary bladder, affecting 10.4% of a sample of 600 cows from these pastures. In a similar study, Alonso-Amelot and Avendano [107] examined the cancer situation of a population of 5.5 million people distributed between three Andean states in westem Venezuela where bracken infests agricultural and pasture land, and a vicinal low laying area surrounding lake Maracaibo where bracken is virtually absent. Other possibly intervening variables such as socio-economic status of the average population of the two groups, nitrates in drink water supplies, diet, alcohol and tobacco use, and availability of health services, all of which have been shown to contribute to the development of various cancers and ensuing death, were taken into account. The incidence of gastric infections by Helycobacter pylori was also examined as there is growing evidence that this bacterium is involved in the development of gastric metaplasias which are an intermediate stage en route to gastric malignancies. Comparison of the 1985-1996 age-sex adjusted mortality rates caused by gastric neoplasias in the two study areas showed that gastric cancer is the main cancer killer among the highlanders and only fifth in the lowlands. Also, the risk of dying from this condition is 3.64 times higher in the Andean uplands than in the Maracaibo lake basin. Cancers of the prostate, breast, and uterus, all occurring also in high rates, were undifferentiated in both regions, in parallel with observations in Costa Rica [64,71]. The only emerging individually acting factor on the higher stomach cancer rates was the presence of bracken fern in those regions. Other studies are not so enthusiastic in questioning bracken as a human hazard, pointing out that it is not wise to extrapolate from animal models, where bracken toxicity is a well established fact, to people. It is also argued that "methodological problems" stand in the way and urge to proceed with further epidemiological research [108]. This is precisely what recent bracken field studies have been pursuing [107].

694 Bracken carcinogenicity may be transmitted through milk Since none of the people in all those areas encompassed by regional epidemiological studies ingested or used bracken parts in any significant scale and still suffered from high mortality rates due to gastric cancer, it was necessary to establish an indirect link between the plant and the affected persons. Likely candidates would be rain-water runoffs passing through bracken thickets and animal products (milk and meat). While the former is awaiting further attention [109], the latter has become a full fledged line of research in recent years. The original source is an early paper by Pamukcu, Olson and Price [110] which reported on the effect of surgically implanting in bladders of mice cholesterol pellets containing an "acidic" fraction of urine (ethyl acetate extract of HCl treated urine to pH 2) from cattle fed bracken supplements. Contingency tables showed a significantly higher proportion of urinary tumors in the treated group as compared with controls that were implanted with pure cholesterol pellets only. The suggestion was clear: a bracken carcinogen had been absorbed gastrically by cows, and excreted by way of urine after passing through the bloodstream. As we shall see later, it is unlikely that the true carcinogen of bracken (see below) survived the extraction procedure because it is highly unstable in acidic medium, but the experiment strongly suggested the presence of bracken metabolites in urine. Evans exploited these results and surmised that the carcinogen might as well be excreted through milk [111]. Milk from cows that received daily supplements of bracken feed, was orally given to a young calf for several weeks. In addition to bloody fecal mucus, indicative of intestinal damage, blood analysis revealed a decrease in leucocyte, neutrophil and platelet counts, typical symptons of bracken poisoning. Further evidence stemmed from feeding bracken to pregnant and lactating mice. Passage of bracken carcinogens through the placenta and /or milk was shovm by the appearance of peripheral pulmonary adenomas after 12-18 months in the offspring who were never allowed to eat the bracken-laden diet. Confirming reports added mammary adenocarcinomas to the record, in 27% of mice bom from bracken-fed mothers and adult mice fed with milk from cows with Pteridium diets [112;113]. Final proof came after chemically detecting the actual carcinogen in milk from bracken-fed cows by HPLC analysis [114]. This point will be resumed later as it makes sense only after a deeper look into bracken bioactive chemistry in the following sections. Direct contact with bracken metabolites may result from other uses. For example, fresh bracken fronds are used in the northern Andes as wrapping of cheese to be smoked, in order to add flavor to the product. Some Wayuu Indians of the Perija Sierra in Western Venezuela bum the dry fi-onds with a limited air supply inside their huts to fend off mosquitoes. Some communities in Colombia use frond infiisions as vermifugues. However, data is still missing to relate these applications to human cancer. But all the well

695 established geographical correlations between bracken prevalence and cancers of the upper alimentary tract suggest that, being stomach neoplasias a consequence of environmental -dietary- carcinogens, somehow bracken secondary allelochemicals -and especially carcinogensfind their way into human diet. This subject will be retaken once we describe the bracken carcinogens. Bracken and ticks Plants may have non-chemical, indirect ways of affecting animal and human well-being. Weeds for one, tend to asphyxiate crops, sequester needed soil nutrients, and harbor pests. The abundant foliage of dense bracken thickets has been shown to foster infestation by ticks that find there protective lodging and thrive by the thousands [115]. Among these ticks is Ixodes ricinus which is a host for the lime disease bacterium Borrelia burgdorferi. Deer in the wilderness, and cattle and horses in managed lands maintain the tick-lime disease cycle active, and it is favored by the bracken stands. Other tick-borne miseries such as piroplasmosis caused by bacteria of the Babesia species, Q-fever, anaplasmosis, tularemia and hemorrhagic fever in cattle are potential side effects of bracken prevalence in tropical areas where ticks have been observed in considerable numbers. Bracken xenobiotic metabolites, an overview. Phenolics Bracken synthesizes a large number of secondary metabolites stemming from the shikimic acid and mevalonic acid pathways, and fi*om the derivatization of certain aminoacids. Shikimic acid itself was isolated from Pteridium, to which part of the carcinogenic properties of this plant were mistakenly attributed in early studies [43,116]. Notorious among the aromatic secondary metabolites are various phenolic acids [117,118], some of which show anti-microbial activity [119]. Of special interest is chlorogenic acid [120], a widely distributed quinic ester of caffeic acid. This compound inhibits the growth of lepidopteran larvae when incorporated in artificial diets [121-123]. 0-coumaric acid and its cyclization product coumarin have been recognized also among the phenolic fraction of P. aquilinum in a sample of the neotropical genotype caudatum [124]. Both groups, aromatic carboxylic acids and coumarins, are known to act as phytotoxins against other plants. A HPLC analysis of water collected under green bracken fronds of the caudatum variety revealed the presence of coumarin as the major component [14]. This suggested that this compound is utilized extemally by bracken, taking advantage of rain-driven leaching to incorporate it into the surrounding soil, and thus contributing to the overall allelopathic potential of this

696 plant [3,8,125-127]. Other studies [128] indicate that the effect of the leachate on certain model plants such as Medicago sativa, Trifolium repens, Eucalyptus haemastoma and Lolium perenne vary according to the frond phenological stage and the cumulative rainfall over the previous three weeks. Thus, rain totaling 13 mm or more in this three-week period, which is a very moderate rate for a tropical environment subject to monsoon rains, reduced almost totally the inhibition of radicle elongation. This result clearly suggests that the supply of allelopathics in the fronds is exhausted and not replenished promptly. Given the genetic flexibility and amplitude of the Pteridium complex, it is to be expected that the allelopathic potential will vary widely among its species, subspecies and varieties, as stated by Gliessman [129]. Allelopathy is far from simple and as a competition strategy, it is subject to strong opposition by target-plant adaptations, soil chemistry, and destruction of semiochemicals by microorganisms in the ground. One example of this is illustrated by the study of Glass [130] who showed that a solution containing p-hydroxybenzoic, p-hydroxycinnamic, vanillic and ferulic acids in the exact proportions reported in the soil around the rhizosphere of P. aquilinum [131] was inactive on barley rootlet development unless calcium sulfate acted as a cofactor. He also concluded that extreme conditions related to temperature and low nutrient content of soil increased the root growth inhibitory effect of the calcium-phenolic acid mixture, thus suggesting that, in nutrient rich soils, bracken allelopathy is likely to break down as a useful strategy. Finally, certain wild species such as Agropyron repens, known to be strongly allelopathic by way of phenolic acid production, was not only resistant to bracken phenolics in the soil but in fact its growth was stimulated by these materials. Other less common phenolics such as p-hydroxystyrene [132] and five glycoside derivatives or ptelatosides [133,134] of unknown activity have been reported in Japanese bracken. Flavonoids are common in ferns [135] and bracken is no exception in being capable of synthesizing a limited series of flavonols [136-140,281]. These compounds purportedly are bioactive against microorganisms. In particular, the widely distributed quercetin, deemed at one time as an inducer of ileal and urinary bladder tumors in rats [141-143,282], antineoplasic [144,145] and a mutagen for prokaryote and eukaryote cells [146-150], does bring about DNA strand rupture [151] in the presence of cupric derivatives and oxygen [152], DNA rearrangements [153], and chromosomal aberrations [154]. Free radicals appear to be involved in these processes. Although quercetin given in oral dosages as high as 20 g per calf per day for several months could not be associated with BEH or papilloma virus-induced lesions in the urinary bladder of cattle [69], quercetin has been linked to other cancers of the upper gastrointestinal tract in bovines. Primary cells infected by bovine papilloma virus type 4 (BPV-4) or by viral transforming-protein E7 are not tumorigenic/>^r se. In fact, the viral lesions or papillomas regress spontaneously. However, the transformation

697 of papillomas into carcinomas may take place under certain conditions [155]. For instance, a single dose of bracken quercetin will suffice to turn the virus-infected cells to cancerous tissue, leading to papillomacarcinoma syndrome of cattle [156]. It has been determined that quercetin trans-activates certain viral transcriptional promoters which lead to increased expression of viral oncogenes, thus fumishing the malignant transformation of the cell [157,158]. This is complimented by damage to genes expressing tumor supressor p53, which spurs the neoplasic process [159,160]. These gene-disrupting processes are believed to operate not only in farm animals but also in humans who eat fern parts. Compelling evidence has been collected from the discovery of human papilloma virus types 16 and 18 in 50% of cancers and precancers of the upper alimentary tract of people from Ouro Preto in Brazil. There, bracken croziers are eaten after relatively brief cooking, a treatment that preserves the molecular integrity of most of its xenobiotic phytochemicals. The detection of viral DNA in tissue from an oesophageal carcinoma of a person from this region lent support to the papilloma virus-bracken-cancer connection. In addition, quercetin and its 0-gluco-rhamnoside, rutin, have been implicated, along with chlorogenic acid, in the inhibition of ineffectivity of Helicoverpa zea single nucleocapsid nucleopolyhedrovirus HzSNPV against larvae of the polyphagus lepidopteran Helicoverpa zea [161], although recent results indicate to the contrary [162]. Of importance to insect feeding is that quercetin and rutin act as modulators of the P450 GST insect digestive enzyme. These flavonoids inhibit this enzyme in the processing of certain suicidal substrates such as alkoxycoumarins, and blocks DNA synthesis and repair [163]. In spite of these remarkable activities, no one has been able to relate the content of flavonoids in bracken to the defensive potential against real predators, based on ecological evidence. Of importance to bracken defensive chemistry, due to the large quantity accumulated in the fronds, are the condensed tannins derived from procyanidin and prodelphinidin [164-166]. Tannins have been a substantial component in chemically based plant defense theories [167169] since they strongly bind to dietary proteins, and reduce the digestibility and nutritive value of plant tissue devoured by insects and vertebrates. To tannin in bracken has been attributed at least part of the carcinogenic potential of this fern because intravascular implantation of bracken tannin induced bladder cancer in mice [137]. Nevertheless, there is a paucity of data concerning the possible synergistic effects that tannins may have with well recognized bracken carcinogens in inducing neoplasias in animals and human consumers of whole plant parts. By contrast, bracken tannins added to the diet could not be associated with cancerous tumors in laboratory rodents [141]. In tropical bracken, the condensed tannins may be in excess of 120 mg/g of frond biomass and are stored chiefly in the vacuoles of the cells in the parenchyma and cuticle of the bracken pinnulettes [14]. Such

698 quantity may be partly responsible for the little interest that insects in general show for this plant [166,170]. Although over one hundred species of arthropods are known to exploit parts of bracken [171-173] and at least two, Conservula cinisigna de Joannis (Lepidoptera: Noctuidae) and Panotima angularis Hampson (Lepidoptera: Pyralidae) have been discovered in South Africa as bracken specialists [174], the damage to the aerial parts by the great majority of exploiters appears to be limited and may not threaten the survival of the plant. Cyanogenic potential Another chemical front against herbivory in bracken is offered through cyanogenesis, a line of defense that is considered fundamental in plants and certain arthropods [175-182]. Hydrogen cyanide is known to be toxic against invertebrates, vertebrates [183], bacteria [184], and fungi [185]. The mechanism of action is believed to consist of interference with the functioning of heme proteins, particularly cytochromes, myo- and hemoglobins, which are rendered inactive by complexation of the cyanide anion with the heme iron atom. If it is true on the one hand that some exploiters may develop digestive strategies to detoxify or even make use of cyanide as a nitrogen source, which results in various degrees of tolerance [186,187], on the other it is also known that cyanogenic genotypes of bracken are less consumed by invertebrate [188] and vertebrate [189] predators than acyanogenic genotypes [190]. Bracken cyanogenesis stems from the mandelonitrile glycoside prunasin [191,192], which is present from the early stages of its development, for the evolution of HCN has been witnessed in gametophytes and young sporophytes [193]. The capacity to release HCN by simple crushing of the fem tissue, as it would occur during chewing, is a clear indication that cyanogenic types also contain prunasinase that is stored in a different cellular compartment. This is a p-glucosidase enzyme which, when rendered in contact with prunasin, excises the glucose portion yielding the unstable a-cyanohydrin mandelonitrile. This compound is then transformed by the concurrence of a second enzyme, oxynitrilase, into HCN and benzaldehyde by p-elimination. The aldehyde is also a pungent chemical for some organisms. The kinetics of the process has been measured recently in bracken croziers of the neotropical variety caudatum [278]. A first order rate constant of ^ = 2.20 ±0.01 x 10"4 s"i for the production of HCN was obtained. The initial velocity was also calculated at 0.0048 ± 0.008 pMol g"^ min"l. Such a slow pace means that the release of HCN will continue to transpire inside the animal's digestive system, but the rate of evolution will depend on the particular pH of the digestive liquor, since the optimum medium for the glucosidase to operate is acidic [194]. Some insects known to possess alkaline media in their midguts [195] are likely to experience the least

699 HCN production in the ingested frond material. In the case of large ruminants, it has been concluded that the cyanogenic principles of bracken offer only a marginal if any contribution to the overall bracken poisoning scenario [196]. In seasonal climates the highest content of prunasin in the bracken fronds occurs in the early to mid spring and decays with the growing season to reach a minimum value by early fall [191,197]. In a tropical environment where no winter diapause takes place, it is the young croziers who show the highest capacity to release HCN throughout the year and it declines as the frond progresses into a mature blade [14]. The decline in the production of HCN with frond age is due to a progressive reduction in prunasin content and not to a lower enzyme activity [198]. The obvious parallel between the temperate and tropical surroundings seems to be a response of the plant directed to protect the fragile meristemic tissue, which is more liable to fall under attack by predators than harder older pinnulettes. However, that the fronds of the variety arachnoideum which grows in higher terrain where insect pressure is less important, contain a significantly higher amount of prunasin than the more heat resistant caudatum variety in tropical environments is not so easily understood. Higher altitudes are generally associated with lower insect pressure, diversity and temperatures. All may be important factors in determining the relative abundance of bracken cyanogenic phenotypes but point in opposite directions. Preliminary measurements in tropical bracken (caudatum) indicate that populations growing at 1050 m above sea level contain around 30% of cyanogenic phenotypes as opposed to 100% for those growing above 1500 m [14]. Also, a detailed study [198] encompassing a total of 141 clones among Australian and New Zealand populations of Pteridium esculentum, established that fronds collected from Queensland -to the north- were 0% cyanogenic with no prunasin albeit they still possessed glucosidase activity, whereas 13.6%) were cyanogenic in those from New South Wales. Additionally, 46.2%) of Victoria samples -to the south of Australia-, were CN active. The mean annual temperature gradient was the only apparent factor in such distribution but the ecological reasons for such phenotype selection remain obscure. Ecdysteroids Among the secondary metabolites in Pteridium produced by the mevalonic acid pathway there are the phytoecdysteroids [199-201]. These compoimds are actual arthropod moulting hormones. Therefore, phytophagous insect larvae that are exposed to critical amounts of ecdysteroids during feeding in principle may be at risk of moulting prematurely and thus jeopardize their development into successfiil reproductive imagos.

700

To the initial stir caused by the discovery of five phytoecdysteroids in bracken, a-ecdysone, ecdysterone, pterosterone, ponasterone-A, and its glucoside, ponasteroside, which counted together, converted bracken into the only known vascular plant to contain that many ecdysteroids, it followed the deception that only trace amounts, too little to cause any hormonal effect in insects, actually accumulated in the fronds. The body of ecdysteroids have been quantified between 0.3 and 53 |ag/kg of fresh frond in northern-type bracken [203]. Although ecdysteroids present in bracken but from another source (prothoracic glands of locusts) accelerated the moulting of the desert locust Schistocerca gregaria upon injection, bracken given as the sole or chief diet at critical times of nymph development did not affect ecdysis, growth or development, save for a reduction in wing size of adults. This was probably related to dietary deficiency but could not attributed to bracken xenobiotics [203]. It is relevant that the active ecdysones were probably dehydroxylated to aecdysone which passed unabsorbed to the fecal pellets. Being bracken a relic of the Mesozoic era when it evolved amidst ancient forest in which insect pressure conceivably may have been considerable, the production of ecdysteroids in today's ferns may be a vanishing vestige of a relatively ineffective line of chemical defense for herbivory pressures found in present-day ecosystems. a-Ecdysone, on the other hand, has been deemed responsible for the induction of neoplasms in toads [204] which, in spite of not being natural enemies of ferns, may function as physiological models for other real-world plant predators. Indanones Another class of terpenoids isolated from bracken is represented by a group of sesquiterpenyl indanones called pterosins and their glycosides, pterosides. These have been known for some time and have relatively simple molecular structures, but it is important to review them in some detail for they bear a close relationship with bracken carcinogenesis. Such compounds are not exclusive of bracken species but are amply distributed among the polypodiaceous ferns [205,206]. The 29 bracken pterosins known [207,211] resuh from variations in oxidation levels at positions C2, C3, and C6 with only one case in which a hydroxyl group is positioned on the aromatic methyl at C5, Table 1. The fact that pterosins arise from reactive illudanes (see below) also present in Pteridium, raises the possibility that some of these indanones may be arctifacts. In particular, exposure of ptaquiloside (1), a likely biosynthetic precursor, to chloroform yields small amounts of pterosin F within a few hours at room temperature. The inference is that the chlorine atom may stem from the chlorohydrocarbon. Along the same the same lines, pterosin O is detected after dissolving ptaquiloside in methanol [212].

701

Table 1. Structures and biological activity of terpenic indanones isolated from Pteridium ssp. [208]. 0 .Ri

Compound

Rl

R2

R3

R4

Cytotoxicity! IC550 mg/ml

1 PterosinA H CH2OH CH2OH CH3 320 H H CH2OH Pterosin B 100 CH3 CH3 1 Pterosin C H OH CH2OH 320 CH3 OH CH2OH CH3 1 Pterosin D >320 1 Pterosin E H H COOH 30 CH3 H H Pterosin F CH2C1 65 CH3 OH 1 Pterosin G H CH2OH 180 CH3 H II Pterosin H CH2CI CH3 CH3 CH3 H 1 Pterosin I CH2OCH3 CH3 OH Pterosin J H CH2CI >100 CH3 CH3 H CH2OH Pterosin K CH2CI >100 CH3 OH Pterosin L CH2OH CH2OH 180 OH H 1 Pterosin N CH2OH 220 CH3 H H 1 Pterosin 0 30 CH3 CH2OCH3 CH3 H 1 Pterosin V CH2OH ND CH2OCH3 H 1 Pterosin Z CH2OH 10 CH3 CH3 CH3 OH CH2OAC Acetylpterosin C H >100 H H CH2OBZ 1 Benzoylpterosin B ND CH3 H H CH2-isoND CH3 Isocrotonylcrotonyl pterosinb CH3 H CH2-palmityl 1 Palmitylpterosin A ND CH2OH CH3 H H CH2-palmityl 1 Palmitylpterosin B >100 1 Palmitylpterosin C ND CH3 OH CH2-palmityl H CH3 OH CH2-phenylND 1 PhenylacetylH pterosin C acetyl CH3 H CH2-glu 1 Pteroside A >320 CH2OH H H 1 Pteroside B CH2OGIU >100 CH3 CH3 CH2OGIU Pteroside C(*) H OH >320 H CH2OGIU Pteroside D(*) >320 CH3 CH3 H 1 Pteroside K ND CH2CI CH2OGIU CH3 H H CH2OGIU Pteroside P (**) ND CH3 CH3 H CH2OGIU 1 Pteroside Z ND CH3 1 ND [1 Wallichoside H CH3 OGlu CH2OH (*) There is a hydroxyl unit on P-C3. (**) The methyl at C5 is oxidized as CH2OH

II

1

702

Bracken pterosins are moderately cytotoxic [213]. Tests with HeLa cell lines [208] (Table 1, column 6), revealed that the same type of cellular lesions were occurring more or less independently of indanone structure. In the end, the authors concluded that, in spite of some morphological changes in the cells, no indication of carcinogenic activity of any of these compounds could be detected, in consonance with earlier extensive inspection [214]. There was no evidence of indanones acting as alkylating agents against cytoplasmatic or nuclear compounds, a prerequisite of carcinogenesis. Nor could it be shown that pterosincontaining bracken extracts induced cattle poisoning of the type observed when the animals ate the fern [215]. Only pterosin F has been found to be moderately active as feeding deterrent against the butterfly Pieris brassicae [190] but this is a specialist of Brassica plants and therefore it is likely to reject many plant metabolites outside this family. Additionally, pterosin F apparently is not a universally distributed indanone in all bracken genets. Contrary to tannins and other phenolics that are increasingly accumulated in cell vacuoles and other plant parts, pterosins in bracken appear in greatest concentration in the young croziers and decline rapidly as the fronds grow into maturity. In temperate regions, the highest accumulation of pterosins occur in May and June, reaching around 24 mg/g of fresh frond weight (pterosin F only), and wane to 4 mg/g of frond by August [216]. Others have reported only 0.015 mg of pterosin F per gram of dry frond in the var. pubescens [111]. This dynamic behavior is similar in the tropics [218,219] except that the time scale is shorter and does not depend on seasonality as such but on the ontogenic stage of the frond. Detected quantities are, however, dissimilar. Thus, in 12-day-old fiddleheads with unbranched rachis measuring 41.8 ± 6.5 cm in length within bracken stands growing at 1850 m above sea level in a rainy valley of the northern Andes during the short dry season, 0.272 ± 0.06 mg/g of frond biomass of pterosin A and 0.66 ± 0.04 mg/g of pterosin B were recorded in the neotropical variety caudatum. The fronds reached fiiU expansion between 20 and 30 days afterwards when the indanone concentration dropped to 0.012 ± 0.001 and 0.084 ±0.005 mg/g of frond biomass, respectively [218]. Pterosin F could not be detected in the caudatum fronds suggesting that the distribution of those indanones, so extensively reported by Yoshihira and Natori's group [208] in the variety latiusculum may not be a universal occurrence in the Pteridium taxonomic complex. Such vertical reduction in pterosin concentration in relation to phenological stage may indicate that indanones are produced temporarily and are diverted to a still unknown destination within the plant. Pterosins are not shed into the environment, so they are possibly a carbon source that is consumed as the blade grows, rather than being accumulated along with other protective secondary metabolites in cell vacuoles. Further bioecological testing of these indanones, of which there is a paucity of data, should shed light on the roles played by pterosins in ferns.

703

A unique proto-illudane A few other terpenoids of special interest in bracken have been discovered in recent years. Of particular relevance to the biosynthesis of bracken illudanes is the finding of the protoilludane pteridanoside (2). This is a minor component (3.2 ppm) of the fronds of the neotropical variety caudatum from the Venezuelan Andes [220]. This is the first tricyclic sesquiterpene possessing afiisedcyclobutane to be discovered in the Pteridaceae family and unique in holding a sugar moiety, although other protoilludanes of higher oxidation states have been isolated recently from the fungal basidiomycete, Laurilia tsugicola [221]. Pteridanoside is moderately toxic eigsdnst Artemia salina with LC50 =250 and 62.5 |Lig/ml after 24 and 48 h of exposure, respectively. This activity may contribute to the overall bracken aggressiveness against vertebrates. The aglycone, pteridanone, obtained by enzymatic hydrolysis with P-glucosidase, exhibited no measurable toxicity, thus showing clearly that the sugar is involved at some critical stage of the mechanism responsible for pteridanoside's toxicity.

GluO OGIu

OGIu . ^ ^ ^

Fig.(2). Ptaquiloside (1) and other terpenoids recently discovered in Pteridium ssp.

704

A novel thymol derivative Additionally, a novel Oi-glucosyl-4-hydroxythymol derivative named pteridioside (3) has also been isolated recently [222]. Interestingly, the glucose portion appears on the most hindered carbon Ci which is ortko to the isopropyl group. In spite of this inherent instability, pteridioside is not part of the toxic fraction of bracken. Further, from the methanol extracts of dried fronds of the New Zealand bracken Pteridium esculentum was isolated a bicyclic sesquiterpene with a cadinene skeleton (4) [132], of which the biological activity is so far undetermined. On the other hand, although diterpenes of the e«r-kaurane, enf-atisane and e«r-pimarane types have been isolated from the related Pteris and Microlepia ferns [206], bracken and other plants from the Dennstatedtiaceae family appear devoid of the pathways that allow their biosynthesis, as these group of compounds has not been detected in these ferns. Other aggressive compounds Bracken toxicity does not end with the described compounds, but includes others of great economic importance such as still chemically uncharacterized antithiamine factors [223-225] that are particularly harmful to monogastric mammals causing varying degrees of paralysis and cerebrocortical necrosis [226]. Direct consumption of bracken parts by humans may also lead to thiamine deficiencies [227]. Fortunately, this factor is 10 to 30 times more abundant in the less accessible rhizome than in the fronds and is likely to be washed away during the culinary preparation of starch from the rhizomes because no antithiamine activity has been recorded in people consuming it. Thiaminase is probably an enzyme since it becomes inactivated by autoclaving. The offensiveness of silicon oxide deposits [228] in mature fronds, which accumulates especially in the epidermal cell layers [229] ought not to be neglected since it not only imparts additional toughness to tissues already hardened by lignin, but its edgy phytolites act as abrasive material on the mandibles of phytophagous insects and mammals teeth. Furthermore, plant silica has been postulated as a possible contributor to carcinogenesis [229]. Other biologically active but still not fully characterized compounds have been observed in bracken. This is the case of braxins Al and A2 of which more is known about their toxicology than their chemistry at the time of this writing. The compounds were isolated from bracken rhizome in up to 0.06% [230]. The braxins affect rat mast cells by swelling and promoting the release of histamine [231,232]. This mechanism may be involved in the episodes of allergic skin rashes that some people suffer after contact with bracken stands. It has also been reported that the effects associated with BEH have been induced by braxin Al in di guinea pig.

705 But its elusive molecular structure and the innards of its physiological effects in laboratory animals still remain to be elucidated. In spite of this powerful battery of chemical weaponry, it is the illudanes which impart the greatest hostility to bracken's defense mechanisms.

Bracken illudanes: the ultimate carcinogens The quest for the bracken carcinogens lasted many years. The many difficulties have been reviewed [36] and in short, were chiefly due to: 1) the lengthy time required for a reliable carcinogenicity test in rats, lasting 6 months or more, to pursue the bio-directed separation of the active component; 2) the low yield of the extraction procedures, 3) the instability of the active material. As a consequence, many dead ends preceded the final isolation of the carcinogen. At one point for example, after a carefiil examination of the water soluble fraction of bracken fronds where the carcinogenic activity could be concentrated, it was believed that the active compound was a C7H8O4 whose partial structure, according to the IR spectrum, was that of a fiiran derivative possessing a CH2-COOH unit [233]. 10 mg of this material administered by intraperitoneal injection in mice, caused their demise only within a few hours. The complete structure was later elucidated as being that of shikimic acid but, interestingly Leach and the Evans [233] determined that the activity was lost by alkaline treatment and "appeared to polymerize when heated under acid conditions". These characteristics were appropriate of the then undiscovered ptaquiloside (1) (see below), the true carcinogen, but whose structural formula and other traits differed completely from the proposed molecular fragments. Further analysis of the aqueous fraction from croziers and rhizomes [234], published about the same time that the structure of ptaquiloside was finally unveiled, furnished 2,3-butanediol, 3-methylbutan-2-ol, mono-methylsuccinate, methyl-5-oxoproline, 2(3H)-dihydrofuranone, trans-2-methyl-cyclohexanol, and other low molecular weight materials [234]. None of these compounds was biologically active individually but a single dose of 4.5 mg of the fresh aqueous extract of the croziers from which most of these materials had been isolated, induced lymphocytic leukaemia in mice. It is also historically piquant that seven years before ptaquiloside was identified, McMorris and co-workers [235] had been looking into the isolation, identification and chemical transformations of femindanones

706 of the bracken pterosin type. Among these compounds was onitin (5), isolated from the fern Onychium auratum. McMorris surmised with foresight that onitin bears a close resemblance to an indanol (7), which he had obtained more than a decade before by simple catalytic hydrogenation of illudin-M (6) [236]. This spyrocyclopropane is a strongly active material isolated from the basidiomycetes Omphalotus illudens and Lampteromyces japonicus [237] (Scheme I). Illudin-M, an acylfulvene, has been a subject of active research for a number of years due to its powerful antitumor activity [238]. McMorris inference established a perfectly reasonable chemical link between the already known but inactive pterosins in bracken and a possible series of illudane precursors of these pterosins. Putting together this information with Leach and Evans' experience on the instability in alkali and acid of their purported bracken carcinogen [233], which are adequate conditions for disconnecting a conjugated cyclopropyl unit, might have allowed researchers to focus their attention on looking for illudanes in bracken, and ease the search for ptaquiloside.

CH3O

McMorris, Liu, and White (1976)

Fig. (3). Chemical relationship between the pterosin-type skeleton and the spyrocyclopropane illudin-M (6), established years before ptaquiloside was discovered [235].

707

Ptaquiloside From the carcinogenic and mutagenic aqueous extract of bracken fronds, ptaquiloside (1) was finally isolated and fully characterized, simultaneously by research groups in Japan and The Netherlands [239,240]. While Niwa, Hirono and collaborators employed the time consuming rat carcinogenesis bioassay to follow the fate of the active fraction, van der Hoeven and co-workers used the expedite mutagenicity analysis employing Salmonella tiphymurium/microsome test with the same end result. An improved extraction procedure was later published [133] and the chemical structure was secured by the x-Ray diffraction study of its tetraacetate derivative [241], which was notably more stable than the natural ptaquiloside. Today, this illudane can be obtained with little loss by brief hot water extraction of the thoroughly crushed fresh croziers, centrifugation and passage through a column of polyamide 6N where pterosins and the non-polar material will be retained. The aqueous fraction is further separated using reverse-phase (ODS) preparative HPLC. Ptaquiloside is unstable to alkaline conditions. 80% of the product will be decomposed in only 10 min at pH 11.5 and room temperature, and will last less that one hour at pH 10. Nor will it survive when subjected to acid, heat, and prolonged contact with active surfaces such as neutral alumina and silica gel, and non-neutral polymer-based chromatographic materials. Simply standing at room temperature and ambient moisture will also promote rapir decomposition [90]. Effective isolation must be performed by reverse-phase HPLC and immediate cooling of the fractions. Freezing to -80 °C or lower must be enforced if storage for long periods of time is desired. Sunlight -especially as pure compound but not so much in the live frond itself- also induces the decomposition of ptaquiloside, but this phenomenon still awaits explanation. However, one may speculate that the considerable accumulation of UV-absorbing tannins and phenolic derivatives in the large vacuoles of the leaf parenchyma cells acts as a sun screen in the live plant, shading ptaquiloside and other light sensitive compounds and organelles from the incoming destructive high energy photons. This chemical lability may be understood better if the molecular structure of ptaquiloside, rather peculiar in various respects, is examined closely. Firstly, by looking at a 3D representation of Fig. (4) one can discern that the cis-AB ring fusion forces out the spyro-cyclopropane in such a way that, while it contributes to block the approach of possible nucleophiles on the carbonyl carbon from the P-side, at the same time it exposes like a tail the cyclopropane methylenes to possible reactants in the medium. The approach to the carbonyl carbon is fiirther impeded from

708 the a-side of the molecular plane by the Cio methyl and the entire glucose fragment, as Fig. (4) depicts. The combination of these two factors makes the cyclopropane moiety more liable to undergo attack than the other electrophilic centers in the molecule, i.e. the ketone.

Fig. (4). Three dimensional sidewise view of a MM4 minimized structure of ptaquiloside (1), showing the protruded cyclopropane ring. The glucose moiety can be seen protecting the a side of the molecular plane.

Additionally, this cyclopropane is set in conjugation with the C5-C6 insaturation in such a way that the bisected geometry derived from its spyrocyclic disposition maximizes the overlap between the n and cyclopropane 3d' molecular orbitals. The setting is ideal for the constitution of a non-classical carbenium ion (8) if a proton is trapped by the C5-C6 double bond, Fig (5) route A. The nucleophilic attack of water from the medium would yield the rupture of the cyclopropane unit, and water elimination from the tertiary alcohol in the resulting diol (9) would furnish a conjugated dienone (10). Aromatization to the observed product, pterosin B (11), would occur by acid-induced glycolysis. Conversely,

709 protonation at the tertiary carbinol on Cg (Scheme II, route B) might lead to another non-classical cyclopropyl carbinyl carbocation (12) that easily leads to pterosin B again along similar steps. These two alternatives would explain the instability of ptaquiloside in acidic aqueous medium.

OGIu

OGIu

"'^oM Pterosin B (11)

12

^oid CXBIu

10

OGIu

13

Fig. (5). Possible decomposition routes for ptaquiloside to pterosin B in an aqueous acidic medium.

Secondly, the glucose fragment is attached to the aglycone at the encumbered C4 angular carbon. Molecular modeling, Fig. (6), reveals that this glucose portion carmot rotate freely for the angular a-proton at C9 will encounter the pyranosyl endocyclic oxygen atom or the hydroxyl oxygen on C2'. This becomes an intramolecular acid-base combination whose components can get as close as 1.62 A, so it is feasible that these conformations might lead to favorable six- and seven-membered cyclic transition states, respectively, as in Fig (6), that could spur the cisperiplanar (3-elimination of glucose to give rise to illudane-dienone (14). The stark contrast between ptaquiloside lability and the stability of its gluco-tetraacetate [241] where the basicity of carbinol oxygens is reduced and the approach of the pyranosyl oxygen is impeded by the additional steric hindrance of the acetates, lends support to the mechanism of Fig (7).

710

Fig. (6). Three dimensional view of ptaquiloside showing a possible rotamer of the glucose moiety such that the endocyclic pyranosyl oxygen and angular C9 hydrogen atoms -shown within circles- are close together.

As opposed to ptaquiloside, illudane-dienone (14) now displays an extended conjugation from the spyrocyclopropane moiety all the way to the carbonyl unit across eight carbon atoms in a fairly planar conformation. Trace of acid would protonate the carbonyl oxygen, strongly assisting the natural electrophilicity of the conjugated cyclopropane. The participation of water on the cyclopropyl methylenes and aromatization would lead again to the observed pterosin B (11). This dienone (14) was first characterized by Niwa et al. in 1983, [239] and may be observed in NMR experiments when non-nucleophilic solvents such as acenotonitrile are used. More importantly, (14) is presently deemed the true carcinogen of bracken (read below). The fast conversion of ptaquiloside in base may occur through a different mechanism that possibly involves attack by base on the a-

711 carbonyl proton at C9 with concommitant glucose elimination to yield again the pivotal illudane-dienone, which would quickly react electrophylically with the hydroxide ion as described above. In fact, other oxygen or nitrogen based nucleophiles, natural or synthetic, may intervene at this stage to yield a variety of pterosins. Protic acid assistance at the carbonyl oxygen may be necessary for the transformation of 14 to pterosin B because the dienone is stable in mild base. This last step lies at the base of the proposed mechanism for ptaquiloside carcinogenesis.

HO / PTEROSIN B

Fig. (7). Intramolecular conversion of ptaquiloside to pterosin B, as proposed AlonsoAmelot [14].

These mechanistic considerations are crucial in understanding the true nature of the toxicity of (1) and (14). A picture is beginning to emerge in which ptaquiloside appears as a pro-toxin, only one self-induced step away from the true toxin, dienone (14). Plants have a difficult time in dealing with their own toxic metabolites because these may interfere with their own metabolic processes, generally similar among living organisms. This is probably why many natural toxins are stored in cellular vacuoles and other special organs away from cytoplasmatic chemistry. However, a more advantageous strategy would be to produce relatively inactive protoxins which may be rapidly converted into the true toxins once they have entered the target organism. Ptaquiloside appears to fulfill this pattern and clearly illustrates that plant evolutionary chemistry takes intriguing and ingenious paths of extraordinary curmingness.

712 Ptaquiloside toxicity Soon after ptaquiloside was discovered, it became clear that it held responsibility for the most important biological activities in mammals attributed to whole bracken fern. Thus, 91 % of the Sprague-Dawley rats given orally 780 mg of ptaquiloside per kg of body weight followed by 100-200 mg/kg once a week for 8 weeks developed mammary tumors as soon as 81 days after the feeding experiment started. Although these amounts are apparently massive, they correspond to about 12 g of fresh bracken fronds per day per animal, a quantity perfectly manageable by laboratory and field rats. Ileal tumors lasted longer to appear. However, histological types of the observed tumors were adenocarcinoma, papillary carcinoma and anaplastic carcinoma, which were the same as those observed in rats fed a whole bracken diet. [242]. Bladder carcinomas were not detected but preneoplasic hyperplasia of the urinary bladder mucosa was observed in 80% of the autopsied animals. Thus, if given enough time, these rodents were likely to develop bladder malignancies. The specific symptoms associated with acute bracken poisoning were reproduced in calves by oral administration of ptaquiloside as a saline solution during 42 days [243]. Dosages reached a peak of 1600 mg/animal/day after day 38. Among the most remarkable effects were severe dilatation of small blood vessels, a vertical increase in leucocyte counts during the first 40 days up to 25.3 x lO^/liter, a sharp drop in neutrophil counts down to 0.1 x lO^/liter, and transmogrification of the femoral bone marrow to fat. No damage to the urinary bladder was observed though, which was thought to be a consequence of ptaquiloside instability during the time required by the experiments. Insufficient experimental time probably was also to blame as it had been shown that the appearance of vesical hematuria requires much longer exposure (several hundred days) [72]. Mode of action of ptaquiloside: how the picture unraveled Further research soon unveiled the ominous irruption of ptaquiloside into genes. The mutagenic capabilities of ptaquiloside were observed early, actually during its isolation and characterization by its codiscoverers van der Hoeven and co-workers (1983) who used the Salmonella typhimurium mutagenicity test to pursue the separations as indicated above. This genotoxicity was later confirmed in a whole [244]. organism such as the Suit fly Drosophila melanogaster Interestingly, Nagao et al. [245] showed that ptaquiloside by itself was not mutagenic at pH 7.4 but pre-incubation under more alkaline conditions (pH 8.4) was a requirement to transform it into a strongly

713

mutagenic substance. According to what we discussed above, this active material was illudane-dienone (14). Other similar compounds (15 - 22) but devoid of a properly activated cyclopropane [245] were inactive. Notably, isoilludin S (15) which possesses a conjugated cyclopropyl ketone with a proper bisected conformation, was not mutagenic to the Ames test, nor was glycol 18 in which the dihedral angle between the cyclopropyl methylene and the tert-carbinol is 93 degrees in the minimized structure [14]. This is very close to the required conformation for maximum orbital overlap and hence facilitation for the departure of the hydroxyl group in the event of carbenium ion formation. This evidence lent support to the idea that the cyclopropyl unit with a proper conjugation like that present in ptaquiloside and its dienone is pivotal in imparting biochemical activity. A different mechanism assists acylfulvenes, illudins M (16) and S (17), in their alkylation of DNA [238] but the opening of the cyclopropyl carbinol is also involved. AcO

16

HO

19

20

21:R = CH2-COOH 22: R = H

Nagao^ra/. 1989. Fig. (8). Structures of cyclopropyl and cyclobutyl perhydroindane derivatives tested for mutagenicity and found inactive [245].

On the other hand, doses as low as 4.5 |Lig/ml of ptaquiloside at pH 7.4 (no dienone present) elicited chromatid exchange-type chromosomal aberrations in hamster lung fibroblast cells [246]. The illudane-dienone

714 caused the same effects at the same concentration and slightly higher pH. That (14) was likely to be involved even under neutral pH was surmised from the need to add as much as 400 |ig/ml of ptaquiloside at pH 5.3 to bring about a comparable clastogenic effect. As opposed to its relative stability under diluted alkali, (14) is very unstable in acid and is quickly transformed into inactive pterosin B. By using the hepatocyte primary culture DNA repair test [247] with liver cells of adult ACI rats, an essay that is capable to discern true carcinogens from solely mutagenic compounds such as quercetin and other bioactive flavonoids, Japanese workers [248] were able to show that ptaquiloside is a genotoxic carcinogen. The key evidence was that ptaquiloside induced DNA repair in the exposed liver cells, which could be interpreted as an intrusion into the DNA molecule that caused enough distortions to trigger the reconstruction mechanism.

H2N.^.N.,^0

14

27

26

Ojikae/a/.(1989). Fig. (9). Covalent binding of adenine to ptaquiloside. An aduct of guanine could also be obtained.

715 Such molecular trespassing of ptaquiloside was in line with the momentous discovery that 1 binds covalently to purine bases in DNA and also cleaves the double helix, exposing dangerously the nucleophile centers of the unveiled bases to the strongly alkylating prowess of ptaquiloside [249], Fig. (9). Under hydrolytic conditions (pH 7.0, 90 ^C, 20 min) Ojika and co-workers isolated guanine and adenine adducts (26) and (27) is small yield. The anti-tumor antibiotic CC-1065 (28) [250], which also displays a highly electrophilic, conjugated spyrocyclopropane in its convoluted structure. It is precisely at the cyclopropyl methylene where the Ns-adenine sites of DNA alkylate CC-1065 as in 29, in clear consonance with ptaquiloside. Fig. (10).

Reynolds e/a/. (1985). Fig. (10). Alkylation of adenine with anti-tumor antibiotic CC-1065, which possesses a Spyrocyclopropane in conjugation with a carbonyl group through four carbon atoms.

Refinement of the DNA aberration experiments under physiological conditions disclosed further intimacies of the molecular mechanism involved. The chemical backdrop here is that DNA contains a nimiber of nucleophilic sites, chiefly carbonyls, hydroxyls, phosphate oxygens, amines and heterocyclic nitrogens, in addition to those illustrated above. All are potentially reactive against strongly electron-deficient agents.

716 Carcinogenesis is believed to start with the 0/N-alkylation of specific DNA nucleosides of a proto-oncogene causing point mutations. Transduction of these mutations as errors in protein synthesis via altered t-RNA may produce defective proteins and cellular disfunction [251,252]. Although ptaquiloside or its dienone are natural electrophiles that react with various biological nucleophiles such as aminoacids, nucleotides and nucleosides, the observed clastogenic effects were a clear suggestion that DNA was ptaquiloside main biological target. Conceiving the illudane-dienone (14) as alkylating reagent, or activated ptaquiloside (APT) as it began to be named among bracken cancer researchers, under the presumption that ptaquiloside may generate APT within the cell by liberation of its glucose at a given point, it was observed that, after 25 h of incubation at 37 °C, APT converted covalently closed circular DNA into an opened form. A model alkylation of a cytosine-guanine-adenine-tyrosine (CGAT) tetranucleoside with APT furnished the A/G debased material [253]. This process was explained by a first step where N3 of adenine and, more slowly, N7 of guanine undergo alkylation by APT, furnishing an unstable adduct (30) that sheds the modified bases by braking the Nglycosidic linkage as in 31 [253]. Consequently, the phosphodiesterpentose backbone becomes so unstable that it brakes apart to 32 via [3elimination, causing severe disruption of the DNA chain. Fig. (11) illustrates this for the alkylation of an adenine site. The cleavage step may also occur naturally by intervention of cellular apurinic endonucleases. The loss of purine bases or depurination is a spontaneous alteration of DNA that occurs at a rate of 10^ depurinations per mammalian cell per day. It plays a key role in spontaneous mutagenesis and possibly aging because these losses inhibit chain elongation of the replicate by DNA polymerases [254,255]. The additional base-blockings and depurinations derived from strongly alkylating agents such as APT may increase this rate to levels that eventually could promote cytotoxicity and the development of malignancies. This picture was confirmed by 32p. postlabelling assay [256]. While this theory was shaping up to account for ptaquiloside carcinogenicity, some inconsistencies appeared. Extensive examination of several alkylating agents revealed the existence of a correlation between the site of alkylation in DNA and the final effect [257]. Alkylation at the 06 of guanine, O2 and O4 of thymine, and O2 of cytosine in the DNA polymer could be correlated with procarcinogenicity. whereas reactions at the nitrogen positions of the heterocycles resulted in enhanced levels of cytotoxicity. Application of the hard-soft acid-base HSAB theory [258,259] to alkylation of DNA nucleosides [260] provided support to this view. APT (14) may be classified as a semi-hard electrophile which is be prone to react with less hard N over harder O nucleophiles. If so.

717

BASE

DEPURINIZED DNA HoO

BASE

^Wo

31 JD^O

>^

1—0

\

ASE

BASE

O

i Y

7

HO

FRAGMENTED DNA

r^\

H,0 33

718 Fig. (11), previous page. Mechanism of DNA depurination promoted by the alkylation of purine bases by ptaquiloside, as proposed by Kushida et al. (1994).

ptaquiloside and APT would lead to cytotoxicity preferentially, which also being present, is nonetheless not considered the main molecular activity of ptaquiloside in the living system, but carcinogenesis. To complicate matters even further, molecular orbital calculations of electron densities of nucleosides within the DNA strand showed that the accumulation of electron density among the various O and N nucleophilic centers and accessibilities to external electrophiles varies greatly with the particular position that the bases hold in the polymer [261]. The disagreement between these predictions and the observed reactions may be reconciled by the gross difference between in vitro reactions and those occurring within the cell. The emerging picture, therefore, appeared to be one in which site specific attack of ptaquiloside or APT and not a generalized non-specific adenine-guanine N-alkylation actually lied at the base of the observed toxicities of these illudanes. The attention thus turned to find APT adducts at specific sites of cancer-related genes such as neu and ras oncogenes and the p53 tumor suppressor gene. Ras proto-oncogenes for instance are activated if mutations occur in codons 12 or 61 by interaction with certain alkylating agents such as nitrosamines and dimethylbenzanthracene. The normal ras genes yield a protein p21 involved in the control of cell proliferation, but the genes with modifications in those specific codons furnish a mutant protein unable to participate in this control and cells begin to divide without regulation. A tumor thus ensues. Urinary bladder carcinomas in the rat [246] and of the upper alimentary tract related to bovine papilloma virus [262] have been shown to be associated precisely to H-ra^ activation [283]. The question of ptaquiloside carcinogenesis turned to find altered H-ras protooncogenes in specific organs where bracken-derived cancer had been observed in animals. This was explored in young calves given a regular diet to which 100 g of powdered bracken, a rather moderate dose, had been added for 28 days [263]. DNA was extracted from various internal tissues selected among those showing visible lesions. This DNA was examined for APT alkylation by way of ^^P post-labeling of the hydrolyzed nucleotides. Indeed, the APT-DNA adducts were found only in the ileum and the urinary bladders of the calves, precisely where malignancies from bracken consumption are well known to occur. As opposed to this, other potentially exposed organs but generally not involved in bracken toxicosis such as liver, kidney, mesenteric lymph, and the epithelium of pharynx and oesophagus were devoid of such adducts. It is no coincidence that the

719 ileum and the urine of mammalian herbivores are alkaline, thus providing adequate conditions for the transformation of ptaquiloside into APT. More importantly, Prakash and coworkers [263] discovered by PCR amplification and sequencing of the H-ra^ gene from the ileum, that a mutation had taken place exactly on the adenine of codon 61. APT alkylation is not base-sequence-specific so several adenines, which occur frequently in DNA, would be expected to become APT-alkylated. However, it is known that the adenine-adduct depurination step is sequence specific and the CAG triad at codon 61 of U-ras fits this pattern. As a consequence, depurination at this codon would rest a the base of the synthesis of defective protein/72i and eventual carcinogenesis of the ileum and urinary bladder. This model was also applicable to laboratory rodents administered intravenously and intragastrically with APT [42,89,283]. Again U-ras activation was witnessed. APT alteration of other cancer related genes such asj^Ji awaits fiiture studies. However, in the presence of bracken quercetin, this gene becomes dysfimctional and looses its transcriptional activity [159,160] as it was described above. The combination of APT and quercetin in the same organism ingesting fern blades, hitherto unexplored together, should lead to profound genotoxicity and a more intense one than with ptaquiloside alone. Thus, intake of whole fern blades is bound to be more toxic than the individual components. Other bracken illudanes Ptaquiloside is not the only illudane present in Pteridium., The clear chemical connection between ptaquiloside and pterosin B, carries within the suggestion that at least some of the myriad indanones that bracken synthesizes is the result of the biosynthetic progress from ptaquilosidelike precursors. Supporting evidence was drawn from the fact that ptaquiloside and pterosin B concentrations vary simultaneously with aging of the fironds [124,218]. With this idea in mind, Castillo and coworkers pursued the search of additional illudanes in this fern. Considering that the largest concentration of ptaquiloside is found in the crozier and that this is the phenological stage of greatest toxicity and carcinogenicity, their attention was concentrated on the aqueous extraction of the young fiddleheads. From the neotropical hybrid caudatum collected in the Venezuelan Andean mountains, three more illudanes similar to ptaquiloside were isolated and fiiUy characterized. These were isoptaquiloside (34), caudatoside (35) [212,280] and Ptaquiloside Z (36) [264]. These compounds were present in 0.4%, 0.6%, and 0.2%, respectively, along with 1% ptaquiloside which is the major illudane present in other samples of the same variety. Also, that the new

720

illudanes were not discovered earlier in other Pteridium sp suggests that these are minor constituents in the genus. Caudatoside and ptaquiloside Z, true CI5 sesquiterpenes and whose spectral characteristics only differ slightly from ptaquiloside (Table 2) are underivatized glucosides and proved to be even less stable than the latter, furnishing rapidly the corresponding dienones and pterosins. Although formal kinetic studies have never been performed in any of these compounds, it is conceivable that the added instability is the result of the steric compression of the additional alkyl group on C2 on the curved trough of the p side of the molecular plane that provides steric acceleration for the conversion to the more planar and considerably less congested dienone. Not surprisingly, these are also toxic with LC50 (48 h) in the order of 8 |ig/ml against brine shrimp nauplii, a value comparable to that of ptaquiloside. However, their recent discovery has not allowed enough time for more ample results to come through and studies on their carcinogenic and mutagenic potential are likely to be under way. HO

,

.. O

OGIu 34 HO

OGIu 35: R = CH2OH

38: R =

36: R = CH3 39:

.=vSa

Saito et al. 1990, Castillo et al. (1997, 1998). Fig. (12). Structure of other illudanes found in neotropical Pteridium aquilinum var. caudatum and other ferns of the

721

Table 2: IH RMN data (400 MHz) of bracken illudanes in C5D5N (ppm) (Castillo et al., 1997,1998). Coupling constants are given in Hz. O V

.10

C proton

3a

2.26 m 2.17m

isoptaquiloside 2.75 m 2.75 m

+++++ 2.51 d,j=13.4

3p

2.68 m

2.41m

3.46dj=13.4

5 9a 10

6.06 s 2.97 j=l.2 0.96 dj=6.5

6.18s 3.27 bs 1.06 s

11

+++++

6.23 s 3.14 bs 1.14 d, j=6.6 +++ +

|2p

ptaquiloside

12 1.44 s 0.47 m 13a 0.84 m 13P 0.88 m 14a 1.08 m 14p 15 1.70 s (*) measured in CD3CN

1.47 s 0.58 m 0.76 m 0.76 m 1.00 m 1.48 s

caudatoside

3.53&4.09,d, J=io 1.45 s 0.62 m 0.80 m 0.86 m 1.16m 1.68 s

ptaquiloside

zn

+++++ 2.15 d, j= 13.4 2.18 d, j= 13.4 5.70 s 2.75 s 1.08 s

1

0.95 s 1.51s 0.66 m 0.81 m 0.53 m 0.84 m 1.20

Ptaquiloside-like illudanes are not unique to Pteridium sp. On the one hand, ptaquiloside itself was discovered in the rock fern Cheilanthes sieberi from the southwestern Pacific [265]. This fern was investigated in view that it thrived in bracken-free farms in parts of Australia that had a history of bovine or ovine enzootic hematuria. The content of ptaquiloside was highest in the crozier and amounted to 0.034% of the fern biomass, only a fraction of the ptaquiloside found in Pteridium but still high

722

enough to cause the disease. This results clearly indicates that the amounts of isoptaquiloside, caudatoside, and pteridanoside found in P. aquilinum var. caudatum are potentially damaging to farm animals. As with bracken, no ptaquiloside was found in the spores of Cheilantes. The ferns Pteris cretica and Histiopteris incisa also proved to contain ptaquiloside [266]. On the other hand, hypoloside A (37) was identified in Hypolepis punctata and hypolosides B (38) and C (39) in Dennstaedtia hirsuta. All are analogs of Ptaquiloside Z except that C8 is epimeric, whereas the aglycone of dennstoside A from rhizomes of Dennstaedtia scabra is identical to caudatoside. These illudanes could be converted easily with acid, base or heat to their corresponding pterosins, showing that they also behave as strong alkylating agents and therefore may act as possible carcinogens. Indeed, hypolosides B and C showed similar mutagenicity in the Salmonella typhimurium test [245], and caused chromosomal aberrations in a Chinese hamster lung fibroblast cell line, just as ptaquiloside did [246], whereas hypoloside A was less active for still unknown reasons. Further research may disclose how DNA-active these spyrocyclopropanes, from bracken or other ferns, are. Ptaquiloside analysis Taking advantage of the quantitative conversion of poorly UV-absorbing ptaquiloside into highly conjugated pterosin B under very mild conditions, various methods of analysis have been devised. Of practical use for semiquantitative studies in low tech or field laboratories is the method of Saito et al. [90]. A sample of the aqueous extract of bracken or other ptaquiloside-containing ferns dissolved in methanol, is spotted on a 20 X 10 cm silica gel TLC plate in a bidimensional array. The first run is performed on the longer side with benzene-acetone (3:7). Then the plate is heated at 110 °C for 2 h to induce the decomposition of ptaquiloside into pterosin B. The plate is run again in the perpendicular direction using benzene and ethyl acetate (1:1) and adding pterosin B standards of known concentration on an unused portion of the plate. Dichloromethane/acetone 1:1 may be used if benzene is to be avoided [14]. Spots are developed using UV photoexcitation at 254 nm or iodine vapor. The presence of ptaquiloside in the sample will be detected by the appearance of a spot above the diagonally distributed compounds, and with the same Rf as the pterosin standard in the second run. The plate may be examined under a TLC optical densitometer to derive a semiquantitative measure, but this procedure may be fraught with some problems [219]. A more sensitive and accurate method for quantitative determinations is the reverse-phase HPLC analysis, albeit more expensive and time

723

consuming, independently developed by Agnew and Lauren [267], Alonso-Amelot et al. [219] and Burkhalter and co-workers [279]. These methods rely on the prior clean-up of the aqueous extract through a solid medium such as polyamide or silica gel columns, of the pterosin B derived from the base treatment of ptaquiloside followed by acid to end the conversion of the illudane-dienone (14). Quantification of pterosin is determined at 260 nm in the flow UV-vis HPLC detector and comparison against a calibration regression. The sharp chromatographic separation of pterosins A, B, and Z allows the quantitation of the other bracken illudanes also in one single HPLC run. The direct observation of unconverted ptaquiloside is also possible at 220 nm [267,279] but it suffers from a 5 to 6-fold reduced sensitivity. It is possible to detect as little as 5 |ag of ptaquiloside-equivalents per gram of plant material. Ptaquiloside and the rest of bracken illudanes have never been synthesized as glycosides so researchers still depend on the natural source to study them further. However, various syntheses of the ptaquiloside aglycone, ptaquilosin, have been reported in recent years [268-272]. This compound has been converted under mild base into illudane-dienone (14) [269] which allows to pursue the genotoxic studies of ptaquiloside using a synthetic material. Interestingly, both enantiomers of ptaquilosin were examined for their carcinogenic potential, and the dienone from the natural enantiomer was found to be more potent with respect to DNA cleavage than that derived from the unnatural precursor [271,272] but the exact mechanism is unknown. That the natural enantiomer is the most genotoxic is an intriguing innuendo that environmental pressure by mammal herbivores has directed the evolution of the biosynthesis of ptaquiloside in a enantiospecific manner. The content of Ptaquiloside in Pteridium^ ecological considerations and human health implications With the advent of appropriate analytical methodology, it was possible to examine more closely the production of ptaquiloside in nature. Surveys indicate that the content of ptaquiloside in fronds varies greatly, from below detection limits to a gross 12.9 mg/g of dry plant material [273]. It was significant that the content distribution was higher than a potentially toxic 5 mg/g for 15% of the samples examined from a population of eastern Australia. Samples drawn from different locations around the world but grown in Sydney along the Australian plants yielded lower levels. Fig. (13), depending on bracken taxa (Table 3). Another source of variation was the developmental stage of the bracken ramets, younger croziers showing the highest concentration of ptaquiloside which rapidly

724

decreased as the fronds matured [124,218]. This finding endorsed the earlier observation that croziers were the most toxic of the fronds [85]. Table 3: Concentrations of ptaquiloside found in various taxa of the Pteridium complex. Data from (a): Smith et al., (1994); (b): Alonso-Amelot et al., (1995). Location

Taxa

Australia New Zealand United aquilinum Kingdom revolutum Indonesia revolutum ' Australia latiusculum (*) arachnoideum (*) arachnoideum Venezuela caudatum (*) Venezuela caudatum Pseudo(*) caudatum decompositum (*) decompositum Venezuela arachnoideum 1 stage 1 2 esculentum esculentum

3 4

1

1

caudatum stage 1 2 3

ref 1

Mean (mg/g) 2.4 1 0.079

SD

n

2.4 0.038

1 23 5

(a) (a)

0.919

0.361

4

(a)

0.369 0.197 0.496 0.363 0.378 1.685 3.005 0.048

0.336 0.186 0.749 0.292 0.317

4 3 7 3

(a) (a)

J^

(b) (a)

0.970

0.956

2 12 1 1

0.378 6.05 X 10-4 2.43 X 10-6 5.41x 10-5

0.317 9.20 X 10-4 4.21 X 10-6 4.80X 10-5

12 3

3.005 0.569 0.677 0.334

0.970 0.174 0.075

12 3 3

(a) (a) (b)

1 1 1

(a) (a)

3

Venezuela

4

1 0.296

[2

(b) (b) (b)

lib)

1

(*)Location not specified.

It has been shown that the content of secondary metabolites varies greatly with the particular microclimatic and edaphic conditions, and a more systematic study in this connection awaits development in bracken. Attempts in this direction, nevertheless, have revealed that Pteridium

725

esculentum accumulated higher concentrations of ptaquiloside as the latitude of collection along a north-south transect following the eastern coast of Australia was higher [273]. Lower mean temperatures associated with this ecological gradient were deemed responsible for part of the effect but genetic factors were also called upon to account for the difference. A parallel was drawn with the variation of frequency of cyanogenic phenotypes of bracken, which also was significantly higher in fern populations of more southerly, hence cooler latitudes [198].

Content of ptaquiloside in bracken fronds 1

>12.0 L •o

c

H World

>8.0 ^ _ |

• Australian

>4.0 ^ 2 ^ ^ H o E

>2.o jjumiig^ >i.o i i m i B i ^ f j U j ^

^^^ ' iHHHtaHHHHH 20

40

60

80

% of the plants tested

Fig. (13). Distribution of ptaquiloside concentration among bracken samples collected in Australia and around the world. Data redrawn from Smith et al., (1994).

The situation is still unclear, however, because the caudatum variety in northern South America accumulates ptaquiloside in inverse proportion to the elevation. Caudatum plants collected in grassland on mountain slopes along a 1000 m altitudinal gradient and similar nutrients content and water regime contained from 0.3 mg/g in the lower tier down to 5.36 3.88 X 10'2 mg/g in specimens near the upper altitudinal limit of this variety, with good concentration/altitude linear correlations [14]. Mean temperatures were 23.3 °C in the lower terrain and 15.6 °C in the upper confines of the transect. However, when the same plants were re-tested after re-growth following wildfires in the area in 1998, values went up to 3.5 mg/g and the altitudinal correlation was diffuse. Therefore, the production and accumulation of ptaquiloside in bracken is still unclear

726 and may respond to still unveiled factors that may be amenable to the rules of the carbon/nutrient balance hypothesis of Coley, Bryant and Chapin [274]. This possibility still remains unexplored in ferns. The ecodynamics of the other known bracken illudanes, quercetin and rutin is completely unknown at this point and should be examined in detail if a holistic picture of bracken toxic and carcinogenic potential is to be ascertained. Ptaquiloside, milk and human health Nowadays, it is well recognized that bracken ptaquiloside finds its way into cows milk [114], a staple in every country. We examined the passage of this compound through the body of cattle from mouth to milk excretion [275] using a set of 6 cows shortly after calving and feeding them daily with well known amounts of ptaquiloside contained in 6 kg of freshly cut fronds of the caudatum neotropical variety. Feeding progressed for five consecutive days and was discontinued thereafter. It took 36 h for ptaquiloside to first appear in milk in detectable quantity, Fig. (14), probably as a result of the time required for digestion of the rumen, absorption, and transport to the milk fluid. The concentration of the carcinogen in milk continued to raise to become stable after 96 h, and began to recede only 48 h after the last bracken feed to finally disappear 38 h thereafter. Clearly, exposure of a cow to bracken for relatively short periods of time will lead to an extended excretion of ptaquiloside in milk. This was tested in a two pulse bracken-feeding experiment, of four and two days with a 72 h pause inbetween. Not unexpectedly, a two pulse excretion curve was obtained, showing that cows in the field exposed to feeding irregularly on Pteridium, as would be the natural situation, will yield pulses of ptaquiloside of varying concentration for extended periods, even when no direct feeding occurs the day of milking. Quantification of these results led to assess the actual amount of carcinogen to which humans may be exposed [276]. Between 8 and 11% of the ingested ptaquiloside may eventually reach milk. It must be said that cattle generally do not favor bracken over commercial and many natural grasses, but they will eat it in times of drought when bracken fronds remain green in the tropics but other fodder pastures are shriveled. Alternative likely scenarios are those places where natural or cultivated grasslands are overgrazed, or when animals stray into dense fern thickets that exclude other edible plants. To translate this situation into manageable figures within a realistic thread, a number of assumptions are necessary. First we presuppose that bracken fronds will harbor a modest average of 0.130 mg of ptaquiloside per fresh gram of tissue, as this is the average we have detected in many

727

caudatum samples. We also assume that, in times of scarcity, cattle may eat 10% of their daily intake as bracken. For a 300 kg cow who eats about 8% of its weight in times of dearth, this means 21.6 kg fodder and 2.4 kg bracken that becomes 312 mg ptaquiloside/day. Let us assume that this takes place 4 days/week. Around 8% of this will go to the total milk production of this cow at the peak of the bell shaped curve of Fig. (14). In the dry season, milk production may be reduced to 25-30% of the normal level due to limited water availability. Thus, a 20 1/day cow may become a 6 1/day animal. Of course ptaquiloside may become more concentrated but we do not know this. With these assumptions at hand, it is easy to calculate a final concentration of as much as 2.38 mg ptaq/1 of this milk or around 0.7 mg per glass [277].

bracken feeding period

o o ^

25-

^

20 -

^ 2 £

15 -

o

M^

30.

10 -

5-

•a oJ

t

w—\

()

24

1

1

48

72

1

1

\

1

1

1

V

1

96 120 144 168 192 216

HOURS AFTER FIRST FEEDING BRACKEN

Fig. (14). Variation in the ratio of ptaquiloside excreted through milk [ptaq]g daily and the total amount of this compound ingested by cows [ptaq]j. Data from Alonso-Amelot et al., 1998.

728 From society's point of view, to have such a quantity of a potent carcinogen in a basic human staple is unacceptable, especially considering the existing correlations between bracken abundance, BEH and gastric cancer in people, as discussed above. From bracken's vantage point, however, the picture is the reverse: ptaquiloside is a exquisitely designed pro-toxin that is delivered to vertebrate herbivores not only mortally damaging them and in the process controlling the population of a plant predator, but also passing the toxin along to the offspring across two trophic levels by way of milk in enough quantity -depending on the appetite of the mother for the now victimized bracken- to cause its demise too. Obviously, further systematic studies will be necessary before a clearer picture emerges. But all the evidence points to bracken as an undesirable plant to share with our managed ecosystems, natural, agricultural or otherwise. The task of controlling bracken excessive growth with herbicides or hand cutting is a costly operation and the content of illudanes and other toxins in specific bracken populations, growing under particular environmental conditions of grazelands and farms, that may propitiate of disfavor the production of these compounds, is a crucial element of judgment at the time of investing on the control of this uttermost dominant element of the plant world [284]. ACKNOWLEDGEMENTS The author gratefully acknowledges grants from the Consejo Nacional de Investigaciones Cientificas y Tecnologicas, CONICIT, (#PC064,#S197001302) and from the Consejo de DesarroUo Cientifico, Humanistico y Tecnologico, CDCHT of Universidad de Los Andes for continued support of our tropical bracken research. Also we are indebted to personal information provided by profs. Saveria Campo, Rinaldo Dos Santos, Uvidelio Castillo, Elida Arellano, Alberto Oliveros, and Arungundrum Prakash, on their latest research developments.

REFERENCES [1]

[2] [3] [4]

Page, C. N.; In: Bracken: Ecology, Land Use and Control Technology. R. T. Smith, R. T.; Taylor, J. A.; Eds.; Parthenon Press. Camforth UK. 1986; pp. 173181. Biggin, P.; Proc. Royal Soc. Edinburg, 1982, 81B, 19-27. Storey, I. N. J.; Fern Gazz., 1991,14(2), 51-53. Tolhurst, K. K.; Tumey, N. D.; Forest Ecol Manag., 1992, 54, 45-67.

729 [5] [6] [7] [8] [9] [10] [11] [12] [13]

[14] [15]

[16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]

Brown R. W. (1986);. In Bracken, Ecology, Land Use and Control Technology. R. T. Smith, R. T.; Taylor, J. A. Eds.; Parthenon Press. Camforth UK. 1986; pp 77-86. Packeman, R. J.; Marrs, R. H.; Biol Cons., 1992, 62, 101-114. Snow, C. S. R.; Marrs R. H.; Biol Cons., 1997, 81, 35-42. Nava-R.; Fernandez, L. E.; Del Amo-R., S.; Agric. Ecosyst. Environ., 1987, 18, 357-379. Evans, I. A.; In Bracken: Ecology, Land Use and Control Technology. R. T. Smith, R. T.; Taylor, J. A. Eds.; Parthenon Press. Camforth UK. 1986; pp. 139146. Tryon Jr., R. M.; Rhodora, 1941, 43(505), 3-31. Conway, E.; 1 Ecol, 1957, 45, 273-284. Conway, E.; Stephens, R.; J. Ecol, 1957, 45, 389-399. Dyer, A. F.; In Bracken Biology and Management. Thomson, J. A.; Smith, R. T.; Eds. AIAS occasional publication No 40. The Austrahan Institute of Agricultural Science, Sydney, Australia. 1989; pp. 35-42. Alonso-Amelot, M. E. et al; Unpublished observations. Caulton, E.; Keddie S.; Dyer A. F.; In Bracken: An Environmental Issue. Smith, R. T.; Taylor, J. A.; Eds. International Bracken Group Special Publication. 1995, (2), 82-89. Caulton E.; Keddie S.; Carmichael R.; In Abstracts, IVth International Bracken Conference. University of Manchester. UK. July 1999. Daniels, R. E.; Weed Res., 1985, 25: 381-388. Alonso-Amelot, M. E.; Rodulfo S.; Vegetatio (Plant Ecology), 1996, 125(2): 137147(1996). Fletcher, W.; Kirkwood R.; In: The Experimental Biology of Ferns. Dyer A., Ed. Academic Press. London. 1979; pp. 591-636. Pitman, J. J.; Abstracts, IVth International Bracken Conference. University of Manchester. UK. July 1999. Calcagno, M. P.; Alonso-Amelot, M. E.; Abstracts, IVth International Bracken Conference. University of Manchester. UK. July 1999. Gliessman, S. R.; Amer. Fern J., 1978, 66(2), 41-44. Sheffield, E.; Wolf, P. G.; Haufler, C. H.; Weed Res., 1989, 29, 455-460. Parks, J. C; Werth, C R.; Am. J. Bot., 1993, 80, 537-544. Watt,A.S.;7VewP/zy/o/., 1970, (5P,431-449. Watt, A .S.; Bot. J. Linn. Soc, 1976, 73, 217-239. Jones, C. G. In Variable Plants and Herbovires in Natural and Managed Systems. Denno, R. F.; McLure, M. S.; Eds.; Academic Press, New York. 1983; chpt 15, pp. 513-558. Fenwick, G. R; J. ScL Food Agric, 1988, 46(2), 147-173. Thomson, J. A.; Abstracts, IVth International Bracken Conference. University of Manchester. UK. July 1999. Thomson, J. A.; Ann. Bot,. 2000, in press.

730

[31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45]

[46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57]

Rhoades, D. F. In: Herbivores, their Interaction with Secondary Plant Metabolites. Rosenthal, G. A.; Janzen D. H.; Eds.; Academic Press, New York. 1979, chptl,pp 3-54. Berenbaum, M. R.; Proc. Natl Acad. ScL USA, 1995, 92, 2-8. Holm, L. G.; Plucknett, D. L.; Pancho, J. V.; Herberger, J. P.; The World's Worst weeds. Distribution and Biology. The University Press of Hawaii. 1977. Penberthy, J.; J. Comp. Pathol, 1893, 5, 266-275. Storrar, D. N.; J. Comp. Pathol, 1893, (5, 276-279. Hirono, I.; Yamada, K.; In Naturally Occurring Carcinogens of Plant Origin. Hirono, I., Ed.; Kodansha-Elsevier, Tokyo-Amsterdam, 1987; chpt4, pp 87-120. Ojika, M.; Sugimoto, K.; Okazaki, T.; Yamada, K. J. Chem. Soc. Chem. Commun., 1989, (22), MlSAlll. Nagao, T.; Saito, K.; Hirayama, E.; Uchikoshi, K.; Koyama, K.; Natori, S.; Morisaki, N.; Iwasaki, S.; Matsushima, T.; Mutation Res., 1989, 275,173-178. Kushida, T.; Uesugi, M.; Sugiura, Y.; Kigoshi, H.; Tanaka, H.; Hirokawa, J.; Ojika, M.; Yamada, K.; J. Am. Chem. Soc, 1994,116, 479-486. Smith B. L., Shaw G., Prakash A. S., Seawright A. A.; In Plant Associated Toxins. Colgate, S. M.; Dorling P. R.; Eds. CAB International, Wallingford UK. 1994;chpt31, 167-172. Prakash, A. S.; Pereira, T. N.; Smith, B. L.; Shaw, G.; Seawright, A. A.; Nat. Toxins, 1996,^,221-227. Shahin, M.; Smith, B. L.; Worral, S.; Moore, M. R.; Seawright, A. A.; Prakash, A. S.; Biochem. Biophys. Res. Commun., 1998, 244(1), 192-197. Evans, I. A.; Bot. J. Linn. Soc, 1976, 73, 105-112. Evans, I. A.; Res. Vet. Sci., 1979, 26, 339-348. Pamukcu, A. M.; Bryan, G. T.; In Naturally-Occurring Carcinogens, Mutagens and Modulators of Carcinogenesis. Miller, E. C ; Miller, J. A.; Hirono, I.; Sugimura, T. Takayama, S.; Eds.; Japan Scientific Societies Press, Tokyo. 1979; pp. 89-99. Hirono, I.; CRC Critical Rev. Toxicol, 1986, 17, 1-22. Cooper-Driver G. A.; Annals Miss. Bot. Card, 1990, 77, 281-286. Smith, B. L. In Bracken Biology and Management. Thomson, J. A.; Smith, R. T. Eds.; AIAS occasional pubhcation No 40. The Australian Institute of Agricultural Science, Sydney, Australia. 1989, pp 227-232. Smith, B. L.; Seawright, A. A.; Nat. Toxins, 1995, 3, 1-5. Shahin, M.; Smith, B. L.; Prakash, A. S.; Mutation Res., 1999, 443(1-2), 69-79. Yasuda, Y.; Kihara, T.; Nishimura, H.; Toxicol Appl Pharmacol, 1974, 28, 264268. Rosenberger, G.; Vet. Med Rev., 1971, 2/3, 189-206. Martincic, M.; Veterinaria (Sarajevo), 1955, 4, 525-620. Pamukcu, A. M.; Ann. N. Y. Acad Sci, 1963,108, 938-947. Smith, B. L.; van der Wouden, M.; A^. Z Vet. J., 1972,20, 56. McKenzie, R. A.; Austral Vet. J., 1978, 54, 61-64. Xu, L. R.; Res. Vet. Sci., 1992, 53, 116-121.

731

[58] [59] [60] [61] [62] [63] [64] [65] [66] [67]

[68]

[69] [70] [71]

[72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82]

Tripathi, S. C; Joshi, H. C; Kumar, M.; Indian J. Animal Sci., 1989, 59, 835837. Somvanshi, R.; Bisht, S. P. S.; Prasad, M. C; Indian Vet. Med J., 1997, 27, 159161. Dawra, R. K.; Sharma, O. P.; Somvanshi, R.; Abstracts, IVth International Bracken Conference, University of Manchester. UK July 1999. De Jongh, F.; PROGAL Hoja Divulgativa, 1978, 3(4), 1-4. Villafane, A. F.; Lichtenberger, E.; Rev. Acovez, 1979, 3, 9-13. Pena, N. E. B.; Villafane, F. A.; Dorado, L. E. V.; Urquijo, G. A. V.; Rev. LC.A., Bogota, 1983, 18, 335-343. Villalobos-Salazar J.; Rev. Cost. Cienc. Med, 1985, 6, 131-139. Dobereiner, J.; Tokamia, C. H.; Canella, C. F. C; Pesq. Agropec. Bras., 1967, 2, 489-504. Brobst, D. F.; Olson, C; J. Am. Vet. Res., 1963,24, 105-111. Pinto, C; Januario, T.; Geraldes, M.; Medeiros, F.; Oliveu-a, V.; Ponte, J.; Louza, A. C; Abstracts, IVth International Bracken Conference, University of Manchester. UK. July 1999. Bento, M. C; Clemente, M. L.; Costa, M. M.; da Paz Gouveia, M.; Spinola, T. M.; Abstracts, IVth International Bracken Conference, University of Manchester. UK, July 1999. Campo, M. S.; Jarrett, W. F. H.; Barron, R.; O'Neil B. W.; Smith, K. T.; Cancer Res., 1992, 52, 6898-6904. Evans, W. C; Patel, M. C; Koohy, Y.; Proc. Roy. Soc. Edinburg, 1982, 8, 6578. Villalobos-Salazar, J.; Meneses, A.; Rojas, J. L.; Mora, J.; Porras, R. E.; Herrero, M. v.; In Bracken Toxicity and Carcinogenicity as Related to Animal and Human Health. Taylor J. Ed.; International Bracken Group Special Publication. Univ. College North Wales, UK. 1989; pp 40-51. Pamukcu, A. M.; Goskoy, S. K.; Price, J. M.; Cancer Res., 1969, 27, 917-924. Price, J. M.; Pamukcu, A. M.;. Cancer Res., 1968, 28, 2247-2251. Evans, I. A.; Mason, J. Nature, 1965, 208, 913-914. Evans, I. A.; Cancer Res., 1968, 28, 2252-2261. Pamukcu, A. M.; Price, J. M.; J. Natl. Cancer Inst, 1969, 43, 275-281. Hirono, I.; Shibuya, C; Fushimi, K.; Haga, M.; J. Natl. Cancer Inst., 1970, 45, 179-188. Ushijima, J.; Matsukawa, K.; Yuasa, A.; Okada, M.; Jpn. J. Vet. Sci., 1983, 45, 593-602. Hirono, L; Also, S.; Yamaji, T.; Niwa, H.; Ojika, M.; Wakamatsu, K.; Yamada, K.; Cancer Lett, 19S4, 22, 152-155. Santos, R. C, Brasileiro-Filho, G.; Hojo, E. S; Brazilian J. Med Bio. Res., 1987, 20, 73-77. Jarrett, W. F. H.; McNeil, P. E.; Grimshaw, W. T. R.; Sehnan, I. E.; Mclntyre, W. I. M.; Nature, 1978,274, 215-217. Smith, B. L.; Prakash, A. S.; Lauren, D. R.; Abstracts, IVth International Bracken Conference, University of Manchester, UK, July 1999.

732

[83] [84] [85] [86] [87] [88] [89] [90] [91] [92]

[93] [94] [95] [96] [97] [98] [99]

[100] [101] [102] [103] [104]

Olson, C; Pamukcu, A. M.; Brobst, D. F.; Kowalczyk, T.; Satter, E. J.; Price, J. M.; Cancer Res., 1959, 7P, 779-783. Olson, C; Pamukcu, A. M.; Brobst, D. F.; Cancer Res., 1965, 25, 840-849. Hirono, I.; Fushimi, K.; Mori, H.; Niwa, T.; Haga, M.; J. Natl. Cancer Inst, 1973,50, 1367-1371. Evans, W. C; Evans, I. A.; Axford, F. R. E.; Threlfall, G.; Humphreys, D. A; Thomas, A. J.; Vet. Rec, 1961, 73, 852-853. Seaward, M. R. D.; Bot. J. Linn. Soc, 1976, 75, 177-185. Evans I. A.; Rev. Environ. Health, 1987, 7, 161-199. Povey A. C; Potter, D.; O'Connor, P. J.; Br. J. Cancer, 1996, 74, 1342-1348. Saito, K.; Nagao, T.; Matoba, M.; Koyama, K.; Natori, S.; Murakami, T.; Saiki, Y.; Phytochemistry, 1989,25, 1605-1611. Jones R. In, Quaternary Studies, Suggate, R. P.; Creswell M. M.; Eds., Royal Society of New Zealand, Wellington. 1975; pp. 21-34. Bowdler, S.; Lourandos, H.; In Coastal Archaeology in Easter Australia. Bowdler S. (ed.). Proceedings of the 1980 Valla Conference on Australian Prehistory. Department of Prehistory, RSPacS. Australian national University, Canberra. 1982; pp. 121-32. Humboldt, A. von, Reise in die Aquinoktial Gegenden des neuen Kontinents. Translation of L. Alvarado. Monte Avila Eds. Caracas, 2nd edition, 1859-60; p. 143. Rymer, L.; Bot. J. Linn. Soc, 1976, 73, 151-176. Kamon, S.; Hirayama, T.; Proc. Jpn. Cancer Assoc, 1975, 34, Hirayama, T.; Nutrit. Cancer, 1979, 1, 67-81. Hu-ono, I.; Shibuya, C; Shimizu, M.; Fushimi, K.; J. Natl. Cancer Inst. 1972, 48, 1245-1250. Dos Santos, R. C; Brasileiro Filho, G.; Hojo, E. S.; Braz. J. Med. Biol. Res., 1987, 20: 73-77. Marliere, C. A.; Santos, R. C; Galvao, M. A. M.; Silva, M. L. C; Kawamoto, M.; Castro, M. C. F. M.; Soares, J. F.; Von Kruger, E. R.; Barretto, J. M.A.; Gomes, R. Q. F.; In Bracken: An Environmental Issue Smith, R. J.; Taylor, J. A.; Eds., University of Leeds, UK. 1995; pp. 99-101. Marliere, C. A.; Santos, R. C; Galvao, M. A. M.; Soares, J. F.; Evangelista, L. M.; Gomes, R. Q. F.; Rev. Brasil. Cancerol, 1998, 3, 225-229. Marliere, C. A.; Galvao, M. A. M.; Wathem, P.; Freitas, S. N.; Abstracts, International Bracken Conference, University of Manchester. UK, July 1999. Dos Santos, R. C; Silva, M. E.; Brasileiro Filho, G.; Recouso, R. C; Lindsey, C. J.; Freitas, R. N.; Marliere, C. A; Galvao, M. A. M.; Abstracts, IVth International Bracken Conference, University of Manchester. UK, July 1999. Santos, R. C; Brasileiro-Filho, G; Silva, M. E.; Ciens. Tecnol. Aliment., 1992, 12, 72-76. Santos R. C, Brasileiro-Filho G.; Silva M. E.; In Bracken Biology and Management. Thomson, J. A.; Smith, R. T.; Eds.; AIAS occasional publication No 40. The Australian Institute of Agricultural Science, Sydney, Australia. 1989; pp. 253-257.

733

[105] Buckley, E; In Bracken Toxicity and Carcinogenicity as Related to Animal and Human Health. Taylor, J. A., Ed.; Special Publication of the International Bracken Group based on Bangor Meeting, September 1988. Geography Dept. UCW, Aberystwyth, Wales U.K. 1989; pp. 58-66. 106] Galpin, O. P.; Whitaker, C. J.; Whitaker, R.; Kassab, J. Y.; Br. 1 Cancer, 1990, 61, 737-740. 107 Alonso-Amelot, M. E.; Avendano, M.; to be published elsewhere. 108 Wilson, D.; Donaldson, L .J.; Sepai, O.; J. Epidemiol. Community. Health, 1998, 57(^72;, 812-817. 109 Galpin, O. P.; Smith, R. M. M.; In Bracken, Ecology, Land Use and Control Technology. Smith, R. T.; Taylor, J. A.; Eds.; Parthenon Publishing, Camforth UK, 1986; pp. 149-157. 110 Pamukcu, A. M.; Olson, C; Price, J. M.; Cancer Res., 1966, 26, 1745-1753. 111 Evans, I. A.; Jones, R. S.; Mainwaring-Burton, R.; Nature,1972, 237, 107-108. 112 Pamukcu, A. M.; Erturk, E.; Yalciner, S.; Milli, U.; Bryant, G. T.; Cancer Res., 1978,38, 1556-1560. 113 Villalobos-Salazar, J.; Meneses, A.; Salas, J.; In. Bracken Biology and Management. Thomson, J. A.; Smith, R. T.; Eds.; AIAS occasional publication No 40. The Australian Institute of Agricultural Science, Sydney, Australia. 1989; pp 247-251. 114 Alonso-Amelot, M. E.; Castillo, U.; De Jongh, F.; Lait, 1993, 73, 323-332. 115 Brown, R.; Abstracts, Ivth International Bracken Conference. University of Manchester, UK, July, 1999. 116 Evans, I. A.; Osmann, M. A.; Nature, 1974, 250, 348-349. 117 Bohm, B. A.; Tryon, R. M.; Can. J. Bot.,1961, 45, 585-593. 118 Glass, A. D.; Bohm, B. A.; Phytochemistry, 1969, 8, 371-377. 119 San Francisco, M.; Cooper-Driver, G. A.; Amer. Fern J., 1984, 74, 87-96. 120 Wang, C. v.; Chiu, C. W.; Pamukcu, A. M.; Bryan, C. T.; J. Natl. Cancer Inst., 1976, 55,33-36. 121 Elliger, C. A.; Wong, Y.; Chan, B. G.; Waiss, A. C, Jr.; J. Chem. EcoL, 1981, 5, 753-758. 122 Felton, G. W.; Donato, K.; Del Vecchio, R. J.; Duffey, S.S.; J. Chem. EcoL, 1989, 15, 2667-2694. 123 Stamp, N. E.; Temple, M. P.; Traugott, M.S.; Wilkens, R. T.; Entomol. Exp. Appl, 1994, 73, 199-210. 124 Alonso-Amelot, M. E.; Rodulfo, S.; Jaimes-Espinoza, R.; Biochem. Syst. EcoL, 1995,25, 709-716. 125 Gliessman, S. R.; MuUer, C. H.; Madrono, 1972,27, 299-304. 126 Gliessman, S. R.; Muller, C. H.; J. Chem. EcoL, 1978, 4, 337-362. 127 Stewart, R. E.; J. Chem. EcoL, 1975, 7, 161-169. 128 Taylor, J. E.; Thomson, J. A; In Bracken Biology and Management. Thomson, J. A.; Smith, R. T.; Eds.; AIAS occasional publication No 40. The Australian Institute of Agricultural Science, Sydney, Australia. 1989; pp. 203-208. 129; Gliessman, S. R.; Bot. J. Linn. Soc, 1976, 73, 95-104. 130 Glass, A. D. M.; Can. J. Bot., 1976, 54, 2440-2444.

734

[131] Whitehead, D. C; A^fl/wre, 1964, 202,417-418. [132] Tanaka, N.; Yuhara, H.; Wada, H.; Murakami, T.; Cambie, R. C; Braggins, J. E.\ Phytochemistry, 1993,52, 1037-1039. [133] Ojika, M.; Kigoshi, H.; Kuyama, H.; Niwa, H.; Yamada, K.;. J. Nat. Prods,, 1985, 48, 634-637. [134] Ojika, M.; Wakamatsu, K.; Niwa, H.; Yamada, K.; Tetrahedron, 1987, 43, 52755280. [135] Wollenweber, E.; Biochem. Syst. Ecol, 1989, 77, 141-144. [136] Swain, T.; Cooper-Driver, G. A.; Bot. J. Linn. Soc.„ 1973, Suppl. 1, 111-134. [137] Wang, C. Y.; Pamukcu, A. M.; Bryan, G. T.; Phytochemistry, 1973, 72, 22982299. [138] Imperato, F.; Phytochemistry, 1995, 40, 1801-1802. [139] Imperato, F.; Phytochemistry, 1996, 43, 1421-1423. [140] Imperato, F.; Minutiello, P.; Phytochemistry, 1997, 45, 199-200. [141] Pamukcu, A. M.; Wang, C. Y.; Hatcher, J. F.; Bryan, G. T.;. J. Natl Cancer Inst., 1980,(55, 131-136. [142] Morino, K.; Mtsukura, N.; Kawachi, T.; Ohgaki, H.; Sugimura, T.; Hirono, I.; Carcinogenesis, 1982, 3, 93-97. [143] Plakas, S. M.; Lee, T. C; Wolke, R. E.; Fed Chem. Toxicol, 1985, 23(12), 10711080. [144] Scambia, G.; Ranelletti, F. O.; Benedeti-Panici, P.; Piantelli, M.; Bonanno, G.; De Vincenzo, R.; Ferrandina, G.; Pierelli, L.; Capelli, A.; Mancuso, S.; Cancer Chemother. Pharmacol, 1991, 28, 255-258. [145] Piantelli, M.; Rinelli, A.; Macri, E.; Maggiano, N.; Larocca, L. M.; Scerrati, M.; Roselli, R.; lacoangeH, M.; Scambia, G.; CapeUi, A.; Ranelletti, F. O.; Cancer, 1993, 71, 193-198. [146] Bjeldanes, L. F.; Chang, G. W.; Science, 1977,197, 577-578. [147] Maruta, A.; Enaka, K.; Umeda, M.; Gann, 1979, 70, 273-276. [148] Stoewsand, G. S.; Anderson, J. L.; Boyd, J.N.; Hrazdina, G.; Babish, J. G.; Walsh, K. M.; Losco, P.; J. Toxicol Environ. Health, 1984,14, 105-114. [149] Nakayasu, M.; Sakamoto, H.; Terada, M.; Nagao, M.; Sugimura, T.; . Mutation Res., 1986,774,79-83. [150] Ishikawa, M.; Okada, F.; Hamada, J.; Hosokawa, M.; Kobayashi, A.; Int. J. Cancer., 1987, 39, 338-342. [151] Fazal, F.; Rahman, A.; Greesill, J.; Ainley, K.; Hadi, S. M.; Parish, J. H.; Carcinogenesis, 1990, 77,2005-2008. [152] Rhaman, A.; Shahabuddin, Hadi, S. M.; Parish, J. H.;. Carcinogenesis, 1990, 77, 2001-2003. [153] Suzuki, S.; Takada, T.; Sugawara, Y.; Muto, T.; Kominami, R.; Jpn. J. Cancer Res., 1991,82, 1061-1064. [154] Ishidate, M.; Data Book on Chromosomal Aberration Tests in Vitro. LICElsevier. Tokyo-Amsterdam. 1988. [155] Campo, M. S.; CNeil, B. W.; Barron, R. J.; Jarrett, W. F. H.; Carcinogenesis, 1994,75, 1597-1601. [156] Campo, M. S.; Vet 1, 1997,154, 175-188.

735

[157] Connolly, J. A.; Morgan, I. M.; Jackson, M. E.; Campo, M. A. S.; Oncogene, 1998, 15, 2739-2746. [158] Campo M. S., Beniston R. G., Connolly J. A.; Grindlay G. J. Abstracts, IVth International Bracken Conference. University of Manchester. UK. July 1999. [159] Beniston, R. G.; Morgan, I. M.; O'Brien, V.; Campo, M. S.; Abstracts, IVth International Bracken Conference. University of Manchester. UK. July 1999. [160] Beniston, R. G.; Morgan, I. M.; O'Brien, V.; Campo, M. S.; to be published elsewhere. [161] Felton, G. W.; Duffey, S. S.; Vail, P. V.; Kaya, H. K.; Manning, J.; 1 Chem. Ecol, 1987, 13, 947-957. [162] Ali, M. I.; Bi, J. L.; Young, S. Y.; Felton, G. W.; J. Chem. EcoL, 1999, 25(9), 2193-2204. [163] Chipman, J. K.; Musonda, C. M.; Lautraite, S.; Abstracts, IVth International Bracken Conference. University of Manchester. UK. July 1999. [164] Voirin, B.; Recherches chimiques, taxonomiques et physiologiques sur les flavonoides des pteridophites. These Doctoral. L'Universite de Lyon. France. 1970. [165] Cooper-Driver, G. A.; Finch, S.; Swain, T.; Bemais, E.A.; Biochem. Syst. EcoL, 1977,5, 177-183. [166] Tempel, A. S.; Oecologia (Berlin) 1981, 51, 97-106. [167] Feeney, P.; Rec. Adv. Phytochem., 1976, 10, 1-40. [168] Appel, H.; J. Chem. Ecol, 1993, 7, 1521-1525. [169] Waterman, P. G.; Mole, S.; Analysis of Phenolic Plant Metabolites. Blackwell Scientific Publications. London. 1994. [170] Bemays, E. A.; Cooper-Driver, G. A.; Bilgener, M.; Adv. Evol. Res., 1989, 19, 1349-1358. [171] Balick, M. J.; Furth, D. G.; Cooper-Driver, G. A;. Oecologia (Berlin) 1978, 35, 55-89. [172] Auerbach, M. J.; Hendrix, S. D.; Ecol. Entomol. 1980, 5, 99-104. [173] Hendrix, S. D.; Amer. Nat., 1980, 775, 171-196. [174] Lawton, J. H.; Rashbrook, V. K.; Compton, S. G; Ann. Appl. Biol, 1988, 772, 479-490. [175] Jones, D. A.; TVa/wre, 1962, 7P5, 1109-1110. [176] Butler, G. W.; Phytochemistry, 1965, 4, 127-131. [177] Jones, D. A.; In Taxonomy and Ecology. Heywood, V.; Ed.; Academic Press, London. 1973, pp 213-242. [178] Ellis W. M.; Keymer, R. J.; Jones, D. A.; Experientia, 1977, 33, 309-310. [179] Bemays E. A.; Internatl. Org. Biol. Control.; Western Palearctic Regional Section Bull, 1977, 3, 123-128. [180] Davis, R. H.; Nahrstedt, A.; Comp. Biochem. Physiol, 1979, 64B, 395-397. [181] Davis, R. H.; Nahrstedt, A.; In. Comprehensive Insect Physiology, Biochemistry and Pharmacology. Verhut, G. A.; Gilbert, L. L; Eds.; Pergamon Press, New York. 1984; Fo/77, 198-221. [182] Conn, E. E.; In The Biochemistry of Plants, Conn E. E.; Ed.; Academic Press, New York. 1981; Vol 7, pp 479-500.

736

[183] Harbome, J. B.; Introduction to Ecological Biochemistry. 2nd ed. Academic Press, New York. 1982. [184] Mcllroy, R. J.; The Plant Glycosides. Arnold, London. 1951. [185] Bell, A. A.; In: Proceedings of the Summer Institute on Biological Control of Plant Insects and Diseases. Maxwell, F. G.; Harris, F. A.; Eds.; University Press of Mississippi, Jackson, Missouri. 1974; pp 407-462. [186] Christensen, H. E;. Registry of Toxic Effects of Chemical Substances. Natl. Inst. Occupational Safety Health, US Public Health Service, Maryland. Publ No NIOSH. 1976; pp 76-191. [187] Scriber, M. J.; Oecologia (Berlin), 1978, 34, 143-155. [188] Schreiner, I.; Nafus, D.; Pimentel, D.; Ecol Entomol., 1984, P, 69-79. [189] Cooper-Driver, G. A.; Swain, T.; Nature, 1976,260, 604. [190] Jones, C. G.; Fim, R. D.; Biochem. Syst. Ecol, 1979, 7, 187-192. [191] Kofod, H.; Eyjolfsson, R.; Tetrahedron Lett., 1966, 12, 1289-1291. [192] Berti, G.; Bottari, F.; Prog Phytochem., 1968, 7, 589-685. [193] Hadfield, P. R. H.; Dyer, A. F.; Biochem. Syst. Ecol., 1988,16, 9-13. [194] Conn, E. E.; Internatl. Dev. Res. Cent. Monographs IDRC, 1973, IDRC-010-e, 55-63. [195] Berenbaum, M. R.; Amer. Nat., 1979, 115, 138-146. [196] Moon, F. E.; Raafat, M. A.; 1 Sci. FoodAgric, 1951, 2, 327-336. [197] Jones, C. G.; ''Chemical content and insect resistance of bracken fern [Pteridium aquilinum (L) KuhnJ". Doctoral Thesis, University of York. England. 1977. [198] Low, V. H. K.; Thomson, J. A.; Bracken Biology and Management. In Thomson, J. A.; Smith, R. T.; Eds.; AIAS occasional publication No 40. The Australian Institute of Agricultural Science, Sydney, Australia. 1989; pp. 105-111. [199] Kaplanis, J. N.; Thompson, M. J.; Robbins, W. E.; Bryce, C. N.; Science, 1967, 157, 1436-1438. [200] Takemoto, T.; Arihara, S.; Hikino, H.; Tetrahedron Lett., 1968, 4199-4202. [201] Takemoto, T.; Okuyama, T.; Jin, H.; Aral, T.; Kawahara, M.; Konno, C; Nabetani, S.; Arihara, S.; Hikino, Y.; Hikino, H.; Chem. Pharm. Bull., 1973, 21, 2336-2338. [202] Jones, C. G.; Fim, R. D.; J. Chem. Ecol., 1978, 4, 117-138. [203] Carlisle, D. B.; Ellis, P. E.; Science, 1968,159, 1472-1474. [204] El-Mofty, M.; Sadek, I.; Soliman, A.; Mohamed, A.; Sakre S.; Nutr. Cancer, 1987, P, 103-107. [205] Tanaka, N.; Satake, T.; Takahashi, A.; Mochizuki, M.; Murakami, T.; Saiki, Y.; Yang, J. Z.; Cheng, C. M.; Chem. Pharm. Bull., 1982, 30, 3640-3646. [206] Murakami, T.; Saiki, Y.; Biochem. Syst. Ecol., 1989,17, 131-140. [207] Yoshihira, K.; Fukuoka, M.; Kuroyanagi, M.; Natori, S.; Chem. Pharm. Bull., 1971,19, 1491-1495. [208] Yoshihira, K.; Fukuoka, M.; Kuroyanagi, M.; Natori, S.; Umeda, M.; Morohoshi, T.; Enomoto, M.; Saito, M.; Chem. Pharm. Bull. 1978, 26, 2346-2364. [209] Fukuoka, M.; Kuroyanagi, M.; Toyama, M.; Yoshihira, K.; Natori, S.; Chem. Pharm. Bull., 1972, 20, 2282-2285.

737

[210] Fukuoka, M.; Kuroyanagi, M.; Yoshihira, K.; Natori, S.; Chem. Pharm. Bull, 1978, 2(5, 2365-2385. [211] Kuroyanagi, M.; Fukuoka, M.; Yoshihira, K.; Natori, S.;. Chem. Pharm. Bull, 1979, 27,592-601. [212] Castillo, U. P.; M. Sc. Thesis. Universidad de Los Andes. 1997, pp 145-149. [213] Kobayashi, A; Koshimazu, K.; Agric. Biol. Chem., 1980, 44, 393-398. [214] Saito, M.; Umeda, M.; Enomoto, M.; Hatanaka, Y.; Natori, S.; Yoshihira, K.; Fukuoka, M.; Kuroyanagi, M; Experientia, 1975, 37, 829-831. [215] Evans, W. C; Kom, T.; Natori, S.; Yoshihira, K.; Fukuoka, M.; J. Pharmacobiodyn., 1983, 6, 938-940. [216] Jones, C. G.; Fim, R. D.; Biochem. Syst. Ecol., 1979, 7, 95-101. [217] McMorris, T. C; Liu, M.; White, R. H.; J. Nat. Prods., 1976, 40, 221-224. [218] Alonso-Amelot, M. E.; Perez-Mena, M.; Calcagno, M. P.; Jaimes-Espinoza, R.; U. Castillo; J. Chem. Ecol., 1992, 18, 1405-1420. [219] Alonso-Amelot, M. E.; Perez-Mena, M.; Calcagno, M. P.; Jaimes-Espinoza, R.; Phytochem. Anal, 1992, 3, 160-164. [220] Castillo, U. F.; Ojika, M.; Alonso-Amelot, M. E.; Sakagami, Y.; Tetrahedron 1999, 55, 12295-12300. [221] Amone, A.; Bambilla, U.; Nasini, G.; Vajna de Pava, O.; Tetrahedron, 1995, 57, 13357-13364. [222] Castillo, U. F.; Wilkins, A.; Lauren, D. R.; McMaughton, D. E.; Smith, B. L.; J. Nat. Prods., 1995, 58, 1889-1891. [223] Thomas, B.; Walker, H. F.; J. Soc. Chem. Ind, 1949, 68, 6-9. [224] Somogyi,].C.',J.Vitamino.,1971, 77,165-174. [225] Evans, W. C; Bot. J. Linn. Soc, 1976, 73,113-131. [226] Evans, W. C; Evans, I. A.; Humphreys, D. J.; Lewin, B.; Davies, W. E. J.; Axford, R. F. E.; J. Comp. Pathol, 1975, 85, 253-267. [227] Samsuatruamphol, S.; Parsons, H. T.; J. Amer. Diet Assoc, 1955, 31, 790-793. [228] Moon, F. E.; Pal, A. K.; J. Sci Food Agric, 1949, 2, 327-336. [229] Parry, D. W.; Hodson, M. J.; Newman, R. H.; Ann. Bot., 1982, 55, 77-83. [230] Saito, T.; Mochizuki, D.; J. Toxicol ScL, 1986, 77, 15-27. [231] Saito, T.; Takeno, K.; Nakamura, S.; Uehara, M.; Jpn. J. Vet. Scl, 1987, 49, 181183. [232] Saito, T.; Kurata, Y.; Takeno, K.; Jpn. J. Pharmacol, 1990, 53, 165-173. [233] Leach, H.; Barker, G. D.; Evans, I. A.; Evans, W. C; Biochem. J., 1971, 72^, 1314. [234] Evans, I. A.; Al-Samarrai, A. M. H.; Smith, R. M. M.; Res. Vet. Sci., 1984, 37, 26\'265. [235] McMorris, T. C; Liu, M.; White, R. H.; J. Nat. Prods., 1976, 40, 221-224. [236] McMorris, T. C; Anchel, M.; J. Am. Chem. Soc, 1965, 87, 1594-1600. [237] Nakanishi, K.; Tada, M.; Yamada, Y.; Ohashi, M., Komatsu, N.; Terakawa, H.; Nature, 1963, 7P7, 292. [238] McMorris, T. C; Kebier, M. J.; Wang, W.; Diaz, M. A.; Estes, L. A.; Taetle, R.; Experientia, 1996, 52, 75-80.

738 [239] Niwa, H.; Ojika, M.; Wakamatsu, K.; Yamada, K.; Hirono, I.; Matsushita, K.; Tetrahedron left., 1983,24, 4117-4120. [240] Van der Hoeven, J. C. M.; Lagerweij, W. J.; Posthumus, M. A.; van Veldhuizen, A.; Holterman, H. A. J.; Carcinogenesis, 1983, 4, 1587-1590. [241] Ohba, S.; Saito, Y.\Acta Cryst, 1984, C40, 1877-1879. [242] Hirono, I.; Also, S.; Yamaji, T.; Mori, H.; Yamada, K.; Niwa, H.; Ojika, M.; Wakamatsu, K.; Kigoshi, H.; Niiyama, K.; Uosaki Y.; Gann, 1984, 75, 833-836. [243] Hirono, I.; Kono, Y.; Takahashi, K.; Yamada, K.; Niwa, H.; Ojika, M.; Kigoshi, H.; Niiyama, K.; Uosaki, Y.; Vet Rec. 1984, 775, 375-378. [244] Sato, T.; Inaba, H.; Kawai, K.; Furukawa, H.; Hirono, I.; Miyazawa, T.; Mutation Res., 1991,257,91-97. [245] Nagao, T.; Saito, K.; Hirayama, E.; Uchikoshi, K.; Koyama, K.; Natori, S.; Morisaki, N.; Iwasaki, S.; Matsushima, T.; Mutation Res., 1989, 275, 173-178. [246] Matsuoka, A.; Hirosawa, A.; Natori, S.; Iwasaki, S.; Sofuni, T.; Ishidate, M., Jr.; Mutation Res., 1989,275, 179-185. [247] Dunkel, V. C; Williams, G. M.; In Proceedings of the Third Life Sciences Symposium on Health Risk Analysys, Richmond, C. R.; Walsh, P. J.; Copenhauer, E. D.; Eds.; Franklin Press, 1981; pp 249-271. [248] Mori, H.; Sugie, S.; Hirono, I.; Yamada, K.; Niwa, H.; Ojika, M.; Mutation Res., 1989, 143, 75-78. [249] Ojika, M.; Sugimoto, K.; Okazaki, T.; Yamada, K.; J. Chem. Soc. Chem. Commun., 1989, (22): MlS-Mll, [250] Reynolds, V. L.; Molineux, I. J.; Kaplan, D. J.; Swenson, D. H.; Hurley, L. H.; Biochem., 1985, 24, 6228-6237. [251] Saffhill, R.; Marginson, G. P.; O'Connor, P. J.; Biophys. Biochim. Acta, 1985, 823, 111-146. [252] BdiYh?iC\d,M., Carcinogenesis, 1986,7,1037-1042. [253] Kushida, T.; Uesugi, M.; Sugiura, Y.; Kigoshi, H.; Tanaka, H.; Hirokawa, J.; Ojika, M.; Yamada, K.; DNA damage by ptaquiloside, a potent bracken carcmogen. J. Am. Chem. Soc, 1994, 116, 479-486. [254] Lawley, P. D.; In Chemical Carcinogenesis, 2""^ ed. Searle, C. E.; Ed.; ACS Monographs # 182. American Chemical Society, Washington, D. C; 1984, Vol 1. [255] Loeb, L. A.; Cell, 1985, 40, 483-484. [256] Smith B. L., Shaw G., Prakash A. S.; Seawright A. A. In Plant associated Toxins, Colgate, S. M.; Dorling, P. R.; Eds. CAB International, Wallingford UK. 1994, chpt 31,pp. 167-172. [257] Bartsch, H.; Terrancini, B.; Malaveille, C; Tomatis, L.; Wahrendorf, J.; Brun, G.; Dodet, B.; Mutation Res., 1983,10, 181-219. [258] Pearson, R. G.; Songstad, J.; J. Am. Chem. Soc, 1967, 89, 1827-1840. [259] Ho, T-L.; Chem. Rev.,1915, 75, 1-20. [260] Carlson, R. M.; Environ. Health Perspect, 1990, 87, 227-232. [261] Pullman, A.; Pulhnan, B.; Quart. Rev. Biophys., 1981,14(3), 289-380. [262] Campo, M. S.; McCaffery, R. E.; Doherty, I.; Kennedy, I. M.; Jarrett, W. F. H.; Oncogene, 1990, 5, 303-308.

739 [263] Prakash, A. S.; Pereira, T. N.; Smith, B. L.; Shaw, G.; Seawright, A. A.; Nat. Toxins, 1996,^,221-227. [264] Castillo, U. F.; Ojika, M.; Alonso-Amelot, M. E.; Sakagami, Y.; Bioorg. Med Chem., 1998, 6, 2229-2233. [265] Smith, B. L.; Embling, P. P.; Lauren, D. R.; Agnew, M. P.; Ross, A. D.; Greentree, P. L.; Austral. Vet. J., 1989, 66, 154-155. [266] Saito, K.; Nagao, T.; Takatsuki, T.; Koyama, K.; Natori, S.; Phytochemistry, 1990, 29, 1475-1479. [267] Agnew, M. P.; Lauren, D. R.; J. Chromatogr., 1991, 538, 462-468. [268] Kigoshi, H.; Imamura, Y.; Niwa, H.; Yamada, K.; J. Am. Chem. Soc, 1989, III, 2302-2303. [269] Kigoshi, H.; Sawada, A.; Imamura, Y.; Niwa, H.; Yamada, K.; Tetrahedron, 1989,^5,2551-2556. [270] Kigoshi, H.; Sawada, A.; Niwa, H.; Yamada, K.; Bull. Chem. Soc. Jpn., 1989, 62, 1639-1642. [271] Kigoshi, H.; Imamura, Y.; Mizuta, K.; Niwa, H.; Yamada, K.; J. Am. Chem. Soc, 1993, 775, 3056-3065. [272] Yamada, K.; Ojika, M.; Kigoshi, H.; Angew. Chem. Int. Ed, 1998, 57, 18181826. [273] Smith, B. L.; Seawright, A. A.; Ng, J. C ; Hertle, A. T.; Thomson, J. A.; Bostok, P. D.; Nat. Toxins, 1994, 2, 347-353. [274] Coley, P. D.; Bryant, J. P.; Chapin, F. T., Ill; Science, 1985, 230, 895-899. [275] Alonso-Amelot, M. E.; Castillo, U.; Smith, B. L.; Lauren, D. R.; Nature, 1996, 382, 587-588. [276] Alonso-Amelot, M. E.; Castillo, U.; Smith, B. L.; Lauren, D. R.; Lait, 1998, 78, 413-423. [277] Alonso-Amelot, M. E.; Castillo, U. F.; Avendano, M.; Smith, B. L.; Lauren, D. R.; Abstracts, IVth International Bracken Conference. University of Manchester. UK. July 1999. [278] Alonso-Amelot, M. E.; Oliveros, A.; Phytochem. Anal, in press. [279] Burkhalter, P. W.; Groux, P. M. J.; Candrian, U.; Hubner, P.; Liithy, J.; J. Nat. Toxins, 1996, 5, 141-159. [280] Castillo, U. F.; Wilkins, A. L.; Lauren, D. R.; Smith, B. L.; Towers, N. R.; Alonso-Amelot, M. E.; Jaimes-Espinoza, R. Phytochemistry,l997, 44, 901-906. [281] Imperato, F.; Phytochemistry, 1998, 47, 911-913. [282] Pamukcu, A. M.; Yalciner, S.; Hatcher, J. F.; Bryan, G. T.; Cancer Res., 1980, 40, 3468-3472. [283] Shahin, M.; Moore, M. R.; Worrall, S.; Smith, B.L.; Seawright, A. A.; Prakash, A. S.; Biochem. Biophys. Res. Commun., 1998, 250, 491-497.

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. Ailrightsreserved.

741

FLAVONOIDS AS NUTRACEUTICALS: STRUCTURAL RELATED ANTIOXIDANT PROPERTIES AND THEIR ROLE ON ASCORBIC ACID PRESERVATION *F.R. M A R I N \ M J . F R U T 0 S \ J.A. P E R E Z - A L V A R E Z \ F M A R T I N E Z - S A N C H E Z \ J.A. DEL RIO^

^Departamento de Tecnologia Agroalimentaria (Division de Tecnologia de Alimentos). Escuela Politecnica Superior de Ingenieros Agronomos. Universidad Miguel Hernandez de Elche. 033J2. Orihuela (Alicante). Spain ^Departamento de Biologia Vegetal (Unidad de Fisiologia Vegetal). Facultad de Biologia. Universidad de Murcia. 30100. Espinardo (Murcia). Spain. "^Author to whom correspondence should be addressed: Francisco R. Marin (Email: [email protected]; Phon.: 34966749667; Fax.: 34966749677. ABSTRACT: Chemicals generically referred to as flavonoids belong to the group of phenolic compounds and constitute an important group of secondary metabolites due to their applications as well as their biochemical properties. The similarity of these phenolic structures to other naturally occurring ones, as in some female insect hormones, does not allow for the use of a proper chemical definition based on structure. Therefore, the best way to define these structures is according to their metabolic origin. Hence, the phenols, and as a subclass, the flavonoids, can be defined as those substances derived from the shikimate pathway and phenylpropanoid metabolism. Flavonoids, which share a common benzo-y-pyrone structure, constitute a kind of compound which are highly ubiquitous in the plant kingdom. Over 4,000 different naturally occurring flavonoids have been discovered, and only in the case of flavones, a specific type of flavonoids, over 36,000 different chemical structures are possible. Flavonoids are present in a wide variety of edible plant sources, such as fruits, vegetables, nuts, seeds, grains, tea and wine. This finding reports a daily intake of flavonoid that ranges from 23 mg/day in the Netherlands to 170 mg/day in the United States. The numerous health related properties of flavonoids, widely described in epidemiological studies, are mainly based on their antioxidant activities. These properties have been found to include anti-inflanmiatory and antiviral activities, effects on capillary fragility, inhibition of human platelet aggregation and anticancer activity. The antioxidant capacity of any flavonoid will be determined by a combination of the catechol structure in the B-ring, the 2,3-double bond in conjunction with a 4-oxo fiinction, and the presence of both hydroxyl groups in positions 3 and 5. Due to their

742

excellent antioxidant properties, flavonoids are capable of preventing the ascorbic acid oxidation generally by reverting the ascorbate radical to ascorbic acid and by supporting the level of vitamin C in foodstuffs. Besides this antioxidant propert}^ some flavonoids, such as myricetin and quercetin, show the interesting ability to inhibit the ascorbate oxidase, preventing the enzymatic oxidation of ascorbic acid. Key Words: Flavonoids, nutraceuticals, antioxidants.

INTRODUCTION Flavonoids constitute a type of compound which is omnipresent in the plant kingdom. Over 4,000 different naturally-occurring flavonoids have been discovered and, unavoidably, they are part of our diet, due to the fact that they constitute up to 2% of the total photosynthesized carbon [1]. Flavonoids are plant secondary metabolites (as they have not been reported as naturally-occurring in animals), aromatics and belong to the group of plant phenols. The term phenol, and consequently the term flavonoid, can be precisely defined from the chemical point of view. However, as Robards and Antolovich reported in 1997, in a very interesting review of the analytical chemistry of flavonoids [1], within our biological context here, this definition is not entirely satisfactory. Basically, this definition may lead to confusion, including some chemicals being referred to as oestron, which is a sexual female hormone metabolically originating in the isopentenil-pirophosphate pathway. As a result, these authors proposed in 1997 a new definition based on the metabolic origin of these substances. Concerning the plant phenols, and consequently the flavonoids, it may be considered as those compounds originating in the shikimate and phenylpropanoid pathways. Notwithstanding, the flavonoids, as a differentiate subgroup inside the phenolic compounds, show a characteristic metabolic intermediate, the naringeninchalcone, from which all the bioflavonoids originate. Exclusively from a chemical point of view the flavonoids are characterized by a skeleton of three units, C6-C3-C6, that forms a cyclic structure in most cases [2]. In this skeleton two aromatic rings, referred to as A and B (in chalcones), can be distinguished, and an additional third ring (C), in the rest of theflavonoids.This last ring appears as a cyclation of chalcones with hydroxyl in 6' position Fig. (1). A and B rings have a different metabolic source. The B ring is formed in the shikimate

743

pathway, while the A ring comes from the condensation of three units of malonil Co-A [3, 4].

Fig. (1). Basic skeleton of theflavonoids.(i) Chalcones. (ii) Phenylbeiizopiran-4-one.

OH

Anthocyanidin

^^ Catecliin

Fig. (2). Chemical stmctiires of the various classes of flavonoids.

744

Different flavonoids can be classified according to the degree of oxidation of the three carbon central segment. From a lower to higher degree of oxidation the flavonoids are usually classified as catechins, chalcones, flavanones, isoflavones, flavan-3, 4-diols, flavones, aurones, flavonols and anthocyanins Fig. (2). The different basic structures can show different substitution patterns. These may include, as more often, hydroxylations, methoxylations, glycosilations and more rarely substituted methylendioxiflavonoids, such as pirano and fiirano flavonoids, with isoprenes, and even more unusually alkaloids, such as ficin, filospadin and litalin [5]. Also, flavanones, flavones, isoflavones, anthocyanins and catechins can be highlighted from a nutritional point of view and according to their importance as health agents and distribution. DIFFERENT TYPES OF FLAVONOIDS AND THEIR NATURAL OCCURRENCES Flavanones The basic structural model of flavanones is the 2-phenylbenzopiran-4-one skeleton [6]. The flavanones are compounds of great interest due to the fact that they are a compulsory step in the metabolic pathway of the other flavonoids. Their metabolic precursors are the chalcones, and the flavones, the dihydroflavonols, and the isoflavones are biosynthesised from the flavanones. Flavanones are widespread throughout the plant kingdom. They are present, as traces, in practically all the plant foods that express the metabolism of the flavonoids. However, it is in citrus plants, and more specifically in the fruits, where they occur as major flavonoids. In this case, in immature fruits, they can reach levels of around 40% in dry weight [7, 8]. These levels are reduced, during fruit growth, until percentages range from 1.5 to 5% of dry weight [9]. Since the very beginning, citrus flavonoids have awakened the interest of chemists. Thus, hesperidin was reported for the first time by Lebreton in 1928 [10]. Nowadays it is widely accepted that hesperidin, neohesperidin, and eriocitrin are the major flavonoids in citrus fruits. Other flavanones that may be found, aUhough in lower percentages, are isonaringin and isosakuranetin. Likewise, other compounds that occur as

745 triglycosides have been reported as traces [11]. The extensive natural occurrence of flavanones in citrus fruits leads to an arousal in interest concerning their presence in citrus products such as juices and marmalades. With the exception of naringenin, the remaining flavanones are practically restricted to citrus fruits. Notwithstanding, their presence has been reported in several foodstuffs, such as in the cuticle of tomatoes, where important amounts of naringenin and prunin have been reported [12, 13]. Peterson and Dweyer [14] reported in their review the presence of other flavanones, such as liquiritigenin in licorice, garbanzol in chickpeas, hesperidin in cumin and in some herbs such as peppermint. However, in the latter case, our results prove that peppermint does not contain hesperidin but isoferulil glucose ester (unpublished results). A phenolic acid is often mistaken for hesperidin, as happens in hyssop and other herbs, when accurate analytical techniques are not used [15]. Isoflavones Isoflavonoids constitute a characteristic and very important subclass of flavonoids. Their structures are based on the 3-phenylcromen skeleton that chemically comes, by an aril-migration mechanism, from the 2phenylcromen skeleton. Structurally, isoflavonoids can be classified according to the oxidation of the C15 skeleton, to their complexity and to the internal formation of heterocyclic rings [16]. Isoflavonoids come from the flavanones metabolically. The central step is the migration from the C2 to the C3 of the aril block, which constitutes the B ring of the flavanone intermediate. This reaction is catalyzed by the 2-hydroxyflavone synthase, a cytochrome P450. At the same time the isoflavones are precursors of an important number of compounds involved in the biosynthesis of phytoalexins and pterocarpanes. Isoflavones are mainly found in legumes and particularly in soy. Soy is the major source of genistein and daizdein, although their presence has also been reported in black beans, green split peas, and clover sprouts. The widespread use of soy products in infant foods, vegetarian formulations and as an ingredient in the composition of several foods leads to its ubiquitous presence in foodstuffs. Other isoflavones with nutritional relevance are brochamin and formononetin, which occur in

746 green beans, chickpeas, lima beans, split peas, alfalfa sprouts, and sunflower seeds [14]. Flavones Flavones have a main structural fingerprint, which is the presence of a double bond between carbons 2 and 3 of the central ring. From all the possible chemically derived products of the basic structure, those with a C3 hydroxylation pattern are specifically referred to as flavonols. This group constitutes a special type of flavones including some widespread compounds such as quercetin. Flavones are a class offlavonoidswith an increasing number of reports of naturally occurring products. In 1975, about 360 structures had already been reported [17]. This number grew to 720 in 1982 [18], to more than 1,000 in 1988 [19], and to over 2,000 in 1991 [20]. Wollenweber reported the estimations of lunuma and Mizuno to determine the possible amount of flavones, which reached 38,627, which can give us an idea of the enormous diversity of possible compounds [5]. Metabolically, the flavones are derived from the flavanones. According to several authors, the flavones are synthesized from glycosilated flavanones by a cytochrome P450 that produces a hydroxylation on C3 with a further dehydration, which leads to a double bond formation [21, 22, 23]. Flavonols are synthesized by a flavolnol synthase, another P450, which hydroxylates the C3 of theflavanone[24]. Both the flavones and flavonols are widely distributed in the plant kingdom. The compounds strictly defined as flavones are often found in grains and herbs. For example, apigenin and chrysoeriol can be found in parsley. In herbs, the flavone diosmin seems to have a ubiquitous distribution at least as a major flavonoid in Thymus sp, Rosmarinus sp. Origanum sp, Mentha sp and Hyssopus sp (unpublished results). The flavone diosmetin, the aglycon of diosmin, together with luteolin and apigenin, have also been reported in several herbs [25, 26, 27]. Apigenin and its glycosides are commonly found in cereal grains and vegetables such as celery and parsley [14]. Luteolin is also found in cereals, lemons, olives, celery, peppers and red grapes, while chrysin is often found in fruit skins [28, 29]. Some polymethoxylated flavones, a particular class of flavones, such as tangeretin, nobiletin, and sinensetin, are found in citrus peel and herb exudates [30, 31].

747

Flavonols are found throughout plant foods. Some flavonols extensively present in the plant kingdom are quercetin, rutin (the glycoside form of quercetin), and kaempferol. As an example, quercetin and its glycosides are present in onions (aprox. 4.8 g/kg dry weight), in apple peel as rutin (aprox. 0,57 gr/kg dry weight), in berries, black grapes, tea and broccoli. Kaempferol is often found in endives, leeks, broccoli, grapefruit, and tea. Isorhamnetin occurs in onions and pears, and myricetin is found most often in berries, maize, and tea [13,14, 29]. Anthocyanins Anthocyanins are water soluble pigments responsible for the red to purple colour in plants. The common basic structure of all these compounds is the cation flavylium, which was proposed for the first time by Wilstaer in 1913 and later confirmed by Robinson in 1922. A special kind of flavonoids related to anthocyanins are the proanthocyanidins, also referred to as leucoanthocyanidins, in the case of monomeric proanthocyanidins, and condensed proanthocyanidins for the polymers of fIavan-3-ols [32]. Metabolically, anthocyanins are built up from the dihydroflavonols by means of a reduction of C4, catalyzed by the dihydroflavonol reductase, which leads to the flavan-2,3-trans-3,4-cis-diols, which are intermediates of proanthocyanidins and anthocyanidins. However, despite all the data in this direction, it has not been possible to obtain in vitro the transformation of leucoanthocyanidins in anthocyanins [33]. Anthocyanins have a considerable potential in the food industry as safe and effective food additives. Their annual world production has been estimated to reach 10,000 tons from grapes alone. However, compared with synthetic colorants they have not been extensively used because of their instability towards a variety of chemical and physical factors. Notwithstanding, polyacylated anthocyanins display a marked stability which may prove to be of particular importance for food technology [33]. Anthocyanins occur widely in the red to blue coloured parts of plants. Therefore, they can be found in several berry fruits, such as cherries and strawberries, in plums, eggplants, red cabbages, and radishes. For example, delphidin can be found in eggplant skins and grapes; cyanidin in grapes, several species of stone fruits, raspberries, and strawberries;

748 pelargonidin in several species of stone fmits and grapes; and malvidin in black grapes and in coloured wines [33]. Catechins Catechins are also known as flavan-3-ols and differ from flavonols in the absence of the double bond between carbons 2 and 3, which is their fingerprint together with a hydroxyl at position 4 of the central ring. Different forms of catechins are epicatechins, gallocatechins, and epigallocatechins. Metabolically, catechins are clearly near relatives of anthocyanins due to the fact that they are formed from dihydroflavonols by the action of a flavan-3,4-cis-diol [24]. Although they can be found in limited amounts in fruits such as apples, cherries and pears, they also occur in considerable amounts in tea, particularly in the green type, and in smaller amounts in black grapes and red wine [1, 29]. The summary in table 1 shows some natural sources of the reported flavonoids and their most relevant properties as nutraceuticals. FLAVONOIDS AS NUTRACEUTICALS Dietary intake The healthy properties of flavonoids have been extensively studied from the epidemiological point of view, directly searching for their effect on enzymatic systems and/or their effect on physiological roles. Whatever the approach to assigning a heahhy property to these compounds, the functionality of the foodstuffs is going to depend on their content, intake, and bioavailability. So far, many studies have been performed on flavanoid content in plants but very few have been done on flavonoid content through food processing and dietary intake. Only a few studies have been done on products that are in the process of being cooked. Hertog et al [34] and Franke et al [35] reported high losses of nonpolar flavonoids during cooking. These losses ranged from 12.8% for kaempferol in Brassica oleracea to 72.5 % for quercetin in Viciafaba. For this kind of compound with such low polarity, any chemist will find it difficuh to accept such a

Table 1. Food sources and nutraceutical properties of the main flavonoids. Where concentrations are available they are quoted as extreme values [l,9,12-15,25311 Class of flavonoid Flavanones

Example

Food Sources

Hesperidin, neohesperidin

Found mainly in citrus h i t s , also in

Antiproliferative

effect

naringin, and isonaringin.

tomato cuticles.

Antiatherogenic

properties

Nutraceutical properties [73,

Concentration (mg/kg) 751,

1,500-5,000

11131,

Cardiovascular properties (prevention of platelet aggregation, haemorrhoids, etc) [96,97]. Flavones

Apigenin, luteolin, diosmin and

Found mainly in herbs, parsley, celery,

diosmetin.

citrus fruits, olives, peppers, red grapes, and some beans.

Antiproliferative Antimitotic

effect

[79],

[67,70-731,

Inactivation

10-4,ooo

of

carcinogenesis mediated by chemicals [90, 91, 921, Inhibition of angiogenesis [108, Cardiovascular properties [99, 100, 1141.

Flavonols

Quercetin, kaempferol and

Widely distributed. Main sources:

myicetin and their glycosides

Onions, apples, tea, red grapes, and

(rutin, etc.)

broccoli, citrus fruits and maize.

Daizein and genistein.

lsoflavonoids are found almost

Down regulation of mutagenic signalling

exclusively in legumes, particularly in

(68, 821, Inhibition of angiogenesis [106.

Antiproliferative effect [69], lnhibition of angiogenesis

[ 1081,

10-500

Cardiovascular

properties 1115, 1161, protection of DNA damage (60, 137, 138.1.

Isoflavones

soybeans. Anthocyanins

Catechins

Cyanidin, delphidin, delphidin and

Coloured berries

their glycosides.

and other fruits

Epigallocatechin, epicatechin

They are the main polyphenols in

gallate, and epicatechin.

green tea. Other sources are apples, cherries, and pears.

1071. Down regulation of mutagenic signalling

3.000-5,000

1841.

20-500

Down regulation of mutagenic signalling

10-300

F41.

750 large amount of flavonoid extraction in an aqueous system. Further studies need to be performed. With respect to industrial food processing, few studies have been performed. Some of our studies show that the flavonoid level in industrial citrus juice is higher than in homemade juice, and this finding is mainly due to the amount of albedo supplied during processing ranging, for orange juice, from 100 to 500 mg/L, according to variety [36]. Results published with respect to dietary intake of flavonoids are also poorly documented, divergent, and restricted to a few classes of compounds. The most frequently reported figures are estimations of 23 mg/day in the Netherlands and 170 mg/day in the United States [37]. The former studies only quantify apigenin, kaempferol, luteolin, myricetin, and quercetin. All other flavonoids were either disregarded or estimated rather than being measured directly, which makes it more clear that these accepted results are not reliable at all^ as they have been considered until now. Older studies situate the dietary intake at 1 g/day, but less precise techniques were used [38]. Other authors range the daily consumption of flavonols plus flavones between 3 and 80 mg, as is the case with quercetin [39, 40]. Whatever the finding may be, it is difficult to disregard the amount of nutraceuticals present in fruits and vegetables. However, it should suffice to follow the recommendation of a minimum of 5 servings of fruit and vegetables per day by the U.S. National Cancer Institute and the U.S. National Research Council [41]. Independent of dietary intake, the aspects that are relatively well documented are the increase in antioxidant capacity of plasma and bioavailability when foodstuffs rich in flavonoids are consumed. Serum antioxidant capacity can increase up to 25% during the 4 hour period following consumption of red wine or strawberries and up to 50% in urine during the 24 hour period following consumption of the same foodstuff [42]. Alcohol-free red wine enhances plasma antioxidant capacity, since it is 20 times more active than alcohol-free white wine [43]. Bioavailability of several flavonoids have been studied, and these studies have mainly focused on those of utmost relevance, genistein, and quercetin. Only a few studies on this topic are found in the literature, and some of the most relevant were compiled by Hollman [44]. Studies performed on genistein (an isoflavone aglycon) absorption by using 4-^'^C-genistein showed that this isoflavone is absorbed in the rat duodenum. It is then taken up by the liver and excreted in the bile as its 7-0-P-glucuronic conjugate with

751 accumulative recovery ranging from 10-15% of the dose [45]. When oral intake of soybean products is studied in man, genistein and daizein are found in plasma 2 hours after feeding and reach their highest concentration 6 hours later. The half-life of these products was 8.36 and 5.79 hours for genistein and daizein, respectively^ with major recovery in faeces [46]. Similar studies performed on humans show that quercetin is detected in plasma 2 hours after consumption and starts to decrease after 7 hours and is detected as its conjugated 3'-methylquercetin [47]. The biotransformation of flavonoids occurs in microsomes by means of the family of CYPlAl of cytochrome P450 and mainly involves hydroxylations, which exclusively occur in the B-ring, and demethylations [48]. Catechins follow the same former pattern. Studies performed on humans showed that catechins are mainly methylated as 3 0-methylcatechin and less than 2% are found as glucouronide and sulphate conjugates [49]. Similar behaviour has been described in rats for epicatechins, which are supposed to be methylated in the liver and kidney, glucuronidated in the intestinal mucosa and sulphated in the liver [50]. Healthy properties (nutraceuticals) Within the last years, flavonoids have been highlighted as possible chemopreventive dietary agents against cancer [37, 51, 52]. Over 200 studies in the epidemiological literature have been reviewed these show an association between cancer incidence and lack of adequate consumption of fruits and vegetables. Consumption of adequate fruits and vegetables is associated with a lowered risk of degenerative diseases, including cancer, cardiovascular diseases, cataracts, and brain malfunction. The estimations of diet responsibility on cancer risk account for a third of the risk [53, 54, 55, 56]. Protective effect on DNA Flavonoids, due to their ability to absorb ultraviolet (UV) radiation, can protect DNA (Deoxyribonucleic acid) from the damage caused by UV radiation. This effect is one of the physiological functions proposed for the flavonoids in plants [57]. In studies performed with UV-B irradiated plasmids, both naringenin and rutin showed protecting activity against DNA damage, induced by UV radiation [58]. Besides a direct protection,

752

by absorbing UV radiation, the ability of the DNA to scavenge free radicals, which can damage the DNA when generated near it, may justify the observed protection in those studies where mice, fed with a rich diet in flavonoids, were y-irradiated [59]. Pretreatment with flavonoids and vitamin C produced a dose-dependent reduction in oxidative DNA damage, and this was the protective effect of those with aglycon structures stronger than for vitamin C [60]. The structurally-related properties will be discussed in the corresponding section. In a similar way flavonoids may protect the DNA by interacting with carcinogens that escape from the cell detoxification machinery, as happens with the chromosome aberrations induced by bleomicin. In this case, it has been observed in vivo and in vitro that treatment of lymphocytes with galangin, a flavonoid metabolic derivative, suppressed the induction of chromosome aberrations by bleomicin in a dosedependent manner [61]. Antiproliferative effects and regulation of involved signalling Flavonoids have shown an antiproliferative effect on several human neoplasic cell lines such as lymphoid and myeloid leukaemia cells [62], gastric cancer cells [63] ovarian cancer cells [64] and prostate cancer cells [65]. Several chalcones display significant cytotoxicity towards murine P388 and L1210 leukaemia cells as well as a number of human tumour cell lines [66]. Recent research has identified the active principles of mandarin King juice that strongly induce differentiation of HL60 cells, promyelocytic leukaemia cells, [67] as nobiletin, 3,3 ,4,5,6,7,8heptamethoxyflavone, natsudain and tangeretin. Genistein inhibits proliferation and differentiation of N2A, JC, SKNSH, MSN, and Lan5 neuroblastoma cell lines and in one (N2A line) induced apoptosis [68]. The flavonol quercetin inhibits proliferation of stellate cells, which play a central role in development of hepatic fibrogenesis when the liver is damaged by viral infection, alcohol and various drugs [69]. Some polymethoxylated flavones, such as nobiletin and tangeretin, inhibit the growth of cancerous cell lines in a dose-dependent manner [70] and show some degree of inhibition of mammary cancer [71, 72]. On the other hand, other citrus flavonoids such as diosmin and hesperidin reduce the incidence and multiplicity of neoplasm in the large intestine of male F344 rats initiated with azoxymethane [73]. In spite of the great amount of

753

published reports on this subject, the involved intracellular mechanisms are still not clearly understood. Flavonoids may affect several enzymatic activities, be they involved or not, in the transmission of mutagenic signalling such as kinases, phospholipases, phosphodiesterases and/or regulating others, which can play a critical role in cell proliferation and growth. Middleton and Kandaswami reviewed in 1993 the effects of flavones and flavonols on 24 different types of enzymes [74]. With respect to those enzymatic activities not involved in carcinogenic processes^ flavonoids, in particular citrus flavonoids, have been correlated (when rats were fed with citrus extracts) with lower 3-hydroxy-3-methyl-glutaryl-CoA reductase and Acyl Co A: cholesterol 0-Acyltransferase, which has been related to lower levels of cholesterol in plasma [75]. Usually the effect of supplying flavonoids leads to an inhibition of glycolysis, which is generally very active in tumours [76]. Other studies performed on leukaemia cell lines point to a depression in lactate production. These effects could be due to the inhibition of lactate transport as well as various membrane ATPases [77, 78]. It has been reported that some phenolic compounds such as resveratrol and flavonol quercetin may suppress inositol phosphate metabolism, tyrosine phosphorylation, and mitogenic-activated protein kinase activation in platelet-derived growth factor/BB-stimulated stellate cells [69]. Another cytotoxic effect which is potentially beneficial from flavonoids is the interference with tubulin polymerization that causes mitotic arrest (which is the mode of action of compounds such as vincristine, vinblastine, and taxol). In this respect, flavones were to a certain extent successful in binding to tubulin. In this case, 3-methoxy substitution is essential and within this group, a 3'hydroxyl and 4'methoxyl substitution pattern is required [79]. However, some synthetic chalcones can be found among some of the most effective drugs to prevent tubulin polymerization. This ability may require some structural requirements to increase activity. Thus, substitution of aryl groups by methyl substitutions, and particularly methylation on C2, greatly enhances the ability of the chalcone to inhibit cell growth. These properties can be structurally correlated with some antimitotic compounds that act by inhibiting tubulin polymerization as combretastatin A-4 and colchicine Fig. (3) [80], Lately, the interference of several flavonoids on tyrosine kinase activity as one of the ways to interfere in cancer development has been highlighted. Enhanced protein tyrosine kinase activity due to

754 overexpression of receptor and/or protein tyrosine kinases leads to a continuous signalling resulting in uncontrolled cell proliferation that produces cancer growth [81]. It has been reported that genistein can down OH OCH3

D

OH

O

CH3O

l<^

.OH

CH3O

HOC

CHsO'

OCH3

OCH3

CH3

}=° kx^N-^"^"^ CH30

CH3

CH30

Fig. (3). Structures of flavonoids that show interference with tubuhn polymerization and antimitotic compounds structurally correlated with the fomier. A: Flavone; B, C: Synthetic chalcones; D: Combretastatin A-4; E: Colchicine.

regulate intrinsic protein tyrosine kinase involved in neuroblastoma development [68]. Treatment with silymarin (a flavonoid present in thistles) resuhed in a significant inhibition of ligand-induced activation of EGRFR (Epidermal Growth Factor Receptor), which is mediated by tyrosine phosphorylation with no change in its protein levels, which may lead to a chemopreventive effect on skin cancer [82]. In a similar way, silymarin has inhibitory effects on cell growth and proliferation of breast carcinoma cells, MDA-MB 468. In this case, it is reported to be associated with a Gi arrest in cell cycle progression concomitant with an induction in the protein expression of cyclin-dependent kinase inhibitor

755 Cipl/p21, which leads to a decrease in kinase activity [83]. Other flavonoids, such as epigallocathechin-3-gallate^ selectively inhibit the intracellular signalling transduction pathway, mediated by tyrosine kinase, and inhibit transformation of fibroblasts [84]. The early components of signal transduction pathways, specifically those of tyrosine kinase, are of utmost significance for the control of cell growth and differentiation [85]. An alteration in this chain can result in a continuous signalling, leading to uncontrolled cell growth and proliferation. Increased expression of the erbB family of RTKs (Receptors of Tyrosine Kinase) has been implicated in a wide variety of human carcinomas [86, 87]. Concerning human PCA (Prostate Carcinoma), the aberrant expression of the erbB family of RTKs has been demonstrated with strikingly high frequency in prostatic intraepithelial neoplasia and in invasive PCA. In this respect recent studies have proven that silymarin inhibits the activation of erbBl signalling and consequently the early stages of PCA [88] Frequently the induction of carcinogenesis depends on the metabolic activation by a specific chemical. Compounds such as heterocyclic amines, and others such as benzoanthracens, resulting from daily cooking, require activation mediated by the citosolic transcription factor AhR (Aril Hydrocarbon Receptor) to induce carcinogenesis [89]. The AhR binds to the aril hydrocarbons and/or polyclorate bifenils, translocating them to the cell nucleus, where they bind to a second protein, the nuclear aril hydrocarbon translocase. These heterodimers bind to specific DNA sequences, leading to overexpression of a great number of genes. One of the best known responses is the induction of the gene CYPlAl, which codifies a P450 lAl enzyme. This enzyme transforms chemicals such as 7,12-dimetilbenzanthracene into others capable of forming abducts with DNA that may trigger the carcinogenic process. In this respect, it has been reported that flavones (and flavonols) can stimulate the response, against xenobiotics, of the gene CYPlAl by enhancing the AhR complex-DNA binding [90]. On the other hand, other flavonoids with a different structure, such as isoflavones, are unable to inhibit the carcinogenesis induced by benzoanthracenes, which may imply some spatial requirements for the inhibition [91]. Both diosmin and aglycon diosmetin display the ability to induce the DNA abduct formation. However, diosmetin inhibits the CYPlAl enzymatic activity and consequently the carcinogenic activation [92]. Despite the fact that diosmin, the glycoside form of diosmetin, is the major form in plant material, it does not display any chemoprotective effect [92]. On the other

756 hand, Ciolino et al [92] reports that, according to Cova et al [93], the diosmetin is the primary form of circulation of this flavonoid in humans and widely accepted by the tissues. This may justify the chemopreventive effect of diosmin on dietary intake. However, these flavones do not have the antagonistic effect of dioxins, which also bind to AhR, due to the fact that they need a 3'-methoxy and 4'-hydroxy pattern of substitution (B ring) as a structural requirement while diosmin and diosmetin show a 3 hydroxy and 4'-methoxy pattern of substitution, which is the naturally occurring pattern for flavonoids Fig. (4) [94]. A _,. B OH OCH .

Fig. (4). Differences between the required pattern of substitution for antagonist effect of dioxin (A) and the naturally occurring pattern (B).

Properties related to vascular disorders The effect of flavonoids on bleeding and capillary fragility was first reported by Rusznyak and Szent-Gyorgyi in 1936, who considered citrus flavonoids to have vitamin activity, which they named vitamin P [95]. These groups of flavonoids have, according to several researchers, effect on platelet aggregation that blocks microcirculation and leads to the corresponding pathology [96, 97]. Capillary damage includes increased permeability, seepage of blood and plasma constituents into the tissues, followed by an inflammatory reaction. Nowadays all these disorders can be treated by many drugs, which are based on flavonoids basically derived from diosmin or hesperidin methylchalcone and hydroxyethylrutosides, which act primarily on the microvascular endothelium to reduce hyperpermeability and edema. In patients with chronic venous insufficiency and/or diabetes, hydroxyethylrutosides improve microvascular perfusion and microcirculation and reduce erythrocyte aggregation. These preparations also alleviate the symptoms of severe haemorrhoids [98]. In particular, the glycosilated flavone diosmin leads to a significant decrease in venous capacitance, venous

757

distensibility, and venous emptying time [99]. Other pharmacological activities of diosmin are interference with edema mechanisms and interference with hyperpermeability induced by bradykinin or ischemia in rats and increased lymphatic flow in anaesthetized dogs [100]. Some researchers have also correlated the increase in vascular tone observed in vivo after treatment with these drugs to the inhibition of amine reuptake. In this case the flavonoids may be acting as antagonists of plasma membrane amine transporters [101] while the mechanism of vasodilatation may be explained by the inhibition of protein-kinase C [102]. Another positive effect of flavonoids on vessels, although not in capillary vessels, is the vasorelaxation in the rat thoracic aorta produced by anthocyanins and oligomeric-condensed tannins present in red wine which may lead to a reduction in cardiovascular disease risk [103]. Angiogenesis is a process intimately bound to vessel growth and carcinomas in which flavonoids can play an important role. The generation of new capillaries from pre-existing vessels is absent in the healthy adult organism except in very few cases. Angiogenesis is a wellregulated and self-limited process. However, pathological angiogenesis is present during the development of some diseases and particularly in tumours. Well-vascularized tumours expand both locally and by metastasis, whereas avascular tumours do not grow beyond a diameter of 1-2 mm [104, 105]. Genistein is a well-known and potent inhibitor of cell proliferation and in vitro angiogenesis [106, 107], which can lead one to think that this property is related to the isoflavone structure. However, recent studies prove that some flavonoids, but not isoflavones, are even better antiangiogenic agents than genistein. Fotsis et al [108] found that several flavonoids such as 3',4'-dihydroxyflavone^luteolin and 3-hydroxyflavone had stronger inhibition of angiogenesis that genistein. Apigenin, eriodyctiol and quercetin showed a similar effect while luteolin glycoside, fisetin, myricetin, hesperetin, and catechins showed a lower effect. These results suggest that a nonhydroxylated ring C with oxo function at position 4 and a double bond between C2 and C3 is required for maximal biological activity. Moreover, a glycosilation pattern seems to imply the lack of these properties. These structural relations with inhibition of angiogenesis seem clearly not to be related to antioxidant properties which are improved with the combination of the former substitutions. Thus, Fotsis et al [108] suggests that this behaviour should be correlated with early events and some enzymatic inhibitory properties such as

758 tyrosine kinases [109] and protein kinases [110, 111]. The structural properties highlighted previously can make them competitive inhibitors with respect to the ATP binding site on a variety of enzymes, a region of considerable homology among kinases [112]. Some flavonoids show an antiadhesive and antiaggregation action against red cell clumping [113]. Platelet aggregation contributes to both the development of atherosclerosis and acute platelet thrombus formation followed by embolization producing cyclic flow reductions in stenosed and damaged arteries. In this respect, methoxylated flavones, such as nobiletin and tangeretin, display much more activity than hydroxylated compounds and their action might be similar to that of acetylsalicylic acid in relation to its properties to inhibit platelet aggregation. Further studies showed that these methoxylated flavones have an effect on platelet aggregation in the range of 30 |LIM while glycosilated flavanones did not show it even up to a 7-fold higher dose [114]. Other researchers have reported consistent results with the aforementioned and suggest that flavonoids act by inhibiting platelet adhesion, aggregation, and serotonin secretion [76]. From several tested flavonoids, only fisetin, kaempferol, and quercetin inhibit platelet aggregation caused by arachidonic acid, with the exception of myricetin, which was effective at a higher dose than the former ones [115]. When researchers compared the effect of several juices with high flavonoid content on platelet aggregation, they found that those containing quercetin, kaempferol, and myricetin (grape juice) had a strong inhibition of platelet aggregation. On the other hand, others such as citrus juices (orange and grapefiuit^ both rich in flavanones but not in flavonols) did not have this same inhibition but a lower one [116], which is also consistent with the results formerly obtained by other researchers. It is widely known that cyclic nucleotides are involved in fundamental processes, which include those related to platelet aggregation. In this way, the inhibition of the enzyme phosphodiesterase could be one such mechanism through its effect on cAMP (Cyclic Adenosin Monophosphate) and cGMP (Cyclic Guanosine Monophosphate) [117]. Another mechanism might involve inhibition of cycloxygenase, with a consequent depression of thromboxane A2 biosynthesis [115] as well as some influence on the inhibition of the intracellular mobilization of Ca'^^ flux and its influx across the plasma membrane, as happens with genistein [118].

759 The possible activity of flavonoids in antiinflammatory and antiallergic responses is well documented by Gabor's review [119]. Recent studies [120, 121] have shown the activity, in a dose dependent manner, of diosmin, hesperidin and other flavonoids as well as their influence on the metabolism of arachidonic acid and histamine release. The former flavonoids significantly inhibit lysosomal enzyme secretion and arachidonic acid release from membranes by inhibiting lipooxygenase, cycloxygenase, and phospholipase A2. Some researchers indicate that the inhibition of arachidonic acid release would provide less substrate for the lipooxygenase and cycloxygenase pathways, leading to a lesser quantity of endoperoxides, prostaglandins, prostacyclins and thromboxanes on the one hand and hydroperoxy and hydroxyeicosatrienoic acids and leukotrienes on the other [119]. The glycosilated flavone diosmin behaves as a powerful protective agent against inflammatory disorders. The flavone diosmin reduces edema formation and inhibits the synthesis for prostaglandin E2, F2, and thromboxane B2 (78%, 45% and 59% respectively). Intravenous injection of diosmin reduces hyperglycaemia induced by alloxan in rats. This effect was related, according to Jean and Bodinier [122], to its ability to scavenge active oxygen radicals. With respect to the antiallergenic activity, the diosmin precursor, (flavanone hesperidin), inhibits histamine release in rat breast cells, which has led one to suggest its use as an effective antiallergenic drug [123]. Other properties Flavonoids also showed, to some extent, some antifungal and antiviral activity. In this case, there is an important structure-activity relationship. The flavonol quercetin and the flavanone hesperidin exhibit inhibition activity towards the infective capacity and/or replication of herpes simplex type viruses and polio viruses^ while the flavanone naringenin totally lacks this ability [124]. For researchers the impossibility to dissociate, the viruses from the flavonol quercetin after 1 hour of interaction, either by dialysis or ultracentrifugation suggests the formation of quercetin-virus complexes, which may have lost the ability to induce infection. With respect to the antiviral activity of the methoxylated flavones, this is strongly related to a substitution pattern based on

760 methoxy substitutions on 4' and 3'-dihydroxyls and a higher activity in those cases with polysubstitutions on A ring [124]. Are the flavonoids totally innocuous? Not all the effects of dietary intake of flavonoids are positive. The excessive consumption of soybean and its products has been considered goitrogenic in humans and animals. Several researchers have reported induction of goitre in iodine-deficient rats maintained on a soybean diet [125, 126, 127]. In some cases, the extreme intake of soybean has been correlated with cancer. Thus, Kimura et al reported the induction of thyroid carcinoma in rats fed on defatted soybean deficient in iodine by up to 40% [128]. Recent studies have highlighted an explanation for this undesirable effect of soy flavonoids^ The function of the thyroid is the synthesis of thyroid hormones, and TPO (Thyroid Peroxidase) catalyzed iodination of tyrosil residues on thyroglobulin and the subsequent coupling of iodotyrosyl residues required for iodothyronine hormone formation. In the presence of iodine ions, genistein and daizein (the major soy flavonoids) block TPO-catalyzed tyrosine iodination by acting as alternate substrates, yielding mono-, di- and triodoisoflavones. Genistein can also inhibit the synthesis of thyrosine using iodinated casein or human goitre thyroglobulin as substrates for the coupling reaction [129]. Genistein and genistin have been known to be strong cytotoxic agents in vitro. This characteristic can be an advantage when target cells are malignant but can be disadvantageous when they are normal cells. Recent experiences carried out with rat myogenic cells (L8) showed that genistein and genistin strongly inhibited in vitro myoblast proliferation and fusion in a dose-dependent manner. Genistein also inhibited protein accretion in myotubes. Decreased protein accretion is largely a result of cellular (myofibrillar) protein synthesis rate while no adverse effect on protein degradation was observed. The results suggest that if sufficient circulating concentrations are reached in tissues of animals consuming soy products, genistein can potentially affect normal muscle growth and development [130]. When soy-based products are promoted as healthy foods, possessing putative beneficial estrogenic and anticancer activity properties, some of them based on their isoflavones, the findings mentioned in the previous

761 paragraph should be highlighted, due to the widespread use of these products in infant food formulas and the consumption of soy products by people with vegetarian diets. Another negative effect of flavonoids, which does not affect humans but does affect bovines, is the action of quercetin as co-carcinogen of BPV-4 (Bovine Papilloma Virus type 4). The transfection with BPV-4 DNA and exposure to a single dose of quercetin leads to tumorigenic transformation of primary bovine cells [131] which can make us think about the possible risk of foods supplemented with flavonoids. Continuing with the effect on animals^ it has been discovered that when heifers were fed with a diet supplement of 500 g/day of green tea (a very good source of catechins), the iron content of muscle was reduced [132]. So far, there are many studies pointing in the direction of flavonoids as heahhy agents. Most of them have been performed on in vitro based research and when performed in vivo, animals have been generally used as subjects. Some heterodox voices are lately disclaiming those putative healthy properties. Thus, recent studies performed on humans reported that flavonoids such as quercetin may not have any effect at all on risk factors for heart disease when supplied orally as supplements in a diet. Quercetin intakes 50-fold greater than in a normal diet did not affect the cardiovascular or thrombogenic risk factors [133]. Similar evidence for genistein and daizein have been found when studies were performed on humans [134]. NUTRACEUTICAL PROPERTIES ANTIOXIDANT ABILITIES.

RELATED

WITH

Free radical formation is associated with the normal natural metabolism of aerobic cells. The oxygen consumption inherent to cell growth leads to the generation of a series of free radicals. The interaction of these species with molecules of a lipid nature produces new species such as hydroperoxides and different kinds of peroxides [135, 136]. This group of radicals (superoxide, hydroxyl, and lipoid peroxides) may interact with biological systems in a cytotoxic manner. In this respect, it has been shown to posses an important antioxidant activity towards these radicals, which is mainly based on the properties of the hydroxyphenolic groups and the structural relations between the different parts of the chemical structure. Together with an ability to capture electrons, these

762 characteristics confer great stability to the flavonoid radical formed by means of a tautomeric dislocation, which prevents the propagating chain reactions of these oxygen-free radicals [137, 138, 139]. As polyphenolic compounds, flavonoids have the ability to act as antioxidants by a free radical scavenging mechanism [140, 141, 142]. This involves the formation of less reactive flavonoid phenoxyl radicals (Eq 1 and 2); on the other hand, through their known ability to chelate transition metals [143, 144], these compounds may inactivate iron ions, and bivalent ROO* + Fl-OH -> ROOH -f Fl-0* HO* + Fl-OH -^ H2O + Fl-0*

(1) (2)

metals, through complexation, thereby suppressing the superoxide-driven Fenton reaction (Eq 3 and 4), which is currently believed to be the most important route to activate oxygen species [145]. 02*" + FeIII ^ 0 2 + F e l l Fe II + H2O2 ~> Fe III + HO* + HO"

(3) (4)

So far it has been widely believed that the antioxidant ability of flavonoids resides mainly in their ability to donate hydrogen atoms and thereby scavenge the free radicals generated during lipid peroxidation. In spite of some flavonoid structures which allow them to form heavy metal complexes, metal chelation has been regarded to play a minor role in the antioxidant activity of these compounds and has not been studied much by researchers Notwithstanding, some recent studies point in this direction. When erythrocytes are in the presence of oxidizing agents, iron is released in a free form (as chelatable desferrioxiamine) and does not bind to specific proteins as it is naturally occurring in living systems. This iron can easily lead to membrane lipid peroxidation and hemolysis, this being the effect of iron inside the cell [146]. Recent studies show that certain flavonoids, such as quercetin, can penetrate the cell and can bind iron, thus preventing lipid peroxidation and hemolysis. Rutin and other flavonoids can also chelate iron, but their penetration inside the cell is lower than that of quercetin [147]. When comparing the effectiveness of several flavonoids (quercetin, rutin, luteolin, chrysin, naringenin and hesperetin) at inhibiting metal-ion-induced peroxidations with peroxidations induced by the water

763 soluble free radical generator AAPM (2, 2'-azobis (2-amidino propane) dihydrochloride), researchers found that these compounds were more effective at inhibiting metal-induced peroxidation than AAPM-induced peroxidation [148]. In this study, quercetin, rutin and luteolin exhibited higher antioxidant activities than in the other flavonoids. This provides more evidence for the highly significant contribution of the 3', 4 dihydroxy phenyl substitution pattern on the B-ring, as will be discussed below with the structural requirements for these chelating properties. As has been mentioned above, flavonoids can protect the DNA from damage by acting as free radical scavengers. In this respect Noroozi et al [60] carried out a study in which the protective effect of vitamin C and several flavonoids (luteolin, myricetin, kaempferol, quercetin, apigenin, quercetin-3-glycoside and rutin) against DNA damage was compared. At the assayed concentrations the protective effect of vitamin C against DNA damage was significantly lower than that of all the flavonoids except apigenin, quercetin-3-glycoside and rutin (all of them glycosilated structures). The data suggest that the free flavonoids are more protective than the conjugated flavonoids. These data are also consistent with the hypothesis that antioxidant activity of free flavonoids is related to the number of hydroxyl groups [137, 138]. Nevertheless, the tremendous lack of antioxidant activity when compounds are conjugated suggests that polarity, as cell permeability, may play a very important role. However, so far the protective mechanism of flavonoid is not entirely clear. When prevention of DNA single strain breakage is evaluated, the chelating mechanism is shown as the main factor. By using an experimental approach based on the notion that iron chelators suppress DNA strand scission and cytotoxicity caused by tert-butylhydroperoxide (a free radical generator), whereas radical scavenging antioxidants prevent only the latter responses, Sestili et al [149] provide experimental evidence indicating that the most prominent activity of the flavonoid quercetin resides in its ability to chelate iron. Further research points in this direction. Thus, when Jurkat T cells are supplemented with green tea (very rich in epigallocatechin gallate) and treated with Fe^^ as prooxidant, they suffered much less oxidative DNA damage (single strand breaks) than those of the control group[150]. Another healthy property of flavonoids related to their antioxidant ability may be the prevention of atheroma formation. Within the last several years, the oxidative theory of atherogenesis has provided another avenue of therapy using antioxidants [151]. According to this theory.

764 antioxidants should protect lipoproteins against oxidative modification and reduce the biological consequences. Vitamin C is the major chainbreaking antioxidant against free radicals and protects LDL (Low Density Lipoproteins) from oxidation in vitro, ahhough it does not bind to LDL [152]. Flavonoids are powerful antioxidants and they act against cupric ion-induced in vitro LDL + VLDL (Very Low Density Lipoproteins) oxidation [153] and also bind to LDL [154]. When citrus extracts, containing flavonoids with and without ascorbic acid^ were tested against oxidation-susceptible lipoprotein, the authors found that citrus extracts containing flavonoids but not ascorbate did not significantly change any lipid contents. However^ citrus extracts containing flavonoids, plus ascorbate^ had a strong beneficial effect on hamster lipids (lower triglycerides and significantly decreased cholesterol and LDL+VLDL compared to the control) [155]. Thus, for the authors there appears to be an in vivo synergism of the flavonoids and ascorbate with respect to the cholesterol, LDL+VLDL, triglycerides, and the atherogenic index. However, no correlations with chelation metal ions were found [155]. Further evidence obtained from in vivo experiences supported the former theory. Thus, when rats fed with PUFA (Polyunsaturated Fatty Acids) or MUFA (Monounsaturated Fatty Acids) were supplemented with flavonoids (a mixture of quercetin and catechins) the amount of dienes produced was significantly reduced and consequently there was lipid peroxidation [156]. Structural requirements for scavenging and chelating properties Bor et al performed studies based on the generation of flavonoid radicals (aroxyl radical) by pulse radiolysis [138, 157, 158]. The observations of fast absorption changes with kinetic spectroscopy allowed the determination of the ratio of the constants, which were obtained with several oxidizing radicals of flavonols such as kaempferol and quercetin. These studies revealed on the one hand that B-ring-localized semiquinones are the major radical species observed after univalent oxidation of dihydroflavonols, flavanones, and flavanols. Simultaneously revealed on the other hand was that a saturated 2,3 bond leads to a break of the 7c-electron system between the carbonyl group and the B-ring. The presence of a 2,3-double bond (flavones and flavonols) led to a dislocation of the scavenged electron across the flavonoid structure. Such

765 results led Bor et al [138] to postulate that it was basically the presence of a B-ring catechol group which stabilized the aroxyl radical. In summary, in the chemical skeleton of the flavonoids there are three basic groups which determine the ability to scavenge radicals: (I) The o-dihydroxy (catechol) structure in the B ring which confers great stability to the aroxyl radicals and which participates in electron dislocation; (II) The double bond between carbons 2 and 3 in conjugation with a 4-oxo function, which is responsible for electron dislocation from the B ring; and (III) The presence of both 3 and 5 hydroxyl groups for maximal radical scavenging capacity and strongest radical absorption [137, 159]. When antioxidant activity of flavonoids is evaluated in liposomal systems, the 3-hydroxy group becomes less importance and an ohydroxylation is considered at A-ring as a requirement for antioxidant activity [148] Fig. (5). (I)

OH

I

OH

OH

HO.

OH

O

(IV)

(III)

OH

HO

OH--- O

OH

O

Fig. (5). Structural criteria tliat enliances the antioxidant activity of flavonoids. 1,11, III: Structural groups proposed by Bor et al [137,138, 157]. IV: Stmctural group proposed by Arora et al [148].

Antioxidant activity of flavonoids and in non-processed foodstuff which are present has been extensively studied [13, 29, 155, 160, 161,

766 162]. Concerning these, and those previously reported, there results a hierarchy of antioxidant activity that has been proposed in which flavanone 4'-hydroxylated keeps the lowest activity and flavonol with catechol structure in the B-ring the highest one. Fig. (6). Table 2 shows the antioxidant activity of some flavonoid families obtained by different researchers. OH

x/^

OH OR

<

Fig. (6). Antioxidant liierarchy, in increasing order, of flavonoid structures. A: Flavanone 4'-OH. B: Flavone 4'-0H. C: Flavonol 4'.OH. D: Flavanone 3'-OH, 4'-OR. E: Flavone 3^-OH, 4^-OR. F: Flavonol: 3'-0H, 4'-0R. R: H or CH3.

The results presented in table 2 show that when 3-hydroxy 1 is absent or substituted, its contribution to electron dislocation is substantially reduced and so is consequently the flavonoid antioxidant activity, although this reduction is smaller when this hydroxyl group is substituted (rutin) than when it is absent (diosmin). These results confirm the importance of the 3hydroxyl group for a better capacity to revert radicals [29, 138, 159, 163]. However, the presence of a double bond between C2 and C3 when 3hydroxyl group is absent (diosmin) does not significantly increase the antioxidant capacity offlavonoidsin respect to those that do not present this double bond (hesperidin) in spite of the fact that the double bond should increase the conjugation of the structure according to the aforementioned [139].

767 Table 2. Relative antioxidant potential of several flavonoids, vitamin C, and vitamin E, measured as the TEAC (Trolox equivalent antioxidant activity) [13, 29, 160,165]. Chemical

Antioxidant activity (mM)

VitaininC'

LO it 0*02

Vitamin E

1.0 ±0.03

Anthocyanins Cyanidin

4.410.12

Delphidin

4.410.11

Flavonols Quercetin

4.710.10

Kaempferol

1.310.08

Rutin

2.410.12

Flavones Apigenin

1.510.08

Diosmin

1,2010,09

Luteolin

2.110.05

Catechins (Epi) catechin

3.810.02

Epigallocatechin gallate

4.810.06

Flavanones Naringin

0,0510,01

Naringenin

1.510.05

Hesperetin

1.410.08

Hesperidin

1.010.03

This phenomenon might be due to the strong intramolecular hydrogen bond between the 5-hydroxyl group (A ring) and the 4-oxo group (Bring), v^hich may reduce the oxo group contribution to the electron

768 dislocation. This seems to have similar consequences for the antioxidant activity of those structures in which 4-oxo group is absent as in the catechins [29, 160]. For these, the absence of the 4-oxo group and the simultaneous presence of 3 and 5 hydroxyls would weaken the hydrogen bond. This should lead to an improvement in the electron dislocation and thereby generate a higher antioxidant activity than the expected due to the double bond lack. Another important structural requirement for the antioxidant activity is the number and distribution of hydroxyl groups [163]. When just one hydroxyl group is present in B ring (naringenin) the antioxidant activity is strongly reduced when compared with other compounds that have catechol structure or a methoxy substitution (hesperidin), in a similar way as discussed above for rutin and diosmin. This suggests that a lone hydroxyl in the B ring does not contribute at all to the antioxidant activity. Glycosilation patterns lead to a reduction in the antioxidant activity with respect to the aglycon original structure (rutin versus quercetin and hesperidin versus hesperetin). On the one hand, the contribution to the molecular weight, of the phenolic structure is lower in the glycoside than in the aglycon^ and on the other hand the glycosilations may lead to conformational changes in the flavonoid structure, which might affect the dislocation ability. According to the results shown in table 2^ a flavonoid hierarchy for the antioxidant activity, when measured as ability to revert the ABTS*^ (2,2Azinobis 3-ethylbenzothiazoline-6-sulphonic acid) radical, can be established as follows: epigallocatechin gallate » quercetina » cyanidin « delphidin > rutin > luteolin > apigenin « hesperetin « naringin > diosmin « hesperidin > naringin. The flavonoids with the highest activity are quercetin, catechins and anthocyanins^ which have an antioxidant activity up to three times higher than vitamin C (TEAC (Trolox equivalent antioxidant activity): 1,12 mM) and slightly higher than licopen (TEAC: 2,9 mM) [29]. Others, such as naringin, have a similar, or even lower, antioxidant activity than vitamin C. As discussed previously flavonoids have the property of chelating bivalent metals. This ability requires a special pattern of substitution which can be supplied by an environment of bivalent negative charges. In the flavonoid structure this requirement is present in the o-hydroxyl substitution, which is often present in the B-ring, and may exist in the

769 simultaneous presence of 3-hydroxyl and 4-oxo groups under some conditions. The flavonoids present a characteristic UV/V (Ultraviolet/Visible) spectra defined by two bands. Band I (320-385 nm) relates to ring B absorption and Band II (250-285 nm) relates to ring A absorption [164]. Brown et al [165] studied the interaction between several flavonoids (quercetin, rutin, kaempferol, and luteolin) with Cu^^ ions finding spectral shifts for all the compounds. However, these shifts were greater and more intense in Band II (ring B) for those compounds with catechol structure (quercetin, rutin and luteolin) than in Band I (ring A) for those with a 4-oxo and 3-hydroxyl substitution (quercetin and OH

Low pH, M +2

Fig. (7). Chelating mechanism proposed by Samia et al [167] for a lone 3-hydroxyl substitution on flavylium cation (B) besides on catechol group (A). M^^: Bivalent metals.

kaempferol). This suggests that chelating properties of flavonoids seem to be based on catechol structure rather than on the 4-oxo group, with either

770

of its adjacent hydroxyl groups [165]. However, another class of flavonoids can chelate bivalent metals with a unique 3-hydroxyl substitution, the anthocyanins. Anthocyanins at low pH ranges exist mainly in the form of flavylium cations and because of the charge distribution, they are susceptible to nucleophilic attack on positions 2 and 4 [166]. According to this and based on spectral shift studies^ it has been postulated that the hydroxylation of an anthocyanin at these positions enhances its chelating ability [167], Fig. (7). Ascorbic acid preservation by flavonoids The antioxidant ability of flavonoids may affect several physiological processes in which redox reactions are involved. In the sixties, it was proposed that the flavonoids might act as protecting agents of ascorbic acid oxidation. Ascorbic acid is the main naturally occurring antioxidant in biological systems and therefore in foodstuffs. Besides this, its importance takes root in the fact that synergistically it may regenerate vitamin E, resulting in an enhancement of antioxidant activity in membranes and non-polar environments [168, 169, 170]. Moreover, it has been proposed that a recycling chain might exist between thiols, glutathione and lipoic acid, in which ascorbic acid and flavonoids might be involved. Thus, flavonoids may play a very important role in preserving antioxidant capacity in biological systems as in foodstuff by working on the recycling chain of antioxidants by protecting the ascorbic acid from its oxidation, for those flavonoids with a lower redox potential than the ascorbate radical/ascorbate couple [171], Fig. (8). Interaction between ascorbic acid and flavonoids has been demonstrated by Bor et al [172] when pulse radiolytic generation of azide radicals was used, which led to the formation of both ascorbate and flavonoid radicals. The same group also investigated the interactions between acorbate radicals and flavonoids when ascorbic acid was oxidized enzimatically with tyrosinase [173] while Cossin et al performed this research by enzimatically oxidizing the ascorbic acid with ascorbate oxidase [171]. This last group reported also that someflavonoids^with a triphenol structure such as myricetin, epicatechins and epicatechins gallates, compete with ascorbic acid for the ascorbate oxidase and therefore are preserving in a double way the ascorbic acid from oxidation

771 [171]. Another way to protect ascorbic acid from oxidation consists of the chelating mechanism. As has been discussed above, flavonoids with ortho-phenol structure have the ability to chelate bivalent metals^ which may undergo the oxidation of ascorbic acid, and this mechanism of protection for anthocyanins is being proposed by Sarma et al [167]. These former reports might explain what was proposed by Rusznyak and SzentGyorgyi more than six decades ago when they observed that the administration of lemon juice decreased the fragility and permeability of the capillary wall, where ascorbic acid alone had no protective effect, and led them to propose flavonoids as vitamin P.

VitE

a-lipoate and othei>>

VitE liliydrolipoate and others

Dehydroascorbate

Fig. (8). Involvement of flavonoids (Fla) in tlie ascorbate recycling pathw^ay. Proposed by Cossins etal. [171].

ABBREVIATIONS ABTS AhR AAPM BPV-4 CAMP CGMP DNA EGRFR LDL

= 2,2'-Azinobis (3-ethylbenzothiazoline-6-sulphonic acid = Aril Hydrocarbons Receptor. = 2, 2'-azobis (2-amidino propane) dihydrochloride. = Bovine Papilloma Virus type 4. = Cyclic Adenosin Monophosphate. = Cyclic Guanosine Monophosphate. = Deoxyribonucleic acid. = Epidermal Growth Factor Receptor. = Low Density Lipoproteins.

772

MUFA PCA PUFA RTK TEAC TPO UV UV/V VLDL

= Monounsaturated Fatty Acids. = Prostate Carcinoma. = Polyunsaturated Fatty Acids. = Receptor of Tyrosine Kinase. = Trolox equivalent antioxidant activity. = Thyroid Peroxidase. = Ultraviolet radiation. = Ultraviolet/Visible. = Very Low Density Lipoproteins.

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15] [16] [17] [18] [19]

Robards, K.; Antolovich, M. Analyst, 1997, 122, 11. Geissman, T.A. In The Chemistry of Flavonoid Compounds; Geissman, Ed.; McMillan Co: New York, 1962; pp. 1-5. Geissman, T.A.; Swain, T. Chem. andlnd., 1957, 5, 984. Shibuta, S.; Yamakazi, M. Pharm. Bull. Tokyo., 1957, 5, 501. Wollenber, E, In: The Flavonoids. Advances in research since 1986; Harbome, J.B., Ed; Chapman and Hall: London, 1993; pp. 259-335. Shimokoriyama, M. In The Chemistry of Flavonoid Compounds; Geissman, T., Ed, McMillan Co.: New York, 1962; pp. 286. Kefford, J.F.; Chandler, B.V. In: The Chemicals Constituents of Citrus Fruits. Advances in Food Research, Academic Press: New York, 1970. Castillo, J., Benavente-Garcia, O.; Del Rio, J.A. Plant. Physiol, 1992, 99, 67. Ortuno, A.; Reynaldo, I.; Fuster, M.D.; Botia, J.M.; Garcia-Puig, D.; Sabater, F.; Garcia-Lidon, A.; Porras, I. Scient. Hortic, 1997, 68, 231. Lebreton, P. J. Pharm. Chim., 1828, 14, 337. Manthey, J.A.; Grohmann, K. In Flavonoids in the Living System; Manthey, J.A,; Busling, W., Eds.; Plenium Press: New York, 1998; pp. 85-101. Hunt, G.; Baker, E.A Phytochemistry, 1980, 19, 1415. Paganga, G.; Miller, N.; Rice-Evans, C. Free. Rad Res., 1999, 30, 153. Peterson, J.; Dweyer, J. Nutrit. Res., 1998, 18, 1995. Marin, F.R.; Ortuno, A.; Benavente-Garcia, O.; Del Rio, J.A. Planta Med., 1998,64, 181. Dewick, P.M. In The Flavonoids. Advances in Research since 1986; Harborne, J.B., Ed.; Chapman and Hall: London, 1993, pp 125-209. Harbome, J.B.; Williams, C.A. In The Flavonoids; Harbome, J.B.; Mabry, T.J.; Mabry, H., Eds.; Chapman and Hall: London, 1975, pp. 376-441. Harbome, J.B.; Williams, C.A.In The Flavonoids. Advances in Research since 1976; Haibome, J.B., Ed., Chapman and Hall: London, 1982, pp. 115-178. Harbome, J.B.; Williams, C.A. In The Flavonoids. Advances in Research since 1980; Haibome, IB., Ed.; Chapman and Hall: London, 1988; pp. 303-328.

773

[20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47]

Christine, A.W.; Harbome, J.B. In The flavonoids. Advances in Research since 1986; Haibome, J.B., Ed.; Chapman and Hall: London, 1993; pp. 337-385. Mcintosh, C.A.; Latchinian, L.; Mansell, R.L. Arch. Biochem. Biophys., 1990, 282, 50. Mcintosh, C.A.; Mansell, R.L. Pytochem., 1991, 29,1533. Benavente-Garcia, 0.; Castillo, J.; Del Rio, J.A. J. Agric. Food Chem., 1993, 41, 1916. Heller, W.; Forkmann, G. In The Flavonoids. Advances in Research since J986; Harbome, J.B., Ed.; Chapman and Hall: London, 1993, pp 499-535 Kulevanova, S.; Stafilov, T.; Mihailo, R; Brkick, D. Croatian Pharm. Soc, 1997, 47,281. Hodisan, V. Clujul. Med., 1997,70, 280. Zaidi, F.; Voirin, B.; Jay, M.; Viricel, M.R Phytochem., 1998,48, 991. Herman, K. J. Food Technol., 1976,11, 433. Rice-Evans, C.A.; Miller, N.J. Biochem. Soc. Trans., 1996,24, 790. Tomas-Baiberan, FA,; Grayer-Barmeijer, R.J.; Gil, ML; Harbome, J.B. Phytochem., 19SH, 21, 2631. Del Rio, J.A,; Areas, M.C.; Benavente-Garcia, O.; Ortuiio, A. J. Agric. Food. Chem., 1998,46, 4423. Timberlake, C.F.; Bridle, P. In The Flavonoids; Harbome, J.B.; Mabry, T.J., Eds.; Chapman and Hall: London, 1975; pp 214-266. Strack, D.; Wray, V. In The Flavonoids. Advances in Research since 1986; Harbome, IB., Ed.; Chapman and Hall, London, 1993; pp 1-19. Hertog, M.G.L.; HoUman, P.C.H.; Katan; M.B. J. Agric. Food Chem., 1992, 40, 2379. Franke, A.A.; Custer, L.J.; Cema, CM.; Narala, K.K. J. Agric. Food Chem., 1994, 42,1905. Del Rio, J.A.; Marin, F.R.; Benavente-Garcia, O.; Martinez, M.; Garcia-Lidon, A.; Porras, L; Ortuiio, A. Lev. Agr, 1999, 347, 193. Hertog, M.G.L.; HoUman, P.C.H.; Katan, M.B.; Kromhout, D. Nutr. Cancer, 1993, 20, 21. Kuhnau, J. World Rev. Nutr. Diet., 1976, 24, 117. Rimm, E.B.; Katan, M.B.; Ascherio, A.; Stamper, M.J.; Willet, W.C. Ann. Intern. Med., 1996, 125, 384. De Vries, J.H.M.; Janssen, K.; Hollman, P.C.H.; Van Staveren, W.A.; Katan, MB. Cancer. Lett., 1991,114, 141. Guidarelli, A.; Clementi, F.; Sciorati, C; Cattabeni, F.; Cantoni, O. J. Pharmacol. Exp. Ther., 1997, 283,66. Cao, G.; Russell, R.M.; Lischner, N.; Prior, RL. J. Nutr., 1998,128, 2383. Serafini, M.; Maiani, G.; Ferro-Luzi, A. J. Nutr., 1998,128,1003. HoUman, P.C.H. Eur. J. Clin. Nutr., 1997, 51 (SI),: 66S. Sfakianos, J.; Coward, L.; Kirk, M.; Bames, S. J. Nutr., 1997,127, 1260. Watanabe, S.; Yamaguchi, M.; Sobue, T.; Takashashi, T.; Miura, T.; Arai, Y.; Mazur, W.; Wahala, K; Adlercreutz, H. J. Nutr., 1998,128,1710. Manach, C; Morand, C ; Crespy, V.; Demigne, C; Texier, O.; Regerat, F.; Remesy, C. FEBSLett., 1998,426, 331.

774

[48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66]

[67] [68] [69] [70] [71] [72] [73] [74]

Nielsen, S.E.; Breinhoult, V.; Justesen, U.; Comett, C; Dragsted, L.O. Xenobiotica, 1998, 28, 389. Donovan, J.L.; Bell, J.R.; Kasim-Karakas, S.; German, J.B.; Walzem, R.L.; Hansen, RJ.; Waterhouse, A.L. J. Nutr., 1999, 129,1662. Piskula, M.K.; Terao, J. J. Nutr., 1998,128, 1172. Hirano, T.; Gotoh, M.; Oka, K. Lif &/., 1994, 55,1061. Elangovan, V.; Sekar, R; Govindasamy, S. Cancer Lett., 1994, 87,107. Key, T.J.A.; Thorogood, M.; Appleby P.N.; Burr, M.L. BMJ., 1996, 313:775-779. Ames, B.N.; Gold, L.S. DrugMet., 1998,30, 201. Eraser, G.E. J. Clin. Nutr., 1999, 70, 532S. Lipkin, M.; Reddy, B.; Newmark, H.; Lamprecht, S.A. Annu. Rev. Nutr., 1999, 19, 545. Stapleton, A.E.; Walbot, V. Plant. Physiol., 1994,105, 881. Kooststra, M. Plant. Mol. Biol., 1994, 26, 771. Shimoni, K.; Masuda, S.; Furogori, M.; Esaki, S.; Kinae, N. Carcinog., 1994, 15, 2669. Noroozi, M.; Angerson, W.J.; Lean, M.E.J. Am. J. Clin. Nutr., 1998, 67, 1210. Heo, H.Y.; Lee, S.J.; Kwon, C.H.; Kin, S.W.; Shon, D.H.; Au, W.W. Mutat. Res., 1994,311,225. Larocca, L.M.; Piantelli, M.; Leone, G.; Sica, S.; Teofili, L.; Benedeti-Panici, P.; Scambia, G.; Mancuso, S.; Capelli, A.; Renelleti, P.O. Br. J. Hematol, 1990, 75, 489. Yoshida, M.; Sakai, T.; Hasokawa, N.; Marui, N.; Matsumoto, K.; Aoike, A. FEBS Lett.,199^,260,10. Scambia, G.; Ranelleti, P.O.; Benedeti-Panici, P.; Piantelli, M.; Bonanno, L.M.; De Vincenzo, R; Ferrandina, G.; Rumi, C; Larocca, L.M.; Mancuso, S. Br. J. Cancer, 1990, 62, 942. Peterson, G.; Barnes, G. Prostate, 1993,22, 335. Dimmock, J.R.; Kandepu, N.M.; Hetherington, M.; Quail, W.; Pugazhenthi, U.; Sudom, A.M.; Chamankhah, M.; Rose, P.; Pass, E.; Allen, T.M.; Halleran, S.; Szyldlowski, J.; Mutus, B.; Taimous, M.; Manavathu, E.K.; Myers, T.G.; De Clerq, E.; Balzarini, J. J. Med. Chem., 1998, 41, 1014. Kawaii, S., Tomono, Y.; Katase, E.; Ogawa, K.; Yano, M. J. Agric Food. Chem., 1999, 47, 128. Brown, A.; Jolly, P.; Wei, H. Carcinog., 1998, 19, 991. Kawada, N.; Seki, S.; Massayasu, I.; Kuroki, T. Hepatol., 1998, 5,1265. Kandaswami, C; Perkins, E.; Soloniuk, D.S.; Drzewiecki, G.; Middleton, E. Cancer Lett., 1991, 56, 147. Guthiere, N.; Gapor, A.; Chambers, A.F.; Carroll, K.K. Asia Pacific J. Clin. Nutr., 1997,6,41. Guthiere, N.; Gapor, A.; Chambers, AF.; Carroll, KK. J. Nutr., 1997,127, 544S. Tanaka, T.; Makita, H.; Kawabat, K; Mori, H.; Kakumoto, M.; Satoh, K.; Kara, A.; Sumida, T.; Tanaka, T.; Ogawa, H. Carcinog., 1997,18, 957. Middleton, E.; Kandaswami, C. In The Flavonoids Advances in Research since 1986; Harbome, J.B., Ed.; Chapman and Hall: London, 1993; pp 619-652.

775

[75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103]

Bok, S.H.; Lee, S.H.; Park, Y.B.; Bae, K.H.; Son, K.H.; Jeong, T.S.; Choi, M.S. J. M//r.,1999, 129, 1182. Manach, C; Regerat, F.; Teixer, O.; Agullo, G.; Demigne, C; Remesy, C. Nutr. Res., 1996,16, 517. Belt, J.A.; Thomas, J.A.; Buchsbaum, R.N.; Racker, E. Biochem., 1979,18, 3506. Shoshan, V.; McLennan, D.H. J. Biol. Chem., 1981, 256, 887. Beutler, J.A., Hamel, E.; Vlietinck, A.J.; Haemers, A.; Rajan, P.; Roitman, J.N.; Cardelin, J.H.; Boyd, M.R. J. Med Chem., 1998,41,2333. Ducki, S.; Forrest, R.; Hadfield, J.A.; Kendall, A.; Lawrence, N.J.; McGown, A.T.; Renninson, D. Bioorg. Med. Chem. Lett., 1998, 8, 1051. Groves, R.W.; Allen, M.H.; McDonald, D.M. J. Cutan. Pathol, 1992,19,66. Ahmad, N.; Gali, H.; Javed, S.; Agarwal, R. Biochem. Biophys. Res. Com., 1998, 284, 294. Zi, X.; Feyes, D. K.; Agarwal, R. Clin. Cancer Res., 1998, 4, 1055. Ahn, H.; Hadizadeh, K.R.; Seul, C; Ymi, Y.; Vetter, H.; Sachinidis, A. Mol. Biol. Cell, 1999,10,1093. Levitzki, A.; Gazit, A. Science, 1995, 267, 1782. Karp, J.E.; Chiarodo, A.; Brawley, O.; Kellof, G.F. Cancer Res., 1996, 56, 5547. Kellof, GJ.; Fay, IR. Steele, V.E.; Lubet, R.A.; Boone, C.W.; Crowell, J.A.; Sigman, C.C. Cancer Epidemiol. Biomark. Prev., 1996, 5,657. Zi, X.; Grasso, A.W.; Kung, H.; Agarwal, R. Cancer Res., 1998, 58, 1920. Edenharder, R., Xiaoyong, T.; Petersdorff, I.; Rolf, R. In Food Factors Cancer Prev. Meet. Hajime, O., Ed.; Springer: Tokyo, 1997; pp. 603-606. Kanazawa, K., Yamashita, T.; Ashida, H.; Danno, G. Biosc. Biotechnol. Biochem., 1998,62, 970. Appelt, L.C.; Reicks, M.M. J. Nutr., 1999,129,1820. Ciolino, H.P.; Wang, T.T.Y.; Yeh, C.C. Cancer Res., 1998, 58, 2754. Cova, D.; De Angelis, L.; Giavarini, F.; Palladini, G.; Perego, R Int J Clin Pharmacol Ther Toxicol, 1992, 30, 29. Henry, E.C;. Kende, A.S.; Rucci, G.; Totleben, M.J.; Willey, J.J.; Dertinger, S.D.; PoUenz, R.S.; Jones, J.P.; Gasiewicz, T.A. Mol. Pharmacol., 1999, 55, 716. Rusznyak, S.; Szent-Gyorgyi, A. Nature, 1936, 27, 138. Vogel, G. In Pharmacognosy and Phytochemistry, Wagner, H.; Horhammer, L., Eds.; Springer-Verlag: New York, 1971; pp. 370-386. Meyers, F.H., Jawetz, E., Goldfien, A. Review of Medical Pharmacology, Lange Medical Publications; Los Altos, California, 1972. Waldworth, A.N.; Faulds, D. Drugs, 1992, 44, 1013. Gerolaukos, G; Nicolaides, A.N. Angiol, 1994, 45, 549. Labrid, C. Angiol, 1994, 45, 524. Sher, E.; Codignola, A.; Biancardi, E.; Cova, D.; Clementi, F. Pharmacol Res., 1992,26:,395. Duarte, J.; Perez-Vizcaino, F.; Utrilla, P.; Jimenez, J.; Tamargo, J.; Zarzuelo, A. Gen. Pharmacol, 1993,24, 857. Andriambeloson, E.; Magnier, C; Haan-Archipoff, G.; Lobstein, A.; Anton, R.; Beretz, A.; Stoclet, J.C.; Andriantsitohaina, R. J. Nutr., 1998,128, 2324.

776

[104] [105] [106] [107] [108] [109] [110] [Ill] [112] [113] [114] [115] [116] [117]

[118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131]

Klagsbnm, M.; Folkman, J. Angiogenesis. In: In Peptide Growth Factors and their Receptorsn\ Spom, M.B.; Roberts, A.B., Eds.;. Springer-Verlag: Berlin, 1990; pp. 549-586. Denekamp, I Cancer Metast Rev., 1990, 9, 267. Fotsis, T.; Pepper, M., Adlercreutz, H.; Fleischmann, G.; Hase, T.; Monlesano, R.; Scheweigerer, L. Proc. Natl Acad Sci. USA., 1993, 90,2690. Fotsis, T.; Pepper, M.; Adlercreutz, H., Hase, T.; Montesano, R; Schweigerer, L. y.iVw/r., 1995, 125,7908. Fotsis, T.; Pepper, M. S.; Aktas, E.; Breit, S.; Raskn, S., Adlercientz, H.; Wahala, K.; Montesano, R,; Schweigerer, L. Cancer Res., 1997, 57,2916. Cunningham, B.D.M., Threadgill, M.D.; Groundwater, P.W., Dale, I.L.; Hickman, ].A.. Anticancer Drug. Des., 1992, 7, 365. Ferriola, P.C; Cody, V.; Middleton, E.J. Biochem. Pharmacol., 1989, 38, 1617. Huang, Y.T., Kuo, M.L.; Liu, J.Y. Huang, S.Y.; Lin, J.K. Eur. J. Cancer, 1996, 32, 146. Graziani, Y.; Erickson, E., Erkson, R.L. Eur. J. Biochem., 1983, 135, 583. Robbins, R.C. Int. J. Vit. Nutr. Res., 1974, 44, 203. Beret, A.; Cazenave, LP. In Plant Flavonoids in Biology and Medicine: Biochemical, Cellular, and Medicinal Properties. Cody, V.; Middleton, E.; Harbome, LB.; Beretz, A., Eds.; Alan R. Liss Inc.: New York, 1988; pp 187-200. Tzeng, S.H.; Ko, W.C, Ko, F.N.; Teng, CM. Thromb. Res., 1991, 64, 91. Osman, H.E.; Maalej, N.; Shanmuganayagam, D.; Folts, ID. J. Nutr., 1998, 128, 2307. Beret, A.; Anton, A.; Cazenave, J.P. In Plant Favonoids in Biology and Medicine: Biochemical, Pharmacological and Structure-Activity Relatinships. Eds.: Cody, V.; Middleton, E.; Haibome, LB.; Beretz, A., Eds.; Alan R. Liss Inc.: New York, 1986; pp 281-296. Ozaki, Y., Yatomi, Y.; Jinnai, Y., Kume, S. Biochem. Pharmacol., 1993, 46, 395. Gabor, M. In Plant Flavonoids in Biology and Medicine: Biochemical, Pharmacological, and Structure-Activity Relationships; Cody, V.; Middleton, E., Harbome, LB., Beretz, A., Eds.; Alan R.. Liss Inc: New York, 1986; pp. 471-480. Galati, E.M.; Monforte, M.T.; Kirjavainen, S.; Forestieri, M.; Trovato, A. // Farmaco, 1994,49, 709. Tordera, M., Ferrandiz, M.L., Alcaraz, M.L Z Naturforsch., 1994,49, 235. Jean ,T.; Bodinier, M.C. Angiol., 1994,45, 554. Matsuda, H. Yakugaku Zasshi, 1991,111,193. Kaul, T.N.; Middleton, E.; Ogra, PL. J. Med Virol., 1985,15, 71. McCarrison, R. Indian J. Med Res., 1933, 21,179. Nordisiek, F.W. Proc. Soc. Exp Biol. Med., 1962, 110, 417. Filiseti, T.M.; Lajolo, F.M. Arch. Latinoam. Nutr., 1981, 31, 287. Kimura, S.; Suwa, J.; Ito, B.; Sato, H. Gann., 1976, 67, 763. Divi, R.L.; Chang, H.C.; Doerge, D.R. Biochem. Pharmacol., 1997, 54, 1087. Ji, S.; Willis, G.M.; Frank, G.R.; Cornelius, S.G.; Spurlock, M.E. J. Nutr., 1999, 129,1291. Connolly, LA.; Morgan, I.M.; Jackson, M.E.; Campo, M.S. Oncogene, 1998, 16, 2739.

777

[132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144]

[145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158]

Zembayashi, M.; Lunt, D.K.; Smith, S.B. Meat Sci., 1999, 53, 221. Conquer, J.A.; Maiani, G.; Azzini, E; Raguzzini, A.; Holub, B.J. J. Nutr., 1998, 128, 593. Hodgson, J.H.; Puddey, I.E.; Beilin, L.J.; Mori, T.A.; Croft, K.D. J. Nutr., 1998, 128, 728. Pryor, W.A. In Lipid Peroxides in Biology and Medicine, Kiinio, Y., Ed.; Academic Press: New York, 1982; pp 1-22. Torel, J.; Cillard, J.; Cillard, P. Phytochem., 1986, 2, 383. Bors, W.; Hellers, W.; Michel, C; Saran, M. In Antioxidants in Therapy and Preventive Medicine; Emerit, Ed.; Plenium: Press. New York, 1990; Vol 7, pp. 165-170. Bors, W.; Hellers, W.; Michel, C; Saran, M. hi Methods in Enzimology; Packer, L.; Glazer, A.N., Eds.;. Academic Press: San Diego, 1990; Vol J86, pp 343-355. Benavente-Garcia, O.; Castillo, J.; Marin, F.R.; Ortimo, A.; Del Rio, J.A. J. Agric. FoodChem,, 1991,45: 4506. Hanasaki, Y., Ogawa, S.; Fukui, S. FreeRadic. Biol Med., 1994, 16, 845. Heilmann, J.; Merfort, L; Weiss, M. PlantaMed., 1995,61,435. Montesinos, M.C.; Ubeda, A.; Terencio, M.C.; Paya, M.; Alcaraz, M.J. Z Naturforsh., 1995, 50, 552. Morel, I.; Lescoat, G.; Cogrel, P.; Sergent, 0.; Pasdeloup, N.; Brissot, P.; Cillard, P.; Cillard, J. Biochem. Pharmacol, 1993, 45, 13. Van Acker, S,A,B,E,; Van den Berg, D.J.; Tromp, M.N.J.L.; Griffioen, D.H.; Van Bennekom, W.P.; Van der Vijgh, W.J.F.; Bast, A. Free Radical Biol Med., 1996, 20,331. Afanav'es, I.B.; Dorozhko, A.I.; Brodskii, A.V; Kostyuk, V.A.; Potapovich, A.I. Biochem. Pharmacol, 1989, 38, 1763. Ferralli, M.; Signorini, C; Ciccoh, L.; Comporti, M. Biochem. J., 1992, 285,295. Ferralh, M.; Signorini, C; Caciotti, B.; Sugherini, L.; Ciccoh, L.; Giachetti, D.; Comporti, M. FEBSLett., 1997, 416, 123. Arora, A.; Nair, M.G.; Strasburg, G.M. FreeRad Biol Med., 1998, 24, 1355. Sestih, P.; Guidarelh, M.; Dacha, M.; Cantoni, O. Free Rad Biol Med., 1998, 25, 196. Elba, D.; Riso, P.; Colombo, A.; Testolin, B. J. Nutr., 1999,129, 2130. Steinberg, D.S.; Parthasarathy, S.; Carew, I.E.; Khoo, J.C; Witzum, J.C. A^. Engl J. Med., 1989, 320, 915. Retsky, K.L.;, Freeman, M.W.; Frei, B. J. Biol Chem., 1993, 268, 1304. Vinson, J.A.; Dabbagh, Y.A.; Serry, M.M.; Jang, J. J. Agric. Food Chem., 1995, 43, 2800. Vinson, J.A.; Jang, J.; Dabbagh, Y.A.; Serry, M.M.; Cai, S. J. Agric. Food Chem., 1995, 43, 2798. Vinson, J.A.; Hu, S.; Jung, S.; Stawki, A.M. J. Agric. Food Chem., 1998,46, 1453. Fremont, L.; Gozzelino, M.T.; Franchi, M.P.; Linard, A. J. Nutr., 1998, 128, 1495. Bors, W., Michel, C; Saran, M. Metk Enzymol, 1994,234,420. Bors, W.; Heller, W.; Michel, C; Stettmaier, K. In Handbook of Antioxidants; Cadenas, E.; Packer, L., Eds.; Marcel Dekker: New York, 1996, pp 409-466.

778

[159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173]

Bors, W., Heller, W.; Michel, C; Saran, M. In Free Radicals and the Liver, Csomos, Q; Feher, J., Eds.; Springer: Beriin, 1992, pp. 77-95. Rice-Evans, C.A.; Miller, N.J.; Paganga, G. Tren. PlantSci., 1997, 2, 152. Cano, A.; Hernandez-Ruiz, J.; Garcia-Canovas, F.; Acosta, M.; Amao, M.B. Phytochem. Analysis, 1998, 9, 196. Amao, M.B.; Cano, A., Acosta, M. Recent Res. Devel. In Agric. Food Chem., 1998,2, 893. Chen, Z.Y.; Chan, P.T.; Ho, K.Y.; Fung, K.P.; Wang, J. Chem. Phys Lipids, 1996, 79, 157. Markham, K.R. Techniques of Flavonoid Identification, Acadecim Press Inc.: London, 1982. Brown, J.E.; Khodr, H.; Hider, R.C.; Rice-Evans, C.A. Biochem. J., 1998, 330, 1173. Brouillard, R. Phytochem., 1983,22,1311. Sarma, A.J.; Sreelakshimi, Y.; Sharma, R. Phytochem., 45,671. Constantinescu, A ; Han, D.; Packer, L. J. Biol. Chem., 1993, 268,10906. Sharma, M.K.; Buettner, G.R. Free Rad Biol. Med., 1993,14,649. Hamre, K.; Waagbo, R.; Berge, R.K.; Lie, O. Free Rad Biol. Med., 1997, 22, 137. Cossins, E.; Lee, R.; Packer, L. Biochem. Mol Biol. Int., 1998,45, 583. Bors, W.; Michel, C; Schikora, S. Free Rad Biol. Med., 1995,19, 45. Bors, W.; Michel, C; Stettmaier, K. J. Mag. Res. Anal, 1997,6.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

779

Natural Products as potential antiparasitic drugs OLIVER K A Y S E R \ A L B R E C H T F . K I D E R L E N ^ SMON L. CROFT^

^Freie Universitdt Berlin Institutfiir Pharmaziey Pharmazeutische Biotechnologie Kelchstrafie 31, 12169 Berlin, Germany Robert Koch-lnstitut Nordufer 20 13353 Berlin, Germany ^London School of Hygiene and Tropical Medicine Department of Infectious and Tropical Diseases Keppel Street London, WCIE 7HT, United Kingdom

ABSTRACT: Pharmaceutical research in natural products represents a major strategy for discovering and developing new drugs. The use of medicinal plants for the treatment of parasitic diseases is well known and documented since ancient times e.g. by the use of Cinchona succiruba (Rubiaceae) as an antimalarial. This chapter provides a comprehensive review of the latest results in the field of antiparasitic drug development from biologic sources (plants, bacteria, fungi and marine organisms) focussing on the treatment of protozoal infections (Plasmodium, Leishmania, Trypanosoma spp.). The status of validated in vitro and in vivo assays is reviewed, discussing their different features, problems and limitations. Because of the high number of natural products tested against the aforesaid protozoa in the last years, we limit the discussion to lignans, phenolics, terpenoids, and alkaloids as defined natural product classes. The review also covers essential research topics of recent publications on specific natural products (e.g. licochalcone A, benzyl- and naphthylisoquinoline alkaloids, and artemisinin) and gives an outlook to semisynthetic approaches of drugs already introduced in clinics or in clinical trial studies.

780

1. INTRODUCTION The fascination of natural products, mostly as used as a preparation from a plant with known medicinal properties, goes back to ancient times. The discovery of pure compounds as active principles in plants was first described at the beginning of the 19^^ century, and the art of exploiting natural products has become part of the molecular sciences. The discovery of quinine (1) from Cinchona succiruba (Rubiaceae) and its subsequent development as an antimalarial drug represented a milestone in the history of antiparasitic drugs from nature for the treatment of all parasitic diseases - not just infections caused by only Plasmodium, Leishmania and Trypanosoma spp.. Early studies on plant products were followed by an era of organic chemistry that led to the development of arsenical and antimonial (melarsoprol, sodium stibogluconate), diamidine (pentamidine) and nitroheterocyclic (metronidazole) antiprotozoal drugs. In the past decades natural products have attracted renewed interest, especially with bacteria and fungi as important sources of biologically active compounds. Recently marine organisms have also been recognized as attractive source of antiparasitic compounds, and it can be expected that in the future other living organisms (for example, insects and amphibians) will provide additional sources. It is, therefore, not surprising that one of the most rewarding frontiers in modem science is the study of the chemistry and biology of natural products. Discovering untapped natural sources of novel antiprotozoal compounds from nature remains a major challenge and a source of novelty in the era of combinatorial chemistry and genomics. Since plants contain a high variety of constituents it is often claimed that the use of a whole plant rather than one single purified product may be more effective therapeutically. Because of the limited space in this contribution we restrict ourselves to defined natural products that have been tested in standard in vitro and in vivo assays.

781 2.

PRESENT SITUATION AND CHEMOTHERAPY OF MALARIA, LEISHMANIASIS AND TRYPANOSOMIASIS

2.1 MALARIA It is estimated that there are 300-500 million cases of malaria annually with 1.75 to 2.5 million deaths. Malaria is a particularly important disease in sub-Saharan Africa, where about 90% of cases and deaths occur, but is also a serious public health problem in certain regions of southeast Asia and South America [1]. Human malaria is caused by four species of Plasmodium, P. falciparum, P. vivax, P. ovale and P. malariae, which are transmitted by female Anopheles mosquitoes. The majority of cases of malaria and deaths are caused by P. falciparum [1]. The life cycle, immunological defence mechanisms, and clinical development of malaria in humans is complex [2, 3]. The sporozoites that develop in the salivary glands of the female mosquito are inoculated into the human when the insect bites to acquire a bloodmeal. The sporozoites travel in the bloodstream to the liver where they invade the hepatocytes, differentiate and undergo asexual division (the exoerythrocytic cycle) and form a schizont (a multiple division form). Mature tissue schizonts release thousands of merozoites after 5-15 days (depending on species). The merozoites invade erythrocytes where they appear initially as a ring stage, followed by a growing trophozoite stage, which develops into a dividing asexual schizont stage. During the erythrocytic stage, which lasts from 48-72 hours depending upon species, the parasite develops in a parasitophorous vacuole surrounded by host cell membrane. The Plasmodium parasite adapts to life within the erythrocyte, depending mainly on glycolysis for energy and altering the erythrocyte membrane with transporters that enable increased uptake of hexoses, amino acids and lipid precursors. During the growth cycle up to 80% of host cell haemoglobin is ingested by and digested in the food vacuole of the Plasmodium trophozoite. The trophozoite divides (schizogony) and the erythrocyte lyses releasing more merozoites to invade further erythrocytes. At some point during the infection the intraerythrocytic stage develops to form sexual stages, male or female gametocytes, rather than merozoites. These sexual stages are taken up

782 in a blood meal by another Anopheles mosquito where fertilisation occurs and the life cycle is completed. Clinical malaria is characterised by periodic fever, which follows the lysis of infected erythrocytes, and caused mainly by the induction of cytokines interleukin-1 and tumour necrosis factor. P. falciparum infection can have serious effects, for example anaemia, cerebral complications (from coma to convulsions), hypoglycaemia and glomerulonephritis. The disease is most serious in the non-immune, including children, pregnant women and tourists. Humans in endemic areas, who have survived an attack of malaria, are semi-immune and disease can be characterised by headache and mild fever. Infection by the other species of Plasmodium is normally self-limiting although relapses may occur, particularly in P. vivax infections. The species of parasite and the age and immune status of the patient are important in considerations of treatment and interpretation of the effects of all medicines. The chemotherapy and prophylaxis of malaria have been undermined by the development of worldwide resistance of P. falciparum to the 4-aminoquinoline chloroquine, first observed in the 1960s, as well as resistance to the antifolates pyrimethamine and cycloguanil. Resistance to quinine and other more recently developed drugs, for example mefloquine, have also been reported [4, 5]. The search for alternative antimalarials is one of the main themes of this chapter. The re-emergence of malaria as a public health problem is due mainly to the development of resistance of P. falciparum to cheap highly effective drugs like chloroquine and pyrimethamine. As a consequence of this problem over 300,000 compounds were tested for antimalarial activity by the Walter Reed Army Institute of Research, USA between 1965 and 1986. The 4-quinolinemethanol mefloquine and the 9-phenanthrenemethanol halofantrine emerged from this programme. Mefloquine (see also section 6.2) has been registered for little over 10 years, but there is already resistance in South East Asia, concern over cross-resistance with quinine and controversy over toxic side effects. Chloroquine is still used in some low resistance areas in Africa and South America and quinine is used for the treatment of cerebral malaria (see section 6.2). The most important recent discovery for the therapy of P. falciparum malaria has been the identification of the sesquiterene lactone artemisinin (qinghaosu) from

783 Artemisia annua (Asteraceae) (see section 6.1). Artemisinin and its' derivatives, for example artemether and artesunate, are rapid acting antimalarials, effective against multidrug resistant P. falciparum, that have been used to treat over 3 million cases in South East Asia [6]. Another new drug, the hydroxynaphthoquinone atovaquone, identified as an antimalarial in the eariy 1980s [7], has proved to be highly effective in clinical trials but has to be used in combination with proguanil (as Malarone®) to prevent the development of resistance. The use of combinations to combat the development of resistance is a current strategy as demonstrated in the clinical use of other combinations for example mefloquine - artemisinin, co-artemether (lumefantrine/ benflumetol - artemether) and Lapdap® (chlorproguanil - dapsone) [8]. A new 8-aminoquinoline, tafenoquine, is on clinical trial as a potential replacement for primaquine to treat P. vivax malaria; it also holds promise as a prophylactic against P. falciparum [9].

2.2 LEISHMANIASIS Protozoa of the genus Leishmania are obligate intracellular parasites of mononuclear phagocytes. Leishmaniasis is a spectral disease, depending on the Leishmania species involved and genetic potential and acute predisposition the hosts defense system. It ranges from selfhealing ulcers (cutaneous leishmaniasis, CL) to progressive nasopharyngal infections (mucocutaneous leishmaniasis, MCL) to disseminating visceral leishmaniasis (VL). While CL poses essentially cosmetic problems, and MCL leads to painful disfiguration, social stigmatization and often highly severe secondary infections, VL is generally lethal if left untreated. Leishmaniasis occurs from tropical to Mediterranean regions where the parasite is transmitted by female sandflies of the genus Phlebotomus in the Old World and Lutzomyia in the New World [10]. In the insect gut as well as in tissue culture media, the parasite exists as extracellular, elongated, flagellated promastigotes. Promastigotes are injected into the skin during a blood meal and rapidly taken up by mononuclear phagocytes where they reside in the parasitophorous vacuole. In contrast to other intracellular pathogens, for example Toxoplasma, Leishmania do not inhibit fusion

784 of lysosomes with the parasitophorous vacuole. Within the phagolysosome the parasites transform into and multiply as ovoid amastigotes, which are specially adapted to the elevated temperatures within the mammalian host and the hostile environment within these potent effector cells [11]. Massive amastigote multiplication leads to host cell disruption and release of parasites to infect freshly recruited host cells. While "resting" macrophages support parasite multiplication and thus spread of infection, these cells can also be activated by elements of the natural and the specific cellular immune response to kill intracellular parasites leading to cure. The prime signal for macrophage activation is the cytokine interferon EFN-y, a glycopeptide released by Natural Killer (NK) cells and T lymphocytes. In the murine model for CL caused by L. major it was shown that distinct populations T cell populations exist with antagonizing effects. While T helper 1-type cells secrete IFN-y and promote cure, Th2-type cells secrete interleukin IL-10, inhibiting macrophage activation and exacerbating disease. The exact factors promoting each type of immune response remain obscure but are in part genetically determined. According to the World Health Organization (WHO) there are 2 million new cases of leishmaniasis per year. With the advent of ADDS, which currently affects over 30 million people in geographical regions largely overlapping with leishmaniasis, the nature of the disease has changed. In Mediterranean countries, where infant VL is endemic, adult VL is now considered a genuine AIDS-related opportunistic disease largely due to reactivation of latent infections by immunosuppression [12]. Although the three disease complexes leishmaniasis, human African trypanosomiasis (HAT, sleeping sickness) and South American trypanosomiasis (Chagas disease) are caused by closely related trypanosomatid parasites, the diseases are treated with different drugs and the parasites themselves have varying sensitivities to many compounds [13, 14]. The recommended drugs for both visceral and cutaneous leishmaniasis are the pentavalent antimonials sodium stibogluconate (Pentostam) and meglumine antimoniate (Glucantime). Both drugs have been used for over 50 years, require long courses of parenteral administration, and have toxic side effects. In addition there has been a dramatic increase in the number of cases

785 of VL in northeastern India that do not respond to antimonials. Alternative treatments for VL include the polyene antibiotic amphotericin B [15] that has highly effective less toxic lipid formulations [9]. A parenteral formulation of the aminoglycoside paromomycin (aminosidine), and the orally available alkylphospholipid miltefosine [16] are also potential treatments for VL. Interest in a new treatment for CL has focussed on different topical formulations of paromomycin. The increasingly observed partial resistance to antimonials has been overcome by higher dose regimens but in general therapy of MCL and VL is becoming increasingly problematic. Experimental studies and treatment of AIDS patients has revealed that successful treatment with some drugs requires the active participation of the immune system [12]. Antimonial drugs have been successfully combined with EFN-y, although costs of such treatmenr render it useless for third world countries.

2.3 AFRICAN TRYPANOSOMIASIS Flagellated protozoa of the genus Trypanosoma infect humans (sleeping sickness; T. brucei rhodesiense, T. b. gambiense) and domestic animals (nagana; T, b. brucei, T. congolense, T. vivax) causing major health and commercial problems in subSaharan Africa. In the 1960s African trypanosomiases was under control, mainly due to eradication of its insect vector, the tsetse fly. According to the WHO, sleeping sickness is again endemic in 36 African countries with over 250 thousand newly infected cases per year [17]. Following the bite of an infected tsetse {Glossina spp.) parasites multiply in the skin for one to three weeks before invading the haemolymphatic system. Early symptoms include high fever, swelling of lymph nodes (neck), hepatosplenomegaly, oedema and diarrhoea. Generalized inflammation of blood and lymph vessels may lead to myocarditis and encephalitis and often to death due to heart failure. Days or many weeks later, parasites invade the central nervous system causing the typical symptoms of sleeping disease: failures in speech and coordination of movement, epileptic episodes, somnolence, apathy, cachexy, and finally - if left untreated - coma and death [18].

786 Most Trypanosoma spp. are found only in wild and domestic animals. While for some species (e.g. antelopes, goats, pigs) infection with T. brucei produces no apparent symptoms, others (e.g. dogs, horses, cattle) often succumb to disease. Cattle show reduction in weight gain, milk yield, reproduction, and general performance. Laboratory strain rats and mice can also be infected, providing useful experimental models. In West African sleeping sickness, especially during epidemics, there is direct human-to-human transmission of T, b. gambiense, whereas the other species that infects humans, T. b. rhodesiense, is mainly transmitted between reservoir mammals. T, brucei spp. are extracellular parasites living in blood and lymph as elongated trypomastigotes. Here, they obtain their energy by glycolysis, whereas in their insect vector they switch to oxidative metabolism using mainly proline. In contrast to intracellular parasites such as Trypanosoma cruzi and Leishmania sp., T brucei are vulnerable for humoral defense mechanisms. Complement-activating antibodies may bind to their surface, facilitating recognition, phagocytosis and destruction by the monocytic phagocyte system. Trypanosoma brucei evade eradication by antigenic variation, sequentially producing new clones differing in their surface glycoproteins which are not susceptible to the prevailing antibody population. This strategy of antigen variation effectively neutralizes the hosts' immune defense mechanisms and has confounded all efforts of vaccine development. Furthermore, T. brucei organisms have been found to activate immunosuppressive macrophage functions. Obviously, any successful strategy to combat the African trypanosomiases must be two-pronged: controlling the insect vector and combating the parasites by prophylactic and therapeutic measures. Chemotherapy of nagana has been reliant for over 40 years on diminazene (berenil), isometamidium and homidium. Due to the intensive use and structural similarities of these drugs, trypanosomes have developed multiple drug resistance in Ethiopia, Kenya, Somalia, and many other African countries. Drugs for the treatment of human African trypanosomiasis are also inadequate. The diamidine pentamidine and the sulphonated naphthylamine suramin have been used for over 50 years, require parenteral administration and are only effective against the early haemolymphatic stage of the disease. The only drug currently available for the treatment of the late stage CNS infection is the

787

trivalent arsenical melarsoprol. Melarsoprol requires parenteral administration, has unacceptable side effects including 5 % mortality due to encephalopathy, and there are an increasing number of patients who are no longer responding to treatment. Eflomithine, an inhibitor of polyamine biosynthesis, proved to be an effective treatment in trials in the late 1980s and was registered in 1990 for the treatment of late stage infections caused by T. b. gambiense (but not T. b. rhodesiense which is refractory). Limited supplies of the eflomithine are available as the drug is no longer manufactured [21]. 2.4 SOUTH AMERICAN TRYPANOSOMIASIS South American trypanosomiasis (or Chagas disease) is caused by Trypanosoma cruzi and is widespread thoughout the subcontinent with an estimated 18 million people infected. The parasite is transmitted by triatomine insects, not directly by by contamination: the insect defaecates whilst taking a blood meal and parasites in the faeces can invade wounds, the eye or mucosal tissues. In mammals the parasite exists in two forms: the extracellular trypomastigote that is an invasive haemolymphatic form and is non-dividing, and the amastigote that divides in the cytoplasm of cells. 7. cruzi trypomastigotes will invade many cell types, in particular macrophages, muscle cells and nerve cells. Chagas disease is characterised by three stages: (a) an early symptomatic acute phase during which the infection spreads throughout the body and up to 30% of deaths occur, (b) an asymptomatic indeterminate phase that may last for many years, and (c) a symptomatic chronic phase when immunopathological reactions to low numbers of parasites in heart and nerve cells cause illness and death. Some cases of transmission of Chagas disease in urban areas have resulted from blood transfusion. Two drugs have been available for the treatment of Chagas disease since the early 1970s: the nitrofuran derivative nifurtimox (now no longer manufactured) and the 2-nitroimidiazole derivative benznidazole. Both drugs are only effective in the acute phase of the disease, have variable efficacy against different strains of T, cruzi, require long courses of (oral) treatment, and have high levels of side effects. Few drugs have proved effective against all stages of the

788 disease. The antifungal sterol biosynthesis inhibitors have shown the most promising activity against T. cruzi inexperimental models. Recently Urbina and colleagues [20] have identified several antifungal triazoles that are active against both acute and chronic T. cruzi infections in rodent models. 2.5 CHEMOTHERAPY OF OTHER PROTOZOAN DISEASES For several other protozoan diseases there is adequate chemotherapy: the 5-nitroimidazoles (for example, metronidazole) for the treatment of amoebiasis, giardiasis and trichomoniasis, the hydroxynaphthoquinone bupravaquone for theileriosis in cattle and other ungulates, and the polyene ionophores (for example monensin, lasalocid, narasin and salinomycin) for the prophylaxis of avian coccidiosis. However, improved therapies are required for some opportunistic parasites that cause disease in immunocompromised humans. Paromomycin and nitazoxanide have some effect in the treatment of cryptosporidiosis and albendazole appears to be effective for microsporidiosis caused by Encephalitizoon intestinalis.

3. IN VITRO AND IN VIVO ASSAYS FOR DETERMINATION OF DRUG ACTIVITY 3.1 ANTIPLASMODIAL ACTIVITY Prior to 1976, when the asexual cycle of P. falciparum was successfully cultured in human erythrocytes, antimalarial drug screening was dependent upon in vivo avian and rodent models, using non-human Plasmodium species. Since 1976 P. falciparum has been used extensively in in vitro screens; techniques to culture the other three human Plasmodium species have been less successful. A semiautomated microdilution assay, in which parasite viability and drug activity is related to the uptake of a nucleic acid precursor [^H]hypoxanthine, was described by [22] and is still widely used in modified 96-well format versions [23]; 384 well format assays are also now being used. The sensitivity (and resistance) of a wide range of P. falciparum strains to the standard antimalarial drugs has been reported using this assay. The technique has some limitations but

789 modifications using synchronous cultures and altering the medium have enabled studies on differential drug effects on trophozoite stages and schizonts [24] or improved sensitivity to antifolates [25]. A colorimetric assay based on the reduction of nitroblue tetrazolium to formazan by lactate dehydrogenase (a glycolytic enzyme essential in Plasmodium) has been used to test drugs against P. falciparum in vitro [26], and is simpler and cheaper to use but less sensitive than the [^H]hypoxanthine assay. Drug sensitivity testing for P. vivax and P. ovale remains a problem as continuous culure of these species is difficult due to nutritional and host cell requirements; only short term cultures have been used in drug tests [27]. In vitro culture of the exoerythrocytic liver stages of malaria infection have been achieved using hepatocytes and hepatomas to culture P. berghei, P. vivax and P. falciparum [28] but these models have not been used widely in drug evaluation studies. None of the four Plasmodium species that cause disease in humans can infect rodents or other animals used in in vivo screens. For the past 40 years in vivo screens have based on rodent malaria models, in particular those using P. berghei, P. yoelii and P. chabaudi infections in mice. Rodent malaria has proved to be an essential part of the process of drug development for detecting blood schizonticidal, tissue schizonticidal and repository activity. The activity of standard antimalarials against a number of P. berghei and P. yoelii strains has been well characterised [29]. The development of drug-resistant strains of these parasites has added to their use in the identification of novel drugs and drug combinations. However, there are differences between the biology of these rodent species and that of P. falciparum and the pharmacokinetic properties of a drug in mouse and humans can be different. Candidate antimalarial drugs are therefore often tested in primate models of P. falciparum infection using Aotus and Saimiri monkeys. 3.2 ANTILEISHMANIAL ACTIVITY In their mammalian hosts, Leishmania parasites exist primarily as amastigotes within phagolysosomes of macrophages. Extra- and intracellular promastigotes occur only during a few hours after infection and extracellular amastigotes appearing between disruption

790 of one host cells and uptake by the next. Antileishmanials must be also be accumulated by infected macrophages enter phagolysosomes and be active under the specific conditions within this compartment (e.g. low pH) and either kill amastigotes or reduce viability. The relative ease of mass cultivation of most Leishmania isolates in the laboratory and the availability of relevant animal models for human CL and VL facilitate adequate screening for novel antileishmanials at different levels of complexity. Toxicity assays for extracellular promastigote are easy to perform and have been used in drug screening. However, they have limited as the intracellular amastigote has different biochemical and molecular properties. Promastigote Leishmania are cultured in a variety of liquid or two-phase semi-synthetic or fully synthetic media generally at 23-27 °C and direct cytotoxic effects can be assessed microscopically as reduced motility, altered morphology (rounded and bloated) or reduced numbers. Growth inhibition assays involving incorporation of radioactive nucleotides (e.g. [^H]-thymidine) by proliferating cells or metabolization of chromophores (e.g. MTT) [30]. Assessments of effects on the intracellular survival of amastigotes is more complex. Primary macrophage cultures or monocytic cell lines are parasitized in vitro with promastigote Leishmania cultures or freshly isolated amastigotes. The former should be cultured at 37 °C at least over night to allow infection and intracellular transformation to amastigotes [31]. After exposure to test compounds, the cultures ells are stained and the average number of intracellular parasites/host cell determined in comparison to untreated controls. For mass screening, parasitized macrophages can be seeded into microtiter plates. After exposure to test substances, the host cells can be selectively lysed releasing intracellular parasites. The relative numbers of surviving transformed parasites can then be determined by the radiometric or colorimetric methods described above [31]. Indirect antileishmanial activity through activation of host macrophages needs to be assessed separately in adequate toxicological and inmiunological assays. For many Leishmania spp. that cause visceral or cutaneous leishmaniasis, rodents are natural hosts and provide excellent laboratory models. Furthermore, genetically defined mouse strains vary according to their spontaneous healing/non-healing capacity and immune response patterns thus providing means for specific experimental design. For CL, highly susceptible, non-curing BALB/c

791

mice can be inoculated subcutaneously with L. major into a footpad or the shaven rump. Progression of disease is monitored by measuring the diameter of the developing lesion. Antileishmanial drugs can be given orally, as topical ointments, or injected into the lesion. As tissue response does not necessarily correlate with parasite numbers, impression smears of tissue samples taken from the periphery of the ulcer should be stained and evaluated microscopically. Alternatively, viable parasite numbers can be estimated from limiting dilution cultures of tissue homogenates. For VL, hamsters or mice should be inoculated intravenously or intracardically with L. donovani. Disease progression can be monitored by the extent of cachexia or development of ascites. Visceral leishmaniasis affects all internal organs, and parasite counts are performed at least on spleen, liver and bone marrow, again using impression smear or limiting dilution culture methods. Sensitive molecular techniques such RT-PCR - while providing little advantage when quantifying an acute infection - are useful for detection of latent infections after parasitological cure.

3.3

ANTITRYPANOSOMAL BRUCEI)

ACTIVITY (TRYPANOSOMA

T. b. bruceiy T.b. rhodesiense and T.b. gambiense extracellular bloodstream form trypomastigotes can be grown axenically in supplemented standard media. Trypanocidal activity may be tested by culturing a constant number of parasites (lOV ml) in serial dilutions of substances for 24-72 hours in microtitre plates at 37 "^C and 5 % CO2. For reference, standard drugs such as melarsoprol (Arsobal®), pentamidine (Pentacarinat®) or suramin (Germanin®) should be included. The % surviving trypomastigotes indrug treated cultures can be estimated colorimetrically using a p-nitrophenyl phosphate as substrate for acid phosphatase [32], the MTT assay (described in section 3.2) or a fluorochrome such as Alamar blue [33]. Plotting % growth inhibition {= [1 - Ssample - 5max kill) / (5max growth " Smax kill)] X 1 0 0 } againSt drUg

concentration, dose-response curves can be generated and EC50 and EC90 values calculated for better comparison. It should be noted, that for some drugs, the two major variants T. b. gambiense and T. b.

792

rhodesiense exhibit in vitro differences in drug sensitivity and some compounds, for example eflomithine are poorly active in vitro . Most animal models are restricted to testing compounds against T. b. brucei or T. b. rhodesiense. In vivo tests against. T. b. gambiense are restricted to Mastomys rats or scid mice. However, some strains of T. brucei can establish chronic CNS infections in mice thus providing models for late-stage sleeping sickness. An effective new drug for sleeping sickness must also be capable to cross the blood/ brain barrier. Monkey models, available for testing of candidate drugs, can also provide data on the penetration of drugs into the CNS.

3.4

ANTITRYPANOSOMAL CRUZI)

ACTIVITY (TRYPANOSOMA

The dividing T. cruzi amastigote can be cultured in vitro in a variety of host cells, most commonly macrophages, fibroblasts and myoblasts. Trypomastigotes used to infect these cells can also be cultured in vitro. Two points about the design and interpretation of the assay: (a) it is normally limited to 3-4 days, as intracellular amastigotes transform to trypomastigotes after this time, and (b) as drug activity is measured by determining the number of amastigotes/host cell in treated and untreated cultures, the effects of compounds on the division rate of host cells as well as amastigotes has to be considered. The standard drugs nifurtimox or benznidazole should be included for comparison, but as mentioned above their activity is variable depending upon the strain of T. cruzi used. This also underlines the importance of testing lead compounds against a number of strains of this parasite. Recently the in vitro amastigote assays have been automated through the use of beta-galactosidase transfected T. cruzi Tulahuen strain [19]. Active compounds can be tested further against the extracellular trypomastigotes over 24 hours at 4°C as a new drug for sterilizing blood transfusions is also required. Inbred mouse strains offer the most useful in vivo models, either for simple suppressive tests or more complex curative tests. Mice are infected by trypomastigotes and treatment starts when the parasitaemia is detectable in tail blood. To determine cure sensitive techniques that can detect low numbers of parasites (haemoculture, serology, immunohistopathology, PCR) have to be performed on

793

blood, muscle and other tissues [20]. Again it is important to include standard drugs and a number of strains of T. cruzi in the process of identifying a lead compound. Chronic infections can be established by infecting with a low number of parasites (10"^) and monitoring the infection by sensitive techniques during and after treatment [20]. 4.

ANTIPROTOZOAL DRUGS FROM NATURE

41 LIPIDS AND RELATED ALIPHATICS 4.1.1 ORGANIC ACIDS, LIPIDS AND ACETOGENINS Long chain hydrocarbons and fatty acids are best known as constituents of waxes and lipophilic compounds. Some representatives of this natural product group show high antiprotozoal activity but mostly combined with a high levels of toxicity to mammalian cells. One example is trans-dicomiic acid (2) that was used in combination with sodium stibogluconate, allopurinol, or pentamidine for experimental visceral leishmaniasis to determine synergistic effects [34]. When these three drugs (50, 15, 8 mM/kg/day, respectively) were used with fmn^-aconitic acid (5 mM/kg/day) the parasite load in BALB/c mice was inhibited by 100, 88, and 100%, respectively. At tested concentration rmn^-aconitic acid itself showed an inhibition of 59.2 %. Four acetogenins from Rollinia emarginta (Annonaceae) were identified with antiprotozoal activity by bioassay-guided screening. Against different Leishmania and T. cruzi strains at 250 fig/mL inhibitory activity up to 89% and 67% was reported for rolliniastatin-1 (3) and squamocin (4) [35]. The leishmanicical activity was related to the number of hydroxy groups on these acetogenins. Maximum activity activity was observed in compounds that possessed three hydroxy groups, for example rolliniastatin-1 and squamocin, while activity was low in acetogenins having four or more hydroxy groups, e. g. sylvaticin and rollidecin.

794

y-WCOOH HOOC COOH

(2) trans-siconiiic acid

H,C»""

(3)rolliniastatin-l QH _

OH

OH CH,

H,C»»*'

(4) squamocin

4.1.2 POLYENES This class of antibiotics is well known because amphotericin B (5) is used as second line drug for the therapy of visceral and mucocutaneous leishmaniasis. Different polyene analogues, related to amphotericin B, also inhibit parasite growth. Polyene antibiotics can be divided into a non-aromatic group, which includes amphotericin B, and an aromatic group, which includes hamycins A and B another potent antiprotozoal agents. Hamycin (6), a polyene antibiotic, now in extensive use in the treatment of candidiasis and otomycosis, is found to be remarkably effective in killing Leishmania donovani promastigotes in a liquid medium at a concentration of 0.1 /xg/mL. Glucose stimulated respiration and the uptake of 2-deoxy-D[U-^^C]-glucose was inhibited in cells treated with the drug at a growth inhibitory concentration. The primary site of action of hamycin on L. donovani promastigote cells appears to be similar to amphotericin B, binding to membrane sterols, disrupting membrane stability with the loss of the permeability barrier to small metabolites. The lower minimum inhibitory concentration (MIC) of hamycin compared to other established drugs warrants further study in the

795 context of increasing reports of clinical resistance to pentavalent antimonials [36]. Despite the high activity of this class of compound there has been no discovery of new potent but less toxic polyene antibiotics in the past decade. Most interest has been focussed on the formulation of amphotericin B in colloidal drug carriers like liposomes, emulsions, micro-, and nanoparticles to improve bioavailability and reduce toxicity (AmBisome®, Ambicer, Amphocil®).

COOH

(5) amphotericin B

^NHCH

COOH

(6) Hamycin

4.1.3 AJOENES Some simple fatty acid or aliphatics show antiprotozoal activity. Ajoenes metabolites (7) are a good example with the naturally

796 occuring allicin that has proved to be active against rodent malaria and Trypanosoma cruzi. These sulfur containing aliphatics, initially isolated from garlic {Allium sativum, Liliaceae), showed significant suppression of Trypanosoma parasitemia in vivo with daily doses of 50 mg/kg over 12 days [37]. Gallwitz et al. [38] identified ajoenes also as potential drugs effecting thiol metabolism by acting as a covalent inhibitor as well as a substrate of human glutathione reductase (GR) and secondly of the Trypanosoma cruzi trypanothione reductase (TR). The interactions between the flavoenzymes and ajoene lead to increased oxidative stress of the respective cell. The antiparasitic and cytostatic actions of ajoene may at least in part be due to the multiple effects on key enzymes of antioxidant thiol metabolism. Urbina et al. also demonstrated an effect on the phospholipoid biosynthesis of Trypanosoma cruzi with an alteration of the lipid composition of parasites from phosphatidylcholine to phosphatidylathanolamine [39]. Ajoenes also inhibited de novo synthesis of neutral lipids and sterols in T cruzi epimastigotes, but these effects are not sufficient to explain antiproliferative effects of the drug.

ft (7) ajoene

4.2

PHENOLS

4.2.1

SIMPLE PHENOLS, COUMARINS

PHENOLIC

ACIDS

AND

Simple phenols that are widely distributed in plants have been tested for their ability to inhibit parasite growth. For example, gallic acid (8) and its derivatives inhibit the proliferation of Trypanosoma cruzi trypomastigotes in vitro with an EC50 value of 15.6 jLtg/mL [40], Higher activities were observed for the gallic acid esters ethyl-gallate and n-propyl-gallate which had EC50 values of 2.28 and 1.47 /ig/mL, possibly due to increased lipophilicity. No in vivo data has been published. It seems unrealistic that such compounds, which form part of the daily diet, will have significant effects. The mechanism of

797 action remain obscure and the authors suggest that the formation of reactive oxygen species might be involved in the galhc acid induced apoptotic cell death [41]. Interestingly, Kayser et al. could not demonstrate any direct toxic effect of gallic acid and related gallotannins on L. donovani in infected macrophages [42], Further studies should be conducted to clarify immunstimulatory activity of gallic acid [43]. Ascofuranone (9), an isoprenoid prenylphenol antibiotic, derived from the fungus Ascochyta visiae, specifically inhibits mitochondrial glycerol-3-phosphate (G-3-P)-dependent electron transport in T. b. brucei [44]. Ascofuranone strongly inhibited both glucose-dependent cellular respiration and glycerol-3-phosphate-dependent mitochondrial oxygen consumption of T. b. brucei bloodstream form trypomastigotes. This inhibition was suggested to be due to inhibition of the mitochondrial electron-transport system, composed of glycerol3-phosphate dehydrogenase and a plant-like alternative oxidase. Ascofuranone noncompetitively inhibited the reduced coenzyme Q l dependent O2 uptake of the mitochondrion with respect to ubiquinol (Ki = 2.38 nM). The site of action was deduced to be the ubiquinone redox machinery that links the two enzyme activities. Further, ascofuranone in combination with glycerol completely blocked energy production, and potently inhibited the in vitro growth of the parasite. Other simple phenols include the hydroquinone derivatives miconidin (10) and espintanol (11), formed from its biosynthesis of a monoterpene to a phenolic, and pholoroglucinol derivatives from Hypericum calycinum (Hypericaceae). Quantitative data are not available for miconidin, but espintanol exhibited an IC90 in the 25-100 jLtg/mL range against twenty different T. cruzi strains [45], and a prenylated phloroglucinol derivative (12), inhibited P. falciparum growth in vitro with an EC50 of 0.88 /xg/mL [46]. The mode of action of these semiquinones is unclear. It is attractive to speculate that the sensitivity of Trypanosoma and Plasmodium is due to oxidative stress resulting from the metabolic oxidation of semiquinone radicals or benzoquinones. The presence of antiplasmodial phloroglucinol derivatives has already been mentioned. Sarothalen B was found to be active in vivo in the 1950s [47] and the biogenetically related phloroglucinol derivative (12), which lacks the cyclohexadienone moiety of sarothalen B, is active in in vitro assays, indicating, that it conserves essential chemical features for this antiprotozoal activity.

798

Oketch-Rabah et al. reported the antiprotozoal activity of 2"epicycloisobrachycoumarinone epoxide (13) and its stereoisomer that has been isolated from Vemonia brachycalyx (Asteraceae) [48]. Both stereoisomers show similar in vitro activities against chloroquinesensitive (CQ-S) and chloroquine-resistant (CQ-R) strains for P. falciparum as well as L. major promastigotes with EC50 values of 0.11/ig/mL and 0.15 /xg/mL for Plasmodium and 37.1 jLtg/mL and 39.2 /xg/mL for L. major, respectively.

v (8) gallic acid

(9) ascofurane H,

(10) miconidin

HsC^As^OM .OMe

OMe CH,

(11) espintanol

(12) phloroglucinol-derivate

(13) 2'-epicycloisobrachy-coumarinone epoxide

799 4.2.2 LIGNANS Lignans are a potent group of natural products with many toxic side effects, best represented by podophyllotoxin derivatives and the antineoplastic drug etoposide. Despite their known biological activities few lignans have been tested against parasitic protozoa. Lopes et al. demonstrated for the potential of the tetrahydrofuran lignans grandisin (14) and veraguensin (15) to prevent the transmission of Chagas disease by blood transfusion [49]. The activity of these terahydrofuran lignans (62 % and 87 % growth inhibition at 2.5 /xg/mL, respectively) was forty times higher than that of the reference drug gentian violet. Lignans isolated from the hexane extract of the leaves of Zanthoxyllum naranjillo (Rutaceae) were tested in both in vitro and in vivo against two strains of Trypanosoma cruzi [50]. The compound (-)-methylpluviatolide (16) was highly effective for chemoprevention in the in vitro assay and healthy animals injected with the tested samples did not develop infection. Moreover, (-)-methylpluviatolide was also highly active against the bloodstream forms of two strains of T. cruzi in an in vivo assay [48].

OCH, 0CH3

HXO H3C0

0CH3

(14) grandisin

OCH, HXO

OCH,

(15) veraguensin

(16) (-)-methylpluviatolide

800 4.2.3 CHALCONES AND AURONES Phlorizidin (17), a naturally occurring dihydrochalcone glycoside from Micromelum tephrocarpum (Rutaceae), was one of first chalcones shown to possess antiparasitic activity. In ethnomedicine it was used for the treatment of malaria because of the bitter taste, a property shared with quinine and other antimalarials. Recent studies provide a rational basis for its antiplasmodial activity. Phlorizidin inhibits the induced permeability in Plasmodium infected erythrocytes to various substrates including glucose. The most promising compound out of this natural product class is licochalcone A (18). This compound was first isolated from Glycyrrhiza glabra (Fabaceae) and is the subject of intensive preclinical studies. The activity of licochalcone A is well documented in vitro and in vivo against a panel of different parasites including P.falciparum L donovani and L. major. The antileishmanial mechanism of action in through the inhibition of the electron transport in the mitochondrion [51]. Further biochemical studies were employed in L. donovani to demonstrated that growth inhibition is mitochondrial specific, and that the main targets are the fumarate dehydrogenase, succinate dehydrogenase and malate dehydrogenase, all essential for parasite viability [51]. Starting with licochalcone A as a lead structure, a large number of chalcones have been synthesized and structure-activity relationships determined with regard to their antiplasmodial, antileishmanial and trypanocidal activity [52]. Kayser et al. [53] demonstrated that structurally related aurones (19) share similar antiparasitic activites with chalcones. It is not kown, if this biogenetically related natural product group inhibits the same target sites as chalcones, but they have a similar size, integrated threecarbon linkers, and similar substituents on both aromatic rings. The main difference lies in the conjugation of the three-carbon linker that in aurones is linked to the B-ring, giving a two-member ring system. A planar structure is typical for all aurones and this conformation has a high similarity to compounds that Li et al. proposed as optimal lead structure of chalcones as protease inhibitors [54]. From molecular modeling studies it appears that chalcones are not only rigid but also adopt an extended structure due to the nature of the conjugated linker. The resulting linear, nearly planar structure, fits perfectly into the active site of Trypanosoma and Plasmodium cysteine proteases [54].

801

These findings suggest that both aurones and chalcones might interact with similar sites in essential parasite enzymes and thus have similar mechanisms of antiparasitic activity.

GIcO

O

(17) phlorizidin (Glc = Glucose)

(18) licochalcone A

(19) aurones (R = H or CH3)

4.2.4 FLAVONOIDS Flavonoids are widespread in the plant kingdom. In contrast to some ethnomedicinally reports up to 1986, there was no scientific evidence of their activity against Leishmania spp., T.cruzi and P, falciparum [55]. However, following the detection of antiplasmodial flavonoids from Artemisia annua (Asteraceae) this natural product group has attracted renewed interest. Elford et al. demonstrated that methoxylated flavonones artemetin (20) and casticin (21) act synergistically with artemisinin against P. falciparum in vitro [56]. The exact mechanism is unclear but tested flavonoids did inhibit the influx of L-glutamine and myoinositol into infected erythrocytes [57]. As a part of a multidiscipHnary research programme in Thailand on antiplasmodial drugs, additional Artemisia species have been screened [58] and exiguaflavanone A (22) and B (23) isolated from Artemisia indica (Asteraceae) exhibited in vitro activity against P. falciparum

802 with EC50 values of 4.6 and 7.1 jLCg/mL, respectively. The flavonoids sakuranetin (24) and 7-methoxyaromadendrin (25) were also reported to be antiprotozoal natural products with inhibition rates of 100 % and 86 % at 500 jLtg/mL in vitro against T. cruzi, respectively. Despite the limited data of the antiprotozoal activity of flavonoids, it can be speculated that the mode of action is linked to the unusual antioxidant pathway. Ribeiro et al. discussed, that the lack of defense mechanisms against oxidative stress makes the parasites susceptible to drugs having an effect on the generation of reactive oxygen species [59]. Recently, Perez-Victoria et al. [60] suggested that specific flavonoids could effect transport mechanisms in Leishmania. The C-terminal nucleotide-binding domain of a P-glycoprotein-like transporter, encoded by the Itrmdrl gene in L. tropica and involved in parasite multidrug resistance (MDR), was overexpressed in Escherichia coli as a hexahistidine tagged protein and purified. The L. tropica recombinant domain efficiently bound different classes of flavonoids with the following affinity: flavone > flavanone > isoflavone > glucorhamnosyl-flavone. The affinity was dependent on the presence of hydroxyl groups at positions C-5 and C-3 and was further increased by a hydrophobic 1,1-dimethylallyl substituent at position C-8. When flow cytometry was used to measure daunomycin accumulation in a L. tropica line, a reversing effect was observed with flavones such as dimethylallyl-kaempferide at low concentrations or apigenin at higher concentration, but not with the glucorhamnosyl derivative rutin nor with the isoflavone genistein (26). The in vivo reversing effect of dimethylallyl-kaempferide was correlated with a high inhibition of the Leishmania cell growth in the presence of daunomycin. The results suggest that flavone inhibition of both daunomycin efflux and parasite growth in the presence of the drug correlates to direct binding of the compound to cytosolic domain of the P-glycoprotein-like transporter [60]. Gale et al. also identified genistein (26) as potent natural compound as modulator of protein phosphorylation with effect on the SPK89 protein kinase in trypanosomes [61]. 4,2.5 NAPHTHOQUINONES Naphthoquinones and other related quinoid compounds are one of the major natural product classes with significant activity against

803

OCH,

(22) exiguaflavanone A (R = H) (23) exiguaflavanone B (R = CH3)

(20) artemetin (R = CH3) (21) casticin (R = H)

R.0

OH

O

(24) sakuranetin (Ri = CH3, R2 = H) (25) 7-methoxyaromadendrin (R = OH) (26) genistein

Leishmania, Trypanosoma and Plasmodium. Many naphthoquinones have been isolated but frequently their potential use has been limited by low bioavailabihty and high toxicity. Wright and Phillipson [62], Sepuvelda-Boza and Cassels [63], Foumet et al. [64] and Akendengue et al. [65] have reviewed much of the literature on naphthoquinones, we focus here on the latest developments and some new structures and their biological activities. The plant derived product hydrolapachol (2-hydroxy-l,4-' naphthoquinone) (27) [66] was shown to have activity against Plasmodium lophurae in ducks in the 1940s (see [67] for details). This observation provided the stimulus for the synthesis of hundreds of analogues including a series of 2-hydroxy-3-alkyl-l,4-naphthoquinones. One of this series, lapinone, synthesised in 1948, showed high activity in experimental models and was used to treat P, vivax infected patients in 1951. Due to high doses required and pharmacological problems interest in naphthoquinones faded. In the 1960s interest was revived and a new compound, menoctone, proved

804 to be highly active in the P. berghei mouse model but disappointing in chnical trials. In the late 1970s and 1980s a series of hydroxynaphthoquinones (HNQs) was synthesised at the Wellcome Laboratories, UK that overcame the problems of poor oral absorption, rapid metabolism and protein binding associated with previous series. These HNQs, with a cyclohexyl ring at the 2-position showed, showed activity against the apicomplexan parasites, Eimeria, Toxoplasma, Theileria and Plasmodium species [68]. This work resulted in the development of parvaquone and buparvaquone for the treatment of theileriosis in cattle and other ungulates and atovaquone for the treatment of malaria (in combination with proguanil) and Pneumocystis carinii pneumonia [69]. Besides this synthetic route, new and structurally interesting naphthoquinones have been isolated. A dimeric naphthoquinone diospyrin (28) from Diospyros montana (Ebenaceae) was found to be active against L. donovani [70]. The inhibition of Type I DNATopoisomerase in this parasite has been suggested as a mechanism of action [70]. Plumbagin and related HNQs have activity against Leishmania spp. in vitro and in vivo [64]. These compounds have been used as the basis for the synthetisis of naphthoquinones designed as subversive substrates of trypanothione reductase. Other rare naphthoquinones have been identified as potential antiparasitic drugs. From Psychotria camponutans (Rubiaceae) the benzisoquinoline-510-dione (29) has been isolated and tested against P. falciparum with an EC50 of 0.84 jLtg/ mL associated with significant cytotoxicity (EC5o= 1.62 jLCg/ mL, KB cells) [71].

o (27) 2-hydroxy-1,4-naphthoquinone

o (28) diospyrin

(29) benzisoquinoline5,10-dione

805 4.2.6 ANTHRAQUINONES AND XANTHONES This natural product group is related to naphthoquinones in structure and biological activity. The main chemical difference between the groups is the tricyclic aromatic system with a para-quinoid substitution. Schnur et al. [71] and Foumet et al. [64] demonstrated that some derivatives have activity in vitro against Leishmania species, but few naturally occuring anthraquinones have been tested. Anthraquinones isolated from the tropical tree Morinda lucida (Rubiaceae) were tested for their antiplasmodial and antileishmanial activtity [72, 73]. In vitro, some (30-32) were more active against L. major promastigotes (EC50 = 9.6 to 185 /xM) than amastigotes and also acive against CQ-R strains of P. falciparum (EC50 = 21.4 to 82.9 jLtM in vitro). The most active compounds have an aldehyde group at C-2, well known as a cytotoxic substructure in other natural products. The activity may also be explained by the cyclic planar structure that makes them potential DNA-intercalators. From the toxicological point, tested compounds showed moderate effects in the lymphocyte proliferation test with all EC50 values over 175 /xM [75]. Antiprotozoal anthraquinones are known from microbial sources [76]. Mycotoxin MT81 (33) and some of its derivatives from Penicillium nigricans showed only moderate antileishmanial activity against L. donovani promastigotes (46% growth inhibition at 250 /xg/mL in vitro). In parallel to the inhibitory effects of naphthoquinones on mitochondria, Majumdar et al. determined the effect of the drugs on the respiration of L. donovani cells [76]. The oxygen uptake was significantly inhibited (inhibition rate > 50 %) by all mycotoxins indicating a similar mode of actions as known for other naphthoquinones. Antiplasmodial xanthones have been isolated from Garcinia cowa (Guttiferae) [77]. Preliminary screening of five prenylated xanthones dmonstrated significant activity against P. falciparum in vitro with EC50 concentrations between 1.5 and 3.0 /xg/mL. Cowaxanthone (34) displayed an antiplasmodial potential (EC50 =1.5 /xg/mL) comparable to that of pyrimethamine (EC50 = 2.8 /xg/mL). Although a number of biological properties are known about xanthones (antibacterial, antifungal and cytotoxicity) there have only been three reports on the

806 antiplasmodial activity of these and only one dealing with a pure natural product.

HX,

0

R3

(30) digitoiutein; Rl = H; R2 = CH3; R3 = OCH3 (33)mycotoxinMT81 (31) rubiadin-1-methyl ether; Rl = OCH3; R2 = CH3; R3 = H (32) damnacanthal; Rl = OCH3; R2 = CHO ; R3 = OCH3

(34) cowaxanthone

4.3 TERPENOIDS 4.3.1 IRIDOIDS The secoiridoid amarogentin (35) isolated from the upper parts of Swertia chirata (Loganiaceae) is a promising compound with leishmanicidal activity. This compound inhibited DNA-Topoisomerase I activity from Leishmania donovani at 30 /xM [79]. This principle is quite interesting as most trypanocidal drugs target type H topoisomerases. There are a few compounds that specifically alter biological functions of toposisomerase I with the enzyme or DNAenzyme complex, an exception being camptothecin an antineoplastic drug [80]. Other natural compounds with the iridoid parent structure also show significant activity against Leishmania parasites. A series of

807 iridoids isolates from Nyctanthes arbortristis (Oleaceae), arbortristosides A, C (36, 37), have antileishmanial activity. Unfortunately, although parasite growth was reduced in vitro at 100 jLtg/mL by 64.5% and 67.4%, respectively, cytotoxicity was noted at 60 jUg/mL [81].

4.3.2 MONOTERPENES Monoterpenes are examples of simple antiprotozoal drugs. Espintanol (11) has already been discussed above and piquerol A (38) was active about 100 /xg/mL [82]. Espintanol, isolated from the bark of Oxandra espinata (Annonaceae) and synthesised [83], is active against Leishmania promastigotes and Trypanosoma epimastigotes. In view of the sensitivity of pathogens from the genus Trypanosoma and Leishmania to oxidative stress, as discussed under section 4.2.1, piquerol A activity may also be due to interaction with the parasite redox cycling system leading to enzyme inhibition and parasite killing.

HO

9OOCH3

H,C OH

(35) amarogentin

"OH HO'

HO

^"

.X^.

(36) arbortristosides A; (38) piquerol A R= p-methoxy cinnamoyl (37) arbortristosides C; R= coumaroyl

4.3.3 SESQUITERPENE LACTONES The antiprotozoal potential of sesquiterpenes is well established since artemisinin (39) and a second endoperoxide sesquiterpene yinghaosu A, were identified as new drugs with high clinical relevance. In addition to artemisinin and other sequiterpene endoperoxides,

808 described in detail in Section 6.1, other series of sequiterpenes with antiprotozoal activity have been described. The sequiterpene lactone parthenin (40) has an EC50 value of 1.29 /xg/mL against P. falciparum in vitro. Although parthenin is described as highly toxic, rats treated with 100 mg/kg/day did not show any signs of toxicity. From these results a series of parthenin derivatives have been synthesized and retested and those with an exocyclic methylene lactone have been identified as active. Exocyclic methylene lactones are well known as the allergic principle in medicinally used plants (mostly Asteraceae, e.g. Arnica montana (Asteraceae)). Pieman et al. [84] demonstrated that parthenin (40) is capable of blocking parasite specific targets responsible for glutathinonylspermidine- and trypanothione synthesis from cysteine and glutathione precursors in Leishmania and Trypanosoma [38]. The sequiterpene lactones brevilin A (41) from Centipeda minima (Asteraceae) and dehydrozaluzanin C (42) from Munnozia maronii (Asteraceae) were discovered from ethnopharmacological screening. From Neuroleaena lobata (Asteraceae), a medicinal plant used in Guatemala for the treatment of Plasmodium infections, activity was documented for germacranolide sesquiterpene lactones as well for furanoheliangolides [85]. These compounds are also active against Leishmania promastigotes and Trypanosoma epimastigotes in vitro. From preliminary structure-activity relationship analysis based on in vitro EC50 data, germanocrenolide sesquiterpenes, like neurolenin A (EC50 = 0.92 /iM) and B (43, 44) (EC50 = 0.62 JLCM), were found to be more potent, than furanoheliangolides as represented by lobatin A and B (EC50 = 15.62 /iM, 16.51 jitM, respectively) (45, 46) [85]. The main reason for the lower activity of the former could best be explained by the shift of the double bond from the 2,3 (neurolenin) into the 3,4 (lobatin) position, suggesting that one of the structural requirement in sesquiterpenes is an a/6-unsaturated keto function. Another approach for discovering antiprotozoal natural products is described by Koshimizu et al. [86]. Wild chimpanzees were observed to chew young stems of Vemonia amygdalina (Asteraceae) from which antiplasmodial sesquiterpenes (vemodalin (47), vemolide, hyroxyvemiladin) have been isolated. Also unusual sesquiterpenes (48, 49) with significant antiplasmodial activities (EC50 < 4 /ig/mL) were isolated from marine red algae {Laurencia implicata.

809

Rhodomelaceae) and brown Rhizophylladaceae) [87].

o (39) artemisinin

algae

6

T

homemannii,

^

(40) parthenin

OH N ^ ^ ^ T T ^ ^

(Portiera

o

(41) brevilin A

T

T

T

(42) dehydrozaluzanin C

. OH

(43) neurolenin A; R = H (44) neurolenin B; R = OAc

(45) lobatin A

(46) lobatin B

"XSO ? (47) vernodalin

(48) 5-isopropyl-3,3,9-trimethylbicyclo-nona-5-en-4ol

(49) 9,10-trisepoxypentadec-12-1,2-diene

4.3.4 DITERPENES Diterpenes from many species are well known for their biological activity and are amongst the most widely distributed terpenes in the plant kingdom. However, most of them combine both high antiparasitic activity as well as high cytotoxicity to mammalian cells. Both jatrogrossidione (50) and jatrophone, isolated from Jatropha

810 grossidentata and J. isabelli (Euphorbiaceae) respectively, showed significant activity against Leishmania promastigotes in vitro (EC50 values of 0.75 /xg/mL (2.4 mM) and 5 jLtg/mL (16 mM)) and L. amazonensis amastigotes in vivo where jatrophone reduced parasite growth at a dose of 25 mg/kg/day. Unfortunately, both proved to have toxic effects of therapeutic doses [88]. A series of 80 labdane derivatives showed significant antileishmanial activity (L. donovani L enriettii, L. major, L. infantum), but cytotoxicity increased in parallel with the antiprotozoal effect [89]. From structure-activity studies overbridged tri- or tetracyclic ring labdanes had more significant activity in comparison to bicyclic labdanes. These results confirmed other data cited in literature, for example the subtype of overbridged labdane derivates (e.g. kauran-, trachyloban-type), as displayed by ent-kaur-16a-ol-19oic acid (51) from Mikania obtusata, ent-kauran-16-en-19-oic acid from Wedelia paludosa (52), and (-)-trachyloban-19-oic (53) from Viguirea aspillioides (Asteraceae). Results of their extraordinary activity against Leishmania and Trypanosoma spp. have been published [90-92]. However, most antiprotozoal labdanes also show low EC50 values in vivo (EC50 < 3.0 /xg/mL), and high toxicity (EC50 < 10 jLtg/mL). One compound isolated from the sponge Acanthella klethra, axisonitrile (54), a sequiterpene derivative, showed potent antiplasmodial activity with no detectable cytotoxic properties [93]. A series of different terpenes have been isolated from marine organisms and are of considerable interest for their unique structural features in antiparasitic drug research. In contrast to plant metabolites most of these 100 plus isolated natural products contain isonitrile, isothiocyanate, and thiocyanate functionalities. Major interest has been focussed on these marine drugs by [94], doubling the information in the literature and identifying novel compounds with EC50 values below 1 jLCg/mL with high selective indices (SI > 50) [95]. From the point of antiparasitic research four diterpene subclasses are of interest: kalihinane diterpenes, e.g. kalihinol A (55) [96], amphilectanes, for example 7-isocyano-ll(20),14-epiamphilectadiene (56) [95], cycloamphilectanes, for example 7-isocyanocycloamphilect-10-ene (57) [95], and isocycloamphilectanes, for example 7,20-diisocanoisocycloamphilectane (58) [95]. The potent and

811 selective biological activities of these compounds represents an exciting advance in the search of novel antiplasmodial agents. In vivo studies are required now to validate the potential of these compounds. Not only novel compounds with unique structural features attract attention, but also well known compounds like macrocyclic terpenes used in other pharmacological fields, like taxol and epothilione (59) (potent antineoplastics) or highly toxic phorbol esters, have been tested in antiparasitic drug screens. Macrocyclic trichothecenes are known as mycotoxins, and a variety of biological activities have been reported. From the fungal culture of Myrothecium verrucaria BCC 112 (Hypomycetes) roridin E (60) show high antiplasmodial activity (EC50 = 0.15 ng/mL) but also significant cytotoxicity (EC50 = 0.5 ng/mL, KB cells (human epidermal nasopharyngeal cancer cells) [97]. It must also be noted that the selective index of roridin E (SI = 12) in comparison to artemisinin (SI > 7,100) was too low to make it useful for further in vivo investigation. The experience of research into antineoplastics with macrocyclic ring systems suggests that it maybe possible to find or synthesize new trichothecene derivatives with high antiplasmodial activity and low toxicity. Oketch-Rabah et al. showed that the macrocyclic germancrane dilactone 16,17-dihydrobrachycalyxolide (61), from Vemonia brachycalyx (Asteraceae), has both antileishmanial and antiplasmodial activity [98]. In in vitro tests the compound is strongly active against L. major (EC50 =17 /xg/mL) and P. falciparum (EC50 =17 /ig/mL), but also inhibits the proliferation of human lymphocytes at the same concentration indicating general toxicity [98].

COOH

(50) jatrogrossidione (51) ent-kaur-16a-ol-19-oic acid

^

COOH

(52) ent-kauran-16-en19-oic acid

812

COOH

(53)

(-)-trachyloban-19-oic

(56)7-isocyano-ll(20),14epiamphilectadiene

CI

(54)

axisonitrile-3

(57) 7-isocyanocycloamphilect-10-ene

O

OH

(55)

kalihinol

(58) 7,20-disiocanoisocycloamphilectane

O

(59) epothilone A

(60) roridin E

A

(61) 16,17-dihydrobrachylocalyxolide

813 4.3.5 TRITERPENES Triterpenes and saponins from plant sources are known for their biological activity (antineoplastic, anthelmintic and antiviral), but they exhibit some toxicity to humans and other mammals. Despite the fact that triterpene action in biological systems is well known, the first rational reports on their antiprotozoal activity were first described late 1970s. Tingenone (62) and pristimerin (63), from species of Celastraceae, have in vitro activity against T. cruzi amastigotes and P. falciparum, Tingenone could act through interaction with DNA or inhibition of DNA synthesis [45]. The lupane-type triterpene betulinic acid (64), also known for its antineoplastic effect, was identified by bioguided fractionation and identified as the antiplasmodial principle of Triphyophyllum peltatum and Ancistrocladus heyneanus (Dioncophyllaceae and Ancistrocladaceae, respectively) [99]. Against P. falciparum in vitro betulinic acid had an EC50 value of 10.46 jLtg/mL and with the exception of human melanoma cells (EC50 =1-5 jLig/mL), only moderate cytotoxicity (EC50 > 20 /xg/mL). Based on saponins isolated from ivy, Hedera helix (Araliaceae), MajesterSarvonin et al. gave a first insight in the structure-activity relationship of different antileishmanial saponin types [100]. Bidesmosides have no effects on the proliferation of either promastigote or amastigotes. In contrast, monodesmosides and hederagin (65) were highly active, especially the sodium salts of a- and 6-hederin (66, 67) are highly active at concentration similar to that of pentamidine. Monodesmosides from Hedera helix damaged macrophages host cells at concentrations between 5 and 25 ixg/mL, but this level of toxicity was in the same range as for Glucantime [100]. The use of saponins as drugs is limited due to poor bioavailability, reduced absorption in the gastrointestinal tract and their hemolytic toxicity when given by parenteral route. It is noteworthy that despite this fact medicinal plants that contain saponins are known. Oketch-Rabah et al. isolated a new steroidal saponin, muzanzagenin (68), from Asparagus africanus (Liliaceae) which had antileishmanial and antiplasmodial activity (EC50 = 70 /xM, L. major, EC50 = 61 /xM, P, falciparum K39) [101].

814

.xCOOCHg

(62) tingenone

(63) pristimerin

COOR, COOH

(64) betulinic acid

(65) hederagin; R,= H, R2 = OH, R3 = H (66) a-hederin; Rl= rham(l-^2)ara (l-^,R2 = OH,R3 = H (67) 6-hederin; Rl= rham(l->2)ara (1->,R2 = H,R3 = 0H

(68) muzanzagenin

4.3.6 LIMONOIDS Bitter terpenoids, known as limonoids, are biosynthetically related to the quassinoids that are produced by species of Meliaceae. One well known representative from this family is Azadirachata indica, the neem tree, widely used as an antiplasmodial plant in Asia. Rochanakij

815

et al. initially identified nimbolide (69) as the active antimalarial principle of the neem tree (EC50 = 0.95 ng/mL, P. falciparum Kl) [102]. Nimbinin, geduin (70) (EC50 = 0.39 ng/mL, P. falciparum D6) and its dihydroderivative were also found to be active in vitro against Plasmodium parasites in the range of EC50 values of 0.72 - 1.74 /xg/ml [103-105]. The mode of action of this natural product group is still unclear. The cytotoxic activity of gedunin was moderate (EC50 = 275 /xg/mL). Geduin derivatives, D-seco-limonoids, do not show significant antiplasmodial activities (EC50 > 100 ng/mL, P. falciparum D6) comparable to the parent structure of geduin [105]. Insufficient geduin-related limonoids have been tested to allow a proper evaluation of this group.

'"OAc

(69) nimbolide

(70) geduin

4.3.7 QUASSINOIDS Quassinoids are biosynthetically related to triterpenes and share the same metabolic precursors. Most of the presently known quassinoids were found in the family Simbaroubaceae, and extracts and isolated natural compounds have been widely tested [106]. As deduced from structure-activity relationship analysis most potent quassionoids have a pentacyclic ring systems with a lactone function and a methyleneoxygen ring bridge linking C-8 and C-13 (e.g. brusatol (71)) or C-11 (e.g. ailanthinone (72)) [107]. Most of the quassinoids do not have a sufficient selective index to be considered as lead structures for clinical drugs. The antiplasmodial activity is high with EC50 values around 0.02 jU-g/mL, but the most active compounds like simalikalactone D (73) from Simaba guianensis (Simaroubaceae) [108], 156-heptylchaparrinone [109] and different sergolide

816

quassinoids [110] were too toxic in vivo. The mode of action of quassinoids seems to be the inhibition of protein synthesis [111]. Quassin (74) is inactive due to the missing methylene/oxygen bridge. A current research aim is to modify the parent compound synthetically, to find semisynthetic quassinoids with reduced toxicity and to improve the toxic/therapeutic ratio. As a potential lead structure for this approach Francois et al. identified chaparrinone (75) derivatives with an improved selective index [112]. The in vitro activities of chaparrinone and 15-desacaetylundulatone (76) were lower than reported activities of certain other quassinoids (EC50 = 0.037 jLtg/mL and 0.047 /xg/mL, P. falciparum NF54, respectively). In vivo, however, when given at 50 mg/kg/day they produced significant reduction of parasitaemia with survival times similar to those of the control group with no signs of acute toxicity [112]. The contradictionary toxic/therapeutic ratio was explained by the occurence of the keto-function at C-2 and missing hydroxy] group at C-14 in a C8->C11 overbridged ring system improving its performance.

(71) brusatol

(72) ailanthinon

rr (13) simalikalactone D

817 0CH3

HXO

(75) chaparrinone, R = H (76) 15-desacaetylundulatone; R = 0-tiglate

4.4. N-CONTAINING NATURAL PRODUCTS (NONALKALOIDS) Besides the large group of alkaloids some nitrogen-containing natural products, do not fall under the definition of alkaloids, and are therefor discussed separately.

4.4.1 STEROIDAL ALKALOIDS The literature contains numerous reports on biological activities of nitrogen-containing steroids of the Solanum-iypQ. Most are quite common derivatives that occur in vegetables and thus in the daily diet; a-solanine (77), tomatine (78) which have been tested for their toxicological potential but their antiparasitic activity has not fully tested. Chataing et al. tested a series of Solanum-iype steroid alkaloids against Trypanosoma cruzi in vitro in comparison to ketoconazole [113]. Glycoalkaloids containing a chacotriose sugar moiety showed trypanocidal activity against epimastigote and against metacyclic trypomastigote forms. The mechanism of action probably is the membrane, followed by structural changes of internal compartments, resulting in destruction of organelles such as mitochondria and glycosomes. The data indicate that steroid alkaloids containig 6chacotriose trisacharide moiety, e.g. a-chaconine (79) and asolarmargine (80), posses antitrypanosomal activity in the range of EC5o = 6.0/xM[113].

818

(77) a-solanine (solatriose)

(79) a-chaconine (chacotriose)

819

(80) a-solarmargine (chacotriose)

4.4.2 OTHER N-CONTAINING COMPOUNDS Piperine (81), a major constitutent of pepper (Piper nigrum, Piperaceae), was tested against L donovani promastigotes [114]. However, as this compound has been part of the daily diet over centuries, is cytotoxic potential, and only the moderate antileishmanial activity, it has not been considered as a potential antiparasitic agent. Recent studies found three halogenated pyrrole-2-carboxylic acids in a Maltese sample of the marine sponge Agelas oroides (Agelasidae) (82-84) [115]. Activity was evaluated against P. falciparum (strains D6 and W2) with EC50 values between 3.3 and 5.3 /xg/mL. Parallel testing for against KB, Lul, LNCaP and ZR-75-1 cells showed cytotoxic activity (EC50 = 2.0 -14.5 /xg/mL).

'^ (81) piperine

(82) 4,5-Dibromopyrolle-2-carboxylic acid; R = OH (83) 4,5-Dibromopyrolle-2-carboxylic methylester; R = OCH3 (84) Oroidin; R = NH^ NH

820 4.5 ALKALOIDS Alkaloids are one of the most important classes of natural product providing drugs for humans since ancient times. Most alkaloids are well known because of their toxicity or use as psychodelic drugs (e.g. cocaine, morphine or the semisynthetic LSD), but many alkaloids have had a deep impact on the treatment of parasitic infections. The outstanding example is quinine (1) from Cinchona succirubra (Rubiaceae) used for the treatment of malaria for more than three centuries.

4.5.1 QUINOLINES Up to the middle of this century quinine (1) was used for the treatment of malaria, and with the widespread development of chloroquineresistant strains of Plasmodium falciparum it has become important again. Quinine has been used for the treatment of malaria for more than 350 years and has its origin in Peru in the early 17^^ century. Quinine was the lead structure in the discovery of new synthetic derivatives like mefloquine that have higher antimalarial activity. This section will focus on other new quinoline alkaloids. The mechanism of antiplasmodial action and resistance of quinolines is well described elsewhere [116]. As the result of an ethnopharmacological search for new antileishmanial drugs aryl- and alkyl-quinolines were isolated from Galipea longiflora (Rutaceae) [117]. These simple natural quinohne derivatives 2-n-propylquinoline (85), chimanine B (86), chimanine D (87), 2-n-pentylquinoline (88), 4-methoxy-2-phenylquinoline (89), and 2-(3,4-methylenedioxyphenyl)-quinoline (90) were tested against strains of parasites causing cutaneous leishmaniasis and exhibited activities of EC50 = 25-50 jLtg/mL or 150 - 300 /xM [117, 118]. Only chimanine B was active in vivo (50 mg/kg, BALB/c mice); twice daily oral treatment results in a decrease of parasite load by 95 %, a similar activity to that of the standard drug Glucantime [119]. No mechanism has been found yet to explain these effects. Two piperidino-4quinolinone alkaloids dictyolomide A (91) and B were identified from Dictyoloma incanescens (syn. D. vandellianum) and D, peruviana (Rutaceae), collected in South America (Bolivia). They induced a

821

lysis of various strains of Leishmania promastigotes in vitro at a concentration of 100 /xg/mL [120].

(85) 2-n-propylquinoline; Ri = C3H7, R2= H (86) chimanine B; Ri= CH=CH-CH3, R2= H (87) chimanine D; Ri XN>/ = , R2 = H

(88) 2-n-pentylquinoline; Ri = C5H11, R2= H (89) 4-methoxy-2-phenylquinoline; Ri = phenyl, R2= OCH3 (90) 2-(3,4-methylenedioxyphenyl)quinoline, \..-^%«^^ R2 = H Ri=

^'j

(91) Dictyolomide A

4.5.2 BENZYL- AND NAPHTHYLISOQUINOLINE ALKALOIDS The chemical structure of this alkaloid group is well known through the widespread and abundant berberine (92). Many antiprotozoal isoquinolines have been isolated from the families Annonaceae, Berberidaceae, Menispaermaceae and Hemandiaceae [65]. Berberine is active at EC50 = 10 /xg/mL against Leishmania amastigotes within murine peritoneal macrophages. Vennerstrom et al. tested berberine and several of its derivatives for antileishmanial activity against L. donovani and L. panamensis in golden hamsters [121]. Tetrahydro-

822 berberine is less toxic and more potent than berberine against L. donovani but was not as potent as meglumine antimonate (Glucantime). Only berberine, the natural product, showed significant activity (greater than 50% suppression of lesion size) against L. panamensis. Berberine was used for cutaneous leishmaniasis in India but was not effective when applied topically [122]. Recently catecholic berberines, (-)-pessione (93) and (-)-spinosine (94), have been isolated and tested for antileishmanial and trypanocidal {T. cruzi) activity in vitro. At a single concentration of 250 /xg/mL 50 % inhibition for T. cruzi is found, indicating low trypanocidal activity [123]. Naphthylisoquinoline alkaloids isolated from tropical llianas have been identified as new promising leads as antiprotozoals. They show remarkable activity against P. falciparum in vitro and in vivo, as well against Leishmania and Trypanosoma species [124]. Extracts from the single species of Triphophyllum peltatum (Dioncophyllaceae) dioncopeltine A (95) and, in particular, dioncophylline B (96) and dioncophylline C (97) exhibited high antiplasmodial activity [124]. Dioncopeltine A is able to suppress parasitaemia almost totally, while dioncophylline C cured infected mice completely after oral treatment with 50 mg/kg per day for 4 days without noticeable toxic effects. Analysis of the dose-response relationship of dioncophylline C revealed an ED50 dose of 10.71 mg/kg/day. Although four daily treatments with 50 mg/kg/day are needed to achieve parasitological cure, one oral dose is sufficient to kill 99.6% of the parasites. Intravenous application of dioncophylline C is even more effective, with an ED50 of 1.90 mg/kg/day and no significant toxic effects. The compound also suppressed more established P. berghei infections when applied orally at day three post infection. Both dioncopeltine A and dioncophylline C are active against the chloroquine-resistant P. berghei Anka CRS parasites. Structure-activity relationships indicate that the presence of a secondary amine function, and the absence of an oxygen substituent at C-6 and R-configuration at C-3 are important. Recently, a novel dimeric antiplasmodial naphthylisoquinoline alkaloid heterodimer, korundamine A (98), has been isolated from another species, Ancistrocladus korupensis belonging to the family Ancistrocladaceae that is biogenetically related to Dioncophyllaceae. Korundamide A is one of the most potent naturally occuring naphthylisoquinoline dimers

823 yet identified in antiplasmodial in vitro screening with an EC50 of 1.1 jLtg/mL against P. falciparum [125].

(93) (-)-pessione; R = H (94) (-)-spinosine; R = CH3

OH = HO- - < ^

(95) dioncopeltine A

>X%:^ -

-

OH =

[I

^OCH,

(96) dioncophylline B

0CH3

OH

(98) korundamine A

(97) dioncophylline C

824 4.5.3 BISBENZYLISOQUINOLINES A number of different bisbenzylisoquinolines with antiprotozoal activity have been identified. Although their antiparasitic activity has been recognised for years, particularly the antiplasmodial activity, the mechanism of action of these alkaloids is still unclear. So far in vivo activity has not been demonstrated. In vitro most bisbenzylisoquinolines exhibit activities in vitro far below 1 /xg, close to the EC50 value of chloroquine (EC50 - 0.2 /iM). Some bisbenzylisoquinolines like gyrocarpine (99), daphnandrine (100) and obaberine (101) are more potent than antimonials and nifurtimox and benznidazole, respectively, against Leishmania and Trypanosoma parasites (EC50 < 50 jtig/mL) [123, 127]. Despite the fact that a large number of bisbenzylisoquinolines has been tested, a clear structureactivity relationship is not clear. Some structural features that appear to be important include the linkage of the heteromers and the number of ether bonds. Studies on Triclisia alkaloids showed that those compounds with two ether bridges (e.g. pycnamine) (102) are more potent than those with three ether bridges such as cosculine (103) (EC50 values of 0.15 /xg and 15.56 ng/mL, respectively) [128]. Recently, Angerhofer et al. published an intensive study on structureactivity/toxicity-relationship of a series of 53 structurally different bisisoquinolines [129]. More than half of the compounds tested against KB cells for cytotoxicity and P. falciparum strains W2 and D6, however, showed selective antiplasmodial activity, with > 100-fold greater toxicity toward one or both of the P. falciparum clones, relative to cultured mammalian cells. The most selective alkaloids were (-)-cycleanine (104), (+)-cycleatjehine (105), (H-)-cycleatjehenine (106), (+)-malekulatine (107), (-)-repandine (108), and (+)temuconine (109). As a result of these studies, an understanding of the relationships between the structures, the stereochemistry, the substitution patterns of these alkaloids and their in vitro antiplasmodial and cytotoxic activities are beginning to emerge. The quatemarization of one or two nitrogen atoms, presence of an acetyl function at N-2', and N-oxide formation leads to a loss of toxicity and antiplasmodial activity. The decrease in lipophilicity (membrane permeability) of all of these alkaloids probably contributes to the lower toxicity observed. Within each subgroup of bisbenzyliso-

825 quinolines a change of configuration of the chiral center, as well as modification of substituents, may lead to independent changes in cytotoxicity and antiplasmodial activity. However, except for the three one-bridged compounds, (+)-neothalibrine (110), (+)-temuconine, and (+)-malekulatine, which show low toxicity and appreciable antiplasmodial activity, the current results do not reveal any clear structure-activity relationship between subgroups of bisbenzylisoquinoline alkaloids. With the exception of the onebridged bisbenzylisoquinolines, all possess a large heterocycle of 18 to 20 atoms, which confers flexibility to the molecule. A study of the conformations assumed by compounds of the same subgroup (e.g., modification of conformation with the change of configuration at C-1 and C-H should give more information on the structure-activity relationship. As the therapeutic index of the most antiplasmodial alkaloids is around 100 and those of chloroquinine, quinine and artemisinin, are 5460, >285 and >4680 respectively, the bisbenzylisoquinolines do not appear to be promising candidates as antimalarial agents. Monomeric benzylisoquinolines do not appear to have potential. The activity of some aporhinoids, like isoguattouredigine (111) (from Guatteria foliosa, Annonaceae) argentinine (112), unonopsine (113) and hydroxynomuciferine (114) show only minor activity against T. cruzi in vitro (EC50 > 250 ixM) [130, 131]. The isoquinoline derivate camptothecin (115), a well known antineoplastic drug and a topoisomerase I inhibitor, showed antiprotozoal activity when tested against L. donovanU T. cruzi and T. b. brucei with EC50 values of 1.5, 1.6 and 3.6 /xM [132, 133]. For these parasites, camptothecin is an important lead for much-needed new chemotherapy, as well as being a valuable tool for further study of topoisomerase I activity.

826 OCH,

PCH, PCH,

R.0

(99) (+)-gyrocarpine

(100) (+)-daphnandrine; Rj = CH3, R2 = H (101) (+)-obaberine; R, = R2= CH3 (108) (-)-repandine; Ri = H, R2 = CH3

.OCH, PCH, PCH,

HO

(102) (+)-pycnamine

o-

(103) (+)-cosculine

^OCH, OCH,

H3CO'

(104) (-)-cycleanine

(105) (+)-cycleatjehine; R = H (106) (+)-cycleatjehinine; R = CH3

827

OCH,

/

"^V—OH OCH3

OCH3

(109) (+)-temuconine; Ri = H, R2 = CH3 (110) (-f )-neothalibrine; Ri = CH3, R2 = H

(107) (+)-malekulatine

H,CO'

(111) isoguattouredigine

(112) argentinine

(114) hydroxynornuciferine

(113) unonpsine

(115) camptothecin

4.5.4 INDOLES Indoles comprise another group of alkaloids with high biological activity. The indole sub-structure is widely distributed in the plant kingdom. Some indole derivatives have been reported to possess antiprotozoal activity. Indoles are biosynthetically derived from

828 tryptophan metabolism, which appears to be important in protozoa such as Leishmania and Trypanosoma, The end products of the tryptophan metabohsm are thought to be involved in carbohydrate metabolism [134]. A simple derivative with antileishmanial activity (L. amazonensis amastigotes) is harmaline (116), often found in indole containing plants, e. g. Peganum harmala (Rutaceae). Harmaline, a harmane-type (117) indole alkaloid, is active at EC50 = 24 jitg/mL, but too toxic for human use. The relevant pharmacological and antiprotozoal action of harmaline and related tryptamine derivatives is intensively discussed by [135]. Other monomeric indole derivatives are olivacine (118) and ellipticine (119). Both were identified as antiprotozoal compounds in the 1970s [136], and both showed in vitro activity against T, cruzi epimastigotes with EC50 values of 2.5 and 5.0 jLtg/mL, respectively. In contrast, both were inactive in vivo, maybe because of inactivation through first pass metabolism. Cryptolepine (120) and related alkaloids, indole-quinolines, have been isolated from Cryptolepis sanguinolenta (Periplocaceae) and were active in vitro against P. falciparum in vitro (EC50 = 27-41 ng/mL, P. falciparum^!, D6, and Kl) but failed in vivo (only 10.8 - 19.4 % suppression of P. yoelii at 100 mg/kg/day) [137]. Among the group of "dimeric" indole alkaloids the tubulin polymerisation inhibitor and antineoplastic agent vinblastine (121) is of experimental interest. In the therapy of protozoa infections its use is limited because of the poor therapeutic ratio against Trypanosoma gambiense, L donovani and P. falciparum [138, 139]. Conodurine (122) and conoduramine from Peschiera van heurkii (Apocyanaceae) (123) showed antileishmanial activity with EC50 value of 50 jiig/mL against L. amazonensis promastigotes in vitro [140]. Conodurine was less active than Glucantime (EC50 = 40 mg/kg/day, BALB/c mice) in vivo against L. amazonensis, and doses of conodurine at 200 mg/kg were toxic [140]. The mechanistically unusual antineoplastic product taxol (124), a diterpene-alkaloid, inhibits depolymerization of tubulin also in Plasmodium, Trypanosoma and Leishmania parasites, acting at concentrations as low as 0.1 /xM. Because of toxicity reasons, this compound does not seem to be a particular attractive candidate for further development as antiparasitic agent.

829

(116) harman; Ri = H, R2 = CH3 (117) harmaline; Ri = OCH3, R2 = CH3

(118) olivacine; R, = CH3, R2 = H (119) ellipticine; Ri = H, R2 = CH3

\ ^ w ^

OH

OCOCH3

(121) vinblastine; R = COOCH3

(120) cryptolepine

C00CH,PH

HjCOOC-

H3COOC

(122) conodurine

(123) conoduramine

830

(124) taxol (paclitaxel)

4.6 OTHER NATURAL PRODUCT CLASSES 4.6.1 NUCLEOSIDES Sinefungin (125), a natural nucleoside isolated from cultures of Streptomyces incamatus and S. griseolus, is structurally related to Sadenosylhomocysteine and S-adenosylmethionine (SAM) (126) [141]. Sinefungin has been shown to inhibit the growth of various fungi and viruses, but its major attraction resides in its potent antiparasitic activity. This natural product has attracted renewed interest since the synthetic S-adenosylmethionin-decarboxylase inhibitor 5-([(Z)-4amino-2-butenyl]-methylamino)-5-deoxyadenosine (MDL 73811), a decarboxylated S-adenosyl-L-methionine analog, was introduced in experimental studies as a new drug for the treatment of Leishmania and Trypanosoma [142, 143]. Sinefungin does not inhibit Sadenosylmethionin-decarboxylase, but its action is focussed on SAMsynthases affecting methylation of macromolecules as nucleic acids and blocking of DNA polymerase by reduction of dATP. NH,

NH

H3

»j»"^N^

NH,

NH^

NH,

Ij^'^^N^

HO

OH

HOOC HO

OH

(125) sinefungin

(126) SAM

831 4.6.2 AMINOGLYCOSIDES The aminoglycoside antibiotic, aminosidine (127), also known as paromomycin and monomycin, was first shown to be active against experimental cutaneous leishmaniasis in the early 1960s [144]. Later studies showed that is was the most potent among a series of tested compounds derived from microbiological sources [145]. Interest in the antileishmanial properties of this compound has been revived by the development of topical formulations for the treatment of cutaneous infections. It was found that topical application of either paromomycin or gentamicin, together with a transdermal enhancing agent, cured the parasite lesion, and that combined treatment with the two compounds had an additive effect [146]. The pharmacology and antiparasitic mechanism of these drugs formulations is discussed in detail by [147].

HO

Q

HO OH OH

(127) aminosidine (paromomycin)

5.

FROM MEDICINAL HERB TO THE DRUG MARKET

5.1 FROM ARTEMISININ TO ARTEMETHER (PALUTHER®), ARTEMETHER (ARTENAM®) AND ARTESUNATE (ARSUMAX®) History. For thousands of years Chinese herbalists treated fever with a decoction of the plant called "qinghao", Artemisia annua, "sweet wormwood" or "annual wormwood" belonging to the family of Asteraccae. In the 1960s a program of the People Republic of China re-examined traditional herbal remedies on a rational scientific basis including the qinghao plant. Early efforts to isolate the active principle

832 were disappointing. In 1971 Chinese scientists followed an uncommon extraction route using diethyl ether at low temperatures obtaining an extract with a compound that was highly active in vivo against P. berghei in infected mice. The active ingredient was febrifuge, structurally elucidated in 1972, called mostly in China "qinghaosu", or "arteannuin" and in the west "artemisinin". Artemisinin, a sesquiterpene lactone, bears a peroxide group unlike most other antimalarials. It was also named artimisinine, but following lUPAC nomenclature a final "e" would suggest that it was a nitrogencontaining compound that is misleading and not favoured today. Chemistry and Pharmacology. The chemistry and pharmacology of artemsinin has been reviewed in detail by Klayman [149], Luo and Shen [150], Woerdenbarg et al. [151], and van Agtmael et al. [152]. The limited stability as well as the poor solubility of artemisinin in water and oil, the two commonly used and approved media for parenteral administration, prompted scientist to prepare semisynthetic derivatives leading to improved solubility in water or higher chemical stability in oil formulations. Artemisinin is poorly soluble in water and decomposes in other protic solvents, probably by opening of the lactone ring. It is soluble in most aprotic solvents and is unaffected by them at temperatures up to 150°C and it shows a remarkable thermal stability. This section will focus on biological and pharmaceutical aspects; synthetic routes to improve antimalarial activity and to synthesize artemisinin derivatives with different substitutiuon patterns are reviewed elsewhere [151, 153]. Most of the chemical modifications were conducted to modify the lactone function of artemisinin to a lactol. In general alkylation, or a mixture of dihydroartemisinin epimers in the presence of an acidic catalyst, gave products with predominantly 6-orientation, whereas acylation in alkaline medium preferentially yields a-orientation (128-132). Artemether (128) as the active ingedient of Paluther® is prepared by treating a methanol solution of dihydroartemisinin with boron trifluoride-etherate yielding both epimers. The main goal was to obtain derivatives that show a higher stability when dissolved in oils to enable parenteral use. The a-epimer is slightly more active (EC50 = 1.02 mg/kg b.w.) than the 6 epimer (EC50 = 1.02 mg/kg) and artemisinin itself (EC50 = 6.2 mg/kg) [154]. Synthesis of derivatives

833 with enhanced water solubility has been less successful. Sodium artesunate, Arsumax®; (132) has been introduced in clinics, is well tolerated and less toxic than artemisinin. The synthetic routes start with dihydroartemisinin treated with succinic anhydride in the presence of DMAP.

(128) (129) (130) (131) (132)

dihydroartemisin; R = H (a + 6) artemether; CH3 (6) arteether; CH2CH3 (6) artelinate; CH2C6H4COONa (B) artesunate; COCHjCHsCOONa (a)

Pharmacokinetics and Pharmacodynamics. A characteristic of artemisinin and its related endoperoxide drugs is the rapid clearance of parasites in the blood in almost 48 hours. Titulaer obtained pharmacokinetic data for the oral, intramuscular and rectal administration of artemisinin to volunteers [155]. Rapid but incomplete absorption of artemisinin given orally occurs in humans with a mean absorption time of 0.78 h with a absolute bioavailability of 15 % and relative bioavailability of 82%. Peak plasma concentrations at a given dose are reached after 1-2 h and the drug is eliminated after 1 to 3 hours. The mean residence time after intramuscular administration was three times that when given orally. Other routes of adminstration, for example rectal or transdermal, are of limited success, but for the treatment of convulsive malaria in children artemeether in a rectal formulation is favoured. Artesunate acts as a prodrug that is converted to dihydroartemisinin. When given orally the first pass mechanism in the gut wall takes places metabolizing half of the administered dose. Oral artemether is rapidly absorbed reaching maximum blood levels (Cmax) within 2-3 hours. Intramuscular artemether is rapidly absorbed reaching Cmax within 4-9 hours. It is metabolized in the liver to the demethylated derivative dihydroartemisinin. The elimination is rapid, with a half-life time

834 (Ti/2) of 4 hours. In comparison, dihydroartemisinin (128) has a Tm of more than 10 hours. The degree of binding to plasma proteins varies markedly according to the species considered. The binding of artemether to plasma protein was 58% in mice, 61% in monkeys and 77% in humans. Radioactive labeled artemether was found to be equally distributed in plasma as well as in red blood cells indicating an equal distribution of free drug between cells and plasma.

Haem (Fe^^)

Haem (Fe^^)

HXi

HX-

Alkylation of proteins Formation of

Fig. (1). Mechanism of action of artemisinin drugs, Active metabolites and formation of reactive epoxide intermediates (according to van Agtmael et al., 1999)

From the toxicological point of view artemisinin seems to be a safe drug for the use in humans. In animal tests neurotoxicity has been documented, but as yet this side effect has not been reported in

835 humans [156]. A major disadvantage of the artemisinin drugs is the occurrence of recrudescence when given in short monotherapy. So far no resistance has been observed chnically although it has been induced in rodent models in vivo. The mechanism of action is different from the other clinically used antimalarials. Artemisinin drugs act against the early trophozoite and ring stages, they are not active against gametocytes, and it affects blood- but not liver-stage parasites. The mode of action is explained by haem or Fe^"*", from parasite digested haemoglobin, catalysing the opening of the endoperoxide ring and forming free radicals. Malaria parasites are known to be sensitive to radicals because of the lack of enzymatic cleaving mechansims. The mechanism of action and the metabolism of reactive artemisinin metabolites is shown in Fig. 1. Prospects. Other indications for malaria for the artemisinin drugs are currently under investigations. Without a final proof, other erythrocyte persisting parasites like Babesia are maybe another interesting target parasite. But also Toxoplasma gondii, Pneumocystis carinii infections in mice have been treated successfully with artemisinin drugs [157159].

6.2 FROM QUININE TO MEFLOQUINE History. Quinine has been listed as one of the six most important plant products that have influenced human history [160]. Cinchona, or "Quinine Bark" is one of the most famous plants from South America and most important discoveries. Legend says that the name "cinchona" comes from the Countess of Chinchon, the wife of a viceroy of Peru, who was cured in 1638 of a malarial type of fever by using the bark of the Cinchona tree. The legend starts with a misspelled name, continues with an extract named mistakenly by Linnaeus in 1742 as "quinquina", and maintains the reputed traditional use of plant extract for a disease probably introduced to that continent by Europeans and their African slaves [see 161, 162, 6]. Quinine bark was used by the Jesuits very early in its history, first advertized for sale in England in 1658 and was made official in the London Pharmacopoeia in 1677.

836 Several years after "Countess's powder" arrived in England, it arrived in Spain where the virtues of the bark were rapidly recognized, from this "tree of fever of the region of Loxa". Rapidly and due to the influence of the Company of Jesus, the "Jesuits' powder" became known all over Europe. Physicians gave credit to the drug, and because of the specificity of its action on malaria, it was recognized officially even when the identity of the producing species remained unknown. Despite the confused history quinine still provides an important treatment for malaria, in particular cerebral malaria, with formulations of interest such as Quinimax, containing quinine and other active isomers, quinidine and cinchonine. Perhaps of greater significance to the history of malaria treatment, by a contorted route, is the work of William Perkins who in 1856 while trying to synthesise quinine discovered the first synthetic dye "mauve". His work led to analine dyes that in turn led to methylene blue, the first compound rationally used for the treatment of malaria by Guttman and Ehrlich in 1891. Methylene blue provided the template for the design of the aminoquinoline drugs mepacrine, primaquine and chloroquine in the 1920s and 1930s [161]. Research on quinine also led to other drugs. In 1944 scientists were able to synthesise and structure elucidate (Fig. 2) the quinine alkaloid in the laboratory in Germany. This led to various synthesized quinine drugs to treat malaria and the use of the common bark and the natural quinine extracted from the bark and sold as antimalarial drugs fell out use. Indonesia and India still cultivates the Cmc/iona-tree, but Zaire has become the top supplier of a world market which is also supplied by other African countries (Burundi, Cameroon, Kenya), and much lower on the list of producers are the South American countries of Peru, Bolivia and Ecuador. Plant products have also played a significant role in the treatment of another parasitic disease, amoebiasis. The root and corm extracts of Cephaelis ipecuanhana were traditionally and empirically for the treatment of dysentry and the active ingredient emetine was isolated in 1817. It was not until 1912 that the antiamoebic activity of the alkaloid emetine was placed on a rational basis. Another product cepahaeline proved to be less active than its methyl ester emetine. In the early 1960s the derivative dehydroemetine was shown to be more active against E. histolytica and less toxic than the parent compound.

837 This compound is still in use. Traditional uses of other plant products led to the identification of other compounds with activity against Entamoeba in the 1940s and 1950s, including the steroidal alkaloid conessine from Holarrhena sp. (Apocyanaceae), glaucarubin from Simarouba amara (Simaroubaceae), and henna from Lawsonia alba (Lythraceae) [163].

\

1 R = H: cinchonine 2 R = OCH3: quinidine ^

R5

3 R = H: epi-cinchonine 4 R = OCH3: epi- quinidine

R

5 R = H: cinchonidine 6 R = OCH3: quinine

7 R = H: epi- cinchonidine 8 R = OCH3: epi-quinine

Fig. (2). The four principal Cinchona-d\ka\oids and their stereochemistry

Chemistry and Pharmacology. Quinine acts as a blood schizonticide although it also has gametocytocidal activity against F. vivax and P, malariae. Its effect is probably because of its properties as a weak base. As a schizonticidal drug, it is less effective and more toxic than chloroquine. However, it has a special place in the management of severe P. falciparum malaria in areas with known resistance to chloroquine [164]. Quinine is readily absorbed when given orally or intramuscularly. Peak plasma concentrations are achieved within 1-3

838 hours after oral dose and plasma half-life is about 11 hours. In acute malaria, the volume of distribution of quinine contracts and clearance is reduced, and the elimination half-life increases in proportion to the severity of the illness [165]. Therefore, the maintenance dose of the drug may have to be reduced if the treatment is continued for more than 48 hours. The drug is extensively metabolised in the liver and only 10% is excreted unchanged in the urine. Quinine is also a potentially toxic drug. The typical syndrome of quinine side effects is referred to as cinchonism and severity is related to size of dose. Mild cinchonism consists of ringing in the ears, headache, nausea and disturbed vision. Functional impairment of the eighth nerve results in tinnitus, decreased auditory acuity and vertigo. Visual symptoms consist of blurred vision, disturbed colour perception, photophobia, diplopia, night blindness, and rarely blindness. These changes are due to direct neurotoxicity, although vascular changes may contribute to the problem. Rashes, sweating, angioedema can occur. Excitement, confusion, delirium are also seen in some patients. Coma, respiratory arrest, hypotension, and death can occur with over dosage. Quinine can also cause renal failure. Massive hemolysis and hemoglobinuria can occur, especially in pregnancy or repeated use. Quinine has little effect on the heart in therapeutic doses and hence regular cardiac monitoring is not needed. However it can cause hypotension in the event of overdose. Quinine reduces the excitability of the motor end plate and thus antagonises the actions of physostigmine. It can cause respiratory distress and dysphagia in patients of myasthenia gravis. Quinine is administered orally at 10 mg/kg 8 hourly for 4 days and 5 mg/kg 8 hourly for 3 days, intravenous at 20 mg of salt/kg in 10 ml/kg isotonic saline or 5% dextrose over 4 hours, then 10 mg of salt/kg in saline or dextrose over 4 hours, every 8 hours until the patient is able complete oral administration or for 5-7 days, and finally intramuscularly at 20 mg/kg stat, followed by 10 mg/kg 8 hourly by deep intra muscular injections for 5-7 days [165]. Mefloquine (133) was developed during the Vietnam war, during a programme to find new antimalarials, to protect American soldiers from the multidrug resistant P. falciparum infection. The Walter Reed Army Institute for Research started preclinical development of mefloquine in 1972, filed and started human studies in the same year. After a successful evaluation of the drug potential further

839 development was done in a cooperation with Hoffman LaRoche leading to a final FDA-approval in 1988. Since then, it has been used worldwide for the treatment [166] and prophylaxis [167] of P. falciparum malaria, known under the trade name Lariam®. The antiplasmodial activity and mechanism of action is unknown. It probably affects the membranes of the parasites. It is effective against the blood forms of P. falciparum malaria, including the chloroquine resistant types. Mefloquine is available for oral administration only. It is absorbed rapidly and is extensively bound to plasma proteins [166]. The elimination half-life is about 2-3 weeks. It is mainly excreted in the faeces. It is generally well tolerated in therapeutic doses up to 1,500 mg. Nausea, vomiting, abdominal pain and dizziness can occur in doses exceeding 1 g [166]. Less frequently it can cause nightmares, sleeping disturbances, dizziness, ataxia, sinus bradycardia, sinus arrhythmia, postural hypotension, and an 'acute brain syndrome' consisting of fatigue, asthenia, seizures and psychosis. It is given at 25 mg/kg in a single dose.

(133) mefloquine (Lariam )

6.

FUTURE DIRECTIONS - DOES NATURE PROVIDE LEADS FOR NEW ANTIPARASITICS?

The widespread opinion that parasitic diseases no longer pose a problem thanks to antibiotics and vaccines is wishful thinking rather than reality. Today we are further away from controlling parasitic diseases than we were 20 years ago. Parasitic protozoa remain a major

840 threat to the health of human population throughout the world. Despite this fact, there are few effective drugs for the treatment of many protozoal diseases. However, the therapies for malaria, leishmaniasis and trypanosomiasis, diseases that threaten more than two billion people in mostly underdeveloped countries, are inadequate. This is now being recognised in multinational programmes such as "Roll back malaria". Traditionally, medicinal plants have already provided valuable leads for potential antiparasitic compounds, including naphthoquinones, terpenoids and alkaloids. The renewed interest in plant products has been stimulated in part by the identification of the antiplasmodial activity of the sesquiterpene lactone artemisinin (qinghaosu). This experience is not to be ignored as plants have frequently provided the template molecules on which to base further novel structures. For more than fifty years important antimicrobial and antiparasitic drugs have been identified from the products of fungi and bacteria and we should look to this source for future novel leads. In recent years marine organisms have been actively investigated and basic information has been made available to evaluate their potential. Although discovery of antiparasitic active compounds was not expected in the first pharmacological studies, promising leads have already been identified with new chemical types and active principles. So what is the potential of natural products as a source of new antiparasitics? From the experiences with artemisinin (qinghaosu) it is wishful thinking to suppose that parasitic diseases can be treated with a single compound. Pure natural products are useful as lead structures, but in most cases high toxicity has restricted their use in humans. Many natural products with the desired activity and low toxicity have been identified, as reviewed here, but they did not progress through preclinical studies for evaluation as a potent drug because of low bioavailability and/or poor solubility. These pharmaceutical problems point towards the need for a rational, preferably mechanistically based, structural modification of chemical leads from nature. Natural products give new inputs to medicinal chemistry to develop new safe and effective drugs. By classical synthetic strategies organic chemists may create safer compounds close to lead to reduce toxicity, side effects or to improve bioavailability. This has been bone very successful for artemether from artemisinin and atovaquone from simple naphthoquinones.

841 Natural products have made an important contribution to antiparasitic drug research and despite all problems there is every indication that they will continue to make a contribution to the efforts to develop new and urgently needed drugs for the future.

References [I] [2] [3] [4] [5] [6] [7] [81 [9] [10] [II] [12] [13] [14] [15] [16] [17] [18] [19]

Mishra, S.K.; Satpathy, S.K.; Mohanty, S. Bull World Health Organ., 1999, 77, 1020 Wakelin, D. Immunity to parasites, Cambridge University Press, 1996,47 Salfelder, K. (eds). Protozoan infection in man, Schwer, Stuttgart, 1986, 99 Meshnick, S.R. Parasitol. Today, 1997, 13, 89-90 Ridley, R.G.; Hudson, A.T.. Curr. Opin. Infect. Dis., 1998, 11, 691 Meshnick, S.R. Med Trop. (Mars,) 1998, 58 (Suppl), 13 Looareesuwan, S.; Chulay, J.D.; Canfield, C.J.; Hutchinson, D.B. Am. J. Trop. Med. Hyg., 1999, 60, 533 White, N.J.; Nosten, F.; Looareesuwan, S.; Watkins, W.M.; Marsh, K.; Snow, R.W. Lancet, 1999, 353, 1965 Peters, W. J. R. Soc. Med, 1999, 92, 345 Berman, J. Curr. Opin. Infect. Dis., 1998, 11, 707 Alexander, J.; Satoskar, A.R.; Russell, D.G. / Cell Sci., 1999, 112 Alvar, J.; Canavate, C; Guiterrez-Solar, B.; Jimenez, M.; Laguna, F.; Lopez-Valez, R.; Molina, R.; Moreno, J. Clin. Microbiol. Rev., 1997, 10, 298 Croft, S.L. Parasitology, 1997, 114 Suppl, S3 Coukell, A.J.; Brogden, R.N. Drugs, 1998, 55, 585 Thakur, C.P., Singh, R.K., Hassan, S.M., Kumar, R., Narain, S., Kumar, A. Trans R Soc Trop MedHyg., 1999, 93, 319 Jha, T.K.; Sundar, S.; Thakur, C.P.; Bachmann, P.; Karbwang, J.; Fischer, C; Voss, A.; Berman, J. N. Engl J. Med, 1999, 341, 1795 World Health Organisation URL: http://www.who.int/emc/ diseases/tryp/index.html Salfelder, K. (eds). Protozoan infection in man, Schwer, Stuttgart, 1986, 21 Buckner, F.S.; Verlinde, C.L.M.J.; La Flamme, A.C.; Van Voorhis, W.C, Antimicrob. Agents. Chemother., 1996, 40, 2592

842 [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39]

Urbina, J.A.; Payares, G.; Contreras, L.M.; Liendo, A.; Sanoja, C; Molina, J.; Piras, M.; Piras, R.; Perez, N.; Wincker, P.; Loebenberg, D. Antimicrob. Agents Chemother., 1998, 42, 771 Barrett, M.P.; Mottram, J.C.; Coombs, G.H. Trends Microbiol 1999, 7,82 Desjardins, R.E.; Canfield, CJ.; Haynes, J.D.; Chulay, J.D. Antimicrob. Agents Chemother. 1979, 16, 710 O'Neill, MJ.; Bray, D.H.; Boardman, P.; Phillipson, J.D.; Warhurst, D.C. Planta Med., 1985, 394 Ter Kuile, R; White, NJ.; Holloway, P.; Pasvol,G.; Krishna, S.; Exp. Parasitol, 1993, 76, 85 Tan-Ariya, P.; Brockelman, C.R.; Menabandhu, C. American J. Trop. Med Hyg., 1987, 37, 42 Basco, L.K.; Marquet, F.; Makler, M.M.; Le Bras, J. Exp. Parasitol, 1995, 80, 260 Basco, L.K.; Le Bras, J. Parasitol. Res., 1994, 80, 262 Janse, CJ.; Waters, A.P. Parasitology Today, 1995, 11, 139-143 Peters, W.; Robinson, B.L. Malaria. In: Handbook of Animal Models of Infection. (Eds Zak.O.). Academic Press, London., 1999, 757 Berman, J.D.; Gallalee, J.V. Antimicrob. Agents Chemother., 1985, 28, 723 Kiderlen, A.F.; Kaye, P.M. J. Immunol Methods 1990, 127, 11 Bodley, A.L.; McGarry, M.W.; Shapiro, T.A. / Inf. Dis., 1995, 172, 1157 Onyango, J.D.; Burri, C; Brun, R. Acta Trop (Basel), 2000, 74:95100 Kar, S.; Kar, K.; Bhattacharaya, P.K.; Ghosh, D.K. Antimicrob. Agents Chemother., 1993, 37, 2459 Fevrier, A.; Ferreira, M.E.; Foumet, A.; Yaluff, G.; Inchausti, A.; Rojas de Arias, A.; Hoquemiller, R.; Waechter, A.-I. Planta Med., 1998, 64, 47 Sarkar, S.; Mukherjee, T.; Patra, P.; Bhaduri, A. Biochem. Biophys. Res. Commun. 1992, 182, 86 Perez, H.A.; De La Rosa, M.; Apitz, R. Antimicrob. Agents Chemother., 1994, 38, 338 Gallwitz, H.; Bonse, S.; Martinez-Cruz A.; Schlichting, I.; Schumacher, K.; Krauth-Siegel, L. J. Med. Chem., 1999, 42, 364 Urbina, J.A.; Marchan, E.; Lazardi, K.; Visbal, G.; Apitz-Castro, R.; Gil, F.; Aguirre, T.; Piras, M.M.; Piras, R. Biochem. Pharmacol. 1993, 45,2381

843

[40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60]

Koide, T.; Nose, M.; Inoue, M.; Ogihara, Y.; Yabu, Y.; Ohta, N. PlantaMed.,1997,63,21 Inuoe, M.; Suzuki, R.; Koide, T.; Sakaguchi, N.; Ogihara, Y.; Yabu, Y.; Biochem, Biophys. Res, Comm., 1994, 204, 898 Kayser, O.; Kolodziej, H.; Kiderlen, A.F. Phytotherap. Res., 2000, in press Kiderlen, A.F.; Kayser, O.; Kolodziej, H. Int. J. ImmunopharmacoL, 2000, submitted Minagawa, N.; Yabu, Y.; Kita, K.; Nagai, K.; Ohta, N.; Meguro, K.; Sakajo, S.; Yoshimoto, A. MoL Biochem. Parasitol, 1996, 84, 271 Goijman, S.G.; Turrens, J.F.; Marini-Bettolo, G.B.; Stoppani, A.O. Medicina (Buenos Aires), 1984, 44, 361 Decosterd, L.A.; Hoffmann, E.; Kyburz, R.; Bray, D.; Hostettmann, K. Planta Med., 1991, 57, 548 Sheagren, J.N. J. Parasitol, 1968, 54, 1250 Oketch-Rabah, H.A.; Lemmich, E.; Dossaji, S.F.; Theander, T.G.; Olsen, C.E.; Comett, C; Kharazmi, A.; Brogger Christensen, S. J. Nat. Prod.,1991, 60, A5% Lopes, N.P., Chicaro, P., Kato, M.J., Albuquerque, S., Yoshida, M., Planta Med., 1998, 64, 667-669 Bastos, J.K.; Albuquerque, S.; Silva, M.L. Planta Med., 1999, 65, 541 Zhai, L.; Blom, J.; Brogger Chrinstensen, S.; Kharazami, A. Antimicrob. Agents Chemotherap., 1995, 39, 2742 Feldbaeck Nielsen, S.; Brogger Christensen, S.; Cruciani, G.; Kharazami, A.; Liljefors, T. J. Med. Chem., 1998, 41, 4819 Kayser, O.; Kiderlen, A.F.; Folkens, U.; Kolodziej, H. Planta Med., 1999,316 Li, R.; Kenyon, G.L.; Cohen, F.E.; Chen, X.; Gong, B.; Dominguez, J.N.; Davidson, E.; Kurzban, G.; Miller, R.E.; Nuzum, E.O.; Rosenthal, P.J.; McKerrow, J.H. J. Med. Chem., 1995, 38, 5031 Khalid, S.A.; Farouk, A.; Geary, T.G.; Jensen, J.B. J. EthnopharmacoL, 1986, 15, 201 Elford, B.C.; Roberts, M.F.; Phillipson, Wilson, R.J.F. Trans. Soc. Trop. Med. Hyg., 1985, 81, 434 Elford, B.C. Parasit. Today, 1986, 2, 309 Chanphen, R.; Thebtaranonth, Y.; Wanauppathamkul, S.; Yuthavong, Y. J. Nat. Prod., 19, 61, 1146 Ribeiro, A.; Pilo-Veloso, D. / Nat. Prod., 1997, 60, 836 Perez-Victoria, J.M.; Chiquero, M.J.; Conseil, G.; Dayan, G.; Di Pietro, A.; Barron, D.; Castanys, S.; Gamarro, F. Biochemistry, 1999, 38,1736

844 [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82]

Gale, M. Jr.; Carter, V.; Parsons, M..J. Cell. ScL, 1994, 107 ( Pt 7), 1825 Wright, C.W.; Phillipson, J. D. Phyterap. Res., 1990, 4, 127 Sepuvelda-Boza, S.; Cassels, B.K. Planta Med., 1996, 62, 98 Foumet, A.; Barrios, A.A.; Munoz, V.; Hoquemiller, R.; Cave, A. Trop. Med. ParasitoL, 1992, 43, 219 Akendengue, B.; Ngou-Milama, E.; Laurens, A.; Hocquemiller, R. Parasite, 1999, 6, 3 Hooker, S.C. J. American Chem. Soc, 1936, 58, 1181 Hudson, A.T.; Lapinone, menoctone, hydroxyquinolinequinones and similar structures. In: Antimalarial Drugs volume n (Eds. Peters,W., Richards,W.H.G.), Springer-Verlag, Berlin, 1984, 343 Hudson, A.T.; Randall, A.W.; Ginger, CD.; Hill, B.; Latter, V.S.; IVlcHardy, N.; Williams, R.B. Parasitology, 1985, 90, 45 Hudson, A.T. Parasit. Today, 1993, 9, 66 Ray, S.; Hazra, B.; Mittra, B.; Das, A.; IVIajumder, H.K.; Mol Pharm., 1998, 54, 994 Solis, P.N.; Langat, C; Gupta, IVl.P.; Kirby, G.; Warhurst, D.C.; Phillipson, J.D. Planta Med, 1995, 61, 62 Schnur, L.; Bachrach, U.; Bar-Ad, G.; Haran, M.; Tashma, Z.; Talmi, Z.; Kathendler, J. Biochem. Pharmacol, 1983, 32, 1729 Obih, P.O.; IVIakinde, M.', Laoye, O.J. Af. J. Med. Medical. Sci., 1985, 14, 45 Kaumaglo, K.; Gbeassor, M.; Nikabu, O.; De Souza, C; Werner, W. Planta Med., 1992, 58, 533 Sittie, A.A.; Lemmich, E.; Olsen, C.E.; Hviid, L.; Kharazami, A.; Nkrumah, F.K.; Brogger Christensen, S. Planta Med, 1999, 65, 259 Majumdar, U.K.; Gupta, M.; Chowdhury, S.; Saha, A.K. Indian J. Exp. BioL, 1993, 31, SSS Likhitwitayawuid, K.; Phadungcharoen, T.; Krungkrai, J. Planta Med., 1998, 64, 70 Abu el Heiga, L.A.; Katzhendler, J.; Gean, K.F.; Bachrach, U. Biochem. Pharmacol, 1990, 39, 1620 Ray, S.; IVIajunder, H.K.; Chakravarty, A.K.; IVIukhopa-Dhyay, S. /. Nat. Prod, 1996, 59, 27 Bodley, A.L.; Shapiro, T.A. Proc. Natl Acad ScL USA, 1995, 92, 3726 Tandon, J.S.; Srivastava, V.; Guru, P.Y. J. Nat. Prod, 1991, 54, 1102 Castro, C; Jimenez, M.; Gonzalez-De-Parra, M. Planta Med., 1992, 58,281

845 [83]

Hocquemiller, R.; Cortes, D.; Arango, G.J.; Myint, S.H.; Cave, A. J. Nat. Prod., 1991, 54, 445 [84] Pieman, A.K.; Rodriguez, E.; Towers, G.H. Chem. Biol. Interact., 1979, 28, 83 [85] Francois, G.; Passreiter, CM.; Woerdenbarg, HJ.; Van Looveren, M. Planta Med., 1996, 62, 126 [86] Koshimizu, K.; Ohigashi, H.; Huffman, M.A. Physiol. Behav. 1994, 56, 1209 [87] Konig, G.M.; Wright, A.D.; Sticher, O.; Jurcic, K.; Offermann, F.; Redl, K.; Wagner, H.; Angerhofer, C.K.; Pezzuto, J.M. Planta Med., 1991, 57 Supplement, Issue 2, A5 [88] Schmeda-Hirschman, G.; Razmilic, L; Sauvain, M.; Moretti, C; Munoz, v.; Ruiz, E.; Balanza, E.; Foumet, A. Phytotherap. Res., 1996, 10, 375 [89] Kayser, O.; Kiderlen, A.F. In Proceedings of the 9^^ International Congress of Parasitology, Monduzzi Editore: Bologna, Italy, 1998, 925-929 [90] Richomme, P.; Godet, M.C.; Foussard, F.; Toupet, L.; Sevenet, T.; Bruneton, J. Planta Med, 1991, 57, 552 [91] Batista, R.; Chiari, E.; de Oliveira, A.B. Planta Med., 1999, 65, 283 [92] Alves, T.M.; Chaves, P.P.; Santos, L.M.; Nagem, T.J.; Murta, S.M.; Ceravolo, LP.; Romanha, A.J.; Zani, C.L. Planta Med, 1995, 61, 85 [93] Konig, G.M.; Wright, A.D.; Sticher, O.; Fronczek, F.J. J. Nat. Prod., 1992, 55, 633 [94] Konig, G.M.; Wright, A.D.; Angerhofer, C.K.; Greenidge, P.; Linden, A.; Desqueyroux-Faundez, R. J. Nat. Prod., 1996, 59, 710 [95] Konig, G.M.; Wright, A.D.; Angerhofer, C.K. /. Org. Chem., 1996, 61, 3259 [96] Miyaoka, H.; Shimomura, M.; Kimura, H.; Yamada, Y. Tetrahedron, 1998, 54, 13467 [97] Isaka, M.; Punya, J.; Lertwerawat, Y.; Tantcharoen, M.; Thebtaranonth, Y. J. Nat. Prod., 1999, 62, 329 [98] Oketch-Rabah, H.A.; Brogger Christensen, S.; Dossaji, S.F.; Theander, T.G.; Comett, C; Watkins, W.M.; Lemmich, E. Planta Med., 1998, 64, 559 [99] Bringmann, G.; Saeb, W.; Ake Assi, L.; Francois, G.; Sankara Narayanan, A.S.; Peters, K.; Peters, E.-M. Planta Med., 1997, 63, 255 [100] Majester-Sarvonin, B.; Elias, R.; Diaz-Lanza, A.M.; Balansard, G.; Gasquet, M.; Delmas, F. Planta Med., 1991, 57, 260

846 [101] Oketch-Rabah, H.A.; Dossaji, S.F.; Christensen, S.B.; Frydenvang, K.; Lemmich, E.; Comett, C; Olsen, C.E.; Chen, M.; Kharazmi, A.; Theander, T. J. Nat. Prod., 1997, 60, 1017 [102] Rochanakij, S.; Thebtaranonth, Y.; Yenjai, C; Yuthavong, Y. Southeast. Asian. J. Trop. Med. Public Health, 1985, 16, 66 [103] Khalid, S.A.; Duddeck, H.; Gonzalez-Sierra, M. J. Nat. Prod., 1989, 52, 922 [104] Badam, L.; Deolankar, R.P.; Kulkami, M.M.; Nagsampgi, B.A.; Wagh, U.V. Indian J. Malariol, 1987, 24, 111 [105] McKinnon, S.; Durst, T.; Amason, J.T.; Angerhofer, C; Pezzuto, J.; Sanchez-Vindas, P.E.; Poveda, LJ.; Gbeassor, M. /. Nat. Prod., 1997, 60, 336 [106] Wright, C.W.; ONeili, MJ.; Phillipson, J.D.; Warhurst, D. Antimicrob. Agents Chemother., 1988, 32, 1725 [107] Wright, C.W.; Anderson, M.M.; Allen, D.; Phillipson, J.D.; Kirby, G.C.; Warhurst, D.C.; Chang, H.R. J. Eukaryot. Microbiol, 1993, 40, 244 [108] Cabral, J.A.; McChesney, J.D.; Milhous, W.K. J. Nat. Prod., 1993, 56, 1954 [109] Robert-Gero, M.', Bachrach, U.; Bhatnagar, S.; Polonsky, J. C R. Acad. Sci. Paris, 1985, t. 300, H, 803 [110] Fandeur, T.; Moretti, C; Polonsky, J. Planta Med., 1985, 51, 20 [111] Kirby, G.C.; ONeill, MJ.; Phillipson, J.D.; Warhurst , D.C. Biochem. Pharm., 1989, 38, 4367 [112] Francois, G.; Diakanamwa, C; Timperman, G.; Bringmann, G.; Steenackers, T.; Atassi G.; Van Looveren, M.; Holenz, J.; Tassin, J.P.; Assi, L.A.; Vanhaelen-Fastre, R.; Vanhaelen, M. Int. J. Parasitol, 1998, 28, 635 [113] Chataing, B.; Concepcion, J.L.; Lobaton, R.; Usubillaga, A. Planta Med., 199^,6^,21 [ 114] Kapil, A. Planta Med., 1993, 59,474 [115] Konig, GM.; Wright, A.D.; Linden, A. Planta Med., 1998, 64, 443 [116] Foley, M.; Tilley, L. Int. J. Parasitol, 1997, 27, 231 [117] Foumet, A.; Hocquemiller, R.; Roblot, F.; Cave, A.; Richomme, P.; Bruneton, J. J. Nat. Prod., 1993, 56, 1547 [118] Gantier, J.C.; Foumet, A.; Munos, H.M.; Hocquemiller, R. Planta Med, 1996, 62,285 [119] Foumet, A.; Ferreira, M.E.; Rojas De Arias, A.; Torres De Ortiz, S.; Fuentes, S.; Nakayama, H.; Schinini, A.; Hocquemiller, R. Antimicrob. Agents Chemother., 1996, 40, 2447 [120] Lavaud, C; Massiot, G.; Vasquez, C; Moretti, C; Sauvain, M.; Balderrama, L. Phytochemistry, 1995, 40, 317

847 [121] Vennerstrom, J.L.; Lovelace, J.K.; Waits, V.B.; Hanson, W.L.; Klayman, D.L. Antimicrob. Agents Chemother., 1990, 34, 918 [122] El-On, J.; Jacobs, G.P.; Weinrauch, L. Parasit, Today, 1988, 4, 76 [123] Queiroz, E.F.; Roblot, R; Cave, A.; Paulo M.Q.; Foumet, A. /. Nat, Prod,, 1996, 59, 438 [124] Francois, G.; Timperman, G.; Eling, W.; Assi, L.A.; Holenz, J.; Bringmann, G. Antimicrob, Agents Chemother., 1997, 41, 2533 [125] Hallock, Y.F.; Cardellina, J.H. 2°^; Schaffer, M.; Bringmann, G.; Francois, G.; Boyd, M.R. Bioorg. Med. Chem. Lett., 1998, 8, 1729 [126] Foumet, A.; Munoz, V.; Manjon, A.M.; Angelo, A.; Hocquemiller, R.; Cortes, D.; Cave, A.; Bruneton, J. J. Ethnopharmacol, 1988a, 24, 327 [127] Foumet, A.; Manjon, A.M.; Munoz, V.; Angelo, A.; Bmneton, J.; Hocquemiller, R.; Cortes, D.; Cave, A. J. Ethnopharmacol, 1988b, 24, 337 [128] Patridge, S.J.; Russell, P.F.; Kirby, G.C.; Bray, D.H.; Warhurst, D.C; Phillipson, J.D.; ONeill, M.J.; Schiff Jr., P.L. J. Pharm. Pharmacol, 1988, 41 (Suppl.), 53P [129] Angerhofer, C.K.; Guinaudeau, H.; Wongpanich, V.; Pezzuto, J.M.; Cordell, G.A. J. Nat. Prod, 1999, 62, 59 [130] Mahiou, V.; Roblot, F.; Hocquemiller, R.; Cave, A. J. Nat. Prod, 1994, 57, 890 [131] Waechter, A.-L; Cave, A.; Bories, C; Munoz, V.; Foumet, A. Phytother. Res., 1999, 13, 175 [132] Bodley, A.L.; McGarry, M.W.; Shapiro, T.A. J. Infect. Dis., 1995, 172,1157 [133] Bodley, A.L., Shapiro, T.A. Proc. Natl Acad. Scl USA, 1995, 92, 3726 [134] Ray, S.; Sadhukhan, P.K.; Mandal, N.B.; Mahato, S.B.; Majumder, M.K. Biochem. Biophysic. Res. Communicat, 1997, 230, 171 [135] Evans, A.T.; Croft, S.L. Phytother. Res., 1987, 1, 25 [136] Benard, J.; Dat-Xuong, N.; Riou, G. C R. Acad. Set Hebd. Seances Acad Scl D., 1975, 280, 1177 [137] Cimanga, K.; De Bruyne, T.; Pieters, L.; Vlietnick, A.J. J. Nat. Prod, 1997, 60, 6SS [138] Ono, T.; Nakabayashi, T. Biken. J., 1979, 22, 117 [139] Dieckmann-Schuppert, A.; Franklin, R.M. Cell Biol Int. Rep., 1989,13,411 [140] Munoz, v.; Moretti, C ; Sauvain, M.; Caron, C ; Porzel, A.; Massiot, G.; Richard, B.; Le Men-Olivier, L. Planta Med., 1994, 60, 455 [141] Nolan, L.L. Antimicrob. Agents Chemother., 1987, 31, 1542

848 [142] Yakubu, M.A.; Majumder, S.; Kierszenbaum, F. J. Parasitol, 1993, 79, 525 [143] Goldberg, B.; Rattendi, D.; Lloyd, D.; Sufrin, J.R.; Bacchi, CJ. Biochem. Pharmacol, 1998, 56, 95 [144] Kellina, O.I. Meditsinisakya Parasitologya i Parataryne Bolezni, 1963, 32, 572 [145] Neal, R.A. Ann. Trop. Med. Parasitol, 1968, 62, 54 [146] Grogl, M.\ Schuster, B.C.; Ellis, W.Y.; Berman, J.D. J. ParasitoU 1999, 85, 354 [147] Neal, R.A.; Murphy, A.G.; Olliaro, P.; Croft, S.L. (1995) Trans. R. Soc. Trop. Med. Hyg., 1995, 88, 223 [148] Maarouf, M.; Lawrence, F.; Croft, S.L.; Robert-Gero, M. Parasitol. Res., 1995, 81,421 [149] Klayman, D.L. Science, 1985, 228, 1049 [150] Luo, X.-D.; Shen, C.C. Med. Res. Rev., 1987, 7, 29 [151] Woerdenbag, H.J.; Pras, N.; van Uden, W.; Wallaart, T.E.; Beekman, A.C.; Lugt, C.B.; Pharm World Sc, 1994, 16, 169 [152] van Agtmael, M.A.; Eggelte, T.A.; van Boxtel, C.J. Trends Pharmacol. ScL, 1999, 20, 199 [153] Ziffer, H.; Highet, R.J.; Klayman, D.L. Fortschr. Chem. Org. Naturst., 1997, 72, 121-214 [154] CCRG-Chinese Cooperative Research Group / Trad. Chin. Med., 1982,2,31 [155] Titulaer, H.A.; Zuidema, J.; Kager, P.A.; Wetsteyn, J.C; Lugt, C.B.; Merkus, F.W. J. Pharm Pharmacol, 1990, 42, 810 [156] Dayan, A.D. Med Trop. (Mars), 1998, 58 (Suppl.), 32 [157] Brun-Pascaud, M.; Chau, F.; Derouin, F.; Girard, P.M. Parasite, 1996, 3, 187 [158] Merali, S.; Meshnick, S.R. Antimicrob. Agents Chemother., 1991, 35, 1225 [159] Berens, R.L.; Krug, E.G.; Nash, P.B.; Curiel, T.J. J. Infect. Dis., 1998,177,1128 [160] Hobhouse, H. Seeds of Change. Harper and Rov^, Nev^ York, 1985 [161] Greenwood, D. J Antimicrob Chemother., 1992, 30, 417 [162] Meshnick, S.R. Parasitol Today, 1997, 13, 89 [163] Woolfe, G. Chemotherapy of Amoebiasis. In: Experimental Chemotherapy Vol 1 (eds. Schnitzer,R.J., Hawking,F.), Academic Press, London, 1963, 355 [164] Foley, M.; Tilley, L. Pharmacol Ther., 1998, 79, 55 [165] Krishna, S.; White, N.J.Clin. Pharmacokinet, 1996, 30, 263 [166] Palmer, K.J.; Holliday, S.M.; Brogden, R.N. Drugs, 1993,45, 430 [167] Schlagenhauf, P. J. Travel Med, 1999, 6, 122

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

849

NATURAL INSECTICIDES: STRUCTURE DIVERSITY, EFFECTS AND STRUCTURE-ACTIVITY RELATIONSHIPS. A CASE STUDY

A. GONZALEZ-COLOMA ^\ M. REINA^ C. GUTIERREZ\ B.M. FRAGA^ ^Centro de Ciencias Medioambientales, CSIC, Serrano 115-dpdo., 28006 Madrid, Spain. Institute de Productos Naturales y Agrobiologia, CSIC, Avda. Astrofisico F. Sanchez, 38206 La Laguna, Tenerife, Spain. Mailing address, e-mail: [email protected], telephone: 34-91-5625020, fax: 34-915640800 ABSTRACT: For the past ten years our research group has been working on bioactive natural products with insecticidal effects. We have focused mainly on the flora of the Canarian Archipelago because of its rich biodiversity (Lauraceae, Compositae, Boraginaceae). Additionally, we have carried out comparative studies of species belonging to these plant families from different geographical origins (Japan, Chile). As a result of these studies we have isolated bioactive compounds belonging to several chemical classes (sesquiterpenes, diterpenes, lignans, diterpenoid alkaloids, pyrrolizidine alkaloids) with selective modes of action and low toxicity. The structure-activity relationships of these compounds have also been established. In this chapter we will present the structures and biological effects of these compounds according to their chemical classes as follows: Diterpene ryanodanes and isoryanodanes from Lauraceae species Lignans from Lauraceae species Sesquiterpenes from Compositae species Diterpene alkaloids from Ranunculaceae species Pyrrolizidine alkaloids from Compositae and Boraginaceae species

INTRODUCTION During the past decades the excessive use of synthetic pesticides has given raise to several problems including the selection of resistant pest

850 populations [1], negative side effects on beneficial parasites and predators and deleterious effects on human health [2]. Therefore, the search for alternative pest control agents with reduced environmental consequences is of great interest [3]. In this context we are studying the phytochemistry of the flora of the Canarian Archipelago [4], searching for environmentally-friendly pesticides. As a result of this ongoing study we have identified several bioactive compounds belonging to different chemical classes. These compounds have been isolated from plant species of the families Lauraceae, Boraginaceae, Ranunculaceae and Compositae. Species of the Lauraceae and Boraginaceae families endemic to Japan and Chile have also been included in this study for comparison purposes. These results emphasize the potential added value of local biodiversity given its wide range of possible agrochemical applications. Diterpene ryanodanes and isoryanodanes from Persea indica (Lauraceae) The Canarian laurel forest is a relic of the Tertiary Mediterranean flora. This forest has suffered from human activity, and is nowadays restricted to small protected areas of the Canarian Archipelago. The uniqueness of the this forest [5] prompted us to investigate the biological value of the dominant species {Laurus azorica Seub., Ocotea foetens Alton, Apollonias barbujana Cay, y Persea indica L.) [6]. Following an insecticidal activity screening, we found that P. indica was a strong antifeedant against several insect species in choice assays (Spodoptera litura, S. littoralis, Macaronesia fortunata, Heliothis armigera, Leptinotarsa decemlineata andMyzus persicae) (table 1). P. indica also inhibited the growth of the lepidopteran species tested (table 1), suggesting a broad range of insecticidal action and emphasizing the phytochemical importance of this plant. The chemical study of the antifeedant fractions of P. indica resulted in the isolation of the following ryanodane-type diterpenes: Ryanodol (1), isolated for the first time as a natural product, the known ryanodanes cinnzeylanol (3) and cinnzeylanine (4) [7, 8] and the new ones cinnzeylanone (6), ryanodol-14-monoacetate (2) and epi-cinnzeylanol (5) [9, 10]. In addition to ryanodanes 1-6, the new isoryanodane type diterpenes indicol (9), vignaticol (7) and perseanol (8) were also isolated [11].

851

Table 1. Antifeedant and growth inhibition effects of P, indica ethanolic extract on several insect species Insect Family

Insect species

FI/Sl'(100|^g/cm^)

% Growth inhibition^

Lepidoptera

1

S. VUura

100

26

S. lUtoralis

100

83

M.fortunata

25

H. armigera

43

Coleoptera

L. decemlineata

93

Homoptera

M. persicae

47

-

1

' %FI/%SI= [1-(T/C)]xl00, where T= consumption/settling of treated disks and C = consumption /settling of control disks. ^ Larval growth inhibition observed in diet-incorporation bioassays (n=20, 0.1%w/wt)

1 R=H 2 R = Ac

3 R=a-OH, H 4

R=a-OAc,H

5 R=p-OH. H 6 R= 0

7 R=:H 8 R = OH

852 Ryanodine-type compounds act primarily at the Ca^^ release channel in both mammals and insects [12]. Ryanodol-type compounds, however, are more selective toxicants for insects than they are for mammals, suggesting a different mode of action for these compounds [13, 14]. To further asses such hypothesis we carried out a comparative study on the antifeedant and insecticidal effects of the ryanodol/isoryanodol-type diterpenes from P. indica (non-alkaloidal type ryanoids 1-9) [9, 10, 11] and ryanodine-type (alkaloidal type) ryanoids (ryanodine/spiganthine ryanoids 10-18), isolated from Spigelia anthelmia (Loganiaceae) [15, 16]. We studied their effects on the feeding behavior, survivorship and performance (biomass gain and food ingestion) ofS. littoralis larvae and L. decemlineata (Colorado potato beetle, CPB) aduhs [17].

^OH

H

10 R1=CH3

R2 = H

R3 = 0 H

11 R1=CH3

R 2 = R3 = H

O

14

12 R1 =CH20H R2 = R3 = H 13 Rl=CH20H R2 = 0 H R3 = H

15 R=CH3 16 R=CH20H

17

853

OH

18

The activity of the test compounds varied depending upon the insect species and type of treatment. In general, the ryanodol/isoryanodol diterpenes are more effective antifeedants and less toxic than the ryanodine/spiganthine ones. Epicinnzeylanol (5), cinnzeylanine (4) and the epoxide 15 were the most promising molecules against S, littoralis, with strong antifeedant (4 and 5, table 2) and/or postingestive effects (5 and 15) Figure (1). Cinnzeylanone (6) was the most active against L decemlineata, with antifeedant and knock-down effects as well as oral toxicity (tables 2 and 3) [17]. The structure-activity study of the ryanoids showed that both C-1 and C14 substituents play an important role in their antifeedant and toxic activity against S. littoralis, as previously shown for their antifeedant effects on S. litura [10]. The (i-stereochemistry at C-1 (5 versus 3) and its O- acetylation (4), increased the toxic and/or antifeedant activity of these compounds. Hydroxylation (1), O- acetylation (2) and pyrrolcarboxylate-esterification (10) at C-14 along with the hydrophobicity of the cyclohexane ring (6) resulted in intermediate activities. Additionally, the presence of a C-2/C-3 epoxide (15,16) increased the toxicity against this lepidopteran (table 2, Figure (1)) [17].

854

S O

U

0^

1 1—\—i \ 10 11 12 13 14 15

r 16 18

Compound Figure (1). Relative consumption (RCR) and growth (RGR) rates of 5". littoralis L6 larvae orally injected with 10 fig of compounds 1-18 with or without PBO pre-treatment (5 jig). Data is expressed as % of control (average+SE). *Denotes a significant difference from the control, p<0.05 LSD test [17].

As is the case with S. litura and S. littoralis, both substituents at C-1 and C-14 are responsible for the antifeedant and toxic effects of the ryanoids on L, decemlineata. However, in this case, the presence of a ketone group at C1 (6) produced strong antifeedant and toxic effects, while the acetylation/hydroxylation of C-1 (4, 3, 5) or C-14 (1, 2) also produced antifeedant and toxic effects on CPB (tables 2 and 3). The presence of a pyrrolcarboxylate ester group at C-14 did not increase any of these biological effects. However, hydroxylation of C-2 (14) and the presence of an epoxide (C-2/C-3,15 and C2/C5,18) conferred antifeedant and postingestive effects respectively (tables 2 and 3) [17].

855 Table 2. Average antifeedant effect (%FI) ± standard error, and effective antifeedant doses (EC50) of the test compounds against S, littoralis sixth-instar (L6) larvae and L, decemlineata adults in choice assays [17]. L. decemlineata

S. littoralis FR(%)^

EC50 ( 9 5 % C L ) '

FR(%)'

EC50 ( 9 5 % C L ) '

(10 ^ig/cm^)

(nmol/cm^)

(10 )ig/cni^)

(nmol/cm^)

1

87.54+10.81

0.52(0.25,1.07)

37.08±13.57

2

90.84+10.81

0.63(0.09,4.41)

78.20+18.08

0.57 (0.03,9.77)

3

96.93+10.81

3.17(0.78,13.02)

61.27+11.05'

4

100.00+10.81

0.01(3.0xlO-\7.0xlO-^)

42.24^+13.98'

-

Compound

5

85.94+10.81

5.9x10"^ (5.7x10-^,0.13)

70.49+14.05

1.87(0.82,26.30)

6

89.11±10.81

1.46(0.21,10.21)

80.16+13.31'

0.22(0.01,2.88)

7

72.87+10.81

3.75 (0.35,38.37)

55.59+13.57

n.a.

8

89.03+10.81

0.44(0.18,1.12)

74.46+13.03

0.63(0.12,3.31)

9

85.54+10.81

8.48 (2.38,30.44)

29.83+13.57

-

10

84.89+10.81

0.93(0.42,1.96)

54.82+9.42'

0.77(0.25,3.71) n.a.

12

57.23+10.81 58.87+10.81

-

57.37+13.57'

13

46.76+13.57

-

14

84.70+10.81

1.27(0.33,4.91)

76.30+13.57

n.a.

15

43.10+10.81

-

28.53+13.57

16

n.a.

n.a.

15.78+13.57

17

85.72+10.81

2.72(9.15,28.00)

30.53+17.52

18

33.67+12.09.

-

32.19+13.57

-

AZA SILPH'

0.7X10-'(0.3X10-',1X10-^) 0.72(0.42,1.23)

|

^ %FI as in table 1 ^ Effective dose (EC50, dose needed to produce a 50% feeding inhibition) and 95% Confidence limits (lower, upper). 'From [32]. "^Knock-down-like effects at 10 |ig/cm^ (%FI values for 2 |Lig/cm^). na, not enough compound available.

Among the isoryanodane-type diterpenes, hydroxylation of C-1 (8) and the polarity of the five-membered A ring (C-8 and C-13) (7 versus 9) determined their antifeedant activity on both insect species (table 2) [17], as previously demonstrated for S, litura [11].

856 Table 3. Hemolymph injection (Inj, lO^g, n=20) and no-choice ingestion (Ing, lOjig, n=10) effects of the test compounds on L. decemlineata adults (% mortality after 48h) without and with synergist (PBO) pre-treatment (5|ig) [17]. Compound 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Inj 54* 16 0 0 0 0 0 0 0 0 0 8 0 15 0 0 0 0

% Mortality' PBO+Inj Ing 42* 52* 42* 61** 22 24 3 22 13 13 n.a. 18 0 3 32* 0 3 2

0 0 0 0 0 40* n.a. 0 0 10 n.a. 10 0 0 0 0 0 0

PBO+Ing 0 2 0 0 0 78** n.a. 0 0 0 n.a. 0 0 2 0 2 0

0

j

1

Corrected according to Abbott [70]. * Denotes a significant difference from the control. Fisher's Exact test for 2x2 contingency tables (*p<0.05; **p<0.005). na, not enough compound available.

Recent studies have shown that the injected knock-down potencies of the alkaloidal ryanoids were generally related to their effectiveness in competing with [^H]-ryanodine at the ryanodine receptor (RYR) of rabbit skeletal muscle. However, ryanodol (1) and didehydroryanodol were found to be more toxic than predicted from their potency at the RYR and may therefore act in a different manner such as the K"^ channel [13, 14]. Our results demonstrated a selective and broader insect antifeedant and toxic action for the ryanodol/isoryanodol-type compounds. Moreover, in this study we have revealed that more than 60% of the non-alkaloidal ryanoids were antifeedants and/or toxicants in contrast to 30% of active alkaloidal ones. Available evidence suggests that a ligand-protein receptor interaction is mediating insect taste transduction [18, 19]. The observed selective

857 antifeedant action of the non-alkaloidal ryanoids suggests a common ligandgated ion channel mediated taste response to these compounds, with the lepidopteran being more sensitive than the chrysomehd beetle, as shown for aconitine, a sodium channel antagonist diterpene alkaloid [20]. Our results support the proposed presence of a ryanodol-type receptor in insects and they indicate that some of the tested compounds (e.g. 4, 5, 6 and 15) are promising leads for a potential new generation of target-oriented insecticides. However, further research is needed in order to characterize this receptor and also to asses the mammalian toxicity and the receptorbinding affinity of these interesting compounds. Lignans from Machilus japonica (Lauraceae) Continuing with the study of bioactive constituents of members of the Lauraceae family, we conducted an investigation of the anti-insect activity of Lauraceae species endemic to Japan. Following an insecticidal-activity screening of Japanese Lauraceae against several pest insect species, we found that a crude extract oi Machilus japonica foliage inhibited the growth of 5". litura larvae [21]. A bioassay-guidedfi^actionationof M. japonica afforded the following bioactive Hgnans: Licarin A (19), the new lignan [(2R,3S,4R,5R)-2-(3,4-dimethoxyphenyl)-3,4-dimethyl-5-piperonyltetrahyd rofuran] (20), (-) Machilusin (21) and [(2S,3S)-2,3-dihydro-7-methoxy-3-methyl-2-(3,4-dimethoxyphenyl)-5-trans - (1-propenyl)- benzofuran] (22), also a new natural product. The comparative insect growth reduction activity of compounds (19-22) is shown in table 4. Compound 22 was the most active one, followed by the other three with a similar effect (table 4, [22]).

OCH.

19

Rl

R2

CH3

0CH3

22 OCH3

CH3

R3 OH OCH3 OCH3 OCH3 21

858 Table 4. Effective growth-inhibition dose of M japonica pure lignans against neonate iS". litura larvae in a seven-day diet incorporation bioassay [21]. Compound

ECso (95% confidence limit)*

19

0.20(0.10,0.5)

20

0.24(0.13,0.65)

21

0.19(0.11,0.45)

22

0.13(0.08,0.28)

Effective dose(%w/wt) to produce a 50% larval growth inhibition and 95% confidence limits (lower, upper).

Lignans are abundant in the plant kingdom and have a broad range of biological activities. Among the ones with anti-insect action are podophyllotoxin analogues, effective insect growth inhibitors; sesamin and sesamolin with weak juvenile hormone activity; p-benzolactone, an insect feeding inhibitor [23], neolignans such as magnolol and a biphenyl ether toxic to non-adapted insects [24], with lignans 19-22 being described in [22] as insect growth inhibitors for the first time. In the course of our experiments we also observed a decrease in the insect growth-reduction activity from the crude extract to the semi-purified fractions or the pure components (table 4). We proposed a potential synergistic action of either all or some of these compounds along with other plant components present in the crude extract of M. japonica, since the biological activity of the fractions containing the four lignans together was similar to the activity of each one individually and significantly lower than the activity of the crude extract [22]. Plants include synergists in their chemical defenses such as mixedfunction oxidase (PSMO) inhibitors [25]. M. japonica contains other lignans such as galbacin, galbegin, licarin B, calopeptin and veragensin [26] as well as many other components that have not been investigated and could account for the overall insect growth inhibition activity of the plant extract. Furthermore, an isomer of licarin A (dehydrodiisoeugenol) has been isolated from Myristicafragans, along with licarin B, as an inhibitor of hepatic PSMO enzymes [27, 28]. Therefore, a synergistic action is proposed for

859

either all or some of these lignans against S. litura when ingested by this insect along with other plant metabolites. Sesquiterpenes from Senecio palmensis (Compositae) The subalpine zone of Tenerife (450-1800 m) includes the pine forest dominated by the endemic Pinus canariensis [4]. This pine forest is an open community with several shrub species including endemic senecios characterized by their content in sesquiterpenes, sesquiterpene lactones and pyrrolizidine alkaloids [29-31]. We have carried out a survey of the pesticidal activity of several of these endemic Senecio species (Compositae). Senecio palmensis Chr. Sm. Proved to be a strong antifeedant against L. decemlineata and M persicae (table 5, unpublished results). Table 5. Antifeedant effects of 5. palmensis ethanolic extract on several insect species. Insect species

FI^ choice

S, littomlis

19

L. decemlineata

85

FI* no-choice

%C^

%T'

61

39*

70

M. persicae

j

' %FI as in table 1. ^ Percent aphids settled on control (C) and treated (T) disks. * Significant difference, p<0.05, Mann-Whitney U-Test.

We proceeded with the bioassay-directed chemical study of this plant. From the active fractions we isolated bisabolene- and silphinene-type sesquiterpenes (23 and 24). These compounds are strong CPB antifeedants in both choice and no-choice assays, the adult insects being the most sensitive to their action (table 6) [32]. Their antifeedant potencies were similar (23) or stronger than that of limonin (24), a CPB antifeedant triterpene [33].

0

23

14

OR

24

R1 Ac

R2 Ang

25

H

Ang

26

H

H

860 Table 6. Relative antifeedant potencies of compounds 23,24 and juglone (ju) against second (L2) and fourth (L4) instar larvea and adult £. decemlineata [32]. Compound

Test'

EC50 (95% CL) L4

L2

Adults

23

A B

20.34(17.03,24.29)' 35.21(21.0,59.05)

4.00(2.73,5.86) 29.86(26.55,33.6)

14.64(3.55,60.40) 11.66(1.65,82.04)

24

A B

12.55(4.87,32.34) 30.90(16.62,57.45)

1.69(1.36,2.11) 14.43(3.01,69.14)

0.27(0.16,0.46) 2.67(1.38,5.18)

0.83 (0.25,2.80) 20.41 (16.93,24.59) Test A, choice; test B, no-choice. Upper and lower 95% confidence intervals of EC50 (|Lig/cm^). A B

ju

1.5

1

0.32(0.06,1.81) 3.14(0.04,9.42)

1

1

1

• V

^ ^

1.0 L

T

10 ng

•

25 Jig 50 \ig 75 Jig

•

• •

o

•

•

•

• • ^^

0.5

> A

•

100 ^g

e

o

-^

T

• 0.0 h^

i •

^ • -0.5 -0.5

1

0.0 1

0.5

1.0

'

1

1

1.5

2.0

2.5

Relative Consumption Rate

Figure (2). Plot of the relative growth rate (RGR) on the relative consumption rate (RCR) for L. decemlineata L4 larvae fed for 24 h on leaf-disks treated with compound 23. The line represents the calibration curve (y= 0.371x-0.220, r=0.84, p<0.0001) [32].

861 To study the mode of action of these sesquiterpenes, fourth-instar (L4) CPB larvae were fed with increasing concentrations of the test compounds and their nutritional indices (relative consumption and growth rates, RCR and RGR) were measured 24h later. These results were compared with those obtained from L4-larvae subjected to increasing levels of starvation (starved control). This study showed that bisabolene 23 reduced RCR and RGR with respect to the control, with growth efficiencies (GE, slope of the regression of RGR on RCR), Figure (2), similar to that of the starved control, indicating that 23 acted as an antifeedant without postingestive effects [32]. Compound 23 also interfered with the settling behavior of the aphid M. persicae an efficient plant virus transmitter. Choice assays showed that this compound had antifeedant potencies slightly lower than the positive control famesol (table 7). Furthermore, the electronic monitorization of the aphid's probing behavior (EPG) showed that 23 reduced the number of intracellular penetrations (table 8) [34]. This is an important parameter when considering the transmission of non-persistent plant viruses by this insect [35]. Table 7. Myzus persicae settling behavior inhibition (%SI) caused by the monoterpene 2,10-bisaboladien-l-one (23) in choice assays. Effective doses (EC50) and geraniol (G), the sesquiterpene precursor farnesol (F) and the sesquiterpene 95% confidence limits (lower and upper) of the test compounds [34]. 1 2 1 Dose (ng/cm ) 0 10 ! 30 60 100 EC50 (95% C.L.)

1 G 21.15+7.38 12.77+4.95 28.86+7.08 51.64+8.92 79.6+6.61 47.13 (22.16,100.25)

%SI (SE) F 21.15+7.38 39.54+10.09 67.17+10.45 90.57+3.04 97.14+1.67 14.91 (12.13,18.33)

23 21.15+7.38 28.99+12.87 63.09+7.28 79.48+8.29 87.86+4.33 20.28 (16.40,25.07)

%SI= l-(%T/%C)xlOO, where %T=percent aphids on treated disk after 24h and %C=percent aphids on control disk after 24h (see methods in [34]).

862 Table 8. Probing behavior variables (mean+SE, n=15) of M persicae feeding on C, annuum plants treated with 23 [34]. Variable No. of probes Total non-probing time (s) Total probing time (s) No. of intracellular punctures (pd) Mean pd duration (s) Time from 1st pd to end of probe

Compund 23 (60|ig/cm )

Control

2.57+0.3 la 330.29+42.14a 269.46+42.16a 1.57+0.44a 6.28+0.8 la 119.02+49.95a

3.07+0.27a 198.20+31.32b 396.08+32.54b 4.21+0.50b 5.03+0.33a 172.93+44.78a

Different letters within rows indicate significant differences between treatments, Fisher's protected LSD test, p<0.05.

Long-term feeding bioassays (>72 h.) showed that 23 had a negative effect on the aphid's reproduction (table 9), without associated adult mortality [34]. These results point out the potential that this bisabolene has for the control of plant virus transmission by non-persistent viruses (International Patent no. 9602748). Table 9. Daily nymphal mortality (%) caused by the natural bisabolene (23) or farnesol (F). Represented are mean values + SE* [34]. Treatment

Dose (Hg/cm2)

Control

0

23

30 60

F

30 60

1

Day 2

4

15.61±5.96a (25,440)^ 35.90±8.27b (24,258) 33.84+8.24b (24,277) 21.33+7.91 a (23,370) 36.89+9.38b (24,172)

27.58±7.48a (25,393) 23.09+8.34a (19,207) 20.55+7.19a (19,225) 7.49+2.08b (19,348) 23.65+9.12a (17,146)

31.57+7.47a (21,340) 11.59+6.42b (17,186) 9.01+4.26b (17,199) 19.43+6.84a (19,325) 21.27+7.49a (14,126)

^ Means within a column followed by the same letter are not significantly different (p>0.05, contingency table analysis). ^ Number of boxes and total insects.

Additionally, the study of the mode of action of silphinene 24 showed that larval RCRs and RGRs decreased significantly. Furthermore, GEs of the treated insects were lower than those of the starved control, Figure (3),

863

suggesting that 24 had a toxic postingestive action in addition to its strong antifeedant effect [32].

-0.5

0.0

0.5

1.0

1.5

2.0

Relative Consumption Rate Figure (3). Plot of the relative grov^ rate (RGR) on the relative consumption rate (RCR) for L decemlineata L4 larvae fed for 24 h on leaf-disks treated with compound 24. Represented are the lines of the calibration curve C (y= 0.376x-0.156, r=0.82, p<0.00001) and the treatment lines with slopes significantly different from the control C [32]. Table 10. Relative antifeedant potencies of compounds 24, 25 and 26 against £. decemlineata [36]. Compound

Test^

ECgQ (95% CL)2 lag/cm^

243

A B

1.69(1.36,2.11) 4.43(3.01,69.14)

25

A B

14.56(6.57,32.27) 15.04(10.80,20.93)

26

A B

22.60(11.45,44.59) 11.32(6.60,10.40)

^ Test A, choice; test B, no-choice. ^ Upper and lower 95%confidence intervals of EC5Q ^ From [32].

864

Given the potential importance of the biological effects of 24, we decided to study its structure-activity relationships. Compounds 25 and 26 were obtained from 24 by chemical hydrolysis [36]. Both silphinene analogs were strong CPB antifeedants with activity levels similar or lower to 24 in choice and no-choice tests (table 10). The study of their mode of action showed that 25 reduced larval growth (RGR) and consumption (RCR) for all the doses tested [36]. GEs of larvae fed with 25 were not different from the starved control, Figure (4), indicating that this compound acts as a repellent. However, 26 significantly reduced RCR and RGR depending on the doses tested [36]. GEs of larvae fed with 26 were significantly lower than those of the starved control at doses >65 |Lig/cm^, Figure (5), indicating that this silphinene analog is both antifeedant and toxic against CPB.

0.5

1.0

1.5

2.0

Relative Consumption Rate

Figure (4). Plot of the relative growth rate (RGR) on the relative consumption rate (RCR) for L decemlineata L4 larvae fed for 24 h on leaf-disks treated with compound 25. The line represents the calibration curve C (y= 0.168x-0.212, r=0.99, p<0.00001) [36].

865

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

Relative consumption rate (RCR)

Figure (5). Plot of the relative growth rate (RGR) on the relative consumption rate (RCR) for L. decemlineata L4 larvae fed for 24 h on leaf-disks treated with compound 26. Represented are the lines of the calibration curve C (y= 0.168x-0.212, r=0.99, p<0.00001) and the treatment lines with slopes significantly different from the starved control [36].

The Relative Efficiency value (GER=GE of treated insects/GE of starved control insects) of a given compound is a toxicity index considered significant when lower than 0.5. GERs of compounds 24-26 showed that silphinene 26 was the most toxic, followed by 24 and 25 (table 11) [36].

866 Table 11. Growth efficiencies (GE) and ratios (GER) of L. decemlineata larvae treated with silphinenes 24, 25 and 26 [36]. GE^

GER^

0 10 25 50 75

0.320 0.350 0.174* 0.125* 0.165* 0.358

0.89 0.97 0.48 0.35 0.46

25

0 30 65 100

0.194 0.263 0.208 0.130

>1.00 >1.00 >1.00 0.77

26

0 30 65 100

0.194 0.113 -0.016* 0.032* 0.168

>1.00 0.67 -0.09 0.19

Compound

Dose (^ig/cin^)

r

^1

sc

SC

^ \

1 From [32]. ^ Growth Efficiency (GE), calculated as the slope of the regression of relative growth rate (RGR) on relative consumption rate (RCR), where RCR=I/(BI)xT (I=mg food consumed, T=feeding period in days, BI=initial insect weight in mg) and RGR=AB/(BI)xT, (AB=change in insect body weight in mg) (see [32] for methods). ^ GER= GE Treatment/GE Starved Control (SC). *Denotes a significant difference from the Starved Control (t-test, p<0.05).

Silphinene analogs 24-26 were also active antifeedants against another Chrysomelid beetle, the western com rootworm Diabrotica virgifera virgifera Le Conte, these behavioral effects being related to the electrophysiological action of these compounds on the insect taste cells, further demonstrating a link between galeal sensillar responses and antifeedant behavior for this insect (table 12) [19].

867 Table 12. Galeal chemosensory responses to silphinenes 24-26 in adult i). virgifera 1191. Mean firing rate+SEM (number of samples)" 1 Concentration ()iM)

1

^1 10 100 1000 Ec^o(^lM)

24

25

26

5.8+1.3(24) 17.5+1.3(32) 25.6+1.5(29) 36.4+1.6(34) 32.4+1.9(34) 3.0

4.6+1.2(25) 16.0+1.8(25) 45.5+3.0 (30) 39.7+2.5 (29) 37.4+2.5 (33) 1.4

6.0+1.7(20) 14.0+2.2(30) 15.6+1.3(30) 31.3+4.5(32) 59.7+2.9 (30)

40 1 Galeal taste cell output in action potentials during first second of stimulation at indicated dose of compound in a lOmM KCl solution containing 10% ethanol. Estimated dose for half-maximal firing rate using regression analysis of log dose versus firing rate (-solvent control) in linear portion of curve.

^

Additionally, a comparative study performed with these silphinene analogs and 12 more compounds including known GABA-gated ion channel inhibitors showed species-related differences in taste sensitivity in relation to the chemical class of the test substances, silphinene 24 being the most active antifeedant against CPB (an alkaloid-adapted insect) and the norditerpene alkaloid aconitine the most active against the com rootworm (non adapted to feed on alkaloid-containing plants) (table 13) [19]. The fact that both species responded to the antifeedant action of compounds with known inhibitory effects at GABA-gated ion channels suggests a shared molecular mode of action for taste regulation in divergent insect species. Furthermore, silphinene 24 has 3D structural features similar to those of picrotoxinin [19], an estabHshed GABA-gated chloride-channel antagonist sesquiterpene [37, 38], indicating that silphinene analogs are good antifeedant models to study the molecular basis of insect taste regulation.

868 Table 13. Feeding deterrency (EC50 and 95% confidence limits) of bioactive phytochemicals on two chrysomelidae beetles [19]. EC50 (nmol/cm2) 1 Chemical class Alkaloids

Terpenoid

Compound

L. decemlineata

D. virgifera

Aconitine

178*(>39)

0.414(0.104,0.804)

GABAA

(+)P-hydrastine

31.2*07.7,121)

2.31 (1.59,3.45)

GABAA

(-)(3-hydrastine

320* (144,1730)

2.74(2.20,3.36)

GABAA

Strychnine

23.4* (17.3,39.8)

1.35(0.95,1.86)

Glycine/GABA

Bicuculline

20.0* (>8.5)

2.95 (2.04,4.07)

GABAA

a-solanine

>100

22.7 (9.55,589)

a-chaconine

>100*

6.58(1.68,13.1)

Picrotoxinin

20.7*(11.8,35.5)

171(102,417)

24

0.72* (0.42,1.23)

38.2(30.4,51.6)

?

Neuroreceptor

GABAA

25

21.6* (7.42,62.6)

2.22(0.80,3.79)

7

26

4 . 8 0 * (1.51,15.3)

67.5(40.8,218)

7

Dehydrosalviarin

20.4* (13.4,48.3)

0.541(0.378,0.761)

Cucurbitacin B

238*(>79)

>1000''

Others Quinone

Juglone

59.6(48.5,74.4)

80.1(51.2,146)

Lignan

Piperonyl butoxide

61.9(38.2,93.3)

13.5(9.0,19.1)

^ Cucurbitacin B was sitimulatoryto D. virgifera feeding. * Species significantly different at p<0.05.

Diterpenoid alkaloids from Delphinium cardiopetalum (Ranunculaceae) The atisine or veatchine system is considered the biogenetic origin of diterpene alkaloids, while aconitine or lycoctonine systems give rise to the norditerpene ones, depending on the oxygen position (C-8 or C-7 and C-8 respectively). These compounds are widely distributed over the temperate regions of the northern hemisphere among plant species of the family Ranunculaceae {Aconitum, Delphinium, Consolidd).

.....^^«

Atisine

Veatchine

Aconitine

R-H

Lycoctonine R=OH

869 Norditerpene alkaloids are toxic compounds [39], however, little is known about the biological effects of diterpene alkaloids. Therefore, as part of one our programs on bioactive compounds, we studied the insect antifeedant and toxic activity of the Delphinium diterpene alkaloids 15acetylcardiopetamine (27), cardiopetamine (30) along with its amino alcohol (31), the p-y unsaturated ketone (28) and the acetylated ketone (29) derivatives. The target insects (S. littoralis and L. decemlineata) were deterred by at least two of the test alkaloids (30 and 27 strongly inhibited the feeding activity oiS. littoralis and L decemlineata respectively, table 14), with cardiopetamine (30) being strongly active in both cases suggesting a shared mode of action for these molecules. Furthermore, S. littoralis was more sensitive to alkaloids 30, 31 and aconitine than the alkaloid-adapted CPB [20].

27 R = OBz; R, = OH; Rj = OAc 28 R = OBz; R^ = O; Rg = OH 29 R = OBz; R, = O; R2 = OAc 30 R = OBz; R, = OH; R2 = OH 31 R = OH; R^ = OH; Rj = OH

870 Table 14. Effective antifeedant doses (EC50) and 95% confidence limits (lower, upper) of compounds 27-31 against 5. littoralis sixth-instar (L6) larvae and L, decemlineata adults in choice tests [20]. Compound

27 j

EC50' (nmol/cm^) SAittomlis

L. decemlineata

>100

12.86(0.16,25.56)

28

>100

-

29

>100

27.25(22.95,31.55)

30

5.48 (3.04,7.92)

22.50(19.73,25.27)

31

23.67(19.37,27.97)

108.30(9.97,116.9)

Aconitine

32.35(19.65,45.05)

178(>39)'

Concentration required to give an antifeedant inhibition of 50%. 'From [19].

These alkaloids were not toxic to S. littoralis, while aconitine negatively affected both feeding indexes (RCR and RGR, table 15). Their toxicity on L. decemlineata was inversely correlated with their antifeedant effects, the amino alcohol (31) derivative being the most toxic (table 15). None of the test compounds proved to be as toxic as aconitine on this insect species [20]. Table 15. Oral and hemolymph injection effects of compounds 27-31 on S. littoralis L6-larvae (72h., RCR and RGR) and L. decemlineata adults (3 days % mortality) respectively [20]. Treatment (10 (xg/insect) Control 27 28 29 30 31 Aconitine

S, littoralis RCR*

RGR^

L. decemlineata % mortality^

100 99.57+4.67 97.17+11.79 113.90+19.61 103.32+4.26 104.54+4.08 62.80+6.24a

100 100.10+4.24 105.89+20.32 87.98+6.65 110.32+6.22 106.46+5.68 42.30+11.22a

0.00 7.85 18.08b 25.16b 15.75b 72.20b 100b

^ RCR=I/(BI)xT, I=mg food consumed, T=feeding period (days), BI=initial insect weight (mg). Data represented as %control. ^ RGR=AB/(BI)xT, AB=change in insect body weight (mg). ^ Corrected according to Abbott [70]. ^'^ Denote a significant difference from the control. 95% LSD test and Contingency table analysis (p<0.05) respectively.

The structure-activity study of the antifeedant action of the test alkaloids showed that the C-13 and C-15 hydroxyl groups are essential features of the active molecule for S. littoralis (compounds 27 and 31), while the presence of a C-13 hydroxy and/or a C-15 acetate determined their antifeedant effect

871 on L decemlineata (compounds 27, 29 and 30). The C-11 benzoate group strongly enhanced this biological action on both insect species [20]. Similarly it has been shown that the most potent norditerpenoid alkaloids acting as inhibitors of mammahan and insect cholinergic receptors have the C-18 anthranihc acid esterification, characteristic of the Delphinium norditerpenoid alkaloid methyllycaconitine (MLA), structurally-related to aconitine [40-42]. MLA also had antifeedant effects against Spodoptera eridania with associated post-ingestive and toxic effects at doses within the effective antifeedant dose range of compound 30 [40], suggesting a similar antifeedant mode of action between MLA/aconitine and the C-20 diterpenoid alkaloids. The toxicity of compounds 27-31 onZ. decemlineata was inversely correlated to their antifeedant effects and directly related to their polarity, suggesting a different mode of action or receptor affinity at the central nervous system [20]. A similar correlation has been shown between antifeedant effects, toxicity and polarity of structurally-related silphinene sesquiterpenes on CPB and Diabrotica virgifera virgifera [19, 36]. Such a lack of behavioral and toxicity relationship has been noted for a broad selection of plant allelochemicals [43-46]. In summary, the Delphinium diterpene alkaloids cardiopetamine (30) and 15-acetylcardiopetamine (27) are potent insect antifeedants active on two insect species with different feeding adaptations (a polyphagous Lepidopteran and an oligophagous Chrysomelid beetle), suggesting a potential broad range of antifeedant action for this class of compounds [20]. Pyrrolizidine alkaloids from Compositae and Boraginaceae species There are numerous plants containing pyrrolizidine alkaloids (PAs) and they mainly belong to the Boraginaceae, Compositae and Leguminosae families [31,47,48]. PAs constitute a class of secondary compounds with high structural diversity (more than 360 known molecules [49]). Toxicologically, these compounds are known to be the cause of liver damage and have been identified as both carcinogenic and mutagenic agents [49-52]. A number of insect species from different taxa have evolved adaptations to sequester [5355], store [56] and utilize plant PAs against insect predators [57, 58], and they are considered plant defenses against generalist insect herbivores [59]. Furthermore, individual PA patterns have been proposed to be under genetic

872 control as the result of evolution under selective pressure [60], suggesting an evolutionary advantage for the structural diversity of PAs. As part of a broad study of bioactive PAs from Chilean, Canarian and European Heliotropium, Senecio and Echium species, we have isolated several pyrrolizidine alkaloids [61-64]. If PA chemical diversity represents an evolutionary advantage, we should expect individual PAs to be specific rather than general plant defenses. In this context we have performed a comparative study of the biological effects of PAs 32-41 on two insect herbivores, S. littoralis (PA-tolerant), and L. decemlineata (not adapted to PAs) in relation to their chemical class (table 16).

o=^==\

32

33 OH ^Me

34

35

873

36

O OMe II OH X HO

H

ay •* 37

38

\

OH 40

CD

R=

OH OR .^VV^^

H

39

41

R=

874 In terms of relative potency values (EC50), S. littoralis had overall higher values than L decemlineata, as expected from their feeding adaptations (a polyphagous and an olyphagous species respectively). Among the macrocyclic diesters (retrorsine, 32; monocrotaline, 33 and senecionine, 34), we observed a significant antifeedant effect for 34 against CPB without associated toxic effects. The secopyrroHzidine diester otosenine (35) was a moderate antifeedant with negative postingestive effects against S. littoralis. The open diester echimidine (36) was a strong antifeedant to L decemlineata without post-ingestive or toxic effects. Among the unsaturated monoesters (lycopsamine, 37; europine, 38; and supinine, 39), 37 and 38 were moderate antifeedants to S. littoralis while 39 moderately deterred L decemlineata. None of these compounds had postingestive or toxic effects. Among the saturated monoesters (3 '-acetyltrachelanthamine, 40 and floridinine, 41), 40 was a strong CPB antifeedant with a relative potency (EC50) similar to echimidine (36) (table 16). Table 16. Comparative antifeedant and toxic effects of PAs on two divergent insect species, a generalist lepidopteran (iS*. littoralis^ S.l.) and a semi-specialist coleopteran {L, decemlineata^ L.d.) adapted to feed on alkaloid-containing plants. Chemical class

Compound

Source

Antifeedant effects S.l

Macrocyclic diester Secopirrolizidinic diester Open diester Unsaturated monoester

Saturated monoester

'Reinaetal., [61; ^ Unpublished results ^ Reina et al., [64; ^ Reina et al., [62;

32 33 34 35 36 37 38 39 40 41

ECsoCng/cm') >50 S. microphyllus^ >50 >50 Aldrich >50 S. erraticus' >50 1.3(0.2,7.9) >50 S glaber^ 13.2(5.7,30.7) E. wildprettf >50 H. megalantum^ 31.6(27.7,35.9) H. bovef -50 H. sinuatum^ -35 H. floridunr' H. floridunv'

>50 >50

Toxic effects S.I.' L.d:

Ld..

RCR*

RGR^

%M'

89.9 84.7 94.4 48.6*

88.2 77.3 113.25 42.0*

0 15 0 9

1.4(0.8,2.1)

118.2

165.5

>100 >50 20.5 (8.2,51.5) 1.8(0.9,3.4) >50

109.0

100.6

0 12

-

-

-

91.6

82.4

3

-

-

5

^ Reina etal, [63] ^ Oral injection (20|ig/insec t) ^ Hemolymph 1njection(10 Insect) ^As in table 15 Percent mortality (%M) corrected according to Abbott [70]

-

875 These results indicate that antifeedant PAs studied here are structure- and species-dependent. The PA-tolerant S. littoralis was deterred by all the unsaturated monoesters tested, and the secopyrrolizidinic diester otosenine (35). Therefore, esterification of the hydroxyl group at C-9 in the presence (37 and 38) or absence (39) of the C-7 hydroxylation and the C-l/C-2 insaturation of the necine base determined the antifeedant action on S. littoralis. Additionally, the antifeedant and postingestive effects of otosenine (35) indicate that the presence of an electrophyllic ketone group in C-8 at the necine base could play a role in determining such activity. CPB sensitivity to these compounds followed a different pattern from that ofS, littoralis. The macrocyclic diester senecionine (34), the open diester echimidine (36) and the saturated monoester 40 were the strongest antifeedants to this insect with similar potencies but few structural features in common to conclude structure-activity relationships. Furthermore, the unsaturated monoester 39 had moderate-low antifeedant effect, suggesting that CPB taste receptors can bind to different PA chemical classes with high molecular selectivity. Similarly to CPB, larvae of the tortricid Chonistoneura fumiferana (a non PA-tolerant insect) were strongly deterred by the open diester lasiocarpine, moderately by the macrocyclic diester senecionine and the unsaturated monoester heliotrine, while otosenine was inactive The antifeedant effects of PAs are proposed to play a key role in plant defense from insect herbivores [59]. However, little is known on the molecular mechanisms modulating PA-insect taste reception. Several PAs, including echimidine, monocrotaline and senecionine, have shown significant binding activity to muscarinic and serotonin receptors, indicating that PAs can affect several molecular targets besides long-term toxicity through DNA alkylation by PA metabolites generated in the liver [66]. Therefore, the interference of PAs with neuronal signal transduction could mediate insect taste regulation as proposed for chrysomelid beetles [18, 19]. Tertiary PAs are deleterious for organisms with a microsomal cytochrome P450, but S. littoralis larvae tolerate PAs. These larvae prevent poisoning by rapid and efficient excretion of the absorbed tertiary alkaloid [56]. However, otosenine (35) had toxic post-ingestive effects on S. littoralis, indicating that the excretion of this compound is not efficient enough and therefore might undergo bioactivation by the insect's cytochrome P-450.

876 None of the PAs tested resulted toxic to the chrysomeUd L. decemlineata. Oreina (Chrysomehdae) beetles are able to take up plant alkaloid N-oxides and eliminate tertiary PAs, but are unable to N-oxidize tertiary PAs [55, 67, 68]. Similarly, L. decemlineata adults could eliminate tertiary PAs efficiently enough to avoid poisoning. In conclusion, PA antifeedant effects are species and structure-dependent with a broad range of action probably related to their ability to interfere with several molecular targets. However, their toxic effects on insect herbivores seem to be more restricted. Conclusions Among the compounds presented here, 34% were insect antifeedants, 22% toxicants and 44% had both effects. These compounds belong to very different chemical classes, have been produced by different plant species and are proposed to act on different molecular targets, suggesting high plasticity for these biological effects at the molecular level. The molecular diversity and multipurpose biological effects of alkaloids has been postulated as the result of "evolutionary molecular modeling" [69]. The data presented in this chapter supports such hypothesis but without restrictions in terms of chemical class as long as the molecule being considered is involved in the defensive strategy of a plant. ACKNOWLEDGEMENTS We gratefully acknowledge A. Santos, Director of the Botanic Garden at La Orotava, A. Fernandez, Director of Garajonay Natl. Park, and S. Carlin for language advice. This work has been partially funded by grants PI 91/17, PB 94-0020-B, PI 13/95, PB97-1226 and PI 1999/076 from the Canarian Government and CICYT-Spain. REFERENCES [1]

[2]

Georghiou, G.P. In Managing Resistance to Agrochemicals: From Fundamental Research to Practical Stratigies; Gree, M.B.; Le Baron, H.M.; Moberg, W.K., Eds.; American Chemical Society: Wasintong D.C., 1990. Metcalf, R.L. In The Pesticide Question. Environment, Economics and Ethics; Pimentel, D.; Lehman, H., Eds.; Chapman and Hall: New York, 1993, pp. 426-430.

877

[3]

[4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]

Duke, S.O.; Menn J.J.; Plimmer J.R. In Pest Control with Enhanced Enviromental Safety; Duke S.O.; Menn J.J.; Plimmer J.R., Eds; American Chemical Society: Washintong D.C., 1993; ACS Symp. Ser. 524, pp. 1-13. Bramwell, D.; Bramwell Z.I. Flores Silvestres de las islas Canarias, Ed. Rueda S.L.: Madrid, 1990. Bramwell, D. In Biogeography and Ecology of the Canary Islands; G.Kunkel Publ: The Hague,1976, pp. 207-240. Gonzalez-Coloma, A.; Escoubas, P.; Reina, M.; Mizutani, J. Appl Zool EntomoL, 1994, 29, 292. Isogai, A.; Suzuki, A.; Murakoshi, S.; Ohashi, Y.; and Sasada, Y. Agr. Biol. Chem., 1976, 40, 2305. Isogai, A.; Murakoshi, S.; Suzuki, A.; Tamura, S. Agric. Biol. Chem. 1977, 41, 1779. Gonzalez-Coloma, A.; Hernandez, M.G.; Perales, A.; Fraga, B.M. J. Chem. EcoL, 1990,16, 2723. Gonzalez-Coloma, A.; Terrero, D.; Perales, A.; Escoubas, P.; Fraga, B.M. J. Agric. Food Chem., 1996,^4^,296. Fraga B.M.; Gonzalez-Coloma A.; Gutierrez C ; Terrero D. / Nat. Prods., 1997, 60, 880. Lehmberg, E.; Casida, J.E. Biochem. Physiol, 1994, 48, 145. Usherwood, P.N.R.; Vais, H. Toxicol. Lett., 1995, 82/83, 247. Jefferies, P.R.; Yu, P.; Casida, J.E. Pestic. Sci., 1997, 57, 33. Achenbach, H.; Hubner, H.; Vierling, W.; Brandt, W.; Reiter, M. J. Nat. Prod., 1995, 58, 1092. Hubner, H.; Achenbach, H.; Vierling, W.; Brandt, W.; Reites, M. Phytochemistry, 1999, in press. Gonzalez-Coloma A.; Gutierrez, C ; Hubner, H.; Achenbach, H.; Terrero, D.; Fraga, B. M. J. Agric. Food Chem,, 1999, 47, 4419. Mullin, C.A.; Chyb, S.; Eichenseer, H.; HolHster, B.; Frazier, J.L. J. Insect Physiol,

\99^,40,9n. [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29]

Mullin, C ; Gonzalez-Coloma, A.; Gutierrez, C ; Reina, M.; Eichenseer, H.; Hollister, B.; Chyb, S. / Chem. Ecol, 1997, 23, 1851. Gonzalez-Coloma, A.; Guadano, A.; Gutierrez, C ; Cabrera, R.; De la Peiia, E.; De la Fuente, G.; Reina, M. J. Agric. Food Chem., 1998, 46, 286. Gonzalez-Coloma, A.; Escoubas, P.; Lajide, L.; Mizutani, J. App. Entomol. Zool, 1994, 29, 289. Gonzalez-Coloma, A.; Escoubas, P.; Mizutani, J.; Lajide, L. Phytochemistry,\994, 35, 607. McRae, W.D.; Towers, G.H.N. Phytochemistry, 1984, 23, 1207. Nitao, J.K.; Johnson, K.S.; Scriber, M.; Nair, M.G. J. Chem. Ecol, 1992,18, 1661. B-Bemard, C ; Philogene, B.J.R. J. Toxicol & Environ. Health., 1993, 38, 199. Takaoka, D.; Watanabe, K.; Hiroi, M. Bull Chem. Soc. Japan., 1976, 49, 3564. Shin, K.H.; Kim, O.N.; Woo, W.S. Arch. Pharmacol Res. (Seoul), 1988, II, 240. Shin, K.H.; Woo, W.S. Korean Biochem. J., 1990, 23, 122. Pieman, A.K. Biochem. Syst. Ecol, 1986,14, 255.

878

[30]

[31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53]

[54] [55] [56]

Mullin, C. A.; Alfatafta, A. A.; Harman, J.L.; Serino, A.A.; Everett, S.L.L. In Naturally Occurring Pest Biorregulators; American Chemical Society: Washington D.C, 1991; ACS Symp. Ser. 449, pp. 278-292. Rizk, A.F.M. Naturally Occuring Pyrrolizidine Alkaloids', CRC Press: Boca Raton, 1991. Gonzalez-Coloma, A.; Reina, M,, Cabrera, R.; Castafiera, P.; Gutierrez, C. J. Chem. EcoL, 1995, 27, 1255. Alford, A.R.; Cullen, J.A.; Storch, R.H.;Bentley, M.D. J. Econ. Entomol, 1987, 80, 575. Gutierrez, C , Fereres, A.; Reina, M.; Cabrera, R.; Gonzalez-Coloma, A. J. Chem. EcoL, 1991, 23, 1641. Powell, G. Ann. Appl. BioL, 1991,119, 313. Gonzalez-Coloma, A.; Gutierrez, C ; Cabrera R.; Reina, M. J. Agric. Food Chem, 1997, 45, 946. Klunk, W.E.; Kalman, B.L.; Ferrendelli, J.A.; Lovey, D.F. 1983. Mol. Pharmacol, 1990,25,511. Ozoe, Y.; Matsumura, F. J. Agric. Food Chem., 1986, 34, 126. H. Wink. In The Alkaloids', Cordell, G.A., Ed.; Academic Press: New York, 1993, Vol. 43, pp. 1-118. Jennings, K.R.; Brown, D.G.; Wright, D.P.J. Experientia, 1986, ^2, 611. Kukel, C.F.; Jennings, K.R. Can. J. Physiol Pharmacol, 1994, 72, 104. Manners, G.D.; Panter, K.E.; Pelletier, S.W. J. Nat. Prods., 1995, 58, 863. Bemays, E.A. Entomol Exp. Appl, 1990, 54, 53. Bemays, E.A. J. Chem. Ecol, 1991,17, 2519. Bemays, E.A.; Comelius, M. Entomol Exp. Appl, 1992, 64, 289. Wmbel, R.P.; Bemays, E.A. Entomol Exp. Appl, 1990 54, 125. Bull, L.B.; Culvenor, C.C.J.; Dick, A.T. The Pyrrolizidine Alkaloids; North-Holland Publishing Company: Amsterdam, 1968. Mattocks, A.R. Chemistry and Toxicology of Pyrrolizidine Alkaloids, Academic Press: London, 1986. Hartmann, T.; Witte, L. In Alkaloids: Chemical and Biological Perspectives', Pelletier, S.W., Ed.; Pergamon: Oxford, 1995; Vol. 9, pp. 155-233. Peterson, J.E.; Culvenor, C.C.J. J. In Handbook of Natural Toxins Voll; Keeler, R.F.; Tu, A.T., Eds.; Marcel Dekker: New York, 1983. Roeder, E. Pharmazie, 1995, 50, 83. Roeder, E. Current Organic Chemistry, 1999, 3, 557. Aplin,R.T.; Rotschild,M. In Toxins of animal and Plant origin; de Vries, A.; Kochva, E., Eds.; Gordon and Breach Sci. Pub: New York, 1972; Vol.2, pp. 579595. Trigo, J.R.; Brown, K.S.; Hartmann, T.; Ernst, L.; Barata, L.E.S. Biol. J, Linn. Soc, 1996, 58, 99. Hartmann, T.; Witte, L. Ehmke, A.; Theuring, C ; Rowell-Rahier, M.; Pasteels, J.M. Phytochemistry, 1997, 45, 489. Lindigkeit, R.; Biller, A.; Buch, M.; Schiebel, H.M.; Boppre, M.; Hartmann, T. Eur, J. Biochem., 1997, 245, 626.

879 [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70]

Dussourd, D.E.; Ubik, K.; Harvis, C; Resch, J.; Mainwald, J.; Eisner, T. Proc. Natl Acad. Sci. USA. 1991, 88, 9224. Rowell-Rahier, M.; Pasteels, J.M.; Alonso-Mejia, A.; Brower, L.P. Animal. Behav., 1995, 49, 709. Van Dam, N.; Vuister, L.W.M.; Bergshoeff, C; De Vos, H.; Van der Meijden, E.D. J. Chem. Ecol, 1995, 21, 507. Hartmann, T.; Dierich, B. Planta, 1998, 206, 443. Reina, M.; De la Fuente, G.; Villarroel, L.; Torres,R. Anales de Quimica, 1993, 89, 387. Reina, M.; Mericli, A.H.; Cabrera, R.; Gonzalez-Coloma, A. Phy to chemistry, 1995, 38, 355. Reina, M.; Gonzalez-Coloma, A.; Gutierrez, C; Cabrera, R.; Henriquez, J.; Villarroel, L. Phytochemistry, 1997, 46, 845. Reina, M.; Gonzalez-Coloma, A.; Gutierrez, C; Cabrera, R.; Henriquez, J.; Villarroel, L.J. Nat. Prod., 1999, 61, 1418. Bentley, M.D.; Leonard, D.E.; Stoddard, W.F.; Zalkow, L.H. Ann. Entomol. Soc. Am., 1984, 77, 393. Schmeller, T; El-Shazly, A.; Wink, M. J. Chem. Ecol, 1997, 23, 399. Ehmke, A.; Rowell-Rahier, M.; Pasteels, J.M.; Hartmann, T. J. Chem. Ecol, 1991, 77,2367. Rowell-Rahier, M.; Witte, L.; Ehmke, A.; Hartmann, T.; Pasteels, J.M. Chemoecology, 1991, 2, 41. Wink, M.; Schmeller, T.; Latz-Briining, B. J. Chem. Ecol, 1998, 24, 1881. Abbott, W.S. A method of computing the effectiveness of an insecticide. / Econ. Entomol. 1925,18, 265-267.

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

881

OCCURRENCE, STRUCTURE AND BIOACTIVITY OF 1,7-DIARYLHEPTANOIDS

PER CLAESON*, UBONWAN P. CLAESON^ PATOOMRATANA TUCHINDA^ AND VICHAIREUTRAKUL^ Division of Pharmacognosy, Department of Pharmacy, Uppsala University, Biomedical Centre, Box 579, SE-751 23 Uppsala, Sweden ^Institute for Bioactive Natural Products, Uppsala Science Park, SE-751 83 Uppsala, Sweden ^Department of Chemistry, Faculty of Science, Mahidol University, Bangkok 10400, Thailand "^To whom correspondence should be addressed Tel: ^4618 471 44 79. Fax: +4618 50 91 OL E-mail: per. claeson @fkog. uu. se ABSTRACT: The 1,7-diarylheptanoids are a class of secondary plant metabolites characterised by the structural motif aryl-C7-aryl. Based on their chemical structures, the diarylheptanoids have been divided into five different groups: non-phenolic linear diarylheptanoids (Type I), phenolic linear diarylheptanoids (Type II), macrocychc biarylheptanoids (Type IE), macrocyclic diaryl ether heptanoids (Type IV) and diarylheptanoids with cyclised C7-chain (Type V). The literature up to October 1993 has previously been covered (Claeson ^r a/. J. hidian Chem. Soc. 1994, 71, 509-521), and presentiy the literature of October 1993 - September 1999 has been reviewed. During the period of review, reports on a total of 114 diarylheptanoid structures (including 75 new natural products) have been retrieved from the literature. These structures, together with pertinent information on plant sources, biological activities and references, have been compiled.

INTRODUCTION The 1,7-diarylheptanoids constitute a class of secondary plant metabolites characterised by the structural motif aryl-Cy-aryl. The diarylheptanoids appear in both linear and cyclic forms. Based on their chemical structures, the diarylheptanoids have been grouped in five major types [1]. The diarylheptanoids of both Type I and n are linear structures, but for practical reasons they have been separated into two groups. The nonphenolic diarylheptanoids (Type I) sometimes appear under the

882 consignation diphenylheptanoids and, historically, the phenolic linear diarylheptanoids (Type H) have also been called curcuminoids after curcumin, the orange pigment of turmeric, Curcuma longa. In the diarylheptanoids of Type HI and IV, the aromatic rings are meta-meta' or meta-para'COimcctQd to form macrocyclic structures with biaryl and diaryl ether moieties, respectively. In the Type V diarylheptanoids, the C7 chain is cyclised to form a pyran-like ring.

Typen

Type I

Type in

m

Type IV

TypeV

The first review of diarylheptanoids covered the literature up to October 1993, and a total of 117 substances were listed [1]. Another review on diarylheptanoids, particularly informative on synthetic aspects, subsequently appeared [2]. In the present review, covering reports appearing in major journals during October 1993 - September 1999, information on 114 diarylheptanoids has been found. Out of these, 75 substances are new natiial products; and for the remaining 39 known compounds, studies on biological activities, re-isolation from other plant sources, synthesis, structure revision etc., have been pubUshed. Tables 1-5 list the diarylheptanoids with trivial names, occurrence in nature and pertinent

883 literature references. Information on biological activities is found in the respective sections. The appearance of reference [1] (the first review) in Tables 1-5 indicates that the substances are not isolated for the first time, but that additional information is reported. Non-Phenolic Linear Diarylheptanoids (Type I) Reports on 11 diarylheptanoids (1-11) of Type I have been published during the time period of review (Table 1). Four of the diarylheptanoids (3, 5,6, and 8) in Table 1 are new natural products. The occurrence of Type I diarylheptanoids is still confined to the families Betulaceae and Zingiberaceae (c/ [1]). Table 1. Non-Phenolic Linear Diarylheptanoids Compound Trivia] name

Occurrence"

Ref.

1

-

Curcuma xanthorrhiza (Z)

3^

2

Dihydroyashabushiketol

Alpinia conchigera (Z) Alnus maximowiczii (Be)

1, 4 5

3

-

Alpinia conchigera (Z) Alnus maximowiczii (Be)

4 5

4

-

Curcuma comosa (Z) Curcuma xanthorrhiza (Z)

1, 6 3

5

-

Curcuma comosa (Z) Curcuma xanthorrhiza (Z)

6

6

-

Curcuma comosa (Z) Curcuma xanthorrhiza (Z)

3" 6 3^

7

-

Curcuma comosa (Z)

1, 6

8

-

Alnus maximowiczii (Be) Alpinia qfficinarum(Z)

9

-

Curcuma comosa (Z) Curcuma xanthorrhiza (Z)

10

Alnustone

5 7 1, 6 3 1, 3

1 1 1 1 1

11

-

3^

1

i Curcuma xanthorrhiza (Z) Curcuma xanthorrhiza (Z)

1

*(Be)=Betulaceae; (Z)=Zingiberaceae ''Semisynthetic compound

A series of naturally occurring (4, 9 and 10) and semisynthetic (1, 5, 6

884 and 11) diaryIheptanoids from Curcuma xanthorrhiza have been evaluated for topical anti-inflanunatory activity in the murine model of ethyl phenyl propiolate-induced ear oedema. A distinct structure-activity relationship could be identified; in which compound 10 was the most potent agent and compound 1 virtually inactive [3]. Compounds 4, 5, 6 and 7 have been isolated from Curcuma comosa as the substances responsible for the nematocidal effect on Caenorhabditis elegans of a methanolic extract of the plant. Compound 9 was also isolated from the same plant, but it was found inactive in the same assay [6].

Rl

R2

Rl

R2

*^1

*^2

1

— H

— OH

4

—H

— OH

2

=

0

—m

5

—H

— OAc

3

=

0

=

6

—"H

=

0

7

— OH

=

0

9

— OH

0

10

=

0

11

—OAc

Phenolic Linear Diarylheptanoids (Type II) The phenolic linear diarylheptanoids (12-59; Table 2) constitute the largest subgroup with reports on 48 substances, of which 28 substances are

885

Table 2. Phenolic Linear Diaryllieptanoids Compd. Trivial

name

Occurrence*

JRef.

|l2

(-)-Centrolobol

Centrolobium sclerophyllum (F)

|l, 8

|l3

Aceroside VII

Betula platyphylla var. japonica (Be)

|l,9

|l4

Aceroside VIII

Betula platyphylla var. japonica (Be)

1,9,10

|l5

Aceroside XIII

Acer triflorum (A)

11

|l6

Aceroside XII

Acer triflorum (A)

11

|l 7

Platyphyllonol

Corylus sieboldiana (Be)

1, 12

|l8

Platyphylloside

Betula platyphylla var. japonica (Be)

|l9

1-

Betula platyphylla var. japonica (Be)

1» 10 1, 10

1 1

{20

Betulaplatoside la

Betula platyphylla var. japonica (Be)

10

1

|21

Betulaplatoside lb

Betula platyphylla var. japonica (Be)

10

1

22

-

Alpinia concigera (Z) Alpinia officinarum (Z)

1,4

1

7

1

|23

-

Alpinia officinarum (Z)

1, 7

1

|24

Rhoiptelol C

Rhoiptelea chilianta (R)

13

1

|25

-

Pinus flexilis (?)

14

1

|26

-

Pinus flexilis (?)

14, 15

1

[27

-

Alnus japonica (Be)

15

1

|28

Hirsutanonol

Pinus flexilis (?)

1, 14

1

|29

Oregonin

Pinus flexilis (?)

1, 14

1

jao

-

Alnus japonica (Be)

1, 15

1

|31

-

A/nw5 japonica (Be)

15

1

32

Oregonoside A

A/nw5 rw^ra (Be)

16

1

33

Oregonoside B

A/n«5 rubra (Be)

16

1

34

Demethoxyyakuchinone B

Alpinia oxyphylla (Z)

17'

35 36

1 Yakuchinone B 1 YPE-01

Alpinia oxyphylla (Z)

1, 17

1 Alpinia oxyphylla (Z)

ir

'(A)=Aceraceae; (Be)=Betulaceae; (F)=Fabaceae; (?)=?inaceae; (R)=Rhoipteleaceae; (Z)=Zingiberaceae ''Semisynthetic compound

1

1

886

Table 2. Phenolic Linear Diarylheptanoids (cont.) Compd. Trivial name

Occurrence'

Ref.

137

1-

Curcuma xanthorrhiza (Z)

18

|38

-

Curcuma xanthorrhiza (Z)

18

|39

-

Alpinia katsumadai (Z)

19

40

-

Alpinia concigera (Z) Alpinia officinarum (Z)

1, 4 7

|41

-

Betula platyphylla var. japonica (Be)

1, 9, 10

|42

-

Betula platyphylla var. japonica (Be)

9

43

Hirsutenone

Alnus japonica (Be) Pinus flexilis (P)

1, 15 14

44

Calyxin A

Alpinia blepharocalyx (Z)

20, 21, 22

45

Calyxin E

Alpinia blepharocalyx (Z)

21, 22

46

Calyxin H

Alpinia blepharocalyx (Z)

22, 23

47

Epicalyxin H

Alpinia blepharocalyx (Z)

22, 23

48

Calyxin B

Alpinia blepharocalyx (Z)

20, 22, 24 1

49

Epicalyxin B

Alpinia blepharocalyx (Z)

20, 22, 24 1

50

Calyxin C

Alpinia blepharocalyx (Z)

22, 24

1

51

Epicalyxin C

Alpinia blepharocalyx (Z)

22, 24

1

52

Calyxin D

Alpinia blepharocalyx (Z)

22, 24

1

53

Epicalyxin D

Alpinia blepharocalyx (Z)

22, 24

1

54

Blepharocalyxin A

Alpinia blepharocalyx (Z)

22, 23, 25 1

55

Blepharocalyxin B

Alpinia blepharocalyx (Z)

22, 23, 25 1

56

-

Alpinia blepharocalyx (Z)

22, 26

1

57

Curcumin

Curcuma longa (Z)

1, 27, 28

1

58

Demethoxy-curcumin Curcuma longa (Z)

1, 27, 28

[

1, 27, 28 26

1 1

59

Bisdemethoxycurcumin

Curcuma longa (Z) Alpinia blepharocalyx (Z)

'(A)=Aceraceae; (Be)=Betulaceae; (F)=Fabaceae; (P)=Pinaceae; (R)=Rhoipteleaceae; (Z)=Zingiberaceae ^'Semisynthetic compound

1

887 reported as natural products for the first time. The majority of these diarylheptanoids have been isolated from the families Betulaceae and Zingiberaceae, but compounds 24-26, 28, 29 and 43 have been isolated from the two families Pinaceae and Rhoipteleaceae, neither of which have previously been reported to contain diarylheptanoids. A novel series of compounds (44-55) with chalcone, flavanoid or chalcone-diarylheptanoid moieties attached to the diarylheptanoid backbone has been isolated from Alpinia blepharocalyx [22-24]. Compound 12 has shown significant anti-leishmanial activity in vitro against the extracellular forms (promastigotes) of Leishmania amazonensis [8]. A series of compounds from the bark of Betula platyphylla var. japonica, 13,14, 18, 19, 20 and 21, has been evaluated in a battery of tests for hepato-protective properties (prevention of CCI4- and D-GalN/LPSinduced liver damage, anti-oxidative activity, 02-scavenging activity). The diarylheptanoids showed varying degrees of activities in the assays [10]. Compounds 22 and 23 also exhibited antioxidant activity [7]. Bioassay-guided fractionation of a CH2Cl2-MeOH extract of Pinus flexilis led to the isolation of 25, 26, 28, 29 and 43. They were aU found to inhibit the enzyme protein kinase C, which has been implicated as a potential target for novel anti-cancer drugs [14]. Compounds 34 and 35 were found to be inhibitors of 5-lipoxygenase and cyclooxygenase-1 and -2. The closely related derivative 36, was, however, found to be a selective inhibitor of 5-lipoxygenase. Topical application of 36 significantly suppressed both arachidonic acid- and 12-0tetradecanoyl phorbol 13-acetate-induced ear oedemas in mice [17]. The two new diarylheptanoids 37 and 38 exhibited significant hypolipidemic action by inhibiting hepatic triglyceride secretion in hamsters [18]. Compounds 44-56 from Alpinia blepharocalyx have all been shown to inhibit nitric oxide (NO) production in lipopolysaccharide (LPS)-activated murine macrophages in vitro. Compound 55 was the most potent inhibitor [22]. Compounds 55 and 59 furthermore inhibited collagen-, arachidonic acid- and adenosine-induced platelet aggregation of human whole blood [26]. Curcunun (57) and its two derivatives 58 and 59 from Curcuma longa exhibited nematocidal activity against second-stage larvae of dog roundworm, Toxocara canis. The nematocidal activity increased remarkably when the curcuminoids were mixed, which was suggested to be due to a synergistic action between them [27].

12

..... OH

— OH

13

••"• 0-P-D-Glc/?

— OH

14

••" 0-p-D-Api/-(l-*- 6)-P-D-Glc/7

— OH

15

•""CM

— O-p-D-Glc/7

16

.....CH

""• o-3-D-Api/--(r

Rl

" 6)-3-D-Glcp

R2

17

=

0

— OH

18

=

0

— O-P-D-Glc/?

19

=

0

— 0-p-D-Api/-(l—»- 6)-P-D-Glc/?

20

•••"OH

— 0-P-D-Glc/?

21

— OH

— O-p-D-Glc/?

R2

R3 CMe

^1

K2

R3

22

—H

= 0

—OH

23

—H

= 0

— O^e

889 R2

R3 OMe

24

•«"0H

""(H

Rl

•••••OH

R2

25

•""(H

—-H

26

•••" 0-p-D'Glcp

—H

27

'""6-0-(3-D-Api-P- D-Glcp

—H

28

= 0

— (B

29

=

— 0-p-D-xylose

30

= 0

— 0-3-D-Glc

31

=o

— 6-O-galloyi-P-D-Glcp

0

32

— 0-I3-D-G1C.6-0-CO-

^ ^ OtAe

33

= 0

— O-P-D-Glc-6-O-CO-. ^

V-CH

890

Ri

R2

34

—H

— H

35

— CMe

— H

36

— OMe

— OMe

Rl

R2

*^1

*^1

K2

37

— H

-—OH

-^H

38

— H

— OH

— OH

39

..... CM

=

— H

0

OMe

40

891

R

41

— H

42

— OH

43

OH

44

45 HO

R

892 OMe

C-7

Rl 46

— H

S

47

— H

R

48

— OH

S

49

— OH

R

C-7

Rl

50

51

0

O^J^^OMe

(Me

893

Rl

52

53

54

55

C-7

894

56

Rj

R3

57

— (Me

— (Me

58

— (Me

—H

59

—H

—H

Macrocyclic Biarylheptanoids (Type HI) Scientific reports have appeared on 24 macrocyclic biarylheptanoids (60-83; Table 3), of which 17 are new natural products. A revised structure of porson (61) has been proposed [29]. Plants of the genera Betula (Betulaceae), Acer (Aceraceae) and Corylus (Betulaceae) are reported for the first time to contain Type III diarylheptanoids. Rhoiptelol A (79) is the first macrocyclic biarylheptanoid fi-om the family Rhoipteleaceae. Compounds 65 and 66 were reported in 1991, unfortunately missed in the previous review (1), but arbitrarily included in the present compilation. No biological activities have been reported for any Type IE diarylheptanoids.

895 Table 3. Macrocyclic Biarylheptanoids Trivial name

Occurrence'

Ref.

Myricanone

Myrica gale var. tomentosa (M)

1, 29

Person

Myrica gale var. tomentosa (M)

1, 29'

I 62

12-Dehydroporson

Myrica gale var. tomentosa (M)

29

1 63 1 64

12 -Hydroxy-myricanone Myrica gale var. tomentosa (M)

29

Myricatomentoside II

Myrica gale var. tomentosa (M)

30

1 65

-

Myrica rubra (M)

66

-

Myrica rubra (M)

3r 3r

1 ^^ '

-

Betula davurica (Be)

32

1 68

-

Betula davurica (Be)

32

69

Acerogenin E

Acer nikoense (A) Betula ermanii (Be) Betula platyphylla var. japonica (Be) 5^f«/fl maximowicziana (Be) B^m/<3 davurica (Be)

33 34 9 35 32

1 70

Aceroside XI

Acer nikoense (A)

33

1

Betula maximowicziana (Be)

35 36

1 1

1, 12 12

1 1

1, 37 35 38

1 1 1 1 1 1 1

Compd.

I 60 61

1 '^^ 1^ '^ Acerogenin K 1^ '^ Alnusonol 1 74 1^ '^ Alnusdiol

Acer nikoense (A) Corylus sieboldiana (Be) Corylus sieboldiana (Be) Betula maximowicziana (Be)

76

-

Betula maximowicziana (Be)

77

Betulatetraol

Betula platyphylla var. latifolia (Be)

78

-

Corylus sieboldiana (Be)

79

Rhoiptelol A

Rhoiptelea chilianta (R)

Alnusone

Corylus sieboldiana (Be)

1

81 82

1I-

Corylus sieboldiana (Be)

1, 12 13 1, 12 39

Corylus sieboldiana (Be)

12

1

83

J-

Corylus sieboldiana (Be)

1, 12

1

1 80

*(A)=Aceraceae; (Be)=Betulaceae; (M)=Myricaceae; (R)=Rhoipteleaceae; (Z)=Zingiberaceae "Revised structure ^'Missed in reference [1]

896

R2

Rl

R3

60

—H

=

0

—(M

61

—(M

=

0

—OMe

62

= 0

=

0

—(Me

63

—m

=

0

—OH

64

—m

=

0

— O-p-D-Glcp

65

—H

—OH

—0-P-D-Glc/?-(l-^ 3)-P-D-Glc/7

66

—H

— OH

— p-D-Glcp-(l-^- 6)-a-L-Ara/

R 67

—H

68

—CMe

897

R3

R2

Rl 69

=

0

H

—(M

70

=

0

H

— O-P-D-Glcp

71

=

0

OH

—(Me

72

— OH

73

-H

74

-OH

— H

75

76

— CM

-m -H« O-P-D-Glcp

-OH —OH

898

77

Mea

79

83

R2

Rl 80

H

H

81

H

(M

82

(M

OH

899 Macrocyclic Diaryl Ether Heptanoids (Type IV) Fourteen new macrocyclic diaryl ether heptanoids have been described (Table 4). The total syntheses of compounds 88, 89, 90, 91, 94 and 101 have also been reported (c/ Table 4). Based on synthetic and spectroscopic evidence, the structure of garuganin HI has been revised to 101 [46]. Most of the Type IV diaryIheptanoids have been found in Acer nikoense (Aceraceae), but plants in both Betulaceae and Juglandaceae are now known to produce this type of compounds (Table 4). No studies on biological effects of Type IV diarylheptanoids have been reported, but antibacterial activities of the garuganins from Garuga species have been predicted based molecular mechanics and molecular orbital calculations [47]. Table 4. Macrocyclic Diaryl Ether Heptanoids

1 Compd. Trivial name Platycarynol 184

Occurrence*

Ref.

Platycarya strobilacea (J)

(+)-Galeon

Myrica gale var. tomentosa (M)

186 187

Myricatomentoside I

Myrica gale var. tomentosa (M)

-

Juglans mandshurica (J)

40 30 30 41

|88

Acerogenin A

Acer nikoense (A)

1, 42', 43' 1

89

Acerogenin B

Acer nikoense (A)

1, 43'

190 191 192 193 194 195 196 197 198

Acerogenin C

Acer nikoense (A)

33, 42', 43' 1 1, 43', 44' 1

85

|99 100

I101

Aceroside IV

Acer nikoense (A)

Acerogenin D

Acer nikoense (A)

1 1 1 1 1

Aceroside V

Acer nikoense (A)

33 33

Acerogenin L

Acer nikoense (A)

36, 43'

1

Acerogenin F

Acer nikoense (A)

Acerogenin J

Acer nikoense (A)

36 36 36 36 37

1 1 1 1 1

Acerogenin I

Acer nikoense (A)

1 Acerogenin H

Acer nikoense (A)

(-)-Maximowicziol A

Betula maximowicziana (Be)

-

Betula platyphylla var. japonica (Be) Betula ovalifolia (Be)

1 Garuganin III

Garuga pinnata (Bu)

1 1

9

45 1,46'^

'(A)=Aceraceae; (Be)=Betulaceae; (Bu)=Burseraceae; (J)=Juglandaceae; (M)=Myricaceae 'Totalsynthesis ^'Revised structure

1 1

900 MdD

MeO

MdD.

^V^^^

84 —H

86

—H

—O-p-D-Glc/7

87

— O-p-D-Glcp

—H

R3

H

—(M

H

OH

—m

0

H

CM

H

0-P-D-Glc/?

— OH

89

—H

R2

Rl 88

85

90

=

91

= 0

92

=

0

OH

—m

93

=

0

CM

— 0-P-D-Glcp

94

H

= 0

OR

901

Rl

R2

R3

R4

95

— H

—H

— OH

—OH

96

—H

—H

—OH

......OH

97

—H

—OH

—H

—OH

98

=

—H

—OH

—H

99

—H

—OH

—OH

— H

0

OMe

100

902 Diarylheptanoids with Cyclised C^-Chain (Type V) The smallest subgroup of diarylheptanoids (Type V; Table 5) has grown considerably since 1993. At that time, only five representatives were known, but during this review period an additional 12 compounds have been described. The compounds 103-106, that were isolated fi-om Alpinia blepharocalyx, are noteworthy in that they have a chalcone or a flavanone moiety attached to the diarylheptanoid motif (cf. Table 5). These compounds, together with 113, all inhibited nitric oxide (NO) production in lipopolysaccharide (LPS)-activated murine macrophages in vitro [22]. A cyclised derivative, 114, of the oldest known diarylheptanoid curcumin (57) has been found in Curcuma longa. This cyclocurcumin (114), together with curcumin (57) and some other curcumin derivatives (58 and 59) showed appreciable nematocidal activity against dog roundworm [27]. (-)-De-(9methylcentrolobin (102) exhibited toxic activity against Leishmania amazonensis [8]. Table 5. Diarylheptanoids with cyclised C7-chain Compd

Trivial name

Occurrence*

Ref.

|l02

(-)-De-O-methyl-centrolobin

Centrolobium sclerophyllum (F)

1, 8

1 103

Calyxin F

Alpinia blepharocalyx (Z)

21, 22 1

|l04

6-Hydroxycalyxin F

Alpinia blepharocalyx (Z)

21,22 1

1 105

Calyxin G

Alpinia blepharocalyx (Z)

21, 22 1

|l06

Epicalyxin G

Alpinia blepharocalyx (Z)

21, 22 1

1 107

-

Zingiber officinale (Z)

48

1

1 108

-

Zingiber officinale (Z)

48

1

1 109

-

Zingiber officinale (Z)

48

1

IllO

-

Zingiber officinale (Z)

48

1

1 ^^^

-

Zingiber officinale (Z)

48

1

|ll2

Rhoiptelol B

Rhoiptelea chiliantha (R)

13

1

|ll3

-

Alpinia blepharocalyx (Z)

22, 49 1

Cyclocurcumin

Curcuma longa (Z)

27

114

'(F)=Fabaceae; (R)=Rhoipteleaceae; (Z)=Zingiberaceae

1

1

903

Rl

102

R2

—H

CMeo 103

CMeo 104

105

106

•'•OH

CMeo

904

MeO.

Rl

R2

107

—OH

—CM

108

•—OH

—CMe

109

^-QAc

—(M

110

•••H.OH

—CM

111

••'•iiQAc

—OH

MeO,

113

(Me

114

905

CONCLUDING REMARKS During the time period of review, diaryIheptanoids have been reported from plants in seven different families (Table 6). This includes the first reports of diarylheptanoids in the Pinaceae and Rhoipteleaceae. Taken together with the previously known taxonomic distribution of diarylheptanoids [1], this class of secondary metabolites are now known to appear in 10 different families, namely Aceraceae, Betulaceae, Burseraceae, Casuarinaceae, Juglandaceae, Fabaceae, Myricaceae, Pinaceae, Rhoipteleaceae and Zingiberaceae, with the largest numbers of representatives in the Betulaceae and Zingiberaceae. Table 6. Taxonomic distribution of diarylheptanoids reported in the literature during October 1993 - September 1999

Plant

family

Genus with diarylheptanoids

Type* and number of diarylheptanoids

|

I

II

Ill

IV

V

Aceraceae

I Acer

-

2

3

8

-

Betulaceae

Alnus

3

6

-

-

-

Betula

-

8

10

3

-

Corylus

-

1

7

-

-

Juglans

-

-

-

1

-

Platycarya

-

-

-

1

-

Myricaceae

Myrica

-

-

7

2

-

Pinaceae

Pinus

-

5

-

-

-

Rhoipteleaceae

Rhoiptelea

-

1

1

-

1 1

Zingiberaceae

Alpinia

3

19

-

-

Curcuma

5

2

-

-

5 1 1 1

Zingiber

-

-

-

-

5

1 Juglandaceae

'Type I: non-phenolic linear diarylheptanoids; Type II: phenolic linear diarylheptanoids; Type III: macrocyclic biarylheptanoids; Type IV: macrocyclic diaryl ether heptanoids; Type V: diarylheptanoids with cyclised C7-chain.

906

As seen in the sections above, several of the diarylheptanoids of Types I, H, and V have been reported to exhibit various types of biological activities. Most of these studies have been performed on biomedically relevant targets in vitro (e.g. protein kinase C [14], LPS-activated macrophages [22], CCI4challenged hepatocytes [10], but in vivo anti-inflanmiatory effects have also been demonstrated for some compounds [3, 17]. Toxic effects on parasites, such as nematodes [6,27] and trypanosomatids [8] have also been recorded for the first time. It is worth noting that still no biological effects have been reported for the macrocyclic diarylheptanoids.

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15]

Claeson, P.; Tuchinda, P.; Reutrakul, V.; / Indian Chem. Soc, 1994, 77, 509521. Keserii, G.M.; N6grM, M. In Studies in Natural Products Chemistry^ Atta-urRahman, Ed.; Elsevier Science B.V: Amsterdam, 1995; Vol. 77, pp. 357-394. Claeson, P.; Pongprayoon, U.; Sematong, T.; Tuchinda, P.; Reutrakul, V.; Soontomsaratune, P.; Taylor, W.C.; Planta Med., 1996, 52, 236-240. Athamaprasangsa, S.; Buntrarongroj, U.; Dampawan, P.; Ongkavoranan, N.; Rukachaisirikul, V.; Sethijinda, S.; Somnantrinda, M.; Sriwub, P.; Taylor, W.C; Phytochemistry, 1994, 37, 871-873. Tori, M.; Hashimoto, A., Hirose, K.; Asakawa, Y.; Phytochemistry, 1995, 40, 1263-1264. Jurgens, T.M.; Frazier, E.G.; Schaeffer, J.M.; Jones, T.E.; Zink, D.L.; Borris, R.P.; Nanakom, W.; Beck, H.T.; Balick, M.J.; J, Nat, Prod. 1994, 57, 230235. Shen, J.; Zhang, H.; Xu, B.; Pan, J.; Nat, Prod. Res. Dev., 1998,10, 33-36. Araujo, C.A.C.; Alegrio, L.V.; Leon, L.L.; Phytochemistry, 1998, 49, 751754. Fuchino, H.; Konishi, S.; Satoh, T.; Yagi, A.; Saitsu, K.; Tatsumi, T.; Tanaka, N.; Chem. Pharm. Bull. 1996, 44, 1033-1038. Matsuda, H.; Ishikado, A; Nishida, N.; Ninomiya, K.; Fujiwara, H.; Kobayashi, Y.; Yoshikawa, M.; Bioorg. Med. Chem. Utt., 1998, 8, 2939-2944. Shiratori, S.; Nagumo, S.; Inoue, T.; Nagai, M.; Chi, H.J.; Chem. Pharm. Bull, 1994, 42, 960-962. Watanabe, N.; Sasaya, T.; Sano, Y.; Mokuzai Gakkaishi, 1994, 40, 12191225. Jiang, Z.; Tanaka, T.; Hirata, H.; Fukuoka, R.; Kouno, I.; Phytochemistry, 1996,43, 1049-1054. Lee, K.K.; Bahler, B.D.; Hofmann, G.A.; Mattem; M.R.; Johnson; R.K; Kingston, D.G.I.; /. Nat.Prod, 1998, 61, 1407-1409. Wada, H.; Tachibana, H.; Fuchino, H.; Tanaka, N.; Chem. Pharm. Bull., 1998, 46, 1054-1055.

907

[16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41]

Gonzdlez-Laredo, R.F; Helm, R.F; Chen, J.; Karchesy, J.J.; / Nat. Prod. 1998, 61, 1292-1294. Yamazaki, R.; Aiyama, R.; Matsuzaki, T.; Hashimoto, S.; Yokokura, T. Inflamm. Res.; 1998,^7, 182-186. Suksamram, A.; Eiamong, E.; Piyachaturawat, P.; Charoenpiboonsin, J.; Phytochemistry, 1994, 36, 1505-1508. Ngo, K.-S., Brown, G.D.; Phytochemistry, 1997, 47, 1117-1123. Kadota, S.; Hui, D.; Basnet, P.; Prasain, J.K.; Xu, G.; Namba,T.; Chem. Pharm. Bull 1994, 42, 2647-2649. Prasain, J.K.; Tezuka, Y.; Li, J.X.; Tanaka, K.; Basnet, P.; Dong, H.; Namba, T.; Kadota, S.; J. Chem. Research (S), 1998, 22-23. Prasain, J.K.; Tezuka, Y.; Hase, K.; Basnet, P.; Dong, H.; Namba, T.; Kadota, S.; Biol. Pharm., Bull., 1998, 21, 371-374. Prasain, J.K.; Li, J.X.; Tezuka, Y.; Tanaka, K.; Basnet, P. Dong, H.; Namba, T.; Kadota, S.; J. Nat. Prod.,199S, 61, 212-216. Prasain, J.K.; Tezuka, Y.; Li, J.X.; Tanaka, K.; Basnet, P.; Dong, H.; Namba, T.; Kadota, S.; Tetrahedron, 1997, 53, 7833-7842. Kadota, S.; Prasain, J.K.; Li, J.X.; Basnet, P.; Dong, H.; Tani, T.; Namba, T.; Tetrahedron Lett. 1996, 37, 7283-8286. Dong, H.; Chen, S.; Xu, H.; Kadota, S.; Namba, T.; / Nat. Prod. 1998, 61, 142-144. Kiuchi, F.; Goto, Y.; Sugimoto, N.; Akao, N.; Kondo, K., Tsuda, Y.; Chem. Pharm. Bull., 1993, 41, 1640-1643. T0nnesen, H.H.; Arrieta, A.F.; Lemer, D.; Pharmazie, 1995, 50, 689-693. Nagai, M.; Dohi, J.; Morihara, M.; Sakurai, N.; Chem. Pharm. Bull, 1995, 43, 1614-1611. Morihara, M.; Sakurai, N., Inoue, T., Kawai, K.-i., Nagai, M.; Chem. Pharm. Bull, 1997, 45, 820-823. Sakurai, N.; Yaguchi, Y.; Hirakawa, T.; Nagai, M; Inoue, T.; Phytochemistry, 1991, 30, 3077-3079. Fuchino, H.; Satoh, T.; Shimizu, M.; Tanaka, N.; Chem. Pharm. Bull. 1998, 46, 166-168. Nagumo, S.; Kaji, N.; Inoue, T.; Nagai, M.; Chem. Pharm. Bull. 1993, 41, 1255-1257. Fuchino, H.; Satoh, T.; Tanaka, N.; Chem. Pharm. Bull 1995, 43, 1937-1942. Fuchino, H.; Satoh, T.; Tanaka, N.; Chem. Pharm. Bull 1996, 44, 1748-1753. Nagumo, S.; Ishizawa, S.; Nagai, M.; Inoue, T.; Chem. Pharm. Bull, 1996, 44, 1086-1089. Hanawa, F.; Shiro,M.; Hayashi, Y.; Phytochemistry, 1997, 45, 589-595. Lee, M.W.; Tanaka, T.; Nonaka, G.I.; Hahn, D.R.; Arch. Pharm. Res., 1992, 15,211-314. Watanabe, N.; Sasaya, T.; Mokuzai Gakkaishi, 1994, 40, 199-203. Tanaka, T.; Jiang, Z.-H.; Kouno, I.; Phytochemistry, 1998, 47, 851-854. Kim, S.H.; Lee, K.S.; Son, J.K.; Je, G.H.; Lee, J.S.; Lee, C.H.; Cheong, C.J.; / Nat. Prod., 1998, 61, 643-645.

908

[42] [43] [44] [45] [46] [47] [48] [49]

Keserii, G.M.; N6grM, M.; Szollosy, A.; Eur. J, Org. Chem. 1998, 521-524. Gonzalez, G.I.; Zhu, J.; J, Org. Chem. 1999, 64, 914-924. Gonzalez, G.I.; Zhu, J.; J. Org. Chem. 1997, 62, 7544-7545. Fuchino, H.; Satoh, T.; Yokochi, M.; Tanaka, N.; Chem. Pharm. Bull. 1998, 46, 169-170. Keserii, G.M.; Dienes, Z.; N6grddi, M.; Kajt^-Peredy, M.; J. Org. Chem. 1993, 58, 6725-6728. Keserii, G.M.; N6grddi, M.; J. Mol. Struct. (Theochem). 1993, 286, 259-265. Kikuzaki, H.; Nakatani, N.; Phytochemistry, 1996, 43, 273-277. Prasain, J.K.; Tezuka, Y.; Li, J.X.; Tanaka, K.; Basnet, P. Dong, H.; Namba, T.; Kadota, S.; Planta Med., 1999, 65, 196.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

909

NITRIC OXIDE : CHEMISTRY AND BIOACTIVITY IN ANIMAL AND PLANT CELLS DAVID WENDEHENNE, ^LAURE DUSSABLY, ^JEAN-FRANCOIS JEANNIN and ALAIN PUGIN Unite Mixte INRA/Universite de Bourgogne, BBCE-IPM Laboratory, INRA BV1540, 17 rue Sully, 21034 Dijon cedex, France; ^EPHE/INSERM, Laboratory of Cancer Immunotherapy and Unit 517, Faculty of Medicine, 7 Bd Jeanne d'Arc, 21033 Dijon cedex, France ABSTRACT: In mammals, nitric oxide (NO) is a reactive free radical involved in diverse physiological functions. NO and its redox-related forms NO^ and NO" react with di(oxygen) and its derivatives, with metalloproteins and thiol-containing proteins. NOmediated nitrosation of proteins represents an important cellular regulatory mechanism. Biosynthesis of NO is catalysed by nitric oxide synthase (NOS). Three isoenzymes representing distinct gene products have been identified: the inducible NOS isoform, the constitutive neuronal and endothelial isoforms. Inducible and constitutive NOSs have the same structural features, but their activities differ in their dependence to calcium and the rate of NO produced. The principal NO-mediated functions in mammals are endothelium-dependent relaxation, neurotransmission and immune response. The role of NO in the antitumor immune response comprises both regulatory and effector functions at the intra- or inter-cellular level. The first function includes inhibition of lymphocyte proliferation or participation in different transduction pathways. The second function includes pro- or anti-tumoral effects and NO-mediated cell toxicity or cell resistance to apoptosis. In plants, NO is produced via non-enzymatic reactions, by the activity of a constitutive nitrate reductase, and by a mammalian-type NOS which corresponding gene has not yet been cloned. NO was found to regulate plant growth, to act as antioxidant, and to function as a signaling molecule in plants challenged by pathogens. Interestingly, key proteins involved in animal NO signaling appear to be modulated by NO in plants. This finding complements the list of evidence suggesting that portions of transduction mechanisms are shared between plants and animals.

INTRODUCTION At the begining of the century, nutrition experiments first suggested that mammals produced oxides of nitrogen [1]. Seventy five years later it was confirmed in germ-fi-ee rats that the host and not its enteric flora produced nitrates (NOs") [2]. This production was enhanced by infection in humans and by lipopolysaccharide (LPS) in rats which implicated the

910 reticuloendothelial system. It was later confirmed that macrophages were the origine of NO3" in LPS treated mice [3]. Then it was shown that arginine was the substrat for macrophage nitrites (NO2") and NOs" production and was required for macrophage-mediated cytotoxicity [4]. Nitric oxide (NO) closely mimicked the cytotoxic effect of activated macrophages, confirming that NO was the active intermediate of the Larginine to NO2" and NO3" cytotoxic pathway [5-7]. In the same time NO was identified as the endothelium-derived relaxing factor (EDRF), a key molecule responsible for vascular smooth muscle relaxation [8-10]. Further studies established NO as a multifunctional effector molecule involved in diverse biological fiinctions including inhibition of platelet aggregation, neurotransmission and immune regulation [for review see 11]. Many insights into the understanding of NO functions came from studies of NO complex chemistry [12] and the identification of nitric oxide synthase (NOS), the enzyme responsible for NO production [13]. The surprising versatility of NO lead Science magazine to recognize NO as "the molecule of the year" in 1992 [14]. Since then, NO has become established as a universal inter/intracellular messenger that transduces its effects by chemical reactions with specific targets such as guanylate cyclase, a major signaling protein in animal cells [for review see 11]. Besides its beneficial effects, NO was also reported to exert deleterious effects and to induce apoptotic cell death when produced at high concentrations under pathophysiological conditions [for review see 15]. Recently, considering the importance and the impact of NO research, the Nobel Prize Commitee awarded the 1998 Prize in Medicine to Drs Furchgott, Ignarro and Murad, pioneers in NO biology. The recent discoveries that plants produce and use NO in diverse biological functions, ranging from development to disease resistance, has also stimulated a plethora of enthusiastic investigations [16-18]. Interestingly, although understanding of NO activities in plant is only at its early stage, first conclusions suggested that some aspects of NO biology are shared between plants and animals. Considering the impressive and diversified literature related to NO in animals, we have had to be selective in the present review. In the first part, we have attempted to give a general view of NO chemistry. Then, in a second part we described the biosynthesis of NO in mammals and detailed its role in the antitumor immune response. The first and second parts serve as background to the third part focalised on the recent advances on NO functions in plants.

911 CHEMISTRY OF NO NO is a gaseous molecule with a single electron in its Ip-n antibonding orbital (according to the International Union of Pharmacology Nomenclature in Nitric Oxide Research, the unpaired electron does not need to be written into the abbreviation for NO as NO- [19]). NO is poorly soluble in water but can diffuse relatively easily across biological membranes. NO diffusion constant is 3300 |Lim^/sec and it needs 5 to 15 seconds to reach a 150 to 300 |am radius (which represents about 20 cells). Because NO is a free radical, it is highly reactive. Thus, once produced, NO is extremely susceptible to both oxidation and reduction and reacts rapidly with (di)oxygen and its various redox forms, other free radicals and transition metal ions [11, 12]. Consequently, the half-life of NO in oxygen-containing aqueous solutions is short, generally admitted to be in the order of seconds. Fig. (1) gives a summary of NO reactions in biological systems. Oxyhemoglobin hemoglobin

NO3 pKa = N2O < ^

NO

< ^

/ )OH ONOOH

NO -^^^^-^

M-Fe(III)

\ NO

+ HO

M-Fe(III) NO

\ RS-NO <

NO NO

^

RO-NO

RNH-NO

Fig. (1). Summary of the chemistry of NO in biological environment [11, 12] (see the text for details)

912 NO reacts with O2 to yield N02- which is converted to NO2' and NOs". During these reactions, the intermediate N2O3 is formed. Relative excesses of N2O3 and N02- account for NO toxicity since both compounds can efficiently oxidize cellular thiols and amines. NOs" is a major stable metabolite of NO and is assumed to be NO's end-metabolite [20]. According to Ignarro et al. [21], the primary decomposition product of NO in biological aqueous solutions is NO2" and its further oxidation to NO3" requires the presence of oxidizing species such as oxyhemoglobin or oxymyoglobin. The reaction of NO with superoxide (02-) results in the generation of peroxynitrite (ONOO"). Many studies have implicated ONOO" as being a potent and destructive oxidant [11, 12, 22, 23]. For example, it has been demonstrated that ONOO" triggers oxidation of sulfhydril functions and lipid peroxidation. Interestingly, although reports describe ONOO" as an inducer of apoptotic cell death, contradictory results claim that the production of ONOO" represents a protective mechanism against 02- toxicity [24]. At physiological pH, ONOO" equilibrates with peroxynitrous acid (ONOOH) (pKa = 6.8). Depending on its conformation, ONOOH decomposes rapidly to NO3" by intramolecular rearrangements {cis conformation) or to N02- and the radical oxidant hydroxyl radical (H0-) by homolytic cleavage {trans conformation) [25]. Moreover, ONOO" has been shown to react with glutathione (GSH), generating the intermediate compound Snitroglutathione (GSNO2) which acts as a NO donor [26]. NO also interacts covalently with a variety of heme or iron sulfur cluster-containing proteins, forming iron-nitrosyl complexes [11]. This interaction leads to either an activation or an inhibition of the activity of the target protein. For example, binding of NO to the mammalian cytosolic guanylate cyclase heme center induces structural modification in the enzyme. This process leads to the activation of the enzyme and subsequently to an increase in the cellular cyclic GMP (cGMP) level [24, 27, 28]. In contrast, when heme or iron sulfur cluster centers participate in the catalysis, attack by NO mainly results in the inhibition of the corresponding protein as reported for the inhibition of aconitase, cytochrome P450s, cytochrome c oxidase, indoleamine 2,3-dioxygenase and NOS itself [for review see 11]. One electron oxidation of NO generates the nitrosonium cation (NO"^). In this reaction, the iron atom of Fe(III)-containing metalloenzymes acts as the electron acceptor. The metal-nitrosyl complex formed (Fe(II)-NO'^)

913 serves as a NO^ carrier, releasing NO"*^ to specific targets. NO"^ reacts with nucleophilic centers in organic molecules. This process, referred to as nitrosation, generates nitroso-compounds including nitrosamines (RNHNO), alkyl or aryl nitrite (RO-NO) and thionitrites (RS-NO), also named S-nitrosothiols. A variety of nitroso-compounds are considered as NO^ carriers that transfer NO"^ (also NO and the nitroxyl anion NO") to specific effector sites in a process referred to as transnitrosation [12, 29]. This is well examplified by S-nitrosothiols which are widely reported to exhibit NO-like biological activities and may provide a means to control NO toxicity [30]. (Trans)nitrosation of sulfhydryl residues, defined as Snitrosylation, has been shown to regulate the function of numerous mammalian proteins, including signaling proteins such as the N-methylD-aspartate (NMDA) receptor, the Ca channel ryanodine receptor (RYR), the small GTP-binding protein p21'^^, protein kinase C and protein tyrosine phosphatases; enzymes such as glyceraldehyde-3phosphate dehydrogenase (GADPH), cathespin B, caspase-3 and yglutamylcysteine synthetase (y-GCS); and nuclear proteins such as the DNA repair protein 0^-methylguanine-DNA-methyl-transferase [for review see 11 and 31]. Protein S-nitrosylation is reversible and recent studies strongly suggest that S-nitrosylation/denitrosylation of target proteins is a regulatory process in signal transduction pathways [32]. Mechanisms by which S-nitrosylation modulates the activity of target proteins include the formation of intramolecular disulfide (Fig. (2)) [12]. This process is accompanied by the release of NO or NO", depending on the physiological conditions. Moreover, it is of importance to specify that multiple sulfhydril groups of proteins are susceptible to S-nitrosylation (this process is defined as poly-S-nitrosylation). For example, poly-Snitrosylation of the Ca^"^ channel RYR triggers the oxidation of 12 -SH groups and the subsequent activation of the channel [33]. One electron reduction of NO generates NO". This reduction is supported by Fe(II) ion and by Fe(II)-containing metalloproteins which act as electron donors. The chemistry of NO" has been less studied than NO or NO^ chemistry. However, NO" is believed to mediate sulfhydryl oxidation of target proteins with the intermediate formation of RSNOH [12]. This process leads to the formation of nitrous oxide (N2O) which is also the result of NO" spontaneous dimerization [12].

914

NO

•

r

RS

RS-SR

RS NO^

•>

RS-NO RS

k

RS-SR

NO

Fig. (2). Simplified representation of NC'-mediated intramolecular disulfide formation [11, 12]

In summary, the chemistry of nitric oxide involves the three redox-related, but chemically distinct, species of NO- (referred as NO), NO", and NO"*". NO" and NO"^ may have bioactivities similar but also different from those of NO. NO can react either directly or by way of nitro-coumpounds with iron and/or thiol-containing proteins. In order to simplify, the signal transduction pathways of NO have been classified as either cGMPdependent (pathways controled by the NO-mediated activation of guanylate cyclase) or cGMP-independent (signaling by S-nitrosylation) [16, 34]. Both processes may represent an important cellular regulatory mechanism in many biological systems. NO IN MAMMALS NO synthesis in mammals In mammals, the biosynthesis of NO is catalysed by the enzyme NOS. So far, three isoenzymes that represent distinct gene products have been identified: the inducible (iNOS) isoform, the constitutive endothelial (eNOS) isoform and the constitutive neuronal (nNOS) isoform (table I). Comparison of the respective amino acid sequences shows that these three isoforms are about 50-60% identical (for review see [35]).

915 Table 1.

'

Classification of nitric oxide synthases [37, 55]

Isoform

Other nomenclatures

Chromosomal Localization (human)

Cellular localization

Molecular mass (kDa)

nNOS

NOS I, ncNOS

12(12q24.2)

Soluble and Particulate

161

eNOS

NOS III, ecNOS

7 (7q35-36)

particulate

133

iNOS

NOSH

17(17cen-ql2)

soluble

131

nNOS was first described in specific neurons of the central and peripheral nervous sytems and has since been found in nonneuronal cell types, including myocytes, epithelial cells and neutrophils (for review see [36, 37]). The human nNOS gene is present as a single copy on chromosone 12 and spans approximatively 200 kb of DNA [38, 39, 40]. It contains 29 exons and consensus sequences for the binding of transcription factors including the activator protein-2 (AP-2), nuclear factor-1 and nuclear factor-KB ( N F - K B ) . The mRNA transcript encodes a polypeptide of 1434 amino acids with a predicted molecular mass of 161 kDa. Soluble and particulate nNOS are found, depending on the cell types studied [41]. This enzyme does not associate directly with membranes, but the NH2-terminal sequence possesses a repeating amino acid sequence, called the PDZ motif (Gly-Leu-Gly-Phe motif), which targets the protein to membrane associated proteins such as the post-synaptic density proteins-93 and -95 (PSD-93 and PSD-95, respectively) [42, 43]. The endothelial NOS was first described in endothelial cells and was later identified in other cell types including neurons, T-cells, myocytes or colon interstitial cells (for review see [36, 37]). The human eNOS gene has been assigned to chromosome 7 and is present as a single copy [44]. The locus is distributed over a region of 21-22 kb [44, 45]. The eNOS promoter contains potential binding sites for transcription factors such as N F K B , nuclear factor interleukin-6 (IL-6), activator protein-1 (AP-1), AP-2 and shears stress-induced transcription factors [45, 46]. The eNOS mRNA is encoded by 26 exons and codes for a protein of 1203 amino acids with a predicted molecular mass of 133 kDa. eNOS is primarily membrane associated and has been shown to be targeted to Golgi and plasma

916

membranes. The molecular targeting of the enzyme is determined by its NH2-terminal fatty acylation sites which contain myristoylation and palmitoylation sites [47, 48]. This membrane association is necessary for the enzyme's activity [49]. Moreover, eNOS has been shown to specifically interact with caveolin isoforms which are transmembrane proteins found in cell caveolae [50]. Conversely, this interaction seems to negatively regulate eNOS activity [51]. Both nNOS and eNOS are considered as constitutively expressed proteins and show transient and weak activation (pmol NO/min/mg protein) upon signals. However, recent data provide evidence that these two isoforms are subjected to transcriptional regulation by various agents such as allergic substances, growth factors, or sex hormones in physiological or pathological conditions [36, 40]. Interestingly, depending on the stimulus and the tissue specificity, cells differentially regulate nNOS transcription through complex processes including cassette exon deletions/insertions, alternate polyadenylation signals and transcriptional initiation by different transcriptional units containing alternative promoters [36, 40]. This leads to diverse nNOS transcripts whose corresponding proteins may have different catalytic properties and cellular localization. Thus, according to the stimulus, cells can adjust nNOS activities and perhaps eNOS activities through similar transcription regulation. Whereas evidence of transcriptional regulation of constitutive NOS has been shown only recently, transcriptional regulation of iNOS has been established for about 10 years. The iNOS isoform is expressed in macrophages and many other cell types in response to endotoxin and inflammatory agents such as the cytokines tumour necrosis factor a (TNF-a), IL-1 and interferon-y (IFN-y) (for review see [37]). Analysis of the human iNOS locus indicated that the iNOS gene is 37 kb in length, contains 26 exons and is located as a single copy on chromosome 17 [52]. Expression of the iNOS gene is complex and is regulated in part by gene transcription. Indeed, the human iNOS promotor contains a number of putative transcription factor binding sites, many of them being probably indispensable for the gene NpKB-dependent induction [24, 37, 53, 54]. Other transcription factor binding sites include the interferon regulatory factor-1 binding site, the gamma-interferon activated site and cyclic AMP (cAMP) response elements. Once the gene is induced, a delay of 6-8 hours is required before obtention of the active iNOS protein. iNOS is a soluble protein of 1153 amino acids (131 kDa) which, once induced, is active from hours to days and generates two to

917

three orders of magnitude more NO (nmol NO/min/mg protein) than the constitutive eNOS and nNOS enzymes [24]. NH3

fatty acylation sites PDZ motif (nNOS)

"\

BH4 binding

Nterminal extension

heme Oxygenas e domain

site confering Ca^"^ -independent CaM binding (iNOS) CaM inhibitor sequence (eNOS and nNOS)

J

CaM-binding

-N

site

FMN domain

FAD domain

I

Reductas e domain

NADPH domain

J

coo

Fig. (3). Functional domains oftheNOSs [55]. CaM: calmodulin.

Each NOS has the same structural feature (Fig. (3)) [35, 37, 55-58]. The C-terminal region (named the reductase domain) shows significant homology (about 58%) with NADPH cytochrome P-450 reductase and contains one binding site each for flavin adenine dinucleotide (FAD), flavin mononucleotide (FMN) and NADPH. The N-terminal region (named the oxygenase domain) contains a cytochrome P-450-type heme center and a binding site for the cofactor tetrahydrobiopterin (BH4). Although the heme group bears characteristics similar to P450s, the NOS N-terminal domain does not reveal any convincing sequence homology with P450s. Recent determination of the oxygenase domain crystal structure has confirmed its unique structure, consisting largely of p-sheets contrary to any other known heme proteins [59]. The oxygenase and

918 reductase domains are connected by a canonical calmodulin-binding site located in the middle of the molecule. The N-terminal extension, which is not essential for catalysis, differs amongst the NOS isoforms. This domain determines the intracellular localization of the enzyme and contains the fatty acylation sites in eNOS and the PDZ domain in nNOS [42, 43, 47, 48]. Another important feature is that the enzyme is active as a homodimer. NOS subunit assembly occurs through the heme domains and binding of BH4 strongly stabilizes the complex by promoting conformational changes [55, 60]. Interestingly, in some studies calmodulin was shown to promote dimerization of the oxygenase domain of eNOS [61]. All NOS isoforms catalyse the hydroxylation of L-arginine to form A^hydroxy-L-arginine, which is subsequently converted to L-citruUine and NO (Fig. (4)). This reaction requires molecular oxygen (O2) and NADPH. NH2

+ NO

^H N^'-hydroxy-L-

COO L-

Fig. (4). The reaction catalysed by NOS [35]

Although the exact chemical steps involved in the conversion of Larginine to L-citrulline and NO are not known, several studies have reported that the heme is directly involved in the oxidation of L-arginine [35, 55, 56, 62]. To go into further detail, electrons are transferred via the flavin FAD and FMN from NADPH to the heme group where O2 is reduced and incorporated into the guanidino group of L-arginine. The electron flow is strictly dependent on calmodulin binding. Indeed, calmodulin functions as a molecular switch between the two domains of NOS and consequently allows the electron to pass from the flavins to the heme iron (Fig. (5)) [62, 63]. Interestingly, it has been shown that besides

919 synthesizing NO, NOS reduces O2 to 02-" under conditions of L-arginine depletion [64-67]. Surprisingly, Or' synthesis from eNOS or nNOS occurs at the heme center of the oxygenase domain whereas iNOSmediated 02-' generation occurs primarily at the flavin-binding sites of its reductase domain. Moreover, Xia et al. [66] reported the ability of iNOS to produce 02*' even in the presence of L-arginine, thus suggesting that iNOS can simultaneously synthesize NO and 02-" which interact to form ONOO" [64]. In addition, the capacity of NOS to produce O2" or ONOO" seems to be closely related to the cofactor BH4. According to Meyer and Hemmens [55], NOS catalyses 02-" production in the absence of BH4, whereas the enzyme with one BH4 per dimer is likely to generate ONOO'. Thus, this cofactor seems to affect both the conformation and the activity of NOS. Much effort is currently being directed to understand the exact function of BH4. Chemical studies and NOS crystal structure analysis have shown that BH4 binds close to the heme group [60, 68]. According to this model, BH4 may exert an electronic influence on heme-bound oxygen and could act as a transient electron donor during NO catalysis. The activity of constitutive NOS isoforms is strictly controlled by Ca^"^ and calmodulin. At resting cytosolic free Ca^^ concentrations (70200 nM), calmodulin is largely free of Ca^"^ and so does not bind to eNOS or nNOS. In response to increased cytosolic Ca^^ concentrations, a Ca^^calmodulin-NOS complex is formed, resulting in the activation of the enzyme. Interestingly, studies have suggested that nNOS could be indirectly associated with the NMDA receptor, which functions as a plasma membrane Ca^"^ channel in neuronal cells [36]. The NMDA receptor contains a C-terminal SXV motif which interacts with PSD95 [69]. Since nNOS has also been postulated to bind to PSD95 [43], PSD95 could enable a physical coupling between the NMDA receptor and nNOS. In this molecular complex, NOS might be rapidly and specifically activated in response to NMDA receptor-mediated localised Ca^"^ influx. An important difference between constitutive and inducible NOS isoforms is that iNOS forms a very tight complex with calmodulin at resting free cytosolic Ca^"^ concentrations. Thus, iNOS binds calmodulin independently of Ca^"^ and becomes fully active even at low Ca^"^ concentrations [70]. Because iNOS activity is independent of elevated Ca^"^, it is sustained contrary to the constitutive NOS activities which depend on transient Ca^"*" increases [37]. iNOS-specific regulation by calmodulin is explained, at least in part, by the absence of a 40-50 amino

920

acid sequence inserted within the FMN-binding domain of the constitutive NOS (Fig. (3)) [55, 71, 72]. It has been proposed that this sequence forms an autoinhibitory domain which regulates constitutive NOS activity by inhibiting electron transfer from FMN to heme in the absence of Ca^"^calmodulin complex and by destabilizing calmodulin binding at low Ca^"^ concentrations. Moreover, it has been proposed that the Ca'^^-independent binding of calmodulin to iNOS is conferred by a specific site in the Nterminal portion of the reductase domain [73]. Additional investigations are necessary to verify this model.

FIMN

NAOPH

FAO

FAD

NADPH

FMN

FAI)<4f+ NADP

L-citrulline n

Fig. (5). Schematic functioning of NOS [56]. NOS is represented as an homodimer (dimerization occurs through the heme domain of each subnit). Calmoduhn (CaM) binding may tether the oxygenase and the reductase domain close together, thus promoting the electron transfer from NADPH to the heme center (H) via FAD and FMN. nNOS and eNOS depend upon Ca^^calmodulin for activation, whereas the activity of iNOS is Ca^^ independent.

921 Roles of NO in mammals The three principal NO-mediated functions in mammals are endotheliumdependent relaxation, neurotransmission and immune response. In vascular system, NO is involved in the regulation of vascular tone, angiogenesis, adhesion of circulating blood cells and platelet aggregation. As neurotransmitter, NO regulates intestinal peristalsis, autonomic and neuroendocrine functions and regulation of behavior. NO is also involved in the pathophysiology of these systems: hypertension, atherosclerosis, heart failure, ischemia, excytotoxicity, neurodegeneration, Parkinson's disease. In the immune responses, NO is effective against various parasites, bacteria and virus and against tumor cells. However some microbial species are NO resistant and in some situations NO has an immunosuppressive effect or promote tumor angiogenesis. NO is involved in inflammation: arthritis, ulcerative colitis, Crohn's disease, erythema and participates in graft rejection, graft versus host disease and sepsis injuries. This specification show^s the role either beneficial or deleterious of NO in organism. Here we will detail the role of NO in the antitumor immune response and show that its ambivalency is determined by its concentration, the duration of its production and its surrounding microenvironment. NO as a mediator of the antitumor immune response

Specific antitumor immune response : Helper T (Th) lymphocytes and cytotoxic T (CT) lymphocytes recognize peptides that are associated with major histocompatibility complex (MHC) molecules on the surface of antigen-presenting cells (APC) or tumor cells. Antigen recognition by T lymphocytes is the initiating stimulus for their activation: proliferation (clonal expansion of antigen-specific T lymphocytes), differentiation (Th or CT lymphocytes), effector functions (cytokine production, expression of surface molecules, tumor cell destruction). Some of the antigen-specific T lymphocytes will develop into memory T lymphocytes. The T cell receptor (TCR) for antigen is composed of membrane proteins and specifically binds to peptide-MHC complexes on the surface of APC or tumor cells. The TCR is expressed in association with transmembrane proteins (cell surface protein 3 (CDS) and CD4 expressed on Th lymphocytes, or CDS

922

expressed on CT lymphocytes) that transduce intracellular signals required for T lymphocyte activation. In addition T lymphocytes express a number of accessory molecules that strengthen their adhesion (Th lymphocytes with APC and CT lymphocytes with tumor cells). Some accessory molecules transduce intracellular signals which enhance the reponses of T lymphocytes. Accessory molecules expressed on T lymphocytes are CD28, VLA-4, LFA-1 and CD2 which bind respectively B7, VCAM-1, ICAM-1 and LFA-3 expressed on APC, and cytokines which bind their receptors. In vitro, activated Th lymphocytes produce IFN-y that stimulates APC iNOS gene [74, 75]. The binding of LFA-1 and CD2 to ICAM-1 and LFA3 induce NO production by macrophages and dendritic cells [76, 77]. NO diffuses in Th lymphocytes and stimulates p2r^ and thus activates specific T lymphocytes proliferation. NO directly activates p2r^ and the src family protein tyrosine kinase p56*^^ in T lymphocytes. It transduces activation signals from TCR and induces transcription of genes coding for cytokines or their receptor. plV^ activation leads to NFK-B translocation and induces, by a positive loop, NO production by T lymphocytes [78, 79]. It can easily be understood that NO has been described as indispensable to T lymphocyte proliferation or survival [80, 81]. However, NO produced by APC can, also by diffusion, inhibit T lymphocytes proliferation without inhibiting IL-2 production and IL-2 receptor expression [82]. In these experiments, phosphorylations of the janus kinase JAK3 and the Signal Transducer and Activator of Transcription protein STATS are inhibited by NO, this inhibition being reversible and cGMP dependent. Tyrosine kinases, which activity is the first step in the TCR signal transduction, can be inhibited by NO which avoids 004"^ lymphocytes proliferation and leads them to apoptosis [83]. Moreover, NO inhibits MHC II molecules expression that are induced by IFN-y on macrophages [84] and astrocytes [85] surfaces, which is also a negative regulation of antigen presentation. Still in vitro, stimulation by antigen of specific Th lymphocytes provokes maximal proliferation and IL-2 production when NO production by lymphocytes is minimal, which is dependent of antigen concentration. When the latter increases, NO production increases as well and Th lymphocytes activation decreases. This effect of NO is observed with Thl lymphocytes (IL-2 and IFN-y producing cells) but is not observed when Th2 lymphocytes (IL-4 and IL-10 producing cells) are stimulated by the antigen [86]. Stimulation by anti-CD3 antibody leads to NO production

923 and to an inhibition of Thl and Th2 lymphocytes: IL-2 and IFN-y synthesis plus IL-4 and IL-10 synthesis inhibition [87, 88]. In all cases, it is a post transcriptionnal NO dependent regulation. Balance between Thl and Th2 lymphocytes is dependent of both the effects of NO on Thl and Th2 which are different, and the effects of Thl and Th2 cytokines on iNOS expression which are also different [89]. Thl cytokines stimulate whereas Th2 cytokines inhibit iNOS expression. Furthermore, Thl and Th2 balance depends on arginine metabolism pathways: NOS and/or arginase [90]. Hence, NO produced by T lymphocytes after TCR/CD3 stimulation has an inhibitory effect on their activation. Furthermore, NO may be cytotoxic via the Fas ligand (FasL; Fas belongs to the tumor necrosis receptor family); T lymphocytes stimulation by anti-CD3 antibody activates iNOS; NO induces FasL expression on T lymphocytes surface; the coexpression of Fas and FasL provokes their apoptosis [91]. Therefore, NO plays a role in antigen presentation and specific T lymphocytes activation, whereas it does not seem to interact with antigen processing and CTL mediated tumor cell death. Non specific immune response : The first article about the role of NO in immunology has been written by Hibbs [5] who shows that NO mediates in vitro macrophage tumor cell toxicity. Numerous articles confirmed this observation. On the other hand the implication of NO in Natural Killer (NK) cell toxicity against tumor cells is controversial. NO would be an effector of NK cells [92-94] and LymphokineActivated Killer (LAK) cells cytotoxicity [92] in some instances, but in others it would inhibit NK activity [95]. NO would be a mediator of LAK cells activation [96] in some cases, but in others it would inhibit LAK cell activation by inhibiting IL-2 receptor expression [97] or by killing precursors [98]. Those differences are independent of the NK cells origins. It has also been demonstrated that NO does not participate to human NK cells activation by LPS nor to NK cells activity against K562 (a human leukemic cell line) target cells [81]. Therefore, NO is a mediator of antitumoral immune responses. If undoubtedly NO is one effector of macrophages mediated tumor cell toxicity in vitro, its role in NK and LAK mediated tumor cell toxicity is doubtfull.

924 NO as effector molecule of the antitumor immune response

NO and cell proliferation : NO has a cytostatic effect by inhibiting ATP synthesis [99] via Kreb's cycle (aconitase inhibition, [100]), glycolysis (GADPH inhibition) and mitochondrial respiration (NAD ubiquinone oxydoreductase and succinate ubiquinone oxydoreductase inhibitions, [101]). Another pathway is the ornithine decarboxylase inhibition. This enzyme is implicated in polyamine production necessary to cell proliferation and its activity is inhibited by NO in human colon cancer cells HT-29 and Caco-2 [102]. Furthermore NO directly inactivates ribonucleotide reductase [103] of TA3 cancer cells (murine breast cancer cells) [104]. This enzyme controlling DNA synthesis catalyses desoxyribonucleotides synthesis, and its inhibition blocks cells in S phase. This inhibition is rapid and reversible in K562 and TA3 cells [105]. Independently of these mechanisms, NO has a cytostatic effect when accumulating the wild type tumor suppressor protein p53 that activates genes encoding the cyclin-dependent kinase inhibitor p21^'^^ [106], the Growth Arrest and DNA Damage protein GADD45 [107] and the oncoprotein MDM2 [108]. By this way NO blocks Gl to S phase progression by inhibiting the Retinoblastoma protein Rb phosphorylation via p21^'^^^ stimulation. This is also by this pathway that NO mediates cytostatic effect of nerve growth factor (NGF) on PC 12 cells (isolated from a pheochromocytoma) and on SH-SY5Y cells (isolated from a neuroblastoma) [109]. p53 would be able to mediate cytostatic effect of NO (stopping in G2 phase) when it inhibits topoisomerase II expression, an enzyme implicated in DNA replication [110]. Other mechanisms have been implicated, NO provokes a cytostase: a) by ADP ribosylation of Ga subunits of G protein in C6 cells (glioma cells, [111]), b) by the inhibition of the proto-oncogene c-myb gene expression in U937 cell [112], c) by the inhibition of the kinase activity of growth factor receptors like EGF receptor [113]. In an other hand, by nitrosylation of p53 protein NO prevents its fixation on specific sites of regulatory DNA region, then removing the control of cell proliferation in MCF7 cells (human mammary cancer cell) [114].

925 NO and cell apoptosis : p53 blocks cells in Gl phase, that leads to DNA repair or apoptosis. Moreover, p53 regulates negatively Bcl-2 expression and positively Bax expression. Bcl-2 is an oncogene product which protects cells from apoptosis and belongs to a family of related proteins. Whereas some members of this family such as Bcl-2, Bclx, Bcl-w, Mcl-l act as death repressors; others such as Bax, Bad, Bak and Bik exert the opposite function (e.g. induce apoptosis). The Bcl-2 active form is a homodimer. Bax and Bak links to Bcl-2 to form inactive heterodimers. In tumor cells isolated from solid tumors, NO via p53 negatively regulates Bcl-2 expression, mRNA and protein synthesis [115, 116] and positively regulates Bax protein expression [117]. Otherwise, NO acts more downstream on caspases such as caspase 1 which activates caspase 3 which inactives itself poly-ADP-ribose-polymerase (PARP), an enzyme participating in DNA repair. NO, when activating caspase 1 and/or 2, inactivates PARP and allows apoptosis [118]. Meanwhile, NO can induce apoptosis of cancer cells by a caspase independent pathway but controlled by Bcl-2 [119]. One pathway or the other is reinforced when NO inhibits DNA-alkyltransferase, an enzyme that participates to DNA repair. In an other hand, apoptosis includes modifications affecting mitochondria, such as the reduction of membrane potential which is an early event in apoptosis. NO induces mitochondrial disfunctions [120] and cytochrome c release [118] that seem to activate caspases. In the already cited articles, NO is produced by NO donors. It can be produced by iNOS activity in macrophages [120, 121], polynuclear cells [122] or endothelial cells [123], which provokes tumor cells apoptosis. NO can also be produced by tumor cells iNOS. When iNOS gene is activated by cytokines: IFNy and IL-ip in rat colon cancer cells, IFNy, ILip, TNFa in human colon cancer cells, TNFa in murine colon cancer cells, IL-la in human ovarian cancer cells, NO has no apoptotic effect. Concerning NO implication in hematopoietic cancer cells apoptosis, the situation is less clear. In U937 and HL60 cells, NO effects are very different although both cell lines have been established in culture from promyelocytic leukemia. NO-induced apoptosis in U937 cells (p53 negative cells) is a p53 and caspase independent pathway but is controlled by Bcl-2 [124]. In these cells NO inhibits caspase 3 by S-nitrosation [125]. In HL-60 cells NO activates caspase 3 and provokes apoptosis [126]. Concerning lymphocytic leukemia, in fresh cells isolated from type B chronic leukemia, when iNOS gene is activated by CD23 ligation, NO

926

inhibits spontaneous apoptosis. NO produced by NOS inhibits the Fas induced apoptosis [127, 128]. Cyclic GMP has been implicated in this effect [129], but the protective effect is likely cGMP independent in the Burkitt's lymphoma [127, 130] and Jurkat T cells [131]. In an other hand, in lymphoma cells, NO is described as a mediator of Fas induced apoptosis [132]. In murine macrophages RAW 264.7, NO generated from NO donor [133] or by iNOS [134] provokes apoptosis in response to p53 accumulation. This effect is inhibited by Bcl-2 [135] and shows an increasing expression of Bax. It is suggested by Briine et al, [136] that NO-mediated apoptosis in macrophages is entirely controlled by the mitochondrial pathway with the implication of p53 accumulation and cytochrome c release. More downstream, NO activates caspase 3 and provokes PARP clivage [137]. Those results confirmed the first experiments on peritoneal macrophages [138, 139]. NO induces apoptosis of murine splenic T lymphocytes [140], and is a mediator of Fas induced apoptosis in hematopoietic cells isolated from bone marrow [141]. The origin of this effect could be a decrease of mitochondrial membrane potential and the generation of active oxygen derivatives [142]. In endothelial cells, NO inhibits apoptosis by tyrosine nitrosylation [143, 144], particularly of caspase-3, which is redox state independent [145]. Weak NO production can induce or protect from apoptosis, this phenomenom is very dependent of target cell. Roughly carcinoma cells and normal macrophages or lymphocytes would be NO sensitive whereas leukemic cells and endothelial cells would be NO resistant. NO induced apoptosis could involve p53 accumulation, mitochondrial dysfunctions and cytochrome c release, with or without caspase activation and Bcl-2 inhibition. The NO protective effect could be due to inhibition of caspase activity by nitrosylation with or without Bcl-2 activation. NO and cell differentiation : NO induces the differentiation of promyelocytes in monocytes, talking about established culture cells (HL-60, U937, THP-1) or freshly leukemic isolated cells. Cell differentiation is shown by the inhibition of cell proliferation and cell adhesion, induction of marker expression (CD 14) and cytokines production (IL-lp and TNFa). This effect is due to NO donors and NO gaz [146], due to iNOS induction by CD23 ligation [112]

927 or by trans retinoic acid and D3 vitamin [147]. NO induced HL-60 cells differentiation is mediated by cGMP; on the other hand, NO induced TNFa production in U937 cells is mediated by a decrease of cAMP concentration [110]. NGF inhibits proliferation and induces differentiation in PC 12 cells (isolated from a pheochromocytoma) and SH-SY5Y cells (isolated from a neuroblastoma [109]. The signal brought by NGF leads to iNOS activation, then to NO-mediated p21 activation [148] following two pathways : NGF - iNOS - p53 - p21^^^ and/or NGF - iNOS - p 2 1 ^ ^ ^ [109]. Neuroblastoma cell differentiation is induced by TNFa via iNOS expression [149]. We can suppose that spontaneous differentiation (polarisation, tight junctions, brush borders, and specific hydrolases) of human colon cancer cells Caco-2 is also due to iNOS. A high NO production is observed in the proliferative phase and disappeared during cell differentiation [150]. Cell resistance to NO In cancer cells, iNOS gene activation by IFNy and TNFa goes with coactivation of 1) y-GCS gene which product is the limiting enzyme of glutathion synthesis, 2) manganese superoxide dismutase gene which product reduces 02-' and so reduces ONOO" formation [151]. As well, iNOS and cyclooxygenase-2 (COX-2) genes are co-activated in macrophages (by LPS, [152]) and in astrocytes (by IL-ip, [153]). COX-2 catalyses prostaglandin E2 synthesis an inhibitor of/W05'transcription. In an other hand, NO inhibits iNOS expression. When NO provokes the p53 accumulation, this one could in return inhibit iNOS gene transcription [154]. Repression is observed in DLD-1 cells (human colon carcinoma) or Calu 6 cells (human pulmonary carcinoma) which express wild type p53. High NO concentration leads to p53 nitrosylation and inhibits the repression [114]. The wild type but not the mutated or the nitrosylated protein binds the iNOS gene promoter. Moreover, exogeneous NO produced by NO donors can inhibit N F K B activity in whole cells and in acellular preparation [155-157]. Those results have been confirmed in endothelial cells in which NO produced by eNOS inhibits N F K B activity and iNOS gene transcription [158]. The inhibition can be explain in parts by N F K B sub-unit p50 nitrosylation [155], but also by its cytoplasmic inhibitor protein iKBa stabilisation (IKB proteins functionally retain NF-

928

kB in the cytoplasm and render it inactive) and by the stimulation of its transcription [157]. Furthermore, NO produced by NO donors induces heat shock protein synthesis (hsp 70) in HEP G2 cells isolated from a human hepatoma. In one hand, hsp 70 inhibits iNOS transcription by reducing NFKB translocation, in an other hand, it prevents ATP depletion due to NO [159]. Consequently, the same signal on the same cell open two pathways: one leads to a cytotoxicity mediated by NO, the other protects from this cytotoxicity. The protective reactions are coactivated by the signals activating iNOS gene or are triggered of by NO itself; they consist either in inhibiting iNOS gene expression or toxic effects of NO. In sum, NO can lead to cytostase, cell apoptosis or cell differentiation, but it can also prolong cellular life [130, 160, 161]. The difference between these pathways seems to be the amount of NO implicated [161, 162], the N0/02-' ratio [163, 164], i.e. the oxidation-reduction (redox) state of the cell and the induction of resitance mechanisms. The difference can also be the nature or the number of mutated genes. Indeed, the mutations transforming cancer cells concern the genes controlling cell proliferation and differentiation. Depending on cancer cells, the sequence and the frequency of mutated gene are different, thus the NO effects are different. For instance when a wild type p53 protein is nitrosylated, its fixation on DNA is inhibited, while NO is without effects on mutated p53 that has already loose its capacity to bind DNA. Conclusion

The data presented here show the complexity of the role of NO in the antitumor immune responses of mammalian cells. All these date are from in vitro experiments. In vivo the complexity is higher than in vitro and few data are available. Using the model of colon cancer in rats, Jeannin et al [165] investigated in vivo the production of NO during tumor growth and tumor regression induced by lipid A. Results show opposite effects of NO on tumor growth depending on the cells producing NO. Without treatment, tumor cells did not not produce NO but activated spleen macrophages that produced NO and inhibited spleen T lymphocytes proliferation, thereby contributing to promotion of tumor growth [166].

929 On the contrary, tumor cells produce NO in response to lipid A [167, 168], leading to tumor cells apoptosis and tumor regression [169]. In the same time the treatment induced desensitization of macrophages without further NO production, via alteration of N F K B binding [170]. Therefore, the role of NO in the antitumor immune response, comprises both regulatory and effector functions at the intra- or intercellular level. The first functions includes inhibition of lymphocyte proliferation or participation in different transduction pathways. The second functions include pro- or anti-tumoral effects and NO mediated cell toxicity or cell resistance to apoptosis. NO IN PLANTS Studies conducted over more than one decade on different plant species have highlighted the ability of plants to produce NO endogenously. First investigations into NO sources in plants demonstrated that NO could be produced from NO2' or NO2 either via non-enzymatic pathways or by the activity of a constitutive NAD(P)H nitrate reductase (NR). These reducing capacities are, for example, illustrated by the ability of ginko (Ginko biloba L.) leaves to liberate up to 70% of absorbed NO2 as NO [171]. In addition, recent studies have shown that plants could also produce NO from L-arginine in a reaction catalysed by a mammalian-type NOS. This discovery stimulated enormous interest in the plant NO area. First analysis of the role of NO in plants shows it functions as a critical effector in plant growth and development. Moreover, NO was reported to counteract cytotoxic processes mediated by reactive oxygen species (ROS) in plant tissues. Furthermore, NO has been shown to play a key role in signaling defense reponses in several plant species when challenged by pathogens. In this context, plants use NO as a signaling molecule via pathways remarkably similar to those found in mammals. This data highlighted previous observations indicating that several individual components of host-pathogen interactions are shared between plants and animals. The purpose of this chapter is to provide an overview of NO synthesis as well as the modes of action of NO in plants. In particular, the central role of NO in plant disease resistance and its interconnection with the plant defense regulators salicylic acid (SA) and ethylene will be discussed.

930 NO synthesis in plants Non-enzymatic sources of NO

Formation of NOx (NO and NO2) gases from plants via non-enzymatic routes has been known to occur for some time. First reports indicated that plant leaves were able to accumulate NO2' and subsequently produce low levels of NOx through a non-enzymatic pathway [172-174]. For example, it was shown that under dark and anaerobic conditions, intact leaves of soybean {Glycine max L.) produced NOx through non-enzymatic reactions between plant metabolites and accumulated NO2' [174]. Moreover, NO could be the result of nitrous acid decomposition. With regard to the low level of NOx produced, it was proposed that non-enzymatic NOx formation probably occured at a lower rate compared to enzymatic processes. It was later reported that non-enzymatic NO formation could also occur through light-mediated reduction of NO2 by carotenoids [175]. This process may prevent damage in plants exposed to the toxic effects of NO2. Thus, although insufficiently characterized, non-enzymatic NO formation pathways could play a significant role in plant detoxication processes. Nitrate reductase-dependent NO production

In plants, nitrogen is essentially absorbed as NO3' which is reduced to ammonium by NR in the cytosol [176]. The first evidence for the capacity of NR to convert NO2" to NO came from the isolation of the soybean mutant nrl whose leaflets lacked both constitutive NR activity and the ability to form NOx [177, 178]. It was later shown that two forms of constitutive NR were present in wild-type soybean leaflets but missing in the soybean nrl mutant [179, 180]. These two constitutive NRs, designated CiNR and C2NR have the same pH optimum of 6.5 but differ in their A^m for NOs" (5 and 0.19 mM, respectively) and their electron donor preferences (NAD(P)H and NADH, respectively) [179]. In addition, when grown on nitrate, wild-type soybean expresses an inducible NR enzyme with a pH optimum of 7.5, a low K^n for nitrate (0.13 mM) and a preference for NADH as an electron donor [179, 181]. ciNR involvement in NOx production was supported by the characterization of two additional soybean mutants whose leaflets lack the C2NR enzyme but contain the

931 ciNR enzyme and the capacity to form NOx [182]. Moreover, the inability to separate the soybean leaflets NOx production activity from the ciNR activity provided further evidence that the soybean CjNR was the enzyme responsible for NOx formation [183]. Interestingly, the ciNR enzyme has a higher affinity for NO2" (Km of 0.49 mM) than for NOs" (5 mM). Furthermore, it was shown that, in addition to NAD(P)H, ciNR was able to use reduced FMN (FMNH2) and reduced methyl viologen as electron donors to convert NO2" to NOx. The ability of soybean NR to produce NO was supported by the work of Caro and Puntarulo [184] who, by EPR spin trapping analysis, measured a close association between the content of NO and NR activity in homogenates of soybean embryonic axes excised from seeds imbibed with increasing concentrations of nitrate. The NO generation was reduced by the NR inhibitor cyanide, thus confirming NR involvement in NO formation. It was believed until recently that NOx production by NR was limited to the Phaseolus tribe of the family leguminosea including soybean or winged bean [183, 185]. However, it has been reported latterly that plant species other than leguminosea may potentially produce NO in a reaction catalysed by NR. Indeed, Yamasaki et al, [186] measured high rates of NO production after the addition of a purified maize NR in a reaction mixture containing NO2' and NADH. To a lesser extent, NO production was also observed when NO3' was used as a substrate in place of NO2". NO production from both sources was reduced by the NO quencher hemoglobin. Since in plants cells, NO2', NO3", NR and NADH are present in the same compartment (i.e. in the cytosol), the authors suggested that NR could be rapidly mobilized to produce NO in response to abiotic and biotic stress. Thus, research on the roles of NO in plants should be also directed towards exploring the features of plant NR. NOS-dependent NO production in plant cells

Over the past 5 years, diverse lines of evidence have converged to suggest the presence of a mammalian-type NOS in plants. Evidence for the presence of NOS activity was first described in the leguminous plant Lupinus Albus [187]. A L-arginine-dependent L-citruUine production, possibly induced by Rhizobium lipopolysaccharides, was measured in roots and nodules using the [^'^C]arginine/citrulline procedure. L-citrulline production was inhibited by the mammalian NOS antagonist N(G)-

932

monomethyl-L-arginine (L-NMMA), suggesting that NOS could be also operative in plants. Similar approaches allowed the detection of NOS activities in tobacco, in pea {Pisum sativum L.) and in the leguminous plant Mucuna hassjoo [188-190]. In parallel, using a NO-specific probe, Leshem and Haramaty [191] measured a NO production sensitive to NOS inhibitors in pea plants which were severed from their roots. Later, Caro and Puntarula [184] showed, by EPR spin trapping analysis, a NOSdependent NO formation in homogenates of soybean embryonic axes excised from seeds. Here too, NO production was partially blocked by competitive substrates of mammalian NOS such as N(G)-nitro-L-arginine and L-NMMA. The specificity of mammalian NOS inhibitors in plants was recently questioned by Allan and Fluhr [192]. The authors demonstrated that application of L-arginine to tobacco cells resulted in a large production of NO and its ROS derivatives. This production was specific for the arginine L-isomer and was blocked by L-NMMA. However, the NO/ROS burst was similarly induced by other amines including L-histidine, L-citrulline, L-glutamine, canavanine, and the polyamines spermidine, spermine and putrescine. Since amines can induce ROS production by acting as substrates for amines oxidases, it was suggested that the L-arginine dependent NO/ROS burst could result from an amine oxidase activity sensitive to mammalian NOS inhibitors such as L-NMMA. Evidence for the existence of NOS in plants was also supported by immunoblot analysis. For example, antibodies raised against mouse macrophage NOS and rabbit brain NOS revealed a protein band of 166 kDa in the soluble fraction of root tips and young leaves of maize [193]. Immunofluorescence microscopy analysis showed that the subcellular localization of the putative maize NOS was dependent on the phase of cell growth. Recently, using antibodies raised against the C-terminal region of murine iNOS, Barroso et al. [190] detected the presence of an immunoreactive protein of 130 kDa in the matrix of peroxisomes as well as in chloroplasts from leaves of pea plants. Similar immunoblot analysis with antibodies made against mammalian NOS allowed the detection of a putative NOS in tobacco, wheat germ and pea embryonic tissue [188, 194-196]. Studies in the plant disease resistance area gave further support to the presence of a mammalian-type NOS in plants. In resistance to fungal, bacterial or viral infections, plants deploy specific mechanisms. These mechanisms are largely based on the specific recognition of pathogen

933 molecules through plant putative receptors. In some plant pathogen interactions, plant resistance is governed genetically by direct or indirect recognition between the product of a plant disease resistance gene and pathogen-encoded avirulence molecules. These specific interactions trigger a battery of cellular responses which may lead to the death of the pathogen [197-200]. In addition, plant responses to infection often include localized cell death (termed HR for Hypersensitive Response) at the site of attempted infection [201]. This cell death is believed to limit the advance of the pathogen. Dumer et al. [202] demonstrated that tobacco mosaic virus (TMV) inoculation of resistant tobacco Xanthi nc (NN), which contains the A^ resitance gene confering resistance to TMV, results in elevated NOS activity partially sensitive to NOS inhibitors including LNMMA and diphenylene iodonium (DPI). This increase in NOS activity was not observed in the TMV-infected susceptible cultivar Xanthi (nn) which does not carry the N resistance gene. Similarly, Delledone et al. [203] reported that inoculation of soybean cell suspensions with the avirulent pathogen Pseudomonas syringea pv. glycinea stimulated NOSdependent NO production. NOS activity was calcium dependent, partially sensitive to NOS inhibitors and detected in the cytosolic fraction of treated soybean cells. Moreover, Delledone and coworkers [203] provided the first evidence that inhibition of NO production might result in expanded pathogen growth. Indeed, co-infiltration of Arabidopsis cultivar Col-0 (which carries the RPMl resistance gene) with Pseudomonas syringae pv. maculicola (containing the corresponding avrRpml avirulence gene) and mammalian-type NOS inhibitors, compromised the hypersensitive response and promoted disease symptoms. A recent study completes these findings by demonstrating that NOS activity is essential for the initiation of the hypersensitive response in tobacco infected by the bacterial pathogen Ralstonia solanacearum [188]. Using antibodies raised against mammalian NOS, the authors detected a tobacco protein of 55 kDa whose intensity increased during Ralstonia solanacearum infection. Taken together, these results suggest that plants produce NO, during the resistance response, using a mammalian-type NOS whose gene could be induced in response to pathogens. Thus, NO may play a central role in the plant response to infection. Despite the above findings which suggest the existence of a plant NOS and its involvement in some physiological processes, the mechanisms for generation of NO in plants are under investigation [18, 204]. Indeed, as

934

evidence for the generation of NO is circumstantial thus far, the identification of the plant NOS is crucial. To date, neither the gene nor the protein responsible for NOS activity has been isolated from plants, although a plant protein homologue of a 10 kDa protein that inhibits specifically an animal NOS has been characterized [205]. Furthermore, there is no evidence that the plant proteins which react with antibodies raised against mammalian NOS correspond to plant NOS [188, 190, 193196]. Indeed, it is surprising that the putative NOS immunodetected in maize, pea and tobacco have estimated molecular masses of 166, 130 and 55 kDa, respectively. Moreover, plant NOS-like activity is only partially sensitive to mammalian NOS inhibitors [184, 187, 202, 203]. These observations suggest that plant and animal NOS could differ in their structure and that animal NOS inhibitors could be less specific and efficient in plants. In this regard, the finding that the NOS inhibitor LNMMA could affect amine oxidases in plants is disconcerting [192]. Similarly, one must be cautious of interpretations based on use of DPI since DPI is widely reported to be a potent inhibitor of NADPH oxidase in plant cells and mammalian phagocyte cells [204]. Thus, there is actually no universal agreement on the source of plant NO and on the enzymes involved in its generation: NR, NOS and maybe amine oxidases could be a source to produce NO in plants. Considering the recent interest of the scientific community in NO in plants, no doubt these questions will be resolved in due course. Roles of NO in plants NO as plant growth regulator

Several lines of evidence have suggested that NO might participate in plant growth regulation. For example, direct exposure of pea to low concentrations (up to 5.10'^ M) of chemically generated NO gas induced a significant leaf expansion [191, 206]. It was proposed that at these concentrations, NO may trigger an increase in cell wall matrice plasticity and in plasma membrane fluidity; both processes being required for cell expansion. As suggested by Leshem [206], this mechanism could provide a plant analogy to benign vasodilation triggered by NO in mammals. At higher concentrations, an opposite effect was observed, NO showing an inhibitory role in leaf expansion. NO-induced inhibition of leaf grovv1;h

935 could be due to changes in membrane permeability and to the disruption of cellular key enzymes. In this respect, chloroplast may represent a major NO target. Indeed, it was found that chlorophyll fluorescence in the 690 nm region of pea guard cell chloroplasts largely increased when exposed to high NO concentrations [207]. The high chlorophyll fluorescence could be the result of impairement of photosynthetic proteins, especially of Fe-S cluster containing proteins which are preferential NO targets in animal and also plant cells [11, 208]. Furthermore, NO could affect the microfluidity of the lipid bilayer matrix whose integrity is essential for the efficiency of photosynthetic proteins. This latter possibility is confirmed by the finding that NO may attack the n component of the C=C double bonds in chloroplast membrane galactolipids [207]. Besides its involvement in leaf expansion, NO might be involved in the control of root growth [193]. Indeed, immunofluorescence detection of the putative NOS in maize root tips showed that in the division zone, the enzyme was present in the cytoplasm whereas the enzyme was located in the nucleus in cells in the elongation zone. This result suggests that NO may be involved in specific pathways controling particular phases of root development. NO involvement in plant growth should also be discussed taking into account probable inter-relations with ethylene. Ethylene is a plant hormone known to be involved in the regulation of developmental processes and stress responses in higher plants including seed germination, fruit ripening, cell elongation, organ senescence, leaf abscission, cell death and pathogen responses [209-211]. Ethylene ability to modulate many plant processes suggests the presence of complex controls regulating its production. Interestingly, NO and ethylene are metabolically interlinked in plants and pharmacological approaches suggest that ethylene production is regulated by NO [191, 206]. Indeed, NO may inhibit ethylene biosynthesis by oxidizing ascorbate or Fe^^ which are required for the activity of 1-aminocyclopropane-l-carboxylic acid oxydase (ACC oxydase), the enzyme responsible for ethylene formation [206, 212]. Thus, NO-mediated inhibition of ACC oxidase might be of biological significance during plant growth and could play a role in the temporal and spatial regulation of ethylene biosynthesis.

936 NO participation in the antioxidant cellular systems

Depending on the physiological circumstances, NO is described as either protective or toxic in animal cells. To substantiate the hypothesis that NO could act as an antioxidant in plants, Caro and Puntarulo [184] studied the effects of exogenously produced NO on the rate of 02*" and HOgeneration by microsomes isolated from soybean axes. In this analytical system, O2" is thought to be generated by reactions catalysed by cytochrome P450. It was demonstrated that NO significantly decreased the O2' generation rate by microsomes but did not alter the microsomal rate of HO- generation. This reduction of 02*" production was shown to be caused by a NO-mediated inhibition of the microsomal cytochrome P450 activity [184]. Since endogenous NO is produced in soybean embryonic axes as shown by EPR studies (see paragraph 3.1.3), the authors conclude from their experiments that the interaction between NO and cytochrome P450 could be a factor of relevance in the control of the oxidative cellular balance at the early stages of plant development. NO participation in the antioxidant cellular system of plants was confirmed by a series of studies conducted by Lamattina and coworkers [213-216]. They analysed NO effects on some oxidative stresses caused in potato {Solarium tuberosum L.) by the methylviologen herbicides diquat and paraquat or by the potato late blight agent Phytophthora infestans, Diquat and paraquat cause ROS generation (H2O2, 02-" and H0-) by extracting electrons from photosynthetic electron transport [216]. Potato treatment by nanomolar concentrations of NO (via the NO donor sodium nitroprusside) reduced or prevented chlorophyll loss, cell ion leakage, DNA fragmentation and apoptotic-like cell death normally induced by the herbicides or Phytophthora infestans. Some of these NO-mediated effects were suppressed by NO scavengers, thus confirming NO involvement in the protection against the ROS. In reference to animal systems, NO antioxidant properties in plants might be explained by NO's ability to combine to ROS through its umpaired electron [184, 215, 217]. For example, the reaction between NO and ROS could prevent membrane damage caused by ROS-mediated lipid peroxidation. NO might also activate enzyme protecting against oxidative stress and inhibit ROS generating enzymes such as cytochrome P450 [184, 215]. Moreover, in order to respond to oxidative stresses, plant cells may have possibly evolved NO-containing compounds which may be mobilized to target NO delivery to ROS generators. This latter hypothesis is supported by the

937 finding that exposure of young or mature soybean root nodules to oxidant stress induces changes in the EPR spectra of a NO-leghemoglobin complex [218]. Thus, as widely described in animals, NO may exert a protective effect in plants by acting as an anti-oxidant. Nitric oxide as a signal in plant defense response

A major advance in the understanding of NO signaling functions in plants comes from the plant disease resistance area. The signal cascade leading to plant defense is induced through recognition of a pathogen avirulence gene product by the corresponding disease resistance gene, or by an elicitor of plant defense responses recognized by a specific receptor (see above) [197, 200, 219-221]. Recognition of either type of signal initiates early signaling events including protein phosphorylation /dephosphorylation, ions fluxes (Ca^^ K^ Cr, H^) across the plasma membrane, plasma membrane depolarization, oxidative burst producing ROS, MAP kinase activation and extra/intra-cellular pH modifications [200, 220, 222]. The elicitor signals are also often amplified through the generation of secondary molecules such as SA, ethylene and jasmonate (JA) [200, 223]. Subsequent transcriptional and/or post-translational activation of transcription factors leads to the induction of plant defense genes. Plant defense genes include genes encoding pathogenesis related (PR) proteins characterized by antimicrobial or insecticidal activities, enzymes involved in the generation of antimicrobial compounds called phytoalexins, enzymes protecting against oxidative stress, enzymes involved in lignification, and others [223, 224]. Accumulating bodies of evidence suggest similarities in the defense mechanisms adopted by plants and animals [225]. In particular, the tobacco disease resistance gene A^ and the flax disease resistance gene L6 products show significant similarity to the IL-1 receptor and Toll, which are critical regulators of disease resistance in mammals and Drosophilia, respectively [197]. Moreover, evidence supports that elicitor-induced Or' synthesis involves, at least in part, a plasma membrane NADPH-oxidase which resembles the NADPH oxidase of mammalian phagocytes [226, 227]. In similar studies, Durner et al. [202] and Delledonne et al, [203] established NO as a key signaling molecule in plant defense responses. The authors reported that incubation of tobacco and soybean cell suspensions with NO donors triggers the accumulation of transcripts

938 encoding phenylalanine ammonia-lyase (PAL), the first enzyme of the phenylpropanoid pathway. In plants, the phenylpropanoid pathway is rapidly induced after pathogen infection and is involved in the synthesis of several defense-related compounds, including phytoalexins and SA [228]. Consistent with NO acting as a signal in PAL gene activation, NO treatment of tobacco leaves and potato tuber tissues was shown to trigger dramatic increases in endogenous SA and phytoalexins, respectively [202, 229]. Moreover, NO was found to activate a common marker of defense gene transcription called PR-l through a SA-dependent pathway [202]. Interestingly, NO-activated PAL expression seems to be mediated by a guanylate cyclase-like protein and acts through cGMP as demonstrated in a set of experiments combining a cGMP-radioimmunoassay and pharmacological approaches [202]. These findings confirm a previous study from Pfeiffer et aL [230] who demonstrated that NO stimulates cGMP formation in spruce needles. In mammals, cGMP serves as a second messenger for NO signaling and the occurrence of cGMP in response to various stimuli has been unambiguously demonstrated in plants although the gene encoding guanylate cyclase has not yet been cloned [231, 232]. Therefore, it seems that there may exist a NOdependent regulation of guanylate cyclase-like activity in plants, as has been described for vertebrates. However, molecular targets of cGMP in plants are poorly characterized. Besides putative candidates, identified on a molecular basis and including inward-rectifying K^-channels and two other putative cyclic-nucleotide-regulated ion channels [233-235], only one cGMP/cAMP-dependent protein has been functionaly characterized in plants [236]. This protein belongs to a particular family of cation channels (cyclic-nucleotide-gated non-selective cation channels) which conduct Na"^, K^ and Ca^"*" currents under the control of cGMP and cAMP in animal cells [236,237]. This latter finding opens the possibility that cGMP may mediate NO signals in plants through the activation of Ca^"^ channels. In certain animal cells, the cGMP generated by NO was reported to stimulate the production of cyclic ADP-ribose (cADPR), a potent Ca^^-mobilizing second messenger (Fig. (6)) [238-240]. This production occurs through the cGMP-dependent activation of a protein kinase which phosphorylates ADP-ribosyl cyclases, the enzyme catalysing the conversion of p-NAD"*^ to cADPR. Then, cADPR directly activates the endoplasmic reticulum

939

Plasma membrane cytoso

J Ca'

^

fc^ Guanylate ^ *^yc'«se

^^ ^,. ^ • f ^ ,—^ P kinase

cellular targets

ADPribosyl 'Ji7^1il cyclase

M

\^ ^ cADPR

Ca^ M

^^^^

ER

RVR —

Fig. (6). NO-mediated increase of free cytoplasmic Ca^"^ concentration in animal cells. NOdependent Ca^^ release from the endoplasmic reticulum (ER) may be mediated by two pathways: a cGMP/cADPR-dependent pathway and/or direct S-nitrosylation of the ryanodine receptor (RYR) by NO (see the text for details). CICR: Ca^'^-induced Ca^"^ release. .2+ (RE) membrane Ca^"^ chamiel RYR which consequently promote a Ca^^ release from the RE. This release is itself potentiated by elevated Ca^"^ cytoplasmic concentrations, a process termed Ca^'^-induced Ca^^ release (CICR) [241]. Alternatively, NO might directly activate RYR by poly-Snitrosylation [33]. In plants, patch clamp studies and Ca^"*" flux measurements have demonstrated cADPR-elicited Ca^"^ release from internal stores [242-245]. The Ca release shows the hallmark characteristics of mediation by ryanodine receptors. Moreover, using a cADPR-microinjection system, Wu et al, [246] and Leckie et al. [247] firmly established a physiological role for cADPR in the signal transduction of abscisic acid, a plant hormone which controls plant response to an array of environmental stresses such as drought, salinity or cold. Based on these studies, a NO-regulated cADPR involvement in plant defense was tentavely assigned by Durner et al. [202]. The authors reported cADPR ability to induce PAL and PR-l gene expression in tobacco leaf discs through a Ca^"^ release mechanism sensitive to antagonists of ryanodine receptors. Interestingly, expression of both genes

940 was amplified when leaf discs were simultaneously treated by cADPR and cGMP. Moreover cADPR appears to act through a SA-dependent pathway for PR-1 activation but a SA-independent pathway for PAL induction (Fig. (8)). However, although informative, these data do not establish whether NO and/or cGMP operate through cADPR to induce PAL and PR-1 genes. Indeed, NO may directly activate Ca^"*" release via Snitrosylation of Ca channels as reported in animal cells [33]. In addition, cGMP could mediate Ca^"^ release in a cADPR-independent pathway by activating cyclic-nucleotide-gated Ca^"^ channels sensitive to antagonists of ryanodine receptors [236, 248]. Bioassay experiments allowing the measurement of cADPR amounts in response to NO and/or cGMP should provide the first answers to these questions. The study of NO signaling effects during plant-pathogen interaction has been extended to the analysis of its putative interaction with aconitase [208]. Aconitase is a Fe-S cluster [4Fe-4S]-containing enzyme which catalyses the reversible isomerization of citrate to isocitrate via cisaconitate. The catalytic activity depends on the presence of the [4Fe-4S] cluster. The enzyme has two isoforms (encoded by two different genes), one located in the cytosol and the other in the mitochondria where it is implicated in the Krebs cycle [249]. In animal cells deprived of iron, a complete dissassembly of the [4Fe-4S] cluster is observed [250]. The resulting apoprotein, named IRP (Iron Regulatory Protein), modulates the translation of mRNA encoding proteins that function in the maintenance of iron homeostasis such as ferritin and the transferrin receptor. More precisely, IRP binds with high affinity to mRNA stem-loops called IRE (Iron Responsive Element) located in either the 5' or 3' untranslated regions of mRNAs [250]. UV-cross-linking and site-directed mutagenesis approaches demonstrated that a sequence of 10 amino acids (residues 121 to 130) and three arginine residues (residues 536, 541 and 780) in human IRP are crucial in IRE binding [251-253]. Extensive studies have shown that binding of IRP in the 5'-untranslated region of ferritin mRNA prevents the mRNA translation, whereas binding of IRP in the 3' untranslated region of the transferrin receptor mRNA triggers an increased stability of the mRNA [250]. Thus, in cells requiring iron, the levels of intracellular ferritin fall while the levels of transferrin receptor rise. Interestingly, a number of observations have shown that NO, ONOO" and also H2O2 induce destruction of the [4Fe-4S] cluster of cytosolic aconitase and enhance IRE binding activity of IRP [254-256]. The physiological relevance of NO-mediated IRP activation remains undefined. In plants,

941 genes encoding cytosolic aconitase have been cloned from Arabidopsis [257], pumpkin [258] and tobacco [208]. The encoded proteins share 85 to 90% identity, and comparison with human IRP revealed 60% identity. Interestingly, plant cytosolic aconitases contain 9 of the 10 amino acid residues (residues 125 to 134 in tobacco cytosolic aconitase) forming the RNA-binding site which, in the human IRP, is believed to directly interact with IREs [208, 257, 259]. Moreover, the three arginine residues involved in IRE binding by IRP are also conserved [208, 252]. Given these similarities, Navarre et al [208] investigated the effects of NO and ONOO" on tobacco aconitase and found that cytosolic and mitochondrial tobacco aconitases, like their animal counterparts, are inhibited by micromolar concentrations of NO and ONOO' in vitro. Moreover, H2O2 was found to dramatically reduce aconitase activity with concentrations as low as 25 |aM. Similar results were previously reported by Verniquet et al [260] who demonstrated that the purified mitochondrial aconitase of potato {Solarium tuberosum L.) was rapidly inactivated by low concentrations of H2O2. Taken together, these results suggest that conditions leading to elevated levels of NO and/or ROS (e.g during plantpathogen interactions) will inactivate aconitase in tobacco. In addition to a decrease in energy production, this inhibition should lead to the accumulation of citrate. In plants, citrate is an inducer of the gene encoding alternative oxidase (a quinol oxidase), the terminal enzyme in the mitochondrial alternative respiratory pathway [261]. Alternative oxidase has been reported to play a putative role in the development of resistance to TMV [262] and may be involved in ROS tolerance in higher plants [263, 264], as suggested by experiments showing that cultured cells from transgenic tobacco with altered levels of alternative oxidase produce significantly higher levels of ROS compared with wild-type cells [264]. Beside disturbing citrate metabolism, NO-mediated cytosolic aconitase inhibition may convert the enzyme to the mRNA binding protein through the disassembly of its Fe-S cluster. Navarre et al. [208] suggested that, in reference to animal systems, this process could contribute to an increase in intracellular free iron levels, which, in turn, may mediate the Fenton reaction by reacting with H2O2 to yield HO- (Fig. (7)). This highly reactive ROS (and other ROS) in addition with elevated iron in cells could contribute to HR cell death by creating a killing

942 Citrate Aconitas pathogen

^

Isocitrate

NO (ONOO)

IRP

Target mRNAs

i

H20

^

Fenton I reactio

T

Killing environment

"

HO

Fig. (7). Hypothetical consequences of NO-mediated inhibition of plant cytosolic aconitase [208]. The interaction between NO and cytosolic aconitase triggers a cluster dissassembly and subsequently the inhibition of the catalytic activity of the enzyme. By analogy to mammalian studies, the resulting apoprotein may act as iron regulatory protein (IRP) and may modulate the translation of mRNA encoding proteins involved in the cellular iron homeostasis. The elevated free iron concentration promotes the Fenton reaction leading to hydroxyl radical (HO) production. Both HO- and high concentrations of iron create a killing environment for host and pathogen.

environment for both host and pathogen. To verify such a mechanism, the hypothesis that plant cytosoUc aconitase serves as a mRNA-binding regulator of iron homeostasis needs to be proved. Preliminary attempts to detect IRP activity in plants has not succeeded [265] and the regulation of iron metabolism may be different in plants and animals [266]. However, this possibility needs to be further analysed and the identification of aconitase mRNA targets could be a major advance in understanding NO signaling functions in plants. NO signaling activities in plant defense should also be discussed in terms of its interconnection with ROS. Several mechanisms for ROS generation exist in plants, including plasma membrane NADPH and NADH oxidases, amine and oxalate oxidases or apoplastic peroxidases

943 [204]. Analogues of the mammalian gp91phox have been cloned from Arabidopsis [267, 268], rice [269] and parsley [204]. Although arguments suggest that plant gp91phox is a target for kinase action linked to Ca^^ fluxes [270], the exact functions of the protein and the mechanisms of its activation remain to be resolved [204]. The oxidative burst initiated by pathogen attack results in the synthesis of 02-' which can spontaneously or enzymatically dismutate to H2O2. A number of possible roles for O2" and H2O2 have been described: direct antimicrobial activity, cross linking of proteins and phenolic compounds to the cell wall, induction of defense genes through intra- and inter-cellular signaling activity, and initiation of host cell death in the HR [271]. Intriguingly, in situ generation of H2O2 or equivalent generation of 02*" in plant cell suspensions closely mimics the oxidative burst induced by avirulent pathogens, but triggers only a weak cell-death response [272]. Interestingly, it was recently demonstrated that NO potentiates the induction of hypersensitive cell death in soybean suspension cells by the reactive oxygen intermediates H2O2 and 02-' [203]. Thus, as observed in the vertebrate native immune system, NO and ROS seem to synergistically promote cell death. This opens the possibility that NO and 02-" produced in plants in response to pathogens, react to yield ONOO". Peroxynitrite could act as a mediator of toxicity and may cause cellular damage by oxidation and nitration reactions. Considering the observation that the cell death associated with the HR in some plantpathogen systems has morphological similarities to animal apoptosis [273], NO and/or ONOO" could promote apoptosis through the activation or inhibition of specific targets. In this regard, there is some evidence for the participation, during the HR, of caspase-like activities known to play a key role in NO-mediated apoptosis in animals [274, 275]. Does NO mediate the activation of caspace-like protease associated with apoptosis in plants ? This possibility is intriguing and merits further investigation. In contrast, the striking complementarity between NO and ROS signal functions was also confirmed by the observation that NO and ROS independently activated complementary gene sets encoding for enzymes involved in cellular protection [203]. Changes in the cellular redox status caused by NO and ROS production and their relative combination could promote transcriptional activation of defense genes. In mammals, increasing evidence is accumulating that NO/ONOO" preferentially alters

944

•

Redox signaling

Ca -dependent

transcription factors

I'

defense genes {t.gPR-2)

Fig. (8). Hypothetical NO-mediated pathway in plant cells challenged by a pathogen. NO may interact with multiple cellular targets including guanylate cyclase, Ca^^ channels and aconitase. Moreover, NO may combine to ROS to modulate the activity of cellular targets that are sensitive to changes in the cellular redox status. NO-activated guanylate cyclase produces cGMP which is a putative second messenger in plants. cGMP may activate Ca^"^ channels directly or through the Ca^"*^-mobilizing agent cADPR. The resulting increase in the cytoplasmic Ca^^ concentration leads to PAL gene expression and the activation of the corresponding enzyme which catalyses SA production. In turn, SA (in)directly modulates cellular targets, including NPRl and the MAP Kinase SIPK (the possible connection between SIPK and NPRl is unknown). These interactions lead to the activation of transcription factors including bZIP transcription factors and a putative NPKB-like protein which remains to identify. Activated transcription factors may then trigger expression of certain defense genes such as PR-J.

945 transcription factors that are sensitive to clianges in the cellular redox status. In particular, the well characterized transcription factors, N F - K B and AP-1 were shown to be (in)directly activated or inhibited by NO, depending on the cell type [24]. Interestingly, the promoter of the Arabidopsis thaliana PR-1 gene contains a consensus N F - K B binding site (at -610) and the as-lAik& cis element (at -640) which was shovm to be essential for SA (and H2O2) induction of gene expression [276, 277]. Although the gene encoding a NFKB-like protein has not been cloned in plants yet, an Arabidopsis thaliana gene showing some homology to the mammalian IKB was cloned recently [278, 279]. The corresponding protein, designated NPRl (also termed NIMl and SAIl), is an important component of the SA signal transduction pathway leading to PR genes expression [280]. In homology with the mammalian proteins, NPRl may interact with the plant N F - K B homolog, thereby inhibiting N F - K B mediated PR genes activation. In response to SA, NPRl could be degraded and N F - K B released to induce PR genes expression. The finding that the promoter oi PR-1 gene contains the a^-7-like cis element is also of great interest. Indeed, this element which was also found in the promoter of the cellular protectant plant gene encoding Glutathione Stransferase (GST), shares homologies with the electrophile-responsive element (EpRE) discovered in the promoters of several mammals' oxidative stress-induced genes [276]. Since EpRE has been shown to bind the transcription factor AP-1 in response to oxidative stress including NO, it is tempting to speculate that a plant homolog to the mammalian AP-1 protein could modulate PR-1 or GST gene expression in response to NO or ONOO". (25-7-like cis elements were shown to bind several plant transcription factors belonging to the family of basic leucine zipper protein (bZIP) transcription factors [276]. Interestingly, using NPRl as bait in a yeast two-hybrid screen, Zhang et al. [281] and Zhou et al. [282] identified bZIP transcription factors interacting strongly and specifically with NPRl. Zhang et al. [281] proposed that, upon SA induction, NPRl might be translocated to the nucleus and would activate bZIP transcription factors, thus leading to PR-1 expression. In this hypothesis, NPRl would act as a transcription co-activator, rather than a repressor. Although speculative, both scenarios provide mechanisms by which NO could activate the PR-1 gene through SA (Fig. (8)). The functional connections between NPRl and NO (ONOO') are not yet established but clearly merit further investigation. In this regard, it is likely that some of the putative

946 NO-modulated transcription factors may be regulated by MAPK, which in animal cells are known to phosphorylate transcription factors in the nucleus. In plants, MAP kinase cascades has been implicated in stress and hormone signal transduction [283], and a direct role of these kinases in the regulation of gene transcription has been suggested [284]. Studies have demonstrated that in tobacco, SA induces a rapid and transient activation of a 48 kDa MAPK designated SIPK (SA-Induced Protein Kinase) [285] which acts either upstream to, or independently of the oxidative burst [286]. SIPK interacts with a MAPKK recently characterized [287], and experimental data suggests that it could participate in the activation of a^-7-like cis elements binding activity [288]. Interestingly, it was recently demonstrated that NO activates SIPK in a SA-dependent manner [289]. By contrast, NO did not affect WIPK (Wounding-Induced Protein Kinase), another tobacco MAPK activated by wounding or by pathogen and pathogen-derived stimuli [290, 291]. Now, the challenge will consist to identify transcription factors phosphorylated by SIPK in response to NO. Additional results confirm that NO and SA act as conspirators. For the past ten years, biochemical, molecular and genetic approaches have been used to address the mechanisms of action of SA in plant defense responses. SA has been shown to play an important signaling function in disease resistance response in several plant species [200, 288, 292]. The finding that SA inhibits catalase and ascorbate peroxidase, two major H2O2 scavenging enzymes, led to the working hypothesis that one of SA's mechanism of action was to elevate H2O2 which then serves as a second messenger to activate defense responses [293, 294]. Supporting this model was the observation that the two synthetic inducers of the PR genes and plant resistance, namely 2,6-dichloroisonicotinic acid and benzothiadiazole, also inhibited catalase and ascorbate peroxidase [294296]. In contrast, several studies have questioned this model and suggested that SA may be acting downstream of H2O2, rather than the reverse [297, 298]. It was later demonstrated that SA-mediated inhibition of catalase and ascorbate peroxydase produces S A free radicals which, in turn, may directly promote lipid peroxidation [299, 300]. Whatever the scenario, concordant results strongly support a model whereby SA and ROS potentiate the HR and defense response [200]. The finding that NO acts through SA and in combination with ROS emphasizes the complexity of NO signaling functions in plants. How are these distinct signals interconnected? At least two scenarios can be envisaged. Firstly, SA can

947 mediate and /or amplify NO's effects by modulating the activity of NOregulated enzymes or the expression of the corresponding genes. Indeed, besides catalase, SA was reported to bind to and inhibit NO-sensitive iron-containing enzymes including aconitase [301] and ACC oxidase [302]. As discussed previously, ACC oxidase is a key enzyme in ethylene biosynthesis. ACC oxidase might be activated in response to pathogens since the ethylene level has been shown to increase during the HR [199]. S A-mediated inhibition of ACC oxidase may reduce ethylene production and senescence; two process known to be also reduced by NO. Thus, SA and NO could act synergistically to inhibit ethylene effects during the HR. This process seems to be controlled by the cellular context since ethylene was also shown to enhance the SA-induced expression of PR-1 in Arabidopsis [303]. Moreover, Sanz et al. [304] recently reported the cloning of a plant cyclooxygenase gene (nammed PIOX for pathogeninduced oxygenase) which is rapidly induced by SA and ROS in the plantpathogen context. In mammals, NO activates cyclooxygenase [305]. Consequently, the NO-activated enzyme mediates the conversion of polyinsaturated substrates to prostanoids which act as signals triggering many cellular processes including the immune response. In a manner similar to mammalian cyclooxygenase, NO-activated PIOX protein could be involved in the synthesis of lipid-derived signal molecules. Thus, according to this model, S A and/or H2O2 may induce expression of genes whose corresponding proteins will be modulated by NO. Secondly, based on mammalian studies, SA might counteract NO effects. Indeed, several reports suggest that the anti-inflammatory mechanism of SA or acetlysalicylate in cardiac diseases involves inhibition of iNOS transcription [306, 307]. In addition, SA was recently reported to inhibit NO-induced oxidative processes by acting as a NO scavenger [308]. In plants, SA activates the alternative oxidase gene and consequently, the capacity of alternative pathway respiration which is insensitive to NO [263, 288, 309]. Although speculative, these processes might compensate NO-mediated inhibition of the respiratory cytochrome c oxidase [263]. Thus, on one hand SA may mediate and/or potentiate the NO signal, and on the other hand, SA may contribute to cell integrity by inhibiting NO biosynthesis and containing some of its effects during the HR. Furthermore, SA could have both a negative and a positive effect on NO biosynthesis since it was also found to enhance NO production in soybean plants [310]. Such intricate mechanisms will critically influence the fate of the plant cells under pathological conditions. Moreover, the complex

948

interconnection between NO, SA and also ethylene might play an important role in the plant's ability to differentially activate distinct defense pathways according to the type of invader encountered [223, 311]. Indeed, recent advances in plant defense research have shown that not all aggressors trigger the same defense genes. In this respect, cross-talk between SA, ethylene (and also JA) may help the plant to prioritize the induction of a specific pathway over another. Studies support the evidence that SA-dependent pathway may confer resistance more effectively against a particular type of pathogen, whereas resistance triggered by ethylene (and JA) in a SA-independent pathway might be directed more against other types of pathogens. Each pathway might regulate the temporal expression and/or amplitude of the other pathway. By these strategies, the plant might fine-tune its defense according to the aggressor. NO's potency to activate SA production and inhibit ethylene production suggests that it could help cells to switch on the SA-dependent pathway after recognition of certain pathogens (Fig. (9)). Interestingly, using a genetic approach. Shah et al. [312] recently reported the characterization of an Arabidopsis thaliana gene named SSIl whose product may function as a switch modulating cross-talk between the SA- and ethylene-mediated defense signal transduction pathway. Cloning of the SSIl gene should provide important information concerning the NO hypothetical role in differentially regulating each pathway. One reaction of plants to pathogen infection is the activation of a longlasting, systemic resistance known as SAR (for systemic acquired resistance). This response leads to whole plant resistance to the inducing agent, as well as to a broad spectrum of other fungal, bacterial, and viral pathogens [313]. Development of SAR involves a mobile signal which is translocated from the infected tissues to the other parts of the plant. SA was considered as a possible mobile signal; however, this possibility is presently controversial [314]. In a recent paper, Alvarez et al. [201] discovered that establishment of SAR requires secondary oxidative bursts in discrete cells in distant tissues and inoculated leaves. These sytemic responses are induced by the primary oxidative burst which occurs at the pathogen penetration site. According to these observations, the primary and then the secondary oxidative burst might generate the mobile signal. The tight complementarity between NO and ROS suggest that NO or a derived metabolite could be the translocated signal. This hypothesis was recently discussed by Dumer and Klessig [18] who proposed that nitrosoglutathione (GSNO) could serve as a long distance SAR signal.

949

cthylenC'ClepenUent pathway

Fig. (9). Simplified model of the possible relationships between NO, SA, ROS and ethylene. NO and ROS trigger SA synthesis and SA might also enhance NO and ROS synthesis. SA, NO and ROS may act synergistically by similarly modulating the activity of common targets (for example, aconitase, ACC oxidase and cyclooxygenase). This amplification process favours the SAdependent pathway. By contrast, NO- and SA-mediated ACC oxidase inhibition prevent ethylene synthesis and consequently the ethylene pathway (SA-independent pathway). SA may also counteract some of the NO effects and may prevent NO deleterious damage (for example accumulation of NO inhibits electron transport through cytochrome c oxidase while SA activates the alternative oxidase [18]).

In mammals, GSNO might serve to facilitate NO transport in the blood and several studies have demonstrated that glutathione accumulates in plants exposed to oxidative stimuli [315]. Moreover, GSNO is a strong inducer of plant defense genes [202]. Thus, in response to pathogens, NO could bind to glutathione and act as a mobile signal that triggers SAR in distal plant cells through the activation of oxidative burst generators. The role of GSNO in activation of SAR remains to be rigorously established. Conclusion Although there is still an ongoing debate about the identity of the NO generators, it is now clear that NO is a key regulator of many plant processes. NO has been found to have a key role in both normal physiological processes (e.g plant growth) and pathological states. As reported in mammalian studies, NO displays anti-apoptotic or pro-

950 apoptotic effects in plants, depending on the cellular context and its production rates. Moreover, formation of peroxynitrite may represent a critical point in cells producing both NO and 02*", leading to either down regulation of the physiological effects of NO and Oi*", or to potentiation of their toxic effects by oxidation of cellular targets. Thus, the balance between NO and 02*' seems to be tightly regulated and may influence critical pathways leading to growth or death. Remarkably, first analysis of NO signaling activities in plants shows that NO triggers the induction of defense-related genes via pathways similar to those found in mammals. In particular, key proteins involved in animal NO signaling including guanylate cyclase, calcium channels, aconitase and MAPK appear to be modulated by NO in plants. This finding complements the list of evidence suggesting that portions of transduction mechanims are shared between plants and animals. Because some interpretations of NO functions relied on pharmacological approaches, significant work still needs to be done in identifying direct targets of NO. Moreover, genetic dissection of the NO signal transduction pathway using mutants compromised in NO synthesis and NO-mediated effects should provide a detailed picture of NO pathways at the molecular level. Such work will hopefully provide some understanding of how NO, SA and ethylene control and potentiate each other's activities. Elucidation of such interconnection should provide crucial information to develop new strategies to control plant processes such as seed germination or pathogenesis responses. ABBREVIATIONS ACC oxydase AP-1/2 APC BH4 bZIP cADPR cAMP CD CICR cGMP COX CTL

= = = = = = = = = = = =

1-Aminocyclopropane-l-Carboxylic Acid Oxydase Activator Protein 1/2 Antigen-Presenting Cells Tetrahydrobiopterin Basic Leucine Zipper Cyclic ADP-Ribose Cyclic AMP Cell Surface Glycoprotein Ca^"^-Induced Ca^"^ Release Cyclic GMP Cyclooxygenase Cytotoxic T Lymphocytes

951 DPI EDRF EpRE ER FAD FasL FMN GADPH Y-GCS GSH GSNO GST HR Hsp IL IFN-y IRE IRP JA JAK LAK L-NMMA LPS MAP kinase MHC NF-KB

NGF NK NMDA NO NOS eNOS iNOS nNOS NOx NR cNR p53 PAL

Diphenylene lodonium Endothelium-Derived Relaxing Factor Electrophile-Responsive Element Endoplasmic Reticulum Flavin Adenine Dinucleotide Fas Ligand Flavin Mononucleotide Glyceraldehyde-3 -Phosphate Dehydrogenase y-glutamylcysteine Synthetase Glutathione Nitrosoglutathione Glutathione S-Transferase Hypersensitive Response Heat Shock Protein Interleukin Interferon-y Iron Responsive Element Iron Regulatory Protein Jasmonate Janus Kinase Lymphokine-Activated Killer N(G)-Monomethyl-L-Arginine Lipopolysaccharide Mitogen-Activated Protein Kinase Major Histocompatibility Complex Nuclear Factor KB Nerve Growth Factor Natural Killer N-Methyl-D-Aspartate Nitric Oxide Nitric Oxide Synthase Endothelial Nitric Oxide Synthase Inducible Nitric Oxide Synthase Neuronal Nitric Oxide Synthase N0andN02 Nitrate Reductase Constitutive Nitrate Reductase a Tumour Suppressing Protein Phenylalanine Ammonia-Lyase

952

PARP PDZ motif PIOX PR PSD-95/93 redox ROS RYR SA SAR SIPK STAT TCR Th TMV TNF-a

Poly-ADP-Ribose-Polymerase Gly-Leu-Gly-Phe Motif Pathogen Induced Oxygenase Pathogenesis Related Post-Synaptic Density Protein 95/93 Oxido-Reduction Reactive Oxygen Species Ryanodine Receptor SalicyHc Acid Systemic Acquired Resistance SahcyHc Acid Induced Protein Kinase Signal Transducer and Activator of Transcription T Cell Receptor Helper T Lymphocytes Tobacco Mosaic Virus Tumor Necrosis Factor a

ACKNOWLEDGEMENTS We thank Drs. Marie-Noelle Binet and Marie-Joe Farmer for helpful comments on this manuscript and Pr. Daniel F. Klessig for sharing results before publication.

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12]

Mitchell, H.H.; Schonle, H.A.; Grindy, H.S.; J. Biol Chem, 1916, 24, 461-490. Green, L.C.; Tannenbaum, S.R.; Goldman, P.; Science, 1981, 212, 56-58. Stuehr, D.J.; Marietta, M.; Proc. Natl Acad. Sci. USA, 1985, 82, 7738-7742. Hibbs, J.B.; Taintor, R.R.; Vavrin, Z.; Science, 1987, 235, 473-476. Marietta, M.A.; Yoon, P.S.; Iyengar, R.; Leaf, CD.; Wishnok, J.S.; Biochem, 1988,27,8706-8711. Hibbs, J.B.; Taintor, R.R.; Vavrin, Z.; Rachlin, E.M.; Biochem. Biophy. Res. Commun, 1988, 157, 87-94. Hibbs, J.B.; Vavrin, Z.; Taintor, R.; J. Immunol, 1987,138, 550-565. Ignarro, L.J.; Buga, G.M.; Wood, K.S.; Byms, R.E.; Chaudhuri, G.; Proc. Natl. Acad. Sci. USA, 1987, 84, 9265-9269. Palmer, R.; Ferridge, A.G.; Moncada, S.; Nature, 1987, 327, 524-526. Arnal, J.; Warin, L.; Michel, J.; J. Clin. Invest, 1992, 90, 647-652. Stamler, J.S.; Cell, 1994, 78, 931 -936. Stamler, J.S.; Singel, D.J.; Loscalzo, J.; Science, 1992, 258, 1898-1902.

953

[13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41]

[42]

Bredt, D.S.; Hwang, P.M.; Glatt, C.E.; Lowenstein, C; Reed, R.R.; Sydner, S.n.; Nature, 1991, 557, 714-718. Koshland Jr., D.E.; Science, 1992, 258, 1861. Schmidt, H.H.; Walter, U.; Cell, 1994, 78, 919-925. Hausladen, A.; Stamler, J.A.; Proc. Natl. Acad Sci. USA, 1998, 95, 1034510347. Dangl, J.; Nature, 1998, 394, 525-527. Durner, J.; Klessig, D.F.; Curr. Opin. Plant Biol, 1999, 2, 369-374. Moncada, S.; Higgs A.; Furchgott R.; Pharmacol Rev, 1997, 49, 137-142. Benthin, G.; Bjorkhem, I.; Breuer, O.; Sakinis, A.; Wennmalm, A.; Biochem J, 1997, 323, 853-858. Ignarro, L.J.; Fukuto, J.M.; Griscavage, J.M.; Rogers, N.E.; Byms, R.E.; Proc. Natl. Acad. Sci. USA, 1993, 90, 8103-8107. Beckman, J.S.; Koppenol, W.H.; Am. J. Physiol, 1996, 271, 1424-1437. Lipton, S.A.; Choi, Y.B.; Pan, Z.H.; Lei, S.Z.; Chen, H.S.; Sucher, N.J.; Loscalzo, J.; Singel, D.J; Stamler, J.S.; Nature, 1993, 364, 626-632. Beck, K.-L.; Eberhardt, W; Frank, S.; Huwiler, A.; MeBmer, U.K., Muhl, H.; Pfeilschifter, J.; J. Exp. Biol, 1999,202, 645-653. Kooy, N.W.; Royall, J.A.; Arch. Biochem. Biophys, 1994, 310, 352-359. Balazy, M.; Kaminski, P.M.; Mao, K.; Tan, J.; Wolin, M.S.; J. Biol. Chem, 1998,273,32009-32015. Ignarro, L.J.; Pharmacol. Toxicol, 1990, 67, 1-7. Ignarro, L.J.; Blood Vessels, 1991, 28, 61-13. Liu, Z.; Rudd, M.A.; Freedman, J.E.; Loscalzo, J.; J. Pharmacol. Exp. Ther, 199S, 284, 526-534. Stamler, J.S.; Curr. Top. Microbiol. Immunol, 1995, 196, 19-36. Stamler, J.S.; Simon, D.L; Osborne, J.A.; Mullins, M.E.; Jaraki, O.; Michel, T.; Singel, D.J.; Loscalzo, J.; Proc. Natl. Acad Sci. USA, 1992, 89, 444-448. Mannick, J.B.; Hausladen, A.; Liu, L.; Hess, D.T.; Zeng, M.; Miao, Q.X.; Kane, L.S.; Gow, A.J.; Stamler, J.S.; Science, 1999, 284, 651-654. Xu, L.; Eu, J.P.; Meissner, G.; Stamler, J.S.; Science, 1998, 279, 234-237. Mohr, S.; Stamler, J.S.; Brune, B.; FEBS Lett, 1994, 384, 223-227. Marietta, M.A.; Cell, 1994, 78, 927-930. Forstermann, U.; Boissel, J.P.; Kleinert, H.; FASEB J, 1998, 12, 773-790. Nathan, C; Xie, Q.W.; Cell, 1994, 78, 915-918. Kishimoto, J.; Spurr, N.; Liao, M.; Lizhi, L.; Emson, P.; Xu, W.; Genomics, 1992,12, 802-804 [erratum; Genomics, 1993,15, 465]. Hall, A.V.; Antoniou, H.; Wang, Y.; Cheung A.H.; Arbus, A.M.; Olson, S.L.; Lu, W.C; Kau, C.L.; Marsden, P.A.; J. Biol. Chem, 1994, 269, 33082-33090. Boissel, J.P.; Schwarz, P.M.; Forstermann, U.; Nitric oxide, 1998, 2, 337-349. Rodrigo, J.; Riveros Moreno, V.; Bentura, M.L,; Uttenthal, L.O.; Higgs, E.A.; Fernandez, A.P.; Polak, J.M.; Moncada, S.; Martinez Murillo, R.; J. comp. Neurol, 1997, 378, 522-534. Ponting, C.P.; Phillips, C; Davies, K.E.; Blake, D.J.; Bioessays, 1997, 19, 469479.

954

[43]

[44] [45]

[46] [47] [48] [49]

[50] [51] [52]

[53]

[54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66]

Brenman, J.E.; Chao, D.S.; Gee, S.H.; McGee, A.W., Craven, S.E.; Santillano, D.R.; Wu, Z.; Huang, F.; Xia, H.; Peters, M.F.; Froehner, S.C; Bredt, D.S.; Cell, 1996, 84,151-161. Marsden, P.A.; Heng, H.H.; Scherer, S.W.; Stewart, R.J,; Hall, A.V.; Shi, X.M.; Tsui, L.C.; Schappert, K.T.; J. Biol Chem, 1993, 268, 17478-17488. Miyahara, K.; Kawamoto, T.; Sase, K.; Yui, Y.; Toda, K.; Yang, L.X.; Hattori, R.; Aoyama, T.; Yamamoto, Y.; Doi, Y.; Ogoshi, S.; Hashimoto, K.; Kawai, C; Sasayama, S.; Shizuta, Y.; Eur. 1 Biochem, 1994, 223, 719-726. Zhang, R.; Min, W.; Sessa, W.C; J. Biol. Chem, 1995, 270, 15320-15326. Busconi, L.; Michel, T.; J. Biol. Chem, 1993, 268, 8410-8413. Liu, J.W.; Hughes, T.E.; Sessa, W.C; J. Cell Biol, 1997,137, 1525-1535. Sessa, W.C; Garca-Cardea, G.; Liu, J.; Keh, A.; Pollock, J.S.; Bradley, J.; Thiru, S.; Braverman, LM.; Desai, K.M.; J. Biol. Chem, 1995, 270, 1764117644. Feron, O.; Belhassen, L.; kobzik, L.; Smith, T.W.; Kelly, R.A.; Michel, T.; J. Biol. Chem, 1996, 277,22810-22814. Michel, J.B.; Feron, O.; Sacks, D.; Michel, T.; J. Biol. Chem, 1997, 272, 15583-15586. Chartrain, N.A.; Geller, D.A.; Koty, P.P.; Sitrin, N.F.; Nussler, A.K.; Hoffman, E.P.; Billiar, T.R.; Hutchinson, N.I.; Mudgett, J.S.; J. Biol. Chem, 1994, 269, 6765-6772. de Vera, M.E.; Shapiro, R.A.; Nussler, A.K.; Mudgett, J.S.; Simmons, R.L.; Morris Jr., S.M.; Billiar, T.R.; Geller, D.A.; Proc. Natl Acad. ScL USA, 1996, 93, 1054-1059. Chu, S.C; Marks-Konczalik, J.; Wu, H.P.; Banks, T.C, Moss, J.; Biochem. Biophys. Res. Commun, 1998, 248, 871-878. Mayer, B.; Hemmens, B.; Trends Biol Sci, 1997, 22, 477-481. Poulos, T.L.; Raman, C.S.; Huiying, L.; Structure, 1998, 6, 255-258. Abu-Soud, H.M.; Feldman, P.L.; Clark, P.; Stuehr, D.J.; J. Biol Chem, 1994, 269, 32318-32326. Wang, J.; Stuehr, D.J.; Ikeda-Saito, M.; Rousseau, D.L.; J. Biol Chem, 1993, 268, 22225-22258. Crane, B.R.; Tainer, J.A.; Science, 1997, 278, 425-431. Crane, B.R.; Arvai, A.S.; Ghosh, D.K.; Wu, C; Getzoff, E.D.; Stuehr, D.J.; Tainer, J.A.; Science, 1998, 279, 2121-2126. Hellermann, G.R.; Solomonson, L.P.; J. Biol Chem, 1997, 272, 12030-12034. Matsuoka, A.; Stuehr, D.J.; Olson, J.S.; Clark, P.; Ikeda-Saito, M.; J. Biol Chem, 1994, 269, 20335-20339. Su, Z.; Blazing, M.A.; Fan, D.; George, S.E.; J. Biol Chem, 1995, 270, 2911729122. Xia, Y.; Zweier, J.L.; Proc. Natl Acad Scl USA, 1997, 94, 6954-6958. Xia, Y.; Tsai, A.L.; Berka, V.; Zweier, J.L.; J. Biol Chem, 1998, 273, 2580425808. Xia, Y.; Roman, L.J.; Masters, B.S.; Zweier, J.L.; J. Biol Chem, 1998, 273, 22635-22639.

955

[67] [68] [69] [70] [71]

[72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94]

Pou, S.; Keaton, L.; Surichamom, W.; Rosen, G.M.; J. Biol Chem, 1999, 274, 9573-9580. Raman, C.S.; Li, H.; Martasek, P.; Krai V.; Masters, B.S.; Poulos, T.L.; Cell, 1998, 95, 939-950. Kornau, H.C.; Schenker, L.T.; Kennedy, M.B.; Seeburg, P.H.; Science, 1995, 269, 1737-1740. Nathan, C; Xie, Q.W.; J. Biol. Chem, 1994, 269, 13725-13728. Salerno, J.C; Harris, D.E.; Irizarry, K.; Patel, B.; Morales, A.J.; Smith, S.M.; Martasek, P.; Roman, L.J.; Masters, B.S.; Jones, C.L.; Weissman, B.A.; Lane, P.; Liu, Q.; Gross, S.S.; J. Biol, Chem, 1997, 272, 29769-29777. Daff, S.; Sagami, L; Shimizu, T.; J. Biol. Chem, 1999, 274, 30589-30595. Lee, S.-J.; Stull J.T.; J. Biol Chem, 1998, 273, 27430-27437. Rolland, C.R.; Walp, L.; Flye, M.W.; J. Immunol, 1994,153, 5453-5464. Lu, L.; Bonham, C.A.; Chambers, F.G.; Watkins, S.C; Hoffman, R.A.; Simmons, R.L.; Thomson, A.W.;y. Immunol, 1996,157, 3577-3586. Siedlar, M.; Marcinkiewicz, J.; Zembala, M.; Clin. Immunol. Immunopathol, 1995,77,380-384. Tian, L.; Noelle, R.J.; Lawrence, D.A.; Eur. J. Immunol, 1995, 25, 306-309. Lander, H.M.; Sehajpal, P.; Levine, D.M.; Novogrodsky A.; J. Immunol, 1993, 150, 1509-1516. Lander, H.M.; Ogiste, J.S.; Pearce, S.F.; Levi, R.; Novogrodsky, A.; J. Biol Chem, 1995,270,1011-7020. Efron, D.T.; Kirk, S. J.; Regan, M.C.; Wasserkrug, H.L.; Barbul, A.; Surgery, 1991,110, 327-334. Reisser, D.; Amould, L.; Maynadie, M.; Belichard, C; Coudert, B.; Jeannin, J.F.; J. Endotoxin Res, 2000, in press. Bingisser, R.M.; Tilbrook, P.A.; Holt, P.G.; Kees, U.R.; J. Immunol, 1998, 160, 5729-5734. Brito, C; Naviliat, M.; Tiscomia, A.C.; Vuillier, F.; Gualco, G.; Dighiero, G.; Radi, R.; Coyota, A.M.; Immunol, 1999, 162, 3356-3366. Sicher, S.C; Vazquez, M.A.; Lu, C.Y.; Immunol, 1994, 153, 1293-1300. Heuschling, P.; J. Neuroimmunol, 1995, 57, 63-69. Taylor-Robinson, A.W.; Immunol, 1997, 75, 167-175. Marcinkiewicz, J.; Grabowska, A.; Chain, B.M.; J. Immunol, 1996, 156, 46174621. Bauer, H.; Jung, T.; Stichtenoth, D.O.; Frolich, J.C; Neumann, C; Immunol, 1997,90,205-211. Taub, D.D.; Cox, G.W.; J. Leukoc. Biol, 1995, 58, 80-89. Munder, M.; Eichmann, K.; Modolell, M.; J. Immunol, 1998,160, 5347-5354. Williams, M.S.; Nogushi, S.; Hemkart, P.A.; Osawa, Y.; J. Immunol, 1998,161, 6526-6531. Cifone, M.G.; Festuccia, C; Cironi, L.; Cavallo, G.; Chessa, M.A.; Pensa, V.; Tubaro, E.; Santoni, A.; Cell Immunol, 1994,157, 181-194. Xiao, L.; Eneroth, P.H.; Qureshi, G.A.; Scand. J. Immunol, 1995, 42, 505-551. Filep, J.G.; Baron, C; Lachange, S.; Perreault, C; Chan, J.S.; Blood, 1996, 87, 5136-5143.

956

95] 96] 97] 98] 99] 100] 101] 102] 103] 104] 105] 106]

107] 108] 109] 110] 111] 112]

113] 114] 115] 116] 117] 118] 119]

Salvucci, O.; Kolb, J.P.; Dugas, N.; Chouaib, S.; Blood, 1998, 92, 2093-2102. Juretic, A.; Spagnolli, G.C.\ Von Bremen, K.; Horig, H.; Filgueira, L.; Lusher, U.; Babst, R.; Harder, F.; Heberer, M.; Cell. Immunol, 1994, 757, 462-477. Kurose, I.; Miura, S.; Saito, H.; Tada, S.; Fukumura, D.; Higuchi, H.; Ishii, H.; Gastroenterol, 1995, 109, 1958-1968. Samlowski, W.E.; Yim, C.Y.; McGregor, J.R.; Nitric Oxide, 1998, 2, 45-56. Nishikawa, M.; Sato, E.F.; Utsumi, K.; Inoue, M.; Cancer Res, 1996, 56, 45354540. Drapier, J.C; Hibbs, J.B.; J. Clin. Invest, 1986, 78, 790-797. Stuehr, D.J.; Nathan, C.F.; J. Exp. Med, 1989,169, 1543-1555. Buga, G.M.; Wei, L.H.; Bauer, P.M.; Fukuto, J.M.; Ignarro, L.J.; Amer. J. Physiol, 1998, 44, 1256-1264. Lepoivre, M.; Chenais, B.; Yabo, A.; Lemaire, G.; Thelander, L.; Tenu, J.P.; J. Biol. Chem, 1990,2(55, 14143-14149. Lepoivre, M.; Fieschi, F.; Coves, J.; Thelander, L.; Fontecave, M.; Biochem. Biophys. Res. Comm, 1991, 179, 442-448. Lepoivre, M.; Bobe, P.; Flaman, J.M.; Lemaire, G.; Henry, Y.; J. Biol. Chem, 1994, 2(5P, 21891-21897. EI-Deiry, W.S.; Harper, J.W.; O'Connor, P.M.; Velculescu, V.E.; Canman, C.E.; Jackman, J.; Pietenpol, J.A.; Burrell, M.; Hill, D.E.; Wang, Y.; Cancer Res, 1994, 54, 1169-1174. Kastan, M.B.; Zhan, Q.; El-Deiry, W.S.; Carrier, F.; Jacks, T.; Walsh, W.V.; Plunkett, B.S.; Vogelstein, B.; Fomace, A.J.; Cell, 1992, 71, 587-597. Wu, X.; Bayle, J.H.; Olson, D.; Levine, A.J.; Genes Dev, 1993, 7, 1126-1132. Poluha, W.; Schonhoff, CM.; Harrington, K.S.; Lachyankar, M.B.; Crosbie, N.E.; Bulseco, D.A.; Ross, A.H.; J. Biol. Chem, 1997, 272, 24002-24007. Wang, Q.; Zambetti, G.P.; Suttle, D.P.; Mol. Cell. Biol, 1997, 17, 389-397. Young, L.T.; Woods, CM.; Wang, J.F.; Asghari, V.; Life Sci, 1997, 60, 12791285. Ouaaz, F.; Sola, B.; Issaly, F.; Kolb, J.P.; Davi, F.; Mentz, F.; Arock, M.; PaulEugene, N.; Komer, M.; Dugas, B.; Debre, P.; Mossalayi, M.D.; Blood, 1994, (^4,3095-3104. Estrada, C; Gomez, C; Martin-Nieto, J.; De Frutos, T.; Jimenez, A.; Villalobo, A.; Biochem. J, 1997, 326, 369-376. Calmels, S.; Hainaut, P.; Ohshima, H.; Cancer Res, 1997, 57, 3365-3369. Xie, K.; Wang, Y.; Huang, S.; Bielenberg, D.; Salas, T.; McConkey, D.J.; Jiang, W.; Fidler, I.J.; Oncogene, 1997, 75, 771-779. Kitamura, Y.; Kamoshima, W.; Shimohama, S.; Nomura, Y.; Taniguchi, T.; Neurochem. Int, 1998, 32, 93-102. Ho, Y.S.; Lee, H.M.; Mou, T.C; Wang, Y.L.; Lin, J.K.; Mol. Carcinog, 1997, 19, 101-113. Uehara, T.; Kikuchi, Y.; Nomura, Y.; J. Neurochem, 1999, 72, 196-205. Okuno, S.; Shimizu, S.; Ito, T.; Nomura, M.; Hamada, E.; Tsujimoto, Y.; Matsuda, H.; J. Biol. Chem, 1998, 273, 34272-34277.

957

[120]

[121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143]

Kurose, I.; Higuchi, H.; Yonei, Y.; Ebinuma, H.; Watanabe, N.; Hokari, R.; Fukumara, D.; Miura, S.; Takaishi, M.; Saito, H.; Nakatsumi, R.C.; Ishii, H.; Gastroenterol, 1996, HI, 1058-1070. Sveinbjomsson, B.; Olsen, R.; Setemes, O.M.; Seljelid, R.; Biochem. Biophys. Res. Comm, 1996, 223, 643-649. Yamashita, T.; Uchida, T.; Araki, A.; Sendo, F.; Int. J. Cancer, 1997, 77, 223230. Geng, Y.J.; Hellstrand, K.; Wennmalm, A.; Hansson, G.K.; Cancer Res, 1996, 56, 866-874. Messmer, U.K.; Reimer, D.M.; Brtine, B.; Eur. J. Pharmacol, 1998, 349, 333343. Mohr, S.; Zech, B.; Lapetina, E.G.; Brtine, B.; Biochem. Biophys. Res. Commun, 1997, 238, 387-391. Yabuki, M.; Kariya, S.; Inai, Y.; Hamazaki, K.; Yoshioka, T.; Yasuda, T.; Horton, A.A.; Utsumi, K.; Free Radic. Res, 1997, 27, 325-335. Mannick, J.B.; Miao, X.Q.; Stamler, J.S.; 1 Biol. Chem, 1997, 272, 2412524128. Sciorati, C; Rovere, P.; Ferrarini, M.; Heltai, S.; Manfredi, A.A.; Clementi, E.; 1 Biol. Chem, 1997, 272, 23211-23215. Liu, L.; Stamler, J.S.; Differenciation, 1999, 6, 937-942. Mannick, J.B.; Asano, K.; Izumi, K.; Kieff, E.; Stamler, J.S.; Cell, 1994, 79, 1137-1146. Dimmeler S.; Haendeler, J.; Sause, A.; Zieher, A.M.; Cell Growth Differ, 1998, 9,415-422. Chlichlia, K.; Peter, M.E.; Rocha, M.; Scaffidi, C; Bucur, M.; Krammer, P.H.; Schirrmacher, V.; Umansky, V.; Blood, 1998, 91, 4311-4320. Messmer, U.K.; Ankarcrona, M.; Nicoreta, P.; Brtine, B.; FEBS Lett, 1994, 355, 23-26. Messmer, U.K.; Brtine, B.; Arch. Biochem. Biophys, 1995, 327, 1-10. Messmer, U.K.; Reed, J.C.; Brtine, B.; J. Biol. Chem, 1996, 271, 20192-20197. Brtine, B.; Fon Knethen, A.; Sando, K. B.; Cell Death Differenc, 1999, 6, 969975. Messmer, U.K.; Reimer, D.M.; Reed, J.C; Brtine, B.; FEBS Lett, 1996, 384, 162-166. Albina, J.E.; Cui, S.; Mateo, R.B.; Reishner, J.S.; J. Immunol, 1993,150, 50805085. Sarih, M.; Souvannavong, V.; Adam, A.; Biochem. Biophys. Res. Commun, 1993,191, 503-508. Okuda, Y.; Sakoda, S.; Shimaoka, M.; Yanagihara, T.; Immunol. Lett, 1996, 52, 135-138. Selleri, C; Sato, T.; Raiolla, A.M.; Rotoll, B.; Young, N.S.; Maciejewski, J.P.; Brit. J. Haematol, 1997, 99, 481-489. Hortelano, S.; Dallaporta, B.; Zamzani, N.; Hirsh, T.; Susin, S.A.; Marzo, I.; Bosca, L.; Kroemer, G.; FEBS Lett, 1997, 410, 313-311. Dimmeler, S.; Haendeler, J.; Nehls, M.; Zeiher, A.; J. Exp. Med, 1997, 185, 601-607.

958

[144] 45] 146] 147]

148] 149] 150] 151] 152] 153] 154]

155] 156] 157] 158] 159] 160] 161] 162] 163] 164] 165] 166] 167]

Rossig, L.; Fichtlscherer, B.; Breitschopf, K.; Haendeler, J.; Zeiher, A.M.; Mulsch, A.; Dimmeler, S.; 1 Bio. Chem, 1999, 274, 6823-6826. Haendeler, J.; Zeiher, A.M.; Dimmeler, S.; Vitam. Norm, 1999, 57, 49-77. Magrinat, G.; Mason, S.N.; Shami, P.J.; Weinberg, J.B.; Blood, 1992, 80, 18801884. Dugas, N.; Mossalayi, M.D.; Calenda, A.; Leotard, A.; Becherel, P.; Mentz, F.; Ouazz, P.; Arock, M.; Debre, P.; Domand, J.; Dugas, B.; Blood, 1996, 88, 3528-3534. Peunova, N.; Enikolopov, G.; Nature, 1995, 375, 68-73. Obregon, E.; Punzon, M.C.; Gonzales-Nicolas, J.; Fernandez-Cruz, E.; Fresno, M.; Munoz-Femandez, M.A.; Eur J. Neurosci, 1997, 9, 1184-1193. Vecchini, F.; Pringault, E.; Billiar, T.R.; Geller, D.; Hausel, P.; Felley-Bosco, E.; Cell Growth Diff, 1997, 8, 261-268. Kuo, H.P.; Ho, T.C.; Wang, C.H., Yu, C.T.; Lin, H.C.; Uber. Lung. Dis, 1996, 77, 468-477. Swierkosz, T.A.; Mitchell, J.A.; Warner, T.D.; Botting, R.M.; Vane, J.R.; J. Pharmacol, 1995, 114, 1335-1342. Mollace, V.; Musculi, C; Rotirot, D.; Nistico, G.; Biochem. Biophys. Res. Comm, 1997, 238,916-919. Forrester, K.; Ambs, S.; Lupoid, S.E.; Kapust, R.B.; Spillare, E.A.; Weinberg, W.C; Felley-Bosco, E.; Wang, X.W.; Geller, D.A.; Tzeng, E.; Billiar, T.R.; Harris, C.C; Proc. Natl. Acad Sci. USA, 1996, 93, 2442-2447. Matthews, J.R.; Bottin, C.H.; Panico, M.; Morris, H.R.; Hay, R.T.; Nucleic Acid Res, 1996, 24, 2236-2242. Park, S.K.; Lin, H.L.; Murphy, S.; Biochem. J, 1997, 322, 609-613. Peng, H.B.; Libby, P.; Liao, J.K.; J. Biol. Chem, 1995, 270, 14214-14219. Togashi, H.; Sasaki, M.; Frohman, E.; Taira, E.; Ratan, R.R.; Dawson, T.M.; Dawson, V.L.; Proc. Natl. Acad Sci. USA, 1997, 94, 2676-2680. Feinstein, D.L.; Galea, E.; Aquino, D.A.; Li, G.C.; Xu, H.; J. Biol. Chem, 1996, 271, 17724-17732. Nitsch, D.; Ghilardi, N.; Muhl, H.; Nitsch, C; Brune, B.; Pfeilschifter, J.; Am. J. Pathol, 1997, 150, 889-900. Genaro, A.M.; Hortelano, S.; Alvarez, A.; Martinez, C; Bosca, L.; J. Clin. Invest, 1995, 95, 1884-1890. Martin-Sanz, P.; Diaz-Guerra, M.J.M.; Casado, M.; Bosca, L.; Hepathology, 1996, 23, 1200-1207. Briine, B.; Gotz, C; Messmer, U.K.; Sandau, K.; Hirvonen, M.R.; Lapetina, E.G.; J. Biol. Chem, 1997, 272, 7253-7258. Sandau, K.; Pfeilschifter, J.; Brune, B.; J. Immunol, 1997,158, 4938-4946. Jeannin, J.F.; Onier, N.; Lagadec, P.; Jeney, N.; Liehl, E.; Gastroenterology, 1991,101,126-133. Lejeune, P.; Lagadec, P.; Onier, N.; Pinard, D.; Ohshima, H.; Jeannin, J.F.; J. Immunol, 1994,152, 5077-5084. Onier, N.; Hilpert, S.; Reveneau, S.; Amould, L.; Saint-Giorgio, V.; Exbrayat, J.M.; Jeannin, J.F.; Int. J. Cancer, 1999, 81, 755-760.

959

[168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181]

[182] [183] [184] [185] [186] [187] [188]

[189] [190]

[191] [192] [193] [194] [195] [196]

Lagadec, P.; Raynal, S.; Lieubeau, B.; Onier, N.; Amould, L.; Saint-Giorgo, V.; Lawrence, D.; Jeannin, J.F.; Am. J. Pathol, 1999, 154, 1867-1876. Onier, N.; Hilpert, S.; Amould, L.; Saint-Giorgio, V.; Davies, J.G.; Bauer, J.; Jeannin, J.F.; Clin. Exp. Metast, 1999, 17, 299-306. Goldring, C; Reveneau, S.; Pinard, D.; Jeannin, J.F.; Eur. J. Immunol, 1998, 28, 2960-2970. Nishimura, H.; Hayamizu, T.; Yanagisawa, Y.; Environ. Sci. Technol, 1986, 20, 413-416. Churchill, K.A.; Klepper, L.A.; Pest Biochem. Physiol, 1979, 12, 156-162. Klepper, L.A.; Atmos. Environ, 1979,13, 537-542. Klepper, L.A.; Plant Physiol, 1990, 93, 26-32. Cooney, R.V.; Harwood, P.J.; Custer, L.J., Franke, A.A.; Environ. Health Perspect, 1994, 102, 460-462. Solomonso, L.P.; Barber, M.J.; Ann. Rev. Plant Physiol. Plant Mol. Biol, 1990, 41, 225-235. Nelson, R.S.; Ryan, S.A.; Harper, J.E.; Plant Physiol, 1983, 72, 503-509. Ryan, S.A.; Nelson, R.S.; Harper, J.E.; Plant Physiol, 1983, 72, 510-514. Streit, L.; Nelson, R.S.; Harper, J.E.; Plant Physiol, 1985, 78, 80-84. Nelson, R.S.; Streit S.L.; Harper, J.E.; Plant Physiol, 1986, 80, 72-76. Beevers, L.; Hageman, R.H. In Inorganic Plant Nutrition, Encyclopedia of Plant Physiology', A. Lauchi, R.L. Bieleski, Eds.; Springer-Verlag: New york, 1983; Vol. 75, pp. 351-375. Streit, L.; Harper, J.E.; Plant Physiol, 1986, 81, 593-596. Dean, J.V.; Harper, J.E.; Plant Physiol, 1988,104, 389-395. Caro, A.; Puntarulo, S.; Physiol. Plant, 1998,104, 357-364. Dean, J.V.; Harper, J.E.; Plant Physiol, 1986, 82, 718-723. Yamasaki, H.; Sakihama, Y.; Takahashi, S.; Trends Plant Sci, 1999, 4, 128-129. Cueto, M.; Hemandez-Perera, O.; Martin, R.; Bentura, M.L.; Rodrigo, J., Lamas, S.; Golvano, M.P.; FEBSLett, 1996, 398, 159-164. Huang, J.-S.; Knopp, J.A. In Proceedings of the second international bacterial wilt symposium', P. Prior, J. Elphinstone, C. Allen, Eds.; INRA Eds, 1998, pp. 218-224. Ninnemann, H.; Maier J.; Photochem. Photobiol, 1996, 64, 393-398. Barroso, J.B.; Corpas, F.J.; Carreras, A.; Sandalio, L.M.; Valderrama, R.; Palma, J.M.; Lupianez, J.A.; del Rio, L.A.; J. Biol. Chem, 1999, 274, 3672936733. Leshem, Y.Y.; Harmaty, E.; J. Plant Physiol, 1996,148, 258-263. Allan, A.C.; Fluhr, R.; Plant Cell, 1997, 9, 1559-1572. Ribeiro, E.A.; Cunha, F.Q.; Tamashiro, W.M.; Martins, I.S.; FEBS Lett, 1999, 445, 283-286. Kuo, W.N.; Ku, T.W.; Jones, D.L.; Jn-Baptiste, J.; Biochem. Arch., 1995, 11, Sen, S.; Cheema, I.R.; Biochem. Arch, 1995, II, 221-227. Kuo, W.N.; Jn-Baptiste, J.B.; Kanadia, R.N.; McNabb, L.D.; Zhai L.; Weeks, K.; Dopson, N.; Chambers, M.C.; Cytobios, 1996, 87,251-263.

960

[197] [198] [199] [200] [201] [202] [203] [204] [205] [206] [207] [208] [209] [210] [211] [212] [213] [214] [215] [216] [217] [218] [219] [220] [221 ] [222] [223] [224] [225] [226] [227]

Staskawicz, B.J.; Ausubel F.M.; Baker B.J.; Ellis J.G.; Jones J.D.G.; Science, 1995,2^5,661-667. Bent, A. F.; Plant Cell, 1996, 8, 1757-1771. Hammond-kosack, K.E.; Jones J.D.G.; Plant Cell, 1996, 8, 1773-1791. Yang, Y.; Shah, J.; Klessig, D.F.; Genes Dev., 1997, 11, 1621-1639. Alvarez, M.E.; Pennell, R.I.; Meijer, P.J.; Ishikawa, A.; Dixon, R.A.; Lamb, C; Cell, 1998, 92, 773-784. Dumer, J.; Wendehenne, D.; Klessig, D.F.; Proc. Natl. Acad ScL USA, 1998, 95, 10328-10333. Delledonne, M.; Xia, Y.; Dixon, R.A.; Lamb, C.; Nature, 1998, 394, 585-588. Bolwell, G.P.; Curr. Opin. Plant Biol, 1999, 2, 287-294. Jeffrey, S.R.; Snyder, S.H.; Science, 1996, 274, 774-777. Leshem, Y.Y.; Plant Gr. Reg, 1996, 18, 155-159. Leshem, Y.Y.; Haramaty, E.; Iluz, D.; Malik, Z.; Sofer, Y.; Roitman, L.; Leshem, Y.; Plant Physiol. Biochem, 1997, 35, 573-579. Navarre, D.; Wendehenne, D.; Dumer, J.; Noad, R.; Klessig, D.F.; Plant Physiol, 2000,122, 573-582. Abeles, F.B.; Morgan, P.W.; Saltveit Jr., M.E. Ethylene in Plant Biology, Academic Press, CA: San Diego, 1992. Pennell, R.L; Lamb, C; Plant Cell, 1997 ,9, 1157-1166. Fluhr, R.; Mattoo, A.K.; Cht. Rev. Plant Sci, 1996, 75, 479-523. Hoffman, N.E.; Yang, S.F.; Ichihara, A.; Sakamura, S.; Plant Physiol, 1982, 70, 195-199. Laxalt, A.M., Beligini, M.V.; Lamattina, L.; Eur. J. Plant Pathol, 1997, 73, 643-651. Beligni, M.V.; Lamattina, L.; Nitric oxide, 1999, 3, 199-208. Beligni, M.V.; Lamattina, L.; Trends Plant Sci, 1999, 4, 299. Beligini, M.V.; Lamattina, L.; Planta, 1999, 208, 337-344. Wink, D.A.; Hanbauer, I.; Krishna, M.C.; De Graff, W.; Gamson, J.; Mitchell, J.B.; Proc. Natl. Acad Sci. USA, 1993, 90, 9813-9817. Mathieu, C; Moreau, S.; Frendo, P.; Puppo, A.; Davies, M.J.; Free radic. Biol Med, 1998, 24, 1242-1249. Parker, J.E.; Coleman, M.J.; Trends Biochem Sci, 1997, 22, 291-296. Nurnberger, T.; Nennstiel, D.; Jabs, T., Sacks, W.R.; Hahlbrock, K.; Scheel, D.; Cell, 1994, 78, 449-460. Nurnberger, T.; Cell Mol Life Sci, 1999, 55, 167-182. Lebrun-Garcia, A.; Bourque, S.; Binet, M.-N.; Quaked, F.; Wendehenne, D.; Chiltz, A.; Schafftier, A.; Pugin, A.; Biochimie, 1999, 81, 1-6. Reymond, P.; Farmer, E.E.; Curr. Opin. Plant Biol, 1998, 1, 404-411. Datta, S.K., Muthukrishnan, S. Pathogenesis-related proteins in plants', CRC press LLC, 1999. Taylor, C.B.; Plant Cell, 1998, 10, 873-875. Van Camp, W.; Van Montagu, M.; Inze, D. Trends Plant Sci., 1998, 3, 330. Pugin, A.; Frachisse, J.-M.; Tavemier, E.; Bligny, R.; Gout, E.; Douce, R.; Guem, J.; Plant Cell, 1997, 9, 2077-2091.

961

[228]

[229] [230] [231] [232] [233] [234] [235] [236] [237] [238] [239] [240] [241 ] [242] [243] [244] [245] [246] [247] [248] [249] [250] [251] [252] [253] [254] [255] [256] [257]

Klessig, D.F.; Durner, J.; Shah, J.; Yang, Y. In Phytochemical Signals and Plant-microbe Interactions', J.T. Romeo, K.R. Downum, R. Verpoorte, eds.; Plenum Press: New York, 1998; Vol. 32, pp. 119-137. Noritake, T.; Kawakita, K.; Doke, N.; Plant Cell Physiol, 1996, 37, 113-116. Pfeiffer, S.; Janystin, B.; Jessner, G.; Pichomer, H.; Ebermann, R.; Phytochem, 1994,5^,259-262. Bowler, C; Neuhaus, G.; Yamagata, H.; Chua, N.H.; Cell, 1994, 77, 73-81. Neuhaus, G.; Bowler, C; Hiratsuka, K.; Yamagata, H.; Chua, N.H.; EMBO, 1997,10, 2554-2564. Sentenac, H.; Bonneaud, N.; Minet, M.; Lacroute, F.; Salmon, P.M.; Gaymard, P.; Grignon, C; Science, 1992, 256, 663-665. Hiroshi, T.; 1 Gen. Physiol, 1995, 105, 309-328. Koler, C; Neuhaus, G.; Plant Physiol, 1998,116, 1604. Leng, Q.; Mercier, R. W.; Yao, W.; Berkowitz, G.A.; Plant Physiol, 1999, 121, 753-761. Dzeja, C; Hagen, V.; Kaupp, U.B.; Frings, S.; EMBO, 1999, 18, 131-134. Galione, A.; White, A.; Willmott, N.; Turner, M.; Potter, B. V. L.; Watson, S.P.; Nature, 1993, 365, 456-459. Galione, A.; Science, 1993, 259, 325-326. Willmott, N.; Sethi, J.K.; Walseth, T.F.; Lee H.C.; White, A.M.; Galione, A.; J. Biol Chem, 1996, 271, 3699-3705. Galione, A.; White, A.; Trends Cell Biol, 1994, 4, 431 -436. Allen, G.J.; Muir, S.R.; Sanders, D.; Science, 1995, 268,1'iS-l^l. Muir, S.R.; Sanders, D.; FEBSLett, 1996, 395, 39-42. Allen, G.J.; Sanders, D.; Adv. Bot. Res, 1997, 25, 217-252. Walden, R.; Curr. Opin. Plant Biol, 1998, /, 419-423. Wu, Y.; Kuzma, J.; Marechal, E.; Graeff, R.; Lee, H.C.; Foster, R.; Chua, N.M.; Science, 1997, 278, 2126-2130. Leckie, C.P.; McAinsh, M.R.; Allen, G.J.; Sanders, D.; Hetherington A.M.; Proc. Natl. Acad Sci. USA, 1998, 95, 15837-15842. Volotovski, LD.; Sokolovsky, S.G.; Molchan, O.V.; Knight, M.R.; Plant Physiol, 1998,117, 1023-1030. Gruer, M.J.; Artymiuk, P.J.; Guest J.R.; Trends Biol. Sci, 1997, 22, 3-6. Klausner, R.D.; Rouault, T.A.; Harford, J.B.; Cell, 1993, 72, 19-28. Basilion, J.P.; Rouault, T.A.; Massinople, CM.; Klausner, R.D.; Burgess, W.H.; Proc. Natl. Acad Sci. USA, 1994, 91, 574-578. Philpott, C.C; Klausner, R.D.; Rouault, T.A.; Proc. Natl. Acad Sci. USA, 1994, P/, 7321-7325. Addess, K.J.; Basilion, J.P.; Klausner, R.D.; Rouault, T.A.; Pardi, A.; J. Mol. Biol, 1997, 274, 72'S3. Drapier, J.-C; Hiding, H.; Wietzerbin J.; Kaldy, P.; Kuhn L.C.; EMBO, 1993, 12, 3643-3649. Hentze, M.W.; Kuhn, L.C.; Proc. Natl. Acad Sci. USA, 1996, 93, 8175-8182. Gardner, P.R.; Costantino, G.; Szabo, C; Salzman, A.L.; J. Biol. Chem, 1997, 272,25071-25076. Peyret, P.; Perez, P.; Alric, M.; J. Biol. Chem, 1995,270, 8131-8137.

962

[258] [259] [260] [261] [262] [263] [264] [265] [266] [267] [268] [269] [270] [271] [272] [273] [274] [275] [276] [277] [278]

[279] [280] [281] [282] [283] [284]

Hayashi, M.; De Bellis, L.; Alpi, A.; Nishimura, M.; Plant Cell Physiol, 1995, 36, 669-680. Henderson, B.R.; Menotti, E.; Bonnard, C; Kuhn, L.; J. Biol. Chem, 1994, 269, 17481-17489. Verniquet, F.; Gaillard, J.; Neuburger, M.; Douce, R.; Biochem J, 1991, 276, 643-648. Vanlerberghe, G.C.; Mcintosh, L.; Plantphysiol, 1996, / / / , 589-595. Chivasa, S.; Carr, J.P.; Plant Cell, 1998,10, 1489-1498. Millar, A.H.; Day, D.A.; FEBSLett, 1996, 398, 155-158. Maxwell, D.P.; Wang, Y.; Mcintosh, L.; Proc. Natl. Acad. Sci. USA, 1999, 96, 8271-8276. Rothenberger, S.; Mullner, E. W.; Kuhn L.C.; Nucleic Acid Res, 1990,18, 11751179. Lescure, A.M.; Proudhon, D.; Pesey, H., Ragland, M.; Theil, E.G.; Briat, J.F. Proc. Natl. Acad Sci. USA, 1991, 88, 8222-8226. Torres, M.A.; Onuchi, H.; Hamada, S.; Machida, C.; Hammomd-Kosak, K.E. Jones, J.D.G. Plant J, 1998,14, 365-370. Keller, T.; Dumude, H.G.; Werner, D.; Doemer, P.; Dixon, R.A.; lamb, C. P/a/7/c^//, 1998, 7^255-266. Groom, O.J.; Torres, M.A.; Fordham-Skelton, P.; Hammond-Kosak, K.E. Robinson, N.J.; Jones, J.D.G.; Plant J, 1996,10, 515-522. Blumwald, E.; Aharon, G.S.; Lam, B.C.H.; Trends plant Sci, 1998, 3, 342-345. Lamb, C.; Dixon, R.A. ; Annu. Rev. Physiol. Plant. Mol Biol, 1997, 48, 251275. Levine, A.; Tenhaken, R.; Dixon, R.; Lamb, C.J.; Cell, 1994, 79, 583-593. Pontier, D.; Balague C.; Roby D.; C.R. Acad Sci. Paris, 1998, 321, 721-734. Del Pozo, O.; Lam, E.; Curr. Biol, 1998, 8, 1129-1132. Solomon, M.; Belenghi, B.; Delledonne, M.; Menachem, E.; Levine, A.; Plant Cell, 1999, 11,431-444. Zhu, Q.; Droge-Laser, W.; Dixon, R.A.; Lamb, C.; Curr. Opin. Genet. Dev, 1996, 6, 624-630. Lebel, E.; Heifetz, P.; Thome, L.; Uknes, S.; Ryals, J.; Ward, E.; Plant J, 1998, 16, 223-233. Ryals, J.A.; Weymann, K.; Lawton, K.; Friedrich, L.; Ellis, D.; Steiner, H.Y.; Johnson, J.; Delaney, T.P.; Jesse, T.; Vos, P.; Uknes, S.; Plant Cell, 1997, 9, 425-439. Cao, H.; Glazebrook, J.; Clarke, J.D.; Volko, S.; Dong, X.; Cell, 1997, 88, 5763. Shah, J.; Tsui, F.; Klessig, D.F.; Mol. Plant Microbe Interact, 1997, 10, 69-78. Zhang, Y.; Fan, W.; Kinkema, M.; Li, X.; Dong, X.; Proc. Natl. Acad Sci. USA, 1999, 96, 6523-6528. Zhou, J.M.; Trifa, Y.; Silva, H.; Pontier, D.; Lam, E.; Shah, J.; Klessig, D.F.; Mol. Plant-Microbe Interact, 2000,13, 191-202. Jonak, C.; Ligterink, W.; Hirt, H.; Cell. Mol. Life Sci, 1999, 55, 204-213. Ligterink, W.; Kroj, T.; zur Nieden, U.; Hirt, H.; Scheel, D.; Science, 1997, 276, 2054-2057.

963 [285] [286] [287] [288]

[289] [290] [291] [292] [293] [294] [295] [296] [297] [298] [299] [300] [301] [302] [303] [304] [305] [306] [307] [308] [309] [310] [311] [312] [313] [314] [315]

Zhang, S.; Klessig, D.F.; Plant Cell, 1997, P, 809-824. Romeis, T.; Piedras, P.; Zhang, S.; Klessig, D.F.; Hirt, H.; Jones, J.D.G.; Plant Ce//, 1999, 77,273-287. Liu, Y.; Zhang S.; Klessig, D.F.; Mol Plant Microbe Interact, 2000, 75, 118124. Shah, J.; Klessig, D.F. In Biochemistry and Molecular Biology of Plant Hormones, P.P.J Hooykaas, M.A. Hall, K.R. Libbenga, eds; Elsevier Science B.V., 1999, pp. 513-541. Kumar D.; Zhang S.; Klessig D.F.; Mol Plant Microbe Interact., 2000, 13, 347351. Seo, S.; Okamoto, M.; Seto, H.; Ishizuka, K.; Sano, H.; Ohashi, Y.; Science, 1995,270, 1988-1992. Zhang, S.; Klessig, D.F.; Proa Natl. Acad Sci. USA, 1998, 95, 7225-7230. Durner, J.; Shah, J.; Klessig, D.F.; Trends Plant Sci, 1997, 2, 266-214. Chen, Z.; Silva, H.; Klessig, D.F.; Science, 1993, 262, 1883-1886. Durner, J.; Klessig, D.F.; Proc. Natl. Acad Sci. USA, 1995, P2, 11312-11316. Conrath, U.; Chen, Z.; Ricigliano, J.; Klessig, D.F.; Proc. Natl. Acad Sci. USA, 1995,92,7143-7147. Wendehenne, D.; Durner, J.; Chen, Z.; Klessig, D.F.; Phytochem, 1998, 47, 651-657. Bi, Y.M.; Kenton, P.; Mur, L.; Darby, R.; Draper, J.; Plant J., 1995, 8,235-245. Neuenschwander, U.; Vemooij, B.; Friedrich, L.; Uknes, S.; Kessmann, H.; Ryals, J.; Plant J., 1995, 8, 227-233. Durner, J.; Klessig, D.F.; J. Biol. Chem, 1996, 277, 28492-28501. Anderson, M.; Chen, Z.; Klessig, D.F.; Phytochem, 1998, 47, 555-566. Ruffer, M.; Steipe, B.; Zenk, M.H.; FEBSLett, 1995, 377, 175-180. Leslie, C.A.; Romani, R.J.; Plant Physiol, 1988, 88, 833-837. Lawton, K.A.; Potter, S.L.; Uknes, S.; Ryals, J.; Plant Cell, 1994, 6, 581-588. Sanz, A.; Moreno, J.L; Castresana, C; Plant Cell, 1998,10, 1523-1538. Salvemini, D.; Misko, T.P.; Masferrer, J.L.; Seibert, K.; Currie, M.G.; Needleman, P.; Proc. Natl. Acad Sci. USA, 1993, 90, 7240-7244. Farivar, R.S.; Chobanian, A.V.; Brecher, P.; Circ. Res, 1996, 78, 759-768. Farivar, R.S.; Brecher, P.; J. Biol. Chem, 1996, 49, 31585-31592. Hermann, M.; Kaplotis, S.; Hofbauer, R.; Exner, M.; Seelos, C; Hald, I.; Gmetner, B.; FEBS Lett, 1999, 45, 212-214. Simons, B.H.; Millenaar, F.F.; Mulder, L.; Van Loon, L.C.; Lambers, H.; Plant Physiol, 1999, 120, 529-538. Klepper, L.; Pest Biochem. Physiol', 1991, 39, 43-48. Pieterse, C.M.J.; van Loon, L.C.; Trends Plant Sci, 1999, 4, 52-58. Shah, J.; Kachroo, P.; Klessig D.F.; Plant Cell, 1999, 77, 191-206. Delanay, T.P.; Plant Physiol, 1997, 775, 5-12. Ryals, J.A.; Neuenschwander, U.H.; Willits, M.G.; Molina, A.; Steiner, H.Y.; Hunt, M.D.; Plant Cell, 1996, 8, 1809-1819. May, M.J.; Hammond-Kosack, K.E.; Jones, J.D.G.; Plant Physiol, 1996, 110, 1367-1379.

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

965

O-AMINOPHENOL-TYPE TRYPTOPHAN METABOLITES: 3-HYDROXYKYNURENINE, 3-HYDROXYANTHRANILIC ACID, AND THEIR ROLE IN LIVING ORGANISMS ANTONIO RESCIGNO* and ENRICO SANJUST Cattedra di Chimica Biologica, Dipartimento di Scienze Mediche, Universita di Cagliari, Cittadella Universitaria, 09042 Monserrato, CagUari, Italy ABSTRACT: Among tryptophan metabolites, 3-hydroxykynurenine and 3-hydroxyanthranilic acid are unique because of their o-aminophenolic structures. Therefore, they share a typical chemical behaviour and are easily (auto)oxidised to a number of interesting derivatives such as ommatins and phenoxazinones. Moreover, their (auto)oxidation usually proceeds via a radical chain mechanism, in which both semiquinonoid species and intermediates in dioxygen reduction are involved. They can also undergo enzymic oxidation. For these reasons, 3-hydroxykynurenine and 3hydroxyanthranilic acid participate in carcinogenesis, neurotoxicity, delignification by white-rot fungi, eye lens senescence, melanogenesis. They also show pro-oxidant as well as anti-oxidant activities.

INTRODUCTION Tryptophan is among the most interesting aminoacids because its biologic role goes far beyond its participation in proteins structure. In mammals, tryptophan is the precursor in the synthesis of the neurotransmitters tryptamine and serotonine, as well as of other bioactive molecules such as kynurenic and quinolinic acids. Along with the usual catabolic pathway of the aminoacid, other essential compounds are found, such as nicotinic acid and melatonin [1-3]. Among the existing degradative pathways for tryptophan, one is known as the kynurenine route. Fig. (1). The word 'kynurenines' is used to define the intermediates of this route which account for more than 90% of the whole tryptophan metabolism in man [4], and for 95% of dietary tryptophan [5]. So, the intermediates of the kynurenine pathway are quite important quantitatively. Among these, kynurenic and quinolinic acids have been extensively studied as regards their neuroactivities [6-13].

966

/-CHNH2COOH

0

Tryptamine Serotonin

H

Melatonin

Trypt(Dphan

f

r-CHNH2C00H

/ -CHNH2COOH

W TV-Formy 1-ky nurenine

- or

COOH

NH2

Kynure

Anthranilic Acid

CHNH2COOH

OH

COOH

- 9^ NH2

COOH OH Kynurenic Acid

OH

3 -Hydroxy kynurenine

3-Hydroxyanthranilic Acid

OH

96-

COOH

COOH COOH

OH Quinolinic Acid Xanthurenic Acid

Fig. (1). The kynurenine route of tryptophan metabolism

967 The kynurenine pathway starts with the oxidative cleavage of tryptophan. Tryptophan 2,3-dioxygenase and indoleamine 2,3-dioxygenase are both able to catalyse the reaction, whose product, namely formylkynurenine, is in turn hydrolysed by formamidase to yield kynurenine. At this point, a branching takes place. A possible route is the cleavage of kynurenine, catalysed by kynureninase, leading to alanine and anthranilic acid. The existence of an anthranilate hydroxylase has been claimed, and so, a possibility exists, of a catabolic convergency with the alternative branch. This latter involves the intervention of kynurenine aminotransferase, leading to kynurenic acid. A third evenience consists in the intervention of a kynurenine 3-hydroxylase, producing 3-hydroxykynurenine (HK), which is in turn cleaved by kynureninase to alanine and 3hydroxyanthranilic acid (HA). The action of 3-hydroxyanthranilate oxidase (dioxygenase) causes a profound rearrangement leading to quinolinic acid, an important neuroactive metabolite. Finally, quinolinic acid is converted into niacinic acid: the starting point for NAD(P) cofactor synthesis. As already noted, some neuroactive metabolites of tryptophan have been extensively studied on account of this relevant physiological feature. In the present review, attention will be focused on two tryptophan metabolites, HA and HK, aiming to discuss their various metabolic implications as well as their role as the precursors of some interesting pigments. These compounds share the unique feature, among tryptophan metabolites, of being o-aminophenols. This structural feature allows for a number of typical reactions to take place. These reactions are important in physiopathology, in that the o-aminophenol structure is not common in Nature, so both HA and HK can undergo a number of fates, either enzyme-catalysed or not, that are worth discussing below. In particular, HA is the direct precursor of cinnabarinic acid (CA, 2-amino-3//-phenoxazinone-l,9-dicarboxylic acid), obtained through an oxidative dimerisation, which is in turn strictly related to the known antineoplastic and antibacterial agents, actinomycins. CA has been found in fungi, moulds, and insects [14-17]. HK, on the other hand, is the precursor of another class of pigments, namely the xanthommatins (XA), found in various insect body parts, and in moulds [18-20]. XAs have been also found in the human cataract [21-24]. So, the general properties of HA and HK will be examined in detail in the present article. It is also worthwhile to briefly describe some common analytical methods, which have been proposed to determine tryptophan and/or its

968 metabolites, also in biological samples. A general method has been described, devoted to general tryptophan metabolism analysis, and based on high-voltage electrophoretic separation on paper [25]. Anthranilic acid, HA, kynurenic acid, kynurenine, niacinic acid, quinolinic acid, xanthurenic acid and unmetabolised tryptophan WQTQ easily separated and identified. A more specific method has been developed, based on HPLC, for simultaneous determination of HA and CA. The samples were extracted with ethyl acetate, after acidification, and directly passed through a CI8 reversed phase column, using an ion-pair technique. Detection took place either electrochemically or photometrically [26]. ENZYMOLOGY OF HA AND HK For a long time, the major interest in tryptophan metabolism has been focused on the well-known neurotransmitter 5-hydroxytryptamine which, however, accounts for only 1% of dietary tryptophan metabolism in peripheral human tissues, more than 95% being transformed via the kynurenine pathway. Nowadays, a renewed interest arises as far as kynurenine and related substances is concerned, for example as potential targets for pharmacological therapies [5]. Therefore, it seems well worthwhile to carry out a somewhat detailed discussion about some aspects of the kynurenine pathway enzymology. The whole subject has been recently reviewed [5]. Kynurenine 3-hydroxylase The key enzyme of HK production is kynurenine-3-hydroxylase (EC 1.14.13.9), a FAD-containing flavoprotein which uses either NADH or NADPH as an external reductant. With NADPH as the cosubstrate, the Km for kynurenine is slightly lower and the Vmax slightly higher than with NADH. Unfortunately, the enzyme is not stable and is therefore of poor utility in mechanistic studies. It appears to be strictly related to other flavin-containing hydroxylases as regards its action mechanism. The enzyme is inhibited by a-ketoacids, derived from branched chain aaminoacids. In particular, 2-oxo-isocaproate, derived from leucine, has been found to be an efficient competitive inhibitor for both kynurenine and NADPH (!). The strongest known inhibitor is 3-(m-nitrobenzoyl)-

969 alanine, with a Kj as low as 0.9 ^xM. It has been suggested that one of the oxygen atoms of the nitro group of the inhibitor mimics one of the oxygen atoms of the dioxygen molecule, therefore explaining the exceptionally high inhibition efficiency. Kynureninase HK, as well as kynurenine, are substrates for kynureninase, a PLP dependent enzyme, which catalyses a unique reaction in metabolism, the hydrolytic p,y-cleavage of aryl substituted y-keto-a-aminoacids. The natural substrates are kynurenine and HK, which are cleaved to anthranilic acid and HA, respectively, the other cleavage product being aminoacid alanine. The enzyme is widely distributed among animal and bacterial sources, and preparations, coming from different organisms, greatly vary in their preferences between the two substrates. For example, the Pseudomonas enzyme cleaves kynurenine five times faster than HK, while the liver enzyme hydrolyses HK twice as rapidly as kynurenine. Specificity studies have demonstrated that artificial substrates, differing from the natural ones for some substitutions on the benzene ring, are cleaved at reasonable rates. Substitution of the aromatic amino group with a nitro or hydroxy group, or its formylation, give compounds which are still substrates, though very poor ones. A detailed mechanism for this unusual cleavage reaction has been established [27], and it fully respects the rule which says that the bond to be cleaved is held perpendicular to the delocalised 7C-system of the PLP co-enzyme, in order to maximise orbital overlap. The compulsory preliminary intervention of a base within the active site of the enzyme, needed for preventive deprotonation of the substrate a-CH group, is not exclusive to the enzyme, but is shared with all other PLP enzymes. In the case of kynureninase, some evidence exists that this base is not a carboxylate but rather the 8-amino group of a lysine residue. It has not yet established whether the nucleophilic attack on the P-carbon of the substrate is performed directly by a water molecule (possibly, by a hydroxide ion) or alternatively by a nucleophilic residue within the active site. Several inhibitors are known for the enzyme, some of which are strongly competitive with respect to kynurenine and HK, and substrates at the same time. Kynureninase inhibitors could be of pharmacological value, even if they are not physiologically important, as they are all synthetic and non-naturally occurring compounds.

970 Kynurenine aminotransferase Kynurenine aminotransferase (EC 2.6.1.7) is a PLP-dependent enzyme that converts kynurenine to the corresponding a-ketoacid, employing aketoglutarate as an electron acceptor. A rapid cyclisation of the reaction product leads to kynurenic acid. By the same way, HK is converted to xanthurenic acid. In the brain, two forms of kynurenine aminotransferase have been found, somewhat differing as regards substrate specificity, affinity, and inhibition. The mechanism of the enzymic transamination of kynurenine and HK has drawn little attention, even if its irreversible character should be an interesting feature. 3-Hydroxyanthranilic acid oxidase HA, arising from kynureninase action on HK, in mammals undergoes a further enzymic reaction, involving the iron-dependent enzyme, known as 3-hydroxyanthranilic acid oxidase, which is perhaps more appropriately named, 3-hydroxyanthranilic acid 3,4-dioxygenase. It is an important target for drug action, as its inhibition is the most direct way to lower quinolinic acid levels. In fact, quinolinic acid arises from the action of this enzyme, through a complex series of rearrangements. The primary oxidation product is the highly unstable a-amino-P-carboxymuconic-ssemialdehyde. The enzyme requires both Fe^"^ and free -SH groups for activity. The enzyme is subjected to strong inhibition by halogenated derivatives of HA, and interestingly the inhibition potency rises sharply from 4-fluoro- to -chloro- and -bromo-HA. This is because the intermediate product is not the semialdehyde but an acyl halide, relatively inert in the case of fluoride and highly reactive in the case of bromide, showing that the acyl chloride has an intermediate reactivity. Somewhat surprisingly, it has also been shown that the inhibition is both competitive and reversible. Anthranilic Acid Hydroxylase Some confusion exists concerning this enzyme, for at least three different activities have been recognised. An enzymic complex, formed by two proteins, which are catalitically inactive when considered separately, uses

971 molecular oxygen and NADPH to convert anthranilic acid into catechol, with a concomitant release of ammonia and carbon dioxide [28]. It has been isolated from a Pseudomonas and its properties are beyond our brief here. The same can be affirmed for the other enzyme [29], also named anthranilic acid hydroxylase, found in Aspergillus niger and capable of oxygenating and deaminating anthranilic acid to 2,3-dihydroxybenzoic acid, with NADPH as a cosubstrate. The third kind of hydroxylase activity is the only one pertinent to the present review and refers to a mono-oxygenase of the hydroxylase type. This catalyses the insertion of a -OH group in place of a hydrogen atom at the 3-position of anthranilic acid, leading to HA. This evenience is often simply ignored by many authors dealing with tryptophan metabolism, so it could seem that the only possibility of HA formation is the action of kynureninase on HK. The actual occurrence of a true hydroxylase, as defined above, could however be considered debatable for a number of reasons, but some authors merely notice the activity without giving any reference to it. Indirect evidence for the existence of such an enzyme arose from the observation [30] that in vitro administration of anthranilic acid to rat neurons causes an increase in intracellular HA concentration. However, that study did not pay due attention to the earlier observation [31], that anthranilic acid is an effective inhibitor (Kj 40 |iM) of the enzyme, devoted to HA oxidation, 3-hydroxyanthranilate oxidase (dioxygenase). Therefore, the increase in HA concentration could well be due to an inhibitory effect of anthranilic acid towards the further catabolism of HA, and not to a direct conversion of anthranilic acid to HA. One study deals with the putative conversion anthranilic acid - • HA [32], in which the production of radiolabeled NAD and NAD? was observed in rat livers in response to administration of radiolabeled anthranilic acid. In the experiments with the hepatic microsomal system, a mixture of 5-hydroxyanthranilic acid and HA was obtained. However, a more recent study exists [33], also based on the use of isotopically labeled anthranilic acid, which demonstrated that the substance is unable to increase HA levels in the liver and in the urines, when administered to rats; on the contrary, such an increase was observed when either kynurenine or HA were administered, suggesting that the pathway, operating in mammals, should be kynurenine - • HK - • HA, and not kynurenine - • anthranilic acid - • HA. In that study, it was also concluded that the preferred route of kynurenine further metabolism was through

972 hydroxylation rather than cleavage; in other words, only a comparatively small fraction of kynurenine pool led to anthranilic acid, and this latter was not hydroxylated to HA. In conclusion, the conversion of kynurenine to HA via anthranilic acid is still awaiting for a conclusive assessment. CHEMICAL PROPERTIES OF HA General HA (2-amino-3-hydroxy-benzoic acid) forms colourless crystals, slightly soluble (<1 mM) in water or aqueous solutions at physiological pH values. Owing to its amphoteric properties, the substance is readily soluble in both acidic and basic solutions. However, HA is easily autoxidised when dissolved in alkaline media. So relatively concentrated solutions should be prepared with the aid of diluted hydrochloric acid, where it is reasonably stable. HA does not show any light absorption in the visible range, whereas has a A^max of 297 nm (s 2496 M'^cm'^) in the UV region, at pH 2. The substance is fluorescent, with X^x 323 nm, Xem 414 at pH 6.4. (Auto)Oxidation of HA The autoxidation of HA, A in Fig. (2), is slight around neutrality but becomes increasingly fast as the pH increases [34], as observed with other phenolics and aromatic amines. This pH dependence of autoxidation is explained by the consideration that the process takes place preferentially towards uncharged, or rather, anionic species, whereas this is comparatively difficult with respect to cationic species. As observed for other phenols and/or aromatic amines, the autoxidation of HA is extremely complex [35]. The main product of this process, at least under physiological pH values, is CA, G in Fig. (2), but other products arise from autoxidation under alkaline conditions, where CA is unstable towards hydrogen peroxide, an intermediate in dioxygen reduction by HA. As regards the autoxidative process, HA could be regarded as an o-aminophenol, so forming a semiquinonoid radical as the first autoxidation product.

973

COOH

COOH

COOH NH2 OH

NH2

-H+ - e -H+ - e -

O-

^o

COOH •H+ - e -

&

+H+ + e -

:

COOH

X^

COOH

NH

Js^^NH2

A + C

^OH ^^^

"OH

D

COOH [OX]

COOH pj

COOH

. N ^ J:^

COOH

COOH

^NH2

COOH

.R>. J ^

^NH2

F i g . (2). The (auto)oxidation pathway of 3-hydroxyanthranilic cinnabarinic acid

acid, leading to

974 The electron-withdrawing carboxy substituent on the aromatic ring is not able to prevent the extraction of one electron from a system which is as the whole electron-rich. Autoxidation could be summarised as follows:

+

QH2 QH2

+

02

-*•

H202 -*

+ +

Q Q

H2O2 2H2O

(1) (2)

where QH2 is HA and Q is its quinonoid counterpart. This scheme is obviously an extreme simplification, in that the intimate mechanism of autoxidation is much more complex and involves a series of monoelectronic steps. In fact, the direct reaction of dioxygen with an oxidisable substrate QH2, leading to H2O2, is spin-forbidden [36-37]. So, very often, 'autoxidations' are indeed metal-catalysed oxidations, and in these cases the autoxidative pathway for a generic QH2 compound could be depicted as follows: QH2

+

QH» QH2 QH»

+ + +

O2 O2 H02' H02'

-* -^ -^ -*>

QHQ QH» Q

+ + + +

UO2* UOj'

H2O2 H2O2

(3) (4) (5) (6)

where the participation of the metal catalyst is omittted for the reasons of clarity. This general scheme [38] has been assessed as a general one for autoxidation of polyphenols and related compounds, and the quantitative relationship among the various steps widely investigated [39-40]. It is worth noting that superoxide is often an intermediate rather than a product of the autoxidation, being produced in reactions (3) and (4) but consumed in (5) and (6). Therefore, in some cases it has been found that addition of superoxide dismutase (SOD) to the polyphenol sharply inhibits the autoxidation process [40], so confirming the role of an intermediate rather than of a product for superoxide. On the other hand, this consideration does not exclude that superoxide could easily migrate away from the reaction site and participate in other chemical processes; moreover, the superoxide-consuming reactions could be comparatively less important, depending on the particular polyphenol, and then a net superoxide production takes place. Once hydrogen peroxide has been formed, it can compete with molecular oxygen for the oxidation of the substrate, as shown by Liang et al [41] in the case of topamine. This competition could be a general phe-

975 nomenon, if one considers that the O2 concentration within aerobic tissues is as low as 0.03 mM. Hydrogen peroxide could obviously arise also from enzymatic and/or non-enzymatic superoxide dismutation, and from other oxidase-catalysed reactions, so its participation in autoxidations should not be underestimated. This is particularly true in the presence of certain ions of transition metals such as iron and copper, that are almost always present in noticeable amounts in practically all living tissues. It is well known that the reaction between hydrogen peroxide and Fe^^ generates hydroxyl radicals •OH (FENTON reaction): H2O2

+

Fe^^

-•

Fe(III)Orf^

+

-OH

Hydroxyl radicals easily attack a number of substrates [42-43], owing to their very high redox potential (estimated +2.74 V in aqueous solution). Moreover, whenever peroxide and superoxide interact, the HABER-WEISS reaction, producing molecular oxygen and hydroxyl radicals, could take place: H02*

+ H2O2

-•

H2O

+

O2

+

-OH

Note that dioxygen is released in the singlet state [44], so it is able to perform oxidations that 'normal' triplet dioxygen cannot carry out. Even if the reaction speed is comparatively low in the absence of any iron ion as a catalyser [45-48], in the living tissues iron is always present at concentrations well suitable for a very effective catalysis. In conclusion, whenever superoxide arises from one-electron oxidation of any substrate, one could reasonably expect the unavoidable concomitant formation of peroxide, hydroxyl, and singlet dioxygen. The effective enhancement of hydroxyl radical production during HA autoxidation has been shown in the presence of iron ions and SOD, while the contrary was observed in the presence of catalase, which removes the hydrogen peroxide, needed by the FENTON reaction [49]. Despite its general validity for the autoxidation of quinonisable phenolics and aromatic amines, the scheme outlined above has to be carefully checked when it has to be applied to a particular class of autoxidisable substrates. Therefore, autoxidation of o- and /7-aminophenols proceeds in a different manner with respect to that outlined above, although quinoneimines are as expected the first non-radical oxidation products, mAminophenols behave obviously very differently for they cannot directly

976 quinonise upon oxidation, but instead undergo an oxidative dimerisation leading to biphenyl derivatives (autoxidation of m- and ;7-aminophenols is beyond the scope of this review and will not be discussed further). As previously noted, HA behaves as a substituted o-aminophenol, and the overall process has been studied in depth [17, 35, 50-55]. The autoxidative process then starts as in (3), but the superoxide produced is not efficiently consumed by reactions such as (5) and (6) and its spontaneous dismutation is comparatively slow, so it could accumulate to a certain extent. The addition of SOD to the reaction mixture therefore accelerates HA autoxidation 4-fold, probably by preventing back reactions between superoxide and the anthranilyl radical B [34, 56]. This could be a case where the reducing power of superoxide becomes evident [37 and therein]. It is worth noting that anthranilyl radical can react with another oxygen molecule to produce another superoxide [17, 34]. Note also that the tricyclic lactone H, Fig (3) has been isolated from autoxidised solutions of HA, so giving further proof of the intermediacy of anthranilyl radicals B on the autoxidation pathway [35]. As for many other cases, the quinonoid oxidation product C of HA is not stable, as it is a very reactive o-quinoneimine. This could simply undergo a hydrolytic cleavage leading to the corresponding 1,2benzoquinone-3-carboxylic acid, which cannot be isolated, and ammonia. However, a typical feature of o-quinoneimines is their high tendency to undergo a nucleophilic attack by still unreacted o-aminophenol [51, 5455]. The nucleophile is the aminogroup of the unreacted aminophenol, and obviously the reaction is inhibited at acidic pH values where this exists as the non-nucleophilic cationic form. The substituted diphenylamine D obtained is oxidised again (most probably, with superoxide production) leading to a A^-phenylquinoneimine E, which cyclises to a substituted dihydrophenoxazine F. This is finally oxidised (always with concomitant superoxide release) to the phenoxazinone, whose structure depends on the starting o-aminophenol. This tendency to nucleophilic dimerisation is typical of and confined to o-aminophenols (possibly, aromatic o-diamines). Conversely, o-diphenols regularly lead to oquinones, but the latter undergo the nucleophilic attack by nucleophiles other than the unreacted o-diphenol [38, 57-58], so no substituted dibenzodioxines have been described among the usual autoxidation products of o-diphenols. The compulsory semiquinonoid intermediates, formed during the steps D -^ E and F - • G are omitted in the Scheme for the reasons of the clarity.

977

2B

H Fig. (3). The tricyclic lactonic homodimer, arising from oxidative coupling between two anthranilyl radicals

The overall autoxidative pathway for HA is shown in Fig. (2). Note that the further autoxidation of the intermediate D and the successive cycHsation and aromatisation reactions leading to the substituted phenoxazinone CA 7 all show the features already pointed out for the autoxidation of HA to its quinonoid counterpart l-benzoquinone-2-imine-3carboxylic acid C [17]. CA is the main, albeit certainly not the only, product of HA autoxidation under a wide range of experimental conditions [17,34]. At pH 7, significant amounts of another quinonisation product, 2- amino- 5- [(2'- carboxy- 6'- hydroxyphenyl)- amino]- 1,4benzoquinone- 3- carboxylic acid, I in Fig. (4) were found.

978 COOH

COOH

NH O

H2O

HO^

J^

^NH2 OH

[OX]

COOH HO.

X

.NH2

HN"

"^^

"OH

HC^J^^^COOH

COOH

"Ti:

yj [OX]

COOH

HN" HO.

X

^COOH

Fig. (4).The quinonoid compound, arising from 3-hydroxyanthranilic acid (auto)oxidation

979 This highly substituted /^-benzoquinone derivative tends to prevail at high pH values and also at pH 7 when the autoxidation products are checked after several hours [35]. It certainly arises fi"om a nucleophilic attack by OH" on the quinoneimine C. This attack is obviously increasingly easy as the pH arises, and leads to 2-amino-3,6-dihydroxy-benzoic acid, which is in turn oxidised to the corresponding quinone. This is attacked at the 5 position by a molecule of still unreacted HA, and the corresponding adduct is finally oxidised to the stable final product I, Fig. (4). CA is not completely stable under the conditions where it is produced, and its concentration during autoxidation of HA reaches a plateau and then slowly decreases [34]. Concomitantly, an oxygen consumption sharply higher than that stoichiometrically expected is found, supporting the idea that the CA degradation is an oxidative one. In particular, the overall autoxidation of HA to CA has been extensively investigated and reviewed [15-17, 34, 59], and the conclusion that CA is much more sensitive to H2O2 than to 02'* seems well ascertained. However, a detailed study has been reported, in which the specific destructive action of superoxide towards CA (but not I) has been conclusively assessed [35]. Following this idea, SOD enhances CA production because it prevents CA bleaching by superoxide. This hypothesis was supported by a series of experiments in which CA destruction by superoxide was found in the presence of a xanthine/xanthine oxidase system. The destruction rate was not significantly affected by the addition of catalase, thus suggesting that is superoxide but not hydrogen peroxide that is the bleaching agent whenever superoxide is generated. Although excess catalase does not accelerate autoxidation, measured as HA consumption, more CA is produced and accumulated, suggesting that catalase prevents CA bleaching by hydrogen peroxide [50]. This has been considered as another point to support the idea that SOD increases both HA consumption and CA formation by a mass action effect rather than because of CA protection against superoxide. In other words, in the presence of SOD more CA was produced as more HA was autoxidised; after a time, depending on the particular experimental conditions chosen, the CA concentration reached a plateau, and then slowly diminished. This plateau was abolished in the presence of catalase, as was expected given the sensitivity of CA towards H2O2 [16, 35]. Peroxyl radicals ROO* have also been tested as putative oxidising agents for HA [17]. These radicals generated in situ cause a much faster

980 CA production, and this was further enhanced in the presence of SOD. So, it has been concluded that reaction (1) is of comparatively little importance when peroxyl radicals are present, and that the superoxide production arises rather from reactions such as (4). It has been suggested that catalase action towards the overall autoxidation process of HA is not limited to removing H2O2 from the reaction environment [17]. A series of experiments has demonstrated that the catalase compound I, generated in situ by means of the glucose / glucose oxidase system, is capable of carrying out the overall autoxidative process from HA to CA, without any intervention of molecular oxygen. In fact, catalase compound I possesses a well-defined 'peroxidatic' activity towards a number of different oxidisable substrates [60-61]. As expected, horseradish peroxidase (HRP) and myeloperoxidase compounds I are very effective in performing the oxidation of HA to CA [17], and this action does not require per se further comments. In the absence of both catalase and HRP, H2O2 generated in situ as above is without significant influence with respect to HA autoxidation, so the ineffectiveness of hydrogen peroxide as an oxidising agent in the whole process can be argued. On the other hand, the presence of molecular oxygen has no effect on HA oxidation rate by compound I, so showing that the intervention of molecular oxygen during the one-electron oxidative steps promoted by compound I does not take place. By way of concluding, one can note that whereas the high affinity of peroxidase compound I for aromatic, electron-rich substrates as HA renders the intervention of peroxidase in CA formation from HA highly probable, the same is far less true in the case of catalase. In fact, catalase compound I has an extremely high affinity for H2O2, therefore it will presumably not be easily available for exerting any peroxidatic action towards HA and/or other potential one-electron donors. The situation outlined above is subject to dramatic changes in the presence of some transition metal ions. Therefore, as seen for many other phenolics, in the presence of small concentrations of copper or iron ions, a steep increase in the autoxidation rate for HA could be expected. The mechanisms by which copper and iron catalyse the autoxidation of phenolics do not follow the same routes [62], and so, for example in the case of 1,2,4-benzenetriol (a toxic benzene metabolite in the liver), iron(III) was found to be a comparatively inefficient autoxidation catalyst in comparison with copper(II). It is worth noting, however, that in the case of that substrate the addition of SOD shows an inhibitory effect towards autoxidation, evidently because superoxide compulsorily participates as

981 an oxidant in the overall autoxidative process (see reactions (5) and (6)). On the contrary, SOD has an enhancing effect on HA autoxidation, for the reasons outlined above. This is another fact speaking against any careless generalisation arising from the autoxidative pathway outlined in reactions (3)-(4)-(5)-(6). Any particular phenolic examined should be carefully studied to reach sound conclusions with respect its autoxidative mechanism. This observation can be extended to autoxidations in the presence of chelated versus free (aquated) metal ions. It is well known that chelation deeply alters the chemistry of transition metal ions, and these alterations have important consequences on autoxidative pathways, closely depending on the particular phenolic considered [63-65]. Finally, great care should be exerted when examining studies on autoxidations, as traces of transition metal ions have a strong influence on the process whenever they are not accurately removed with for example chelating resins prior to such experiments. Generally speaking, two alternative mechanisms could operate in metal-ion-assisted autoxidations: a) the metal ion in a lower oxidation state gives rise to a one electron transfer to dioxygen: M"^

+ 0 2 - ^

M^"^^^^

+

02--

Alternatively, b) the metal ion in a higher oxidation state could withdraw one electron from the oxidisable substrate: M"^

+

QH2 -^

M^"'^^^

+

QH* +

H^

The relative importance of the mechanisms a) and b) depends on a number of factors: i) the reducing power of the metal ion towards dioxygen, taking into account that the redox potential of the metal ion is only one among various factors determining its reactivity as an one electron reductant for dioxygen; ii) the reducing power of QH2 towards metal ions; iii) the pH, because QH' and a fortiori Q^" are much stronger reductants than QH2; iv) the ability of the metal ion to form a peroxidic adduct with dioxygen; v) the presence of any complexing agent, other than water, capable of altering the behaviour of the metal ion. Taking into account that the obtention of a strictly demetallated reaction environment is a target hard to reach, the presence of catalytic amounts of transition metal ions can only seldom be excluded. This is also true in vivo, where copper and iron ionic species are always present

982 in significant concentrations. So their presence can profoundly alter the 'autoxidative' pathway of several oxidisable substrates, and certainly HA is among those. Therefore, also this factor must be considered when examining the apparently contradictory results obtained when autoxidations are studied in relation to the participation of superoxide as a reduction product of dioxygen [66-67]. It is not easy to draw a generally valid conclusion from the huge number of studies concerning autoxidations of very different substrates, both in the presence or in the absence of transition metal ions, metal ion chelators, SOD, catalase, and/or •OH scavenger. Moreover, any information about the effective presence of transition metal ions in still catalytically active traces, is often omitted. So, very often the reported results are only apparently contradictory, merely reflecting little differences in experimental conditions which, however, lead to dramatic changes in results. At least in the case of 1,2,4-benzenetriol [62] a series of experiments has shown the enhancement of the 'normal' autoxidative pathway (reactions (3)-(6)) by Fe"^^ ions, which act as moderately effective catalysts, whereas Cu^^ ions change the autoxidation mechanism, most probably because cupric ions are more prone to form a chelate complex with o-diphenols (and, noticeably for the scope of this review, with 6^-aminophenols). It is somewhat surprising that there is a lack of detailed works aimed at elucidating the intimate mechanism(s) by which transition metal ions such as cupric and ferric ions interfere with and catalyse HA autoxidation. However, one detailed study has been reported [68] on the action of hydroxyl radicals, generated in situ in a reaction mixture containing HA, hydrogen peroxide and ferric ions. In these conditions, -OH production can only arise when at least a fraction of ferric ions are reduced to the ferrous state, in accordance with the FENTON reaction scheme: Fe^^

+

H202-^

Fe(III)OH

+

-OH

Moreover, •OH production will continue only when an efficient system for recycling ferric ions to the ferrous state is operating. In the system studied, ferrous ions are provided by the following reaction: Fe^^

+

QH2

-•

Fe^^

+

QH^

+

H^

where QH2 is HA and QH^ is an anthranilyl radical. In other words, ferric ions could be able to promote autoxidation of HA in a stoichiometric

983 manner, if present in suitable amounts and bearing mind the stoichiometry for the complete (i.e., leading to CA) HA oxidation. In this context, however, the fate of HA was not under investigation, but rather attention was focused on hydroxyl production. The influence of a number of phenolics, some of which are capable of efficiently chelating iron ions, on the -OH production, was tested. Interestingly, some non-phenolic iron chelators, such as EDTA and diethylenetriamine-A/;A^,A^'A^''A/^''-penta-acetic acid, strongly enhanced •OH production, most probably because they maintain iron ions into the aqueous solutions, and are able to chelate both ferrous and ferric ions, thus ensuring a high water solubility during the redox cycling. The fact that some o-diphenols, particularly caffeic (3,4dihydroxycinnamic) acid, inhibit at least in part the generation of hydroxyl, has been interpreted as a consequence of the strong chelating effect of those diphenols [68]. In reality, the situation is probably more complicated, as if the chelating effect on iron ions is the criterium to assess the negative effect of a phenolic in hydroxyl production, one should expect for example a high activity against •OH production for salicylic acid. This compound, well known as an iron chelator, is, however, ineffective. A closer inspection of the overall hydroxyl generating mechanism, has shown that caffeic acid (and possibly other related catechols) do no directly inhibit the FENTON reaction, and so the function of the compound must be the chelating one, thus preventing Fe^"^ generation at the expense of HA monoelectronic oxidation. On the other hand, 1,4dihydroxybenzene notably enhances •OH production, probably being able to reduce Fe(III) to Fe(II) with high efficiency. Catechols and related aromatics are not the only substance able to inhibit HA autoxidation. It has been reported [69-70] that a faulty tryptophan metabolism is related to malignant bladder tumors, owing also to increased plasmatic levels of HA. Moreover, it has been suggested that HA is not a carcinogen per se, but only after oxidation [70], and it was also found that ascorbic acid is capable of preventing, at least in vitro, the autoxidation of HA [71]. So, a logical consequence of these observations led to in vivo experiments, during which different doses of ascorbic acid were administered to rats implanted with known amounts of HA in their bladders. This study confirmed the hypothesis that also in vivo ascorbic acid could effectively prevent HA autoxidation and therefore tumor insurgence [70]. HA carcinogenicity has been associated with its metabolic conversion to CA. But 5-hydroxyanthranilic acid, which cannot give CA

984 upon oxidation, is also a carcinogen, thus the hypothesis has to be rejected. Other studies have dealt with HA autoxidation in the presence of some transition metal ions such as iron, manganese and/or copper ions [72]. By using the damage extent of DNA both double and single strands in vitro, the authors found that Mn^"^ strongly enhanced DNA damage when HA plus Cu^"^ ions were present. This was related to the known stimulating effect of manganese ions on HA autoxidation [16,34] through is SOD-like activity. The addition of catalase abolished the DNA damage, suggesting the compulsory participation of H2O2 to the process. The same effect was seen when bathocuproine was present, indicating the existence of a redox cycling Cu(II)/Cu(I) as a responsible for the damage. On the contrary, when Fe(III) was substituted for Cu(II) in the reaction mixture, no enhancement of DNA damage was observed, confirming the different mechanisms by which Fe and Cu ions interact with both autoxidisable phenolics (HA in this case) and dioxygen and dioxygen-derived species. A closer inspection of the features of the DNA damage effect in the presence of HA, has led to the formulation of a redox cycle by which manganese ions destroy superoxide:

Oi* Oi'*

+ +

in" + Mn^"

Mn'"

- » •

-•

H2O2 O2

+ +

Mn^" Mn^"

Destruction of superoxide is also operating when a moderately alkaline solution of HA is exposed to the air in the presence of either Mn^"^ or Mn^"^: a strong acceleration in autoxidation is observed, because of the already noted positive influence of superoxide subtraction on the overall equilibrium of autoxidation [34]. Somewhat unexpectedly, it has been also reported [34] that a strong oxidiser such as Mn "^ is unable to oxidise HA in the absence of molecular oxygen. The same was verified for Fe(III)EDTA. The lack of any dismutase-like activity for the couple Fe(II)/Fe(III) indicates that another mechanism is involved when superoxide and iron ions interact: in this case, the HABER-WEISS reaction is a highly probable evenience. However, the inactivity of iron ions towards DNA damage mediated by HA confirms that such a reaction does not take actually place under the conditions specified above. In other words, the redox couple Fe(II)/Fe(III) is unable to catalyse superoxide dismutation. As a special point of interest, it could be noted that 5-hydroxyanthranilic acid, a minor tryptophan metabolite, is capable of causing DNA damage

985 in the same manner described for HA in the presence of copper(II), although it cannot produce CA. This is an indirect proof which argues against the involvement of CA in HA carcinogenity [72]. Hemoglobin / HA interactions In recent years, several biochemists have proposed a novel view of metabolic biochemistry, according to which, at least in some cases, it is possible to assign functions, other than those traditionally considered for a given biomolecule. Among these cases, hemoglobin seems without doubt to be a promising candidate owing to its wide diffusion among vertebrates (and also in other organisms). The unique chemical features of its ferroheme moiety, capable of reversible bonding with dioxygen but also oxidisable to the corresponding ferriheme under certain physiological and pathological conditions, has stimulated a number of studies focused on the putative participation of hemoglobin in (bio)chemical processes, other than dioxygen transport. Therefore, a possible 'enzymic' role of hemoglobin, studied as an 'oxidase' for drugs, xenobiotics and other molecules of physiological as well as pathological relevance, has been claimed by various researchers. In particular, the induction of methemoglobinemia by o-aminophenol is well known [73], and this feature is shared not only by substituted o-aminophenols, but also by a number of different compounds, such as aromatic amines like aniline. Since 3-HA can be regarded as a carboxy-substituted o-aminophenol, and is moreover a tryptophan metabolite usually found both in healthy individuals as well as in some pathological states [74-75], it is not surprising that a series of studies have dealt with (bio)chemical interactions between HA and hemoglobin. The interaction of autoxidising HA and hemoglobin has been checked, and a positive influence on the autoxidation rate has been opined [76]. In particular, HA regularly gave rise to CA, whereas at the same time oxyhemoglobin was oxidised to ferrihemoglobin. A redox cycle has been proposed, in which hemoglobin shuttles between its ferro- and ferriforms. Accordingly, CA formation was inhibited, when CO was added to the reaction medium containing ferrohemoglobin, and conversely also when cyanide was added to the same mixture, but containing ferrihemoglobin. A further study [77] demonstrated that only HA and HK show a marked redox interaction with hemoglobin, and this has been related to the o-aminophenol structure of these compounds. The suppression of the

986 redox cycle involving hemoglobin, when catalase was added to the reaction mixture, was also noted. Deoxyhemoglobin was not capable of participating in any redox cycle, involving HA or HK. The interactions between hemoglobin and the unsubstituted o-aminophenol have been also studied [78], fully confirming the results already obtained on HA and HK. As expected, 2amino-3-phenoxazinone was the product of o-aminophenol autoxidation. Again, ferrohemoglobin action was nearly blocked by CO, and conversely ferrihemoglobin action was totally suppressed by CN" and N3". A fairly similar work [79] was conducted, in which 2-amino-5methylphenol was autoxidised in the presence of both ferro- and ferrihemoglobin, and also in the presence of intact erythrocytes. The most interesting finding of this work is that, the tendency of o-aminophenols to oxidative dimerisation to substituted phenoxazines is so strong, that 2amino-4,4a-dihydro-4a,7-dimethyl-3//-phenoxazin-3-one was quickly formed. This is in spite of the sacrifice of a substantial fraction of resonance energy, because the presence of the methyl substituent at the bridgehead carbon 4a hinders the completion of the autoxidation reaction otherwise leading to the 2-amino-3-phenoxazinone nucleus. However, the experimental data reported are not fully convincing, also because they sometimes lack a control experiment series; therefore, the topic has been thoroughly re-examined [34]. Neither oxyhemoglobin nor ferrihemoglobin were found to be able to increase HA autoxidation. However, CA production increased in the presence of both hemoglobin and SOD, in a manner comparable to that observed in the presence of SOD plus catalase. Also the plateau in CA concentration was abolished by the combination SOD / hemoglobin. UV / Vis spectroscopic analysis showed that there was a ferrihemoglobin / oxyhemoglobin interconversion, but within one hour protein deterioration began to be evident, while approximately one half of oxyhemoglobin was converted into ferrihemoglobin. The conversion rate increased when HA autoxidation was enhanced by added SOD. The interconversion between ferro- and ferrihemoglobin during HA autoxidation was interpreted as being due to the presence of a one-electron interaction between both the forms of hemoglobin and one-electron donors and acceptors such as HA, anthranilyl radical, quinoneimine and superoxide. It was concluded that neither HA autoxidation nor CA formation are substantially affected by either oxyhemoglobin or ferrihemoglobin, even if a chemical interaction between autoxidising HA and the protein was evident. The apparent influence of both oxy- and ferri-hemoglobin is best explained by noting their ability in

987 scavenging hydrogen peroxide. This is indirectly confirmed by the observation that a mixture of either oxy- or ferri-hemoglobin and SOD behaves almost identically as the mixture catalase / SOD during the overall autoxidation process. Obviously, these conclusions do not contradict the long-time-known and well-ascertained finding that many oxidisable substances under certain conditions actively promote the oxidation of oxyhemoglobin to the ferri- form. A highly conceivable explanation has been given [80], that considers the following reaction: HbFe(II)02

+

QH2 - •

HbFe(III)^ +

QH* +

H02"

In this reaction, QH2 represents a readily oxidisable (or rather, quinonisable) substrate, whose apparently paradoxical effect is that of a reductant producing oxidised hemoglobin. HO2' represents the monoanionic form of hydrogen peroxide, which is efficiently scavenged by both oxy- and ferri-hemoglobin [34]. The 'final' phenoxazinones are no longer destroyed, so the net effect of increased autoxidation is achieved. Autoxidation of 3-hydroxy-kynurenine HK, J in Fig. (5), is the immediate catabolic precursor of HA, so this can be reasonably discussed first. However, its structure is more complicated, and consequently its behaviour more versatile as it regards different modes of degradation and/or transformation. Therefore, its interactions with the living matter from a (bio)chemical point of view will be dealt here, only simplifying the discussion whenever the similarity with HA behaviour could render further insight superfluous. In fact, the COCH2CHNH2COOH moiety of HK, substituted for -COOH in HA, does not significantly alter the autoxidative properties of the former, so any discussion about features and mechanism(s) of HK autoxidation is unnecessary. On the other hand, that r-oxo-3'-amino-3'-carboxypropyl substituent actively participates in the condensation reactions of the semiquinoneimine K, so no CA arises from HK autoxidation, but more complex compounds, whose (bio)chemistry is worth summarising, owing to the wide diffusion of such condensation and oxidation products in living organisms.

988 CHNH2COOH

CHNH2COOH .0

NH2

K

COOH

R = - COCH2CHNH2COOH

COOH

COOH

M

Fig. (5). 3-Hydroxykynurenine (J), the corresponding quinoneimine (K), xanthommatin (L), dihydroxanthommatin (M), dihydroxanthommatin semiquinonoid radical (N)

989 Isolated HK is a crystalline solid, with no well-defined melting point, and slightly yellowish (kmax 228, 267 and 368 nm, 8358 3650 M'^cm"^ [81] or 4050 M'^cm"^ [82]). It is not very soluble in water, but solubility is good enough to prepare solutions, suitable for chemical as well as biochemical experiments. In insects, kynureninase is much less important than in mammals, so nearly no HA arises fi-om tryptophan metabolism, but rather HK which is (auto)oxidised to ommochromes. The ommochrome pathway could be a defensive strategy against tryptophan toxicity, developed by organisms that have lost the ability to completely degrade this aminoacid. It has to be noted that no cyclisation products, comparable with dihydroxyindole derivatives obtained from DOPA autoxidation, arise form HK (auto)oxidation [83]. On the contrary, also from HK the phenoxazinone nucleus arises [84], even if obviously CA is not the 'final' product. A tetracyclic system, that of XA, is the main product of the whole process. The entire set of compounds is depicted in Fig. (5). Also note that the one-electron consecutive steps necessary for the 'complete' HK oxidation could be performed by hexacyanoferrate(III) (ferricyanide) as a one-electron oxidising agent [84]. One could reasonably expect that (auto)oxidation of HK follows the same criteria already indicated for HA autoxidation as it regards superoxide and/or peroxide formation and intervention in the whole process. A typical feature of HK (auto)oxidation is as already noted the formation of a tetracyclic nucleus, namely xanthommatin, XA, L in Fig. (5). This highly resonance-stabilised heterocycle is often formed together with its dihydro counterpart, and a stable dihydroxanthommatin-derived semiquinonoid radical HXA», N, with an oxidation state just between the two forms has been detected with the aid of e.p.r. techniques [83, 85]. The two products, XA (a yellow pigment) and dihydroxanthommatin (H2XA, a red pigment, M in Fig. (5)), appear to be a redox couple and are easily interchangeable by means of a number of oxidising or reducing agents, thus resembling a typical quinol / pquinone as well as a catechol / o-quinone couple. Therefore, XA, derived from HK autoxidation or oxidation by ferricyanide [83-86], is readily reduced to its dihydro counterpart by metabisulfite [19-20, 87-88]. The dihydro form easily autoxidises in neutral or, rather, in alkaline solutions, and is also a good substrate for HRP and hydrogen peroxide [85]. In any case, XA is the oxidation product of H2XA. When H2XA is oxidised with H2O2 in the presence of HRP, the semiquinonoid radical already cited is formed, and is still detectable after 5 hours of incubation. A number of

990 different mesomeric formulas can be drawn for HXA», which show a highly electronic delocalisation and a somewhat 'abnormal' chemical inertness. However, a stoichiometric excess of H2O2 over H2XA merely leads to XA, without any radical accumulation. The radical is probably too unstable towards oxidation by hydrogen peroxide to allow it being accumulated to a detectable extent under the conditions described. Besides HRP, ferrihemoglobin and also hematin are able to promote the formation of HXA» from H2XA [85]. In spite of being readily (auto)oxidised, H2XA is a well-defined chemical species, capable of existing for a long time especially when its dihydro structure is stabilised by means of proper substitutions in the -OH group which is oxidised to = 0 leading to the phenoxazinone functionality typical of XA. So, a p-Dglucoside and a 0-sulfate have been described as rhodommatin and ommatin D, respectively. Xanthommatin is not the only 'final' product of HK (auto)oxidation. Therefore, other pigments, widespread among organisms, especially arthropods, have been found and at least in part characterised [89]. They are usually comprised within the general definition of 'ommochromes', which underlines their importance as pigmented substances found in arthropod and especially insect eyes. Among 'ommochromes', ommatins such as XA and H2XA have been characterised, but no exhaustive characterisation has ever been done for other HK-derived eye pigments, namely ommins and ommidins. These latter seem to be essentially polymeric or at least oligomeric species, whose chemical characterisation is quite difficult. They constantly contain sulfur, and ommidins differ form ommins for having relatively lower molecular weights. Generally speaking, ommochromes have a special place in insect biochemistry as there is convincing evidence that tryptophan catabolism is channelled almost exclusively towards them, with the virtual exclusion of all alternative pathways. In insects, ommochromes are not confined to eyes: they have been found also in cuticle, wings, internal organs and eggs. Ommochromes have often been referred to as 'terminal' tryptophan metabolites, and accordingly they have been found in insect excretions. Also intermediates in the kynurenine pathway of tryptophan metabolism, such as just HK, have been found in various insect body districts, and for example in the butterfly Papilio, tryptophan, kynurenine and HK have been identified from the tissues with kynurenine being excreted in significant quantities, particularly at adult emergence. HK has been also found in malpighian tubules of Drosophila, and under certain conditions it is stored in large

991 amounts, prior to its conversion to ommatins and ommins. HK is contained in the pupae of the silkworm Bombyx mori, as the corresponding glucoside. Some caution must be, however, exercised when interpreting analytical data related to tryptophan metabolite content in insect excreta. Therefore, 8-hydroxyquinaldic acid no longer appeared in the excreta of the cockroach Periplaneta americana, fed with tryptophan and treated with neomycin sulfate to destroy its intestinal bacterial flora. Both enzymic and non-enzymic reactions could take place soon after excretion and significantly alter excreta composition before analysis. In the case of Periplaneta, a microbial transformation of the 'final' catabolite xanthurenic acid to 8-hydroxyquinaldic acid takes place. It is somewhat surprising that, despite the wide occurrence of HKderived pigments in the living world, only a relatively few works deal with the intimate mechanism of HK (auto)oxidation. However, the available information is sufficient to delineate a pathway for autoxidation, even if not all its steps have been studied as regards superoxide and/or peroxide and/or hydroxyl involvement in this process. Also, transition metal ion and/or enzyme participation is still awaiting a thorough assessment. However, a detailed study has been reported [90] in which the catalytic properties of ferrihemoglobin in the HK oxidation by H2O2 were investigated. It was found that neither H2O2 nor ferrihemoglobin alone was capable of oxidising HK, whereas when they are present together, a very fast and complete oxidation to xanthommatin takes place. In the reaction, ferrihemoglobin substantially behaves as a peroxidase: HbFe(III)

+

H202-^

Hb-''Fe(IV)=0

+

H2O

The ferryl derivative of hemoglobin then oxidises HK, finally returning to the ferri- state. •OH scavengers such as DMSO and ethanol had no effect on the reaction, while desferrioxamine substantially inhibited it, and EDTA was without action. These results suggest that hydroxyl radicals are not involved in the ferrihemoglobin-mediated oxidation of HK by H2O2, but that 'free' iron ions, arising from a slow hemoglobin decomposition by excess H2O2, could prime -OH generation. One important difference between HA and HK autoxidation must be underlined: H2XA is readily (auto)oxidised, but the process produces a surprisingly stable semiquinonoid radical HXA», and only in the presence of an excess oxidant does that radical finally give the ommatin XA. These

992 considerations could suggest that HK toxicity should be essentially due to dioxygen reduction intermediates such as superoxide, peroxide and hydroxyl, as HXA* and XA are too stable to exert significant damage. On the other hand, in the case of HA (auto)oxidation, the participation of semiquinonoid species, intermediates in CA formation, cannot be ruled out, in that the phenoxazinone nucleus alone is not capable of stabilising radical species at an extent comparable with the tetracyclic nucleus of ommatins like xanthommatin. Some studies [77, 91-92] confirm the idea that reactive intermediates in dioxygen reduction are responsible for HK (neuro)toxicity. The observation that vitamin B6 deficiency substantially impairs HK catabolism, so leading to massive accumulation of the substance into CNS, where it exerts its convulsant effects, drew attention to the chemical mechanism of this toxic effect. Since the in vitro toxicity of HK is lowered by antioxidant treatments and is abolished when catalase is added to the incubation medium, it was suggested that hydrogen peroxide plays a key role in HK toxicity [91]. The actual source of hydrogen peroxide in the presence of HK does not need to be identified, as H2O2 can readily cross the cellular membranes. It could arise from oxidase activities, not pertaining to tryptophan metabolisms, or can also derive from (auto)oxidation of the same HK, possibly in the presence of transition metal ions. In the presence of iron ions, -OH production also must be taken into account. The participation of hydrogen peroxide, regardless to its origin, in the production of a toxic quinoneimine by HK oxidation, cannot be at first ruled out. The experiments indicate that the intracellular pool of hydrogen peroxide is strongly linked to HK-mediated neurotoxicity, although it is impossible to discriminate between intracellularly produced H2O2 and that entering the cell from the outer medium. The addition of catalase to the HK-treated cells abolished toxicity, but this observation does not allow for discriminating between a 'direct' H2O2 toxicity and the formation of the toxic quinoneimine derived from HK. Since experimentally increased peroxidase levels within the cells are linked with a toxicity decrease, it can reasonably be concluded that H2O2 exerts a 'direct' toxic effect rather than producing a toxic quinoneimine by acting on HK. Moreover, the addition of desferrioxamine led to a lowered toxicity, suggesting that hydrogen peroxide, obtained as the main product of dioxygen reduction during (auto)oxidation of HK, is not the true toxic agent, but (neuro)toxicity is attributable rather to the in situ generation of hydroxyl radicals during HK (auto)oxidation, in the presence of iron ions. However, even if the above experimental design appears to be well-

993 conducted and highly convincing, the involvement of another radical species, most probably the semiquinonoid compound derived from the one-electron oxidation of HK (a 'S-hydroxykynureninyl' radical, analogous to the 3-hydroxyanthranilyl radical involved in HA autoxidation) cannot be ruled out. The neurotoxic activity of HK seems to be well ascertained [93-95]. In fact, the substance showed neurotoxic properties when added to neuronal cell cultures [91] and provoked seizures when injected intracerebroventricularly [93]. As a significant and substantial increase of HK concentrations in brain samples taken post mortem from HUNTINGTON'S disease patients was observed, a metabolic disorder of tryptophan degradation was associated with the disease. This disorder seems to be relatively specific to HUNTINGTON'S disease, and normal levels of HK were found under the same conditions in ALZHEIMER'S disease patients. The entire topic of HK neurotoxicity has been very recently reexamined and reviewed in the light of several new experimental findings [96]. Some important conclusions have been drawn form this latest study: a) HK is a potent neurotoxin [97] with respect to striatal neuronal cells. This neurotoxicity is still perceived at concentrations as low as 1 |iM. It has been also conclusively ascertained that the neuotoxin is effectively HK itself and not one or more of its catabolites. Among these, only HA shows a neurotoxic effect, but much less potent than that of HK. b) HK neurotoxicity is mediated by oxidative stress [97]. The generation of reactive species arising from oxygen partial reduction appears to be crucial on HK neurotoxicity. This toxicity was partially or totally abolished in a dose-dependent manner when selected anti-oxidants were added to the incubation media. So, a-tocopherol, its water soluble analogue trolox, ascorbic acid, A^-acetyl-cysteine, and the spin-trap TEMPO, all were effective in completely preventing HK neurotoxicity. When HK concentration was raised to 100 |LIM, only A^-acetyl-cysteine was still able to give some protection. However, such high concentrations of HA within the brain are not likely, so they are not significant for brain pathology [95]. c) Cell death, caused by HK, is apoptosis and not necrosis. A careful histological investigation showed that HK causes neuronal apoptosis rather than necrosis, a feature considered as typical of cells, killed by reactive intermediates of dioxygen reduction [98-101]. d) HK neurotoxicity is highly selective with respect to brain regions, so cortical and striatal neurons were most vulnerable. Hippocampal neurons were less sensitive but not unaffected by the neurotoxin, whereas cerebellar neurons were

994 found to be completely resistant to HK. It seems obvious to compare the vulnerability of neurons to HK with their vulnerability to oxidative stress. However, the direct administration of hydrogen peroxide to neurons coming from the four brain regions showed no selectivity as regards the toxic effects; moreover, HA, whose neurotoxic action could conceivably be quite similar to that of HK, showed no selectivity in its neurotoxicity, e) HK toxicity is dependent on its cellular uptake. There is evidence that HK is transported into the cells by aminoacid carriers that recognise bulky neutral aminoacids such as tryptophan, phenylalanine and so on [102]. That the difference in neurotoxicity towards different types of neurons depends on the entity of HK uptake has been shown by competitively inhibiting HK uptake. As expected, neuronal death induced by 10 |LIM HK was completely abolished by co-application of tryptophan or phenylalanine 10 mM in the incubation medium, leucine is also effective but to a lesser extent. Acidic or basic aminoacids were ineffective in preventing HK uptake and therefore HK toxicity. Specific inhibitors of aminoacid carriers, both Na'^-dependent and -independent, namely 2-aminobicycloheptane-2-carboxylic acid and 2-amino-isobutyric acid respectively, were tested to assess their ability of preventing HK toxicity. Results showed that only the Na^-dependent aminoacid carrier is responsible for HK uptake by cells. However, the inhibitor cannot prevent HK toxicity when the toxin concentration was as high as 100 |iM, suggesting that a carrierindependent uptake, or rather, a passive permeation of HK takes place at higher concentrations. The triptophan uptake ability of neuronal cells of different regional origin as indicated above, was proportional with the different vulnerability of the cells to the toxin, so confirming the idea that neurotoxicity was mediated by toxin uptake. Only two tryptophan metabolites, namely HK and HA, show a significant (neuro)toxicity, and they share the o-aminophenol structure which is a structural requisite for easy autoxidation with a concomitant release of superoxide and peroxide. Generation of both superoxide and peroxide in HK autoxidation has been shown earlier [72, 85] whereas the generation of these species during HA autoxidation has been discussed above. The involvement of these intermediates of dioxygen reduction is strongly supported by the observation that antioxidants suppress the toxic effects exerted by HK. In particular, the participation of ^OH radicals is suggested by the fact that the well-known hydroxyl radical scavenger, Nacetyl-cysteine, is the most effective preventer of neurototoxicity, while superoxide seems to play a secondary role, being unaffected by A^-acetyl-

995 cysteine [103]. •OH radicals can well arise from an intracellular FENTONlike reaction. However, a very important question remains unanswered. Whenever the toxicity of any readily oxidisable substance, leading to reactive intermediates of dioxygen reduction, is ascribed just to these intermediates, namely superoxide, peroxide and hydroxyl, why should one admit a significant accumulation of such species in aerobic cells, usually well rich in both SOD and catalase? Moreover, the intracellular average concentration of molecular oxygen is around 0.03 mM, in other words, surprisingly low: only a fraction of the total dissolved dioxygen will be reasonably available for superoxide, peroxide, and/or hydroxyl generation. Therefore, even if in some cases the 'final' quinonoid product(s) of autoxidation are relatively unreactive (this is precisely the case of HK, producing the chemically inert xanthommatin), the contrary is true for the semiquinonoid counterparts. So, even if this semiquinonoid radical is - as already underlined - exceptionally stable, the situation is very different for the highly reactive semiquinone arising from one-electron oxidation of HK, which cannot be isolated and whose substantial participation in (neuro)toxicity signs cannot be underestimated. COVALENT INTERACTIONS WITH AMINOACIDS, PEPTIDES AND PROTEINS Since many years now the possible covalent interactions between HA and aminoacyl residues of insect proteins has been well documented [104105]. Later, Manthey and colleagues [106] elucidated some aspects of this topic, and found that during HA (auto)oxidation at alkaline pH, significant amounts of 2-amino-l,4-benzoquinone-3-carboxylic acid are produced. The 5-position of the latter is highly reactive towards nucleophiles such as amines and thiols, so its covalent coupling with lysine- and cysteine-containing proteins appears to be highly possible. Fig. (6). As expected, at pH around 11, free lysine reacts through its more nucleophilic 8-amino group, also when the a one is available. The formation of an adduct with lysine has been also observed when HA is (auto)oxidised in the presence of polylysine. Also free glycine, alanine and proline can form covalent adducts with HA, so any free -NH2 group can in principle form such adducts. However, proline forms a particular adduct, whose

996

importance is very slight, unless high concentrations of the free aminoacid are present during HA autoxidation. COOH 0

NH2

V-^0 R - SH

R-NH2

RNH

O^

[OX]

[OX]

COOH

COOH

Js.

,NH2

RNH"

Fig. (6). Oxidative coupling of lysine and cysteine to 3-hydroxyanthranilic acid

NH2

997 As it concerns lysine residues involvement in protein covalent modification (tanning) by (auto)oxidised HA, a w^ide range of proteins were examined [107]. Chemical modification experiments w^ith bovine serum albumine suggested that lysine residues WQVQ the main sites of reaction, and lysyl-p-quinone adducts were the products of tanning. A synthetic polypeptide (lys, glu)n was tanned at a substantially reduced rate, suggesting that electrostatic repulsive forces, exerted by carboxylate moieties, could be responsible for the inertness of lysozyme to HA-induced tanning. In principle any protein, carrying available lysine residues, is susceptible to tanning by HA. However, aminoacid analysis of HA-tanned proteins revealed that only a small percentage of available lysyl residues are actually engaged in covalent modification. This is in agreement with the observation that in human senile cataracta the lysine content of lens proteins is substantially unchanged even when tanning becomes evident [107, 108]. In another study [109], after incubation of HA with lens proteins, a brown colour rapidly arose, with a concomitant -SH depletion. These modifications are similar to those observed in senile cataracta, and it is important to note that lens proteins are not subjected to any turnover, so the damage will be permanent. Coloured products are also observed concomitant to tanning in senile cataracts and are not exclusively the above mentioned quinonoidal adducts of lysine residues. It has been suggested that the observed yellow products are mainly made of XA and/or other similar ommatins, arising form HK (auto)oxidation. Metabolism of HK within the eye lens is not confined to (auto)oxidation: for many years it has been well known [81] that a yellowish substance is physiologically present in the human eye lens, whereas in lenses of other diurnal mammals such as squirrels it is replaced by a somewhat analogous compound, A^-acetyl-HK [110-112]. The substance from humans has been unequivocally identified by a number of analytical methods as being HK-0-|3-D-glucoside HKG [81]. Both the glucoside and the acetyl derivative chemically differ from the parent compound for being much less susceptible to autoxidation, and have been suggested as possessing anti-oxidant properties. An extensive study has shown that the concentration of the glucoside in the normal human eye lens steadily decreases from birth to 30 - 40, after which it tends to remain substantially unchanged. The presence of some tryptophan metabolites in human eye lenses has been known for many years [109, 113], but what role, protective or destructive, they may have on the physiology and pathology of the lens, is far from being completely clear. Among the

998 putative functions, proposed for HKG, the most intriguing is that of being a harmless UV-A filter [114]. This would be in agreement with the modifications observed in human lens with aging, concomitant with the already noted decrease in HKG concentration. However, the actual filtering ability of eye lenses increases with age, this increase being due to unknown substances, different from HKG. These unknown yellow substances, which are most probably derived from HK alteration, are also good UV-A screens, but time-resolved spectroscopic studies suggest that they are able to prime (photo)chemical reactions, potentially harmful to the lens. The chemical nature of the pigment present in brunescent cataractous lenses was suspected to be some tryptophan degradation product, until it was identified as XA by spectrophotometric comparison with an authentic sample of XA [24]. This was shown by means of particular spectrophotometric techniques and by comparison with authentic samples of XA. The direct involvement of kynurenine pathway enzymes is not necessary to explain XA formation within lenses, taking into account the finding that tryptophan solution, irradiated with UV light for long periods, are changed to an ommochrome-like pigment [115]. It has also been shown that tryptophan contents of proteins in cataractous lenses is significantly decreased when compared with normal lenses, and the aminoacid is changed mainly to A^'-formyl-kynurenine and to kynurenine [22]. A further browning of insoluble proteins takes place, when they are incubated for several hours with HK and irradiated with UV light [24]. The whole topic of the presence and physiopathology of HK and its derivatives within the eye lens and its interactions with lens proteins have been recently reviewed and re-examined [116]. The structure of eye lens undergoes modifications with aging, with the proteins tending to form aggregates, and therefore with an increase in the insoluble protein fraction. This causes an increase of light scattering [117], which is a typical feature of cataractous pathology. The human lens contains fluorescent, lowmolecular weight compounds, that absorb UV light and reduce chromatic aberration [110]. Two of these molecules, arising from tryptophan catabolism, namely HK and HKG, have been extensively studied [113, 118120]. HK has been shown to react with lens proteins, resulting in a tanned material. Moreover, it has been shown that also free glycine, as well as calf crystallins, in vitro react with HK, leading to fluorescent products [121-122]. Prolonged UV irradiation of the lens can lead to HKG photochemical depletion and yellowing of the lens [123]. Insurgence of cataract

999 has also been ascribed to XA, arising from oxidative dimerisation of HK [124], or to oxidised xanthurenic acid [125]. It has been proposed that HK as well as HKG, owing to their anti-oxidant activity [126-128] together with their near-UV screening properties in the range 295 - 400 nm [114], exert a well-defined benign role in lens protection against both UVinduced and oxidative- stress-linked damages. However, a prolonged exposure to UV light and oxidising conditions could lead to (photo)oxidation of both HK and HKG, resulting in tanning of lens proteins which become at first yellow and with time, owing to increased light absorption, and then evolve towards cataractous state. The hypothesis has been checked under different conditions as regards pH, oxygen concentration, and light intensity [116]. At neutral pH, in the presence of air, HK is readily autoxidised, leading to a complex mixture of coloured products, absorbing in the range 400 - 500 nm. Also covalent modification of crystallins takes place, involving predominantly amino groups [118]. It has been suggested that HK, besides being an intermediate in the UV screen compound HKG, hsis per se anti-oxidant activity and can therefore act as a protective agent within eye lens [112,127]. Contrary to this, a prooxidant activity of HK has been claimed [124], and a photodynamic reaction of HK, leading to various pigments and in particular XA, has been proposed. However, the formation of XA in both healthy and cataractous lenses has not been assessed, as the claims of the presence of XA within lenses are based on indirect evidence. All these observations suggest that not HK, but rather its oxidation products, are responsible for lens discoloration and/or damage. As expected, it was found that HK autoxidation becomes faster at higher pH, but surprisingly the extent of protein tannning remains substantially unchanged. It is worth noting that absorbance measurements do not necessarily allow for a simple correlation with tanning extent, as the products of HK autoxidation and covalent coupling to proteins vary considerably in their extinction coefficients. The development of a blue/green fluorescence in aged human lenses [129-130] is well known. This rise in fluorescence was reproduced in vitro by incubating HK with crystallins under aerobic conditions [116]. Subsequent protein digestion and aminoacid analysis confirmed the covalent modification. The extent and pattern of protein modification of these experiments are quite similar to those obtained with cataractous lenses. A photo-oxidative mechanism operating in age-related cataract development [123, 131-134] is therefore confirmed [116]. Further ex-

1000 periments have allowed for the assessment of oxygen's influence on crystallin modification with HK. It has been conclusively shown [116] that the presence of molecular oxygen is a conditio sine qua non for protein modification, and that UV irradiation enhances this process. It has been suggested [135] that superoxide may be involved in the chemical modification of lens proteins by HK and/or its (auto)oxidation products. However, also in the virtual absence of molecular oxygen a very slow protein modification takes place, suggesting that aged lenses are modified by HK under the nearly anoxic conditions of the lens centre. Further experiments showed that GSH is capable of preventing oxidative coupling between (auto)oxidised HK and lens proteins [116]. This observation, taken together with the known decrease of GSH concentration within lens with age [136-137 ], leads to the reasonable conclusion that GSH may prevent lens damage by HK and/or (auto)oxidised derivatives. The mode of GSH action is mainly an adverse effect on HK (auto)oxidation, also confirmed by the increased rate of GSH oxidation in the presence of autoxidising HK. Reaction between tyrosine residues and (auto)oxidising HA leads to benzocoumarin derivatives [138]. These benzocoumarins arise also when in vitro incubation of HA and tyrosine-containing proteins takes place around pH 7 and in the presence of air [139]. The mechanism of HA/TYR oxidative condensation is still partly unclear. Given the radical coupling mechanism between tyrosyl radical and 3-hydroxyanthranilyl radical, one could speculate whether tyrosyl radicals arise from a redox reaction between tyrosine and HA-derived quinoneimine, or alternatively, tyrosyl radicals could be the product of superoxide (generated during HA autoxidation) on tyrosine residue. The modified tyrosine residues, showing the above mentioned benzocoumarin structure, are acid-stable, and therefore can be isolated after total acid hydrolysis of the modified protein. Their formation is also of some biological importance, as they have been isolated from the pupal cocoons of different butterflies [104-139]. A systematic exploration of the reactivity of different aminoacid residues with (auto)oxidising HA was carried out also with the aid of homopolyaminoacids. Among these, only polylysine and polytyrosine were found to become coloured after incubation in the presence of air with HA in a neutral buffered solution. A new absorption band, centred at 340 nm, appeared in tanned polytyrosine, and it can be used as a measure of the tanning extent. A benzocoumarin arising from radicalic oxidative dimerisation of HA is well known as one of the major products of HA

1001 (auto)oxidation [35]. Therefore, the hypothesis of a possible crossdimerisation between anthranilyl radical and tyrosyl radical has been checked and fully confirmed by isolating and characterising the putative benzocoumarin expected as the product of the reaction. Interestingly, only the homodimer H, Fig. (3) and the cross-dimer between anthranilyl and tyrosyl radicals were found. Fig. (7), whereas no dityrosines were isolated after complete acid hydrolysis of the tanned polypeptide. Also surprising is the stability of the mixed benzocoumarin (which is still an oaminophenol) towards (auto)oxidation. This stability could be due to the quasi-planar structure of the tricyclic nucleus, and the related enhanced electron delocalisation. XOOH

OH

Fig. (7). The tricyclic lactone, arising from oxidative coupling between tyrosine and 3-hydroxyanthranilic acid

Generally speaking, the role of HA as a modifying agent for proteins seems reasonable also in processes in which it is not yet assessed, such as cuticle hardening and ootheca formation, processes that are thought to pass through covalent oxidative coupling (tanning) of phenols on to proteins.

1002 ANTIBACTERIC ACTIVITY OF HA The true metabolic role of HA per se (and not as mere precursor of other substances such as quinolinic acid and so on) is still awaiting a conclusive definition. However, HA 'toxicity' seems to be related not to the compound itself, but rather to other substances, arising form its (auto)oxidation. As ever, one can speculate about the chemical nature of those species, therefore some evidence exists in favour of the profound involvement of reactive intermediates in dioxygen reduction, namely superoxide, peroxide and hydroxyl, whereas other indications suggest the participation of anthranilyl and/or HA quinoneimine in the toxicity mechanism. One study has been reported [140] in which HA was allowed to autoxidise in the presence of a Salmonella typhimurium strain. A bactericidal effect was observed. To understand the mechanism of this effect, SOD was added to the incubation medium, obtaining an enhancement of the toxic action. This could be conceivably ascribed to the increase of hydrogen peroxide concentration, arising from superoxide dismutation. However, SOD is capable of stimulating HA autoxidation by means of a mass action effect [34], also a concomitant stimulation in anthranilyl radical production takes place, and the latter could well be the true bactericidal species. In an attempt to choose between these alternative hypotheses, catalase was also added to the incubation mixture. The result was a protective effect, so it was argued that H2O2 could be the toxic species. However, H2O2 concentrations, analogous to those generated by HA autoxidation in the presence of SOD, did not show bactericidal effects towards S. typhimurium. The possible involvement of hydroxyl radicals, arising from a FENTON reaction, owing to the unavoidable presence of iron ions in the incubation mixture, was ruled out because the addition of •OH scavengers was of little effect towards toxicity. The most reasonable conclusion is that, a synergistic effect of various intermediates and products of HA and molecular oxygen reaction are implicated in bactericidal effect of autoxidising HA. HA AS A CARCINOGEN HA has been recognised as a carcinogen. Since early studies, which appeared in the 1950s, a relation between some metabolites of tryptophan

1003 and some neoplastic forms was established [141-142]. A significant increase in bladder cancer incidence in rats fed with DL-tryptophan plus 2-acetamidofluorene has been seen in comparison with control rats, fed with the same doses of 2-acetamidofluorene alone [143]. Abnormally high concentrations of HA and other tryptophan metabolites were found in the urine from patients suffering from bladder cancer [144]. High concentrations of HA were also found in urine of patients with Bilharziainduced bladder cancer [145]. The possibility, that HA embedded in cholesterol pellets, surgically implanted within bladder, could cause bladder cancer in mice, has been assessed [72], while on the contrary it has been excluded on the basis of other analogous studies [146]. It was finally concluded that even if pellets of cholesterol alone can cause a low incidence of bladder cancer when implanted, when they also contain HA the incidence of cancer is significantly higher [69]. It seems to be well ascertained that HA is capable of causing an alteration of the cells at the bladder wall, so that cancer could arise. However, the mechanism of HAinduced carcinogenesis is still uncertain. So again, one asks: is }lAper se capable of cancer induction, or rather its (auto)oxidation products? And again, are the reactive intermediates in dioxygen reduction, or rather, are the semiquinonoid as well as the quinonoid species, both arising form the (auto)oxidative process, the actual causal agents of cancer insurgence? To answer these questions, a study [70] was conducted on mice, where in one control group cholesterol pellets, containing HA, were implanted, and in another group, the same pellets were implanted, but containing the HA derivative 5- hydroxy- benzoquinone- l-(2'- hydroxy-6'- carboxyanil)- 4imide [51]. As the latter compound is among the (auto)oxidation products of HA, it could in principle belong to the carcinogenic metabolites of tryptophan, as an oxidation product of HA. On the contrary, no significant increase in bladder cancer was observed in mice after implantation of cholesterol pellets, filled with that compound. Moreover, oral administration of large doses of ascorbic acid to mice, whose bladders have received HA embedded within cholesterol pellets, prevented bladder cancer insurgence, associated with HA implantation. These results suggested that the (auto)oxidative process may be responsible for bladder cancer insurgence, and not at least the quinonoid oxidation product studied. This observation obviously does not exclude the active participation of at least another HA oxidation product in carcinogenesis. Moreover, as ascorbic acid is mainly excreted in urine, a reasonable hypothesis on its beneficial action is re-

1004 lated to the inhibition of HA (auto)oxidation by competing for molecular oxygen. A further elucidation about the actual carcinogenic power of HA came from studies on DNA damage by (auto)oxidising HA, possibly in the presence of certain transition metal ions. The topic has already been discussed above, in the Paragraph on HA autoxidation. However, it is noteworthy that the intervention of •OH radicals in that damage should be excluded, when taking into account their extremely short halflife (10"^ sec). Accordingly, they have a highly reduced mean free path within the cell, so they should have only a few chances of damaging DNA [147148]. Moreover, it has been found that in the presence of autoxidising HA, relatively specific cuts are observed along DNA strands, corresponding to 5'-GTC-3' sequence [72]. Instead, it is well known that -OH radicals inflict a non-specific damage on DNA, as they are able to modify nucleotides indiscriminately [149]. In conclusion, the failure of hydroxy 1 scavengers to protect DNA against HA-induced damage in the presence of air, is not surprising. As already noted, catalase and the copper chelator bathocuproin effectively protected DNA against that damage [72]. The latter observation supports the idea that a redox cycle Cu(II)/Cu(I) is much involved in this autoxidation (as in many others, see for example [58]). NEUROTOXICITY OF HA Contrary to what has been found for other tryptophan metabolites, HA has not been recognised as a neurotoxic agent. Whereas in peripheral tissues the origin of the compound is certainly to be attributed to tryptophan degradation by means of the kynurenine pathway, its origin in the brain is much less clear. In fact, the two enzymes that promote the reactions leading to HA biosynthesis, namely kynureninase and kynurenine-3hydroxylase, both show a low activity within the rat brain [150-151]. Moreover, attempts to demonstrate kynurenine metabolism to quinolinic acid have failed [152], thus seriously undermining the theory that within the brain tryptophan could be metabolised to HA. HA concentration in the rat brain is somewhat low. In fact, HA is apparently unable to penetrate within neurons from the outside, and it can only cross the hematoenkephalic barrier by means of passive diffusion. Moreover, this crossing is hampered by a strong binding to plasma proteins. Studies carried out on

1005 rat brain slices have shown that upon HK administration, only a slight increase in intracellular HA concentration could be seen. This is in contrast with what was observed under the same conditions when HK was administered to spleen slices, where an intense activity of the kynurenine pathway in tryptophan catabolism is well documented [30]. The ease by which antliranilic acid enters within brain tissue is in agreement with the observation that HA levels rapidly increase when anthranilic acid is administered [30], even if this could be an indirect effect, not implying a direct conversion of anthranilic acid to HA, as noted above (Chapter 2). The observation moreover contrasts with the difficulty experienced in cell wall crossing by anthranilic acid. On the contrary, it has been already noted [153] that HK enters the cell, carried by a transport system specific to neutral aminoacids.

HA AS AN ANTIOXIDANT Undoubtedly, HA as well as HK can be considered as pro-oxidants, despite the opposite views concerning the exact role of superoxide during their autoxidation, and the actions of both superoxide and peroxide on the 'final' oxidation products CA and XA respectively. Different views also exist as regards the mechanism(s) of the actions exerted by enzymes such as SOD and catalase, as well as the precise involvement of various transition metal ions. On the other hand, other authors have dealt with the antioxidant properties of HA, that appear to be well documented [112,127, 154-158]. When mixed lymphocyte cultures undergo allogenic stimulation, in their sumatant a fluorescent substance, produced by macrophages in response to stimulation, can be evidentiated [159-160]. Although the substance was hard to identify, owing to its low concentration and sensitivity to autoxidation, it has been conclusively identified as HA [161]. This finding has allowed a new physiological role for HA to be hypothesised, namely that of being involved in defence mechanisms in living organisms. It is common knowledge that many low molecular weight compounds, such as some vitamins, play a defensive role against oxidative stresses [162-166]. Among these, certain tryptophan metabolites have shown a sharp anti-oxidant activity [167 ]. In particular, HA and HK show an anti-oxidant potency as high as that of ascorbic acid and pyridine

1006 nucleotides [127]. Moreover among tryptophan metabolites, under certain experimental conditions equimolar amounts of either HA or HK were more effective than ascorbic acid and trolox (a water-soluble tocopherol analogue) in protecting phycoerythrin-B against damage by peroxyl radicals. This is because every HA or HK molecule scavenges at least two peroxyl radicals [127]. The anti-oxidant activity of both HA and HK is related to their o-aminophenolic functionality, and therefore their nonhydroxylated analogues do not show any anti-oxidant activity [127]. Most probably, they can give an H» to other radicals, leading to semiquinones that dimerise and in turn can scavenge other radicals. Obviously, under different conditions, either experimental or physiopathological, the same semiquinone intermediates could behave as pro-oxidant agents and also be the actual toxic species arising from both HA and HK (auto)oxidation. The potential inhibition by interferon-y (IFN) of the first steps in LDL oxidation, mediated by human monocytes and macrophages, in a medium containing physiological concentrations of tryptophan, has been studied [154]. The inhibition effect was studied by measuring the consumption of the main anti-oxidant contained in LDL, namely a-tocopherol (TOH), the target molecules for the oxidative process (EFA moieties C 18:2 and C 20:4), and the accumulation of cholesteryl ester hydroperoxide, which is the main oxidation product. The results obtained demonstrated that IFN strongly lowered LDL oxidation mediated by monocytes and macrophages. The involvement of tryptophan metabolites in the process was demonstrated by the release of HA when the cells had been exposed to IFN. The results are in accordance with the well documented anti-oxidant properties of HA [17, 127, 156]. It is also noteworthy that LDL oxidation was inhibited when either tryptophan was present in the incubation medium and when exogenous HA was added. These considerations have led to the conclusion that HA is among the main anti-oxidants that protect LDL against oxidation. As regards its mode of action, HA can exert an important role as a coanti-oxidant, together with ascorbic acid and ubiquinol-10, in that it is able to reduce the a-tocopheryl radical T0» back to a-tocopherol TOH. Again, it must be underlined that this ability is due to the presence of the -OH substituent at the 3-position of the HA molecule. In fact, anthranilic acid, without -OH, is completely unable to regenerate TOH from TO* [156]. The anti-oxidant role of HA is not confined to the plasmatic environment. Therefore, it has been found [112] that HA (and also HK) is capa-

1007 ble of protecting crystallins form eye lens against oxyradicals, generated when the reaction mixture contains riboflavin and is irradiated. It is well known that riboflavin is an effective photosensitiser, and is widely used for photocrosslinking of proteins. Under the conditions studied, both HA and HK inhibit the oxidative damage towards lens proteins, showing photoprotective properties. Interestingly, also anthranilic acid is capable of exerting a protective effect, though it lacks the -OH substituent and therefore is not an o-aminophenol. Perhaps the radicals arising from riboflavin upon irradiation are so reactive that also anthranilic acid is capable of scavenging them, thus stopping the propagation of the chain reaction. Another aspect of the anti-oxidant properties of HA concerns it presence in some foods. Food anti-oxidants have recently drawn much attention because of a possible protective role against damage caused by free radicals, including the rising of some neoplastic forms as well as cellular senescence. These anti-oxidants can be used to slow down the process by which oils and fats grow rancid, and to protect liposoluble vitamins. HA has been identified as one among the several composants of the okara koji (OK), a particular okara fermented by Aspergillus oryzae [158]. OK prevents in vitro oxidation of lipids [168] and abolishes the increase of plasmatic cholesterol as well as the accumulation of lipids in rat liver [169-170]. Moreover, rats fed with oxidised oil and other foods, void of TOH, maintained normal levels of glutathione peroxidase both plasmatic and hepatic, and showed no alterations related to the normal growth, when they were fed with OK [158]. Another 'traditional' food, for which anti-oxidant properties have been claimed, is tempeh. This traditional Indonesian food, whose use has been documented since 16^^ century, is now far more widespread. It is obtained from fermented soybeans. Dried tempeh powders have been sometimes used as an anti-oxidant to treat freshly caught fish [171]. It has been reported that tempeh is exceptionally stable towards growing rancid [172], and 6,7,4'-trihydroxy-isoflavone was proposed as the actual antioxidant contained in that food. Later, a micro-organism, the mould Rhizopus oligosporus, was isolated from tempeh and found responsible for a high content of another anti-oxidant, different from the above isoflavone, and absent in the non-fermented tempeh [155]. It was also found that only tempeh, fermented by Rh oligosporus, shows distinct anti-oxidant properties [155], even if inspection of the isoflavone structure renders its putative anti-oxidant properties as highly probable, especially when one

1008 considers the well-known anti-oxidant properties of flavones such as quercetin. The 'new' anti-oxidant was then identified as HA, whose concentration rises during fermentation until its reaches a maximum (about 50 mg/lOOg dry material) after two days of fermentation. Accordingly, the maximum anti-oxidant power is reached after the same period of time. From these results, it was concluded that HA and not the isoflavone is the main substance responsible for the anti-oxidant properties of tempeh. However, no data are available on the actual beneficial effects deriving from usual eating of tempeh by humans, because no data exist about the fate of HA after ingestion nor about variations of plasmatic HA concentration in tempeh eaters. It has also been proposed [157] that HA concentrations within the organism could not reach values high enough to allow for a distinct antioxidant activity. On the contrary, HA could act as a pro-oxidant. Both low and high concentrations of HA have been studied as it concerns their effects on biological systems. The obtained results show that high HA concentrations actually exert a sharp anti-oxidant effect towards lipid oxidation in its initial step, and on the contrary low concentrations are cytotoxic against malignant human cells. Moreover HA could induce oxidative stress because it seems to promote SOD inhibition and GSH oxidation. Because oxidative stress has been proposed as a factor promoting cell death through apoptosis [98], this could in part explain the cytotoxic effects of HA. Such an action of HA could be exerted through a multiple-way-induced superoxide production together with GSH depletion. The fact that despite of its widespread use, well documented for centuries, tempeh has never provoked visible toxicity symptoms, argues against any negative effect against normal cells. On the other hand, pre-malignant cells show a more pronounced tendency to apoptosis in comparison with the normal ones [173]. Therefore, one very intriguing hypothesis is that large doses of HA could have a somewhat selective killing effect on premalignant cells, by inducing their preferential apoptosis and consequent death, while respecting the healthy ones. This hypothesis, however, is still awaiting a definitive assessment.

1009

HA AS A PRECURSOR OF PHENOXAZINE NUCLEUS Actinomycins are chromopeptides, generally containing the same chromophore, based on the tricyclic system of phenoxazinone (phenoxazone), and known as actinocin, Fig. (8). The chromophore is responsible for the intense yellow, orange or red colours of actinomycins. The first known actinomycin was isolated from a culture medium of an actinomyces, Streptomyces sp. [174], and caused some interest on account of its antibiotic activity. Many other actinomycins were isolated later and characterised by their antibiotic and/or antitumor activities [175]. Among these, the best known is actinomycin D (Dactinomycin), Fig. (8). COOH

COOH

CONHR

CONHR NH2

Q CH3

^O"

^ CH3

^O

Q

^

^o-"

CH3

^

^o

CH3

R = - THR-DVAL-PRO-SAR-MeVAL O-

V

ig. (8). The structures of actinocin (left) and actinomycin (right)

Naturally-occurring actinomycins show as a common feature the presence of the phenoxazinone chromophore, doubly-linked to a pentapeptide which forms a macrocyclic structure. Differences among actinomycins arise from variations in composition of the pentapeptide. The precursors of the aminoacids, forming the pentapeptide, have been singled out [176178]. Regardless of the peptide moieties, it is now clear that biosynthesis of the chromophore compulsorily passes via HA as a key intermediate [179].

1010 HA is specifically methylated to form 3-hydroxy-4-methyl-anthranilic acid [180-181]. This reaction is catalysed, in Streptomyces antibioticus, by a specific methyltransferase [182]. The enzyme appears to be bound to cell membrane, and is rather specific: for example, HK is a non-substrate. The pentapeptide moiety is not attached to the cromophore until actinocin is formed. In Str. chrysomallus, the whole enzyme complex responsible for the modification and polymerisation of the aminoacids which constitute the peptide chain has been isolated and characterised as being composed of three fractions (actinomycin synthetase I, II, and III) [183184]. The final reaction, leading to the active molecule, is catalysed by another enzyme, phenoxazinone synthase. This enzyme has been isolated from Str. antibioticus [55] and can exist in a dimeric form, with lower activity, and in a hexameric form, with higher activity, depending on the age of the cultures [185]. The enzyme, cloned in Str. lividans [186] contains 4 - 5 copper ions. However, only 3 copper ions are functionally active, and can accept electrons from substrates [187]. Given the close structural analogy between CA and actinocin, much interest has grown up around the enzymes, involved in their biosyntheses. Various studies account for the existence of an enzyme activity, capable of transforming HA into CA. It has been reported that such an activity, relatively non-specific, is present in the liver of various poecilotherm animals [188]. Another study deals with the existence of an enzyme, promoting the conversion of HA into CA, in rat liver [189-190]. This 'cinnabarinate synthase' activity seems to be confined to the nuclear fraction, and shows biochemical features somewhat differing from that reported for the Streptomyces enzymes [55]. Also a similar activity was found in some plants [191]. The enzyme, purified from rat liver, is rather specific, being inactive towards o-aminophenols, other than HA, and does not for example work on HK. Moreover it absolutely requires Mn^^ for its catalytic activity [192], such as does the enzyme from Tecoma stans (a plant from southern India) [193]. Also in the polyporaceous fungus, Pycnoporus cinnabarinus, the conversion of HA to CA has been shown, even if there is no Mn^^ requirement for the process [194]. The same was seen also in extracts of spinach leaves [195]. Nuclear fractions, coming from other rat organs, were found able to catalyse the conversion of HA to CA [195]. Nucleotide bases, and in particular guanine, behave as competitive inhibitors for the conversion, and this fact, together with the strict structural analogies between

1011 CA and actinocin, and the intercalating properties of the latter, prompted Nair [195] to hypothesise the existence of a regulatory mechanism for RNA synthesis, involving the oxidative dimerisation of HA. Following this suggestion, CA could specifically interact with DNA, preferably on guanine residues, such as actinomycin D does [196]. This was confirmed by the observation, that either preformed CA and HA (converted to CA by the nuclear fraction) were actually bound by DNA, in correspondence with guanine residues. However, no data exist about the actual formation and concentration of CA within the rat liver. The existence of a specific cinnabarinate synthase or synthetase in the nuclear fraction is still subjudice. It has been noted [197] that the nuclear fraction is capable of releasing small amounts of hydrogen peroxide. Therefore, H2O2 plus Mn^^ ions could be an efficient, non-enzymic system for HA conversion to CA [17]. Moreover, purification and characterisation of a true cinnabarinate synthase from that source have never been reported. HA AS AN ENZYME SUBSTRATE: ITS INTERACTIONS WITH LACCASE, PEROXIDASE, CERULOPLASMIN, AND TYROSINASE It is well ascertained that HA conversion to CA can take place either enzymically or non-enzymically. So, the process may be described as an (auto)oxidation, involving reactive intermediates in dioxygen reduction as well as transition metal ions [35, 50, 72]. Alternatively, the conversion can be enzyme-mediated. In Pycnoporus cinnabarinus, a ligninolytic white-rot fungus, a laccase (p-diphenol : O2 oxidoreductase, water producing, EC 1.10.3.2) converting HA to CA, has been identified. This is not unexpected if one considers the relatively low substrate specificity of laccases, that oxidise a number of quinonisable substrates such as o- and /7-diphenols, -aminophenols, and -diamines, either natural or unnatural [198-199]. Laccases have been identified and often also purified and characterised from plants and fungi, in particular from white-rot fungi, that require the enzyme for the delignification of their wooden growth media. The physiological role of laccases is not confined to delignification: some relationship exists, depending on the particular vegetal or fungal organism considered, between laccase production and activity on the one hand, and sexual differentiation, pigmentation of fruit bodies.

1012 detoxification of xenobiotics, on the other. A similar copper protein, ceruloplasmin, may be considered the counterpart of laccase in the animal world [200]. The laccase from P. cinnabarinus [201-202] is able to convert HA to CA even at pH 4, where the autoxidation of HA is negligible. The fungus has the unusual feature of accumulating HA in the culture medium, and then releasing extracellular laccase. A decrease in HA concentration, paralleled by an increase in CA formation, is strictly dependent on laccase excretion into the culture medium [201]. Also other related fungi, P. sanguineus and P. coccineus, are effective laccase producers [203-204]. Their names are due to the red or orange-red colours [205-207] which they show at their sporocarps (the same is of course true for P. cinnabarinus). These fungi actively produce two pigments, cinnabarin and tramesanguin respectively. Fig. (9). COOH

CHO

.N.

JL

CH2OH

.NH2

COOH .N.

A

.NH2

Fig. (9).The structures of cinnabarin (left) and tramesanguin (right)

Also Coriolus sanguineus and Trametes cinnabarina produce tramesanguin [205-206, 208]. It is noteworthy that some confusion exists about the systematics of Polyporaceae, so for example P. cinnabarinus and T. cinnabarina are actually synonymous. This ambiguity renders it somewhat difficult to assign a given pigment unequivocally to one fungus species, and the following observations unavoidably suffer from this. Moreover, tramesanguin had been also named cirmabarine [204]: the name is quite unsuccessful, as confusion with cinnabarin is highly probable, and accordingly will no longer be utilised here. The existence of partially reduced aromatic compounds (in this case, cirmabarin and tramesanguin), deriving from substituted benzoic acids, and therefore carrying formyl- and/or hydroxymethyl- substituents replacing a carboxy-

1013 substituent, is not confined to the genus Pycnoporus. Sporotrichum pulverulentum (anamorph of Phanerochaete chrysosporium) for example reduces vanillic acid to vanillin and vanillyl alcohol, and the same has been shown as the general behaviour of white-rot fungi [209-211]. Another white-rot fungus, Phlebia radiata, actively reduces veratric acid (3,4-dimethoxybenzoic acid) to veratraldehyde and then to veratryl alcohol [212]. Therefore, the reduction of CA by typical white-rot fungi such as those pertaining to the genus Pycnoporus (=Trametes=DaedaIea) is a quite normal evenience. The already noted antibacterial properties of phenoxazinones have been confirmed also in the case of P. cinnabarinus [213]. The fungus, grown in a liquid medium, concomitantly produces laccase and CA (formed by enzymic action of laccase on HA, excreted by the fungus in the culture medium). Proportional to laccase and CA production, a sharp antibacterial activity arose in the culture medium, higher towards gram(+) and lower towards gram(-) bacteria; the activity has been conclusively ascribed to CA. As expected, a P. cinnabarinus mutant, lacking laccase, is void of any antibacterial activity. In nature, the fungus lives on decayed wood and not in liquid media, and CA has been found within the sporocarps rather than being ubiquitous in the hyphae, so a reasonable hypothesis about its biological activity is that of protecting the growing fruit bodies against bacterial infections. Both the physiology and enzymology of P. cinnabarinus have attracted much attention in relation to both HA excretion and CA production by the fungus [201-202, 214-215]. The delignification process by white-rot fungus is an extremely complex one, and not yet fully elucidated [214]. It is commonly believed that a lignin peroxidase is deeply involved in lignin molecular breakdown, because redox potentials of the known laccases (that vary widely from one fungal genus to another [216217]) are too low for these enzymes to be efficient catalysts of nonphenolics oxidation. Apart from laccases, a serious limitation for the action of any ligninolytic system is the highly polymeric nature as well as the complete insolubility of native lignins, so even lignin peroxidases can hardly perform lignin oxidative breakdown with any reasonable efficiency. The problem is solved by the existence of a very effective redox mediator, namely a veratryl alcohol-derived cation radical, formed by lignin peroxidase and capable of freely diffusing close to to-be-cleaved lignin. An alternative strategy involves the production of another peroxidase, manganese peroxidase. This produces Mn(III) chelates, that exert

1014 the same strong oxidising function as the above mentioned cation radical. However, also among white-rot fungi, that are strictly specialised in lignin depolymerisation and utilisation, some of them lack any detectable peroxidase, although they produce very active laccases. The genus Pycnoporus is among these peroxidase-lacking white-rot fungi, whereas it is an exceptionally efficient laccase producer [203-204]. Laccases do not produce reactive intermediates of dioxygen reduction, as the dioxygen molecule is tightly bound to the type III binuclear copper cluster until it is fully reduced, by means of successive one-electron steps, to water. Therefore, any indirect delignifying action of laccase cannot proceed through the reactive intermediates of dioxygen reduction. Some studies have demonstrated that purified fungal laccases (that show redox potentials higher than plant laccases, and therefore should be more efficient oxidisers) are substantially ineffective in depolymerising lignins and semisynthetic analogues [214]. Laccases absolutely require at least one phenolic hydroxy substituent on to an aromatic ring to exert their activity, and so their incapability of efficiently depolymerising lignins is not unexpected, taking into account that the phenylpropanoid units that compose lignin are largely methylated. These theoretical considerations have recently received a convincing experimental confirmation [214]: three lignin model compounds, indicated as I, II and III in Fig. (10), were tested as substrates for both purified lignin peroxidase isozyme H8 [218] from Phanerochaete chrysosporium, and purified laccase from P. cinnabarinus. As expected, the lignin peroxidase with high efficiency cleaved all three model compounds, regardless of their structural difference, and owing to its high redox potential. On the contrary, laccase quickly oxidised model compound I, but showed a comparatively low efficiency towards II, whose guaiacyl moiety is subjected to both -I and - M effects of a carbonyl group. Model compound III is a non-substrate for laccase, lacking any phenolic hydroxy substituent. How does P. cinnabarinus oxidatively depolymerise Hgnin, if it lacks any peroxidase activity? One answer was suggested by the discovery that the synthetic chromogenic laccase substrate ABTS (2,2'-azino-6/5'-(3-ethylthiazoline-6-sulfonic acid), diammonium salt) could act as a redox mediator in lignin depolymerisation [219220]. Also the fungi of the genus Pycnoporus possess their own redox mediator: HA. As reported above, HA had earlier been identified as the immediate precursor of CA, cinnabarin and tramesanguin in Pycnoporus, but its role in the biology of the genus is much more crucial than that of being merely the parent compound of phenoxazinone-based pigments. To

1015

assess the actual role of HA as a redox intermediate in lignin oxidative depolymerisation, the lignin model compounds I, II and III were therefore incubated with laccase, in the presence of HA: after 48 hours, more than half of III were cleaved to produce veratric acid and guaiacol. Concomitantly, HA was converted to CA. A particular HA/III concentration ratio was also found to be optimal for the degradation of III; variations of this ratio led to a dramatic drop in degradation efficiency. This suggests that the regulation of biosynthesis and excretion of HA by P. cinnabarinus could be a modulation mechanism for lignin degradation by the fungus. It is noteworthy that also the polymeric lignin model compound known as a dehydrogenative polymerisate is satisfactorily cleaved by the mixture laccase/HA, while laccase alone was substantially ineffective. CH3O

HOCH^

CH3O

^CHOH

HOCH

OCH3 OH

CH30

HOCH

Y

^0CH3

OH

II

y ^

-0CH3

OCH3

III

Fig. (10). The three model compounds, used for assessing the redox mediating role of 3-hydroxyanthranilic acid in delignification by laccase

Interestingly, ABTS was much less effective as a redox mediator than HA when P. cinnabarinus laccase acted on III. Moreover, veratryl alcohol (which is a non-substrate for laccase alone) was efficiently converted to veratraldehyde by laccase when ABTS was added to the reaction mixture, whereas less than 10% was oxidised to the aldehyde when the mediator was HA. The reason of this mechanistic difference in regioselectivity of

1016 oxidation is unclear. For unknown motives, the cation radical arising from ABTS oxidation preferentially abstracts a benzylic hydrogen from veratryl alcohol, whereas hydroxyanthranilyl radical obtains an aromatic electron from the benzene ring. In this behaviour, the mediator strictly resembles both lignin peroxidase and manganese peroxidase action. Obviously, some other mediators of lignin depolymerisation, either natural or unnatural, will be presumably discovered; indeed, some are already well known [221-222]. As a concluding remark, it can be noted here that the presence of HA as the redox mediator could be of taxonomic value, helping to disentangle the complex classification of the genus Pycnoporus and of some others strictly related to it. Also the blue copper protein, ceruloplasmin, which is considered the laccase counterpart in the animal kingdom, could oxidise HA. The seric protein is widespread in animals and has been studied in depth, but its actual physiological role is far from being understood. It shows a wide substrate specificity [223-224]. The ability of ceruloplasmin to oxidise HA has been demonstrated [201], even if this oxidation is slow and stops when unaltered HA is still present. It has been suggested that the protein could be inactivated by some reactive intermediates in HA oxidation. As noted elsewhere [38], it is reasonable to conclude that HA / ceruloplasmin interaction, if existing, would be of very little physiological importance. Another enzyme capable of using HA as a substrate is tyrosinase. Tyrosinases (monophenol, o-diphenol : O2 oxidoreductase, water producing, EC 1.14.18.1) are copper enzymes catalysing two different reactions: a) hydroxylation of monophenols to o-diphenols (monophenolase or cresolase activity), and b) oxidation of o-diphenols to o-quinones (diphenolase or catecholase activity). In both cases, the enzyme uses molecular oxygen as the final electron acceptor. Tyrosinase are widespread in the living world, having been found in micro-organisms, plants, animals and man. It is well known that in mankind tyrosinase is responsible for the formation of melanic pigments, which are localised in the melanocytes, and in the epithelial cells of the eye retina. Moreover, seasonal variations in enzyme activity have been observed in human blood [225]. The usual substrates of tyrosinases are as noted above, either mono- and o-di-phenols, but recently it has been shown that the enzyme from Neurospora crassa is capable of oxidising also the nitrogencontaining isologues of the 'normal' substrates, i.e. aromatic amines and

1017 aminophenols, that are finally converted to quinoneimines [226]. As could be expected in the case of HA, the compound is oxidised to the corresponding quinoneimine, which in turn undergoes the reactions outlined in the Paragraph on autoxidation. By comparing the Km values obtained for some o-aminophenols and for the o-diphenoHc counterparts, no significant differences were found [226], as a confirmation of structural and electronic analogies between the two classes of substances. The position of the -COOH substituent related to the 6^-aminophenol and odiphenol moieties were also studied with respect to differences in affinities. A close proximity was found between -NH2 and -COOH groups which favours the binding of the substrate to the active site of the enzyme, whereas the contrary was seen when the -COOH group was moved to the ortho position to the -OH group. Because the spatial structure of the active site of tyrosinases is not yet known, it cannot be assessed whether these differences are attributable to electrostatic interactions among the ring substituents or to an asymmetry between the two copper ions of the enzyme. Another interesting aspect of tyrosinase activity is represented by the typical lag time the enzyme shows when acting on nonophenols. It has been suggested that this lag time is actually the time needed by the enzyme to reach its stationary state, with respect to o-diphenol concentration [69]. An alternative hypothesis explains the existence of the lag time by invoking an autocatalytic mechanism depending on the DOPA production when the monophenolic substrate is just tyrosine [227]. Both the existence and the duration of the lag time depend on several factors, among which is the presence of reductants or o-diphenols, apart from that arising from the enzyme action on a particular monophenol [228-229]. Recently it has been found that purified mushroom tyrosinase can catalyse the oxidation of HA, leading to CA formation [230]. The affinity of the enzyme for HA is very considerable and is comparable to that shown for odiphenols [231-232]. The interaction between tyrosinase and HA goes beyond its function as a '^wa^'/'-diphenolic' substrate. So, HA is capable of reducing the lag time when the enzyme is working on A^-acetyl-tyrosine or 4-/^r^butyl-phenol. Substrates chosen, as the corresponding quinonoid products are relatively stable and cannot actively participate in nucleophilic cyclisation reactions leading to catechols, capable of further complicating the reaction mechanism [230]. This observation has allowed the hypothesis that there is a direct interaction between the active site of the enzyme and HA. HA could be an activator with relation to the lag time.

1018 This behaviour of HA is in agreement with a kinetic mechanism [229, 233, 234], which considers three different states for the enzyme, Emet, Eoxy and Edeoxy. Edeoxy somcwhat resembles deoxy-hemocyanin and has a dinuclear copper(I) cluster and can be considered as the reduced counterpart of Emet, which contains a binuclear copper(II) cluster. A significant fraction of the enzyme in the resting state is in the met form, which is unable to act on polyphenols; HA draws this form in the deoxy state, while being oxidised to the quinoneimine. Interestingly, HA cannot completely abolish the lag time (difference from o-diphenols) even in high concentrations; this is due to the low value of A:cat (in other words, HA interacts with high affinity with tyrosinase, but is only slowly oxidised by it). The role of HA as a tyrosinase substrate and activator is of potential physiopathological relevance, since it is well known that many diseases exist, in which the tryptophan metabolism is disturbed, and therefore noticeable variations of HA concentrations could take place. Human tyrosinase is not confined to melanocytes, but has been found also in serum of both healthy and ill indviduals, and particularly in the presence of either melanoma or melanosis, and in cultivated melanoma cells [235-237]. HA, HK AND MELANINS From a chemical point of view, and in a very simplified way, in the animal kingdom the so-called eumelanins (or melanins tout-court), black or dark brown, and completely insoluble in water, differ from pheomelanins, that contain sulfur, varying from yellow to brown-red, and alkali-soluble. Both the kinds of melanin can be present at the same time within the same organism. For a deeper insight in melanin occurrence and properties, see [238]. It has been hypothesised [239] that eumelanins derive biosynthetically from tyrosine, whereas pheomelanins of hair involve the participation of other monomers besides tyrosine, for example HK which can polymerise to a yellow pigment (an ommochrome-like substance?), owing to interferences with the main biosynthetic pathway of eumelanin synthesis. Some attempts to assess tryptophan and/or some metabolites in melanin biosynthesis were carried out in the late 1960s [240, 241]. Tryptophan and HK were tested as tyrosinase substrates, leading to polymeric materials, whose colour, infusibility, heterogeneity and difficult purification resembled those of 'true' eumelanins. On the contrary, no melanin-like

1019 products were obtained when HA was used as a tyrosinase substrate [242]. Later experiments, performed by using radiolabeled tryptophan, led to the idea that tryptophan itself rather than its metabolites are involved in (eu)melanin biosynthesis, at least under pathological conditions such as melanoma [243]. However, any attempt to assess any structural difference between 'normal', tyrosine-derived melanin and that putatively containing tryptophan, have so far failed [244]. Other studies have suggested that tryptophan could have a regulatory role in melanin biosynthesis, and it has been reported that the aminoacid is capable of either stimulating or inhibiting the biosynthetic process, depending on the particular experimental conditions [245]. Experiments, carried out on patients, suffering from vitiligo, a skin disease characterised by localised hypochromic spots, and whose etiopathology is still unclear, showed that after oral administration of significant doses of tryptophan, a reduced excretion of metabolites such as HK and HA was seen, in comparison with normal subjects [74]. On the contrary, the same experiments, conducted on patients, suffering from skin hyperpigmentation, showed a sharp increase in excretion of both HK and HA [75]. Indeed, the involvement of some tryptophan metabolites, and particularly HK and HA, in melanin biosynthesis, has been recently proposed. However, convincing proofs of that assertion are still lacking. A more probable involvement of both HK and HA in melanin biosynthesis concerns the so-called plasma soluble melanins, (PSM). PSM can be formed either in vivo or in vitro [246] as the result of oxidation/polymerisation reactions, starting from certain aromatic compounds, present in plasma, and bearing either -OH and -NH2 substituents, or both, such as HK, HA, homogentisic acid, catecholamines and so on. PMS, whose functions in living organisms are still obscure, can be associated with proteins, mucoproteins, lipids (in this case, lipofuscins arise), or with non-lipo-proteins, leading to soluble melanoproteins [247]. PSM have been found in urines of healthy individuals, indicating that they are water-soluble regardless of their conjugation with proteins, and that the kidneys can perform a selective elimination, therefore retaining the protein fraction [246]. HA-derived PSM have been identified in plasma from healthy individuals, regardless of age, sex and race [248]. Also human hemoglobin binds a little amount of PSM [249]. PSM formation has been related to some degenerative processes dependent on aging, chronic renal diseases, degenerative cardiovascular diseases [250]. Abnormally high amounts of PSM have been also found in alcaptonuria patients with

1020 cardiovascular complications [250]. In all these cases it has been suggested that abnormal situations, leading to enhanced PSM formation, can cause or at least aggravate pathological states through oxidative mechanisms (the implications of HA and HK (auto)oxidation have been thoroughly discussed above), and the significance of the antioxidant-induced protection has been emphasised. ABBREVIATIONS HK HA CA XA HXA* H2XA SOD HRP HKG IFN TOH OK ABTS PSM

Hydroxykinurenine Hydroxyanthranilic acid Cinnabarinic acid Xanthommatins Dihydroxanthommatin semiquinonoid radical Dihydroxanthommatin Superoxide dismutase Horse radish peroxidase HK-O-yS-D-glucoside y-Interferon a-Tocopherol Okara koji 2,2*-Azino-bis-(3 -ethylthiazoline-6-sulfonic acid) Plasma soluble melanin

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II]

Hayaishi, O.; Annu. Rev. Biochem., 1994, 63, 1-24. Heine, H.; Radke, M.; Wutzke, K.D.; Amino Acids, 1995, 9, 191-205. Sainio, E.L.; Pulkki, K.; Young, S.N.; Amino Acids, 1996, 10, 21-47. Schwarcz, R.; Biochem. Soc. Trans., 1993, 21, 77-82. Dotting, N.P.; Chem. Soc. Rev., 1995, 24, 401-412. Stone, T.W.; Perkins, M.N.; Eur. J. Pharmacol., 1981, 72, 411-412. Schwarcz, R.; Whetsell, W.O.; Mangano, R.M.; Science, 1983,219, 316-318. Beal, M.F.; Kowall, N.W.; Ellison, D.W.; Mazurek, M.F.; Swartz, K.J.; Martin, J.B.; Nature, 1986, 321, 168-171. Schwarcz, R.; Brush, G.S.; Foster, A.C.; French, E.D.; Exp. Neurol., 1984, 84, 117. Perkins, M.N.; Stone, T.W.; Brain Res., 1982, 247, 184-187. Kessler, M.; Terramani, T.; Lynch, G.; Baudry, M.; J. Neurochem., 1989, 52, 1319-1328.

1021

[12] [13]

[14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37]

Moroni, F.; Rossi, P.; Lombardi, G.; Beni, M.; Carla, V.; J. Neurochem., 1988, 51,177-180. Schwarcz, R.; Du, F.; Schmidt, W.; Turski, W.A.; Gramsbergen, J.B.P.; Okuno, E.; Roberts, R.C. In Neurotoxins and Neurodegenerative Disease; Langston, J.W. and Young, A , Eds.; Ann. N. Y. Acad. Sci.: New York, 1992; Vol. 648, pp. 140153. Gripenberg, J.; Honkanen, E.; Patoharju, O.; Acta Chem Scand., 1957, 11, 14851492. Ogawa, H.; Nagamura, Y.; Ishiguro, 1.; Hoppe-Seyler's Z Physiol Chem., 1983, 364, 1059-1066. Ogawa, H.; Nagamura, Y.; Ishiguro, I.; Hoppe-Seyler's Z Physiol Chem., 1983, 364, 1507-1518. Christen, S.; Southwell-Keely, P.T.; Stocker, R.; Biochemistry, 1992, 31, 80908097. Butenandt, A.; Schiedt, U.; Biekert, E.; Kommann, P.; Liebigs Ann. Chem., 1954, 586,217-228. Butenandt, A.; Biekert, E., Kubler, H.; Linzen, B.; Hoppe-Seyler's Z Physiol C/zew., 1960,319,238-256. Ferre, J.; Silva, F.J.; Real, M.D.; Mensua, J.L.; Biochem. Genet., 1986, 24, 545269. Van Heyningen, R.; Nature, 1971, 230, 393-394. Pirie, A.; Isr. J. Med Sci., 1972, 8, 1567-1573. Grover, D.; Zigman, S.; Exp. Eye Res., 1972, 13, 70-76. Tomoda, A.; Yoneyama, Y.; Yamaguchi, T.; Kakinuma, K.; Kawasaki, K.; Yonemura, D.; FEBS Lett., 1987, 219, 472-476. Sainio, E.L.; Hynnien, P.H.; J. Chromatogr., 1987, 420, 263-274. Christen, S.; Stocker, R.; Anal Biochem.,1992, 200, 273-279. Palcic, M.M.; Antoun, M.; Tanizawa, K.; Soda, K.; Floss, H.S.; J. Biol Chem., 1985,260,5248-5251. Kobayashi, G.; Kuno, S.; Itada, N.; Hayaishi, O.; Kozuka, S.; Oae, S.; Biochem. Biophys. Res. Commun., 1964, 16, 556-561. Subba Rao, P.V.; Sreleela, N.G.; Premkumar, R.; Vaidyanathan, C.S.; Biochem. Biophys. Res. Commun., 1968, 31, 193-196. Baran, H.; Schwarcz, R.; J. Neurochem., 1990, 55, 738-744. Vesica, A.; Di Prisco, G.; J. Biol Chem., 1962, 237, 2318-2324. Ueda, T.; Otsuka, H.; Goda, K.; Ishiguro, I.; Naito, J,; Kotake, Y.; J. Biochem., 1978, 84, 687-696. Bender, D.A.; McCreanor, G.M.; Biochim. Biophys. Acta, 1982, 717, 56-60. Dykens, J.A.; Sullivan, S.G.; Stem, A.; Biochem. Pharmacol, 1987, 36, 211-217. Manthey, M.K.; Pyne, S.G.; Truscott, R.J.W.; Biochim. Biophys. Acta, 1990, 1034,207-212. Uri, N. In Autoxidation and antioxidants', Lundberg, W.O., Ed.; Interscience: New York, 1961; Vol. 1, pp. 55-106. Ingraham, L.L.; Meyer, D.L.; Biochemistry of Dioxygen, Plenum Press. New York, 1985, pp. 7-20.

1022

[38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69]

Rescigno, A.; Rinaldi, A.C.; Sanjust, E.; Biochem. Pharmacol, 1998, 56, 10891096. Cohen, G.; Heikkila, R.E.; MacNamee, D.; 1 Biol. Chem., 1974, 249, 2447-2452. Heikkila, R.E.; Cabbat, F.; Anal. Biochem., 1976, 75, 356-362. Liang, Y.O.; Wightman, R.M.; Adams, R.N.; Eur. J. Pharmacol., 1976, 36, 455458. Heicklen, J.; Int. J. Chem. Kinet., 1981, 13, 651-665. Jeong, K.M.; Kaufman, F.; J. Phys. Chem., 1982, 88, 1816-1821. Kobayashi, G.; Ando, W.; Biochem. Biophys. Res. Commun., 1979, 188, 676-681. Czapski, G.; Ilan, Y.A.; Photochem. PhotobioL, 1978, 28, 651-663. Koppenol, W.H.; Butler, J.; vanLeeuwen, J.W.; Biochem. Biophys. Res. Commun., 197S, 188,655-658. Ferradini, C; Foos, J.; Houee, C; Pauchault, J.; Photochem. PhotobioL,191^, 28, 697-700. Weinstein, J.; Bielski, B.H.J.; J. Am. Chem. Soc, 1979, 101, 58-62. Iwahashi, H,; Ishii, T.; Sugata, R.; Kido, R.; Biochem. J., 1988, 251, 893-899. Dykens, J.A.; Sullivan, S.G.; Stern, A. In Superoxide and Superoxide Dismutase in Chemistry, Biology and Medicine', G. Rotilio, Ed.; Elsevier Science B.V.: Amsterdam, 1986; pp. 92-94. Butenandt, A.; Keck, J.; Neubert, G.; Liebigs Ann. Chem., 1957, 602, 61-72. Suatoni, J.C.; Synder, R.E.; Clarck, R.O.; Anal. Chem., 1961, 33, 1894-1897. Job, D.; Dunford, H.B.; Eur. J. Biochem., 1976, 66, 607-614. Barry, C.E.; Nayar, P.G.; Begley, T.P.; Biochemistry, 1989, 28, 6323-6333. Katz, E.; Welssbach, H.; J. Biol. Chem., 1962, 237, 882-886. Ishii, T.; Iwahashi, H.; Sugata, R.; Kido, R.; Fridovich, 1.; Arch. Biochem. Biophys., 1990, 276, 248-250. Patai, S.; Rappoport, Z.; The Chemistry of Quinonoid Compounds, Wiley: New York, 1988. Rinaldi, A.C.; Porcu, M.C.; Curreli, N.; Rescigno, A.; Finazzi-Agro, A.; Pedersen, J.Z.; Rinaldi, A.; Sanjust, E.; Biochem. Biophys. Res. Commun., 1995, 214, 559-567. Feinstein, R.N.; Proc. Soc. Exp. Biol. Med., 1978, 158, 398-401. Nicholls, P.; Schonbaum, G.R. In The Enzymes; Boyer, P.D.; Lardy, H.; Myrback, K., Eds.; Academic Press: New York, 1963; pp. 147-225. Oshino, N.; Oshino, R.; Chance, B.; Biochem. J., 1973, 131, 555-563. Zhang, L.; Bandy, B.; Davison, A.J.; Free. Radic. Biol. Med,\996, 20, 495-505. Gee, P.; Davison, A.J.; Arch. Biochem. Biophys., 1984, 231, 164. Sanjust, E.; Rinaldi, A.C.; Rescigno, A.; Porcu, M.C.; Alberti, G.; Rinaldi, A.; Finazzi-Agro, A.; Biochem. Biophys. Res. Commun., 1995, 208, 825-834. Luzzatto, E.; Choen, H.; Stochkeim, C; Wieghardt, K.; Meyerstein, D.; Free Radic. Res., 1995, 23, 453-463. Bandy, B.; Moon, J.; Davison, A.J.; Free. Radic. Biol. Med., 1990, 9, 143-148. Ishii, T.; Fridovich, I.; Free. Radic. Biol. Med, 1990, 8, 21-24. Iwahashi, H,; Ishii, T.; Sugata, R.; Kido, R.; Arch. Biochem. Biophys., 1990, 276, 242-247. Bryan, G.T.; Brown, R.R.; Price, J.M.; Cancer Res., 1964, 24, 596-602.

1023

[70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95]

Schlegel, J.U.; Pipkin, G.E.; Nishimura, R.; Shultz, G.N.; J. Urol, 1970, 103, 155-159. Pipkin, G.E.; Nishimura, R.; Banowsky, L.; Schlegel, J.U.; Proc. Soc. Exp. Biol. Med., 1967, 126, 702-704. Hiraku, Y.; Sumiko, I.; Oikawa, S.; Yamamoto, K.; Tada, S.; Nishino, K.; Kawanishi, S.; Carcinogenesis, 1995, 16, 349-356. Heubner, W.; Meier, R.; Naunyn-Schmied Arch. Exp. Pathol. Pharmacol. 1923, 100, 137-148. Costa, C; Cardin, E.L.; De Antoni, A.; Vanzan, S.; Allegri, G.; Boll. Soc. It. Biol. 5p^r., 1981,57, 1933-1942. Costa, C; Cardin De Stefani, E.L.; Giorn. It. Chim. Clin., 1984, 9, 107-111. Tomoda, A.; Shirasawa, E.; Nagao, S.; Minami, M.; Yoneyama, Y.; Biochem. J., 1984, 222, 755-760. Tomoda, A.; Shirasawa, E.; Yoneyama, Y.; Hemoglobin, 1986, 10, 33-48. Tomoda, A.; Yamaguchi, T.; Kojima, H.; Amemiya, H.; Yoneyama, Y.; FEBS Lett, 1986, 196,44-48. Tomoda, A.; Hamashima, H.; Arisawa, M.; Kikuchi, T.; Tezuka, Y.; Koshimura, S.; Biochim. Biophys. Acta, 1992, 1117, 306-314. Wallace, W.J.; Houtchens, R.A.; Holt, J.M.; Caughey, W.S. In Biochemical and Clinical Aspects of Hemoglobin Abnormalities', Caughey, W.S., Ed.; Academic Press:New York, 1978. Bando, M.; Nakajima, A.; Satoh, K.; J. Biochem., 1981, 89, 103-109. Gray, D.O.; Weitzman, P.D.J. In Data for Biochemical Research, Dawson, R.M.C.; Elliott, D.C.; Elliott, W.H.; Jones, K.M., Eds.; Oxford University Press: London, 1969, 2nd Ed., pp. 1-63. Bolognese, A.; Bonomo, R.P.; Chillemi, R.; Sciuto, S.; J. Heterocyclic Chem., 1990,27,2207-2208. Bolognese, A.; Liberatore, R.; Riente, G.; Scherillo, G.; J. Heterocyclic Chem., 1988,25, 1247-1250. Ishii, T.; Iwahashi, H.; Sugata, R.; Kido, R.; Arch. Biochem. Biophys., 1992, 294, 616-622. Kotake, M.; Sakan, T.; Senoh, S.; J. Am. Chem. Soc, 1951, 73, 1832-1834. Butenandt, A.; Schiedt, U.; Biekert, E.; Liebigs Ann. Chem., 1954, 588, 106-116. Butenandt, A.; Schiedt, U.; Biekert, E.; Cromartie, R.J.T.; Liebigs Ann. Chem., 1954, 590, 75-90. Candy, D.J.; Kilby, B.A.; Insect Biochemistry and Function, Chapman and Hall: London, 1975. Ishii, T.; Iwahashi, H.; Sugata, R.; Kido, R.; Free Radical Biol. Med, 1992, 13, 17-20. Eastman, C.L.; Guilarte, T.R.; Brain Res., 1989, 495, 225-231. Eastman, C.L.; Guilarte, T.R.; Neurochem. Res., 1990, 15, 1101-1107. Ogawa, T.; Matson, W.R.; Beal, M.F.; Myers, R.H.; Bird, E.D.; Milbury, M.S.; Saso, S.; Neurology, 1992, 42, 1702. Pearson, S.J.; Reynolds, G.P.; Neurosci. Lett., 1992, 144, 199-201. Guilarte, T.R.; Eastman, C.L.; J. Neurol. Sci., 1993, 116, 227-228.

1024

[96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [Ill] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129]

Okuda, S.; Nishiyama, N.; Saito, H.; Katsuki, H.; J. Neurochem., 1998, 70, 299307. Okuda, S.; Nishiyama, N.; Saito, H.; Katsuki, H.; Proc. Natl. Acad ScL USA, 1996,93, 12553-12558. Buttke, T.M.; Sandstrom, P.A.; Immunol. Today, 1994, 15, 7-10. Ratan, R.R.; Murphy, T.H.; Baraban, J.M.; J. Neurochem., 1994, 62, 376-379. Portera-Cailliau, C; Hedreen, J.C; Price, D.L.; Koliatsos, V.E.; J. Neurosci., 1995, 15,3775-3787. Dragunow, M.; Faull, R.L.; Lawlor, P.; Beilharz, E.J.; Singleton, K.; Walker, E.B.; Mee, E.; Neuroreport, 1995, 6, 1053-1057. Eastman, C.L.; Guilarte, T.R.; Lever, J.R.; Brain Res., 1992, 584, 110-116. Aruoma, O.I.; Halliwell, B.; Hoey, B.M.; Butler, J.; Free. Radic. Biol Med., 1989, 6, 593-597. Brunet, P.C.; Coles, B.C.; Proc. R. Soc. Lond B Biol. ScL, 1974, 187, 133-170. Brunet, P.C; Endeavour, 1976, 26, 68-74. Manthey, M.K.; Pyne, S.G.; Truscott, R.J.W.; Tetrahedron, 1989,45, 2803-2810. Benavente, M.G.; Truscott, R.J.W.; Arch. Biochem. Biophys., 1991, 290, 451457. Truscott, R.J.W.; Faull, K.; Augusteyn, R.C.; Ophthalmic Res., 1977, 9, 263-268. Truscott, R.J.W.; Martin, F.; Exp Eye Res., 1989, 49, 927-940. Van Heyningen, R. In The Human Lens in Relation to Cataract, Elsevier, Science, B.V.: Amsterdam, 1973; pp. 151-171. Zigman, S.; Paxhia, T.; Exp. Eye Res., 1988, 47, 815-824. Luthra, M.; Balasubramanian, D.; Exp. Eye Res., 1992, 55, 641-643. Truscott, R.J.W.; Wood, A.M;. Ophthalmic Res., 1994,26, 214-218. Dillon, J.; Atherton, S.J.; Photochem. Photobiol, 1990, 51,465-468. Zigman, S.; Schultz, J.B.; Yulo, T.; Grover, D.; Isr. J. Med. Sci., 1970, 8, 15901598. Aquilina, J.A.; Carver, J.A.; Truscott, R.J.W.; Exp. Eye Res., 1997, 64, 727-735. Dillon, J.; Spector, A.; Nature, 1976, 259, 422-423. Stutchbury, G.M.; Truscott, R.J.W.; Exp. Eye Res., 1993, 57, 149-155. Wood, A.M.; Truscott, R.J.W.; Exp Eye Res., 1993, 56, 317-325. Wood, A.M.; Truscott, R.J.W.; Vision Res., 1994, 34, 1369-1374. Dillon, J.; Lens Res., 1983, 1, 133-145. Ellozy, A.R.; Wang, R.H.; Dillon, J.; Photochem. Photobiol., 1994, 59, 479-484. Dillon, J.; Doc. Ophthalmol., 1994, 88, 339-344. Tomoda, A.; Yoneyama, Y.; Yamaguchi, T.; Shirao, E.; Kawasaki, K.; Ophtalmic /?g5., 1990,22, 152-159. Malina, H.Z.; Martin, X.D.; Graefes Arch. Clin. Exp Ophthalmol., 1995, 233, 38-44. Truscott, R.J.W.; Augusteyn, R.C.; Biochim. Biophys. Acta, 1977, 492, 43-52. Christen, S.; Peterhans, E.; Stocker, R.; Proc. Natl, Acad Sci. USA, 1990, 87, 2506-2510. Wadano, A.; Takenaka, Y.; Matsumoto, M.; Aldrichina grahami. Insect Biochem. Mol. Biol., 1993, 23, 919-925. Zigler, J.S.J.; Goosey, J.; TIBS, 1981, 5, 133-136.

1025

[130] Bessems, G.J.H.; Keizer, E., Wallensak, J.; Hoenders, H.J.; Invest. Ophthalmol. Vis.ScL,l9H7,2S, 1157-1163. [131] Zigman, S.; Surv. Ophthalmol., 1983, 27, 317-325. [132] Taylor, H., West, S.; Rosenthal, F.; N.E.J.M., 1988, 319, 1429-1433. [133] Wegener, A.R.; Doc. Ophthalmol, 1994, 88, 221-232. [134] Linetsky, M.; Ortwerth, B.J.; Photochem. PhotobioL, 1995, 62, 87-93. [135] Atherton, S.J.; Dillon, J.; Gaillard, E.R.; Biochim. Biophys. Acta, 1993, 1158, 7582. [136] Harding, J.J.; Biochem. J., 1970, 117, 957-960. [137] Kamei, A.; Biol. Pharm. Bull, 1993, 16, 870-875. [138] Manthey, M.K.; Pyne, S.G.; Truscott, R.J.W.; J. Org. Chem., 1990, 55, 45814585. [139] Manthey, M.K.; Pyne, S.G.; Truscott, R.J.W.; Proc. Natl. Acad. Sci. USA, 1992, 89, 1954-1957. [140] Ishii, T.; Iwahashi, H.; Sugata, R.; Kido, R.; Free Rad. Res. Commun., 1991, 14, 187-194. [141] Boyland, E.; Watson, G.; Nature, 1956, 177, 837-838. [142] Allen, M.J.; Boyland, E.; Dukes, C.E.; Homing, E.S.; Watson, J.G.; Br. J. Cancer, 1951, 11,212-228. [143] Dunning, W.F.; Curtis, M.R.; Maun, M.E.; Cancer Res., 1950, 10, 454-459. [144] Boyland, E.; Williams, D.C.; Biochem. J., 1956, 64, 578-582. [145] Abul-Fadl, M.A.M.; Khalaeallah, A.S.; Brit. J. Cancer, 1961, 15, 479-482. [146] Clatson, D.B.; Jull, J.W.; Bonser, J.M.; Brit. J. Cancer, 1958, 12, 222-226. [147] Pryor, W.A.; Annu. Rev. Physiol., 1986, 48, GSl-eei. [148] Kaur, H.; Halliwell, B. In Methods in Enzimology, Tabor, H. and Tabor, C.W. Eds.; Academic Press: London, 1994, Vol. 233, pp. 67-82. [149] Kawanishi, S.; Inoue, S.; Sano, S.; J. Biol. Chem., 1986, 261, 5952-5958. [150] Battle, C ; Verity, M.A.; J. Neurochem., 1981, 36, 1308-1310. [151] Nishizuka, Y.; Hayaishi, O.; J. Biol. Chem., 1963, 238, 3369-3377. [152] Speciale, C ; Ungerstedt, U.; Schwarcz, R.; Neurosci. Lett., 1989, 104, 345-350. [153] Speciale, C ; Hares, K.; Schwarcz, R.; Brookes, N.; J. Neurosci., 1989, 9, 20662072. [154] Christen, S.; Thomas, S.R.; Garner, B.; Stocker, R.; J. Clin. Invest., 1994, 93, 2149-2158. [155] Esaki, H.; Onozaki, H.; Kawakishi, S.; Osawa, T.; J. Agric. Food Chem., 1996, 44, 696-700. [156] Thomas, S.R.; Witting, P.K.; Stocker, R.; J. Biol. Chem., 1996, 271, 3271432721. [157] Matsuo, M.; Nakamura, N.; Shidoji, Y.; Muto, Y.; Esaki, H.; Osawa, T.; J. Nutr. Sci. VitaminoL, 1997, 43, 249-259. [158] Matsuo, M.; Biosci. Biotech. Biochem., 1997, 61, 1968-1972. [159] Fuchs, D.; Hausen, A.; Huber, C ; Magreiter, R.; Reibnegger, C ; Spielberger, M.; Wachter, H.; Hoppe-Seyler's Z. Physiol. Chem., 1982, 363, 661-664. [160] Huber, C ; Fuchs, D.; Hausen, A.; Magreiter, R.; Reibnegger, C ; Spielberger, M.; Wachter, H.; J. Immunol, 1983, 130, 1047-1050.

1026

[161] Werner, E.R.; Lutz, H.; Fuchs, D.; Hausen, A.; Ruber, C; Niederwieser, D.; Pfleiderer, W.; Reibnegger, G.; Troppmair, J.; Wachter, H.; Biol Chem. Hoppe 5'^j;/^r, 1985,366,99-102. [162] Ames, B.N.; Cathcart, R.; Schwiers, E.; Hochstein, P.; Proc. Natl, Acad. ScL USA, 1981, 78, 6858-6862. [163] Stocker, R.; Yamamoto, Y.; McDonagh, A.F.; Glazer, A.N.; Ames, B.N.; Science, 1987,235, 1043-1046. [164] De Lange, R.J.; Glazer, A.N.; Anal. Biochem., 1989, 177, 300-306. [165] Stocker, R.; Glazer, A.N.; Ames, B.N.; Proc. Natl, Acad. Sci. USA, 1987, 84, 5918-5922. [166] Frei, B.; Stocker, R.; Ames, B.N.; Proc. Natl, Acad Sci. USA, 1988, 85, 97489752. [167] Cadenas, E.; Simic, M.G.; Sies, H.; Free Radical Res. Commun., 1989, 6, 11-17. [168] Matsuo, M.; Nippon Nogeikagaku Kaishi, 1989, 63, 1765-1770. [169] Matsuo, M.; Nippon Nogeikagaku Kaishi, 1992, 66, 899-904. [170] Matsuo, M.; Biosci. Biotech. Biochem., 1996, 60, 571-574. [171] Gyorgy, P.; Murata, K.; Sugimoto, Y.; J. Am. Oil Chem. Soc, 1974, 51, 377-379. [172] Gyorgy, P.; Murata, K.; Ikebata, H.; Nature, 1964, 203, 870-872. [173] Bursh, W.; Paffe, S.; Putz, B.; Barthel, G.; Schultze-Hermann, R.; Carcinogenesis, 1990, 11,847-853. [174] Waksman, S.A.; Woodruff, H.B.; Proc. Soc. Exp. Biol. Med., 1940, 45, 609-615. [175] Waksman Conference on Actinomicyns: Their potential for cancer chemotherapy. Cancer Chemother. Rep., 191 A, 58, 1. [176] Sivak, A.; Meloni, M.L.; Nobili, F.; Katz, E.; Biochim. Biophys. Acta, 1962, 57, 283-289. [177] Katz, E.; Welssbach, H.; J. Biol. Chem., 1963,238, 666-675. [178] Mason, K.; Shaw, G.J.; Katz, E.; Arch. Biochem. Biophys.,1977, 180, 509-513. [179] Katz, E. In Antibiotics II; Gottlieb, D. & Shaw, P.D., Eds.; Springer-Verlag: New York, 1967, p. 276. [180] Troost, T.; Katz, E.; J. Gen. Microbiol., 1979, 111, 121-132. [181] Keller, U.; J. Biol. Chem.,19S4, 259, 8226-8231. [182] Jones, G.H.; J. Bacteriol., 1987, 169, 5575-5578. [183] Keller, U.; Kleinkauf, H; Zocker, R.; Biochemistry,\9M, 23, 1479-1484. [184] Keller, U.; J. Biol. Chem.,l9S7, 262, 5852-5856. [185] Choy, H.A.; Jones, G.H.; Arch. Biochem. Biophys., 1981, 211, 55-65. [186] Jones,G.H.;Hopwood,D.A.; J. Biol. Chem., 1984,259, 14151-14157. [187] Freeman, J.C.; Nayar, P.G.; Begley, T.P.; Villafranca, J.J.; Biochemistry, 1993, 32, 4826-4830. [188] Morgan, L.R.; Weimorts, D.M.; Biochem. Biophys. Acta, 1964, 82, 645-646. [189] Subba Rao, P.V.; Jegannathan, N.S.; Vaidyanathan, C.S.; Biochem. Biophys. Res. Commun., 1964, 16, 145-149. [190] Subba Rao, P.V.; Jegannathan, N.S.; Vaidyanathan, C.S.; Biochem. J., 1965, 95, 628-632. [191] Madhusudanan Nair, P.; Vaidyanathan, C.S.; Biochim. Biophys. Acta, 1964, 81, 507-516. [192] Subba Rao, P.V.; Vaidyanathan, C.S.; Biochem. J., 1966, 99, 317-322.

1027

[193] Subba Rao, P.V.; Vaidyanathan, C.S. In Methods in Enzimology\ Tabor, H. and Tabor, C.W. Eds.; Academic Press: London, 1970, Vol. 77, pp. 554-559. [194] Nair, P.M.; Vining, L.C.; Can. J. Biochem., 1964, 42, 1515-1525. [195] Nair, ?M.\Arck Biochem. Biophys. 1972, 153, 139-146. [196] Goldberg, I.H.; Amer. 1 Med., 1965, 39, 722-731. [197] Chance, B.; Sies, H.; Boveris, A.; Physiol Rev., 1979, 59, 527-605. [198] Mayer, A.M., Harel, H.; Phytochemistry, 1979, 18, 193-196. [199] Milstein, O.; Nicklas, B.; Hutterman, A.; Appl. Microbiol. BiotechnoL, 1989, 31, 70-74. [200] Messerschmidt, A.; Huber, R.; Eur. J. Biochem., 1990, 187, 341-352. [201] Eggert, C ; Temp, U.; Dean, J.F.D.; Eriksson, K.E.L.; FEBS Lett., 1995, 376, 202-206. [202] Eggert, C ; Temp, U.; Eriksson, K.E.L.; Appl. Environ. Microbiol, 1996, 62, 1151-1158. [203] Oda, Y.; Adachi, K.; Aita, I.; Ito, M.; Aso, Y.; Igarashi, H.; Agric. Biol Chem., 1991,55,1393-1395. [204] Esposito, E.; Innocenti-Mei, L.H.; Ferraz, A.; Canhos, V.P.; Duran, N.; J. Biotechnol, 1993, 29, 2\9-22S. [205] Gripenberg, J.; Acta Chem. Scand., 1951, 5, 590-593. [206] Cavill, G.W.K.; Ralph, B.J.; Tetaz, J.R.; Werner, R.L.; J. Chem. Soc, 1953, 525529. [207] Sullivan, G.; Henry, E.D.; J. Pharmacol Sci., 1971, 60, 1097-1098. [208] Cavill, G.W.K.; Clezy, P.S.; Tetaz, J.R.; Werner, R.L.; Tetrahedron, 1959, 5, 275-280. [209] Ander, P.; Eriksson, K.E.L.; Yu, S.; Arch. Microbiol, 1983, 136, 1-6. [210] Ander, P.; Hatakka, A.; Eriksson, K.E.L.; Arch. Microbiol, 1980, 125, 189-202. [211] Buswell, J.A.; Eriksson, K.E.L.; Gupta, J.K.; Hamp, S.G.; Nordh, I.; Arch. Microbiol, 1982, 131, 366-374. [212] Lundell, T.; Leonowicz, A.; Rogalski, A.; Hatakka, A.; Appl. Environ. Microbiol, 1990, 56, 2623-2629. [213] Eggert, C ; Microbiol Res., 1997, 152, 315-318. [214] Eggert, C ; Temp, U.; Dean, J.F.D.; Eriksson, K.E.L.; FEBS Lett., 1996, 391, 144-148. [215] Temp, U.; Eggert, C ; Appl Environ. Microbiol, 1999, 65, 389-395. [216] Rheinhammar, B.; Biochim. Biophys. Acta, 1972, 275, 245-254. [217] Xu, F.; Shin, W.S.; Brown, S.H.; Wahleithner, J.A.; Sundaram, U.M.; Solomon, E.I.; Biochim. Biophys. Acta, 1996, 1292, 303-311. [218] Tien, M.; Kirk, T.K.; In Methods in Enzymology, Tabor, H. and Tabor, C.W. Eds.; Academic Press: London, 1988, Vol. 161, pp. 238-249. [219] Bourbonnais, R.; Paice, M.G.; FEBS Lett., 1990, 267, 99-102. [220] Bourbonnais, R.; Paice, M.G.; Reid, I.D., Lanthier, P.; Yaguchi, M.; Appl Environ. Microbiol, 1995, 61, 1876-1880. [221 ] Mucke, M. (December 1994) World patent application WO 94/29510. [222] Novo Nordisk A/S (January 1995) World patent application WO 95/01426. [223] Young, S.N.; Curzon, G.; Biochem. J., 1972, 129, 273-283. [224] Harris, E.D.; Proc. Soc. Exp. Biol Med., 1991, 196, 130-140.

1028

[225] Sonesson, B.; Eide, S.; Rorsman, H.; Rosengren, E.; Acta Derm. Venereal., 1995, 75, 283-286. [226] Toussaint, O.; Lerch, K.; Biochemistry, 1987, 26, 8567-8571. [227] Cooksey, C.J.; Garrat, P.J.; Land, E.J.; Pavel, S.; Ramsden, C.A.; Riley, P.A.; Smit, N.P.M.; J. Biol. Chem., 1997, 272, 26226-26235. [228] Burton, S.G.; Catal. Today, 1994, 22,459-487. [229] Sanchez-Ferrer, A.; Rodriguez-Lopez, J.N.; Garcia-Canovas, F. Garcia-Carmona, F.; Biochim. Biophys. Acta, 1995, 1247, 1-11. [230] Rescigno, A.; Sanjust, E.; Soddu, G.; Rinaldi, A.C.; Sollai, F.; Curreli, N.; Rinaldi, A.; Biochim. Biophys. Acta, 1998, 1284, 268-276. [231] Pomerantz, S.H.; Warner, M.C.; J. Biol. Chem., 1967,242, 5308-5314. [232] Kahn,; V. J. FoodBiochem., 1990, 14, 177-188. [233] Cabanes, J.; Garcia-Canovas, F.; Lozano, J.; Garcia-Carmona, F.; Biochim. Biophys. Acta, 1987, 923, 187-195. [234] Ros-Martinez, J.R.; Rodriguez-Lopez, J.N.;Castellano, R.V.; Garcia-Canovas, F.; Biochem. J., 1993, 294, 621-623. [235] Sohn, N.; Gang, H.; Gumport, S.L.; Deppisch, L.M.; Cancer, 1969, 24, 897-903. [236] Jergil, B.; Lindbladh, C ; Rorsman, H.; Rosengren, E.; Acta. Derm. VenereoL, 1983, 63, 205-208. [237] Karg, E.; Hultberg, B.; Isaksson, A.; Rorsman, H.; Rosengren, E.; Acta. Derm. VenereoL, 1990, 70, 286-290. [238] Prota, G.; Melanins andMelanogenesis, Academic Press:San Diego, 1992. [239] Fitzpatrick, T.B.; Brunet, P.C.J.; Kukita, A.; in The Biology of Hair Growth, W. Montagna, Ed.; Academic Press: New York, 1958, pp. 255-303. [240] Chen, Y.M.; Chavin, W.; Anal. Biochem., 1965, 13, 234-258. [241] Nicolaus, R.A.; Melanins, Herman, Paris, 1968, pp. 30, 73-74. [242] De Antoni, A.; Allegri, G.; Costa, C ; Gazz. Chim. ItaL, 1970, 100, 1039-1049. [243] Allegri, G.; De Antoni, A.; Baccichetti, F.; Costa, C ; Gazz. Chim. ItaL, 1972, 102,426-430. [244] De Antoni, A.; Allegri, G.; Costa, C ; Bordin, F.; Experientia, 1974, 30, 600-601. [245] Chakraborty, A.K.; Chakraborty, C.P.; Int. J. Biochem., 1993, 25, 1277-1280. [246] Hegedus, Z.L.; Nayak, U.; Arch. Int. PhysioL Biochim. Biophys., 1993, 101, 417423. [247] Hegedus, Z.L.; Kuttab, S.H.; Altschule, M.D., Nayak, U.; Arch. Int. PhysioL Biochim. Biophys., 1981, 89, 393-398. [248] Hegedus, Z.L.; Nayak, U.; Arch. Int. PhysioL Biochim. Biophys., 1994, 102, 311313. [249] Hegedus, Z.L.; Nayak, U.; Arch. Int. PhysioL Biochim. Biophys., 1993, 101, 289295. [250] Hegedus, Z.L.; Nayak, U.; Arch. Int. PhysioL Biochim. Biophys., 1994, 102, 175181.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 26 © 2002 Elsevier Science B.V. All rights reserved.

1029

NEW CONTRIBUTIONS TO THE STRUCTURE ELUCIDATION AND PHARMACOLOGY OF STRYCHNOS ALKALOIDS P. RASOANAIVO\ *M-T. MARTIN^ E. GUITTET^ AND F. FRAPPIER' ^Laboratoire de Phytochimie et de Pharmacologie Cellulaire et Parasitaire, Institut Malgache de Recherches Appliquees, B.P. 3833, 101'Antananarivo, Madagascar. ^Laboratoire de RMN, Institut de Chimie des Substances Naturelles, CNRS, 91198 - Gif-sur-Yvette, France. ^Laboratoire de Chimie des Substances Naturelles, ESA 8041 CNRS, Museum National d'Histoire Naturelle, 63 rue Buffon, 75231 Paris Cedex 05, France. ABSTRACT: This paper gives first an outline of the taxonomy of indigenous and endemic Strychnos species of Madagascar and describes their ethnopharmacology with focus on their ethnomedical uses as herbal remedies to enhance chloroquine action during the sudden emergence of malaria in Madagascar in 1980's. This led to the discovery of the chemosensitizing activity of the new Nb-C(21) secocuran Strychnos alkaloids. The alkaloid content of Strychnos species are reviewed and their chemotaxonomy briefly described. The second part of the paper reports new NMR strategies for structure elucidation, exemplified by structure analysis of monoindole and bisindole alkaloids published during the period 1994-2000. The third part of the paper reports the biological activities of Strychnos alkaloids reported over the same period with emphasis on the chemosensitizing effect of the Nb-C(21) secocuran alkaloids.

INTRODUCTION The word Strychnos was first coined by the Greeks to term toxic Solanaceae species, then used by the Swedish botanist Linne to name dangerous and toxic plants with reference to two typical species, S. nuxvomica and S. colubrina. [1, 2]. This toxic property originally stimulated intensive research on Strychnos species, leading to the successful isolation of strychnine (1) from Asian Strychnos and curare alkaloids from South-American Strychnos. For a long time, the tetanising or the curarizing activities dominated the research devoted to Strychnos.

1030 However, as investigation progressed, ethnobotanical and pharmacological works have revealed that alkaloids of Strychnos species display a broad spectrum of biological activities [3, 4]. This has attracted the interests of phytochemists over the last three decades as evidenced by the continued publication of several review articles devoted to this series of alkaloids [5, 6, 7]. While 150 years were necessary to completely elucidate the structure of 1, structure determination has considerably benefited from the progress of NMR techniques, especially ^H-^H and ^H-^^C 2D-NMR experiments. Based on their particular structures, the Strychnos alkaloids have been the focal point of considerable spectroscopic efforts. More recently some selected Strychnos alkaloids, 1 and brucine (2), and their N-oxide derivatives, have been used successfully as model compounds for the application of the ^^N NMR methods [8, 9].

1

strychnine

R1=R2=H

2

brucine

Rl=R2=0Me

12

2-hydroxy-3-methoxystrychnine Rl=OH;R2=OMe

One characteristic feature that has stimulated research on Madagascan Strychnos is the use of their decoction in association with chloroquine to reinforce the action of this standard antimalarial drug in the treatment of chronic malaria [10], This has resulted in the isolation of a new sub-type of Strychnos indole alkaloids, the Nb-C(21)-secocuran group. These alkaloids, endowed with a resistance-modulating activity in Plasmodium malaria, have been isolated hitherto from endemic and indigenous Strychnos of Madagascar. This paper will focus, after comments of the taxonomy, on ethnopharmacology and alkaloid contents of Madagascan Strychnos, on recent advances in the structure elucidation of Strychnos indole alkaloids for the period 1994-2000, exemplified by our own results. One section will summarize the relevant pharmacological data on Strychnos alkaloids for the same period with particular emphasis on the chemosensitising activity of the secocuran alkaloids.

1031

TAXONOMY, ETHNOPHARMACOLOGY AND ALKALOID CONTENTS OF MADAGASCAN STRYCHNOS. The Strychnos genus, the largest genus of the Loganiaceae family, is pantropical and comprises nearly 200 species. Apart S. potatorum, all species are geographically segregated in the three continents: 75 species in Africa including Madagascar, 73 in America and 44 in Asia. Several review articles have been devoted to the ethnobotany, pharmacology and alkaloids contents of the African, Asian and South-American Strychnos [3, 4 and references cited herein]. Although Madagascar is currently considered as part of Africa, its flora has evolved in comparative isolation from main land species since the island was separated from Africa some 140 million years ago and has acquired a pronounced individuality. This long isolation together with culture differences, different perception of diseases and ways to treat them in various ethnic groups, have influenced the traditional knowledge of herbs and plants in Madagascar. In this first section of the present paper, we stress this specificity in terms of ethnomedical uses and alkaloids contents of the indigenous and endemic Strychnos. Then we give useful suggestions for future work directed to these Strychnos species. To start, we summarize in Table 1 up-dated data on the distribution of Madagascan Strychnos within the taxonomic sections, their geographical distribution, their medicinal uses retrieved from computerised database [11] and their alkaloid contents according to the geographical origin of the species investigated [12-36]. Taxonomy The genus Strychnos is classified in the Strychneae tribe of the Loganiaceae family and is further sub-divided into twelve sections. Fourteen species are present in Madagascar, of which four are endemic to the island and ten shared in common with Africa [37].

1032

TABLE 1. Outline of taxonomy, ethnomedical uses and alkaloid contents of Strychnos species of Madagascar AbbreviatJons: LV, leaves; SB, stem barks; ST, stem; RB, root bark; RT, roots; SD, seeds; FR, fruits. Sections

Rouhamon

Breviflorae

Species (geographical distribution)

Ethnomedical uses Plant parts, origin: alkaloid contents [ref.]

S. decussata (Pappe) Gilg (Eastern and Southern Africa, Madagascar)

Tea (RT): externally appUed to treat inflammation in wound and abscesses ; Infusion (LV) for stomach complaints, for malaria prophylaxis or treatment as chloroquine substitute.

Madaeascan origin, LV: gluco-alkaloid [12]; SB: rouhamine (5,6-dehydro-decussine), decussine, 3,14dihydro-decussine, malindine, macusine B, 10hydroxy-3,14-dihydro-decussine, 0-methyl-macusine B, macusine A or C, akagerine, akagerine lactone, 170-methyl-akagerine, 10-hydroxy-akagerine, 10hydroxy-17-0-methyl-akagerine, 10-hydroxy-17-0methyl-kribine, 10-hydroxy-epi-17-(9-methyl-kribine, bisnordihydrotoxiferine [13-15].

S. floribunda Gilg (Western and Central Africa, Madagascar)

NO REPORT

Ivory Coast origin. LV: angustine [16]; SB: akagerine, decussine, strychnocarpine, deacetylisoretuline, rouhamine, bisnordihydro-toxiferine, isorosibiUne, [17,18].

5. potatorum Linne f. (Eastern and Southern Africa, India, Sri-Lanka, Birmanie, Madagascar)

NO REPORT

African, Asian origin. LV: angustine, angustidine [16]; Tanzanian, Indian origin. LV, SB, SD: diaboline, Acetyl-diaboUn Zairean origin. RB: harmane carboxamide, cantleyine, 18-19 dihydroe [19]; -usambarensine, polyneuridine, norharmane, akuanmiidine, Nor-C-fluorocurarine, ochrolifuanine A, bisnordihydro-toxiferine, ochroUfuanine E, normacusine B, normavacurine, henningsamine, 11methoxy-henningsamine, 19,20-dihydro-longicaudatin, 19,20-dihydro-longicaudatin Y, Antirhine, (20R)- and (20S)-dihydro-antirhine, 11 -methoxy-12-hydroxydiabohne, diaboUne, 11-methoxy-diaboline, deacetylretuline, diabohne A^-oxide [20].

5. henningsii Gilg (Congo, Angola, Eastern and Southern Africa, Comoros, Madagascar)

Tea (LV) for malaria; Tea (SB) for stomach complaints and vomiting. Tea (FR) for dysentery and vomiting.

Zairean origin. SB: holstiine, holstiline, rindhne, retuline, condensamine, diabohne, henningsamine, Henningsohne, 0-acetyl-henningsohne, 11-methoxydiaboline, 2,16-dehydro-diabohne, 2,16-dehydro-l 1methoxy-diaboline [21-25]. Madaeascan origin. LV: SD: tsilanine, 10-methoxy-tsilanine, 0-demethyltsilanine, 0-demethyl, 10-methoxy-tsilanine [26]. Madaeascan origin. Twigs: tsilanimbine, N^deacetyl-isoretuline, N(a)-acetyl-l 1-methoxystrychnosplendine, 18-hydroxy-isoretuline, N^deacetyl, 18-hydroxy-isoretuhne, N^-deacetyl-18hydroxy-17-methoxy-isoretuUne [27].

1033

Zairean origin, LV: 0-acetyl-retuline [28], Tanzanian origin. RB: holstiine, splendoline, 23hydroxy-spermostrychnine, 19-epi-23-hydroxyspennostrychnine, retuline, henningsiine, deshydroxyacetyl-henningsiine, 0-acetyl-henningsiine, 3-hydroxy-henningsiine, henningsiine-N(4)-oxide, diaboline, 23-hydroxy-spermostrychnine -N(4)-oxide, 17,23-dihydroxy-spennostrychnine, cyclostrychnine, spennostrychnine, henningsamide, 0-acetylhenningsamide, deshydroxyacetyl-henningsamide [29].

Breviflorae (continued)

Kenian origin: RB: diaboline, holstiine, Na-acetylstrychnosplendine, Na-acetyl-11-methoxystrychnosplendine [30]. S. mitis S. Moore No report (Eastern Africa, South Africa, Comoros, Madagascar)

Spinosae

S. spinosa Lamarck (Tropical Africa, South Africa, Madagascar, Seychelles, Mascareignes)

Tea (LV, ST) for blennorrhagia; Infusion (LV) externally apphed for skin diseases, especially scabies.

S. panganensis Gilg

No report

S. madagascariensis Poiret (Eastern and Southern Africa, Madagascar)

Infusion (LV, ST) for malaria; Tea (ST) for veneral diseases

Lanigerae

Densiflorae

South African origin: RB: cantleyine, tetrahydrocantleyine, strychnovoline, tubotaiwine, Onicotinoyl-7-tetrahydrocantleyine, 16(S)-Eisositsirikine, 16(R)-E-isositsirikine, tubotaiwine; SB: cantleyine, tetrahydrocantleyine, strychnovohne, tubotaiwine; LV: cantleyine, tetrahydrocantleyine, O-nicotinoyl-7tetrahydrocantleyine, strychnovoline, tubotaiwine, tubotaiwine N-oxyde [31], Congolese origin. RB: cantleyine, tetrahydrocantleyine, 4-carbomethoxy-l-methyl-2,7naphtyridine, 5-carbomethoxy-4-methyl-2,7naphtyridine, scaevodimerine, O-nicotinoyl-7tetrahydrocantleyine, 3-carbomethoxy-5-(lhydroxyethyl)-pyridine, 3-carbomethoxy-4-methyl-5(1 -hydroxyethyl)-pyridine, 11 -methoxy-diaboline, 11 methoxy-12-hydroxy-diaboUne, 11 -methoxyhenningsamine, 12-hydroxy-l 1-methoxyhenningsamine [31]. Nigerian origin. LV, SB: akagerine, lO-hydroxyakagerine, kribine, 11-methoxy-diaboline [32]; Tanzanian origin. RB: matopensine, 16-(S)-E-isosotsirikine, 12-hydroxy-l 1-methoxy-diaboline, N-deacetyl-retuline, N-deacetyl-isoretuhne, N-deacetyl-spermostrychnine, 12-hydroxy-11 -methoxy-nor-C-fluorocurarine, 12-hydroxy-11 -methoxy-N-acetyhior-C-N-acetyl-norC-fluorocurarmine, panganensines 19'R and 19'S, panganensines X and Y. South African origin. LV: 16(S)-E-isositsirikine, 9methoxy-16(S)-E-isositsirikine, 9-methoxy-16(R)-Eisositsirikine, Normacusine B, 16(R)-E-isositsirikine. SB, SR: 16(S)-E-isositsirikine, 9-methoxy-16(S)-Eisositsirikine, 9-methoxy-16(R)-E-isositsirikine, Normacusine B, 16(R)-E-isositsirikine [31].

1034 S. bifurcata Leeuwenberg (Endemic to Madagascar)

No report

Not investigated

5. diplotricha Leeuwenberg (Endemic to Madagascar)

Infusion (ST) for malaria in association with chloroquine; tea (ST) externally applied for wound, leprosy

Madagascan origin. SB: strychnobrasiline, strychnofendlerine, malagashanine, 3-ep/-myrtoidine, 11 -demethoxy-12-hydroxy-3-^/7i-myrtoidine, myrtoidine, 11-demethoxy-myrtoidine, 11-demethoxy3-^/?/-myrtoidine [34].

5. mostueoides Leeuwenberg (Madagascar and probably Congo)

Infusion (ST) for malaria in association with chloroquine

Madagascan origin. SB: strychnobrasiline, malagashanine, malagashine, W.G. aldehyde, strychnofendlerine, Normacusine B, deacetylstrychnobrasiline, spermostrychnine [35].

S. myrtoides Gilg & Busse (Tanzania, Mozambique, Madagascar)

Infusion (ST, LV) for malaria in association with chloroquine. Tea (LV) for colic.

Madagascan origin. SB: strychnobrasiline. malagashanine, malagashanol, 12-hydroxymalagashanine, 12-hydroxy-19-^pi-malagashanine, myrtoidine, 11-demethoxy-myrtoidine [36]; LV: strychnobrasiline, malagashanine, 3-epimyrtoidine, 11 -demethoxy-12-hydroxy- 3 -epimyrtoidine, 11-demethoxy-3-^p/-myrtoidine [34].

5. pentantha Leeuwenberg (Endemic to Madagascar)

No report

Not investigated

S. trichoneura

No report

Madagascan origin. LV: angustine, angustidine, angustoline [16].

Penicillatae

Leeuwenberg (Endemic to Madagascar)

Ethnopharmacology Medicinal plants form the basis of traditional or indigenous health system used by the majority of the people in developing countries. Schultes defined ethnopharmacology as the observation, identification, description and experimental investigation of the ingredients and the effects of indigenous drugs [38]. Today, ethnopharmacology is also associated with the need for conservation and sustainable uses of herbal resources. The ethnomedical data on Madagascan Strychnos previously reported [39] are outlined here in a simplified manner. From Table 1, it appears that there are no ethnomedical reports on seven species, suggesting that either they

1035 are toxic or rare. However, one characteristic feature is the use of six species of indigenous and endemic Strychnos of Madagascar as herbal remedies to combat malaria. Three species, namely S. decussata, S. myrtoides and 5. diplotricha are used as chloroquine substitutes for both prophylactic and curative treatment of malaria. Three other species, 5. henningsii, S. mostueoides and 5. madagascariensis are used only as curative herbal antimalarials. While some Strychnos species are used merely as curative antimalarials in other continents, ethnomedical data in Madagascar include specific uses as prophylactic agents to protect against malaria. They all contain alkaloids. From this observation and in line with current belief in Madagascar, the degree of antimalaria activity of a given medicinal plant has been linked to its bitterness [40]. Crude alkaloid extracts of these six Strychnos species are found to be of moderate antiplasmodial activity in the in vitro experimental model, with an inhibitory concentration (IC50) ranging from 8 to 60 /xg/ml (Table 2). In the same conditions, chloroquine, as standard antimalarial drug, has an IC50 ranging from 0.01 to 0.1 jLtg/ml depending on the susceptibility of the Plasmodium strain used. TABLE 2. In vitro antiplasmodial activity of crude tertiary alkaloids from 6 Strychnos species of Madagascar traditionally used to treat malaria. Botanical names

Plan part

IC50 (lig/ml)

References

S. myrtoides

Stem barks

57.6

10

S. decussata

Leaves

19.7

Unpublished results

S. mostueoides

Stem barks

8.5

S. madagascariensis

Stem barks

12.8

41

S. tienningsii

Stem barks

16.9

Unpublished results

S. diplotricha

Stem barks

58.3

34

Unpublished results

Discrepancies between traditional knowledge and their confirmation in experimental models turn out to be a major problem in drug discovery involving medicinal plants. We suggest some plausible explanations as a basis for future work in Strychnos species and other plant species dealing with malaria. A first explanation is that these alkaloids have nothing to do with antiplasmodial activity. Their claimed antimalarial activity may be due to other biological effects including antipyretic, anti-inflammatory and

1036 immunomodulatory actions. It is also possible that the efficacy of Strychnos treatment is modulated by pre-existing immunity malaria. A second explanation is that the biological tests are inappropriate to confirm the activity. Indeed, Strychnos species used as prophylactic herbal antimalarials may act on the hepatic stage of the malaria parasites. To the best of our knowledge, very little is known about plant extracts that may inhibit the hepatic stage of Plasmodium malaria although data exist in the ethnomedical literature. Plant extracts have been screened hitherto for their in vitro/in vivo antiplasmodial activity in the erythrocytic stage of the malaria parasites. Furthermore, mechanismbased assays in malaria including glutathion reductase inhibition [42], haem polymerization inhibition and protease inhibition [43] are all designed to find active extracts/compounds that act in the asexual stage of the malaria parasites. As in vitro/in vivo tests to assess the effect of plant extracts on the sporozoite stage of malaria parasites may be available for routine screening, this opens a new route for the evaluation of herbal antimalarials as prophylactic agents, especially those claimed to be efficient for combating malaria but found to be ineffective in the in vitro/in vivo antiplasmodial tests. Further studies on the Strychnos species of Madagascar are needed to exploit this possibility. A third explanation is that the method of extraction fails to isolate the active ingredients. This can be illustrated by the following observations. Complex indole alkaloids, including glucoindole alkaloids, are derived from secologanin and tryptophan. These alkaloids are restricted to a few iridoid containing families namely Loganiaceae, Rubiaceae and Apocynaceae [44]. Two species of the Rubiaceae family, Nauclea latifolia and Pauridiantha (= Urophyllum) lyalii , widely used to treat malaria respectively in Central and West Africa [45] and in Madagascar [40], contain glucoindole alkaloids [46-50] and unusual trinitrogenated alkaloids [51-56]. The use of anmionia in the extraction procedure and the ease with which dihydrovincoside lactam is converted into dihydroangustine (3) [57], raises the possibility that the trinitrogenated alkaloids might be artefacts (Scheme 1). Polar compounds appear to be the active constituents of Nauclea latifolia [58]. S. decussata contains the only glucoindole alkaloid isolated from Strychnos of Madagascar [12], together with trinitrogenated alkaloids. It is possible that similar glucoindole alkaloids may be present in other Strychnos species of Madagascar and may be responsible for their reported antimalarial activity. These hydrophilic compounds

1037 deserve further attention as antimalarials. This stresses also the importance of using decoction of medicinal plants as they are used in real life by healers in preliminary screening instead of more elaborate methods of extraction.

3

OGlc

dihydroangustine

Scheme 1

Going back to the antimalarial activity of the six Strychnos species, the chemosensitising effect of some alkaloid extracts appeared to be hitherto the most important finding in this series of alkaloids. The ethnomedical story behind this discovery dates back to the 1980^s. For unknown reasons, malaria re-emerged in the Central Highlands in the 1980s as the most devastating of the country's tropical diseases. Local populations have given the name bemangovitra (disease of great shivering) to this previously unknowned infection. It is unfortunate that this event was poorly documented from an epidemiological point of vue. The only relevant publication on the subject reported a dramatic increase in Plasmodium falciparum resistance to chloroquine in two regions of Madagascar [59]. Statistics published by the Ministry of Health provided an estimated 15,000 deaths/year for the period 1984-1988. It is likely that significantly more deaths occured than in these statistical data since only records from hospital facilities and primary health care system were maintained. The bemangovitra phenomenon is still unexplained but might have resulted from selection of resistant parasites following an uncontrolled increase in use of chloroquine by the local population. Vendors on the roadside or in the grocery stores sold chloroquine pills one at a time, without doctors or prescriptions involved. The pills are usually favoured over traditional remedies because their effect is quick.

1038 After only a few doses however, most people either run out of money or felt better and then stopped taking pills. After the scenario repeats itself a few times, their bodies become factories for drug-resistant parasites. Ignoring all facts about the selection of drug-resistant parasites due to the misuse of chloroquine, some people believed that bemangovitra was an unnatural disease and consulted traditional healers. Therefore they went back to the massive use of herbal remedies as attachment of traditional culture. In this particular context, we learned for the first time of the use of sub-curative doses of chloroquine tablets in association with decoction of medicinal plants for the treatment of bemangovitra. Three Malagasy Strychnos species, namely 5. mostueoides, S. myrtoides and S. diplotricha, were used to enhance chloroquine action in different regions of Madagascar by culturally different ethnic groups. They all showed significant chloroquine enhancing action in the experimental models. This will be commented upon in the section devoted to biological activities. It is worthwhile to note that local populations in Madagascar had used empirically drug reversal combination to overcome drug resistance in malaria before the advent of the rational approach of chemosensitization. While current practices in herbal remedies include the association of two or several plants to treat a disease, it is very rare to find data in which medicinal plants are combined with western medicines for the same purpose. The only work that we are aware is the use of Aloe vera juice to enhance the action of glibenclamide in the treatment of diabetes mellitus [60]. To the best of our knowledge, Boiteau pioneered this practice in Madagascar by using association of quinine and extracts of Burasaia madagascariensis for the treatment of chronic malaria during World War II [61]. The bemangovitra event has disappeared and consequently the empirical recipe to treat it tends to be forgotten by the local populations. As ethnopharmacology is a dynamic process, the documentation of ethnomedical data in a specific context is therefore of particular relevance [62]. Alkaloid contents If all alkaloids isolated from Strychnos of different geographic origins are taken into consideration, about 120 alkaloids of high chemical diversity have been identified from indigenous and endemic Strychnos of Madagascar. It should be pointed out however that, based on our on-

1039 going investigation and taking into account the plant parts investigated, discrepancies have been observed in the alkaloid constituents of some Strychnos species collected in Madagascar and in Africa. One typical example is Strychnos henningsii collected in Madagascar [26] and in Tanzania [20]. Assuming that botanical authentication was correct, ecological factors such as neighbouring plant species, seasonahty, climatic and soil conditions, play a decisive role in the biosynthesis of these alkaloids. The effect of ecological factors on chemical constituents is well illustrated by the example of Cinnamomum camphora. Essential oils from the Asian species contain nearly 70% of camphor while the species growing in Madagascar does not contain any camphor but has a unique chemical composition with strong antimicrobial and antiviral activities, justifying its large scale export [63]. It is therefore difficult to draw any chemotaxonomic conclusions from the present alkaloid survey. In the opinion of the authors, indigenous Strychnos of Madagascan origin should be phytochemically re-investigated for their alkaloid constituents. Only in that way can a good picture of Madagascan Strychnos alkaloids be drawn and their chemotaxonomy determined. One of the prominent phytochemical results that have come over the past five years is the isolation of the Nb-C(21) secocuran alkaloids from two indigenous Strychnos species of Madagascar, 5. myrtoides and 5. mostueoides, and one endemic species, S. diplotricha [36]. To the best of our knowledge, they have been isolated only from Strychnos of Madagascar. Here also, it is particularly important to investigate 5. myrtoides from Tanzania or Mozambique and S. mostueoides from Congo to know whether the Nb-C(21) secocuran alkaloids are present in species of African origin or restricted only to species of Madagascan origin. From a chemotaxonomy point of view, malagashanine, the parent compound of the Nb-C(21) secocuran alkaloids, is shared in common with three species of the Penicillatae section, namely S. diplotricha, S. mostueoides and S. myrtoides. This compound appears to be a key alkaloid of this section. The alkaloid constituents of S. trichoneura fit better in the Rouhamon section than in the Penicillatae section. 5. bifurcata and S pentantha are both endemic to Madagascar and remain to be phytochemically and biologically investigated. For the purification of alkaloids, the counter-current distribution technique in a Craig apparatus using chlorinated solvents as stationary phase and a buffer at discontinuously decreasing pH as mobile phase appears to be one of the

1040

most useful methods that have been applied successfully to the separation of several Strychnos alkaloids [64]. Furthermore, new and powerful NMR techniques, which are the subject of the next section, are available for rapid and unambiguous structure elucidation. NEW NMR STRATEGIES IN STRUCTURE ELUCIDATION NMR has become a standard tool for structure determination and, in particular, for these of Strychnos alkaloids. The last general article in this field was authored by J. Sapi and G. Massiot in 1994 [65] and described the advances in spectroscopic methods applied to these molecules. More recently, strychnine (1) has even been used to illustrate newly introduced experiments [66]. We conmient, here, on their advantages and sum up the principles of usual 2D experiments in Fig. (1) and Fig. (2) (COSY : correlation Spectroscopy, TOCSY: TOtal Correlation Spectroscopy, NOESy : Nuclear Overhauser Enhancement Spectroscopy, ROESy: Rotating frame Overhauser Enhancement Spectroscopy, HMQC: Heteronuclear Multiple Quantum Coherrence, HMBC : Heteronuclear Multiple Bond Correlation). This section updates two areas of research in the field: new ^H and ^^C NMR experiments with gradient selection or/and selective pulses, ^^N NMR, and microspectroscopy. To take these data into account, another section comments on the structure elucidation of new compounds isolated from Strychnos. It covers the literature from 1994 to early 2000. SigniHcant improvements in NMR spectroscopy New experiments in ^H and ^^C NMR. Gradient selection

The development of NMR experiments with gradient selection has proven to be of fundamental importance for homo- and heteronuclear 2D applications. One of the main problems encountered in classical 2D spectroscopy is the differentiation between wanted and unwanted coherences.

1041

COSY experiment detects coupled pairs of protons usually separated by two or three bonds (V, ' / ) .

TOCSY experiment can, in principle, gives a total correlation of all protons of a chain with each other.

dipolar coupling

NOESY, ROESY yield correlation signals which are caused by dipolar cross-relaxation between nuclei in a close spatial relationship.

Fig.(l). The principles of usual 2D homonuclear experiments, (bold lines represent typical correlations obtainable through the corresponding experiments)

1042

HMQC experiment yields cross signals for all protons and '^C nuclei which are connected by a •^C,H coupling over one

Me H Q

bond (VH-13C)..

-^H HMBC experiment yields cross signals for C,H spin pairs connected by two or three bond couplings (or more in conjugated p electron systems).

H ^ V ^

^ N

f A

H'*v

/ / ^ ,

OH

^^

HMQC-TOCSY experiment is a combinaition of HMQC and TOCS Y experiments OJ'^^C-H'^JH-H)

Fig. (2). The principles of usual 2D heteronuclear experiments, (bold lines represent typical correlations obtainable through the corresponding experiments)

1043

For example in a COSY, TOCSY, NOESY or ROESY experiment, one has to achieve frequency discrimination in Fi, and eliminate axial signals. In standard HMQC and HMBC experiments, one has to distinguish between protons bound to ^^C and ^^C. Time-consuming phase cycling previously performed these tasks and with standard 2D experiments a number of scans higher than four was usually required. The use of pulsed field gradients represents a new way to achieve all these tasks [67]. The selection of the desired coherences already happens in the probe-head and, usually, only one single transient (or two for NOESY) with no phase cycling is sufficient to record high quality data. Since the NMR receiver now detects only the desired signals, its gain can be set much higher. Therefore ^H, ^^C correlation experiments using pulsed field gradients are performed in a fraction of the time formerly needed, provided the available quantity of material is sufficient. Moreover, any instability during an experiment now takes on decreased importance due to its reduced total length. These instabilities are among the sources of '7i noise". For the spectroscopist, this noise from strong signals (solvent, methyl groups) produces "stripes" in the F\ domain of a 2D experiment, possibly preventing the observation of less intense crosspeaks. Gradient selection experiments are not only less time-consuming than phase cycling experiments but they are also cleaner. For all these reasons the special hardware required for gradient experiments (a special probe-head with self-shielded gradient coils and a gradient amplifier), is a must for structural analysis of natural products. In the following discussion, gs (gradient selection) in front of the name of experiments, indicates the use of this technique. Use of selective pulses [66]

Instead of recording the full 2D matrix, one "row" can be simply measured by replacing the first 90° pulse of a 2D experiment with a soft pulse, thus looking only for spin couplings that affect the particular proton excited. A further advantage over 2D techniques is the high digital resolution. Furthermore the ID experiment can be stopped by operator intervention whenever the signal-to-noise ratio is sufficient for a particular aspect. Prior to performing this kind of experiments, a certain amount of calibration and preparation has to be performed. They thus

1044 cannot be used as simply as routine ID or 2D standard pulse sequences. They require the skills of an experienced spectroscopist. The selective COSY method yields the same connectivity information as the homonuclear decoupling technique. In contrast to the latter, however, the multiplets due to the coupling partners are still visible and can be easily evaluated. Because this is a ID experiment, it can be performed at high resolution. The method works well even when the Jcoupled spin overlaps with other signals, but the selectively pulsed signal must be well resolved[68]. In the same way, gs-SELTOCSY (Selective Total Correlation Spectroscopy), also called HOHAHA (Homonuclear Hartmann-Hahn), gives a total correlation of all the protons of a chain with each other [66]. Heteronuclear long-range correlations are invaluable for investigating the connectivity, configuration and conformation. However, the cross-peak multiplets appear with complex phase properties in both dimensions and the extraction of coupling constants is not straightforward. Difficulties arise because the long-range proton coupling VHH is of the same order of magnitude as VCH- In this case selective ID sequences are useful alternatives. Stelten and Leibfritz [69] describe semi-selective ID multiple-bond VCH correlation with ^H detection. A normal ^H spectrum is compared with the resultant spectrum. The difference in signal multiplet width yields the active VCH coupling constant. The traditional nuclear Overhauser difference spectroscopy can be replaced by DPFGSE-NOE (Double Pulsed Field Gradient Spin Echo method-NOE spectrometry) experiment. The desired NOE effects can be recorded without interference from other signals and without phase distortion [66]. Three dimensional NMR spectroscopy and improved pulse sequences

Three dimensional techniques should only be applied in case of overlap in optimally recorded 2D spectra. The basic principle of these experiments is simply to merge two 2D experiments. The addition of a third frequency domain increases the spectral resolution, and gives additional information. The 3D HMQC-COSY spectrum, for example, consists of a set of 2D ^H-^H COSY maps separated according to the ^^C chemical shift along the third frequency dimension. Thus moving across ^^C, ^H and ^H, ^H planes makes unequivocal assignments possible [66].

1045 An interesting alternative to three dimensional NMR techniques is to suppress one of the evolution times while retaining the basic 3D pulse sequence. The spectral resolution is no longer increased, which is usually not a problem with smaller molecules, but the extra information is still available. The gs-HMQC-TOCSY experiment represents one such experiment. The combination of the HMQC method with the TOCSY sequence leads, in principle, to a 3D technique. However, if the ti evolution period of the TOCSY part is omitted, a 2D sequence is obtained which provides a ^^C edited TOCSY spectrum. Starting from each HMQC cross-signal, one finds in the same row additional signals which are caused by a TOCSY transfer. When the structure elucidation is difficult, this experiment can fruitfully complement the HMBC experiment. This last experiment, which gives long range through bond connectivities, is crucial for the structure elucidation. A significant improvement of this experiment was recently proposed in the form of the ACCORD-HMBC (ACCORDion-HMBC) experiment. This experiment has two distinct advantages over the standard gs-HMBC. It uses a dual stage low pass filter to suppress more effectively all V correlation signals. In addition it uses the ACCORDION principle to sample a range of V coupling constants, thus more correlation signals will appear as compared with the HMBC method with a fixed polarisation delay. Besides the usual ti acquisition time, in ACCORDION spectroscopy there are two other different time variables [70]. The HMBC delay decreases with increasing t\ time, thus keeping the overall delay after the first excitation and final detection to a minimum [71]. But in this experiment, ^H-^H coupling modulation introduces an Fi modulation or a 'skew" of responses in the second frequency domain. A new experiment, IMPEACH-MBC (Improved Performance Accordion Heteronuclear Multiple-Bond Correlation), suppresses this artefact [72]. Both the ACCORD-HMBC and IMPEACH-HMBC experiments provide much better experimental access to four-bond long-range couplings which are important in the elucidation of molecules having proton deficient regions in their structures. Now a simple, general and pratical method is available for direct measurement of V(CH) coupling constants [73]. ID and 2D heteronuclear single quantum multiple bond correlation (HSQMBC) experiments are easily performed in a routine manner. These experiments are based on the evolution of heteronuclear single quantum coherence (SQC) in

1046

combination with a trim pulse and a gradient-enhanced heteronuclear z,zfilter to destroy unwanted magnetization, the spectra are devoid of phase complications due to the evolution of the ^H- H homonuclear couplings and provide pure absorption, antiphase multiplets for the long-range ^H^^C heteronuclear correlations. ^^NNMR.

There has been burgeoning interest in this topic. A growing body of reports has appeared using gradient long-range ^H-^ N heteronuclear chemical shift correlation experiments at natural abundance. The first of those appeared in 1995 [9] and a review has just been published this year [74]. Their use should rapidly increased because of hardware improvements (submicrotechniques) and because of the additional structural information they provide. These points will be discussed in the following paragraphs. Submicro techniques

*^N spectra were usually acquired using sample quantities in the range of 20-40 mg (-0.1 mmole) in 5 mm NMR probes with a typical acquisition time ranging from a few hours to overnight. These experiments were performed using 500 or 600 MHz spectrometers. G.E. Martin and coworkers [75, 76, 77] recently reported results obtained using a 1.7 nmi submicro inverse-detection gradient or SMIDG NMR probe. Data can be acquired overnight on -3 |Limole of compound (-1 mg) but acquiring the data over a weekend yields higher quality data. High quality correlations can be observed on 1 |xg of strychnine (1) with the combined utilisation of gradient 2D and selective ID techniques. A 2D experiment, with a fast acquisition of a coarsely F\ digitised spectrum, allows the definition of the ^^N chemical shifts. Afterwards, the overnight acquisition of a selective long-range ID ^H-^^N GHMBC (Gradient Heteronuclear Multiple Bond Correlation) spectrum affords high-quality, long-range correlation data [77]. Recently these authors have compared the performance characteristics for 3 nrai gradient micro inverse probes with those for 1.7 nrni SMIDG NMR probes [78]. The results obtained with the 1.7 mm submicro probe were ~ 2.3 fold better in signal to noise ratio

1047 than with the 3 mm micro probe. The authors point out that sample handling at the micromole level and below requires a considerable degree of care to insure that high-quality data can be obtained from these experiments. Long-range couplings

G.E. Martin and co-workers recently demonstrated the possibility of exploiting long-range ^H-^^N coupling as a structural probe. Their analysis can be performed by using a gradient-enhanced variant of the HMQC pulse sequence. The experiment affords the ability to correlate ^H to ^^N chemical shifts either through one-bond (VNH), or through multiple bonds (VNH) couplings, obviating the need for a second experiment to establish the chemical shifts of protonated nitrogen resonances. Wide ^^N chemical shift dispersion and the limited number of nitrogens contained in most molecules are such that the overlap of direct and long-range correlations is very unlikely. The relative intensity of the correlations can be severely modulated as a function of the optimisation of the long-range coupling delay [79]. This technique requires several experiments optimised for different coupling values: 2, 5 and 10 Hz, for example. Often, it is very difficult to observe long-range coupling when dealing with "ortho'' coupling from aromatic proton to a nitrogen not contained in the same ring, as in the case of the correlation between H-4 and N-9 of 1 [75]. Recently, Koshino and Uzawa [80] have demonstrated the feasability of observing four-bond long-range correlations to nitrogen in heteroaromatic systems by optimisation of the long-range delays in the ^H-^^N GHMBC experiment for couplings of < 2 Hz. In conclusion, the ACCORD-HMBC experiment seems to be very useful for detecting long-range couplings of different sizes [71]. The values of the long-range ^H-^^N couplings can give significant informations on geometrical parameters. Martin and Crouch [81] note that the orientation of a proton capable of a VNH coupling with a given nitrogen relative to the lone electron-pair of that nitrogen plays a key role in the intensity of the responses observed. A weak crosspeak correlating protons to the N-6' in the velbanamine subunit of navelbine (4) can be observed within about a weekend of data acquisition while the response to N-16' is observable within about 4 hours as do those N-1 and N-9 in the vindoline subunit of 4. Much larger VNH couplings are expected when

1048

the proton engaged in a two-bond coupling is synclinal with the nitrogen lone pair [9]. velbanamine subunit

6

vinorelbine

MeO

1049 Three-bond long-range ^H-^^N couplings obey a Karplus relationship with maxima near 0° and 180^ [81]. TABLE 3.

^^NNMRdata N-9

N-19

Ref.

Strychnine 1

148.0 149.1

35.0 35.1

9,77 76

strychnine N-oxide

146.5

136.3

Brucine 2

151.0

37.0

bnicine N-oxide

145.3

135.5

N-1

N-4

146.5

39.5

Holstiine 5

9

N-1

N-9

N-6'

N-16'

Navelbine 4

66.0

55.3

43.0

138.2

81

Vinorelbine 6

66.0

55.3

43.0

138.2

79

Chemical shifts are reported downfield from Hquid ammoniac.

Chemical shift analysis : the example of A^-oxidation

Despite their wide potential, *^N chemical shifts are still rather rarely used for structure determination. ^^N NMR data of few Strychnos alkaloids are given in Table 3. Though a few natural compounds have been isolated as A^-oxides [82, 83] no structure elucidation was done with ^^N NMR. This technique would have clearly been of interest as demonstrated by the following study. ^^N NMR studies were performed on 1 and 2 following conversion to their corresponding N-19 oxides using hydrogen peroxide [8]. Upon A^-oxidation, the N-19 resonance is shifted downfield by about 100 ppm. The N-9 resonance undergoes a slight upfield shift of -1.5 ppm. This difference is minimal and may be attributed to concentration and/or slight temperature differences rather than electronic or conformational changes. There remains a relative lack of information regarding ^^N chemical shift of iminoxides. So we have studied differences between N-1

1050

chemical shift of 1,2-didehydroaspidospermidine (7) and the oxydation compound 8 obtained from strychnobrasiline (9) with pchloroperbenzoic acid (F. Trigalo, personal communication).

II 144.2 283.0

1 T O

7

1,2-didehydroaspidospenmdine

compound

I 153.4

8

/ H

V"

Fig.(3) 15N and ^^C chemical shift

The structural characterisation of the two compounds was achieved through the concerted analysis of COSY, NOESY, HMQC, and HMBC data. The VN,H long range coupling was optimised for 10 Hz. We observed a upfield shift for both ^^N and ^^C resonances (Fig. (3)) as previously observed by Yavari and Roberts [84]. An examination of N data shows that replacement of the lone pair of electrons on the nitrogen atom of pyridine by a bond to another atom leads to an upfield shift of the ^^N resonance. The magnitude of ^^N-shielding change for A^oxidation of pyridine is about -23 ppm. Structure elucidation of new compounds

Though all these experiments have been applied to indole alkaloids, they are not yet commonly used. Some of these (COSY, TOCSY, HMQC, HMBC) have allowed: -The acquisition of complete ^H and ^^C correlation spectral data for known compounds such as 9 isolated from Strychnos myrtoides [85], bisnordihydrotoxiferine (10) and longicaudatine (11) and the main tertiary indole alkaloids previously isolated from Strychnos, atlantica [86],

1051

9

strychnobrasiline

10

11

longicaudatine

bisnordihydrotoxiferine

13

isostrychnine R1=R2=H; X=N

14

isobrucine

15

isostrychnineN-oxide Rl =R2=H; X=NO

16

isobrucineN-oxide Rl =R2=0Me; X=NO

Rl=R2=0Me;X=N

1052

-The elucidation of the structure of new natural compounds Monoindole alkaloids:

Sixteen alkaloids have been isolated from the seeds of Strychnos vomica L [82]. This journal (Acta Pharmaceutica Sinica) being difficult to access, we report in Table 4 the ^^C NMR chemical shifts for 2-hydroxy3-methoxystrychnine (12), isostrychnine (13), isobrucine (14), isostrychnine-A^-oxide (15) and isobrucine-A^-oxide (16) which are new compounds. TABLE 4.

^^C NMR data for compounds 12,13,14,15 and 16.

c

12->

1 2 3 4 5 6 7 8 10 11 12 13 14 15 16 17 18 20 21 22 23 OMe OMe

108.3 147.1 143.4 100.7 134.3 123.8 51.5 59.9 170.0 50.0 77.0 47.8 31.1 26.2 59.5 41.8 49.4 52.1 139.3 123.8 64.3 55.8

(90 MHz)s a): In CDCI3-CD3OD.

13'-> 120.5 124.3 128.3 114.6 142.3 134.8 52.3 67.5 168.5 46.3 120.5 141.1 34.7 25.9 63.5 36.9 52.9 54.0 137.7 126.5 58.0

b): In CDCI3.

14 "> 105.8 146.2 149.1 99.5 135.2 125.1 52.4 67.6 167.7 45.6 120.5 137.4 34.7 25.7 62.9 36.7 52.7 53.9 142.2 126.7 57.9 56.1 56.5

15-> 124.7 126.9 131.6 116.7 142.9 130.5 52.9 66.8 171.2 44.3 125.0 140.8 34.5 25.2 81.2 44.9 68.4 68.7 133.3 136.8 59.4

16'^ 108.6 149.2 152.0 101.5 136.7 125.4 53.3 67.7 170.7 44.3 124.0 141.8 34.9 25.1 81.5 42.6 69.8 69.9 132.5 135.1 59.4 57.9 57.5

c): In CD3OD.

Twelve alkaloids, seven of which are new, have been isolated from the root bark of Strychnos panganensis collected in Tanzania. The novel monoindole alkaloid are A^-desacetylspermostrychnine (17), 12-hydroxy11 -methoxy-nor-c-fluorocurarine (18), 12-hydroxy-11 -methoxy-A^acetyl-nor-C-fluorocurarimine (19) [33].

1053 Phytochemical investigation of the minor alkaloids of Strychnos myrtoides resulted in the isolation of four new alkaloids of the Nb-C(21) secocuran series, viz.,myrtoidine (20), 11-demethoxymyrtoidine (21), which showed an a-|3 unsaturated y-lactonic ring hitherto not found among Strychnos alkaloids, 12-hydroxy-19-epi-malagashanine (25) and malagashanol (26). Except for 26, the other Nb-C(21) secocuran alkaloids described have an uncommon H-3P configuration [36].

MeO

17

18

12-hycIroxy-11 -methoxy-nor-Cfluorocurarine R1=H; R2=CHO

19

12-hydroxy-11 -methoxy-N-acetyl-nor-Cfluorocurarimine Rl=Ac; R2=CHNH

N-desacetylspermostrychine

o<^^^

20

myrtoidine

21

11 -demethoxymyrtoidine R1 =R2=H

22

11 -demethoxy-12-hydroxymyrtoidine R1=H;R2=0H

23

malagashanine R1=H; R2=C02CH3; 3bH; 19,20aH

24

12-hydroxy-maIagashamne R1=0H; R2=C02Me; 3bH; 19,20aH

25

12-hydroxy-l 9-epi-nialagashanine R1=0H; R2=C02Me; 3bH; 19bH; 20aH

2(5

malagashanol R1=H; R2=CH20H; 3aH;19aH;20bH

Rl=OMe; R2=H

1054

A suggested biogenetic pathway which can explain the formation of the Nb-C(21) secocuran alkaloid and the cooccurence of alkaloids epimer at C-3 and C-20 is reported in Fig.(4). Me

malagashanine derivatives Fig.(4). Suggested biogenesis for Nb-C(21) secocuran alkaloids. Reproduced with permission of Elsevier Science [36].

1055 Compound 9, by electrophilic attack of proton at C-20 and subsequent hydrolysis of the resulting unstable ammonium ion with fission of the NbC(21) bond, gives an aldehyde group and by rearrangement a ahydroxypyrrolidinic ring. The new ammonium ion, resulting from the attack of a proton with the elimination of a water molecule, gives rise by non stereoselective reduction to the H-a and H-P secocuran alkaloids. ^H and ^^C NMR spectra of strychnobrasiline (9) illustrate the peculiar structure of the Nb-methyl-sec-pseudo series of Strychnos alkaloids. From the ^H NMR spectrum, the H-21 olefinic proton appears as a singlet deshielded at 6 ppm. This is explained by an intramolecular exchange interaction between the double bond and the C-3 carbonyl. Furthermore in the ^^C spectrum, the C-3 carbonyl signal is greatly shielded (192.5 ppm) due to a non-bonded interaction between Nb and C3. These structural features explain the poor reactivity of the C-20, C-21 double bond which does not undergo typical reactions of the neostrychnine double bond. Consequently, an unexpected molecular rearrangement in the lithium aluminium hydride reduction of (9) was observed [87]. Bisindole alkaloids

The complete stereostructure of divarine (27), a new bisindole alkaloid from Strychnos divaricans has been determined, by use of 2D NMR techniques [88]. The structural determination of guianensine (28), an alkaloid isolated from Strychnos guianensis was described. The alkaloid has a zwitterionic structure with the negative and positive charges on the two parts of an asymetrical bis-indole alkaloid [89]. Recently, a reinvestigation performed by the same authors resulted in the isolation of guiaflavine (29), a new quaternary bisindole [90]. Lastly, beside monoindole, four dimeric indole alkaloids of the reticuline-diaboline type have been isolated from the root bark of Strychnos panganensis: two couple of stereoisomers on C-19' or C-19 named: panganensine R (30) and S (31), and panganensine X (32) and Y (33) respectively [33]. It has not been possible to assign the C-19 configuration for 32 and 33.

1056

OMe 27

divarine

28

guianansine

OMe 29

guiaflavine

1057

30

32

panganensine R

panganensine X

31

33

panganensine S

panganensine Y

Though NOESY spectroscopy has proved to be a useful method for studying the spatial proximity of protons and for the investigation of the stereochemistry of molecules [33, 36, 85], some authors propose assigning relative configurations based on biogenetic grounds and on coupling constant values [89, 90]. This method is not always reliable [36]. NOESY spectra are also very useful for studying the conformational equilibrium which can occur in N-acetylindoline alkaloids. This conformational equilibrium is due to the presence of restricted rotation around the amide. At room temperature, it often is in slow exchange compared to the NMR time-scale as found in 9 [85] Fig.(5). At -40°C, the two conformations are frozen and no interconversion is observed in

1058

(a)

(b)

II

A-^

AJV^LAA->i_

-j—I—I—I—I—|—I—I—p—I—J—I—I—I—\—(—I—(—I—I—j—I—»—1—I—I—I—I—I—I—I—>—I—I—I—

7.0

6.0

5,0

4.0

3.0

2.0

PPM

the phase-sensitive NOESY spectrum. The ratio of Z and E isomers has been estimated to be 70:30. Fig.(5). IH NMR spectrum of strychnobrasiline 9 recorded at (a) - 4 0 and (b) 27°C. Reproduced with permission of John Wiley & Sons [85].

In each case the acetyl carbonyl oxygen deshielded the proton towards which it is directed: H-12 for the Z isomer and H-2 for the E isomer [85]. On the other hand, a NOESY experiment allows to point out the change of the conformation of the cyclohexane D ring from chair conformation (24) to boat conformation (25) when the C-19 relative configuration is inverted. So Me-18 is always in an equatorial conformation Fig. (6) [36].

1059

,Me C-19

. ^ ^

Me

Fig.(6). Conformation of compounds 24 and 25 deduced from NOES Y spectra.

BIOLOGICAL ACTIVITIES Although several hundred alkaloids have been isolated from Strychnos species, little work has been devoted to the evaluation of their biological activities. For the last five years, the biological results, in which preliminary structure-activity relationships can be drawn, include anticancer, antiprotozoal and chemosensitizing activities. They are reviewed here with focus on structure-bioactivity relationships whenever possible.

1060

Anticancer activity

Two types of Strychnos alkaloids have been evaluated for their potential as anticancer drugs. These include the monoindole alkaloids whose structure comprises a harmane skeleton 34, Fig. (7) and the bisindole alkaloids of 5. usambarensis, Fig. (8). Matadine (35) isolated from 5. grossweilleri, serpentine (36) found in Rauwolfia serpentina as well as in S. grossweilleri or S. camptoneura, together with the non-Strychnos alkaloid cryptolepine (37) from Cryptolepis sanguinolenta (Asclepiadaceae), were evaluated for their ability to interact with DNA and topoisomerase II [91]. These alkaloids were found to bind tightly to DNA and to behave as typical intercalating agents. Compound 37 was shown to bind 10-fold more tightly, and to be more cytotoxic toward B16 melanoma cells than were the other two alkaloids. Probably the planar harmane skeleton component of 35 and 36 is a structure requirement for DNA intercalating activity. Strychnopentamine (38), a bisindole alkaloid isolated from 5. usambarensis, was found to exhibit powerful cytotoxicity on B16 melanoma and on non-cancer human fibroblasts cultured in vitro [92]. Subsequently it was shown to be active in vivo on an Ehrich ascites tumour growing in the mouse [93]. Inhibition of RNA synthesis is believed to be a possible mechanism of action of (38) [94]. However, there appears to be no report on structure-activity relationships in this series of alkaloids. Antiprotozoal activity

Seven monoindole Strychnos alkaloids, namely matadine (35), serpentine (36), alstonine (39), melinonine F (40), normelinonine F (41), 5,6dihydroflavopereirine (42) and strychnoxanthine (43) as well as the hemisynthetic Nb-methyharmalane, together with cryptolepine (37) as reference compound, were tested for their in vitro antiplasmodial activity [95]. All alkaloids were found to have little activity compared to the reference compound. The strong DNA intercalating activity of 37 is believed to be responsible for its high degree of antiplasmodial activity [96]. The moderate intercalating activity of the harmane-derived alkaloids may thus explain their weak antiplasmodial action.

1061

34

harmane

35

matadine

r^^^ Me

36

serpentine

39

alstonine (20-epi-serpentine)

40

melinonine F

41

normelinonine F R=H

^^^^^^, 37

cryptolepine

42

5,6-dihydroflavopereirine

43

strychnoxanthine

Me

R=Me

Fig.(7). Structure of Strychnos monoindole alkaloids evaluated for their anticancer and/or antiprotozoal acrtivites. Structure of cryptolepine and ellipticine are given for comparison purpose.

1062

r^^^N

38

strychnopentamine 45

usambarensine

44

5',6'-dihydrousambareiisine

r^^^

^

46

usambarine

47

18,19-dihydrousambarine

^

Fig.(8) Structure of Strychnos bisindole alkaloids evaluated for their anticancer and/or antiprotozoal activities.

1063 These alkaloids appear not to be promising candidates for further development as anticancer or antimalarial drugs. Several Strychnos alkaloids were evaluated for their antiprotozoal activities [97, 98]. The bisindole alkaloids of 5. usambarensis appear to display the most interesting effects in terms of potency and selectivity, strychnopentamine (38) and 5',6'-dihydrousambarensine (44) were found to be the most active in vitro against Plasmodium falciparum but were inactive in vivo against P, berghei. Usambarensine (45), usambarine (46) and 18,19-dihydrousambarine (47) were shown to exhibit high activity against Entamoeba histolyca in vitro while 44 was found to be the most selective against Giardia intestinalis. Although selectivity was observed in connection with antiprotozoal activities, no clear relationships allowing to associate some structural features with an increase in activity was evidenced. Chemosensitising activities

The resistance reversing activity of Madagascan Strychnos alkaloids was discovered as a scientific follow-up of their empirical use as herbal remedies to enhance chloroquine action during the sudden resurgence of malaria in 1980's. By bioassay-guided fractionation, two alkaloids, the Nb-C(21) secocuran alkaloid malagashanine (23) and the Nb-C(3) secocuran alkaloid strychnobrasiline (9), were found to be the major bioactive constituents [99]. They exhibited a weak in vitro antiplasmodial activity but when combined with chloroquine at dose levels much lower than required for antiplasmodial activity, they significantly enhance chloroquine action against resistant strains of Plasmodium malaria. One characteristic feature that distinguishes these two alkaloids was their in vivo activity: 23 was shown to be active in vivo when combined with subcurative doses of chloroquine whereas 9 was devoid of any in vivo activity even at concentrations approaching the toxic doses [99]. The strong lipophilic property of 23 was claimed to play an important role in its in vivo activity. It is possible that poor bioavailability of 9 may contribute to the lack of in vivo activity. Since the pioneering work of S.K. Martin and co-workers with verapamil [100], drugs that reverse chloroquine resistance attracted much attention in the 1990's [101]. However, the lack of effect of the two powerful synthetic chemosensitisers, desipramine and cyproteptadine, in

1064 clinical studies has somewhat decreased interest in this class of therapeutic drugs [102, 103]. In the opinion of the authors, naturallyoccurring chemosensitisers, especially those having ethnobotanical backgrounds, deserve special attention as biochemical tools that might contribute to the understanding of chloroquine resistance and its reversal and also as an alternative therapy to overcome chloroquine resistance. At this point, pharmacological results obtained with 23 suggest that this may be a promising compound [99] but further investigations are still needed to clarify its mechanism of activity compared to verapamil. TABLE 5. Interaction factors (IF) and relative reversal percentage of naturallyoccurring and hemisynthetic test alkaloids Test alkaloids

IF

Strychnobrasiline 9

200

Reversal %*

_

Myrtoidine 20

4.6

12.5

1 l-demethoxymyrtoidine 21

9.0

24.3

11.1

30.0

1 l-demethoxy-12-hydroxymyrtoidine 22 Malagashanine 23

9.1

24.5

Malagashanol 26

16.7

44.9

5.1

14.6

Na-deacetylstrychnobrasiline 49

37.0

100.0

Na-deacetyl-Na-methylstrychnobrasiline50

Strychnofendlerine 48

37.0

100.0

Compound 51

2.1

5.6

Compound 52

3.6

9.6

Compound 53

2.0

5.5

Spermostrychnine 54

9.1

24.5

3-methoxyspermostrychnine 55

9.1

24.6

1 l-methoxy-12-hydroxyspermostrychnine 56

9.5

25.7

Strychnospemiine 57

9.1

24.5

Compound 58

29.0

78.4

Compound 59

5.5

14.8

*With respect to Na-deacetylstrychnobrasiline

Presently, it is premature to draw structure-activity relationships for chemosensitising activities in this series of alkaloids. Preliminary conclusions from published and unpublished work are reported here. The interaction factor (IF) has been introduced to quantify the in vitro

1065 interaction of a given compound towards chloroquine [104]. The IF equals 2, < 2 or > 2 for additive, antagonistic or synergistic effects respectively. The higher the IF value, the higher its enhancing activity on chloroquine. IF values of test alkaloids are given in Table 5. Structures of naturally-occurring test alkaloids and those of hemisynthetic test alkaloids are given in Fig (9). Thus four minor alkaloids of S. myrtoides structurally related to 23, namely malagashanol (26), myrtoidine (20), 11-demethoxymyrtoidine (21) and ll-demethoxy-12-hydroxymyrtoidine (22) were found to exhibit a significant variation of the chemosensitising activity in comparison to the parent compound. In the Nb-(C3) seco series, strychnofendlerine (48) (20,21-dihydrostrychnobrasiline) was shown to be four times less active than 9. Several hemisynthetic derivatives of 9 were also assessed for chloroquine-potentiating action [87]. Na-deacetylstrychnobrasiline (49) and the methyl derivative 50 were found to be twice more potent than the parent compound. In contrast, ring constraints in compounds 51 and 52 led to a significant decrease of activity. Furthermore, the disappearance of the nitrogenated heterocycle in compound 53 led also to a drastic loss of activity. Lastly biogenetic precursors of the Nb-C(3) and Nb-C(21) secocuran alkaloids : spermostrychnine (54), 3-methoxy-spermostrychnine (55), 11-methoxy12-hydroxy-spermostrychnine (56) and strychnospermine (57) were shown to exhibit comparable activity to 23 [105]. All this suggests that the Nb-nitrogen should be included in a rigid system and should also fulfil length requirements with respect to the indole nucleus to be active. A preliminary conclusion that stems from the critical analysis of these results is that many Strychnos alkaloids and probably the aspidosperman indole alkaloids may reverse in vitro chloroquine resistance with different degrees of activity. Instead of testing all of these alkaloids, it would be now useful to determine the minimal structure requirement for in vitro and in vivo chloroquine-enhancing activity.

1066

48

strychnofendlerine R=Ac (20,21 -dihydrostrychnobrasiline)

49

Na-deacetylstrychnohrasiline R=H

50

Na-deacetyl-Na-methylstrychnobrasiline R=Me

52

54

R=H

spermostrychnine R1=R2=R3=H

55

3-methoxyspennostrychmne Rl=OMe;R2=R3=H

56

11 -methoxy-12-hydroxyspermostrychmne Rl=H;R2=OMe;R3=OH

57

strychnospermine Rl=R3=H;R2=OMe

Fig.(9) Structure of Strychnos alkaloids and hemisynthetic derivatives tested for their in vitro chemosensitising activity. Compounds 51 to 53 and 58-59 have been obtained after traitment of strychnobrailine by LiAlH4 [87].

1067

CONCLUSION These data demonstrate the interest of the Strychnos alkaloids as pharmacological agents. They are in line with the renewed interest in active compounds stenmiing from traditional medicine. The screening of previously described compounds and the reinvestigation of their structure have certainly to be performed, along with the isolation and extensive analysis of new compounds. The efficiency of NMR has considerably benefited from its recent developments, both in terms of speed and reliability, and clearly represents a key asset towards this task. ABBREVIATIONS IC50 COSY TOCSY NOESy ROESy HMQC HMBC gs SELTOCSy HOHAHA DPFGSE NOE ACCORD IMPEACHMBC HSQMBC SMIDG GHMBC IF

= = = = = = = = = = = = = =

Inhibitory Concentration fifty correlation Spectroscopy TOtal Correlation Spectroscopy Nuclear Overhauser Enhancement Spectroscopy Rotating frame Overhauser Enhancement Spectroscopy Heteronuclear Multiple Quantum Coherrence Heteronuclear Multiple Bond Correlation gradient selection SELective TOtal Correlation Spectroscopy HOmonuclear HArtman-HAhn Double Pulsed Field Gradient Spin Echo method Nuclear Overhauser Effect ACCORDion IMproved PErformance ACcordion Heteronuclear Multiple-bond Correlation = Heteronuclear Single Quantum Multiple Bond Correlation = SubMicro Inverse-Detection Gradient = Gradient Heteronuclear Multiple Bond Correlation = Interaction Factor

ACKNOWLEDGEMENTS. Some of the work described in this paper was supported by the Centre National de la Recherche Scientifique (CNRS-France) and Aupelf-Uref under convention ARC n''X/7.10.04/Palu. The authors wish to express sincere thanks to their co-workers Drs F. Trigalo, H. Rafatro and Ramanitrahasimbola who took part in the data described in this article.

1068

We wish to thank Dr K. Gehring for reading the manuscript and for helpful english suggestions. REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

Linnaeus, C. Materia Medica, 1749. Linnaeus, C, Species Plantarum, Vol I, Stockholm, 1753. Ohiri, F. C; Verpoorte, R.; Baerheim Svendsen, A. J. EthnopharmacoL, 1983, 9, 167-223. Quetin-Leclercq, J.; Angenot L.; Bisset, N. G. J. EthnopharmacoL, 1990, 28, 1-52. Delaude, C; Delaude, L. Bull. Soc. R. Sci. Liege, 1997,66, 183-286. Bosch, J.; Bonjoch, J.; Amat, M. Alkaloids, Academic Press, 1996; Vol. 48, pp. 75-189. Sapi, J.; Massiot, G. In The Monoterpenoid Indole Alkaloids; J.E.Saxton, Ed.; John Wiley 1994; pp. 279-353. Hadden, C. E.; Kaluzny, B. D.; Robins, R. H.; Martin, G. E. Magn. Resort. Chem., 1999,37, 325-327. Martin, G. E.; Crouch, R. C ; Andrews, C. W. J. Heterocycl. Chem., 1995, 32, 1759-1766. Rasoanaivo, P.; Ratsimamanga-Urverg, S.; Milijaona, R.; Rafatro, H.; Galeffi, C; Nicoletti, M. Planta Med., 1994,60,13-16. Rasoanaivo, P. CD Compilation of Medicinal and Aromatic Plants of Madagascar, IDRC-funded project, 1999. Petitjean, A.; Rasoanaivo, P.; Razafmtsalama, J.-M. Phytochemistry, 1977,16, 154-155. Rolfsen, W.; Olaniyl, A.; Hylands, P. J. Nat. Prod. ,1980, 43, 97-101. Olaniyl, A.; Rolfsen, W. J. Nat. Prod., 1980,43, 595-597. Rolfsen, W.; Olaniyl, A.; Verpoorte, R.; Bohlin, L. /. Nat. Prod., 1981, 44, 415-421. Phillipson, J. D.; Hemingway, S.; Bisset, N. G.; Houghton, P.; Shellard, E. Phytochemistry, 1974,13, 973-978. Verpoorte, R. Planta Med., 1980, 39,236. Verpoorte, R.; loose, F. T.; Groenink, H.; Svendsen, A. B. Planta Med., 1981, 42, 32-36. Singh, H.; Kapoor, V. K.; Phillipson, J. D.; Bisset, N. G. Phytochemistry, 1975, 14,587-588. Massiot, G.; Thepenier, P.; Jacquier, M. J.; Le Men-Olivier, L.; Delaude, C. Phytochemistry, 1992,31,2873-2876. Bosly, J. J. Pharm. Belgique, 1951,6,150-177; 243-266.

1069

[22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41]

[42] [43] [44]

Janot, M. M.; Goutarel, R.; Bosly, J. C R. Acad. Sc. Paris, 1951, 232, 853858. Bisset, N. G.; Bosly, J.; Das, B. C; Spiteller, G. Phytochemistry, 1975, 14, 1411-1414. Spiteller-Friedman, M.; Spiteller, G. Ann. Chem., 1968, 712, 179-184. Grossert, J. S.; M. Hugo, J.; Von Klemperer, M. E.; Warren, F. L. J. Chem. 5oc., 77,1965,2812-2814. Sarfati, R.; Pais, M.; Jarreau, F. X. Phytochemistry, 1970, 9, 1107-1113. Koch, M.; Fellion, E.; Plat, M. Phytochemistry, 1976, 75, 321-324. Angenot, L.; Tits, M. Planta Med., 1981,41, 240-243. Massiot, G.; Thepenier, P.; Jacquier, M. J.; Renin, J; Le Men-Olivier, L.; Delaude, C. Phytochemistry, 1991, 30, 3449-3456. Chapya, W.; Wycliffe, A.; Galeffi, C.; Sperandei, M.; Msonthi, J. D.; Nicoletti, M.; Messana, I.; Marini-Bettolo, G. B. Gazz. Chim. Ital, 1983,113,113-115. Delaude, C.; Thepenier, P.; Jacquier, M. J.; Massiot, G.; Le Men-Olivier, L. Bull. Soc. R. Sci. Liege, 1992, 61,429-440. Oguakwa, J. U.; Galeffi, C, Nicoletti, M.; Messana, I.; Patamia, M.; MariniBettolo, G. B. Gazz. Chim. Ital, 1980,110, 97-100. Nuzillard, J.M.; Thepenier, P.; Jacquier, M. J.; Massiot, G.; Le Men-Olivier, L.; Delaude, C. Phytochemistry, 1996,43, 897-902. Rasoanaivo, P.; Palazzino, G.; Galeffi, C; Nicoletti, M. Phytochemistry, 2000, (submitted). Rasoanaivo, P.; Galeffi, C.; De Vicente, Y.; Nicoletti, M. Rev. Latinoamer. Quim., 1991,22, 32-34. Martin M.-T.; Rasoanaivo, P.; Palazzino, G.; Galeffi, C.; Nicoletti, M.; Trigalo, F.; Frappier, F. Phytochemistry, 1999, 57,479-486. Leeuwenberg, A. J. M. In: Flore de Madagascar et des Comores, 16T"^^ Famille, Loganiacees, H. Humbert, Ed;, Paris, 1984, pp. 70-104. Schultes, R. E. In: The Conservation of Medicinal Plants, O. Akerele, V. Heywood and H. Synge, Eds; Cambridge University Press, 1991; pp. 65-75. Rasoanaivo, P., Ratsimamanga-Urverg, S.; Frappier, F. Cahiers Sante, 1996, 6, 249-253. Rasoanaivo, P.; Petitjean, A.; Ratsimamanga-Urverg, S.; and RakotoRatsimamanga, A. J. EthnopharmacoL, 1992,37,117-127. Ratsimamanga-Urverg, S.; Rasoanaivo, P.; Le Bras, J.; Ramiaramanana, L.; Rakoto-Ratsimamanga, A.; Coulaud, J. P. Discovery and Innovation, 1991,3, 81-83. Schirmer, R. H.; Muller J. G.; Krauth-Siegel, L. Angew. Chem., 1995,34,141154. Ridley, R. G. J. Pharm. Pharmacol, 1997,49,43-48. Jensen, S. R.; Nielsen, B. J.; Dahlgren, R. Bot. Notiser., 1975, 725,148-180.

1070

[45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62]

[63] [64] [65] [66]

Kerharo, J. La Pharmacopee Senegalaise Traditionnelle: Plantes Medicinales et Toxiques, Vigot Freres: Paris, 1974. Hotellier, F.; Delaveau, P.; Pousset, J.-L. Plantes Med. Phytoter., 1977, 77, 106-108. Brown, R. T.; Chappie C. L.; Lashford, A. G. Phytochemistry, 1977, 76, 16191620. Levesque, J.; Pousset, J.-L.; Cave, A. C R. Acad. Sc. Paris, 1975, 280, 593595. Levesque, J.; Pousset, J.-L.; Cave, A. Fitoterapia, 1977,48, 5-7. Levesque, J.; Jacquesy, R.; Foucher, J. P. Tetrahedron, 1982, 38,1417-1424. Hotellier, F.; Delaveau, P.; Pousset, J.-L. Phytochemistry, 1975,14,14071409. Hotellier, F.; Delaveau, P.; Besselievre, R.; Pousset, J.-L. C R. Acad. Sc. Paris, 1976,282, 595-597. Hotellier, F.; Delaveau, P.; Pousset, J.-L. Planta Med.,1979, 35, 242-246. Pousset, J.-L.; Bouquet, A.; Cave, An.; Cave, Ad.; Paris, R.-R. C R. Acad. Sci. Paris, 1971,272, 665-667. Levesque, J.; Pousset, J.-L.; Cave, Ad.; Cave, An. C R. Acad. Sci. Paris, 1974, 278,959-961. Levesque, J.; Pousset, J.-L.; Cave, A. C. R. Acad. Sci. Paris, 1974,279, 10531055. Brown, R. T.; Charalambides, A. A; Cheung, H. T. Tetrahedron Letters, 1973, 4837-4838. Benoit-Vical, F.; Valentin, A.; Cournac, V.; Pelissier, Y.; Mallie, M.; Bastide, J.-M. J. Ethnopharmacol, 1998, 61, 173-178. Lepers, J. P.; Deloron, P.; Mouden, J. C; Le Bras, J.; Coulanges, V.Trans. R. Soc. Trop. Med. Hyg., 1989, 83,491-492. Bunyapraphatsara, N.; Yongchaiyudha, S.; Rungpitarangsi, V.; Chokechaijaroenporn, O. Phytomedicine, 1996, 3,245-248. Boiteau, P.; Mouton M.; Rakoto-Ratsimamanga, A. Elements de Pharmacopee Malagasy, Imprimerie Nationale, Tananarive, 1969, Tome I, pp. 172-193. Balick, M. J.; Cox, P. A. In: Medicinal Plants for Forest Conservation and Health Care, G. Bodeker, K. K. S. Bhat, J. Burley and P. Vantomme, Eds; 1997;Vol. ll,pp. 12-23. Rasoanaivo, P.; De La Gorce, P. Herbalgram , 1998,43, 31-39. Galeffi, C. J. Ethnopharmacol., 1980,2, 129-134. Sapi, J.; Massiot, G. In The Monoterpenoid indole Alkaloids, Saxton, Ed.; John Wiley & Sons Ltd, 1994; Vol. 7, pp. 313-314. Braun, S.; Kalinowski, H.O.; Berger, S. 150 and More Basic NMR Experiences, VCH : Weinheim, 1998.

1071

[67]

[68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91]

Piotto, M. In New Advances in Analytical Chemistry, Atta-ur-Rahman, Ed.; Elsevier Science: Amsterdam, 2000, in press (private communication from the author). Bernstein, M.A.; Trimble, L.A. Magn. Reson. Chem., 1994,32,107-110. Stelten, J.; Leibfritz, D. Magn. Reson. Chem., 1995, 33, 827-830. Bodenhausen, G.; Ernst, R.R. J. Am. Chem. Soc, 1982,104, 1304-1309. Wagner, R.; Berger, S. Magn. Reson. Chem., 1998, 36, S44-S46. Hadden, C.E.; Martin, G.E.; Krishnamurthy, V.V. J. Magn. Reson., 1999,140, 274-280. Williamson, T.; Marquez, B.L.; Gerwick, W.H.;K6ver, K.E. Magn. Reson. Chem., 2000, 38, 265-273. Martin, G.E.; Hadden, C.E. J. Nat. Prod., 1999, 62, 238-240. Hadden, C.E.; Martin, G.E.; Tackie, A.N.; Schiff Jr., P.L. J. Heterocyclic Chem 1999,56,1115-1117. Hadden, C.E.; Martin, G.E. J. Nat. Prod., 1998,61, 969-972. Hadden, C.E.; Martin, G.E. Magn. Reson. Chem., 1999, 37, 385-388. Martin, G.E.; Hadden, C.E. Magn. Reson. Chem., 1999, 37, 721-729. Martin, G.E. J. Heterocyclic Chem., 1997, 34, 695-699. Koshino, H.; Uzawa, J. Kagaku to Seibustu, 1995,33,252-258. Martin, G.E.; Crouch, R.C. J. Heterocyclic Chem., 1995, 32,1839-1842. Cai, BC; Wu, H.; Yang XW; Hattori, M.; Nainba, T. Acta Pharmaceutica Sinica, 1994, 29,44-48. Mukherjee, R.; da Silva, B.A.; Das, B.C.; Keifer, P.A.; Shoolery, J.N. Heterocycles, 1991, 32, 985-950. Yavari, I.; Roberts, J.D. Org. Magn. Reson., 1979,12, 87-91. Martin, M.T.; Frappier, P.; Rasoanaivo, P.; Randrianarivelojosiq, M. Magn. Reson. Chem., 1996, 34,489-491. Mukherjee, R.; da Silva, T.M.S.; Guimaraes, J.B.L.; Oliveira, E.J.; Keifer, P.A.; Shoolery, J.N. Phytochem. Anal. 1997, 8, 115-119. Mukherjee, R.; Das, B.C.; Keifer, P.A.; Shoolery, J.N. Heterocycles, 1994,38, 1965-1970. Trigalo, F.; Martin, M.-T.; Blond, A.; Brouard, J. P.; Rafatro, H.; Ramanitrahasimbola, D.; Frappier, F. Tetrahedron, 1999, 55, 6139-6146. Quetin-leclerq, J.; Llabres, G.; Warin, R.; Belem-Pinheiro, M.L.; MavarManga, H.; Angenot, L. Phytochemitry, 1995,40,1557-1560. Penelle, J.; Tits, M.; Christen, P.; Brandt, V.; Frederich, M.; Angenot, L. J. Nat. Prod., 1999,62, 898-900. Dassoneville, L.; Bonjean, K.; De Pauw-Gillet, M.-C; Colson, P.; Houssier, C ; Quetin-Leclercq, J.; Angenot, L.; Bailly, C. Biochemistry, 1999,38,77197726.

1072

[92] [93] [94] [95] [96] [97] [98] [99]

[100] [101] [102] [103] [104]

[105]

Quetin-Leclercq, J.; De Pauw-Gillet, M.-C; Angenot, L.; Bassleer, R.; Chem.Biol Interact, 1991, 80, 203-211. Quetin-Leclercq, J.; Bouzahzah, B.; Pons, A.; Greimers, R.; Angenot, L.; Bassleer, R.; Barbason, H. Planta Med,, 1993, 69,59-62. Bonjean, K.; De Pauw-Gillet, M.-C, Quetin-Leclercq, J.; Angenot, L.; Bassleer, R. Anticancer Res,, 1996,16, 1129-1137. Wright, C. W.; Phillipson, J. D., Awe, S. O.; Kirby, G. C; Warhurst, D. C ; Quetin-Leclercq, J.; Angenot, L. Phytoter. Res., 1996,10, 361-363. Kirby, G. C ; Paine, A.; Warhurst, D. C; Noamese, B. K.; Phillipson, J. D. Phytoter Res., 1995, 9, 359-363. Wright, C. W.; Bray, D. H.; O'Neill, M. J.; Warhurst, D. C; Phillipson, J. D.; Quetin-Leclercq, J.; Angenot, L. Planta Med., 1991,57, 337-340. Wright, C. W.; Allen, D.; Cai, Y.; Chen, Z.; Phillipson, J. D.; Kirby, G. C ; Warhurst, D. C.; Tits, M.; Angenot, L. Phytoter. Res., 1994,8,149-152. Rafatro, H.; Ramanitrahasimbola, D. Rasoanaivo, P.; Ratsimamanga-Urverg, S.; Rakoto-Ratsimamanga, A.; Frappier, F. Biochem. Pharmacol, 2000,59, 1053-1061. Martin S. K.; Oduola, A. M. J.; Mihous, W. K. Science, 1987,235, 899-901. Rasoanaivo, P.; Ratsimamanga-Urverg, S.; Frappier F. Current Med. Chem., 1996,3,1-10. Bjorkman, A.; Willcox, M.; Kihamia, C. M.; Mahikwano, L. F., PhillipsHoward, P. A.; Hakansson, A.; Warhurst, D. Lancet, 1990,336,59-60. Warsame, M.; Wernsdorfer, W. H.; Bjorkman, A. Trans. R. Soc. Trop. Med. Hyg., 1992, 86, 235-236. Frappier, F.; Jossang, A.; Soudon J.; Calvo, F.; Rasoanaivo, P.; RatsimamangaUrverg, S.; Saez, J.; Schrevel, J.; Grellier, P. Antimicrob. Agents Chemother., 1996,40,1416-14S1. Rasoanaivo, P.; Ratsimamanga-Urverg, S.; Ramanitrahasimbola, D.; Rafatro, H.; Rakoto-Ratsimamanga, A. J. EthnopharmacoL, 1999, 64, 117-126.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

1073

OCCURRENCE OF BIOLOGICALLY ACTIVE 2-THIOXOPYRROLIDINES AND 3,5-DISUB. STITUTED 2-THIOHYDANTOINS FROM THE PUNGENT PRINCIPLE OF RADISH (Raphanus sativus L.) YASUSHI U D A I , YOSHIO OZAWA2, AND KOICHI Y 0 N E Y A M A 3 Department of Bioproductive Sciences^, and Center for Research on Wild Plants^, Utsunomiya University, Utsunomiya, 321-8505 Japan, and Department of Food and Nutrition^, Gunma Women's Junior College, Takasaki, 370-0033 Japan ABSTRACT: The pungent principle of radish (Raphanus sativus L.), 4-(Methylthio)(£',Z)-3-butenyl isothiocyanate (MTBI), was found to be significantly labile in the presence of water to form 2-thioxopyrrohdines, such as 3-(hydroxy)methylene-2thioxopyrrolidine (HMTP) and 3-(mediylthio)methylene-2-thioxopyrrohdine (MMTP). Also, MTBI reacted easily with L-tryptophan to generate the tetrahydro-P-carbohne derivative (IS*, 3S*, 3'R*)-l-(2'-pyrrohdinethione-3'-yl)-l,23,4-tetrahydro-p-carbohne3-carboxylic acid (PTCC) and its (IR*, 3S*, 3'R*)-isomer, and it was reactive with other amino acids to form their corresponding 3,5-disubstituted 2-thiohydantoins. By using mass (MS), nuclear magnetic resonance (NMR), infrared (IR), and ultraviolet (UV) spectrometric analyses, we were able to characterize the structures of these reaction products and their formation pathways. Subsequently, we assayed the reaction products and their related compounds for their antimicrobial and antimutagenic activities and examined their action mechanisms. HMTP, the major product formed from MTBI in aqueous media, showed not only antimicrobial activities against a wide range of microorganisms but also antimutagenic activities against food-bom heterocyclic amines, IQ and Trp-p-1. The 3,5-disubstituted 2-thiohydantoins derived from MTBI with leucine and phenylalanine also exhibited remarkable antimutagenic activities against the heterocyclic amines. The antimicrobial action of HMTP was due to its sporicidal and bactericidal activities. The antimutagenic mechanisms of HMTP and the 2-thiohydantoins were suggested to be attributable to their inhibition of cytochrome p-450s-mediated activation of the heterocyclic amines.

INTRODUCTION A variety of isothiocyanates are generated through an enzymatic (thiogluco-

1074 side glucohydrolase, E.C.3.2.3.1) hydrolysis of their corresponding glucosinolates and consequent Lossen rearrangement of an intermediate, thiohydroxamate-O-sulfate [1-3] (see Fig. (1)). Since Roubiquet and Bourtron isolated the glucosinolate sinalbin from white mustard (Sinapis alba L.) [4], approximately one hundred different isothiocyanates and/or their corresponding glucosinolates have been found in approximately eleven botanical famihes [5, 6], mostly in Cruciferae. The products formed from glucosinolates depend upon the nature of the glucosinolate as well as the reaction conditions, such as the pH or the presence or absence of specific metal ions [7-9]. For example, isothiocyanates are preferably formed at a neutral pH, while the formation of nitriles is definitely preponderant in the presence of ferrous ion or thiol compounds at pH 5.0-5.5 [8-10]. Several reviews that deal with the biosynthesis, chemical nature, and enzymatic degradation of glucosinolates are available [1-3, 11-13]. As for the biological and physiological activities of the glucosinolates and their breakdown products, many investigations have focused on the harmful role of such products as oxazolidine-2-thiones (goitrogens) and cyanoepithioalkanes in mammals [2, 14-17]. Also, there have been many studies on the antimicrobial activity of both naturally occurring and synthetic isothiocyanates [18-20]. Today, isothiocyanates are receiving considerable attention because of their possible protective action against cancer. For example, isothiocyanates have been linked to the inducibility of cancerprotective phase II enzymes, such as glutathione S-transferase, UDPglucronosyl transferase, and quinone reductase [21-23], and the inhibitory activity of these enzymes on tumors in mammals [24-27]. A few researchers have examined the metabolic fate of isothiocyanates in mammals [28-31]. D-Glucose /S-CgHnOs

^

R-C "^N-O-SOs"

SH R-C

R~c:

-^^

N-O-SO3-

Myrosinase

Glucosinolates

NH-O-SO3-

Thiohydroxamate-O-sulfonate

R-NCS Isothiocyanates

A—S Oxazolidine2-thione

R-CN Nitriles

H2C

-C—(CH2]t—CN ri

1 -Cyanoepithioalkanes

F i g . ( l ) . Enzymatic breakdown of glucosinolates.

1075 On the other hand, isothiocyanates are known to be reactive with a variety of food constituents, such as amines, amino acids, proteins, thiols, and alcohols, due to their electrophilic properties to give a variety of compounds, including thioureas and dithiocarbamates [32]. Despite their chemical nature, only limited information has been made available regarding the stability of isothiocyanates and the biological activity of reaction products of isothiocyanates with food constituents [33-35].

\ (£)-MTBI

/

^

^

\

.

(Z)-MTBI

Fig.(2 ). Structures of the pungent principle of radish, (E)- and (Z)-4-(methylthio)3-butenyl isothiocyanates.

Among the cruciferous plants, only the genera Raphanus sativus L. (radish) and Matthiola fruticulosa (gilly-flower) have been known to generate 4(methylthio)-(£,Z)-3-butenyl isothiocyanate (MTBI) (see Fig. (2)) [36, 37]. (£',)-MTBI was shown to be mainly responsible for the characteristic taste and flavor of the radish [36]. According to our investigations on the relative amounts of volatile isothiocyanates found in radish roots of Japanese and Kenyan origin [38], at least eight volatile isothiocyanates, including such minor compounds as 4-(methylpentyl), 5-hexenyl, benzyl, 2-pheneylthyl, 3-(methylthio)propyl, 4-(methyIthio)butyl, and 5-(methylthio)pentyl, were detected in either the Japanese or the Kenyan radish, but both had MTBI as the major pungent constituent. These observations have also been supported in quantitative studies of the glucosinolates in radish cultivars [39,40]. In respect to other pungent components of the radish, 4(methylsulfinyl)butyl and 4-(methylsulfinyl)butenyl isothiocyanates, which prevail in the seed, have been identified [37]. The parent glucosinolate of (£)-MTBI was recently isolated from the roots of the radish, and its spectroscopic properties have been characterized [41]. Radish root is primarily cultivated in Asia, and it is a particularly important crop for the people in East Asia, where it is consumed after undergoing a variety of cooking and processing methods. In Japan, at least 20 varieties of radish are cultivated through the year to produce about 2 million tons per year in total, and the radish roots are eaten as a freshly cooked or grated vegetable and as a salted or pickled product. The grated or lightly salted roots have a pungent flavor mainly due to MTBI formed during treatment. Average daily intake is estimated to be about 16 g per person in Japan. The average contents of glucosinolates and MTBI in fresh

1076 radish roots have been reported to be about 1080 and 340 mg per kg, respectively [40]. Taking into consideration the abundant production and consumption of the radish, it is important to carry out research on the interaction between MTBI and food constituents as well as on the biological and physiological activities of the reaction products. This should provide useful information to both consumers and the food industry. To contribute to this research, we have investigated the chemical and biological behavior of the pungent principle of the radish. We describe below the characteristic reactivity of MTBI along with some biological activities of its reaction products generated in aqueous and alcohol-containing media and in processed radish roots. Formation of thioxopyrrolidines from MTBI in aqueous and alcohol-containing media The labile property of MTBI in the presence of water There have been only a few studies of the decomposition of isothiocyanates in the presence of water. Kawakishi et al showed that p-hydroxybenzyl isothiocyanate, the major pungent factor in white mustard seeds, was so labile in a moistened paste that the isothiocyanate degrades into benzyl sulfide and benzyl alcohol within a few hours [35, 42, 43]. Allyl isothiocyanate is also unstable in an aqueous medium and decomposes to form a variety of sulfur-containing products such as diallyl mono-, di-, tri-, and tetrasulfides, 1,2-dithiolene, and 1,2,3-trithiin [44]. We therefore studied the stability of MTBI in aqueous media [45], in which a mixture {ca. 7:1) of (£)- and (Z)-MTBI obtained by an enzymatic hydrolysis of their corresponding glucosinolates extracted from radish seedlings was treated in the following three different ways: sonicated in a buffer (pH 6.5) at 37° C; shaken in the same buffer at 3T C; and spread on an agar-plate (pH 6.5) at 37"" C. As shown in Table 1, most of the (£)-MTBI, which is the major form of MTBI occurring naturally in chopped or grated radish root, was lost during the 3-h shaking and 5-h sonication. (£)-MTBI was mostly degraded even on the agar-plate after the 6-h treatment. The labile property of (£)-MTBI was considered to be comparable to that of p-hydroxybenzyl isothiocyanate. In contrast, the (Z)-isomer was much more stable. As expected, phenyl isothiocyanate, which was used as the reference compound, did not decrease in quantity at all. These results demonstrated that MTBI, particularly the (£)-isomer, is significantly unstable in the presence of water.

1077 Table 1. Labile property of MTBI in aqueous media (Modified from Udaera/. [45]) Remaining MTBI {%f^ Treating time (h)

Shaking (P)-MTBI (Z)-MTBI 1.3 0.5

Sonication (£)-MTBI (Z)-MTBI

Spreading on agar-plate (E)-MTBI (Z)-MTBI PHEITC '

64.4

43.6

84.0

- c)

72.4 62.3

-

49.5

26.4 12.4 0.6

29.4

-

-

-

47.7

99.8

19.7 4.1

77.4 38.8

0.6

25.4

100 100 100 100

a) Detrmined by FPD-GC analysis. b) Phenyl isothiocyanate as the reference. c) Not examined.

Identification of the products generated fi'om MTBI in the presence of water The labile properties of MTBI led us to study the chemical structure of products that should be formed from MTBI in aqueous media, their reaction mechanisms or pathways, and their biological activities. These studies were considered to be important when evaluating the use of radishes and radish products in the food industry, because MTBI is readily formed in the tissues of the radish during cooking and processing. Thus, we studied the MTBI-derived products in the presence of water [46,47]. When MTBI was sonicated or shaken in a buffer at pH 6.5, methanethiol was generated. This was confirmed by an IR spectral analysis of the mercaptide obtained by trapping methanethiol with mercuric cyanide [46], which suggested that MTBI underwent breakdown to form other products in the presence of water. The consequent survey of the products showed that the aqueous phase of the reaction mixture contained a major as well as some minor compounds. The major compound was successfully isolated as a pale yellow powder by cross-linked polyvinylpyrrolidone column chromatography, and the powder was identified as 3-(hydroxy)methylene2-thioxopyrrolidine (HMTP) by UV, IR, MS, and NMR spectral analyses [46]. The MS spectrum was recorded at 70 eV, and the 1H- (400 MHz) and l^C-NMR (100 MHz) data were measured in dimethylsulfoxide-^5 using TMS (tetramethylsilane) as the internal standard. Figure (3) shows the structure of HMTP, and its spectral data are shown in Table 2 .

1078

Fig. (3). Structure of 3-(hydroxy)methylene-2-thioxopyrrolidine (HMTP).

Table 2. UV, IR, MS, and NMR spectral data for HMIP UV X ^ ^ (H2O) nm

270; 330

I^^max (KBr) cm"^

2970 (w); 2910 (w);1735 (m);1655 (m);

EI-MSm/z(%) lH-NMR6(DMSO-t/6) ppm

129(M+.100);101 (96);100 (92);71 (18) 2.67 (2H, m, 3-CH2);3.48 (2H, m. 5-CH2); 7.51 (IH, t. 7=2.4 Hz, 6-CH); 9.60 (IH, s, OH) 23.13 (4-CH2); 45.74 (5-CH2);l 17.69 (3-C); 150.55 (6-CH); 195.20 (2-C)

13c-NMR6(DMSO-fif6)ppm

More than seventy years ago, Nakamura described the generation of methanethiol in freshly grated as well as variously processed radishes [48]. Our investigation ascertained that MTBI was readily converted to HMTP accompanied by the release of methanethiol in an aqueous medium. Accordingly, we proposed a formation pathway for the major product, HMTP, as shown in Fig. (4). Further investigations to identify the minor products formed from MTBI in aqueous media were then carried out [47]. A reverse phase HPLC analysis provided evidence that the minor products were composed of at least five different compounds. Figure (5) shows the HPLC-separation of the products obtained by the sonication of MTBI in a citric acid-disodium hydrogen phosphate buffer (pH 7.0) containing 0.1 % Triton X-100 for 5 h. Six major peaks (P1-P6) were detected. Their relative ratios were: PI, 46.2%; P2, 6.6%; P3, 17%; P4, 1.3%; P5, 7.7%; and P6, 21%. The most abundant component, PI, was easily identified as HMTP from the spectral data obtained, but the other five components were too small to identify. Optimal reaction pHs were then studied to aid in the identification. Among the six products, the components PI, P2, and P3 were most readily formed in an acidic media (pH 3-5), although these were detected at all pHs (3-9) examined. In contrast, the components P4, P5, and P6 were abundantly formed above pH 6. Hence, the reaction was carried out at pH 3.0 to obtain both P2 and P3, and pH 9.0 was used to collect the latter three components.

1079

H HsC-

H C

on

CH2

/ S=C

\ CH2 H

CH3SH

yA -o

I OH

HMTP Fig. (4). Proposed pathway for the formation of HMTP from MTBI in aqueous media Modified from Uda et al. [46].

1080

PI

P6

P3

A P5

ft

A

P2 P4

A 10

15

20

25

30

Retention time (min Fig. (5 ). An HPLC profile of the products generated from MTBI in an aqueous medium at pH 7.0. Modified from Matsuoka et al. [47]. P1-P6: the reaction products. The HPLC analysis was done on a LiChrosorb RP-18 column (4.6 mm i.d. x 250 mm) with a mobile phase of 30% acetonitrile in 25 mM phosphate buffer (pH 6.5) at a flow rate of 0.8 ml/min. Detection was made by a UV detector at 270 nm.

These components were then subjected to chromatographic separations on reverse-phase and normal-phase silica gel columns, which gave sufficient amounts of purified compounds for their identification. The HR-EI-MS, LR-EI-MS, IH-NMR, and l^C-NMR data obtained were: the compound P2, HR-EI-MS m/z (M+) : found, 159.0145; calcd for C6H9NS2, 159.0176; LR-EI-MS m/z (%): 159 ( M+ ,57), 144 (100), 100 (8), 85 (24), 71 (8); IH-NMR 6 (CDCI3, ppm): 2.39 (3H, s), 3.00 (2H, dt, 7=7.4 & 1.7 Hz), 3.62 (2H, dt, 7=7.4 & 1.0 Hz), 6.74 (IH, s), 7.71 (IH, s); l^CNMR 5 (CDCI3, ppm): 20.6, 30.6, 46.3, 130.9, 140.4, 194.8; and the compound P3, HR-MS m/z (M+) : found, 159.0175; calcd for C6H9NS2, 159.0176; LR-EI-MS m/z (%): 159 ( M+ ,57), 144 (100), 100 (8), 85 (24), 71 (8); 1H-NMR6 (CDCI3) ppm: 2.52 (3H, s), 2.79 (2H, dt, J=7.3

1081

8L 2.4 Hz), 3.67 (2H, dt, J=7.3 & 1.0 Hz), 7.58 (IH, s), 7.99 (IH, s); l^C-NMR 6 (CDCI3) ppm: 17.5, 26.4, 45.8, 133.6, 139.3, 194.5. From the spectral data, the compounds P2 and P3 were identified to be (Z)-3(methylthio)methylene-thioxopyrrolidine ((Z)-MMTP) and its (£)-isomer ((£)-MMTP), respectively. The two compounds were presumed to have been formed through an intermolecular cyclization of MTBI. The amount of (£)-MMTP formed in the pH 7.0 medium was 2.5-fold higher than that of the (Z)-isomer. This was probably due to the predominant amount of (£)MTBI in the reaction mixture. Table 3. MS and NMR data for the compounds P4, P5, and P6 HR-MS m/z (M+): found, ; 209.0355; calcd for CyH^gNS3, 209.0367 EI-MS m/z (%) : 209 ( M+,8), 176 (8). 162 (100), 114 (16), 91 (28) ^ H-NMR 6(CDCl3) ppm : 1.68 (2H. dt, Jb7.1 & 6.4 Hz), 1.79 (2H, quint, Jb7.1 Hz), 2.01 (3H, s), 2.54 (2H. t, Jb7.1 Hz), 7.99 (IH. s), 2.64 (3H, s). 3.78 (IH, t, v^6.4 Hz), 3.76 (IH. t, Jb6.8 Hz), 7.05 (IH, s) ''^C-NMR 6 (CDCI3) ppm : 15.5. 18.2. 26.3, 27.3, 33.8. 46.8, 199.1 pg

HR-MS m/z (M+): found, ; 207.3839; calcd for C7H13NS3. 207.3849 EI-MS m/Z{%) : 207 ( M+ .8), 160 (39), 100 (61), 91 (35). 85 (100) ^ H-NMR 6 (CDCI3) ppm : 2.31 (3H, s), 2.48 (2H, dt, Jt=6.8 & 6.4 Hz), 2.62 (3H, s), 3.80 (1H, t, J=6.8 Hz), 3.82 (1H. t, J=6.4 Hz). 5.54 (1H, dt. Jt9.3 & 7.3 Hz). 6.09 (IH, d, J=9.3 Hz), 7.15 (IH, s) ""^C-NMR 6 (CDCI3) ppm : 17.0, 18.1. 27.8. 46.5,122.8, 130.6. 198.9

pg

HR-MS m/z (M+): found, ; 207.3844; calcd for C7H13NS3, 207.3849 EI-MS m/z (%) : 207 ( M+ .8), 160 (39), 100 (61), 91 (35), 85 (100) •"H-NMR 6 (CDCI3) ppm : 2.20 (3H, s), 2.37 (2H. dt,Jb7.6 & 7.3 Hz). 2.51 (3H. s). 3.60 (IH, t, J=7.3 Hz), 3.61 (1H, t, J=7.Z Hz), 5.35 (1H, dt, JS=15.1 & 6.8 Hz). 6.15(1H, d, J=15.1 Hz), 9.89(1 H,s) ^^C-NMR 6 (CDCI3) ppm : 15.0, 18.3, 32.0, 46.7, 121.9,127.7, 199.2 The MS data were recorded at 70 eV, and the ^H- (400 MHz) and ^^C- (100 MHz) NMR spectra were measured in CDCI3 using TMS ^^C-(100 sp as the internal standard.

The spectral data obtained for the compounds P4, P5, and P6 (Table 3) allowed to identify them as methyl 4-(methylthio)butyldithiocarbamate (MBDC), (Z)-methyl 4-(methylthio)-3-butenyldithiocarbamate ((Z)-

1082 MBDC), and the (£)-isomer of (Z)-MBDC ((£)-MBDC), respectively. It is well known that isothiocyanates are converted to their corresponding dithiocarbamates by a reaction with thiol compounds [23,32].

H3C-S (2)-and(£)-MTBI HoO

-v

r —

rY"

OH

pH3-6

L

H HMTP

\

J

^N=:C=S

H.,C-

(2)-and(£)-MTBI

lpH3-9

pH6-9

'f

- CHgSSCI^

\

H3C-S / ^s^X-V^N^^/S-CH3 1 ( £)-MBDC

\ 2H^

H3C-S

pH3-9

+ (Z)-MB

DC

II

Fig. (6 ). Pathways for the formation of the reaction products of MTBI in aqueous media Modified from Matsuoka et al. [47].

1083 Thus, both (Z)- and (£)-MBDC were thought to have been formed by a reaction with MTBI and methanethiol evolved from MTBI during the reaction. The saturated MBDC, on the other hand, was presumed to have been generated via another reaction pathway, because GC-MS analysis showed that the MTBI used in the reaction had only a negligible amount of 4-methylthiobutyl isothiocyanate as an impurity of MTBI. It is known that methanethiol is easily oxidized to form dimethyldisulfide, during which protons should be released. This led us to examine the possibility that saturated MBDC is generated by the addition of hydrogen to unsaturated (Z)- and (£)-MBDC. To confirm or disprove this, gaseous methanethiol was introduced into a buffer (pH 3.0 or 9.0) containing purified (£)-MBDC under an agitation at room temperature. The saturated MBDC was found only in the presence of methanethiol. The formation of dimethyldisulfide was also confirmed in this circumstance. The alkaline medium was much more conducive to the formation of the dimethyldisulfide than was the acidic medium. These results demonstrated that the methanethiol released in the process of the generation of HMTP reacted with MTBI to yield (Z)- and (£)-MBDC, and that it was also involved in the formation of the saturated dithiocarbamate through an oxidative formation of dimethyldisulfide [47]. Based on these findings, we proposed the reaction pathways of MTBI in the presence of water that are shown in Fig. (6). We conclude that MTBI is highly reactive in food systems such as freshly grated or processed radish and that it is probably converted to the 2-thioxopyrrolidines and/or dithiocarbamates in the human gastric and intestinal tract during digestion of freshly grated radish roots. The formation of 2-thioxopyrrolidines from MTBI and its synthetic analogue, 4-methoxy-3-butenyl isothiocyanate (MBI) The finding that MTBI can be easily converted to HMTP and (£',Z)-MMTP in aqueous media suggested that alkoxy-substituted 2-thioxo-pyrrolidines should be formed in alcohol-containing media. In fact, Shibayama et al isolated 3-[a-methoxy(a-methylthio)]methyl-2-thioxopyrrolidine from a methanol extract of the fresh roots of radish as an antihypertensive and anesthetic factor, and the compound was named raphantin [49]. They also described that when ethanol was used for the extraction instead of methanol, the product was 3-[a-ethoxy(a-methyl-thio)]methyl-2thioxopyrrolidine. Hasegawa et al have reported the presence of a lightinduced growth inhibitor in the methanol extract of radish seedlings, which they have identified as 3-methoxy-4-methyl-thio-2-piperithione (raphanusanin) [50]. Raphanusanin was later shown to be composed of (3R*, 6R*)-3-[a-methoxy-(a-methylthio)methyl]-2-thioxo-pyrrolidine and its (3R*, 6S*)-isomer [51]. Kosemura et al subsequently reported that the

1084 precursor of these compounds is MTBI [52]. In our preliminary experiments, however, MTBI was converted to HMTP as the major product in 20-60% methanol-containing media, although the main product formed in 60-80% methanol-containing media was raphanusanin (unpublished data). On the basis of the structural similarity of the compounds, we believed that a synthetic analogue of MTBI, i,e,, 4-methoxy-3-butenyl isothiocyanate (MBI), should afford HMTP and alkoxy-substituted analogues of HMTP in an alcohol-containing aqueous media. In a buffer (pH 2.0-7.0), HMTP was formed with about 90% yield, but the major products formed in 80% methanol and 80% ethanol were 3-(a,a-dimethoxy)methyl-2thioxopyrrolidine and 3-(a-ethoxy-a-methoxy)meth-yl-2-thioxopyrrolidine, respectively (Fig. (7)) [53].

3-(a,a-Dimethoxy)methyl2-thioxopyrrolidine (ca. 49%)

HMTP(Ca. 90%)

3 - (a- Ethoxy-a- me thoxy) me thy 12-thioxopyrrolidine (Ca. 36.%)

Fig. (7 ). Formation of HMTP and its structurally related 2-thioxopyrrolidines in aqueous or alcohol-containing media. Modified from Matsuoka et al. [53].

When radish roots or seedlings were extracted with acetone or methanol-^4^, no raphanusanin was detected in the acetone extracts, but CDs-substituted raphanusanin was found in the methanol-^4 extracts [54]. Furthermore, when 70% methanol extract of blanched seedlings and roots of radish was treated with or without myrosinase, both (3R*, 6R*)- and (3R*, 6S*)raphanusanins were detected only in the reaction mixtures with myrosinase

1085 [54]. These results clearly demonstrated that the formation of raphantin or raphanusanin depends upon the extracting solvents or reaction media. This suggests that these alkoxylated 2-thioxopyrrolidines are not naturally occurring products in the radish. The reaction products ofMTBJwith amino acids Asian people consume a variety of traditionally cooked and/or processed radishes. In Japan, there are some varieties of salted radishes that are annually produced in amounts of about 500,000 tons. It has been known for many years that the salted and fermented product of white radish roots turns a bright yellow during manufacturing. Also, the yellow color fades easily after exposure to light, including sunlight, which is a significant problem for the food-processing industr>^ For that reason, we studied the yellowing reaction in the salted roots of the radish [55-59]. At the beginning of our studies, we found that the yellowing reaction could be produced by concentrating freshly prepared radish juice at a temperature below 40° C but not when using a root blanched at 100°C for 3 min. When the blanched juice was concentrated under the same conditions in the presence of an enzyme fraction of radish, a yellow color was formed, which suggested that the yellowing reaction was produced enzymatically [55]. A survey for the precursor(s) of the yellowing located a fraction that turned yellow in the presence of the enzyme fraction of the radish, and consequently 4(methylthio)-(E,Z)-3-butenyl glucosinolate was identified as the precursor [56]. The optimal temperature and pH for the yellowing reaction were coincidental to those for myrosinase action. Thus, we confirmed that a hydrolysis of the glucosinolate by myrosinase was essential for the formation of the yellow color in radish roots. Further investigations were conducted to isolate the yellow pigments formed in salted radishes which included extraction with various solvents and chromatographic separations. Although all the attempts to isolate the pigments were unsuccessful due to their complexity, a main route for the formation of the yellow color was found by identification of a potent intermediate of the pigmentation [56]. A pale yellow compound was isolated from a bright-yellow product that had been aged for over 6 months after salting. Its UV, IR, EI-MS, FAB-MS, iH-NMR, and l^C-NMR spectral data were as follows. UV Xmax (MeOH) nm: 220, 270, 280, and 290 (shoulder); IR v ^ax (KBr) cm-l; 3400, 3300, 3200, 3000, 1750, 1650, 1600, 1550, 1470, 1420, 1350, 1310, 1240, and 750; EI-MS miz (%): 101 (74), 144 (72), 168 (90); FAB-MS m/z\ 315 (M-f-H)+ and 338 (M+Na)+: Elemental analysis: found: C, 54.55%; H, 6.06%; N, 11.5%; calcd. for C16H17N3O2S-2H2O: C, 54.69%; H,6.02%; N,11.96%; iH-NMR 6 (DMSO-^^) ppm: 1.64 (IH, m), 2.24 (IH, m), 2.65 (lH,dd,y= 15.3 & 11.1), 3.01 (IH, dd, 7 = 15.3 & 3.9), 3.31 (2H, m), 3.47 (IH, m), 3.60 (IH, dd, 7 = 11.1 & 3.9), 4.5 (IH, b), 4.97 (IH, s).

1086 6.98 (IH, dt, 7 = 7.8 & 1.2), 7.02 (IH, dt, 7 = 7.8 & 1.2), 7.40 (IH, d, 7 = 7.8; IH. d, J = 7.8), 9.63 (IH, s), and 10.40 (IH, s); 13C-NMR 6 (DMSO-d6) ppm: 23.98 (4'-C), 25.16 (4-C), 47.07 (5'-C), 54.87 (1-C), 55.90 (3'-C), 56.73 (3-C), 109.50 (11-C), 111.71 (8-C), 117.37 (5-C), 118.57 (7-C), 120.87 (6-C), 126.46 (12-C), 132.16 (10-C), 136.16 (13C), 173.39 (14-C), and 202.68 (2'-C). These spectral features provided evidence allowing us to identify the compound as l-(2'-pyrrolidinethione3'-yl)-l,2,3,4-tetrahydro-p-carboline-3-carboxylic acid (PTCC). FTCC was later confirmed to be composed of a mixture of (IS*, 3S*, 3'R*)- and (IR*, 3S*, 3'R*)-isomers (Fig. (8)) [60].

(1S*,3S*,3'R*)-PTCC

(1R*,3S*,3'R*)-PTCC Fig. (8). Structure of (IS*, 3S*, 3R*)-PTTC and (IR*, 3S*, 3R*)-PTTC.

These compounds, which have not been found elsewhere, gradually

1087 became deep yellow in color, accompanied by the formation of some yellow compounds possessing the tetrahydro-p-carboline skeleton. The major secondary product derived from PTCC was identified as l-(4'-amino-3'butenal-2'-yl)-p-carboline [57]. The presence of at least three pigments possessing the tetrahydro-p-carboline skeleton was ascertained in the salted roots of the radish [58]. Tetrahydro-p-carbolines are known to be formed from carbonyl compounds and L-tryptophan via a condensation reaction, i.e., the Pictet-Spengler reaction, under physiological conditions [61]. Since tetrahydro-p-carbolines can function as neurotransmitters and/or neuromodulators [62], many studies of the analysis of tetrahydro-pcarbolines in fermented foods and alcoholic beverages have been reported [63-68]. In relation to changes in the color of these compounds, Chu and Clydesdale have provided a detailed discussion, pointing out that certain tetrahydro-p-carbolines turn yellow when they are held at room temperature or during thermal treating [69]. Tetrahydro-p-carboline-3-carboxylic acids undergo an oxidative decarboxylation [70]. As mentioned above, the precursor to the yellowing of radish roots was the glucosinolate of MTBI, which was converted to HMTP in the presence of water. HMTP was thought to be formed to maintain equihbrium between the enolic and keto forms. Thus, it was presumed that HMTP could react with L-tryptophan to form PTCC. This assumption was clearly supported by stirring MTBI in a buffer with L-tryptophan, upon which approximately 55% of the MTBI used reacted with L-tryptophan to form PTCC within 60 min. A pH of 6.5 was most effective for the formation of PTCC [59]. From these results, the pathways for the formation of PTCC in radish roots can be given as shown in Fig. (9). Because too small a quantity of L-tryptophan was contained in the radish root, the major source for the production of PTCC was thought to be attributable to the growth of such L-tryptophan-producing yeasts as Sacchromyces sp., Candida sp., and Hansenula sp. during the aging of the salted radish product [71]. In addition, in the pungent principle of radish, MTBI has been found to be involved in other yellowing reactions [72]. It was observed that MTBI can react to form a greenish-yellow substance with L-ascorbic acid and such o-dihydroxyphenolic compounds as catechin, Ldopa, and caffeic acid. Among them, L-ascorbic acid was the most reactive with MTBI. When 10 other aliphatic and aromatic isothiocyanates were employed for the reaction instead of MTBI, no yellow-colored substance was formed, suggesting that the reaction was specific to MTBI. In the salted roots of the radish, MTBI is detectable for at least one month after salting [73], and L-ascorbic acid and some o-diphenolic compounds are contained in the radish [74, 75]. Therefore, the yellowing reaction with Lascorbic acid or o-dihydroxyphenohc compounds is more than likely involved in the formation of the yellow color during the early stages of salting. However, these colored products have not yet been identified. As for the reaction of isothiocyanates with amino acids, the Edman degradation of proteins and/or peptides is well known [76].

1088 .S-C6H„05 H3CS -

N-o-sq 4-Methylthio-3-butenyl glucosinolate

Myrosinase

D-Glucose

H.CS

OH

H

HNfTP

COOH

Condensation L-Trp

COOH

PTCC

Fig. (9). Proposed pathways for the formation of PTCC.

1089 The kinetics and mechanism of the formation of 3,5-disubstituted 2thiohydantoins were described by Drobnica and Augustin [77, 78]. It is thought that during the cooking or processing of food and also in the human digestive system, isothiocyanates in foods react partly with free amino acids to form their corresponding 2-thiohydantoins. Therefore, we attempted to identify the 2-thiohydantoins formed by the reaction of MTBI with some selected amino acids, i.e., L-phenylalanine (Phe), L-methionine (Met), and L-leucine (Leu). When a mixture of MTBI containing in total 1.0 mmol of the isomers and the equivalent amount of amino acid was incubated at pH 8.0 for 48 h, a sufficient amount of product to identify by UV, IR, MS, ^H-NMR, and l^C-NMR was obtained as a crystal. The products (Fig. (10)) obtained from the reaction mixtures of MTBI with Leu, Phe, and Met were respectively identified as 3-[4(methylthio)-3-butenyl]-5-isobutyl-2-thiohydantoin (1) (UV Xmax (EtOH) : 267 nm; HR-EI-MS m/z (M+) : Calcd. for C12H20ON2S2, 272.1017; found, 272.0969; LR-EI-MS m/z (%, rel. intens.) : 272 (M+, 19), 225 (57), 100 (91), 85 (100); IR vmax (KBr) cm-l; 3186 (NH), 1749 (C=0), 1527 (N-C=S), 1163 (C=S); ^H-NMR 6 ((CD3)2CO) ppm: 9.11 (IH, s), 6.08 (lH,d, 7=15.1 Hz), 5.33 (lH,dt, 7=7.3 & 14.7 Hz), 4.21-4.25 (IH, m), 3.79 (2H, t, 7=7.3 Hz), 2.42-2.52 (2H, m), 2.21 (3H, s), 1.89-1.98 (IH, m), 1.57-1.70 (2H, m), 0.97 (3H, d, 7=6.8 Hz), 0.96 (3H, d, 7=6.4 Hz); 13C-NMR 6 ((CD3)2CO) ppm: 184.4, 175.7, 127.4, 122.5, 58.3, 41.4, 40.8, 32.1, 25.2, 23.4, 21.9, 14.4), 3-[4-(methylthio)-3-butenyI]-5-benzyl-2thiohydantoin (2) (UV Xmax (EtOH) : 269 nm; HR-EI-MS m/z (M+) : calcd. for C15H18ON2S2, 306.0861; found, 306.0859; LR-EI-MS m/z (%, rel. intens.): 306 (M+, 6), 259 (27), 100 (43), 91 (15), 85 (100), 77 (4), 65 (4), 51 (2); IR vmax (KBr) cm-l; 3203 (NH), 1747 (C=0), 1519 (NC=S), 1162 (C=S); iH-NMR 6 (CDCI3) ppm: 7.70 (IH, s), 7.32 (IH, t, 7=6.8 Hz), 7.27 (IH, t, 7=7.3 Hz), 7.20 (IH, d, 7=6.8 Hz), 6.02 (IH, d, 7=15.1 Hz), 5.26 (IH.dt, 7=7.3 & 14.6 Hz), 4.30 (IH, dd, 7=3.9 & 8.8 Hz), 3.77 (2H, t, 7=7.3 Hz), 3.28 (IH, dd, 7=3.9 & 14.2 Hz), 2.87 (IH, dd, 7=8.8 & 14.2 Hz), 2.30-2.42 (2H, m), 2.21 (3H, s); 13C-NMR 6 (CDCI3) ppm: 183.6, 173.4, 134.7, 129.2, 129.2, 129.0, 127.6, 126.9, 121.2, 60.4, 40.5, 37.6, 31.2, 14.2), and 3-[4-(methylthio)-3-butenyl]-5-[2(methylthio)-ethyl]-2-thiohydantoin (3) (UV Xmax (EtOH) : 268, 228 nm; IR vmax (EtOH) cm-l: 3567 (NH), 1740 (C=0), 1541 (N-C=S) , 1157 (C=S); LR-EI-MS m/z (rel. intens. %) : 290 (M+, 2.6), 243 (15.8), 100 (34.2), 85 (100), 72 (3.8), 61 (19.1); ^H-NMR 6 ((CE>3)2CO) ppm: 9.12 (IH, s), 6.09 (IH, d, 7=15.1 Hz), 5.35 (IH, dt, 7=7.3 & 14.7 Hz), 4.354.38 (IH, m), 3.81 (2H, t, 7=6.8 Hz), 2.66 (2H, t, 7=7.3 Hz), 2.42-2.53 (2H, m), 2.21 (3H, s), 2.11-2.19 (IH, m), 2.09 (3H, s), 1.97-2.03 (IH, m); 13C-NMR6 ((CD3)2CO) ppm: 184.6, 175.2, 127.4, 122.6, 58.6, 40.9, 32.1,

1090 31.5, 29.6, 14.9, 14.4) [79]. We studied the effect of pH on the formation of 2-thiohydantoins using allyl isothiocyanate instead of MTBI. The formation was best at alkaline pH's, but the 2-thiohydantoins could be formed at neutral and weakly acidic (pH 5.0-7.0) pH's, suggesting that a certain amount of 2-thiohydantoin derivatives can be formed in isothiocyanate-containing foods or seasonings, and even in the human digestive tract [79]. We are now investigating this finding to determine the time required for the 2-thiohydantoin derivative's formation.

Rv.

^NH, coo-

Ami no acid

R: -CH2-CH-(CH3)2 (1)

-CHrf^

(2)

-CH 2-CH2-S-CH3 (3) Fig. (10). Formation of 3,5-disubstituted 2-thiohydantoins from MTBI and amino acids.

1091 Biological activities of MTBI and MTBI-derived products A series of our studies has shown that the pungent principle of radish, i.e., MTBI, can be labile and reactive in food systems and in the human digestive tract to form various sulfur-containing compounds. We subsequently describe below some of the biological activities of the reaction products and their action mechanisms. Antimicrobial activity of MTBI There have been many investigations into the antimicrobial properties of naturally occurring and synthetic isothiocyanates [18-20]. Most of the investigations were done under such conditions as directly mixing the test compounds with a liquid or a moistened agar-medium. Esaki and Onozaki have reported their study of the antimicrobial activity of MTBI, in which this activity was evaluated using MTBI-containing agar-plates [80]. Considering the labile property of MTBI in the presence of water, the activity should be assayed in a manner that will minimize the reaction of MTBI with water and the components of the medium. Recent investigations have shown that the antimicrobial activity of volatile compounds was much higher when some microbial strains grown on agar plates came into contact with the compounds vaporized within the dishes [81, 82]. Accordingly, we assayed the antimicrobial activity of MTBI in a manner similar to the technique for gaseous contact, using five molds, three yeasts, and four Gram-negative and four Gram-positive bacteria [45]. In these experiments, MTBI was dissolved in acetone and added to a 10-mm diameter paper disk. The paper disk with MTBI was then placed on a small aluminum pipe which had been fixed on the inside of the dish's lid , on which the microorganism-containing agar-plate was placed and sealed tightly with a plastic film. The bacteria and fungi were then incubated at 37° C for 48-60 h and at 25"" C for 120-144 h, respectively. The resulting growth-inhibitory zones were measured. Simultaneously, the concentration of MTBI vaporized in the head space of the dish was monitored by FPD-GC (flame photometr}^ detector-gas chromatography) during the incubation. All the fungal growth examined was inhibited by MTBI at a dose greater than 2.5 [xmol (397.5 fxg), in which MTBI showed a highly inhibitory activity against the growth of Cladosporium colocasiae, Alternaria helianthi, Eurotium chevalieri, and Candida valida. As for the antibacterial activity, it was lower than that on the fungi, except in the case of Bacillus cereus and Enterobacter cloacae. On the other hand, the concentration of MTBI vaporized within the head space of the dishes was unexpectedly low. For example, the concentration in the vapor within the dishes added with 7.5

1092 nmol MTBI increased to a maximum after 12 h of incubation at 25° C, at which time the average amount estimated was ca. 12 nmol/dish (0.16% of the dose), which corresponded to approximately 30 ppb. This may be partly due to a high boiling temperature of MTBI. In the previous report by Esaki and Onozaki, the amount of MTBI required to completely inhibit the growth of microbes, including Escherichia coli, was over 23.5 [xmol /plate, although this value was obtained by using MTBI-containing media [80]. Obviously, MTBI vapor has a significantly high antimicrobial activity, being nearly comparable to that of allyl isothiocyanate [83]. The antimicrobial activity ofHMTP, MMTP, and dithiocarbamates On the basis of the inconsistencies between the labile properties of MTBI in moistened media and the results of Esaki and Onozaki, we presumed that the products formed from MTBI also had an antimicrobial activity. We next attempted to confirm this presumption [47, 84, 85]. For the antimicrobial studies, we conducted a preliminary examination in order to determine the pH-stability of the reaction products of MTBI by HPLC analysis of the remaining amounts of the reaction products in some media. Only HMTP, which has a weakly acidic hydroxy group within the structure, was unstable in a neutral or an alkaline medium. In the pH 5.0-6.0 media containing heart-infusion broth, soy bean-casein-digest broth or potato-dextrose agar, HMTP was stable in the pH 5.0 media for at least 196 h at 25° C, and a few percent of the initial amount was lost at pH 6.0 after 24 h at 37° C [85]. Thus, the minimum inhibitory concentrations (MICs) of the MTBI-derived 2-thioxopyrrolidines and dithiocarbamates were determined against fungi at pH 5.0 and against bacteria at pH 6.0 [47, 85]. The results are summarized in Table 4. Among the six MTBI-derived products, HMTP was found to be the most active toward a wide range of the microbes. The MICs of HMTP ranged from 50 to 400 [xg/ml, providing evidence that this degradation product of MTBI acts as an antimicrobial agent. Because HMTP has watersoluble properties but lacks any odor [84], it is possible that HMTP could be used for food processing and/or preservation. In contrast, a significantly reduced activity was observed with the structurally related products (£)- and (Z)-MMTP, suggesting that a hydrophilic hydroxy group contained within the HMTP molecule plays an important role in antimicrobial action. On the other hand, the three dithiocarbamates were more prominently active on the fungi than on the bacteria: their MICs toward the fungi ranged from 100 to 400 [Ag/ml, but those against the Gram-negative and Gram-positive bacteria were 400 |xg/ml or more. Dithiocarbamates have been known to be antifungal agents and are used for soil fungicides. With respect to the 2thioxopyrrolidines like HMTP and MMTP, however, no modes of action have been reported. We have studied the modes of antimicrobial action of

1093 HMTP using a fungal strain, E. chevalieri, and a bacterium, 5. epidermidis, because of their relatively higher sensitivities to HMTP [85]. The effects of HMTP on the growth of the fungal and the bacterial strains are shown in Fig. (11) and Fig (12). When the fungal spores were seeded in a medium with HMTP corresponding to the MIC, the growth was completely inhibited. Table 4. Antimicrobial activity the MTBI-derived products. (Modified from Matsuoka etal. [47]) MIC (mg/ml)^ of the Products Microorganism HMFP

(Z )-MMTP

{E )-MMTP

MBDC

(Z)-MBDC

{E )-MBDC

(Molds) ^ A. fumigatus

200

>800

800

200

200

200

C. colocasiae

200

400

800

200

200

100

E. chevalieri

200

400

400

200

100

100

-

-

-

-

M

racemosus

400

. d

(Yeasts)*' C. albicans

400

-

-

-

-

200

S.pombe

200

400

800

200

200

200

400

(Bacteria) ^ 50

400

400

>400

400

100

>800

>800

>400

400

400

S.typhimurium

200

-

-

-

-

>400

S. epidermidis

100

800

>800

>400

>400

>400

B. subtilis E. coli

a) Maximum dose was 800 \iglm\ for the thioxopyrrolidines and 400 MS^ml for the dithiocarbamates. b) Determined in glucose-peptone-broth at pH 5.0. c) Determined in nutrient broth at pH 6.0. d) Not determined.

However, neither delay in their growth curve nor the inhibition was observed when the same levels of HMTP were added to the medium after 48 h of growth, at which the fungal mycelia had apparently grown. When a half-MIC (100 fig/ml) of HMTP was added to the medium at the start of culture of the fungal spores, about 24 h delay in their growth curve was shown (data not shown). On the other hand, a significant reduction in the number of bacterium-survivors was observed after the addition of HMTP corresponding to the MIC (100 jxg/ml).

1094

60 4

40 4

20 4

^.

0 4

s "S 40 •c Q

20

Culturing (h) Fig. (11). Effect of HMTP on the growth of spores (A) and mycelia (B) of the fungal strain E. chevalieri. Modified from Matsuoka et al. [85], The fungal strain (lO"* spores) was grown in a glucose-peptone broth (pH 5.0) for 144 h at 25° C, and the growth was evaluated by measuring the weight of dried mycelium. Symbols: • , without HMTP; and O , with 200 fig/ml HMTP. The arrow indicates the time at which the corresponding amounts of HMTP were added.

These data clearly demonstrated that the antifungal activity of HMTP was due to its sporicidal action, and the antibacterial activity was attributed to its bactericidal property. We also investigated the effects on the cellular

1095 biosynthesis of DNA, RNA, proteins, lipids, and cell-wall peptideglycans using radioactive precursors, [methyl-^H]-thymidine, [5,6-^H]-uridine, [U14C]-leucme, [l,3-14c]-glycerol, and [l-l^CJ-glucosamine [85]. We used an HMTP-sensitive bacterium, S. epidermidis, for the experiment.

10»

-,

107

10^

u 10 5

J

104

J

10-^

1 0

1 1

1 2

r 3

Incubation time (h Fig. (12). Bactericidal effect of HMTP on the strain S. epidermidis. Modified from Matsuoka et al. [85]. The bacterial strain (10^ cells/ml) was incubated in a nutrient broth (pH 6.0) for 4 h at 37** C, during which the survivors were determined on nutrient broth-containing agar plates. Symbols: • , without HMTP; and O , with 200 ^ig/ml HMTP. The arrow denotes the time at which the corresponding amounts of HMTP were added.

A nearly identical pattern of inhibition was revealed in the biosynthesis of these biologically important molecules, suggesting that the biosynthetic systems of the molecules were damaged by the incorporation of HMTP (Fig. (13)). Also, oxygen uptake by the bacteria was completely inhibited within 15 min of the addition of 200 fxg/ml HMTP, in which little change in

1096 number of the bacterial survivors was observed; thus, the decrease in oxygen uptake was not attributable to bacterial death [85]. Although further studies are required to identify a more detailed mode of action, HMTP appears to be a nonspecific and multi-action growth-inhibitor for the microorganisms. 100

"5 o o o c

50

100

150

200

0

50

100

150

200

Dose of HMTP (mg/ml) Fig. ( 1 3 ) . Inhibitory effect of HMTP on incorporation of labeled compounds in S. epidemddis. Modified from Matsuoka et al. [85]. Symbols: • , [methyl-^H]-thymidine; • , [5,6-3H]-uridine; O , [U-l^CJ-L-leucine; A , [l,3-^^]-glycerol; and O , [l-^^C]-D-glucosamine. The incubation of the bacteria with these compounds lasted 40 min, and the incorporated radioactivity was measured by a scintillation counter.

The relationships between the structure and antimicrobial activity of 2thioxopyrrolidines The significant difference in the antimicrobial activities between HMTP and MMTP suggested that the side chain bound at the C3-position of the thioxopyrrolidine ring plays an important role in the antimicrobial activity. To compare the effects of differences in the side chain structure on antimicrobial activities, we synthesized fifteen HMTP- and MMTP-related thioxopyrrolidines consisted of seven 3-[(a-alkoxy)-a-methoxy]methyl-2thioxopyrrolidines (alkoxy group: methoxy; ethoxy; n-propoxy; n-butoxy; n-pentyloxy; n-hexyloxy; and n-heptyloxy), four 3-[(a-alkoxy)-amethylthio]methyl-2-thioxopyrrolidines (alkoxy group: methoxy; ethoxy; npropoxy; and n-butoxy), and four 3-(alkoxy)methylene-2-thioxopyr-

1097 rolidines (alkoxy group: methoxy; ethoxy; n-propoxy; and n-butoxy) [86]. MICs for these compounds were assayed against the fungal strains Eurotium chevalieri and Schizosaccharomyces pombe and the bacterial strain S. epidermidis [86]. The fungi (5 x 10^ colony forming unit (CFU) per ml) were cultured in a glucose-peptone broth (pH 5.0) for 72-120 h at 25"" C, and the bacteria (1 x 10^ CFU/ml) were grown in a nutrient broth (pH 5.0) for 24 h at 36° C. The compounds dissolved in dimethylsulfoxide were serially diluted so that the final concentrations were 0-800 [xg/ml in the media. The results are summarized in Table 5. Ikble 5. Relationship bet^veen structure and antimicrobial activitiy of the synthetic thioxopyrrolidines. (ModiHed from Matsuoka et al. [86]) M I C ( i mg/ml) Structure (A) 0CH3

L

y^oK

/' / ^

\

\ ^

R

E. chevarieli ^^

S, pombe^^

S. epidermidis"^

methyl ethyl n-propyl n-butyl n-pentyl n-hexyl n-heptyl

>800c) >800 800 800 400 100 50

>800 200 200 200 50 25 25

>800 >800 >800 800 800 400 200

methyl ethyl n-propyl n-butyl

>800 >800 800 400

800 200 400 >800

400 400 200 100

methyl ethyl n-propyl n-butyl

>800 >800 800 400

>800 >800 100 200

800 800 800 800

H SCH3

(B)

y^oR

/

\

( H

(C)

. - ^ ° ^ 1

/ \

\

v / ^

^r

a) Eurotium chevalieri and Schizosaccharomyces pombe were grown in a glucose-peptone broth at pH 5.0. b) Staphylococcus epidermidis was cultured in a nutrient broth at pH 6.0. c) Maximum dose was 800 mg/ml. (A): 3-[(a-alkoxy)-a-methoxy]methyl-2-thioxopyrrolidines;(B): 3-[(a-alkoxy)-a-methylthio]methyl-2-thioxopyrrolidines; and (C): 3-(alkoxy)-methylene-2-thioxopyrrolidines.

1098

Among the seven a-alkoxy-a-methoxy-substituted derivatives, dimethoxyderivative shov^ed no growth inhibition against the microorganisms tested at the highest dose examined. Similarly, ethoxy, n-propoxy, and n-butoxy derivatives had only weak antimicrobial activity against the fungal strain E, chevalieri and showed no inhibition of the growth of the bacterial strain, 5. epidermidis. The derivatives n-hexyloxy and n-heptyloxy showed a much higher growth inhibition activity against the fungi and bacteria. (A)-heptyloxy r-,

3.5

(A)-hexyloxy I

3.0

(B)-butoxy (Q-butoxy A^

I <

2.5 (A)-propoxyH (O-propoxy A^

2.0

I 1.0

(A)-pentyloxy

0(B)-propxy (A)-butoxy

I 1.5 Hydrophobic ity [Log (tj^)]

I 2.0

Fig. (14). Relationships between the antifungal activity and the hydrophobicity of the 2-thioxopyrrolidines. Modified from Matsuoka et al. [86]. The antifungal activity and the hydrophobicity are expressed as logarithmically converted reciprocals of MICs (mM) for the fungus E. chevalieri and logarithmically converted retention times from the ODS-HPLC analysis (column, LiChrosorb RP-8 (4.6 i.d. x 250 mm); mobile phase, water-acetonitrile-acetic acid (70 : 30 : 1); and detection, at 270 nm) of the compounds, respectively. The symbols denote the compounds grouped into (A), (B), and (C) in Table 5: B , 3-[(a-alkoxy)-a-methoxy]methyl derivatives ((A), alkoxy; n-propoxy, n-butoxy, n-pentyloxy, n-hexyloxy, and n-heptyloxy); • , 3-[(a-alkoxy)-a-methylthio]methyl derivatives ((B), alkoxy: n-propoxy and n-butoxy); and A , 3-(alkoxy)methylene derivatives ((C), n-propoxy and n-butoxy).

In the case of these a-alkoxy-a-methoxy-substituted derivatives, antimicrobial activity increased by lengthening the alkoxy chain, indicating that

1099 the activity depends largely on the hydrophobicity of the side chains of thioxopyrrolidines. On the other hand, a-alkoxy-a-methylthio-substituted derivatives were more active against the bacteria than against the fungi, and the compounds having longer alkoxy chains showed stronger activities. However, the growth-inhibitory activity of these a-alkoxy-a-methylthiosubstituted derivatives against the yeast appeared not to depend merely on the hydrophobicity of the molecules. Hence, the activity of the a-alkoxy-amethylthio-substituted derivatives may be affected not only by the hydrophobicity of the molecules but also by some steric effect of the sulfur atom in the side chain. In general, 3-(alkoxy)methylene-2-thioxopyrrolidines showed rather weak antimicrobial activities, although the activity against E, chevalieri increased by lengthening the alkoxy chain. Figure (14) shows the relationships between the antifungal activity and the hydrophobic properties of the 3-substituted 2-thioxopyrrolidines. Here, we can see that the antifungal activity of the 3-substituted 2-thioxopyrrolidines against E, chevalieri depends upon the hydrophobic properties of the molecules (correlation factor, r=0.953), except for several inactive compounds. Other factors such as the electric and steric characters of the compounds may also play an important role in the growth inhibition, in particular, against 5. pombe. There was no difference in the inhibitory activities between the (3R*, 6R*)- and (3R*, 6S*)-isomers of 3-[(a-n-butoxy)-a-methylthio]methyl-2-thioxo-pyrrolidine (data not shown), indicating that the difference in the antimicrobial activities between diastereomers of the other 2-thioxopyrrolidines is probably small [86]. Antimutagenic activity of HMTP, MMTP, and their structurally related 2thioxopyrrolidines In recent years, much attention has been focused on foodstuffs of plant origin because of their involvement in cancer prevention [87-89]. Among them, isothiocyanate-producing vegetables such as broccoli, cabbage, and cauliflower are considered to be possibly involved in anticarcinogenesis [23, 89]. As was mentioned above, the major isothiocyanate of radish roots was so labile in the presence of water that its role in antimutagenesis and anticarcinogenesis has been unclear. We studied the antimutagenic properties of the MTBI-derived product, i.e., HMTP and MMTP, and their structurally related 3-(alkoxymethylene)-2-thioxopyrrolidines (alkoxy group: methoxy; ethoxy; n-propoxy; and n-butoxy) [90, 91]. We evaluated the antimutagenic activity of these compounds with a mutation frequency assay using Salmonella typhimurium TA98 and/or TAIOO in the presence or absence of a rat liver-microsome fraction (S9) [92]. In the investigation, two heterocyclic multi-site carcinogens, i.e., IQ (2-amino-3-methy-limidazo[4,5-f|quinoline) and Trp-p-1 (3-amino-l,4-dimethyl-5H-pyrido[4,3-

1100 b]indole) [93,94], were used as the positive controls. These heterocyclic amines are metabolized by cytochrome p-450s, such as the P-450 lAl and 1A2 contained in the S9 mix to their corresponding ultimate mutagenic and carcinogenic forms that have a high reactivity with DNA [95-99].

100 H

T

(A)

75 J 50 J 25-

-O-

125H

Control

130

260

390

HMTP(ng/mI)

520

(£,Z)- Meth Eth n-Pro n-But MMTP MMTPand alkoxylated analogs at 500 ng/ml

Fig. (15). Antimutagenic activities of HMTP, (£,Z)-MMTP, and their structurally related 3-(alkoxy)methylene-2-thioxopyrrolidines against IQ (A) and Trp-p-1 (B). Modified from Uda et al. [90, 91]. Symbols: Ell, antimutagenic effect on IQ (A); and H , that on Trp-p-l (B). Control means the absence of the test compounds. The abbreviations Meth, Eth, n-Pro, and n-But denote methoxymethylene, ethoxymethylene, n-prop>oxynietliylene, and n-butoxymethylene derivatives of 3-(alkoxy)methylene-2-thioxopyrrolidines shown in Table 5. Salmonella typhimuriumTA9S was simultaneously treated with IQ (0.5 ^g), Trp-p-1 (0.5 \ig), and the test compounds (HMTP: 130-520 ^ig/ml; other compounds: 500 fxg/ml) in the presence of S9-mix for 20 min at 37° C. After 48 h incubation, the bacterial survivors and revertants were counted, and ratios of the numbers of revertants against those of the survivors were calculated. The ratio obtained from the control (treated with only IQ or Trp-p-1) is defined as 100% mutagenicity. A statistically significant difference between the control and the treating was determined by t-test (* (p < 0.05) and ** (p < 0.01).

1101 HMTP, which had been pretreated with S9 for 20 min at 37° C before contact with the bacterial strain TA98, had no mutagenic or bactericidal activity[, even at 500 ng/ml dose, but without the S9-treatment, this compound behaved as a slightly mutagenic agent at a dose over 390 ^g/ml dose, the level at which a cytotoxic effect was also shown. On the strain TAIOO, the spontaneous mutation frequencies that resulted with or without the S9 were 9- to 13-fold higher than those in TA98. Based on these results, we concluded that HMTP has a weak mutagenic property, but that it could be converted to form nonmutagenic and noncytotoxic product(s) by the action of the S9. MMTP and other 3-alkoxymethylene-derivatives had no cytotoxic or mutagenic activity against the bacterial strains with or without S9 at a 500 [xglml dose. The antimutagenic activities of the MTBIderived products against the heterocyclic amines in S. typhimurium TA98 are shown in Fig. (15).

100-

(B)

?

T-

-7 75a

s.

O

-r

A

LS

C

1 25-

1

15

1

30

1

45

1

n _

^ "

60 Incubation time (min)

1

1

0

15

30

45

60

Fig. (16). Inhibitory effects of HMTP and (£,Z)-MMTP on the S9-mediated metabolism of the heterocyclic amines IQ (A) and Trp-p-1 (B). Modified from Uda etal. [90, 91]. A mixture of 5.0 ^g of IQ or Trp-p-l in 25 ^.1 of dimethylsulfoxide (DMSO), 250 \ig of HMTP or (E,Z)-MMTP in 25 jxl of DMSO, and 450 ^il of S9-mix (S9 : co-factors = 1 : 9 ) was incubated for 60 min. The amounts of IQ and Trp-p-1 were periodically determined by a reverse phase-HPLC analysis. Symbols: w, IQ alone; O , Trp-p-l alone; B , IQ -i- HMTP; • , Trp-p-1 + HMTP; A, IQ + (£,Z)-MMTP; and A , Trp-p-l + {E,Z)-UUTP.

1102 HMTP showed a suppressing effect on the mutation frequency of IQ and Trp-p-1 at a dose over 390 fxg/ml. At a dose of 500 ng/ml, MMTP showed an antimutagenic activity against Trp-p-1, but its effect on the mutagenicity of IQ was much lower than on that of Trp-p-1, suggesting that MMTP behaves differently from HMTP against both heterocyclic amines and/or their S9-mediated metabolic activation system. The synthetic HMTP-related compounds exhibited antimutagenic activity against the heterocyclic amines at 500 \ig/ml dose, where the activity of the alkoxylated compounds appeared to decrease with a lengthening in the carbon chain of the alkoxy group. These results clearly demonstrated that HMTP as well as its structurally related 2-thioxopyrrolidines can act as an antimutagenic agent against the heterocyclic amines in the presence of S9. We subsequently studied the effects of HMTP and MMTP on the S9-mediated decline in the amount of heterocyclic amines and changes in the amounts of HMTP and MMTP during their incubations with the S9 mix. As shown in Fig. (16), IQ (5.0 fxg) and Trp-p-1 (5.0 ^ig) were largely reduced during the incubation with the S9 mix in the absence of HMTP. When the same amounts of the heterocyclic amines were incubated with the S9 mix in the presence of 250 ^ig HMTP, the reduction rates of both heterocyclic amines were considerably decreased. On the other hand, HMTP (250 \xg) was decreased to about 60% of the initial amount during 15 min of incubation with the S9 mix, but the reduction rate slowed thereafter. These results suggested that HMTP was partially metabolized with the S9 mix to form an inhibitor of the S9 fraction, by which HMTP inhibited the S9-mediated activation of the heterocyclic amines. This was considered to be the reason for the reduction of the mutation frequencies of IQ and Trp-p-1 in the presence of HMTP. MMTP also inhibited the S9mediated reduction of both IQ and Trp-p-1, though the reduction rates of IQ were lower than those of Trp-p-1. This may support the above results that MMTP showed a little effect on the mutagenicity of IQ. To investigate the mode of the inhibition of the S9-mediated reduction of the heterocyclic amines, Lineweaver-Burk plots were obtained for combinations of various amounts of IQ (1.25-20 ^ig) and HMTP (0-300 \xg) or MMTP (0-300 [ig). As Fig. (17) shows, an uncompetitive inhibition of the metabolic decrease of IQ was observed in the presence of 50-100 \ig HMTP. With 200-300 [ig HMTP, however, the IQ-metabolism was so strongly inhibited that no V'max value was obtained. In contrast, MMTP showed a noncompetitive inhibition at 50-300 fxg/ml doses. Since cytochrome P-450s are involved in the conversion of the heterocyclic amines into their activated Nhydroxy-forms, both HMTP and MMTP may be able to act as blocking agents for the P-450-mediated activation of the heterocyclic amines by such a mode of action as binding to the enzyme(s)-IQ complex or binding to the enzyme(s) [90,91].

1103 7n

60 80 1/ IQ (mM)

100

Fig. (17). Lineweaver-Biuk plots showing an uncompetitive and a noncompetitive inhibitions of S9-mediated reduction of IQ by addition of HMTP (A) and MMTP (B). Modified from Uda et al. [90, 91]. Symbols denote the amounts of HMTP and MMTP added • , 0 ^g ; O , 50 ^g; O , 100 ^g; A , 200 ^g; and ttj, 300 ^g. IQ was incubated with S9 in the presence or absence of represented amounts of HMTP or MMTP at 37° C for 15 min, and then the remained amount of IQ in the reaction mixtures was measured by an ODS-HPLC.

1104

The antimutagenic activity of 2-thiohydantoins Since the 3,5-disubstituted 2-thiohydantoins may be produced in isothiocyanate-containing foods and in the human digestive tract, we evaluated their antimutagenic activities with a Salmonella test using S. typhimurium TA98 [79]. The MTBI-derived 2-thiohydantoins, i.e., 3-[4(methylthio)-3-butenyl]-5-isobutyl-2-thiohydantoin, 3-[4-(methylthio)-3butenyl]-5-benzyl-2-thiohydantoin, and 3-[4-(methylthio)-3-butenyl]-5-[2(methylthio)ethyl]-2-thiohydantoin were studied for their antimutagenicity against the food bom-heterocyclic amine, IQ.

500 Dose (^ig/ml)

Fig. (18). Antimutagenic activities of the 2-thiohydantoins. Modified from Takahashi et al. [79]. (A), 3-[4-(methylthio)-3-butenyl]-5-isobutyl-2-thiohydantoin; (B), 3-[4-(methylthio)-3butenyl]-5-benzyl-2-thiohydantoin; and (C), 3-[4-(methylthio)-3-butenyl]-5-[2-(methylthio)ethyl]-2-thiohydantoin. Symbols: , Salmonella typhimurium TA9S was simultaneously incubated with IQ (0.5 \ig) and the 3,5-disubstituted 2-thiohydantoins (0-500 jAg/ml) for 20 min at 37° C and then cultured for 48 h at the same temperature; , the bacterial strain was treated with the same amount of IQ for 20 min at 37° C, rinsed with a buffered saline to eliminate the mutagen, and incubated with the test compounds for an additional 20 min. The bacteria were cultured for 48 h at 37° C. Mutagenicity is expressed in terms of the percentage in the revertant number found with or without the test compounds, where the mutagenicity of IQ is defined as 100%. Symbols: I—I and™, 3-[4-(methylthio)-3-butenyl]-5-isobutyl-2-thiohydantoin; A and A, 3-[4-(methylthio)-3-butenyl]-5-benzyl-2-thiohydantOLn; a n d O and w , 3-[4-(methylthio)3-butenyl]-5-[2-(methylthio)ethyl]-2-thiohydantoin.

1105 As Fig. 18 shows, the former two compounds, which were formed by the reaction of MTBI with Leu and Phe, were able to inhibit the IQ mutagenicity at concentrations of 125-500 ^ig/ml when incubated simultaneously with IQ in the presence of S9, and the compound derived from MTBI and Met showed an antimutagenic activity at a dose of 500 Hg/ml by simultaneous treatment with IQ. However, no inhibition of the IQ mutagenicity was observed when the test compounds were incubated with the bacteria that had been preincubated with IQ to induce mutation. This suggests that these 2-thiohydantoins are not involved in the repair of DNA lesions, but they are involved with the inactivation of IQ or the inhibition of S9-mediated metabolic activation of the mutagen. Dijferences in the mutagenic and antimutagenic properties of stereoisomers of tetrahydro-p-carbolines (PTCCs)

the

PTCC, which can be derived from MTBI and tryptophan, has two stereoisomers, (IS*, 3S*, 3'R*)-PTCC and (IR*, 3S*, 3'R*)-PTCC. In a fermented radish product, the amount of (IS*, 3S*, 3'R*)-isomer was much higher than that of the (IR*, 3S*, 3'R*)-isomer. We studied the difference in their mutagenic and antimutagenic properties using Salmonella typhimurium TA98 and TA100. In this study, both isomers were isolated from the reaction mixture of MTBI and L-tryptophan, and their FAB-MS, iH-lH-COSY, iH-l^C-COSY, and NOESY spectral data were compared along with their UV and IR spectra [60]. The spectral data showed that both PTCC-isomers were successfully obtained, and we were then able to carry out an Ames test. Neither PTCC had any bactericidal activity toward the bacteria at the highest dose (600 ^ig/plate) examined. An Ames test was then done in the presence or absence of S9. (IS*, 3S*, 3'R*)-PTCC had no mutagenic activity in a dose range of 0-600 ng/plate toward both tester strains, regardless of the presence or absence of S9. On the contrary, (IR*, 38*, 3'R*)-isomer showed obvious mutagenic activity, depending on its dose, toward S, typhimurium TA98 with S9, but not in the absence of S9. This isomer was not mutagenic toward S. typhimurium TAIOO in the same dose range with or without S9. However, the mutagenicity of (IR*, 3S*, 3'R*)-PTCC against TA98 was not considered strong, because the revertant number was below 100 even at a 600 jxg/plate dose. Since pcarboline derivatives enhanced the mutagenicity of IQ and Trp-p-1 [100], we examined the effect of PTCCs on the mutagenicity of IQ. Neither PTCC had an effect on the mutagenic or antimutagenic properties against IQ. In conclusion, the pungent principle of radish, MTBI, is a class of compounds that is extremely labile in the presence of water and can change into a variety of products having biological and physiological activities. There have so far been no studies on the activity that induces cancer-

1106 protective phase II enzymes like quinone reductase and glutathione Stransferase. Further investigations of this type should be conducted because of their potential importance to consumers and the food industry. ABBREVIATIONS (£)-MBDC (Z)-MBDC

= (£)-Methyl 4-(methylthio)-3-butenyldithiocarbamate. = (Z)-Methyl 4-(methylthio)-3-butenyldithiocarbamate. I H - I ^ C - C O S Y = Proton-Carbon Correlation Spectrometry. 1 H - 1 H - C 0 S Y = Proton-Proton Correlation Spectrometry. CFU = Colony Forming Unit. FAB-MS = Fast Atom Bombardment Mass Spectrometry. FPD-GC = Flame photometry detector-gas chromatography. HMTP = 3-(Hydroxy)methylene-2-thioxopyrrolidine. HR-EI-MS = High Resolution-Electron Impact lonization-Mass Spectrometry. IQ = 2-Amino-3-methylimidazo[4,5-f]quinoline. IR = Infrared Spectrometry. KBr = Potassium Bromide. LR-EI-MS = Low Resolution-Electron Impact lonization-Mass Spectrometry. MBDC = Methyl 4-(methylthio)butyldithiocarbamate. MeOH = Methanol. MIC = Minimum Inhibitory Concentration. MMTP = 3-(Methylthio)methylene-2-thioxopyrrolidine. MS = Mass Spectrometry. MTBI = 4-(Methylthio)-(£,Z)-3-butenyl isothiocyanate. NMR = Nuclear Magnetic Resonance Spectrometry. NOESY = Nuclear Overhauser Enhancement and Exchange Spectrometry. ODS = Chemically Octadecylated Silica Gel. PTCC = l-(2'-Pyrrolidinethione-3'-yl)-1,23,4-tetrahydro-b-carboline-3carboxylic acid. TMS = Tetramethylsilane. Trp-p-1 = 3-Amino-l,4-dimethyl-5H-pyrido[4,3-b]indole. UV = Ultra Violet Spectrometry. ACKNOWLEDGMENT The authors would like to thank Dr. H. Matsuoka, Gunma Women's Junior College, and Dr. Y. Yamada, Utsunomiya University, for their

1107 measurements of the NMR spectra. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]

Kjaer, A. In Progress in the Chemistry of Organic Natural Products; W. Hetz; H. Grisebach; G.W. Kirby, Ed.; Springer-Verlag: New York, 1960; pp. 122-176. VanEtten, C.H. In Toxic Constituents of Plant Foodstuffs; I.E. Liener, Ed.; Academic Press; New York, 1969; pp. 103-142. Fenwick, G.R.; Heaney, R.K.; MuUin, W.J. CRC Crit. Rev. Food Sci. Nutr., 1983, 18, 123. Roubiquet, P.J.; Boutron, F. J. Pharm. Chim., 1830, 96, 443. Ettlinger, M.G.; Kjaer, A. In Recent Advances in Phytochemistry; T.J. Mabry; R.E. Alston,; V.C. Runeckles, Ed.; Appleton-CenturyCrofts: New York, 1968; Vol. 1, pp. 59-144. Kjaer, A.; Skrydstrup, T. Acta Chem. Scand., 1987, B41, 29. Ettlinger, M.G.; Dateo, G.R; Harrison, Jr., B.W.; Mabry, T.J.; Thompson, C.P. Proc. Nat. Acad. Sci. USA., 1961, 47, 1875. Tookey, H.L.; Wolf, I.A. Can. J. Biochem., 1970, 48, 1024. Uda, Y.; Kurata, T.; Arakawa, N. Agric. Biol. Chem., 1986, 50, 2735. Uda, Y; Kurata, T; Arakawa, N. Agric. Biol. Chem., 1986, 50, 2471. Lee, C.-J; Serif, G.S. Biochim. Biophys. Acta, 1968, 165, 569. Kjaer, A.; Larsen, P.O. In Biosynthesis; T.A. Geissman, Ed.; The Chemical Society: London, 1973, Vol. 2, pp. 71-105. Rausch, T.; Butcher, D.N.; Hilgenberg, W. Physiol. Plant., 1983, 58,93. Astwood, E.B.; Greer, M.A.; Ettlinger, M.G. J. Biol. Chem., 1949, 181, 121. VanEtten, C.H; Gagne, W.E.; Robbins, D.J.; Booth, A.N.; Daxenbichler, M.E.; Wolff, I.A. Cereal Chem., 1969, 46, 145. Nishie, K.; Daxenbichler, M.E. Food Cosmet. Toxicol., 1980, 18, 159. Searle, L.M.; Chamberlain, K.; Rausch, T.; Butcher, D.N. J. Exp. Botany, 1982, 33, 935. Virtanen, A.I. Angew. Chem. Internat., 1962, 1, 299. Drobnica, L.;Zemanova, M.;Nemec, P.; Antos, K.; Kristian, P.; Stullerova,A.; Knoppova, V.; Nemec, Jr., P. Appl. Microbiol., 1967, 15,701. Chew, F.S. In Biologically Active Natural Products for Potential Use in Agriculture; H.G. Cutler, Ed.; American Chemical Society:

1108

[21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43]

Washington, DC, 1988; pp. 155-181. Talalay, P.; De Long, M.J.; Prochaska, H.J. Proc. Natl. Acad. Sci. USA., 1 9 8 8 , 8 5 , 8 2 6 1 . Zhang, Y.; Talalay, P.; Cho, C-G.; Posner, G.H. Proc. Natl. Acad. Sci. USA., 1992, 89, 2399 Zhang, Y.; Talalay, P. Cancer Res. Suppl., 1994, 54, 1976s. Wattenberg, L. W. Cancer Res., 1981, 4 1 , 2991. Wattenberg, L.W. Carcinogenesis, 1987, 8,1971. Morse, M.A.; Eklind, K.I.; Hecht, S.S.; Jordan, K.G.; Chok, CI.; Desai, D.H.; Amin, S.G.; Chung, F-L. Cancer Res., 1991, 5 1 , 1846. Verhoeven, D.T.H.; Verhagen, H; Goldbohm, R.A.; Van Den Brandt, P.A.; Van Poppel, G. Chemico-Biol. Interact., 1997, 103, 79. Brusewitz, G.; Cameron, B.D.; Chasseaud, L.F.; Gorier, K.; Hawkins, D.R.; Koch, H.; Mennicke, W.H. Biochem. J., 1977, 162,99. Gorier, K.; Krumbiegel, G.; Mennicke, W.H.; Siehl, H-U. Xenobiotica, 1982,12, 535. Mennicke, W.H.; Gorier, K.; Krumbiegel, G. Xenobiotica, 1983, 13, 203. Shapiro, T.A.; Fahey, J.W.; Wade, K.L.; Stephenson, K.K.; Talalay, P. Cancer Epidemiol., Biomarkers & Prevention, 1998, 7, 1091. Ashworth, M.R.F. The Determination of Sulfur-containing Groups, Academic Press: New York, 1972. Kawakishi, S; Namiki, M. J. Agric. Food Chem., 1982, 30, 618. Kawakishi, S.; Goto, T.; Namiki, M. Agric. Biol. Chem., 1983, 47, 2071. Kawakishi, S.; Muramatsu, K. Agric. Biol. Chem., 1966, 30, 688. Friis, P.; Kjaer, A. Acta Chem. Scand., 1966, 20, 698. Daxenbichler, M.E.; Spencer, G.F.; Carlson, D.G.; Rose, G.B.; Bnnker, A.M.; Powell, R.G. Phytochemistry, 1991, 30, 2623. Kjaer, A.; Madsen, J.0.; Maeda, Y.; Ozawa, Y.; Uda, Y. Agric. Biol. Chem., 1978, 42, 1715. Carlson, D.G.; Daxenbichler, M.E.; VanEtten, C.H.; Hill, C.B.; Williams, P.H. J. Am. Soc. Hortic. Sci., 1985, 110, 634. Ishii, G.; Saijo, R.; Nagata, M. Jpn. J. Food. Sci. Technol. {Nippon Shokuhin Kogyou Gakkaishi), 1989, 36, 739. Visentin, M.; Tava, A.; lori, R.; Palmieri, S. J. Agric. Food Chem., 1992; 40, 1687. Kawakishi, S.; Namiki, M.; Watanabe, H.; Muramatsu, K. Agric. Biol. Chem., 1 9 6 7 , 3 1 , 8 2 3 Kawakishi, S.; Namiki M. Agric. Biol. Chem., 1 9 6 9 , 3 3 , 4 5 2 .

1109 [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66]

Chen, C.-W; Ho, C.-T. J. Agric. Food Chem., 1998, 46, 220. Uda, Y.; Matsuoka, H.; Kumagami, H.; Shima, H.; Maeda, Y. Jpn. J. Food Sci. Technol. {Nippon Shokuhin Kogyo Gakkaishi), 1993, 40, 743. Uda, Y.; Ozawa, Y.; Ohshima, Y.; Kawakishi, S. Agric. Biol. Chem., 1990, 54, 613. Matsuoka, H.; Toda, Y.; Yanagi, K.; Takahashi, A.; Yoneyama, K.; Uda, Y. Biosci. Biotech. Biochem., 1997, 61, 2109. Nakamura, N. Biochem. Z., 1925, 164, 31. Shibayama, F.; Yamada, S.; Kikuchi, H.; Yamada, Y. Jpn. Kokai Tokkyo Koho JF 81, 1981, 104, 864. {Chem. Abstr. 1982, 96, 35076p.) Hasegawa, K.; Shiihara, S.; hvagavva, T.; Hase, T. Plant Physiol, 1982, 70, 626. Harada, N.; Hagiwara, H.; Ono, H.; Uda, H.; Ohba, S.; Kubota, M.; Nishiyama, S.; Yamamura, S.; Hasegawa, K.; Sakoda, M. Tetrahedron Letters, 1991, 32, 6757. Kosemura, S.; Yamamura, S.; Hasegawa, K. Tetrahedron Letters, 1993. 34, 481. Matsuoka, H.; Uda, Y.; Mitani, K.; Yoneyama, K.; Maeda, Y. Biosci. Biotech. Biochem., 1996, 60, 914. Matsuoka, H.; Toda, Y.; Yoneyama, K.; Uda, Y. Phytochemistry, 1998,47, 975. Maeda, Y.; Ozawa, Y.; Uda, Y. Jpn. J. Food Sci. Technol. {Nippon Shokuhin Kogyo Gakkaishi), 1982, 29, 522. Ozawa, Y.; Kawakishi, S.; Uda, Y.; Maeda, Y. Agric. Biol. Chem., 1990,54, 1241. Ozawa, Y.; Uda, Y.; Kawakishi, S. Biosci. Biotech. Biochem., 1992, 56, 144. Ozawa, Y.; Uda, Y.; Kawakishi, S. Jpn. J. Food Sci. Technol. {Nippon Shokuhin Kogyo Gakkaishi), 1992, 39, 490. Ozawa, Y.; Uda, Y.; Kawakishi, S. Agri. Biol. Chem., 1990, 54, 1849. Ozawa, Y.; Uda, Y.; Matsuoka, H.; Abe, M.; Kawakishi, S.; Osawa, T. Biosci, Biotechnol. Biochem., 1999, 63, 216. Whaley, W.H.; Govindachari, T.R. In Organic Reactions Coll.; R. Adams Ed.; Wiley; New York, 1951;.Vol. 6, pp. 151-X. Buckholtz, N.S. Life Sci., 1980, 27, 893. Beck, O.; Holmstedt, B. Food. Cosmet. Toxicol., 1981, 19, 173. Beck, O.; Bosin, T.R.; Lundman, A. J. Agric. Food Chem., 1983,31,288. Wakabayashi, K.; Ochiai, M.; Saito, H.; Tsuda, M.; Suwa, Y.; Nagao, M.; Sugimura, T. Proc. Natl. Acad. Sci. USA., 1983, 80, 2912. Bosin, T.R.; Krogh, S.; Mais, D. J. Agric. Food Chem., 1986,

1110

[67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78]

34, 843. Adachi, J.; Mizoi, Y.; Naito, T.; Ogawa, Y.; Uetani, Y.; Ninomiya, I. J. Nutr. 1991, 121, 646. Herraiz, T. J. Agric. Food Chem., 1996, 44,3057. Chu, N.T.; Clydesdale, F.M. J. Food ScL, 1976, 4 1 , 891. Bobbitt, J.M.; Willis, J.P. J. Org. Chem., 1980, 45, 1978. Kato, S.; Nakase, A. Jpn. J. Food Sci. Technol. (Nippon Shokuhin Kogyo Gakkaishi), 1 9 8 6 , 3 3 , 659. Ozawa, Y.; Uda, Y.; Ohshima, T.; Saito, K.; Maeda, Y. Agric. Biol. Chem., 1990, 54, 605. Kj£er, A.; Madsen, J.0.; Maeda, Y.; Ozawa, Y.; Uda, Y. Agric. Biol. Chem., 1978, 42, 1989. Nielsen, J.K.; Olsen, O.; Pedersen, L.H. ; Sorensen, H. Phytochemistry, 1984, 23, 1741. Harborne, J.B.; Corner, J.J. Biochem. J., 1961, 81, 242. Edman, P. Acta Chem. Scand., 1950, 4, 283. Drobnica, L ; Augustin, J. Collect. Czech. Chem. Commun., 1965, 30, 99. Drobnica, L.; Augustin, J. Collect. Czech. Chem. Commun.,

1965,30,1221. [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91]

Takahashi, A.; Matsuoka, H.; Ozawa, Y.; Uda, Y. J. Agric. Food Chem., 1998, 46, 5037. Esaki, H.; Onozaki, H. Jpn. J. Food Nutr. (Eiyo Shokuryo Gakkaishi), 1982, 35, 207. Inoue, S.; Goi, H.; Miyauchi, K.; Muraki, S.; Ogihara, M.; Iwanami, Y. J. Antibact. Antifung. Agents, 1983, 11, 609. Goi, H.; Inoue, S.; Iwanami, Y. J. Antibact. Antifung. Agents, 1985, 13, 199. Issiki, K.; Tokuoka, K.; Mori, R.; Chiba, S. Biosci. Biotech. Biochem., 1992, 56, 1476. Uda, Y.; Matsuoka, H.; Shima, H.; Kumagami, H.; Maeda, Y. Jpn. J. Food Sci. Technol. (Nippon Shokuhin Kogyo Gakkaishi), 1993, 40, 801. Matsuoka, H.; Takahashi, A.; Yanagi, K.; Uda, Y. Food Sci. Technol. Int. Tokyo, 1997, 3, 353. Matsuoka, H.; Takahashi, A.; Y oneyama, K.; Uda, Y. Biocontrol Sci., 1 9 9 9 , 4 , 101. Graham, S. Cancer Res. SuppL, 1983, 43, 2409s. Caragay, A.B. Food Technol. 1992, April. 65. Block, G.; Patterson, B.; Subar, A. Nutr. Cancer, 1992, 18, 1. Uda, Y.; Shimizu, A.; Hayashi, H.;Ozawa, Y. Proceeding for 5th International Congress on Mechanisms of Antimutagenesis and Anticarcinogenesis (Okayama Univ. Japan), 1996, Page 119. Uda, Y.; Hayashi, H.; Takahashi, A.; Shimizu, A. Food Sci. Technol., 2 0 0 0 , 3 3 , 37.

nil [92]

Nhomi, T.; Matsui, M.; Ishidate, M. Toxicol. Forum, 1986, 9, 189. [93] Aeschbacher, H.U.; Turesky, R.J. Mutation Res., 1991, 259, 235. [94] Ohgaki, H; Takayama, S.; Sugimura, T. Mutation Res., 1991, 259, 399. [95] Ishii, K.; Yamazoe, Y.; Kamataki, T.; Kato, R. Cancer Res., 1980, 40, 2596. [96] Yamazoe, Y.; Shimada, M.; Kamataki, T.; Kato, R. Cancer Res., 1983, 43, 5768. [97] Soucek, P.; Gut, I. Xenobiotica, 1992, 22, 83. [98] Masson, H.A.; loannides, C.; Gorrod, J.W.; Gibson, G.G. Carcinogenesis, 1983, 4, 1583. [99] Aldrick, A.J.; Rowland, I.R. Mutation Res., 1985, 144, 59. [100] Friedman, M.; Cuq, J.L. J. Agric. Food Chem., 1988, 36, 1079.

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. Allrightsreserved,

1113

STRUCTURE, OCCURRENCE AND ROLES OF CARBOHYDRATES WITH THE HEXO-D-MANNO CONFIGURATION

NORMAN K. MATHESON

Department ofAgricultural Chemistry and Soil Science, The University of Sydney N.S W, Australia, 2006 ABSTRACT: Aspects of the occurrence, structure, properties, metabolism, biological activity and uses of naturally-occurring carbohydrates based on the hexo-D-manno structure have been discussed. These include D-mannose, D-mannitol, oligosaccharides, a-mannoproteins of fungi, glycoproteins, p-mannan, p-glucomannan, a-galacto-Pmannan, galacto-P-glucomannan, alginic acid, as well as bacterial polysaccharides and plant gums and mucilages containing D-mannose. Some in vitro syntheses of mannans and manno-oligomers have also been described.

INTRODUCTION The most abundant naturally-occurring sugar is D-glucopyranose, with equatorial hydroxyl groups on C-2, 3 and 4 and an equatorial hydroxymethyl group on C-5 when in the usual ^Ci conformation. It is the thermodynamically most stable aldohexose structure. Manno- and galacto- configurations in the "^Ci conformation, which also occur extensively, have one axial OH - at C-2 in mannose - and both avoid a 1:3 diaxial interaction as either anomer. A six carbon chain is the shortest that can provide an unstrained hemi-acetal ring (six-membered) in which all carbons are substituted with hydrophilic -OH groups and a -CH2OH. The biosynthesis of D-mannose involves the conversion of the D-gluco isomer via the D-fructo (as the 6-phosphates) by isomerization at C-2; differing from the conversions of other sugars (including D-galactose) which proceed via nucleoside diphospho sugars, the formation of which requires a supply of nucleoside triphosphates. However, despite the widespread distribution of D-mannose and its utilisation in glycolysis and

1114

biosynthesis in most species, under some circumstances it can be deleterious to growth and even toxic at low concentrations. The related hexitol D-mannitol also occurs extensively. It is stored in fungi and plants and is readily utilised. Mannosyl glycosides and oligosaccharides, particularly lipid-linked, occur and phosphoric esters are participants in metabolism. Related monomeric structures, such as amino sugars, deoxy-sugars, ethers, esters and uronic acids have also been detected. Polysaccharides containing D-manno-hexo structures can be broadly grouped into those in which these monomers constitute the majority of the polymer and those in which they contribute a minority of residues, aMannans and a-phospho-mannans are found as proteoglycans in the cell walls of fungi, and a-mannans and a-glucomannans as fibrous material in the walls of algae. The glycoproteins, which are of general distribution, also contain a-mannosyl Unkages. Fungal mannoproteins and the glycoproteins are related structurally and biosynthetically. P(Gluco)mannans and their substituted polymers - a-galacto-p-mannan and galacto-P-glucomannan - are associated with plants. A major sugar in the seaweed polysaccharide alginic acid is D-mannuronic acid. The viscosity and gel-forming characteristics of a-galacto-P-mannan, p-glucomannan and alginic acid, which is also a polyelectrolyte, leads to their use in the food, pharmaceutical, paper and textile industries. D-mannose provides a hexose with many properties similar to those of D-glucose, even reacting with the same enzyme (hexokinase) in one instance, but clearly differentiated in other enzymic reactions. This allows the production of a different set of oligomeric and polymeric structures with (1-3), (1-4) and (1-6) inter-residue linkages, having many properties similar to glucans but quite separate modes of biosynthesis and cellular recognition. The conformations of (1-3), (1-4) and (1-6) linked mannans resemble those of the equivalent glucans. Structural polymers, although having many similarities to glucans with the same linkage, have somewhat different physical properties, and energy reserve polysaccharides are utilised (apart from hexokinase) by a separate set of enzymes. a(l-2) Mannan provides a linkage of two axial bonds which is not available for the gluco conformation. D-mannose is also present in smaller amounts in other polysaccharides, such as bacterial lipo- and exopolysaccharides, and the plant gum exudates and mucilages. Chemically synthesized oligomers and polymers have also been prepared.

1115

In this text, biochemical style abbreviations - the condensed system of symbolism - have been used for the formulae of carbohydrates [1]. Where more convenient M has also been used as a symbol for D-mannopyranose. Unless described otherwise, chiral compounds are D- and glycosidically linked sugars are pyranose. In oligosaccharide formulae a reducing-end sugar is italicised. Italicised reference numbers indicate reviews and these provide early information and lists of references. P represents a phosphate group. DISCUSSION 1. Metabolism of D-Mannose Man enters metabolism as Man 6-phosphate (Man 6-P) in a reaction catalysed by hexokinase [EC 2.7.1.1] (reaction A) in which the phosphate donor is MgATP and the other product of reaction MgADP [2,5,4].

Man+ MgATP Man 6-P Man6-P

A -> B ^ D ^

Man 6-P -f MgADP C Fru6-P ^ Glc6.P Manl-P

Man 6-P can then be converted into Fru 6-P by phosphomannoisomerase (PMI) [EC 5.3.1.8] [5,6] in an equilibrium reaction of 40% Man 6-P and 60% of Fru 6-P (reaction B). Isomerisation of Fru 6-P by phosphoglucoisomerase [EC 5.3.1.9] (reaction C) in another equilibrium reaction gives Glc 6-P and entry into glycolysis. Free Man can also be recovered through phosphorylation to Man 6-P by hexokinase. Man 6-P is converted to Man 1-P by phosphomannomutase [EC 5.4.2.8] [7-10a] (reaction D) for further metabolism in the D-manno configuration. Although some kinases specific for Glc have been detected [2,3], hexokinase generally phosphorylates both Glc and Man even when obtained from sources in which free Man is absent. A number of legume seeds (fenugreek, guar, carob and honey locust) contain a reserve agalacto-p-mannan. After germination, when depolymerisation occurs, the released Man is readily utilised. In honey locust {Gleditsia triacanthos) attempts to separate gluco- and manno-kinase activities were unsuccessful

1116

[4]: the rates of loss of activity of the purified enzyme by slow heat inactivation were similar with both substrates. The Michaelis constants (Km) were similar for both hexoses, as also were the specificities [Vmax/Km] with the P-pyranose anomers. Despite the general occurrence of Man in cellular glycoprotein and its widespread presence in structural polysaccharides of plants, as well as the ability of hexokinase to phosphorylate both Glc and Man - which extends to analogues such as 2deoxy-D-glucose (2dGlc), D-glucosamine and 1,5-anhydro-D-glucitol [4,11] - Man (and analogues) are deleterious to many organisms, such as mammalian cells [12], yeast mutants [13], bees [14] and a range of plants and plant parts [3]. Man and Glc analogues act at the same site on the enzyme as Glc, both as substrates and competitive inhibitors to Glc. Inhibitor constants for Man (Ki, inhibitor concentration giving 50% inhibition) versus Glc for the hexokinase from brain [15] and honey locust seeds [4] of .036 and .031mM have been estimated. Toxic effects are associated with undetectable or very low levels of endogenous PMI, leading to an inability to further metabolize the Man 6-P synthesized by hexokinase. Where Man and 2dGlc are deleterious their 6-phosphates accumulate [14, 16, 17]. When 2dGlc, for which there is no available further reaction for the 6-phosphate (except hydrolysis), is supplied even more species are affected [18, 19-21]. One proposal for the cause of this effect of Man and Glc analogues not further metabolised in species with undetectable or very low levels of PMI, is that the accumulation of Man 6-P (or analogue-P) sequesters inorganic phosphate (Pj), lowering endogenous concentrations of Pi, and so denying Pj for the synthesis of Glc 6-P. Reductions in Pi have been demonstrated in many cells [5, 12, 16-18, 20, 22, 23]. In 5 minutes [Pi] decreased markedly in ascites tumour cells [12] maintained in 1 mM 2dGlc and, in yeast [18] in 20 mM 2dGlc, dropped from 34 to 25 |amole per g in 4 hr. The Pi content of wheat leaf fragments [22] in 10 mM Man was lowered in 6 hr from 4.4 nmole per g fresh weight to 2.9. Depletion has also been noted in pear fruits [17] and spinach leaves [3], ^^P n.m.r. spectra of maize leaves [23] incubated in 25 mM Man for 1 hr had a lowered [Pi]. In spectra of maize root tips [20], incubated for 2 hr with 50mM 2dGlc, the peak attributable to cytoplasmic Pi virtually disappeared, while vacuolar Pi decreased slowly. In another study [16], at a much lower concentration of Man (2mM), after 30 hr, cytoplasmic [Pi] remained constant but Pi disappeared from the vacuole.

1117 The very sensitive inhibition of mammaUan hexokinase by the product Glc 6-P (vs Mg ATP, Ki 0.04 mM) suggests the possibility of inhibition by accumulating Man 6-P or analogue-P. However, although in yeast [24] mixed inhibition by Glc 6-P does occur, it is only at much higher levels (20-30mM). In wheat germ hexokinase [25], with Man as substrate, Glc 6-P inhibited non-competitively to MgATP (Kj 16.2 mM). With the enzyme from potato tubers [26], it acted non-competitively against Glc (Ki 4.1 mM) and, since a cellular concentration of Glc 6-P of approximately 3-7 mM was estimated, it was concluded that a capacity for a moderate effect may be present. When analogue inhibition occurs, although [Glc 6-P] decreases, there is an increase in the concentration of analogue-P, which may inhibit hexokinase. ^^P n.m.r. spectroscopy indicated large peaks ascribed to Man 6-P in potato slices after 12.5 hr in 10 mM Man [16] and in maize root tips after 30 hr in 2 mM Man [20]. When pear fruits [17] were infiltrated with 100 mM Man for 2 days, the [Man 6-P] rose to 491 nmole per g fresh weight, and in 300 mM Man to 550 nmole per g. Bees fed Man [14] accumulated 11 ^mole per g of 6phosphate after 4 hr. Similarly, in yeast [18] in 2 mM 2dGlc, the 6phosphate content rose to 50 |imole per g; P n.m.r. spectroscopy [19] indicated that in 25 mM Glc there was a 25% increase in the content of 2dGlc 6-P when the external concentration of 2dGlc was increased from 5 to 10 mM. A large spectral peak attributable to 2dGlc 6-P appeared in maize root tips [20] after 2 hr in 50 mM 2dGlc and rat heart perfUsed with 2dGlc attained a level of 55 mM 2dGlc 6-P[21]. With Glc as substrate, yeast hexokinase [15] was inhibited by Man 6P - with mixed inhibition versus ATP - with a Kjs value of about 2-3 mM. The enzyme from beef heart [21] with 2dGlc as substrate, was inhibited by 2dGlc 6-P against ATP and non-competitively to 2dGlc with a Ki of 1.4 mM. However, when phosphate transfer from ATP and phosphorylation of Glc were measured for hexokinase from honey locust seeds [27] in the presence of Man 6-P, negligible inhibition against Glc and a very slight inhibition against ATP were observed (Kjs 52 mM). The results indicate that, whereas there could be a possible effect after significant accumulation of analogue-P, it was unlikely that analogue phosphates were associated with the initiation of inhibition. The supply of Man to sensitive species also leads to decreases in phospho compounds such as ATP, polyphosphate, Glc 6-P and phosphoenolpyruvate, significant high-energy intermediates in fiirther metabolism. The decreases of these are more pronounced than those of

1118

Pj. Where measured, much smaller decreases in ADP have been noted [12, 22, 28]. In bees[14] fed Man, [ATP] changed from 1.6 to 0.3 ^imole per g in 4 hr. In maize scutellum slices [28] after 1 hr in 50 mM Man, [ATP] decreased from 0.40 to <0.05 |imole per g fresh weight, while the ADP level was maintained: Glc 6-P and Fru 6-P also almost completely disappeared. In wheat leaf fragments [22], held in 10 mM for 6 hr, ATP content diminished from 30 to 8 nmole per g fi-esh weight, while ADP showed a smaller decrease. In roots [29] with 1 mM Man, [ATP] decreased to 46% of the control after incubation for 3hr. Decreases in [ATP] have also been detected in pear fruits [17] and sugar beet leaves [30] in the presence of Man. ^^P n.m.r. spectra of maize leaves [23], incubated in 25 mM Man for 1 hr, showed that spectral peaks due to Glc 6-P and phosphoenolpyruvate disappeared, together with large decreases in those of adenylates. The ATP content of ascites tumour cells [12] in 1 mM 2dGlc decreased (from 3.2 to 0.5 |imole per ml of cells in 5 min), but ADP increased. Polyphosphate in yeast [18], incubated in 20 mM 2dGlc, dropped from 41 to 5 |Limole per g in 4 hr, and in a study with "^^P n.m.r. spectra [19] incubation of yeast in 10 mM 2dGlc reduced polyphosphate to an undetectable level. Decreases of [ATP] in Chlorella [31] have been recorded with 2dGlc and GICNH2. Hexokinases are inhibited by MgADP. Ki values at the millimolar level were obtained for enzymes from brain, muscle and liver [32]. Inhibition has also been seen with calf brain enzyme [15] and in yeast [33]. MgADP competitively inhibits against ATP and non-competitively with Glc. It reduced activity with pea seed [34, 35] and maize kernel enzymes [36]. In a wheat fraction [25] a Kj of ImM was determined in non-competitive inhibition against Glc. ADP and 5^ AMP were competitive inhibitors with respect to ATP for the enzyme from loranthus leaves [37]. Two hexokinases from potato [26] were inhibited competitively (Kj 0.04 and 0.108) against ATP. These values were 2 to 3 fold less than the Km values for Mg ATP. It was considered that MgADP would be unlikely to be an effective inhibitor under aerobic conditions, when high ATP:ADP ratios are expected, but that, if the ratio fell below unity (as under anaerobic conditions) inhibition would become significant. A particulate maize enzyme [38] inhibited non-competitively with regard to both ATP and Glc (Kj 0.034 mM). From a consideration of these inhibition results another proposal for the effect of Man on species that have an undetectable or very low level of PMI is that substrate analogues of Glc compete for hexokinase, reducing the level of Glc 6-P. In cells transport and phosphorylation are closely

1119 linked and both may be involved in this competition. When white mustard and w^heat seeds [27] were imbibed in the presence of analogues of Glc that were substrates for hexokinase - Man; 2dGlc; 1,5-anhydro-Dglucitol; 2,5-anhydro-D-talitol - the degree of inhibition of germination and seedling growth was inversely related to the phosphorylation coefficient (Vmax/Km analogue ^Vmax/Km Glc). If the seed contained a significant level of PMI - as in mung beans and fenugreek - Man was not an inhibitor. 2dGlc inhibited germination in all four species. Then, with 2dGlc and if Man 6-P cannot be converted to Fru 6-P at a sufficient rate, the level of glycolytic turnover leading to regeneration of ATP by both substrate level and oxidative phosphorylation would decrease. Nonmaintenance of the ATP level relative to ADP would cause inhibition of hexokinase and an even lower rate of synthesis of Glc 1-P, leading to a spiral of diminishing production of ATP. Since ATP is also necessary for so many other cell processes - polynucleotide metabolism, protein synthesis, transport and phosphoprotein conversion - this reduction would have many further consequences. Carbon catabolite repression, the repression of the expression of genes that code for enzymes involved in the metabolism of other carbon sources, has been extensively studied in microorganisms. Glc, Man and Fru as well as related sugars are effective repressors [39]. It has also been observed with higher plant cells [40, 41]. However,in germinating seeds, deficient in PMI, the toxic effects of Man and analogues that were substrates for hexokinase were strongly present at concentrations at which sucrose, Glc and Gal were without effect. In yeast, extensive metabolism of the sugar causing repression is not necessary [39]. A kinase able to phosphorylate the sugar is required, so a possible role for Glc 6-P has been suggested, and if it is involved then analogues may also be effective. However, electroporation of Glc 6-P into plant protoplasts gave no repression [41]. 2. D-Mannitol D-mannitol is widely distributed in plant families from photosynthetic bacteria to fungi, brown and green algae, mono- and dicotyledons, as well as animals [42-45], It is the most abundant hexitol and concentrations of more than 1% dry weight are commonly found; often much higher. In seaweeds it accumulates in the summer, whereas in many cases in higher plants the levels are higher in winter. It serves as a reserve carbohydrate.

1120

but is also an early product from photosynthetic reduction of CO2 [46] being involved in photoassimilate partitioning [47]. In some instances higher incorporation of CO2 into D-mannitol than into sucrose has been found. In celery protoplasts the majority was in the vacuole with some in the cytosol [48]. In higher plants it can be a major form of translocated carbohydrate, when it is transported in the phloem [48, 49]. Other roles are as an osmoregulator [50] and in cryoprotection. A more active role for mannitol in fungi has been proposed [51] in the infection of plants by these organisms. Mannitol quenches reactive oxygen species superoxide, O2" and H2O2 - that are produced by plants and animals in defence against pathogens. Evidence has been presented that phytopathogenic fungi use mannitol to suppress this defence. Increased mannitol production by yeast has been shown to protect it from oxidative injury. Also, in plants that do not normally contain mannitol, the enzyme mannitol dehydrogenase (H) is induced, providing a mechanism to counteract the effect of secretion of mannitol by invading pathogens. D-Mannitol and D-mannitol 1-phosphate undergo a number of enzymic reactions:

Man 6-P

E ^=^ mannitol 1-P; co-factor NADP"^, catalysed by mannose 6-P reductase [EC 1.1.1.-]

mannitol 1-P

F -> mannitol + Pj, catalysed by mannitol 1-P phosphatase [ECS. 1.3.22]

mannitol 1-P

G ^-? Fru 6-P, co-factor NAD^, catalysed by mannitol 1-P dehydrogenase [EC 1.1.1.17]

H mannitol *«? Fru, catalysed by mannitol dehydrogenase co-factor NADP^ [EC 1.1.1.138] co-factorNAD^ [EC 1.1.1.67] co-factor cytochrome [EC 1.1.2.2]

1121

I mannitol *=^

Man, co-factor NAD"^, catalysed by mannitol 1 -oxido reductase [EC 1.1.1.-]

Also, when many bacterial cells are grown in a medium containing Dmannitol, an enzyme, for which the source of phosphate is phosphoenol pyruvate, phosphorylates and transports D-mannitol (and other sugars) into the cytoplasm [52]. Reactions E and F have been studied in celery [46, 47, 53]. The enzymes are located in the cytosol. The oxidation of Dglyceraldehyde 3-P by glyceraldehyde 3-P dehydrogenase provides a source of reduced co-factor to drive reaction E from left to right. Glyceraldehyde 3-P is derived from dihydroxy acetone-P, exported from the chloroplast. It has been proposed that mannitol fimctions as a temporary store of carbon and reducing power and in translocation, similarly to sucrose: in celery both reactions with sucrose and mannitol are present [48]. Mannitol 1-P dehydrogenase (reaction G) has been described from bacteria, algae and fimgi [54-56]. The NAD^ and NADP"^ dependent oxidations of Man to Fru (reaction H) have also been studied [56-59]. The enzyme catalysing the oxidation of D-mannitol to mannose (reaction I) has been found in celeriac and celery [60, 61]. The conversions of D-mannitol to keto-hexose (Fru) by mannitol dehydrogenase (H), or to aldo-hexose (Man) by mannitol 1oxidoreductase (I), are reactions that after fiuther phosphorylation return D-marmitol to metabolism. Mannitol dehydrogenase in cultured celery cells is repressed by sugars [62]. D-mannitol is also found as a terminating group at the reducing end of the seaweed P-glucan laminaran [63-65]. The glycosidic linkage is to a primary hydroxyl of the hexitol. Laminaran from brown algae serves as a food reserve, but may also have a structural role. It is a p(l-3) glucan of relatively low d.p. with some degree of branching through 0-6 as well as some (1-6) intrachain linkages. If D-mannitol is present it terminates about 40-75 per cent of the chains. Mass spectrometry [66] of eight samples of algal species showed the presence of chains both containing and free of D-mannitol. Laminaria, Alaria, Chorda and Sphaerotrichia spp. contained D-mannitol but it was absent from two species of Cystoseira. Glycosides of D-mannitol, P-linked from Glc to the 1 or 3, or the 1 and 6 hydroxyls of D-mannitol, and p-linked from Fru to the 1 or the 1 and 6 positions have been reported for fungi, protozoans, algae and lichens [42\.

1122

3. Mannose Oligomers, Derivatives and Related Structures Man linked to lipid through a phosphodiester bridge is widely distributed since it is a donor molecule in the formation of intermediates in the biosynthesis of glycoprotein and mannoprotein and of some bacterial polysaccharides [67-69]. It is synthesized from GDPMan and its anomeric linkage is p. Manno-oligosaccharides, joined to two GlcNAc residues and then through diphospho to lipid, and which are intermediates in the formation of glycoproteins, are also common [70] {see section 4}. Phosphatidyl-inositol-mannosaccharides with up to six Man residues (1) which are associated with lipoarabinomannan {section 7(a)}, have been extracted from Mycobacterium spp. [71]. In species such as M. tuberculosis Mana 1 -2Mana 1 -2Mana 1 -6Mana 1 -6Man al 6 wyo-inositol-P-lipid 2 al Man they are involved in adhesion to non-phagocytic cells [72]. The major species-specific surface antigen of M. malmoense is a lipid-linked octasaccharide with a Mana l-3Mana 1-^2 group at the non-reducing terminus [73]. The Mana 1-3Mana 1- group has been found in oligolipid of Micrococcus lysodeikticus [74], M.luteus [75] Arthrobacter spp. and Microbacterium lacticum [76]. The Man residue joined to lipid is also acylated on the 6-OH. The oligosaccharide fraction from the freshwater bivalve Corbicula sandai contained the ceramide glycosides (2) and (3) [77]. Manp 1 -4 Glcp 1 -Cer 2

Manp 1 -4 ManP 1 -4 Glc p 1 -Cer 3

1123

More complex structures including an octa-heteroglycosyl ceramide containing the D-mannobiosyl segment (4) were also isolated. PO3 (CH2)2NH2 6 -2Manal-3Manpi->4 2

I al Xyl 4 Hyriopsis schlegelii produces a tetrasaccharyl glycoside GlcNAcpi2Manal-3Manpl-4Glcpl-Cer, as well as ceramide penta-, octa- and nona saccharides containing this unit. Mannosyl glycosphingolipids have also been detected in insects and plants. In wheat, (2) and (3) have been found [78]. Manal-4Glcal-2 m>^o-inositol accumulated in cell-suspension cultures of the rose [79]. These glycolipids are components of membranes. Part of the mannoprotein structure, as found in the yeast Saccharomyces cereviseae [80] contains a series of unbranched oligosaccharide chains of degree of polymerisation (d.p.) 1-5, with a(l-2) and a(l-3) linked Man residues, that are a-glycosidically linked to the OH of serine or threonine in the polypeptide, and which can be released by base catalysed P-elimination. Methods of determination of structures of oligosaccharides have included n.m.r. and mass spectrometry, deacylation with hydrazine, sequential glycosidase hydrolysis, partial enzymic and acidic hydrolysis and acetolysis, polarimetry and methylation analysis. Oligosaccharides (d.p. 4-8) prepared by hydrolysis of galacto-P(l4)glucomannan were inhibitors at 10"^ to 10|LIM of hormonal stimulated elongation of pea stem segments, and also affected the viability of spruce embryos on media supplemented with indole-3-acetic acid: zeatin [81]. Methyl ethers of Man, mono-substituted at the 2, 3, 4 or 6 positions have been reported in the lipopolysaccharides of photosynthetic procaryotes [82^ 83], The reducing end of the mannans from Mycobacterium smegmatis and Streptomyces griseus have a terminal a0-methyl group on C-1 and Man30Me in the polymer chain. The lactic acid ether 4-0-[-(S)-l- carboxyethyl]-D-mannose is a component of extracellular polysaccharides of Mycobacterium spp., and a-amino-3-0-[-

1124

R-l-carboxyethyl]-2-deoxy-D-mannose is found in the peptidoglycan of Micrococcus lysodeikticus, partly replacing the usual gluco-analogue. Acetals of 4,6-linked pyruvic acid are present in some Klebsiella exopolymers. Acetyl esterified Man residues occur in many polysaccharides. The 1- and 6-monophosphates of Man and the 1- of Dmannitol, as well as the nucleoside diphospho sugar GDP-Man are intermediates in the metabolism of D-manno-hexo-sugars. Man 1,6bisphosphate is a co-factor in the reaction catalysed by phosphomannomutase [7,8,10a]. Phosphodiesters occur in fungal mannoprotein and some bacterial polysaccharides. ManNAc is found in a number of lipo- and exo-cellular polysaccharides of bacteria. ManNAcA is another D-manno-hexo structure found in bacterial polysaccharides, and an extracellular polysaccharide of an Arthrobacter sp. contains ManA. The rarer D-isomer of Rha (6dMan) is a component of the Oantigens of some Pseudomonas spp. 4. a-Mannans a-Marmans occur as the carbohydrate portion of cell wall components of fungi, of glycoproteins and in seaweeds as xylomannan sulphate. (a) Fungal Polysaccharides In fungal cell walls, mannoprotein, in combination with glucan, forms the strong cell wall. Exocellular and capsular polysaccharides are also present. Mannoproteins can be enzymes, such as invertase and acid phosphatase in the cell periplasm or hydrolytic enzymes like carboxypeptidase, located in the vacuole, and they are found on the cell surface as sexual agglutination factors of yeasts. As well as contributing to structure, the oligomannosyl side chains are immunologically active and provide adhesion to the host in the initial stages of infection. Released 0-linked oligosaccharides from Candida albicans amannoprotein are inhibitors of lymphoproliferation initiated by the parent mannan. The a-mannan molecules of yeasts [69, 80, 84-87] have been dissected by alkaline P-elimination to release manno-oligosaccharides joined to serine and threonine in the polypeptide [88]; very mild acidic hydrolysis to release chains linked by phosphodiester bonds [89-91] and

1125 to provide core residues [92-95]; hydrazinolysis and enzymic hydrolysis [96, 98, 99] to separate glycan joined to asparagine from the polypeptide; fragmentation by acetolysis, when a(l-6) links undergo preferential scission [88, 91, 92, 100-104]. Oligosaccharides can also be produced by acidic hydrolysis [105]. Released oligomers have been examined by n.m.r. and mass spectrometry and by methylation analysis. Sequential enzymic degradation with endo a(l-6) mannanase, exo a(l-2) mannosidase, a-mannosidases and alkaline phosphatase [88, 98, 99, 106] has also provided information. Enzyme-linked immunosorbent assay (ELISA) and agglutination are rapid methods for detecting structural components [91, 102, 103, 107-109]. Methylation analysis and n.m.r. and mass spectroscopy have also been applied to the polysaccharides. Biosynthetic pathways and genetic analysis are consistent with the results of structural determinations [69, 86, 87, 97]. The a-mannan of the yeast Saccharomyces cereviseae and mutants have been extensively studied [69, 80, 84-87, 88, 97-100, 105,124].

H

I

Mal-(6Mal)d - 6Mal - 3Mpi - 4GlcNAcpl - 4GlcNAcpl - N - Asp 2

1

al

1 al (M)b

(M)c

outer chain

inner core

(Mal)a-O-Ser (Thr) 0-linked oligomers

A generalised model for the glycan section (5) of the wild type consists of 0-Hnked oligomeric chains and an N-linked chain. The 0-linked chains, which comprise about 10% of Man residues, are joined to serine or threonine in the peptide. They can be selectively released by pelimination with mild alkali [88] and consist of unbranched a-man oligomers (a = 1 to 5) linked (1-2) and (1-3), with (1-3) if present, on the terminal fourth and fifth units [97]. The N-linked chain divides into two sections, the inner core, which is joined by an N-glycosyl bond to the amide N of asparagine and the outer chain. The latter consists of an

1126

a(l-6) mannan chain which has side chains of a(l-2) and a(l-3) Man residues. The structure was derived from enzymic and partial depolymerisation [80, 88, 100, 105, 106], A mutant, lacking side chains in the outer chain was depolymerised by endo-a(l-6) mannanase [EC3.2.1.101] and partial acetolysis or mild acidic hydrolysis of the wild type gave a series of manno-oligomers [80, 100, 105] showing that branches (c = 0 to 3) were linked a(l-2) to the a(l-6) chain (d - 40) and that (1-3) links, if present, were terminal. Mild acetolysis [100] produced a low amount of Man, indicating a high level of branching. Man residues in branches may have a phosphodiester bond to an a-Man residue, with a further possible Man joined through 0-3 to this unit.Mutants have also been used to establish the structure of the inner core [98, 99]. Sequential glycosidase hydrolysis, combined with n.m.r. spectroscopy established a structure in which Man residues are linked a(l-2), a(l-3) and a(l-6) in a branched structure [99, 124] (b ~ 6 to 13) with the Man at the potential reducing end linked a(l-4) to chitobiose which is joined to asparagine. The extensive use of yeasts in bread-making and brewing has ensured continuing interest in their structure and metabolism. The role of pathogenic fungi in infection of immunocompromised patients has extended this interest. Candida spp. [89-91, 101-104, 107-109] have many structural similarities to S. cereviseae. Very mild acidic hydrolysis (.OlM HCl, 100^ Ihr) of C glabrata [91] released a P-linked disaccharide, Man pi-2 Man. In C albicans [89] hydrolysis of phosphodiester bonds gave a series of oligomers from a d.p of 1 up to at least 14. Those from 2-7 and higher [90] contained only P(l-2) bonds. These P(l2) oligomannosides are involved in host-pathogen interactions, inducing antibodies and binding to macrophages and stimulating these cells to produce tumour necrosis factor a. When an antigen on the cell wall surface and plasma membrane was solubilized, [108] activity was lost on periodate oxidation but not on proteolysis or heating. Mild acidic hydrolysis gave a hexasaccharide. Activity was determined by a sequence of four unbranched P(l-2) Man residues. The antigen was involved in the attachment of C albicans to mouse spleen marginal macrophages. Both solubilized antigen and tetrasaccharide inhibited binding. The presence of a(l-6)Man groups branching from the side chains was revealed by mild acetolysis [101]. In C. albicans and stellatodea these were attached to O3 linked Man residues [107] and this structure was associated with antigenic activity. In C guilliermondii [102] another type of P(l-4)

1127

linkage was detected in which side chains connected to the main chain were extended by two P(l-2) Man residues and this structure was also associated with antigenic activity. Similar groups were detected in Saccharomyces kluyveri. A third antigenic structure in Candida spp. [103] was associated with extension by a single Man P(l-2) residue. The extra-cellular phosphomannan produced by Pichia holstii [93], grown in excess phosphate, appears to be a chain of a repeating hexamannosyl phosphate units (6) of which 10% of the polymer is composed. P-6 Man al-3 Man al —2 Man al

I

Man al

I

2 2 [6Manal — 6Manal]n 6 The majority of chains attached to this as phosphodiesters are composed of at least ten repeating unbranched pentamannosyl phosphate units. Chains with d.p. 2-6 also occur. The Man linkages are a(l-2) and a(l-3) and the phospho group is on 0-6. Several Microsporum and Trichophyton spp. [110] contain a-mannans consisting of an a(l-6) main chain with branches of single a(l-2) Man groups on some main chain residues. Phosphomannoproteins of Kluyveromyces lactis with different flocculation behaviour showed no differences in chemical composition and molecular size [111] but differed in their recognition of a lectin extracted from the flocculent strain. The oligosaccharides derived by acetolysis of this strain contained equivalent levels of Man a\-2Man and Man a 1-3A/aw and of Man aX-lMonal'lMan and Manal-3Manal2Man, whereas in the non-flocculent these consisted of a(l-2) linked biose and triose. The terminal a(l-3) linked Man residue appears to be recognized by the secreted lectin, contributing to flocculation. Many other fungal mannans are heteropolymers based on an amannan chain [84, 55]. a-L-Rhamno-a-mannans with L-Rha al-->3, LRha al-2-L-Rha al->3 and L-Rhal-4GlcApi-2-L-Rhaal^3 side chains have been described. a-Gluco-a-mannans with single Glc residues attached to the mannan chain have been isolated from Ceratocystis spp. The capsular polysaccharides of the serotype of an encapsulated pathogenic yeast, Cryptococcus neoformans, have a main chain of a(l-3)

1128

Man residues, partly acetylated, on 0-6, to which are attached single Xyipi-^ groups to 0-2 and 0-4 of Man units in this chain, as well as GlcApi--> to 0-2 [112, 113]. A variety of galacto-a-mannans have been reported [84,85] with Gal/pi, Gal/ a l , Galal and Galpl units. Cladosporium spp. produce a phosphogalacto-a-mannan and Penicillium charlesii an exocellular phospho-galacto-a-mannan containing consecutively linked p(l-5) Gal/sections. A polysaccharide from the cell wall of Neurospora crassa consists of an a(l-6) mannan chain to which Gal/pi-residues are linked to 0-2 [114]. A polymer from the cell walls of two Trichoderma spp. has a more complex structure, with side chains containing P(l-5) and (1-6) Gal/residues joined to the 0-2 of Man. GlcAla- and Glcla- groups are also present [95]. The cell wall galactoa-mannan of two Trichophyton spp. [94] has an a-mannan chain made up of a repeating unit of four a(l-2) linked units joined a(l-6) and Gal/pi^ groups linked to 0-3 of Man. The extracellular galactomannan of Aspergillus fumigatus has a similar mixed linkage main chain, with side chains of four to five Gal/(l-5) units linked to 0-6 and 0-3 of Man [115]. The (Gal/)n chains are immunogenic. In Schizosaccharomyces pombe Gal deficient cells did not flocculate non-sexually [116]. After removal of the lipophilic phytotoxic activity of a fungal pathogen (Phomopsis foeniculi) of fennel [117] one of the two exocellular polysaccharides in the aqueous phase was shown to be a mannan composed of a(l-6) chains of Man with Manal-^, Manal-2Manal~> and Manal-3Manal-2Mana ~> side chains attached to 0-2 of Man in the main chain. It, and an accompanying fijranogalactan caused stem wilting and browning of fennel, and necrosis of tomato and tobacco leaves. They appeared to act through interfering with water movement by forming mechanical plugs, and this may have been related to molecular size and viscosity, although host specificity was also considered to be a possibility. Differences in the chemical structure of mannans and hetero-mannans of yeasts and other fungi can be used as chemotaxonomic markers in their classification [95]. A comparison [118] of n.m.r. spectra of heteromannans of the fungal part of lichens from 13 genera has allowed a classification of the lichens into five groups based on common signals. The galactomannans extracted from lichens have generally been found to be composed of a(l-6) mannan chains substituted variously with Gaipi and Gal/1 to 0-4 and Manal and Galal to 0-2 of Man residues in these chains [119-121].

1129

Yeast offers a recombinant protein-expression system for the production of biopharmaceuticals. However, glycosylation may occur and produce expressed protein antigenic to humans. Also, its Man residues may react with the mannose receptor and clear the protein too rapidly. A recombinant tryptase inhibitor [122], derived from leeches was expressed in strains of Saccharomyces cereviseae secreting glycosylated and non-glycosylated forms. The glycan chains were located opposite the protein binding loop involved in the reaction site and exclusively a-0glycosylated to three serines by a single Man and 0-glycosylated to threonine by manno-oligomers of d.p. between 1 and 3 or 1 and 13 according to strain. The solution conformation of this glycosylated recombinant tryptase was derived from n.m.r. spectroscopy. The conformation around unsubstituted hydroxy amino acids indicated that Hbonding to -OH or poor access blocked the transfer of Man. Comparison of glycosylated and non-glycosylated conformations showed only slight differences: substitution at only one serine gave small changes in the overall folding. The a(l-6) main chain found in many fungal mannans with pyranose rings joined by two atoms not involved in ring structures, allows flexibility, with a large zone of low conformational energy in the contour diagram [123], This is in contrast to the a(l-2) link, which involves a single oxygen atom joined via two axial bonds between pyranose rings, offering a small zone of low conformational energy in the calculated contour diagram, indicating a more restricted conformation. Both the a and P(l-2)Man linkages give oligomeric and polymeric conformations distinct from any available to glucans and galactans. Fungal mannoproteins share structural affinities with some glycoproteins of higher eukaryotes - the N-glycosyl linkage of GlcNAc to the amide N of asparagine, the -4GlcNAcpl-4 GlcNAc- segment, the p linkage between this and the first Man residue and then a core region of a-linked Man residues. The biosynthesis of both has similarities. A significant structural difference is the long a(l-6) mannan sequence found in fungal mannoproteins. (b)

Glycoproteins

The widespread occurrence of glycoproteins [124-131] across species, in different types of proteins and at different cellular locations, combined with the expenditure of metabolic energy required for glycosylation,

1130

implies a significant and general role. They are involved in a number of cell processes [87, 132] - interactions between cells, responses of cells to external factors such as hormones and antigens; cell recognition, differentiation and growth; protein transport and sorting; protection from proteolysis, increased thermal stability and life span of protein and the facilitation of protein chain folding. The glycan may interact with the polypeptide chain. Blood group type is specified by the carbohydrate part of certain glycoconjugates and the glycan chains of the related mannoproteins of pathogenic fungi are antigenic. The carbohydratedeficient glycoprotein syndrome type-IA is an autosomal disorder [10a] leading to psychomotor retardation, cerebellar dysfunction, peripheral neuropathy, liver insufficiency and abnormal adipose tissue distribution. It is characterised by a decrease in the number of glycan chains on glycoproteins and is due to a deficiency of the enzyme phosphomannomutase (reaction D). Man 1-P is the precursor of GDPMan, the substrate for the addition of Man residues to glycan chains. Inhibition of glycan chain processing by indolizidines causes toxicity in animals [127], A major group of glycoproteins contain Man and are joined to asparagine in the polypeptide. Their glycan chains range from those in which the constituent sugars are Man and GlcNAc (oligomannose) to those that contain high proportions of other sugars - Gal, L-Fuc, sialic acid and Xyl - which are called complex. Hybrid types have GlcNAc (Gal) and Man in the outer chains. The section linked to polypeptide usually contains the pentasaccharyl sequence (7) linked by an Nglycosylamine bond to the amide N of asparagine. H I Manal - 6Manpi - 4GlcNAcpl - 4GlcNAcpi -N - Asp 3 I al Man In the oligomannose type further substitution by a-Man residues, which can be linked (1-2), (1-3) and (1-6), occurs. The a(l-3) Man in the pentasaccharide core is commonly further linked a(l-2) and the a(l-6) in the core by a(l-3) and a(l-6). In complex types further substitution (by GlcNAc and Gal in sequence and then sialic acid) occurs. L-Fuc can be a

1131

substituent on the GlcNAc residue joined to asparagine. Sialic acid is not found in plants; Xyl may be joined to the p-linked Man residue. Particular glycan structures are found in different glycoproteins and a single glycoprotein usually contains a number of different chains. The same asparagine in the amino acid sequence in different molecules of the same protein can have different glycan chains. These are called glycoforms. In biosynthesis [68,70,124,126,127,129], initially a lipid-diphosphooligo-saccharide, GIC3 Man9 GlcNAc2-P-P-dolichol is made (8- ON represents GlcNAc and G, Glc. Bold M correspond to Man residues in the pentasaccharide 7). Mal-2M al

I 3

Mai - 2Mal ~ 6Mal - 6Mpi - 4 GNpl - 4 GNpi-P-P-Dol 3

I al Gal-2 Gal-3G al-3M al-2Mal-2M 8 Its formation involves reaction of UDPGlcNAc and dolichol-P to form GlcNAc-P-P-Dol. The second GlcNAc and Manpi are transferred from UDPGlcNAc and GDPMan. Then the first four aMan residues are transferred from GDPMan and remaining Man and Glc residues from sugar-P-Dol. The oligosaccharide part of 8 is then transferred to asparagine and processed by removal of the Glc and some Man residues. In the complex type, six Man residues are removed and up to four Plinked GlcNAc residues added, followed by Gal (p), NeuAc (a) and LFuc(a). GlcNAc is joined 2 or 4 to the a(l-3) linked Man in 7 and 2 or 6 to the a(l-6) linked Man. A bisecting GlcNAc can be connected (pi-4) to the P-linked Man. The d.p. is mainly in the range 6-20 and four oligosaccharyl antennae (tetra-antennary) can be formed. Further substitution with single side chains is possible. Oligomannase type chains are processed by the removal of the three Glc and some Man residues. The d.p. is generally from 5 to about 12. In yeast mannoprotein, the

1132 oligosaccharide Glc3Man9GlcNAc2 is trimmed to Man8GlcNAc2 and then elongated by mannosyl transferases to form the long a-mannan outer section of the main chain and its branches. Xyl appears on plant glycoproteins, where it is P-linked to 0-2 of the P-linked Man residue. A variety of modes of substitution is found, Table (1). All except P-linked Table 1. Linkages of Monosaccharides in Mannose Type Glycans Joined to Asparagine in Glycoproteins Sugar Man (M)

Linkages Mal^2 Mal->3 M a l ^ 6 2Mal^2 3Mal^2 2Mal->3 2,4Mal-^3 2,6 lVlal->3 2,6Mal->6 3,6Mal->6 2,4,6Mal-^6

2Mal^6

6Mpi-^4 6Mpi->6 2,6Mpi^4 3,6Mpi->4 236Mpi^4 3A6Mpi-^4 GNAc (GN)

GNal^6 GNpi->2 GNpi-->4 GNP1^6 3GNpi->4 4GNpi->N 4GNpi->2 4GNpi->3 4GNpi->4 4GNpi->6 3,4GNpi^N 4,6 GNpl->N 3,4GNpi->2 3,4 GNP 1-^4 4,6GN^2 4,6GNpi->4

"Cai

Gal/xl->2 Galal-^3 Gaipi-^3 Gaipi^4 3Gaipi->4 6Gaipi->4

L-Fuc

L-Fucal->3

Xyi

Xyipi^2

Neu Ac

Neu Aca2->3

GalNAc

GaiNAcpl-»4

L-Fucal->6

Neu Aca2->6

Man can be non-reducing end groups: Gal can be both pyranose and fiiranose [133]. Determination of the structure (monosaccharide sequence and linkage) of glycoproteins involves initial separation of glycan from protein [96]. Hydrazinolysis provides a chemical method, but enzymic hydrolases are mostly used. Endo-P-N-acetyl glucosaminidase [EC 3.2.1.96] hydrolyses between the two GlcNAc residues linked to asparagine, releasing the glycan section less one GlcNAc. Ease of reaction depends on the source of enzyme, glycan type and number of antennae. Glycopeptide N-glycosidase [EC 3.2.2.18] and glycopeptidase [EC 3.5.1.52] hydrolyse between GlcNAc and asparagine, releasing the carbohydrate chain with both GlcNAc residues attached. Fractionation of the released glycan chains can be accomplished by size exclusion, lectin affinity, ion exchange and reverse phase chromatography and by gel and

1133

capillary electrophoresis. Structures of the individual oligosaccharides can be derived from methylation analysis [134], and n.m.r. and mass spectrometry [96, 130, 131, 135]. Sequential glycoside hydrolysis has also been used [96, 106]. The enzymes are of two types, glycosidases that hydrolyse one anomeric linkage to all positions on the next sugar, such as a-mannosidase [EC3.2.1.24] and those like exo 1,2-1,3-a-mannosidase [EC3.2.1.77] that are limited in hydrolysis by the linkage of the next sugar. Estimation of released monosaccharide provides a sequence from the reducing end. The unravelling of the structural requirements controlling sugar addition by transferases in biosynthesis [68, 70, 87, 106, 124, 126, 127, 129, 136\ has also been important in the understanding of glycoprotein structure. Oligosaccharide structures have been simply and rapidly established from the chromatographic mobilities of fluorescently labelled derivatives (with 2-aminopyridine) on size exclusion and reverse phase columns, by plotting their elution rates in two dimensions and comparing the co-ordinates with standard oligomers [137]. The structure can be confirmed by then hydrolysing the oligosaccharide with an appropriate glycosidase and re-chromatographing the residual oligosaccharide. An examination of ovalbumen by this procedure detected two new oligosaccharides. When applied to an endopolygalacturonase from the culture filtrate of the fungus Stereum purpureum (apple silverleaf disease) [138], one major oligosaccharide, a heptasaccharide (9), was identified. Man al-6Man al-6Man pi-4GlcNAcpl-4 GlcNAc 3 3

I

I

al Man 9

al Man

Two other recent examples of structure determination are studies of rat liver cathepsin L [139] and horseradish peroxidase [140]. The oligosaccharides of cathepsin L were released with glycopeptidase, fractionated and purified by size exclusion and reversed phase chromatography. N.m.r. and mass spectrometry identified a series of high mannose oligomers of d.p. 4-8 which had been produced by sequential endogenous a-mannosidase hydrolysis. The minor glycan chains of horseradish peroxidase, released by glycopeptidase, were examined by

1134 mass spectrometry and methylation analysis. These contained Fuc and Xyl and there was heterogeneity in structure at each substitution site. Matching of the linkage point of particular glycan chains with specific asparagines in the polypeptide sequence initially requires a determination of the amino acid sequence of the polypeptide. The glycoprotein is then fragmented with a protease and the sugar-peptide fragments fractionated. Provided there is sufficient length of peptide, the position of an asparagine in the original polypeptide chain can be decided by sequencing the peptide. A secreted recombinant human a-L-iduronase was found [141] to contain six N-glycosyl sites. There were complex chains at two sites; bisphosphorylated oligomannose (P2.Man7.GlcNAc2) at two; oligomannose (Man9.GlcNAc2 - with some monoglucosylated) at one; and a mixture of complex and oligomannose at one. Mass spectrometry, combined with glycosidase hydrolysis with neuraminidase, a-Lfiicosidose and p-galactosidase [142] showed that glycosylation at two asparagine sites on mouse scrapie protein was heterogeneous, comprising some sixty glycoforms of complex glycan chains. Soluble lysosomal enzymes are transported by Man 6-P receptors to the lysosome from the endoplasmic reticulum, where they are synthesised, through the Golgi complex [143]. The Man receptor on macrophages has been implicated in endocytosis of glycoproteins, phagocytosis of organisms with surface a-mannans and in antigen uptake [144]. With the Man-6-P receptors, oligosaccharides with P-6-Man linked a(l-2) have a higher affinity than those linked a(l-3) and a(l-6). Man 6-P is a poor ligand, and Fru 1-P but not Glc 6-P are competitive inhibitors of binding. Affinity is increased in multivalent ligands. These can be antennary chains on the same glycan or different chains on the same glycoprotein. The structure of the Man-6-P receptors is highly conserved from reptiles to mammals [145]. Mice deficient in either of the two Man 6-P receptors (300 kDa and 46 kDa) had much higher levels of glycoproteins with 6-P on terminal Man groups and there were specific differences in the patterns of these, indicating each receptor has distinct functions for targeting lysosomal enzymes [146]. The Man receptor is involved in host defence against infection by organisms such as Mycobacteria, Pseudqmqnas and Chlamydia and Candida spp. [147]. Ingestion of Candida albicans reduces the expression of Man receptor on rat macrophages [148]. Binding affinity varies among glycoproteins [144]. The toxin ricin consists of two forms. When separated, the form with two oligomannose chains was taken up by

1135

macrophages more rapidly than the other with one chain [149]. The protein moiety of the glycoprotein may also affect binding, possibly by its influence on the conformation of the oligosaccharide [144]. The macrophage Man receptor is a lectin and the stereochemistry of binding of Man and other sugars has been investigated by n.m.r. spectroscopy and site-directed mutagenesis [150]. Insects lack a non-self recognition system.and use innate immunity for defence against micro-organisms, which is probably mediated by lectins [151]. Lectins [152] interact with sugar residues by hydrophobic forces; water molecules and cations are also involved [153, 154]. They can be proteins or glycoproteins [155] and are widely distributed. There is specificity for binding to one anomer of a sugar or several related sugars. In solution, with a polymer containing multiple ligands (multivalent) they give aggregates, the formation of which can be inhibited by ligands of low molecular size. They adhere to cell surfaces and endogenous lectins provide adhesion sites. One group, which includes one of the most familiar, concanavalin A, binds both aGlc and aMan residues. The crystal structure of the complex between concanavalin A and a compound with high affinity, Manal-6(Manal3)Manal-0Me, which is similar to the trimannosyl core of glycoproteins - the three Man residues of (7) - shows that all three Man residues are bound [156] and that the tetrameric structure of the protein is somewhat flexible. However, lectins have also been described that are specific for only aMan, recognising internal [157] and terminal residues [158]. A lectin fi'om orchid leaves {Listera ovata) [157] interacted with yeast amannans and galacto-a-mannans but not a-glucans. It reacted with unbranched a(l-3) mannan and the degree of binding was inhibited by alinked manno-oligosaccharides in the order a(l-3)>a(l-6)>a(l2)>Manal-OMe. Periodate-oxidised and then reduced a(l-3) pentasaccharide also inhibited, indicating that the lectin recognised internal a(l-3) linkages. The lectin from bulbs of Crocus vernus selectively interacted with terminal Manal-3Man- groups [158]. Garlic {Allium sativum) lectins [159] bound strongly to oligomannose chains, shoving a preference; Man9 (8 with asparagine replacing -P-P-Dol) > >Man5 with a(l-6) and a(l-3) links > Manal-6Manal-3A/a«>Manal3Man, Two GlcNAc residues at the reducing end of triose or pentaose enhanced activity. Glycoproteins with compound chains, like fetuin, were ligands: removal of terminal sialic acid residues increased binding. Mycobacteria contain a lectin that adheres to macrophages [160]. The fimbriae of more highly virulent strains oi Escherichia coli, which act as

1136

adhesive organelles, expressed a lectin [161] that interacted with single Man residues, whereas those of less virulent strains expressed a lectin specific for mannotriose. Arcelin-1 from wild varieties of the bean, Phaseolus vulgaris [162], a lectin that has insecticidal properties, does not interact with monosaccharides, but binds to complex type glycan chains. An insect-derived mutant of the toxin ricin, modified at three sites to give a 1,000-fold reduced galactoside avidity [163], when supplied to mouse macrophages was still lethal, showing that intracellular binding to galactose is not required for toxicity. The function of a biological macromolecule in solution is dependent on its conformation. The determination of three-dimensional structures in solution, in conjunction with the dynamic behaviour, contributes to understanding of the associations between the polypeptide and its substituent glycan chains, and between the glycoprotein and interacting molecules, as well as the significance of glycoforms. Restrictions in flexibility imposed by the hydrodynamic volume of chains and the possibility of interaction with the polypeptide are factors in the conformation. Molecular modelling [725, 750, 757], calculations of minimum energy conformations [164-166], n.m.r. spectroscopy [757,164, 167-169], molecular dynamics simulations [164, 169-171], molecular orbital calculations [164, 169] and fluorescence energy transfer measurements [172] have been used to study these aspects of glycoproteins. Molecular modelling [125] of a dianntenary complex structure suggested a compact pentasaccharide core, with the diequatorially linked P(l-4)-Man-GlcNAc-GlcNAc- segment as a flat ribbon, and the two antennae as helical chains with two possible orientations. Preferred positions have been proposed [165, 166] for the sugar residues in oligomannose chains and a bisected pentasaccharide. Time resolved fluorescence energy transfer measurements [172] on a dansyl-labelled triantennary complex glycan indicated two populations of conformers when two of the antennae were labelled, but only one when the third was labelled. N.m.r. spectroscopy, molecular orbital calculations and molecular dynamics simulations [169] of an oligomannose type heptasaccharide and an undecasaccharide indicated that the Manal3ManP linkage was highly restrained and the Manal-6Manp and Manal6Mana linkages disordered. The Man al-6ManP linkage was more restrained in the undecasaccharide. Comparison of a hybrid heptasaccharide with the bisected octasaccharide structure showed that the presence of the bisecting GlcNAc restricted the dynamic behaviour of the

1137

Manal-6ManP and Manal-3ManP bonds [170]. Molecular dynamics simulation of an oligomannose chain (Man5.GlcNAc2) attached to immunoglobulin M, and a complex chain in immunoglobulin G [171] indicated the availability of various conformational minima and differences in mobilities of the various glycosidic linkages. From these multiple minima and the presence of the same glycans in different proteins, it was suggested that the same chain sequence might have different dominant conformations in different protein environments. The calculations were consistent with results from n.m.r. spectroscopy and fluorescence energy measurements. Examination of the adhesion domain of a cell surface glycoprotein (CD2) of human T cell lymphocytes [173], which contains a single oligomannose type glycan, indicated that this chain increases rigidity and stability by counterbalancing a cluster of positive changes about lysine 61. Removal of this N-linked chain caused loss of binding activity, with unfolding of the protein. Reduction to a single GlcNAc unit reduced stability. Protein with lysine 61 replaced with glutamic acid required no glycan chain to be stable. It was found that through interaction with the polypeptide the conformational mobility of parts of the attached glycan can be restricted relative to the free oligosaccharide. Bovine pancreatic ribonuclease B has five glycoforms in which a series of high mannose oligomers (Mans GlcNAc2 - Man9 GlcNAc2) are attached to the single Nglycosylation site [174]. Exo-glycosidase digestion gave individual glycoforms. Varying glycosylation affected dynamic stability and the glycoforms showed increased resistance to proteolysis compared with the unglycosylated enzyme. There was a fourfold variation in enzymic activity among glycoforms and molecular modelling suggested that steric factors were responsible. Glycan chains-oligosaccharides MansGlcNAc and Man3(Xyl)GlcNAc(L-Fuc)G/cA^^c - at concentrations of 0.5 and 5|ag/mL (2|iL/g fresh weight) promoted fruit ripening when infiltrated into tomatoes [175]. (c) Seaweed a-Mannans Sulphated and xylo-a(l-3) mannans have been extracted from seaweeds. A water-soluble (1-3) mannan has been reported from Urospora penicilliformis [176], A sulphated a-mannan was isolated in more than 30% yield (dry weight) from Nemalion vermiculare by extraction with hot water. Cetavlon gave a precipitate containing 15.5% sulphate.

1138 Methylation analysis of the desulphated polymer showed an unbranched (1-3) linked mannan chain and the positive optical rotation indicated aglycosidic links [64, 65]. Water extraction of the red seaweed Nothogenia fastigiata, gave a sulphated xylomannan [177]. Methylation analysis and n.m.r. spectroscopy showed a structure with an a(l-3) mannan main chain, substituted at 0-2 with single Xylpl- groups. The Man residues were also substituted at 0-2 or 0-6 with sulphate, with some substituted at both. The polysaccharide inhibited viral replication and this activity was considered to depend on its ability to block virus receptors necessary for adsorption to target cells. 5. P-Mannans A number of plant [106, 178-183] and algal [64, 176] polysaccharides have structures based on a P(l-4)mannan main chain - P(l-4)maiman, a(l-6)galacto-P(l-4)-D-mannan and the related P(l-4)-D-glucomannan and galacto-P(l-4)glucomannan. They serve as cell wall structural material and as carbohydrate reserve polymers. An unbranched p(l-2) mannan of low molecular weight has been reported from the insect flagellate Herpetomonas samuelpessoai [184]. (a) J3(l'4)-D'Mannan P-Mannan is water-insoluble and solubilizing it requires very strong alkali, cuprammonium or derivatization. X-ray and electron diffraction show that mannan has a ribbon-like conformation, similar to cellulose, with the 2-OH axial instead of equatorial. In seeds (palm, coffee, caraway) a few (<5%) of a-Gal groups are attached to the hydroxymethyl and in seaweeds a small percentage (<5) of Glc residues may occur in the chain. A pure p(l-4) mannan has been extracted from cell walls of the siphonous green alga Codium latum as the methylol mannan with paraformaldehyde: dimethylsulphoxide. It was purified by size exclusion chromatography in dimethyl sulphoxide and recovered by the addition of water or methanol and could be re-dissolved in hot dimethyl sulphoxide. Infra-red and n.m.r. spectroscopy and periodate oxidation confirmed the unbranched P(l-4) structure [185].

1139

(b) 6-a'GalactO'P(l-4) Mannan When seeds containing gaiactomannan germinate, metabolic substrate is derived initially from oligosaccharides and then from gaiactomannan, which supports early seedling growth. It also provides a humid environment. Legume seeds such as guar (Cyamopsis tetragonolobus) carob (or locust bean, Ceratonia siliqua) fenugreek {Trigonella foenumgraecum) honey locust {Gleditsia triacanthos) Senna and Leucaena spp. contain galactomannans in their endosperm. They find commercial use [186\ as thickeners, stabilizers, sizing and finishing agents, in mineral processing and some are used in food preparations in conjunction with other polysaccharides, like xanthan, agarose and carrageenan, with which they interact to form gels. Galactomannans consist of P(l-4) mannan chains in which single Gal groups are attached to the 6-OH of Man. The degree of substitution varies among plant species [187]. The pattern of substitution also differs among species. This, like most plant polysaccharides, in contrast to bacterial polysaccharides where the mechanism of biosynthesis produces regularly repeating units, is not regular. Nor is it usually statistically random: it can be described as nonregular. More recent structural analyses have involved determining the distribution of substituents. Carob gaiactomannan is very suitable for the formation of gels with other polysaccharides like xanthan and this is due to the level and distribution of Gal substituents. There is interest in enzymically modifying the structure of more readily available polymers, such as fi-om guar, to produce a gaiactomannan resembling carob. Another approach is to genetically modify the biosynthesis. The galactomannans have been depolymerized with highly purified endo-Pmannanases [188], whose action pattern is fully described [189], and the oligosaccharide products identified and estimated. This has been linked to computer studies that compare these experimental products with those of simulated hydrolyses of modelled biosynthetic polymers in which the simulated action pattern followed that of enzymic hydrolysis. The biosynthesis [182,183] of gaiactomannan proceeds by the sequence D *-^Manl-P J Man 1-P + GTP -^ GDPMan + P-P Man6-P

1140

mGDPMan + nUDPGal

K,L -^ (Gal al-6)n [4 Manpl-]m)

catalysed by phosphomannomutase (reaction D) [7-10a] present in Cassia corymbosa seeds and konjac tubers; GTP:a-mannose-l-phosphate guanylyl transferase [EC 2.7.7.13] (reaction J), found in Gleditsia macracantha seeds; GDP-mannose:galactomannan 1,4-P-mannosyl transferase (reaction K), and galactomannan 1,6-a-galactosyl transferase (reaction L), for which activities have been detected in seeds of fenugreek, guar and senna [190-192]. The favoured conformation of the P(l-4) mannan chain of galactomannan in solution [106, 182] is probably an average related to that in the solid state [193], giving segments of extended ribbon-like structure in which alternate Gal groups (or the unsubstituted CH2OH groups) lie on opposite edges of the ribbon. Due to the low solubility of P(l-4) mannan, substitution of pre-formed mannan is unlikely. A similar type of polymer a(l-6)-D-xylo-P(l-4)-D-glucan was shown to be formed by concurrent incorporation of Xyl and Glc [194] and the same process occurs with galactomannan [190]. It has been proposed [106, 182, 195] that steric effects of the extending galactomannan chain on the membranebound mannosyl and galactosyl transferases affect the substitution pattern. In (10), where M represents the non-reducing terminal Man available for substitution, CH2OH or Gal

I M — M— M— [M]m (Gal)n -> I I reducing end CHjOHorGal 10 the groups that influence substitution are those on the neighbouring (nearest) and next-nearest Man units, which may carry either -CH2OH or -CHiOGal. The substituent on the next-nearest neighbour offers more steric hindrance than that on the neighbouring Man, since the former lies on the same edge of the ribbon-like conformation as the position for a new substituent. The ease of substitution is then dependent on steric interactions between the nucleoside diphospho sugar:transferase enzyme:membrane complex and the elongating chain. In seeds of the

1141

Faboideae (lucerne, fenugreek),where Gal substitution is nearly complete no steric inhibition would be encountered. In seeds of the Caesalpinioideae (carob, honey locust, Leucaena, Senna) the degree of reduction of substitution and the distribution of substituents would depend on the level of steric interaction. The action pattern oi Aspergillus niger P-mannanase [EC 3.2.1.78] was established from the structures of oligosaccharides released by hydrolysis of the hot-water-soluble fraction of carob galactomannan [189]. Mannobiose and triose and ten galactomannosaccharides (d.p.3, 4, 7-9) were separated and identified. They showed that the enzyme binds to five contiguous Man residues on one edge of the ribbon-like conformation of the polysaccharide and hydrolyses between the second and third of these xmits. Binding and scission can only occur if Man units lack Gal substitution on that edge: hydrolysis occurs with (11) and (12) but not (13). Gal Gal Gal Gal Gal

I I I

->

II

M->M->M-^M->M ->

-^M-^M-^ M->M->M->M-^

L_

I

I Gal

11 Gal

12 Gal

I

I

->M ~ - > M - > M - ^ M - > M - > M - >

I

Gal 13 Simulated structures of galactomannan were computed with a chainextending sequence [196]. With substitution affected by the presence or absence of Gal groups on the nearest and next-nearest Man, four situations exist. Both or neither can be substituted, or either can be substituted and the other unsubstituted. Probabilities for substitution were assigned for each of these situations and a polymer generated. This structure was then subjected to simulated hydrolysis following the known action pattern of a p-mannanase and the degree of hydrolysis, amounts and structures of released oligosaccharides and galactose content of the generated polymer found. Probabilities were then adjusted until these parameters matched the experimental values. The distributions of

1142

substituents in carob galactomannan and its hot and cold-water-soluble fractions were quite different from regular, random or block patterns. They could be described as non-regular. There was a high proportion of two contiguous substituents (couplets) and single isolated Gal groups but few sequences of alternate substituted-unsubstituted units. There was a low amount of triplets, an absence of substituted blocks and a higher proportion of unsubstituted blocks of intermediate size than in a random structure. The probabilities of substitution were: nearest-neighbour only substituted > no substitutent > next-nearest-neighbour substituted > both substituted, indicating that steric hindrance was not the only factor. If the nearest-neighbour has just been substituted the enzymes and substrate are in proximity and the probability that the enzymes will continue to react with the same chain is elevated. The degree of substitution of guar galactomannan with two thirds of Man residues substituted makes the application of this procedure difficult, but it appeared to also have a nonregular structure. The non-regular but not random model also fitted the structures of the galactomannans of Ceratonia siliqua, honey locust and Caesalpinia vesicaria, but not of Sophora japonica and Caesalpinia pulcherima [197] whose structures did not show a significant difference from the random model. Those of honey locust and C vesicaria, although both non-random, differed in their distribution of substituents. Guar and Leucaena leucocephala galactomannans have similar Gal contents but the degree of hydrolysis of guar by p-mannanase is half that of L leucocephala. The hydrolysates of the latter indicated the presence of frequent regions of alternating substituted and unsubstituted residues, whereas the former appears to contain a high proportion of doublets. After the Gal content of fenugreek galactomannan had been reduced by hydrolysis with a-galactosidase, the oligosaccharides then produced by Pmannanase contained a high proportion of the branched trisaccharide Man-(Gal)Ma«, indicating that a-galactosidase removed substituents sequentially from one edge of the mannan ribbon-like chain and not in a random manner. Cell particulate preparations containing both mannosyl and galactosyl transferase activities from developing seeds of fenugreek and guar incorporated label from GDP [U-^ C] Man, giving a polymer with the properties of a p(l-4)mannan [190]. No transfer was observed with UDP [U-*^C] Gal, but with labelled UDPGal and GDPMan the product had the properties of a-galacto-P-mannan. The level of substitution by Gal could be varied by supplying saturating amounts of UDPGal and changing the

1143

concentration of GDPMan. Incorporation of Gal and Man was concurrent. Pre-formed Man sequences did not subsequently substitute either with UDPGal alone or mixed with GDPMan. The patterns of change in levels of Gal substitution of developing of seeds of Gleditsia triacanthos and Senna occidentalis differed from that of guar and fenugreek. In the first and second the Gal content decreased at later stages of development, wticreas in guar and fenugreek it remained constant [191, 198]. The first two plants belong to the Caesalpiniodeae, which contain galactomannans with lower levels of Gal than those of the Faboideae to which fenugreek and guar belong. The Caesalpiniodeae are more primitive than the Faboideae. The polysaccharide occurs as a thickening of the secondary wall and those in the Caesalpinioideae have more of the characteristics of structural polymers - low levels of substitution and solubility. The higher levels of Gal in the Faboideae may reflect an evolutionary change to a more soluble material with better properties for a storage polysaccharide [187]. From analysis of the oligomers released by P-mannanase from the galactomannans extracted fi-om G, triacanthos at different stages of development [198] it was considered that the observed decrease in Gal content in late samples was probably not due to hydrolysis by the small amount of a-galactosidase in seeds near maturity. This enzyme sequentially removes Gal groups that lie on one edge of the P-mannan ribbon [196], leading to an absence of doublets, whereas the oligosaccharides produced by depolymerization of the galactomannan of G, triacanthos revealed the polymer had a high level of doublets but not of single substituents [198]. A limiting supply of UDPGal or changes in levels of transferase near maturity provide possible causes of the observed decrease in Gal content. Another study [191] concluded the decrease was due to a-galactosidase action. In S. occidentalis, guar and fenugreek, the ratios of the two biosynthetic transferases were correlated with the Gal content. In the first (with a lower level of Gal) the ratio of mannosyl to galactosyl transferase in extracts was higher than in guar or fenugreek and increased during seed development. In guar and fenugreek the levels of activity of the two transferases were similar and remained constant with development. Isolated, membrane-bound preparations were presented [192] with a mixture of saturating levels of UDPGal and decreasing levels of GDPMan, when a series of galactomannans containing from 14-35% Gal were produced. The Gal contents of fenugreek and guar galactomannans are 48 and 38%. The high level in fenugreek galactomannan does not

1144

allow any structural analysis with p-mannanase and limited information can be obtained from guar. The synthetic polysaccharides, which undergo significant hydrolysis, allowed investigation of their biosynthesis. The polymers were hydrolysed with P-mannanase and the products matched to models derived from simulations of synthesis and hydrolysis. In S. occidentalis with a low GalrMan ratio, a low probability of substitution applied when there was a Gal on the next-nearest-neighbour or on both nearest and next-nearest-neighbours. In guar the probabilities were similar, except for a somewhat higher value if the nearest-neighbour carried a Gal residue; this would produce a preference for doublets. In fenugreek, when the next-nearest-neighbour or both Man residues were substituted,the probabilities were significantly higher. When the most recently added Man residue has just been substituted, bringing the enzyme-membrane complex and terminus of the polymeric substrate into proximity, and steric hindrance is not encountered, further substitution of the next Man residue added would be favoured over dissociation of the enzyme and reaction with another chain. The probabilities for biosynthesis of S. occidentalis galactomannan from mature seeds were similar to those for enzymically synthesized chains. If polymer in the seed undergoes a-galactosidase hydrolysis near maturity this would only be expected if hydrolysis were random. However, the oligosaccharides obtained from p-mannanase hydrolysis of fenugreek galactomannan, in which the Gal content had been decreased by a-galactosidase [196], indicated that hydrolysis was not random but proceeded sequentially along one edge of the ribbon-like mannan chain. (c) P(1'4) Glucomannan and 6a-Galacto-P(l-4) Glucomannan p(l-4) Glucomannan is a secondary wall component (hemi-cellulose) of wood [775, 199-201], and also occurs in tubers, bulbs, seeds, roots and leaves of monocotyledonous plants, such as Liliaceae, Amaryllidaceae and Orchidaceae [181-183, 202]. It may be partly acetylated and this increases solubility in water. Polymers with a small amount (<5%) of a(l-6) linked Gal residues have been prepared from seeds. Glucomannan serves as a structural element and as food reserve. The Glc content varies: it is 23% in the polysaccharide from Orchis morio, 43% from bluebell seeds and 49% from Dracaeno draco seeds. P(l-4)Mannanase hydrolyses them to oligosaccharides. Reaction with Aspergillus niger p-mannanase involves binding to five sugar residues [189]. Scission is always at a

1145

glycosidic bond of a Man residue. Binding can occur if the sugar residues (P and 8) on either side of the hydrolysed Man are either Man or Glc (14). (Glc) (Glc) (Glc) Man - Man - Man - Man - Man .. .Man a B Y I 5 g 14 The next sugar towards the reducing end of the polymer (s) must be Man but towards the non-reducing end (a) can be Man or Glc. Where examined, the products of hydrolysis have indicated a non-regular distribution of Glc and Man in the chain. Galacto-P(l-4)glucomannan is a significant secondary wall component of gymnosperms and also occurs in smaller amounts in angiosperms. It has been found in the endosperm of seeds of the Liliaceae and Iridaceae and Cercis siliquastrum. It has also been extracted from primary walls and is secreted extracellularly by suspension-cultured cells. [178, 183, 199, 200, 203], The ratio of the three constituent sugars is variable but generally Man is at least 50%. Man content is higher in polymer from whole tissue than from suspension-cultured cells. The distributions of Glc and Man in the main chain and the Gal substitution are non-regular and acetyl esterification occurs. A cell-wall polysaccharide fraction containing Gal, Glc and Man in the ratio 1:2:4 was prepared from the mid-rib of tobacco (Nicotiana tabacum) [204]. Suspension-cultured cells gave a polymer with these sugars in approximately equal proportions [205]. Methylation analysis showed a structure of a p(l-4)glucomannan main chain with Galal- and Gaipi-2Galal- side chains linked to 0-6 of Man residues. Small amounts of arabinosyl, as well as Xyl- and 1,2 linked Xyl residues were also detected. The structures of gluco-manno- and galacto-glucomannosaccharides, released by hydrolysis with Streptomyces Pmannanase or partly purified Trichoderma viride cellulase, supported the structure. In the polysaccharides from fruit and suspension-cultured cells of Actinidia deliciosa (kiwi fruit) [206] Man content was higher in the former and pentose, Gal and Glc contents and the ratios of Glc to Man were lower. Methylation analysis indicated similar sugar linkages to the galactoglucomannan of tobacco. About 30% of Man residues in the fruit polysaccharide were substituted at 0-6, whereas in the polymer from cultured cells about 84% were substituted. Examination of a purified

1146

polymer from suspension-cultured cells of Nicotiana plumbaginifolia [207] gave a Gal:Glc:Man:Xyl:arabinose ratio of 1.0:1.1:1.0:0.04:0.1, with traces of acetyl groups attached to the primary -OH of Gal and Man residues. Methylation analysis showed a P(l-4) glucomannan main chain with 1-6 Gal and 1-2 linked Gal, terminal Xyl- and 1-2 linked Xyl, as well as terminal furanose and pyranose arabinosyl residues substituting the primary -OH of Man. Hydrolysis by a-galactosidase or P-galactosidase, or with both, and estimation of the amounts of Gal released, as well as methylation analysis of the residual polysaccharides, indicated that only Man residues were substituted and P-galactosidase removed almost all the 2-linked Gal without exposing any new 6-OH groups on Man residues. It was calculated that about 38% of side chains were Galal, 42% Gaipi2Galal- and the remaining 20% were terminal pentose. Total substitution of the P(l-4) main chain was 35%. Digestion with P-mannanase produced a series of oligosaccharides and those from d.p. 2-7 were fractionated. The reducing-end sugar was always Man and oligomers with a hetero-chain of P(l-4)Glc and Man with side chains of Galal-, Xyll-, D-Araj^pi- and Gaipi-2 Galal-, always linked to 6-OH of Man, were identified by methylation analysis and n.m.r. and mass spectroscopy. In these oligomers the reducing-end Man could be substituted with any of these groups except Galpl-2Galal-, indicating that the disaccharide side chain prohibits binding of the P-mannanase to the y Man to allow hydrolysis. These substituents, including Gaipi-2Galal-, were found on the Man residue corresponding to the a position, as expected from the binding pattern. (d) Conformation and Interactions of Polymers with a P(1'4) Mannan Main Chain The conformation of these polymers is significant in their selfinteraction, association with other polysaccharides [123, 164, 208-210] and in enzymic reactions. Interacting polysaccharides have a role in plant structure and substituted p-mannans and glucomannans form gels with other polysaccharides that are used in food products. X-ray [211, 212] and electron diffraction [213] of crystalline P-mannan show an extended ribbon-like molecule, in which alternate hydroxymethyl groups lie on opposite edges of the ribbon: the chains are aligned and anti-parallel. In 6-a-galacto-P(l-4) mannans in the solid state [193] the mannan chains are

1147 also highly extended, ribbon-like chains with neighbouring ~CH20H or -CH20Gal groups on opposite edges. There are 2.75 water molecules per monomer repeat. A 3-OH ...0-5 hydrogen bond links adjacent Man residues and 0-5 of a Gal residue is at H-bonding distance from 0-3 of the adjoining Man residue towards the reducing end. The galactomannan chain has a sheet-like structure of total width of almost 15.4A. There are H-bonds between adjacent helices and Gal groups and water molecules are also involved in linkages between chains. A number of observations indicate that elements of this structure are present in aqueous solution, so that molecules have a somewhat extended average conformation. Viscosity is very high with non-ideal flow [179, 210, 214, 215]. Size exclusion chromatography [216] - in both neutral and alkaline solution shows a large molecular size in solution relative to molecular weight. On ultra-centrifiigation [217, 218] there is a non-linear dependence of sedimentation rate on concentration, with hyperfine boundary sharpening. Periodate oxidation gives a rapid and very large decrease in viscosity [216]. For carob galactomannan the specific viscosity decreased to 0.05 of the original value, whereas for a(l-6)dextran and a(l-4)glucan (amylose) at the same concentration the decreases were to 0.80 and 0.48 of those of the unoxidised polymers. Oxidation converts the P(l-4) linked ring structure, in which the bonds of all the atoms involved in the chain structure except one (the glycosidic oxygen) have very restricted rotation, into a flexible 1,3 dioxotrimethylene substituted with hydroxymethyl and C-formyl groups, in which full rotation is theoretically possible in all bonds. Comparison of elution volumes on size exclusion chromatography of the original and periodate-oxidised, borohydride-reduced galactomannans showed a large decrease in molecular size, consistent with a change from a partly extended to a globular shape. Although the poly-aldehyde formed by periodate oxidation forms hemi-acetal linkages between polymer chains [219], borohydride reduction converts these to hydroxyl groups. Experimental results of enzymic synthesis and depolymerization of galactomannan, glucomannan and galactoglucomannan have fitted models of binding between sugar transferases and pmannanase with segments of extended conformers with alternate 5CH2OH or -CHiOGal groups on opposite edges of an extended ribbon [106, 182, 189, 192, 195-197, 207] Conformational energy calculations of galactomannans and p-mannans have indicated [209, 220, 221] that the characteristic ratios (Coo) and persistance lengths (a) lie within the range of the extended ribbon type. When Gal groups are attached to the mannan

1148

chain these values are lower and they decrease as the substitution is increased. They are also affected by the pattern of substitution. Block substitution gives the stiffest chain, an alternating sequence the most flexible, with intermediate values for a random distribution. From n.m.r. spectra of Manpl-4A/a« and Galal-6Man in solution and potential energy calculations, [222] three extended helical galactomanno-oligomer conformations have been generated, one of which was similar to the adopted shape of the p-mannan chain in the solid state as shown by X-ray and electron diffraction. These calculations found that substitution by Gal caused a kink in the backbone, which was enhanced if substitution was consecutive. Some galactomannans (and glucomannans) form gels with xanthan and de-acetylated acetan and increase the gel strength of kappacarrageenan and agarose [179, 208], and these are used in food preparations. Polymer associations are also involved in plant cell wall structure, where cellulose interacts with other polysaccharides, among which glucomannan and galactoglucomannan are based on a P-mannan chain: the combination provides flexibility with strength [223]. Galactomannans with high Gal contents, such as lucerne and fenugreek, do not form gels. However, as well as the Gal content [224] the distribution of Gal substituents is also a factor [197]. Although galactomannan from Leucaena leucocephala and guar have similar Gal contents, only the former gels. Oligomers released on p-mannanase hydrolysis of L leucocephala indicated more frequent regions in which every second Man residue is substituted, giving more stretches of chain in which substituents lie on the same edge of the ribbon-like conformation. Although initial models attributed gel formation to mixed junction zones between unsubstituted regions of mannan chain and the second polymer to form a network, a later modification [224] allowed association if the region of the chain had Gal substitution on only one edge of the ribbonlike conformation. As well as rheological measurements [214, 225-229] this gelation behaviour has been studied by a range of methods, including X-ray fibre diffraction [225, 230, 231] n.m.r. and electron spin resonance spectroscopy [226, 232, 233] molecular modelling [231, 234] X-ray scattering [235] polarising microscopy [236] electron microscopy [229] chiral properties [225, 227] and differential scanning calorimetry [227, 232, 233]. X-ray diffraction of stretched gels of carob galactomannan with kappa-carrageenan [230] gave the same pattern as carrageenan alone, suggesting that mixed-junction zones did not form, whereas with

1149

xanthan,and also with xanthan and glucomannan [225] a new pattern appeared, as would be expected with binding. Results from differential scanning calorimetry and electron spin resonance spectroscopy [233] of the gelation of kappa-carrageenan and glucomannan were interpreted in terms of the formation of mixed aggregates of helices of the former and glucomannan, possibly involving self-aggregated helices covered with surface adsorbed glucomannan. However, n.m.r. spectra of agarose and carob [226] were interpreted to be the consequence of a coupled network with specific junction zones: self-association of galactomannan appeared at high concentrations. For kappa-carrageenan with carob galactomannan or konjac glucomannan [232] much enhanced line broadening was observed, and it was suggested that the gels contained bundles of selfaggregated carrageenan helices covered with surface-adsorbed, disordered maiman chains. Line broadening was not seen with guar. Molecular modelling [234] of kappa-carrageenan and mannan indicated that intermolecular bonding was possible with some change from the conformation with the lowest energy found in the crystal structure. Mixing at room temperature of galactomannans of low Gal content with swollen kappa-carrageenan particles demonstrated binding but at a very much lower level of saturation than after heating [237] and with guar binding did not occur. X-ray diffraction pattems of stretched fibres of xanthan and guar galactomannan have given evidence of an ordered structure and modelling produced a number of associations that were spatially acceptable [231]. Electron microscopy of gels of carob galactomannan and xanthan [229] produced a model in which a network structure of supermolecular strands of xanthan was interconnected and bound together by segments of the mannan chain of galactomannan. Xray diffraction pattems of gels of de-acetylated acetan and glucomannan provided evidence of intermolecular binding involving the formation of novel mixed helical structures [238]. N.m.r. spectroscopy of carob galactomannan on the surface of cellulose crystallites indicated that the majority (70%) of Man and no Gal residues were involved in interacting with the cellulose [239]. A solution of carob galactomannan, when passed through columns of cellulose, adhered as strongly as xyloglucan and P-glucan [240]. Electron microscopy [241] of bacterial cellulose that had been deposited into solutions of glucomannan or three galactomannans with a range of Gal contents, demonstrated the formation of composites. Glucomannan, and galactomannans with low levels of Gal substitution, caused a coalescence of cellulose fibrils that was likened to secondary plant cell wall formation.

1150

N.m.r. spectroscopy suggested that unsubstituted mannan segments bound to cellulose by undergoing a conformational change. Galactomnnans with medium levels of substitution formed crosslinks between cellulose fibrils. 6. Alginic Acid Alginic acid occurs in brown seaweeds where it has a structural role [64, 65, 176, 242-244]. The location is intercellular and it is associated with divalent cations like Ca^"^ and Mg^^. Some species of Gram negative bacteria produce it as an exo-cellular polymer. It consists of an unbranched co-polymer of p-(l-4)D-mannuronosyl and a-(l-4)Lguluronosyl residues. The pyranosyl rings of ManA are in the "^Ci conformation, whereas those of L-GulA are ^€4: the carboxyl groups of both are equatorial. Bacterial alginic acid is acetylated and the acetyl groups are linked to ManA on the 2 or 3, or the 2 and 3 hydroxyls. N.m.r. spectroscopy (^H and ^^C) of polymer partly hydrolysed by mild acid gives the ratio of ManA to L-GulA, the frequencies of single ManA and L-GulA imits, the four nearest-neighbour and eight triad frequencies. The average block length can then be calculated. Polymannuronate and polyguluronate lyases can also be used to selectively depolymerize [106, 245-248]. As lyases, they do not promote transglycosylation but the A4,5 unsaturated uronosyl residue formed at the non-reducing end is the same for both L-GulA and ManA residues. In seaweeds, there has been a wide range of ManA contents determined for alginates from different sources (10-90% ManA) as well as from different parts of the plant. The distribution of the two monomers is not regular. Blocks of ManA and LGulA, as well as heteropolymeric blocks that contain high proportions of ManA and L-GuIA are present. In a series of southern hemisphere seaweeds [249] well-defined blocks were only present in alginates with high ManA or L-GuIA contents. Those with intermediate contents (ManA 60-70%) appeared to contain mainly alternating ManA and LGulA sections of heteropolymeric blocks. Very dilute solutions of the unbranched polyelectrolyte are highly viscous and gels form at low concentrations. Alginic acid in seaweeds is a structural component. Maximum rigidity of gels of isolated polymers is associated with poly-LGulA regions [250]. Poly-ManA regions provide less rigidity and mixed segments the least. Higher molecular weights give higher rigidity. With Ca^^ (and other divalent ions) present even more rigid gels form. Ions mediate binding between the polysaccharide chains [208, 209], In eleven

1151 New Zealand seaweeds the L-GulA content was inversely related to yield [251]. Poly-ManA blocks of d.p. -30 have been prepared from Pseudomonas alginate by utilising lysis of acetylated alginate by polymannuronate lyase. Alternating blocks with a d.p. > 20 were obtained by lysis after de-acetylation [248]. Strains of Pseudomonas aeruginosa are the main pathogen in chronic lung infections of cystic fibrosis patients, where alginate is a major virulence factor [252] and accumulates. Suggested roles include adhesion and protection against phagocytosis and dehydration. Immobilizing of living cells - from bacteria, algae, fimgi, plant protoplasts and animals - by entrapment in beads of calcium alginate gels provides biocatalysts for a wide range of processes such as production of monoclonal antibodies, cell implantation and yeast fermentation [244], Alginate stimulates cytokine production in human monocytes and the amount of ManA residues appears to be significant. Derived oligosaccharides act as elicitors with growth promoting properties for plants [248]. The biosynthesis of alginic acid [69, 243, 244] has been investigated in both seaweeds and bacteria. Azotobacter vinelandii produces exopolysaccharide in a high carbon-low nitrogen environment and hexoses, disaccharides and D-mannitol can all provide a carbon source. This bacterium contains enzymes that convert hexose phosphates into GDPMan (reactions A to D and J) which then undergoes oxidation and polymerisation M GDPMan -> GDP ManA. N GDP Man A + (Man A)n -> (Man A)n+i + GDP Reaction M is catalysed by GDP-mannose dehydrogenase [EC 1.1.1.132] for which the co-factor is two moles of NAD^ and reaction N by GDPmannuronic acid mannuronan mannosyl transferase [EC 2.4.1.33]. ManA residues in mannuronan are convertible to L-GulA residues by a C-5 epimerase for which the substrate is polymeric [243, 244, 253]. At high concentrations of Ca^^ a polymer with a high L-GulA content is produced. There is evidence of a multiple attack mechanism. The epimerase is located extra-cellularly and it does not react with ManA residues in the polymer that are acetylated. In a cell free system of the brown seaweed

1152 Fucus gardneri the presence of enzymes capable of synthesizing mannuronan from Fru 6-P has been demonstrated. Mannuronan 5epimerase activity has been detected in extracts from brown seaweeds. A nucleoside diphospho sugar, identified as GDP-L-GulA, was also found, but there was no evidence that it was incorporated into polymer. When sorbitol 6-P (synonym D-glucitol 6-P = L-gulitol 1-P) was injected into Sargassum muticum [254], after 24 hours ManA blocks were unlabelled but L-GulA were. Mixed ManA:L-GulA blocks were also labelled but only in the L-GulA component. After 72 hours ManA residues were still unlabelled but the ManA component of the mixed segments was radioactive. A pathway was proposed in seaweeds fi*om Fru 6-P to L-gulitol 1-P to GDP-L-Gul and GDP-L-GulA, followed by incorporation into alginic acid. From the slow incorporation of label into some ManA residues, reversibility of the C-5 epimerase was proposed. It has been noted [244, 249] that the occurrence of alternating sugars in a polymer is more readily interpreted in terms of co-polymerization involving two specific transferases than an enzyme modifying a polymer. However, a recombinant mannuronan epimerase from A. vinelandii [255] produced long sections of alternating ManA and L-GulA sequences and during reaction the fraction of the L-GulA-ManA-L-GulA triad increased linearly as a fraction of total L-GulA residues. 7. Polysaccharides with a Low Content of Mannose Residues As well as polymers in which a majority of the sugar residues are Man, a number of other hetero-polysaccharides contain smaller amounts of this sugar. These include bacterial polysaccharides and plant gums and mucilages. However, occasionally a member of the first of these has a high Man content. (a) Bacterial Polysaccharides There are several locations for bacterial polysaccharides [82, 83, 256\. The exo-cellular may occur as a capsule or form an extra-cellular slime and are produced by both Gram positive and negative bacteria. Lipopolysaccharides are embedded in the membrane of Gram negative bacteria. Most of the bacterial polysaccharides consist of repeating hetero-oligosaccharyl units which may be branched: lipopolysaccharides

1153

have a core with a distinctive structure. A great variety of sugars occurs and Man is common. Other sugars with the manno-hexo configuration Table 2. Glycosidic Linkages to D-Manno-Hexo-Sugars in Bacterial Polysaccharides Linkages

Sugar Mana

Manal->4 Manal->2 2Manal-^3^ 2Manal-^4^ 3Manal-^2 3Manal->3 4Manal->5 6Manal->4

6ManaI->2

2(3)Manal->2^ 2(3)Manal->3 2(3)Manal->4 3(2)Manal-^2 3(2)Manal-^3 3(6)Manal->64(6)Manal->24(3)Manal->3 4(3)Manal->2 6Manal-^4 4(2,3)Manal->4

Manp

Manp 1^4

Manpl->6

2Manpi^2

2Manpi-^3

3 Manpi->4^

4Manpi->3^

4Manpi^4 2(3)Manpi^2 2(3)Manpi->3 3(2)Manpi^4 3(4)Manpi^4 4(3)Manpl->3 4(2) Manp 1 ->4 4(3) Manp 1 ->4 4(3,6) Manpl->4

ManNAc

3ManNAcal^3 4ManN Acp 1 ->3 4(3)N AcP 1 ->3

D-Rha

2D-Rhaal->2 3-D-Rhaal-^3

ManNAcA

3ManNAcApi->4

6ManNAcal-P-6 4ManN Acp 1 ->4

3ManNAcpi->4

6(3,4)ManN AcP 1 ->4

2-D-Rhaal^3

4(3)ManN AcAp 1 ^ 3

2-D-Rhaal->4

4ManNAcApi->4

3-D-Rhaal->2

4ManNAcApi->6

4(3)ManNAcAp 1 ->4

ManAal->4* 4ManApl-)'4 D-Rha2NAc3NFopl->3*

2-D-Rha4NH2al->2

* side chain compounds ^ side chain and main chain components numbers are side chain linkages

^ bracketed

1154

that are found, but much less frequently, are ManNAc, D-Rha, ManNAcA, ManA, D-Rha4NH2, D-Rha2NAc3NFo, as well as monomethyl Man substituted at 0-1, 0-2, 0-3, 0-4 or 0-6. 4,6-Pyruvyl Man derivatives have the asymmetric carbon in the S configuration. Acetyl esters are common. Exocellular polymers generally contain an acidic group - uronic acid, pyruvyl acetal (pyr) or phosphate. In contrast to plant heteropolysaccharides their structures have a regularly repeating unit and this is a consequence of the mode of biosynthesis [69, 256]. The number of sugars in the repeating unit usually ranges from two to eight. The polymers consist of a main chain with most having a repeating unbranched side chain. In some cases two repeating unbranched side chains are attached. These side chains vary in d.p. from one to four residues. Table 2 shows types of Man linkages to other sugars from a survey of 90 polysaccharides. Both a- and P- anomeric linkages are found and Man may be in the main chain or the side chain. Three illustrative examples of repeating structural units are 15,16 and 17. -4Glcpi-4Glcpi3

I al pyr4,6ManP 1 -4GlcAp 1 -2Man6 Ac Xanthomonas campestris extra-cellular polysaccharide

15 -3Mana 1 -3Mana 1 -2Mana 1 -2Mana 1 -2aMana 1 Klebsiella 03 lipopolysaccharide

16 -4FucNAca 1 -3 GlcNAcP 1 -4ManN AcP 1 3 3

I

I

al GalNAc

al Glcal-2Glc

Streptococcus pneumoniae XyxiQ 12A polysaccharide

17 The high number of linkage types of Man, combined with the number of other component sugars [83, 256] and the possible variations in

1155

numbers of sugars and sequence in the repeating oligosaccharide, provide a great diversity of polysaccharide structures. Conformational aspects of a nonasaccharide segment from the polysaccharide of a Bradyrhizobium sp. have been determined by molecular dynamics and n.m.r. spectroscopy [257]. A quite extended main chain with a significant amount of flexibility was indicated. The side chain was highly flexible. A few polysaccharides [82], such as the 0-antigens of Klebsiella 03 and 05 and Pseudomonas diminuta consist of a majority of Man residues. Micrococcus spp. produce an unbranched lipo-a-mannan in the cell wall, which is linked 0-2, 0-3, 0-6 (2:2:1) and esterified with succinic acid. Thermoplasma acidophilum has a lipoglucomannan composed of a repeating tetrasaccharide unit-3Manal-2Manal-4Manal-3Glcal. The cell wall of Mycobacterium smegmata and Streptomyces griseus contains a polymer of a(l-4) linked Man30Me. Mycobacterium tuberculosis and leprae and other Mycobacteria [258, 259] contain a lipo-arabinomannan interspersed in the cell wall. The mannan portion has an a(l-6) mannan chain to which are linked many single Manal~->2 groups. Also attached to 0-2 is a polymer of a(l-5)D-Ara/* chains branched through 0-3 and terminated by D-Ara/pi->2 groups, or with these further substituted by Manpi-^5, Manal-^2 Manpi-^5 or Man al-2 Man pi-»5. The reducing end the a(l-6) mannan chain is linked to wyo-inositol at 0-6, which in turn is linked at 0-2 to a Manal-group and through phosphodiester to glycerol lipid, showing similarities in structure to lipooligosaccharides from Mycobacteria spp. [71, 72] {section 4}. A lipomannan, consisting of the mannan section has also been described. Bacterial polysaccharides are antigenic and provide vaccines [260]. Some are used industrially [7(?6]. One of these, xanthan (15), contains Man. In Proteus mirabilis swarming is facilitated by a colony migration factor, which is a capsular polysaccharide containing ManA [261]. Biosynthesis by a block mechanism is usual [69, 256] although some are made by a regular sequence of addition from nucleoside diphospho sugars. In the block mechanism, the sugars are assembled sequentially as polyprenol diphospho diesters, to form a lipo-oligosaccharide, and then transferred as a block to the growing polymer, leading to a regularly repeating structure. Direct addition from nucleoside diphospho sugars, where it occurs, also leads to regularity, due to the strictness of the transferase enzymes for the donor nucleoside diphospho sugar and the acceptor sugar, as well as for the anomeric linkage produced and the OH in the acceptor to which it is joined [69, 256, 262],

1156 Methods of determination of structure include complete hydrolysis and identification of component sugars (which may be unusual) and any further substituents such as pyruvate or esters; methylation analysis; partial hydrolysis (by acid, acetolysis or enzymically) to oligosaccharides followed by n.m.r. spectroscopy, which has also been applied to whole polymers; modification of uronosyl residues - by reduction prior to or after methylation, and by p-elimination reactions; Smith periodate degradation. N.m.r. spectroscopy of whole polymers avoids problems associated with partial hydrolysis and acetolysis - incompleteness, low yields and differential loss of sugars. Enzymic partial depolymerization [106, 263] which may be by endo-hydrolase or lyase action, provides high yields of oligosaccharide related to the repeating unit and leaves intact acid-sensitive substituents such as pyruvyl acetal, as well as often cleaving at a different glycosidic bond to that split by hydrolysis or acetolysis. Many of the depolymerases are derived from bacteriophages some from other microorganisms - and they may be associated with the production of exo-polysaccharide. (b) Plant gum exudates and mucilages Plant gum exudates [264-267] are formed in response to disease, insect and mechanical damage and physiological stress such as drought. They appear to seal and protect damaged sections. The isolated gums form viscous solutions and find use in pharmaceutical and food preparations. The structures are highly branched, and characterised by the presence of uronic acids (GlcA, GlcA40Me and GalA) and the absence of Glc. They can be grouped into substituted arabinogalactans, glucuronomannans, galacturonorhamnans and glycanoxylans. The chains comprising the branched structure divide into the single basal chain, which contains the reducing-end sugar, those joined to it by glycosidic links - first tier chains - and then additional tiers joined sequentially. Their structures have some resemblance to polymers found normally in plant cells - arabinogalactan, glucuronomannan, pectic substances and xylans. Protein may be covalently linked. A diversity of component sugars, glycosidic linkages and branching increase the difficulty of depolymerisation by the enzymes of pathogens. Due to their highly branched structure, the number of constituent sugars and the limited regularity of structural elements, specific structures cannot be assigned - only generalised formulae. Methods of determining structural elements include methylation analysis,

1157 Smith periodate degradation, partial acidic hydrolysis, reduction of uronosyl residues, acetolysis, as well as base elimination reactions and other degradation procedures that can be applied to uronosyl residues, and n.m.r. spectroscopy [261]. Increasing availability of purified enzymes with well-defined action patterns should assist future studies [106, 262]. Structural features can also be recognised by monoclonal antibodies [268]. Table 3. Chain Structure of Gum Ghatti Chain position Basai First tier

Second tier

Structures (4GlcApi-2Manal)„ L-Ara/1L- Ara/1-2 (or 3 or 5) L-Ara/GicAal-(6Galpl)„-3L-Ara Gaipi-(6Gaipi)„-3L-Ara L-Ara/1L-Ara/1 -2 (or 3 or 5) L-Ara/1 -

Linkage to lower positioned chain

— -^6 Man ->6 Man ->3 Man ^ 3 Man ->3Gal ->3Gal

A number of plant gums contain Man, Anogeissus (gum ghatti), Combretum, Encephalartos (cycad) spp., Chohsia speciosa [269] and Grevillea robusta [270] have been extensively examined, but Man also occurs in Albizia, Virgilia, Hakea, Brabeium and Prunus spp. Determination of Man content in 16 species of cycads gave considerable variation [271] but it was always a minor component (<10%). These gums have as a structural unit in the basal chain - 4GlcApl-2 Manal-. Since the glycosidic linkages of glycuronosyl residues are much more resistant to acidic hydrolysis than those of neutral sugars, on hydrolysis the aldobiuronic acid GXck^X-lMan is recovered. When Smith periodate degradation has been performed Man has persisted: in G. robusta gum it remained after three degradations, consistent with its location in the basal chain. In Anogeissus latifolius gum [264-267] Table (3) the Man residues carry substituents on 0-3 and 0-6 and the GlcA residues are unsubstituted. The P(l-6) Gal sequences in the first tier can consist of six or more residues. The structure of Encephalartos longifolius gum [272] has the same repeating unit -4GlcApl-2Manal- in the basal chain as gum ghatti. Some differences in the structure are substitution of Man residues at 0-4 with L-Rha30me or L-Rha groups, as well as no substitution at 06. There is also substitution of GlcA in the basal chain with L-Ara/ groups at 0-3. The Gal chains in the first tier contain P(l-3) linkages. Branching occurs in these at 0-6 by GlcApl-6Galpl-> units, which can be further extended through 0-4 of the GlcA residues by (-L-Rha l-^)n

1158

where n=l-3, or branched at 0-3 of Gal residues by L-Ara groups. Some L-Rha residues are 2,4 disubstituted with L-Rha units. Encephalartos friderici-guilielmi gum [273] also contained P(l-3) linked Gal residues, L-Rha and L-Rha30Me residues and the same basal chain. An exudate of Ceratozamia spinosa (cycad) also had this basal chain [274] with L-Rha, L-Rha30Me or L-Ara/linked to 0-3 of some GlcA residues. The 0-4 of Man was substituted with GlcApi-. The 0-4 of these GlcA residues were substituted with L-Rhaal-, L-Arayfecl- or Xyipi-. 0-3 of the same GlcA residues could be substituted with these sugars to give a second tier. Plant mucilages are associated with normal metabolism and are seed components [275'] but are also found in fruits, leaves and inner bark. They are used medicinally. Polymers with a repeating 4GlcApl-2Manal unit, in which the amount of this aldobiuronosyl residue is higher than in gums, and sufficient in most to be described as a main chain, have been detected. A polysaccharide secreted by leaves of Drosera capensis contained 64% of Man plus GlcA with L-Ara/'and Xyl groups attached to 0-3 of GlcA, and Gal groups to 0-3 of Man residues [276]. Previously, an extra-cellular polymer from suspension-cultured cells of tobacco {N. tabacum) containing 49% of Man plus GlcA, with L - A r a / a l - groups linked to 0-3 of some GlcA and L-Ara/pi- joined to 0-3 of some Man residues [277, 278] had been described. The polysaccharide from the stem pith of Actinidia deliciosa consisted of less aldobiuronic acid (20%) [279]. It had (1-3) Gal chains joined to 0-3 of Man and L-Ara/to 0-3 of GlcA. The (1-3) Gal chains can be terminated with D-Aral-5-L-Ara/l->3 or carry a second tier of L-Fucl->2 (or 1-^6), L-Ara/1~>2 (or 1~>6) or Gal 1^6. The leaves of Dicerocaryum zanguebaricum - an African plant - have specialised mucilagenous hairs. Water extraction affords a polysaccharide [280] with a simple pattern of substitution of the glucuronomannan chain. This main chain is substituted at the 0-3 of Man and some GlcA units with single Xyipi- or Galal groups. The carboxylreduced polymer contained 56% of Man plus Glc. 8. Synthesis of Mannans and Manno-oligomers by in vitro Methods. The stereo-regular, chemical synthesis of a-manno-pyranans with (1-2), (1-3) or (1-6) linkages by a mechanism of chain growth from monomeric derivatives, as well as a singly branched a(l-6) polymer from a disaccharide derivative has been reviewed [209, 281] and the synthesis of

1159 heteropolymers discussed [282]. Oligomers with specific linkages have been prepared by sequential addition of appropriately protected monomers [283, 284] and also by enzymic addition with glycosyl transferases and nucleoside diphospho sugar. [J36, 262]. Block addition can also be used. The oligomers provide the glycan segment for the synthesis of neoglycoconjugates [96, 283] which consist of carbohydrate joined synthetically to other molecules such as proteins, lipids and synthetic polymers. The carbohydrate can also be derived from naturally occurring glycoconjugates. These compounds find a range of uses in biochemical separations and purification, assays and the study of cellular processes. An unbranched a(l-6) mannan was synthesized from 1,6-anhydro-Dmannopyranose, with benzyl ether protecting groups on 2, 3 and 4-OH, by polymerisation at -78° initiated by phosphorus pentafluoride, followed by debenzylation with Na in liquid NH3. Results from periodate oxidation and optical rotation were consistent with an a(l-6) structure, and from the viscosity average d.p. values of up to about 1,000 were estimated [285]. The polymer cross-reacted with an a(l-6) mannan from Trichophyton rubrum. Polymerising in the presence of dimethoxymethane reduced the average d.p. and allowed the synthesis of oligomers [286]. Cationic catalysts polymerised l,3-anhydro-2,4,6-tri-0-benzyl D-mannopyranose to a product with an average d.p. of 60-90, which, after debenzylation, was shown by n.m.r. and infra-red spectroscopy to be virtually completely an a(l-3) mannan with an average d.p. of 30-60 [287]. Polymerisation of the benzyl protected 1,2-anhydro sugar with potassium tert-butoxide complexed with a crown ether, followed by debenzylation, gave a stereoregular product with a low average d.p. [288]. Starting with a benzyl protected anhydro disaccharide - l,6-anhydro-2,3-di-0-benzyl-4(2,3,4,6-tetra-0-benzyl-a-D-mannopyranose)-p-D-mannopyranose - and polymerising at -60° with PF5, followed by debenzylation, gave an a(l-6) mannan with single Man groups on the 4-OH. The structure was supported by n.m.r. spectroscopy. The average d.p. was much lower (56) than that obtained for the unbranched a(l-6) linked polymer [289]. P(l4) Mannan (from ivory nut) and a glucomannan (konjac) have been branched by reacting them with the 3,4,6 acetylated 1,2-ethylorthoacetate of P-D-glucopyranose in the presence of 2,6-dimethyl-pyridine perchlorate [290]. Unbranched hetero-polysaccharides have been prepared by condensation of hetero-oligomers with a l,2-0-(lcyanoethylidene) group on the reducing-end sugar, a triphenylmethyl

1160 ether at the reacting OH, and non-reacting groups protected with benzyl (for OH) or as phthalimide (for NH2). Reaction of the protected repeating tetrasaccharide unit of Shigella flexneri with triphenylmethylium perchlorate in dichloromethane, followed by removal of the protecting groups gave a product with an average d.p. of about 40. Its n.m.r. spectrum was identical to that of the native 0-antigenic polysaccharide [291]. Earlier, polymerisation of the cyanoethylidene derivative of Manpi-4-L-Rhaal-3Ga/, in which the 6-OH of Man was substituted as a triphenylmethyl ether and all other free hydroxyls were acetylated gave, after removal of protecting groups, a fraction with an estimated d.p. of 30 [292]. The P(l-6) linked polymer from this trisaccharide is the Oantigenic polysaccharide of Salmonella newington, and in haemagglutin tests the synthetic polymer gave a positive reaction in which it was eightfold times as active as the trisaccharide. The synthetic polymer in which the anomeric linkage of Man was a was practically inactive. Unbranched a-D-mannopyranans with (1-6) and (1-4) linkages with an average d.p. of 10 and 15 have been prepared by this method [283]. A synthesis of a lipid-mannodisaccharide-phosphate, P-6Manal2Manal-0-octyl, a partial structure of phosphorylated glycoproteins that are targets for the Man 6-P transporter, consisted of the following sequence [293]. Octanol was reacted with 2-0-acetyl-3,4,6-tri-0-benzyla-D-mannosyl chloride. Removal of acetyl and glycosylation with the ethyl thioglycoside of 6-0-acety 1-2,3,4-tri-O-benzyl-a-D-mannopyranose gave a disaccharide, which, after removal of acetyl, was phosphorylated with chlorodiphenylphosphate. Hydrogenolysis gave the dimannosephosphate. The a(l-3), a(l-6) branched trisaccharide (linked to lipid) which is part of the common core of glycoproteins has been conveniently prepared [294] by reacting unprotected mannoside-0-lipid with 2 moles of the 2,3,4,6-tetra-O-benzoyl-l-trichloroacetamide, followed by debenzoylation. The unbranched pentasaccharide segment (containing DRha) of a antigenic glycopeptidolipid of Mycobacterium arium was made by block synthesis with the thioether of the protected disaccharide and the trisaccharide with one available OH: the promoter was N-iodosuccinimide and trifluoromethanesulphonic acid [295]. Synthesis of the pentasaccharide Mana 1 -2[Mana 1 -6]ManP 1 -4[GlcNAcp 1 -6]GlcNAcp 1 0-/7-methoxyphenyl and comparison of the n.m.r. spectrum with that of the core structure from a glycan of Chinese hamster ovary suggested that the latter had a GlcNAc (1-6) linked branch with a novel a-linkage [296].

1161 The P-Man residue was joined to the 4-OH of GlcNAc with silver alumina-silicate as promoter. The enzymic synthesis of oligosaccharides avoids the need for the addition and removal of protecting groups and, with Man residues, overcomes the difficulty of synthesising the (J-configuration. The core trisaccharide of Salmonella group Ei has been made in a reaction of GDPMan with L-Rhaal-3Ga/, catalysed by a recombinant P(l-4) mannosyl transferase. Reaction was rapid and the yield was high [297]. Glycoprotein oligosaccharides have also been prepared by removal of sugars by glycosidase hydrolysis of native oligomers [137]. Sequential addition of specifically substituted monosaccharides has given a tri-antennary undecasaccharide with the structure of a glycan chain of the N-linked glycoprotein of calf thyroglobulin (18). Synthesis started from the trisaccharide Manpi-4GlcNH2 ^\-4GlcNH2 [298]. Mal-2M al 6 Mal-2Mal-3M al

I 6 Ma 1-2 Ma 1-2 M a 1-3 MP 1-4 GlcNAcpi-^G/cA^^ic 18 The NH2 groups were protected as phthalimides, and all OH groups except the 3 and 6 on the terminal Man, as benzyl ethers. Selective protection of the more reactive primary 6-OH with chloracetyl chloride allowed glycosylation of 3-OH. Alternatively, the 6-OH could be selectively silylated with a bulky tert-butyl dimethyl silyl group. Glycosylation with the 1-trichloro-acetamide of 2-0-acetyl-3,4,5-tri-0benzyl-a-D-mannoside, with BF3 in ether as promoter, gave the tetrasaccharide with the first Man residue of the trimannosyl antenna in posifion. Removal of the silyl group with HF and reaction with 1-chloro3,6-di-O-acety 1-2,4-di-O-benzyl-a-D-mannoside added the second branched Man residue. De-acetylation and glycosylation with 3 molecules of 1 -chloro-2-0-acetyl-2,4,6-tri-0-benzyl-a-D-mannoside gave an octasaccharide and repetifion of de-acetylation and glycosylation gave the

1162 protected undecasaccharide. Hydrazine hydrate in ethanol and then acetic anhydride converted phthalimide into N-acetyl and hydrolysed 0-acetyl. Hydrogenolysis then gave an undecasaccharide, for which the n.m.r. and mass spectra were in agreement with those of the native glycan. A procedure for the rapid synthesis of Man glycans that removes the need for the unmasking of protecting groups between glycosylation steps, by utilising the difference in reactivity of seleno- and thio-glycosidic groups when reacted with N-iodosuccinimide and trifluoromethanesulphonic acid (combined with protection of the 3 and 4 OH groups as the cyclohexane di-acetal) has been described [299]. A nona-mannosaccharide containing a(l-2), (1-3) and (1-6) linkages, which is a glycan chain found in the viral coat of HIV-1, has been prepared. The synthesis involved only one operation with a protecting group and several glycosylations could be combined, removing the need for the isolation of intermediates. ABBREVIATIONS Cer Dol d.p. Fo P PMI Pyr

= = = = = = =

ceramide dolichyl degree of polymerisation formyl phosphate phosphomannoisomerase pyruvyl acetal

REFERENCES [1]

lUB-IUPAC Joint Commission on Biochemical Nomenclature, J.Biol.Chem., 1982,257,3347-3354. [2] Purich, D.L.; Fromm, H.J.; Rudolph, F.B.; Adv. EnzymoL, 1973, 39, 249-326. [3] Herold, A.; Lewis, D.H.; New PhytoL, 1977, 79, 1-40. [4] Myers, D.K.; Matheson, N.K.; Phytochemistry, 1994, 37, 957-969. [5] Topper, Y.J. In The Enzymes, Boyer, P.D.; Lardy, H,; Myrback K., Ed.; Academic Press: New York, 1961, vol. 5, pp. 435-44 L [6] Lee, B.T.; Matheson, N.K.; Phytochemistry, 1984, 23, 983-987. [7] Small, D.M.; Matheson, N.K.; Phytochemistry, 1979, 18, 1147-1150. [8] Guha, S.K.; Rose, Z.B.; Arch. Biochem. Biophys., 1985, 243, 168-173. [9] Zielinski, N.A.; Chakrabarty, A.M.; Berry, A.; J. Biol. Chem., 1991, 266, 97549763. [10] Ye, R.W.;Zielinski,N.A.;Chakrabarty, A.M.; J. BacterioL, 1994, /7d,48514857.

1163

[10a] Pirard, M.; Achouri, Y.; Collet, J-F.; Schollen, E.; Matthijs, G.; Van Schaftingen, E.; Biochem. J., 1999, 557, 201-207. [11] Mizuno, H.; Morita, M.; Akanuma, H.; J. Biochem. (Tokyo), 1995, 77(5, 411-417. [12] Yushok, W.D.;y. Biol Chem.. 1911,246, 1607-1617. [13] Markovitz, A.; Sydiskis, R.J.; Lieberman, M.M.; J. Bacteriology, 1967, 94, 14921496. [14] De la Fuente, M.; Peftas, P.F.; Sols, A.; Biochem. Biophys. Res. Commun., 1986, 7^^,51-55. [15] Fromm, H.J.; Zewe, V.; J. Biol. Chem., 1962, 237, 1661-1667. [16] Kime, M.J.; Ratcliffe, R.G.; Loughman, B.C.; J. Exp. Bot., 1982, 33, 670-681. [17] Watkins, C.B.; Frenkel,; C. Plant Physiol., 1987, 85, 56-61. [18] Van Steveninck, J.; Biochim. Biophys. Acta, 1968,163, 386-394. [19] Herve, M.; Wietzerbin, J.; Lebourguais, O.; Tran-Dinh, S.; Biochimie, 1992, 74, 1103-1115. [20] Roberts, J.K.M.; Ray, P.M.; Wade-Jardetzky, N.; Jardetzky, O.; Nature, 1980, 283, 870-872. [21] Chen, W.; Gueron, M.; Biochimie, 1992, 74, 867-873. [22] Miginiac-Maslow, M.; Hoarau, A.; Z PfanzenphysioL, 1982. 107, 427-436. [23] Harris, G.C.; Gibbs, P.B.; Ludwig, G.; Un, A.; Sprengnether, M.; Kolodny, N.; Plant Physiol, \9U,82, 1081-1089. [24] Fromm, H.J.; Eur. J. Biochem., 1969, 7, 385-392. [25] Higgins, T.J.C.; Easterby, J.S.; Eur. J. Biochem., 1974, 45, 147-160. [26] Renz, A.; Stitt, M.; Planta, 1993,190, 166-175. [27] Matheson, N.K.; Myers, D.K.; Phytochemistry, 1998, 48, 241-248. [28] Garrard, L.A.; Humphries, T.E.; Phytochemistry, 1969, 8, 1065-1077. [29] Stenlid, G.; Physiol Plantarum, 1971, 25, 397-401. [30] Lucas, WJ.; Wilson, C ; Plant Physiol, 1987, 85,423-429. [31] Thurston, C.F.; John, P.C.L.; Syrett, P.J.; Arch. Mikrobiol, 1973, 88, 135-145. [32] Grossbard, L.; Schimke, R.T.; J. Biol Chem., 1966, 2^7, 3546-3560. [33] Fromm, H.J.; Zewe, V.; J. Biol Chem., 1962, 237, 3027-3032. [34] Turner, J.F.; Chensee, Q.J.; Harrison, D.D.; Biochim. Biophys. Acta, 1977, 480, 367-375. [35] Turner,J.F.; Copeland, L.; Plant Physiol, 1981, 68, 1123-1127. [36] Doehlert, D.C.; Plant Physiol, 1989, 89, 1042-1048. [37] Baijal, M; Sanwal, G.G.; Phytochemistry, 1976, 15, 1859-1863. [38] Galina, A.; Reis, M.; Albuquerque, M.C.; Puyou, A.G.; Puyou, M.T.G.; de Meis, L.; Biochem. J., 1995, 309, 105-112. [39] Gancedo, J.M.; Eur. J. Biochem., 1992, 206, 297-313. [40] Graham, I.A.; Denby, K.J.; Leaver, C.J.; Plant Cell 1994, 6,161-112. [41] Jang, J-C; Sheen, J.; Plant Cell, 1994, 6, 1665-1679. [42] Bielski, R.L. In Encyclopedia of Plant Physiology; Loewus, F.A.; Tanner, W., Ed.; Springer-Verlag: Berlin, 1982, Vol. 13A, pp. 158-192. [43] Lewis, D.H. In Storage Carbohydrates in Vascular Plants', Lewis, D.H., Ed.; Cambridge University Press: Cambridge, 1984, pp. 157-179. [44] Beck, E.; Hopf, H. In Methods in Plant Biochemistry, Dey, P.M.; Harborne, J.B., Ed.; Academic Press: London, 1990; vol. 2, pp. 235-289. [45] Loescher, W.H.; Physiol Plantarum, 1987, 70, 553-557.

1164

[46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72]

Rumpho, M.E.; Edwards, G.E.; Loescher, W.H.; Plant Physiol., 1983, 73, 869873. Everard, J.D.; Franceschi, V.R.; Loescher, W.H.; Plant Physiol.. 1993, 102, 345356. Keller, F.; Matile, P.; New PhytoL, 1989, 113, 291-299. Salmon, S.; Lemoine, R.; Jamai, A.; Bouche-Pillon, S.; Fromont, J.C; Planta, 1995, 7P7, 76-83. Stoop, J.M.H.; Pharr, D.M.; Plant Physiol, 1994, 106, 503-511. Jennings, D.B.; Ehrenshaft, M.; Pharr, D.M.; Williamson, J.D.; Proc. Natl Acad. ScL USA, 1998, P5, 15129-15133. Henstra, S.A.; Tuinhof, M.; Duurkens, R.H.; Robillard, G.T.; J. Biol. Chem., 1999,27^,4754-4763. Loescher, W.H.; Tyson, R.H.; Everard, J.D.; Redgwell, R.J.; Bieleski, R.L.; Plant Physiol., 1992, 98, 1396-1402. Horwitz, S.B.; Kaplan, N.O.; J. Biol. Chem., 1964, 239, 830-838. Kiser, R.C.; Niehaus, W,G.;Arch. Biochem. Biophys., 1981, 211, 613-621. Richter, D.F.E.; Kirst, CO.; Planta, 1987, 170, 528-534. Ruffner, H.P.; Rast, D.; Tobler, H.; Karesch, H.; Phytochemistry, 1978, 17, 865868. Morton, N.; Dickerson, A.G.; Hammond, J.B.W.; J. Gen. Microbiol., 1985, 131, 2885-2890. Slatner, M.; Nidetzky, B.; Kulbe, K.D.; Biochemistry, 1999, 38, 10489-10498. Stoop, J.M.H.; Pharr, D.M.; Arch. Biochem. Biophys., 1992, 295, 612-619. Stoop, J.M.H.; Pharr, D.M.; Plant Physiol., 1993, 103, 1001-1008. Prata, R.T.N.; Williamson, J.D.; Conkling, M.A.; Pharr, D.M.; Plant Physiol., 1997,77^,307-314. Manners, D.J.; Sturgeon, R.J. In Encyclopedia of Plant Physiology; Loewus, F.A.; Tanner, W., Ed.; Springer-Verlag: Berlin, 1982; Vol. 13A, pp. 472-514. Painter, T.J. In The Polysaccharides; Aspinall, G.O., Ed.; Academic Press: New York, 1983; Vol. 2. pp. 195-285. Percival, E.; McDowell, R.H.; In Biochemistry of Storage Carbohydrates in Green Plants; Dey, P.M.; Dixon, R.A., Ed.; Academic Press: London, 1985; pp. 305-348. Chizhov, A.O.; Dell, A.; Morris, H.R.; Reason, A.J.; Haslam, S.M.; McDowell, R.A.; Chizhov, O.S.; Usov, A.I.; Carbohydr. Res., 1998, 310, 203-210. Lezica, R.P.; Daleo, G.R.; Dey, P.M.; Adv. Carbohydr. Chem. Biochem., 1986, ^^,341-385. Elbein, A.D.; Kaushal, G.P. In Methods in Plant Biochemistry; Dey, P.M., Ed.; Academic Press: London, 1990; Vol. 2, pp. 79-110. James, D.W.; Preiss, J.; Elbein, A.D. In The Polysaccharides; Aspinall, G.O., Ed.; Academic Press: Orlando, Florida, 1985; Vol. 3, pp. 107-207. Elbein, A.D. In Encyclopedia of Plant Physiology; Tanner, W.; Loewus, F.A., Ed.; Springer-Verlag: Berlin, 1981; Vol. 13B, pp. 166-193. Severn, W.B.; Furneaux, R.H.; Falshaw, R.; Atkinson, P.H.; Carbohydr. Res., 1998, i(?^, 397-408. Hoppe, H.C.; de Wet, B.J.M.; Cywes, C; Daffe, M.; Ehlers, M.R.W.; Infect. Immun., 1997, 65, 3896-3905.

1165

[73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99]

McNeil, M.; Tsang, A.Y.; McClatchy, J.K.; Stewart, C ; Jardine, I.; Brennan, P.J.; J. BacterioL, 1987,169, 3312-3320. Lennarz, W.J.; Talamo, B.; J. Biol Chem., 1966, 241, 2707-2719. Bultel-Ponce, V.; Debitus, C ; Blond, A.; Cerceau, C ; Guyot, M.; Tetrahedron Lett, 1997,3(^,5805-5808. Niepel, T.; Meyer, H.; Wray, V.; Abraham, W-R.; Tetrahedron, 1997, 53, 35933602. Kochetkov, N.K.; Smirnova, G.P.; Adv. Carbohydr. Chem. Biochem., 1986, 44, 387-438. Laine, R.A.; Renkonen, O.; Biochemistry, 1974, 13, 2837-2843. Smith, C.K.; Hewage, C M . ; Fry, S.C; Sadler, I.H.; Phytochemistry, 1999, 52, 387-396. Cohen, R.E.; Ballou, C.E. in Encyclopedia of Plant Physiology; Tanner, W.; Loewus, F.A., Ed.; Springer-Verlag: Berlin, 1981; Vol. 13B, pp. 441-458. LiSkova, D.; Auxtova, O.; Kakoniova, D.; Kubackova, M.; Karacsonyi, S.; Bilisics, L.; Planta, 1995, 196,425-429. Kenne, L.; Lindberg, B. In The Polysaccharides; Aspinall, G.O., Ed.; Academic Press: New York, 1983; Vol. 2, pp. 287-363. Lindberg, B.; Adv. Carbohydr. Chem. Biochem., 1990, 48, 279-318. Gorin, P.A.G.; Barreto-Bergter, E. In The Polysaccharides', Aspinall, G.O., Ed.; Academic Press: New York, 1983; Vol. 2 pp. 365-409. Barreto-Bergter, E.; Gorin, P.A.J.; Adv. Carbohydr. Chem. Biochem., 1983, 42, 67-103. Lehle, L. in Encyclopedia of Plant Physiology; Tanner, W.; Loewus, F.A., Ed.; Springer Verlag: Berlin, 1981; Vol. 13B, pp. 459-483. Tanner, W.; Lehle, L.; Biochim. Biophys. Acta, 1987, 906, 81-99. Nakajima, T.; Ballou, C.E.; J. Biol. Chem., 1974, 249,1619-16^:16^5-169A. Faille, C.; Wieruszeski, J-M.; Michalski, J-C; Poulain, D.; Strecker, G.; Carbohydr. Res., 1992,236, 17-27. Trinel, P.A.; Lepage, G.; Joualt, T.; Strecker, G.; Poulain, D.; FEBS Lett., 1997, 416, 203-206. Kobayashi, H.; Mitobe, H.; Takahashi, K.; Yamamoto, T.; Shibata, N.; Suzuki, S.; Arch. Biochem. Biophys., 1992, 294, 662-669. Teixeira, A.Z.A.; lacomini, M.; Gorin, P.A.J.; Phytochemistry, 1992, 31, 34673470. Parolis, L.A.S.; Parolis, H.; Kenne, L.; Meldal, M.; Bock, K.; Carbohydr. Res., 1998, 309, ll'Sl. Ikuta, K.; Shibata, N.; Blake, J.S.; Dahl, M.V.; Nelson, R.D.; Hisamichi, K.; Kobayashi, H.; Suzuki, S.; Okawa, Y.; BiochemJ., 1997, 323,297-305. Prieto, A.; Leal, J.A.; Poveda, A.; Jimenez-Barbero, J.; Gomez-Miranda, B.; Domenech, J.; Ahrazem, O.; Bernabe, M.; Carbohydr. Res., 1997, 304, 281-291. Lennarz, W.J.; Hart, G.W. Ed., Methods EnzymoL, 1994, Vol. 230. Lussier, M.; Gentzsch, M.; Sdicu, A-M.; Bussey, H.; Tanner, W.; J. Biol. Chem., 1995,270,2110-2115, Hernandez, L.M.; Ballou, L.; Alvarado, E.; Tsai, P-K.; Ballou, C.E.; J. Biol. Chem., 1989, 264, 13648-13659. Ballou, L.; Hernandez, L.M.; Alvarado, E.; Ballou, C.E.; Proc. Natl. Acad Sci. USA, 1990, 87,336^-3312.

1166 [100] Kobayashi, H.; Shibata, N.; Watanabe, M.; Komido, M.; Hashimoto, N.; Hisamichi, K.; Suzuki, S.; Carbohydr. Res., 1992, 257, 317-323. [101] Shibata, N.; Ikuta, K.; Imai, T.; Satoh, Y.; Satoh, R.; Suzuki, A.; Kojima, C; Kobayashi, H.; Hisamichi K.; Suzuki, S.; J. Biol. Chem., 1995, 270, 1113-1122. [102] Shibata, N.; Akagi, R.; Hosoya, T.; Kawahara, K.; Suzuki, A.; Ikuta, K.; Kobayashi, H.; Hisamichi, K.; Okawa, Y.; Suzuki, S.; J. Biol. Chem., 1996, 271, 9259-9266. [103] Shibata, N.; Onozawa, M.; Tadano, N.; Hinosawa, Y.; Suzuki, A.; Ikuta, K.; Kobayashi, H.; Suzuki, S.; Okawa, Y.; Arch. Biochem. Biophys., 1996, 336, 4958. [104] Kobayashi, H.; Watanabe, M.; Komido, M.; Matsuda, K.; Ikeda-Hasebe, T.; Suzuki, M.; Shibata, N.; Hisamichi, K.; Suzuki, S.; Carbohydr. Res., 1995, 267, 299-306. [105] Ogawa, K.; Nishikori, J.; Ino, T.; Matsuda, K.; Biosci. Biotech. Biochem., 1994, 58, 560-562. [106] McCleary, B.V.; Matheson, N.K.; Adv. Carbohydr. Chem. Biochem., 1986, 44, 147-276. [107] Shibata, N.; Senbongi, N.; Hosoya, T.; Kawahara, K.; Akagi, R.; Suzuki, A.; Kobayashi, H.; Suzuki, S.; Okawa, Y.; Eur. J. Biochem., 1997, 246, 477-485. [108] Li, R-K.; Cutler, J.E.; J. Biol. Chem., 1993, 268, 18293-18299. [109] Kobayashi, H.; Suzuki, J.; Tanaka, S.; Kiuchi, Y.; Oyamada, H.; Iwadate, N.; Suzuki, H.; Shibata, N.; Suzuki, S.; Okawa, Y.; Arch. Biochem. Biophys., 1997, 341,10-14. [110] Jimenez-Barbero, J.; Prieto, A.; Gomez-Miranda, B.; Leal, J.A.; Bernabe, M.; Carbohydr. Res., 1995,272, 121-128. [Ill] Bilang, M.; Attioui, F.; Loppinet, V.; Michalski, J-C; Bonaly , R.; Carbohydr. Res., 1996,2(5^,303-313. [112] Bacon, B.E.; Cherniak, R.; Kwon-Chung, K.J.; Jacobson, E.S.; Carbohydr. Res., 1996,255,95-110. [113] Sheng, S.; Cherniak, R.; Carbohydr. Res., 1997, 301, 33-40. [114] Leal, J.A.; Jimenez-Barbero, J.; Gomez-Miranda, B.; Prieto, A.; Domenech,J.; Bernabe, M.; Carbohydr. Res., 1996, 283, 215-222. [115] Latge, J-P.; Kobayashi, H.; Debeaupuis, J-P.; Diaquin, M.; Sarfati, J.; Wieruszeski, J-M.; Parra, E.; Bouchara, J-P.; Fournet, B.; Infect. Immun., 1994, 62, 5424-5433. [116] Tanaka, N.; Awai, A.; Bhuiyam, M.S.A.; Fujita, K.; Fukui, H.; Takegawa, K.; J. BacterioL, 1999, 181, 1356-1359. [117] Corsaro, M.M.; De Castro, C; Evidente, A.; Lanzetta, R.; Molinaro, A.; Mugnai, L.; Parrili, M.; Surico, G.; Carbohydr. Res., 1998, 308, 349-357. [118] Teixeira, A.Z.A.; lacomini, M.; Gorin, P.A.J.; Carbohydr. Res., 1995, 266, 309314. [119] lacomini, M.; Schneider, C.L.; Gorin, P.A.J.; Carbohydr. Res., 1985. 142, 237251. [120] Gorin, P.A.J.; lacomini, M.; Carbohydr. Res., 1985,142, 253-267. [121] Da Silva, M. de L.C.; Gorin, P.A.J.; lacomini, M.; Phytochemistry, 1993, 34, 1\5-1\1. [122] Bergwerff, A.A.; Stark, W.; Fendrich, G.; Knecht, R.; Blommers, M.J.J.; Maerki, W.; Kragten, E.A.; Van Oostrum, J.; Eur. J. Biochem., 1998, 253, 560-575.

1167

[123] Brant, D.A. In The Biochemistry of Plants; Preiss, J., Ed.; Academic Press: New York, 1980; Vol. i , pp. 425-472. [124] Montrieul, J.; Schachter, H.; Vliegenthart, J.F.G., Ed.; Glycoproteins, Elsevier: Amsterdam, 1995. [125] Montreuil, J.; Adv. Carbohydr Chem. Biochem., 1980, J7, 157-223. [126] Kornfeld, R.; Komfeld, S.; Ann. Rev. Biochem., 1985, 54, 631-664. [127] Elbein, A.D.; Plant Physiol, 1988, 87, 291-295. [128] Selvendran, R.R.; O'Neill, M.A. in Encylopedia of Plant Physiology, Loewus, F.A.; Tanner, W., Ed.; Springer-Verlag: Berlin, 1982; Vol. ISA, pp. 515-583. [129] Kaushal, G.P.; Szumilo, T.; Elbein, A.D. in The Biochemistry of Plants, Preiss, J., Ed.; Academic Press: San Diego, 1988; Vol. 14, pp. 421-463. [130] Hounsell, E.F.; Adv. Carbohydr. Chem. Biochem., 1994, 5^, 311-350. [131] Vliegenthart, J.F.G.; Dorland, L.; van Halbeek, H.; Adv. Carbohydr. Chem. Biochem., 1983,^7,209-374. [132] Olden, K.; Parent, J.B.; White, S.L.; Biochim. Biophys. Acta, 1982, 650, 209-232. [133] Takayanagi, T.; Kimura, A.; Chiba, S.; Ajisaka, K.; Carbohydr. Res., 1994, 256, 149-158. [134] Ciucanu, I.; Kerek, F.; Carbohydr. Res., 1984, 131, 209-217. [135] Stenutz, R.; Jansson, P-E.; Widmalm, G.; Carbohydr. Res., 1998, 306, 11-17. [136] Beyer, T.A.; Sadler, J.E.; Rearick, J.I.; Paulson, J.C; Hill, R.L.; Adv. Enzymoi, 1981,52,23-175. [137] Tomiya, N.; Awaya, J.; Kurono, M.; Endo, S.; Arata, Y.; Takahashi, N.; Anal. Biochem., 1988, 77/, 73-90. [138] Hasui, Y.; Fukui, Y.; Kikuchi, J.; Kato, N.; Miyairi, K.; Okuno, T.; Biosci. Biotech. Biochem., 1998, 62, 852-857. [139] Towatari, T.; Miyamura, T.; Kondo, A.; Kato, I.; Inoue, M.; Yano, M.; Kido, H.; Eur. J. Biochem., 1998,255, 163-169. [140] Yang, B.Y.; Gray, J.S.S.; Montgomery, R.; Carbohydr. Res., 1996, 287, 203-212. [141] Zhao, K-W.; Faull, K.F.; Kakkis, E.D.; Neufeld, E.F.; J. Biol. Chem., 1997, 272, 22758-22765. [142] Stimson, E.; Hope, J.; Chong, A.; Burlingame, A.L.; Biochemistry, 1999, 38, 4885-4895. [143] Hille-Rehfeld, A.; Biochim. Biophys. Acta, 1995, 72^7, 177-194. [144] Noorman, F.; Barrett-Bergshoeff, M.M.; Rijken, D.C.; Eur. J. Biochem., 1998, 257,107-113. [145] Nadimpalli, S.K.; Hille-Rehfeld, A.; von Figura, K.; Comp. Biochem. Physiol., 1997, 118B, 805-809. [146] Sohar, I.; Sleat, D.; Liu, C-G.; Ludwig, T.; Lobel, P.; Biochem. J., 1998, 330, 903-908. [147] Swanson, A.F.; Ezekowitz, R.A.B.; Lee, A.; Kuo, C-C; Infect. Immun., 1998, 66, 1607-1612. [148] Shepherd, V.L.; Lane, K.B.; Abdolrasulnia, R.; Arch. Biochem. Biophys., 1997, 344, 350-356. [149] Riccobono F.; Fiani, M.L,; Carbohydr. Res., 1996, 282, 285-292. [150] Hitchen, P.G.; Mullin, N.P.; Taylor,M.E. Biochem. 1, 1998, 333, 601-608. [151] Wilson, R.; Chen, C ; Ratcliffe, N.A.; J. Immunol., 1999, 162, 1590-1596. [152] Kennedy, J.F.; Palva, P.M.G.; Corella, M.T.S., Cavalcanti, M.S.M.; Coelho, L.C.B.B.; Carbohydr. Polym., 1995, 26, 219-230.

1168 [153] Swaminathan, C.P.; Surolia, N.; Surolia, A.; J. Am. Chem. Soc, 1998, 120, 51535159. [154] Ng, K.K-S.; Park-Snyder, S.; Weis, W.I.; Biochemistry, 1998, 57, 17965-17976. [155] Debray, H.; Wieruszeski, J-M.; Strecker, G.; Franz, H.; Carbohydr. Res., 1992, 236, 135-143. [156] Loris, R.; Maes, D.; Poortmans, F.; Wyns, L.; Bouckaert, J.; J. Biol. Chem., 1996, 277,30614-30618. [157] Saito, K.; Komae, A.; Kakuta, M.; Van Damme, E.J.M.; Peumans, W.J.; Goldstein, I.J.; Misaki, A.; Eur. J. Biochem., 1993, 217, 677-681. [158] Misaki, A.; Kakuta, M.; Meah, Y.; Goldstein, I.J.; J. Biol. Chem., 1997, 272, 25455-25461. [159] Dam, T.K.; Bachhawat, K.; Rani, P.G.; Surolia, A.; J. Biol. Chem., 1998, 273, 5528-5535. [160] Goswami, S.; Sarkar, S.; Basu, J.; Kundu, M.; Chakrabarti, P.; FEBSLett., 1994, 355, 183-186. [161] Sokurenko, E.V; Chesnokova, V.; Doyle, R.J.; Hasty, D.L.; J. Biol. Chem., 1997, 272, 17880-17886. [162] Fabre, C; Causse,H.; Mourey, L.; Koninkx, J.; Riviere, M.; Hendriks, H.; Puzo, G.; Samama, J-P.; Rouge, P.; Biochem. J., 1998, 329, 551-560. [163] Frankel, A.E.; Fu, T.; Burbage, C; Tagge, E.; Harris, B.; Vesely, J.; Willingham,M.C.; Carbohydr. Res., 1997, 300, 251-258. [164] French, A.D.; Brady, J.W., Ed.; Computer Modeling of Carbohydrate Molecules, American Chemical Society: Washington, 1990. [165] Biswas, M.; Sekharudu, Y.C.; Rao, V.S.R.; Carbohydr. Res., 1987, 160, 151-170. [166] Paulsen, H.; Peters, T.; Sinnwell, V.; Heume, M.; Meyer, B.; Carbohydr. Res., 1986, 156, 87-106. [167] Gumming, D.A.; Shah, R.N.; Krepinsky, J.J.; Grey, A.A.; Carver, J.P.; Biochemistry, 1987, 26, 6655-6663. [168] Gumming, D.A.; Carver, J.P.; Biochemistry, 1987, 26, 6664-6676; 6676-6683. [169] Homans, S.W.; Pastore, A.; Dwek, R.A.; Rademacher, T.W.; Biochemistry, 1987, 26, 6649-6655. [170] Rutherford, T.J.; Homans, S.W.; Biochemistry, 1994, 33, 9606-9614. [171] Mukhopadhyay, C; Biopolymers, 1998, 45, 177-190. [172] Wu, P.; Rice, K.G.; Brand, L.; Lee, Y.C.; Proc. Natl. Acad Sci. USA, 1991, 88, 9355-9359. [173] Wyss, D.F.; Choi, J.S.; Li, J.; Knoppers, M.H.; Willis, K.J.; Arulanandam, A.R.N.; Smolyar, A.; Reinherz, E.L.; Wagner, G.; Science, 1995, 269, 12731278. [174] Rudd, P.M.; Joao, H.C.; Coghill, E.; Fiten, P.; Saunders, M.R.; Opdenakker, G.; Dwek, R.A.; Biochemistry, 1994, 33, 17-22. [175] Priem, B.; Gross, K.C.; Plant Physiol., 1992, 98, 399-401. [176] Percival, E.; McDowell, R.H. in Encyclopedia of Plant Physiology, Tanner, W.; Loewus, F.A., Ed.; Springer-Verlag: Berlin, 1981; Vol. 13B, pp. 277-316. [177] Kolender, A.A.; Pujol, C.A.; Damonte, E.B.; Matulewicz, M.C.; Cerezo, A.S.; Carbohydr. Res., 1997, 304, 53-60. [178] T\mt\\,T.E.',Adv. Carbohydr. Chem. Biochem., 1965,20,410-483; 1964, 19, 247-312. [179] Dea, I.C.M.; Morrison, A.; Adv. Carbohydr. Chem. Biochem., 1975, 31, 241-312.

1169

[180] Dey, P.M.; Adv. Carbohydr. Chem. Biochem., 1978, 35, 341-376. [181] Meier, H.; Reid, J.S.G. In Encyclopedia of Plant Physiology; Loewus, F.A.; Tanner, W., Ed.; Springer-Verlag: Berlin, 1982; Vol. ISA, pp. 418-471. [182] Matheson, N.K. In Seed Physiology, Murray, D.R., Ed.; Academic Press: Sydney, 1984; vol. / , p p . 167-208. [183] Matheson, N.K. In Methods in Plant Biochemistry, Dey, P.M., Ed.; Academic Press: London, 1990, Vol. 2, pp. 371-413. [184] Mendonga-Previato, L.; Gorin, P.A.J.; Previato, J.O.; Biochemistry, 1979, 18, 149-154. [185] Kaihou, S.; Hayashi, T.; Otsuru, O.; Maeda, M.; Carbohydr. Res., 1993, 240, 207-218. [186] Sandford, P.A.; Baird, J. In The Polysaccharides', Aspinall, C O . , Ed.; Academic Press: New York, 1983; Vol. 2, pp. 411-490. [187] Buckeridge, M.S.; Panegassi, V.R.; Rocha, D.C.; Dietrich, S.M.C.; Phytochemistry, 1995, 38, 871-875. [188] McCleary, B.V.; Methods Enzymol, 1988, 160, 596-610. [189] McCleary, B.V.; Matheson, N.K.; Carbohydr. Res., 1983, 119, 191-219. [190] Edwards, M.; Bulpin, P.V.; Dea, I.C.M.; Reid, J.S.G.; Planta, 1989, 178, 41-51. [191] Edwards, M.; Scott, C ; Gidley, M.J.; Reid, J.S.G.; Planta, 1992, 187, 61-1 A. [192] Reid, J.S.G.; Edwards, M.; Gidley, M.J.; Clark, A.H.; Planta, 1995, 195, 489495. [193] Chandrasekaran, R.; Bian, W.; Okuyama, K.; Carbohydr. Res., 1998, 572, 219224. [194] Hayashi, T.; Matsuda, K.; J. Biol. Chem., 1981, 256, 11117-11122. [195] McCleary, B.V.; Matheson, N.K.; Small, D.M.; Phytochemistry, 1976, 75, 11111117. [196] McCleary, B.V.; Clark, A.H.; Dea, I.C.M.; Rees, D.A.; Carbohydrate Res., 1985, 139, l-M'ieQ. [197] Dea, I.C.M.; Clark, A.H.; McCleary, B.V.; Carbohydr. Res., 1986, 147, 275-294. [198] Mallett, I.; McCleary, B.V.; Matheson, N.K.; Phytochemistry, 1987, 26, 18891894. [199] Aspinall, G.O. In The Biochemistry of Plants; Preiss, J., Ed.; Academic Press: New York, 1980, Vol. 3, pp. 473-500. [200] Aspinall, G.O.; Adv. Carbohydr. Chem. Biochem., 1959, 14, 429-468. [201] Kato, K. In Encyclopedia of Plant Physiology, Tanner, W.; Loewus, F.A., Ed.; Springer-Verlag: Berlin, 1981; vol. 13B, pp. 29-46. [202] Bewley, J.D.; Reid, J.S.G. In Biochemistry of Storage Polysaccharides in Green Plants, Dey, P.M.; Dixon, R.A., Ed.; Academic Press: London, 1985; pp. 289304. [203] Bacic, A.; Harris, P.J.; Stone, B.A. In The Biochemistry of Plants; Preiss, J., Ed.; Academic Press: San Diego, 1988, Vol. 14, pp. 297-371. [204] Eda, S.; Akiyama, Y.; Kato, K.; Takahashi, R.; Kusakabe, I.; Ishizu, A.; Nakano,; Carbohydr. Res., 1984, 757, 105-118. [205] Eda, S.; Akiyama, Y.; Kato, K.; Ishizu, A.; Nakano, J.; Carbohydr. Res., 1985, 757, 173-181. [206] Fischer, M.; Wegryzn, T.F.; Hallett, I.C; Redgwell, R.J.; Carbohydr Res., 1996, 295, 195-208. [207] Sims, I.M.; Craik, D.J.; Bacic, A.; Carbohydr. Res.; 1997, 303, 79-92.

1170

[208] Rees, D.A.; Morris, E.R.; Thorn, D.; Madden, J.K.; In The Polysaccharides', Aspinall, CO., Ed.; Academic Press: New York, 1982; Vol. 7, pp. 195-290. [209] Yalpani, M.; Polysaccharides: Syntheses, Modifications and Structure/ Property Relations, Elsevier: Amsterdam, 1988. [210] Gidley, M.J.; Robinson, G.; In Methods in Plant Biochemistry, Dey, P.M., Ed.; Academic Press: London; 1990; Vol. 2, pp. 607-642. [211] Atkins, E.D.T.; Farnell, S.; Burden, C; Mackie, W.; Sheldrick, B.; Biopolymers, 1988,27, 1097-1105. [212] Chandrasekaran, R.; Adv. Carbohydr. Chem. Biochem., 1997, 52, 311-439. [213] Chanzy, H.; Perez, S.; Miller, D.P.; Paradossi, G.; Winter, W.T.; Macromolecules, 1987, 20, 2407-2413. [214] Goycoolea, P.M.; Morris, E.R.; Gidley, M.J.; Carbohydr Polym., 1995, 27, 6971. [215] Kapoor, V.P.; Milas, M.; Taravel, F.R.; Rinaudo, M.; Carbohydr. Polym., 1994, 25, 79-84. [216] Tao, L.; Matheson, N.K.; Carbohydr. Polym., 1993, 20, 269-277. [217] McCIeary, B.V.; Matheson, N.K.; Phytochemistry, 1975, 14, 1187-1194. [218] Seifert, A.; Heinevetter, L.; Colfen, H.; Harding, S.E.; Carbohydr. Polym., 1995, 28, 325-332. [219] Painter, T.; Larsen, B.; Acta Chem. Scand, 1970, 24, 2724-2736. [220] Sundararajan, P.R.; Rao, V.S.R.; Biopolymers, 1970, 9, 1239-1247. [221] Petkowicz, C.L.O., Reicher, P.; Mazeau, K.; Carbohydr. Polym., 1998, 37, 25-39. [222] Bergamini, J-F.; Boisset, C; Mazeau, K.; Heyraud, A.; Taravel, F.R.; New J. Chem., 1995, 19, 115-127. [223] Sellen, D.B.; Symp. Soc. Exp. Biol., 1980, 5^, 315-329. [224] McCIeary, B.V.; Carbohydr. Res., 1979, 71, 205-230. [225] Brownsey, G.J.; Cairns, P.; Miles, M.J.; Morris, V.J.; Carbohydr. Res., 1988, 77^,329-334. [226] Turquois, T.; Taravel, F-R.; Rochas, C; Carbohydr. Res., 1993, 238, 27-38. [227] Goycoolea, P.M.; Richardson, R.K.; Morris, E.R.; Gidley, M.J.; Biopolymers, 1995,5(5,643-658. [228] Bresolin, T.M.B.; Sander, P.C.; Reicher, P.; Sierakowski, M.R., Rinaudo, M.; Ganter, J.L.M.S.; Carbohydr. Polym., 1997, 33, 131-138. [229] Lundin, L.; Hermansson, A-M.; Carbohydr. Polym., 1995, 26, 129-140; 28, 9199. [230] Cairns, P.; Miles, M.J.; Morris, V.J,; Brownsey, G.J.; Carbohydr. Res., 1987, 75^,411-423. [231] Chandrasekaran, R.; Radha, A.; Carbohydr. Polym., 1997, 32,201-208. [232] Piculell, L.; Zhang, W.; Turquois, T.; Rochas, C; Taravel, F-R.; Williams, P.A.; Carbohydr. Res., 1994, 265, 281-290. [233] Williams, P.A.; Clegg, S.M.; Langdon, M.J.; Nishinari, K.; Piculell, L.; Macromolecules, 1993, 26, 5441-5446. [234] Turquois, T.; Rochas, C; Taravel, F-R.; Tvaroska, I.; J. Mol. Recognition, 1994, 7, 243-250. [235] Turquois, T.; Rochas, C; Taravel, F-R.; Doublier, J.L.; Axelos, M-A-V.; Biopolymers, 1995, 36, 559-567. [236] Schorsch, C; Gamier, C; Doublier, J-L.; Carbohydr. Polym., 1995, 28, 319-323.

1171

[237] Parker, A.; Leiimousin, D.; Miniou, C; Boulenguer, P.; Carbohydr. Res., 1995, 272,91-96. [238] Ridout, M.J.; Cairns, P.; Brownsey, G.J.; Morris, V.J.; Carbohydr. Res., 1998, 309, 375-379. [239] Newman, R.H.; Hemmingson, J.A.; Carbohydr. Polym. 1998, 36, \61A12, [240] Mishima, T.; Hisamatsu, M.; York, W.S.; Teranishi, K.; Yamada, T.; Carbohydr. Res., 1998,56/(5,389-395. [241] Whitney, S.E.C.; Brigham, J.E.; Darke, A.H.; Reid, J.S.G.; Gidley, M.J.; Carbohydr. Res., 1998, 307, 299-309. [242] Percival, E.; McDowell, R.H. In Methods in Plant Biochemistry, Dey, P.M., Ed.; Academic Press: London, 1990; Vol. 2, pp. 523-547. [243] Shankar, S.; Ye, R.W.; Schlictman, D.; Chakrabarty, A.M.; Adv. Enzymol, 1995, 76/, 221-255. [244] Skjak-Br^k, G.; Biochem. Soc. Trans., 1992, 20, ll-^'i. [245] 0stgaard, K.; Larsen, B.; Carbohydr. Res., 1993, 246, 229-241. [246] Heyraud, A.; Colin-Morel, P.; Girond, S.; Richard, C; Kloareg, B.; Carbohydr. Res., 1996, 291, 115-126. [247] Ochi, Y.; Takeuchi, T.; Murata, K.; Kawabata, Y.; Kusakabe, I.; Biosci. Biotech. Biochem., 1995, 59, 1560-1561. [248] Heyraud, A.; Colin-Morel, P.; Gey, C; Chavagnat, F.; Guinard, M.; Wallach, J.; Carbohydr. Res., 1998, 308, 417-422. [249] Panikkar, R.; Brasch, D.J.; Carbohydr, Res., 1997, 300, 229-238. [250] Draget, K.I.; Skj^k-Brask, G.; Smidsrod, O.; Carbohydr. Polym., 1994, 25, 31-38. [251] Miller, I.J.; Phytochemistry, 1996, 41, 1315-1317. [252] Simpson, J.A.; Smith, S.E.; Dean, R.T.; Biochem. Mol. Biol. Inter., 1993, 30, 1021-1034. [253] Franklin, M.J.; Chitnis, C.E.; Gacesa, P.; Sonesson, A.; White, D.C.; Ohman, D.E.; J. BacterioL, 1994, 176, 1821-1830. [254] Quillet, M.; de Lestang-Bremond, G.; Phytochemistry, 1985, 24, 43-45. [255] H0idal, H.K.; Ertesv^g, H.; Skj^k-Brask, G.; Stokke, B.T.; Valla, S.; J. Biol. Chem., 1999, 274, 12316-12322. [256] Shibaev, V.N.; Adv. Carbohydr. Chem. Biochem., 1986, 44, 277-339. [257] Poveda, A.; Santamaria, M.; Bernabe, M.; Prieto, A.; Bruix, M.; Corzo, J.; y\m€v\QZ-BdivhQro,].', Carbohydr. Res., \991, 304,209-2X1. [258] Brennan, P.J.; Nikaido, H.; Ann. Rev. Biochem., 1995, 64, 29-63. [259] Chatterjee, D.; Lowell, K.; Rivoire, B.; McNeil, M.R.; Brennan, P.J.; J. Biol Chem., 1992, 267, 6234-6239. [260] Bishop, C.T.; Jennings, H.J. In The Polysaccharides; Aspinall, G.O., Ed.; Academic Press: New York, 1982; Vol. 1, pp. 291-330. [261] Rahman, M.M.; Guard-Petter, J.; Asokan, K.; Hughes, C; Carlson, R.W.; J. Biol. Chem., 1999,27^,22993-22998. [262] Matheson, N.K.; McCleary, B.V. \x\ The Polysaccharides', As^m?i\\,G.O., Ed.; Academic Press: Orlando, Florida, 1985, Vol. 3, pp. 1-195. [263] Sutherland, I.W.; Carbohydr. Polym., 1999, 55, 319-328. [264] Aspinall, G.O.; Adv. Carbohydr. Chem. Biochem., 1969, 24, 333-379. [265] Stephen, A.M. In The Polysaccharides', Aspinall, G.O., Ed.; Academic Press: New York, 1983; Vol. 2, pp. 97-193.

1172

[266] Stephen, A.M. In Methods in Plant Biochemistry; Dey, P.M., Ed.; Academic Press: London, 1990; Vol. 2, pp. 483-522. [267] Smith, F.; Montgomery, R.; The Chemistry of Plant Gums and Mucilages, van Nostrand-Reinhold: Princeton, New Jersey, 1959. [268] Yates, E.A.; Knox, J.P.; Carbohydr Polym., 1994, 24, 281-286. [269] Di Fabio, J.L.; Dutton, G.G.S.; Moyna, P.; Carbohydr. Res., 1982, 99, 41-50. [270] Churms, S.C; Stephen, A.M.; Carbohydr. Res., 1987,167,239-255. [271] De Luca, P.; Moretti, A.; Sabato, S.; Gigliano, G.S.; Phytochemistry, 1982, 21, 1609-1611. [272] Vogt, D.C.; Stephen, A.M.; Carbohydr. Res., 1993, 238, 249-260. [273] Vogt, D.C.; Stephen, A.M.; Carbohydr. Res., 1993, 241,211-226. [274] Barone, G.; Corsaro, M.M.; De Castro, C.; Lanzetta, R.; Mangoni, L.; Parrilii, M.; Carbohydr Res., 1994, 260, 259-270. [275] Whistler, R.L.; Smart, C.L.; Polysaccharide Chemistry, Academic Press: New York, 1953, pp 327-334. [276] Gowda, D.C.; Reuter, G.; Schauer, R.; Carbohydr Res., 1983, 113, 113-124. [277] Akiyama, Y.; Eda, S.; Mori, M.; KatO, K.; Agric. Biol. Chem., 1984, 48, 403-407. [278] Aspinall, G.O.; Khondo, L.; Puvanesarajah, V.; Carbohydr Res., 1989, 188, 113120. [279] Redgwell, R.J.; O'Neill, M.A.; Selvendran, R.R.; Parsley, K.J.; Carbohydr Res., 1986,755,97-106; 107-118. [280] Barone, G.; Corsaro, M.M.; Gianattasio, M.; Lanzetta, R.; Moscariello, M.; Parrilii, M.; Carbohydr Res., 1996, 280, 111-119. [281] Schuerch, C; Adv. Carbohydr Chem. Biochem., 1981, 39, 157-212. [282] Kochetkov, N.K.; Tetrahedron, 1987, 43, 2389-2436. [283] Lee, Y.C.; Lee, R.T., Ed. Methods Enzymol, 1994, Vol 242 and 247. [284] Flowers, H.M.; Methods Enzymol., 1987, 138, 359-404. [285] Frechet, J.; Schuerch, C; J Am. Chem. Soc, 1969, 91, 1161-1164. [286] Ota, N.; Hatanaka, K.; Macromolecules, 1998, 57, 2123-2128. [287] Kong, F.; Schuerch, C; Macromolecules, 1984, 77, 983-989. [288] Yamaguchi, H.; Schuerch, C; Biopolymers, 1980, 19, 297-309. [289] Kobayashi, K.; Nomura, K.; Okada, M.; Oshima, R.; Carbohydr. Res., 1993, 242, 161-166 [290] Matsuzaki, K.; Yamamoto, I.; Sato, T.; Makromol. Chem., 1986, 187, 325-331. [291] Kotchetkov, N.K.; Byramova, N.E.; Tsvetkov, Y.E.; Backinowsky, L.V.; Tetrahedron, 1985, 41, 3363-3375. [292] Kochetkov, N.K.; Betaneli, V.I.; Ovchinnikov, M.V.; Backinowsky, L.V.; Tetrahedron, Suppl.l, 1981, 37, 149-156. [293] Hallgren, C; Hindsgaul, O.; Carbohydr Res., 1994, 260, 63-71. [294] Becker, B.; Furneaux, R.H.; Reck, F.; Zubkov, O.A.; Carbohydr Res., 1999, 575, 148-158. [295] Bajza, I.; K5ver, K.E.; Liptak, A.; Carbohydr Res., 1998, 308, 247-258. [296] Guo, Z-W.; Ito, Y.; Nakahara, Y.; Ogawa, T.; Carbohydr Res., 1998, 306, 539544. [297] Zhao, Y.; Thorson, J.S.; Carbohydr. Res., 1999, 57P, 184-191. [298] Nakahara, Y.; Shibayama, S.; Nakahara, Y.; Ogawa, T.; Carbohydr Res., 1996, 280, 67-84.

1173

[299] Grice, P.; Ley, S.V.; Pietruszka, J.; Osborn, H.M.I.; Priepke, H.W.M.; Warriner, S.L.; Chem. Eur. J., 1997, J, 431-440.

This Page Intentionally Left Blank

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

1175

THE CHEMISTRY OF LITHISTID SPONGE: A SPECTACULAR SOURCE OF NEW METABOLITES M. VALERIA D'AURIA, ANGELA ZAMPELLA AND FRANCO ZOLLO* Dipartimento di Chimica delle Sostanze NaturalU Via D. Montesano 49 80131 Napolh Italy ABSTRACT: Among marine invertebrates, sponges belonging to the lithistid order are certainly the most spectacular source of many classes of compounds, some of them showing interesting pharmacological activities. The extreme diversity of lithistid sponge metabolites, macrolides and polyketides, cyclic and linear peptides, alkaloids, pigments and novel sterols, as well as their structural similarity to those reported from microorganisms, gave well founded the hypothesis that they may be produced by symbiotic bacteria and miaoalgae. The purpose of this review is to offer a comprensive survey of the chemistry of lithistid sponges with a particular emphasis to the pharmacological properties and to the mechanisms of action of their metabolites, and to report the total syntheses of representative macrolides and peptides isolated from the sponges belonging to this so fascinating order.

INTRODUCTION Some of the most exciting marine metabolites from a structural point of view have been isolated in recent years from sponges of order Lithistida: Theonella, Discodermia, Jereicopsis, Neosiphonia, Reidispongia, Callipelta, Corallistes, Aciculites and Microscleroderma. These natural products, encompassing a wide range of molecular diversity, such as linear and cyclic peptides, macrolides, lipids, alkaloids and steroids, possess a variety of pharmacological properties, including insecticide, antimicrobial, antiviral, antitumor and tumour promotive, antiinflammatory and immunosuppressive. Some of them represent lead compounds in drug design and in the study of human biochemical pathways. The presence of an extreme diversity of bioactive metabolites appears surprising taking into account that lithistid sponges have not only hard bodies, but also a shield of fused siliceous spicules which offers good physical protection against predators, and so they don't need chemical defence. The extreme diversity of Uthistid sponge metabolites as well as the resemblance between these metabolites and those isolated from micro-organisms made well founded the hypothesis of their symbiotic origin. To date, two excellent reviews have been published on the chemical composition of lithistid sponges [1,2], The 1993 review "Bioactive

1176 Sponge peptides" by Fusetani and Matsunaga contains a detailed description of the peptide derivatives isolated from lithistid sponges whereas the recent review "Lithistid sponges: star performers or hosts to stars" by Bewley and Faulkner reports a summary of the chemistry of lithistid sponges, with a particular emphasis to the relationship between these sponges and their symbionts. Our review constrains its scope to a more thorough discussion of the chemistry, total synthesis and biological activity of secondary metabolites from lithistid sponges. The material is organised in two main sections. A first section is devoted to the structure and the chemical distribution of the secondary metaboUtes isolated from each genus belonging to the Uthistid order, A second section will focus on the total synthesis of representative metaboUtes from Uthistid sponges. CHEMICAL STRUCTURES In this section the secondary metaboUtes from Uthistid sponges are described foUowing the taxonomic classification of extant Uthistid sponges reported by Faulkner [2]. Metabolites from Theonella sp. and Theonella swinhoei Lithistid sponges identified as Theonella sp. are often superficiaUy similar to Theonella swinhoei except for the orange colour of their interior tissue due to the presence of aurantoside A (1) and related polyenes.

loH

OH

O/,

't.

Aurantoside A (1)

'y—o OH

Sterols The pioneering paper on metaboUtes from sponges of Theonella genus appeared in Uterature in 1981 deaUng with the isolation of two new and unusual sterols with the unprecedented 4-methylene nucleus, conicasterol (2) and theoneUasterol (3) [3]. Subsequently, the 3-keto derivatives, conicasterone (4) and theoneUasterone (5), together with a Diels Alder type dimeric steroid, named bistheoneUasterone (6), were reported from the Okinawan marine sponge Theonella swinhoei [4]. It is noteworthy that

1177

Conicasterone (4)

o

Theonellasterone (5)

.-

Bistheonellasterone (6)

Bisconicasterone (7)

5 and 6 were seen under an optical microscope as crystals deposited in the tissue of fresh marine sponge. In 1994 bisconicasterone (7) was isolated from a Theonella swinhoei specimen collected off Hachijo-jima Island and its absolute stereochemistry determined by X-ray crystallography [5]. It is intriguing that no trace of theonellasterone (5) and its bis-derivative 6 has been found in this collection of Theonella swinhoei in contrast with

1178

the finding of Kitagawa who reported theonellasterone (5) as a major component of the same species collected at Okinawa [4]. This suggests a site dependence of the ingredients of marine organisms belonging to the same species. A further investigation of this specimen of Theonella swinhoei afforded in the isolation of seven oxygenated conicasterol derivatives along with the unusual sterol 8 with 7,8-epoxy functionaUty [6].

HO

R= f Swinhosterol A (9) R= I Swinhosterol B (10)

^0CH3

Swinhosterol C (11)

During the investigation of an Okinawan T. swinhoei, it was found that the apolar extract contained three new steroids, swinhosterols A-C (9-11) [7] which structures were determined by spectroscopic investigation. Compounds 9 and 10 combine the unusual 4-methylene and seco features. Cyclopeptides Small cyclopeptides, particularly rich in non-conventional aminoacids, are very promising and interesting to medicinal chemistry because of their unique features. In fact the presence of non-ribosomal aminoacids, including D-series and N-alkylated versions of the natural ones as well as strongly structural modified aminoacids, produces pronounced effects on the pharmacological activities of these metabolites. For example, Nalkylation and the lack of N and C termini enhance the hydrophobicity of depsipeptides, thus determining a more facile crossing of biological membranes and an improved stability to enzymatic degradation. In

1179 addition N-alkyl residues exhibit reduced preferences to form the trans conformation of the amide bond leading to biologically important P-tum structures. Cyclization reduces peptide conformational flexibility that results in a higher receptor binding affmity and often offers the possibility to determine their three dimensional structures. All these features make cyclopeptides promising lead compounds in drug discovery. In 1990 Fusetani et al. isolated from different collections of the same

A^C}~°"

HN—C

NH

>=0

°

)—NH

y

I

Cyclotheonamide A (12) R=H Cyclotheonamide B (13) R=Me

Specimens of a Theonella sponge collected off Hachijo-jima Island, a mixture of aurantoside A (1) and B together with an important family of cyclic peptides, named cyclotheonamides [8]. Cyclotheonamide A (12) is a cycUc pentapeptide containing the vinylogous tyrosine (V-Tyr), the a-

OH

V

NHCHO

HgN

^N

H2N NH

Cyclotheonamide C (14)

NH

Cyclotheonamide D (15)

%HCHO

1180

ketohomoarginine (k-Arg) and a diaminopropionic acid as unusual aminoacid residues. This peptide is a potent serine protease (thrombin and trypsin) inhibitor and its mechanism of action has been clarified by X-ray study of the complex with human a-trombin. This study disclosed that cyclotheonamide A (12) binds to the catalytic triad of the enzyme by forming a covalent bond between its a-keto group and the hydroxyl group of a serine residue of one of the triad residues [9]. Cyclotheonamide B (13) is the acetamido derivative of 12 whereas cyclotheonamide C (14) [10] possesses an additional insaturation in the V-Tyr residue. The aminoacid constitution of cyclotheonamide D (15) [10] is quite different from cyclotheonamide A (12) with a Leu residue in place of D-Phe. From a morphologically different sample of Theonella swinhoei, cyclotheonamide E (16) has been isolated [10]. This latter includes D-Ile and a phenylacetyl-Ala amide groups in place of D-Phe and formyl groups in 12. Recently, a recollection of Theonella sp. off Tanegashima Island, afforded cyclotheonamides E2 (17) and E3 (18) [11], closely related to 16 from which they differ in the N-acyl group of the alanyl side chain. These structural modifications affect the biological activity. Infact the side chain modified cyclotheonamides 16, 17 and 18 showed an enhanced specificity toward thrombin due to the bulky acylated alanyl R

Y^

Cyclotheonamide E (16)

Cyclotheonamide E2 (17)

HgN

IT

rr

Cyclotheonamide E3 (18)

NH

residue that increases the hydrophobic interaction with the enzyme. On the other hand the properties of the aminoacid residues present in cyclotheonamides A-D (12-15) gave more favourable interactions with trypsin than trombin [12]. Immediately after the publication of the first paper on the isolation of cyclotheonamides A (12) and B, these compounds attracted considerable synthetic interest. The first total synthesis of cyclotheonamide B (13) was reported by Schreiber and Hagihara in 1992 and led to a stereochemical reassignment of the stereocenters in V-Tyr and k-Arg residues for which

1181 the (S) configuration was assigned [13]. Further total syntheses of cyclotheonamides were reported by Wipf [14], Shioiri [15], Maryanoff [16], and Ottenheijm [17]. Recent reviews provide a comprensive survey on the syntheses of cyclotheonamides in particular [18] and on the coupling methods for the incorporation of noncoded aminoacids in peptides in general [19] thus these topics will not be discussed in the Total Syntheses Section. The presence of a-ketoaminoacids appears to be the main feature of the peptides isolated from sponges of the genus Theonella. The cytotoxic orbiculamide A (19) (IC50 4.7 |ig/ml), isolated from the same marine sponge, Theonella sp., that contained cyclotheonamides, represents another example of cyclic peptide containing three new aminoacid residues, theoleucine, theonsJanine and 2-bromo-5-hydroxytryptophan [20].

H

11

L

'^'^V V

NH

Orbiculamide A (19)

From a structural point of view, keramamides B-H (20-26) and J (27), a series of novel cyclopeptides isolated from Okinawan marine sponges Theonella sp. [21-23], are closely related to orbiculamide A (19). Keramamides B-E (20-23) [21,22] are composed by a cyclic moiety and a side chain. The cycUc hexapeptide portion contained three modified aminoacid residues linked each other to form a segment containing an a,P-unsaturated amide group conjugated with an oxazole as well as a aketo-P-aminoacid moiety constituting a homoleucine-like unit. In keramamides F-G (24-25) [22,23] and H-J (26-27) [23] a residue of Omethylseryl thiazole replaces the alanyl (or homoalanyl) oxazole present in keramamide B-E (20-23). Keramamides B-D (20-22) inhibited the superoxide generation induced by the chemiotactic peptide, N-formylMet-Leu-Phe whereas keramamides E (23) and F (24) showed cytotoxicity against human epidermoid carcinoma cells (IC50 of 1.6 and 1.4 |Ltg/ml, respectively) and against murine lymphoma L1210 with IC50 values of 1.55 and 2.0 |LLg/ml respectively. Keramamides G-J (25-27)

1182 showed weak cytotoxicity (IC50 >10 |xg/ml). Recently, the total synthesis of a molecule possessing the structure proposed for keramamide J (27) OH =

H

Keramamide B (20) Keramamide C (21) Keramamide D (22) Keramamide E (23)

n

-^

^

HN

R,

Rj=CrT2CH3 R2=Crl2CH3 Rj=CH2CH3 R2=CH2CH3 R2=Crl3 R2=CH2Cri3 Rj=Cri3 R2=CH2Crl2Cri3

was reported [24]. Spectroscopic data indicated that the synthetic compound is not the same compound reported by Kobayashi and coworkers [23]. A detailed analysis of ^H and ^^C NMR data would suggest that the two compounds are epimeric at the homoleucine chiral Ca.

.0CH3

"

Keramamide F (24) *S Keramamide G (25) *R

Keramamide A (29) was quite different in the aminoacid constituents and its most important feature was the presence of an ureido bond consisting of the a-NH of the lysine and the a-NH of the phenylalanine

1183 residues [25]. Keramamide A (29) showed no cytotoxicity against various human and murine carcinoma cells whereas exhibited inhibitor activity

.0CH3

Keramamide H (26) Keramamide J (27) Keramamide K (28)

Ri=Br, R2=0H, R3=H Ri=R2=R3=H Ri=R2=H, R3=CH3

against sarcoplasmic reticulum Ca^"^"ATPase (IC50 3 x 10"^ mol dm'^). Further investigation on the extracts of the Okinawan Theonella sponge resulted in the isolation of two new keramamide derivatives, K (28) and L (30) [26]. Keramamide K (28), closely related to 27, is a new

Me

COOH

Keramamide A (29) R=OH Keramamide L (30) R=H

thiazole containing cyclic peptide with the rare 1-Me-Trp residue while keramamide L (30) is the first peptide with a MeCtrp residue from natural origin. These compounds are cytotoxic in vitro against L1210 murine leukaemia cells (IC50 0.72 and 0.46 ^g/ml, respectively) and KB cells (IC50 0.42 and 0.90 |Lig/ml, respectively).

1184

Other cyclic peptides including konbamide (31), from the keramamide derivatives containing Theonella[21], and mozamides A (32) and B (33), from a Theonellid collected in the southern Mozambique [28], incorporate an ureido linkage. In the cyclic hexapeptides 32 and 33 an

Me

X 'NH

COOH

Konbamide (31)

isoleucine residue is joined to the a-amino group of a lysine giving the ureido moiety. In the konbamide's side chain the leucine takes isoleucine's place. Although keramamide A (29), mozamides (32-33) and konbamide (31) are structurally related, only 31 showed a calmodulin antagonistic activity in mammalian cells. Me I

COOH

Mozamide A (32) R=H Mozamide B (33) R=Me

The cytotoxic cyclic peptide, oriamide (34), containing the new 4propenoyl-2-tyrosylthiazole unit, has been isolated from the blue marine sponge, Theonella sp., collected in Sodwana Bay [29]. This unusual aminoacid is reminiscent of the thiazole moiety early reported as a constituent of keramamide F (24) [22]. In 1989, Fusetani group reported the isolation and the complete structure of theonellamide F (35), a novel antifungal peptide, from a

1185 OH

OH

I

O

^

/

^

'^^^ NH

\ NaOgS—^

/

-OH

]\

Oriamide (34)

Theonella sp. collected off Hachijo-jima Island [30]. Theonellamide F (35) is a bicyclic peptide containing twelve aminoacids. In this cyclic peptide a dehydroalanine residue is masked by intramolecular Michael addition, giving rise to the elaborate and unprecedented histidino-alanine bridge. Further separation of the antifungal fraction of the sponge extract afforded five related peptides, theonellamides A-E [31]. Theonellamide A HO. HOgC HN

AAA A L " T H

/ ^OH

O

Theonellamide F (35) Ri=R2=H, R3=R4=Br Theonellamide B (36) R^OH, R2=Me, R3=Br, R4=H Theonellamide C (37) Ri=R2=R3=H, R4=Br

(38), B (36) and C (37) are closely related to 35, from which they differ in three aminoacid residues. Additionally, 38 bears a p-D-galactose linked to the free imidazole nitrogen. Theonellamide D (39) and E (40) are the P-Larabinoside and the P-D-galactoside of 35, respectively. Theonellamides A-F were moderately cytotoxic against P388 murine leukaemia cells with IC50 values of 5.0, 1.7, 2.5, 1.7, 0.9 and 2.7 |Lig/ml, respectively. Moreover, theonellamide F (35) inhibited growth of various pathogenic

1186 fiingi at concentrations of 3-12 |Lig/nil. Indeed theonegramide (41), isolated from a Theonella swinhoei collected at Antolang, Negros Island, the FiMppines, represents the first glycopeptide from lithistid sponge to be fiiUy characterized [32]. The structure of this antifungal bicyclic glycopeptide closely resembles those reported for 35. HO

0°-^' H

l/^OH

Theonellamide A (38) RpOH, R2=Me, R3=H Theonellamide D (39) Ri=R2=H, R3=Br Theonellamide E (40) Ri=R2=H, R3=Br O

HO.

o ^V^''"r^N

•02C

?

X J

NH

/ ^ * ^ \ ^

X=p.D-Gal X=P-L-Ara X=p-D-Gal

L ^

\-'^^

yj ,

°

H2NOC

Theonegramide (41) X=3-D-Ara Theopalauamide (42) X=p-D-Gal

During their studies on symbiotic filamentous eubacteria found in the interior tissue of Theonella swhinoei from Palau, Faulkner et al. isolated a bicycUc glycopeptide, theopalauamide (42) [33]. Its molecular formula, deduced by HR mass measurements, differs from that of theonegramide

1187 (41) by addition of CH2O in agreement with the replacement of the pentose sugar (D-arabinose) in 41 by a hexose in theopalauamide (42). The sugar was identified as galactose by NMR analysis on the natural product whereas its absolute stereochemistry as D was determined by analysis of the hydrolysis products by chiral GC-MS. However the structure determination of 42 was complicated for the presence of a minor peptide, isotheopalauamide later identified as a stable conformational isomer occurring during the acid isolation procedure [33]. Both 42 and its isomer inhibited the growth of Candida albicans with a 10 ^,g/disk and a 50 ^ig/disk dose, respectively. The butanolic extract of a sponge of the genus Theonella, collected off Perth, Australia, was found to inhibit the binding of the ligand [^^^I]IL-ip to the intact receptor, the EL4.6.1 cells. Thus, the bioassay-guided purification of the active components was undertaken and the perthamide B (43) was isolated [34]. Perthamide B is a cyclic octapeptide that showed moderate activity in the above binding assay but this activity

could not be differentiated from the toxicity of the compound. During a period of five years, Kitagawa group reported the isolation and the structural elucidation of five cyclic depsipeptides, the theonellapeptolides, from an Okinawan collection of Theonella swinhoei [35]. Theonellapeptolide la, lb, Ic, Id, le (44-48) are tridecapeptide lactones particularly rich in N-methyl and D-aminoacids. In these peptides the N-terminus is protected with a methoxyacetyl group and the C-terminus is connected through a lactone linkage to the p-hydroxyl group of a threonine residue. Among these five peptides, 47 has iontransport activities for Na"", K"", and Ca^"" ions, 48 exhibited iontransporting activities for Na"^ and K"^ ions across human erythrocyte

1188 membranes. Theonellapeptolides 45-48 showed moderate cytotoxicity against L1210 cells (IC50 16, 1.3, 2.4 and 1.4 |ig/ml, respectively). Following the characterisation of theonellapeptolide la-Ie (44-48), another new tridecapeptide lactone, theonellapeptolide lid (49), was isolated from the Okinawan sponge Theonella swhinoei. From NMR and mass evidences, the structure of 49 has been determined as an L-Ala

R4

TheonellapeptoUde la (44) RpOCHa, R2=H, R3=CH3, R4=H, R5=CH3, R6=CH3 TheonellapeptoUde lb (45) RPOCH3, R2=H, R3=H, R4=CH3, R5=CH3, R6=CH3 TheonellapeptoUde Ic (46) Ri=OCH3, R2=H, R3=CH3, R4=CH3, R5=CH3, R^=H TheonellapeptoUde Id (47) Ri=OCH3, R2=H, R3=CH3, R4=CH3, R5=CH3, R^=CH3 TheoneUapeptoUde le (48) Ri=OCH3, R2=CH3, R3=CH3, R4=CH3, R5=CH3, R^=CH3 TheoneUapeptoUde lid (49) RPOCH3, R2=H, R3=CH3, R4=CH3, R5=H, R6=CH3 50 Ri=SOCH3, R2=H, R3=CH3, R4=CH3, R5=CH3, R6=CH3 51 R2=H, R2=CH3, R3=CH3, R4=CH3, R5=CH3, R^=CH3

analogue of theonellapeptoUde Id (47), which possesses L-MeAla as the tenth aminoacid from the N-terminus [36]. Despite their strong structural similarity, surprisingly, theonellaptoUde lid (49) didn't show Ca^^ ion binding and ion-transport activities. Infact, it seems that the lack of NMeAla produces a significant change in the three-dimensional peptide conformation, responsible for the biological activity mentioned above. TheonellapeptoUde lid (49) prevented fertilisation of sea urchin Hemicentrotus pulcherrimus at a concentration of 25 ^ig/ml while has no effects on early embryonic development of fertilised eggs up to the gastrula stage. This remarkable effect is restricted to 49 whereas 47 did not affect fertilisation of sea urchin gametes even at 100 ^g/ml. Investigation of another Theonella sp., collected off Kerama Island, Okinawa, resulted in the isolation of two new theonellapeptolides (50-51)

1189 [37]. These compounds are new congeners of theonellapeptolides and possess a methylsulfinylacetyl and an acetyl group, respectively, at their N-terminus, replacing the methoxyacetyl group found in all theonellapeptolides previously isolated [35,36]. Compounds 50-51 showed antimicrobial activity against some Gram-positive bacteria and against fungi with MIC in the range 8-60 |Lig/ml. It should be noted that in 1998 a new peptide lactone, theonellapeptoUde Ille (52) has been isolated

I

O

II

Ma

r

II

I

A"••'X.Jl heonellaDeDtolide Ille (52) Theonellapeptolide

I

from a sponge, Lamellomorpha strongylata [38], belonging to an order taxonomically distant from that of Theonella swinhoei. Theonellapeptolide Ille (52) possesses the same molecular formula as theonellapeptoUde le (48) [35] and a similar peptide backbone but differs in four aminoacid residues. In particular the substitutions of L-Ala for PAla results in a 36-, rather than 37-membered peptolide ring for 52. Isolation of closely related compounds from taxonomically remote species suggests the involvement of symbionts in the production of theonellapeptolides. Another evidence for the symbiotic origin of theonellapeptolide family has been reported in a recent Crews' paper [39] in which seventeen specimens of the Papua New Guinea collections of Theonella swinhoei and T. conica were examined. This study has resulted in the isolation of theonellapeptolide Id (47) [35] and motuporin (53), as

Motuporin (53)

1190

the two major metabolites of these collections of Theonella swinhoei and conica, both heavily infested with cyanobacteria. Interestingly one of these specimens of Theonella swinhoei afforded the new antifungal cycUc decapeptide cyclolithistide A (54) [39]. Its structure, which contains the unique aminoacids 4-amino-3,5-dihydroxyhexanoic acid, formyl-leucine and chloroisoleucine, was elucidated through a combination of CI

II

Me

-.

o H2N^

CycloUthistide A (54)

spectroscopic techniques. The cycUc pentapeptide motuporin, firstly isolated from a Theonella swinhoei [40] represents an extremely potent PPl protein phosphates inhibitor belonging to a family of structural related cyanobacterial toxins, including nodularins [41] and the heptapeptide microcystins [42] Specifically, these peptides all contain the unusual aminoacid Adda and an a,p-unsaturated aminoacid, judged, on the basis of X-ray studies [43] the essential features required for binding to PPl [44]. Although their potent and competitive inhibition of protein phosphatases PPl and PP2A [45,46], motuporin shows antineoplastic activity [40] whereas nodularins and microcystins promote tumour formation in mice [47,48]. This interesting paradox in the biochemical profiles of these peptides suggests that the biological outcome of phosphatase inhibition is dependent upon cellular localisation and the specific phosphatase that is targeted. The structural differences between microcystins and motuporin may contribute to the different localisation of these compounds in the cells. A new cycUc heptapeptide cupolamide A (55) was isolated from two samples of the sponge Theonella cupola collected in Indonesia and Okinawa [49]. Three of the aminoacid residues, fran5-4-hydroxyprohne, homoarginine and 2,4-diaminobutanoic acid are of uncommon occurrence; in particular the hydroxy-proline is a constituent of collagen and occasionally has been found in terrestrial biota [50] and in marine

1191

cyanobacterial metabolites [51]. Cupolamide A (55) was active against NH H2N—^ HN-

O O HO.

^NH ^,1

\

I O

//

O

Cupolamide A (55)

^ ^

^

Q^T)-..iOS03Na

,^^0^°"

P388 murine leukaemia cells with an IC50 value of 7.5 |ULg/ml. Analysis of the extracts obtained from Papua New Guinea collections of Theonella mirabilis and Theonella swinhoei resulted in the isolation of four novel cycUc depsipeptides, papuamides A-D [52]. The structures of papuamide A (56) and B (57) were determined from a combination of spectroscopic analysis, chemical degradation and derivatization studies.

O^NH Papuamide A (56) Ri=R2=H, R3=CH3 Papuamide B (57) Ri=R2=R3=H

O V ' " O

=

OH

1192

These peptides contain a number of unusual aminoacids including 3,4dimethylglutamine, P-methoxytyrosine, 3-methoxyalanine, 2,3diaminobutanoic acid or 2-aniino-2-butenoic acid residues. Papuamides are also the first natural marine-derived peptides reported to contain the 3hydroxyleucine and homoproline residues, as well as an unprecedented acid moiety N-linked to the terminal glycine residue. Papuamide A (56) was cytotoxic against a panel of human cancer cell lines with a mean IC50 value of 75 ng/ml; papuamide A (56) and B (57) inhibited the infection of human lymphoblastoid cells by HIV-1 in vitro with an EC50 of 4 ng/ml. Linear peptides Further examination of the same sponge that contained the thrombininhibitory cyclotheonamides [8], led to the isolation of a thrombininhibitory linear tetrapeptide, nazumamide A (58) [53]. It is the first naturally occurring peptide possessing the N-2,5-dihydroxybenzoate terminus. The discovery of thrombin inhibitors is of extreme interest due •\^^

OH

OH

^

^^

Nazumamide A (58)

to their role in promoting the coagulation process. Recently Nienaber et al [54] reported the crystal structure of 58 complexed with human thrombin. This structure demonstrated that 58 binds to thrombin with a novel retromanner but with low selectivity and specificity. Although the weak binding proprieties of this natural linear peptide make it a nonviable anticoagulant, 58 may serve as a usefiil template for the design of more potent and specific thrombin inhibitors via synthetic [55] and combinatorial methods. Thus a large library of NAZA analogous has been prepared using 25 natural and unnatural aminoacids [56]. These studies led to the identification of 2,5-dihydroxylbenzoyl-lysyl-isoleucylphenylalanyl-arginine as a novel thrombin inhibitor 25 times more potent (IC50 1.9 p,M) than the natural peptide 58 (IC50 53 |iM). In 1999, koshikamide Ai (59), a new cytotoxic (IC50 2.2 ^g/ml) linear decapeptide, has been isolated from a recollection of Japanese Theonella sp. [57]. Its aminoacid constituents have been determined by NMR analysis and chemical methods and the sequence was elucidated through

1193 an extensive HMBC analysis and supported by FAB-MS/MS, Unique features of 59 are the methoxyacetylation of the phenylalanyl N-temiinus, C0NH2 O

^ H

Me

Me O

O

z

" z » O ^y^^ Me

/

.CONH2 ^ O

H 5

O

I / . / ', ^

^

Koshikamide Ai (59) reminiscent of theonellapeptoUdes [35], as well as the unprecedented presence of five contiguous N-methyl aminoacids. Another class of linear peptides is represented by polytheonamides

PH

o=< w

NH

NHMe

° H SVK S^K'^^^K'J 0=< NHMe

%

HO

H,N^O

OH

o

0=< NH2

NH2

NHMe

NHMe

Polytheonamide A (60) R=H Polytheonamide B (61)R=Me (60-61), isolated from the Upophilic extract of Theonella swinhoei [58]. The structures of highly cytotoxic polytheonamides 60-61 were assigned to be linear 48-residue peptides with N-terminus blocked by a carbamoyl group.

1194 A recent reinvestigation of 1993 collection of Theonella swinhoei collected off Hachijo-jima Island has resulted in the isolation of six new peptides related to cyclotheonamides named pseudotheonaraides Aj, A2, B2, C, D (62-66) and dihydrocyclotheonamide A (67) [59]. HO.

HN

I

I

Q^iJs^NH HN / ^ ^

HN

„

HO / N//

^^

Pseudotheonamide A^ (62)

Pseudotheonamides 62-64 are linear pentapeptides embracing the rare piperazinone and piperidinoiminoimidazolone ring systems. Pseudotheonamide C (65) contains a residue of vinylogous tyrosine instead of a piperazinone ring. Pseudotheonamide D (66) is a linear tetrapeptide, which lacks the C-terminal a-ketohomoarginine unit whereas the dihydrocyclotheonamide A (67) is a reduction product of the above-mentioned cyclotheonamide A (12). These linear congeners of

H' -O O Pseudotheonamide A2 (63) cyclotheonamide family also showed a serine protease activity; in particular pseudotheonamides 62-66 and 67 inhibited thrombin and trypsin with IC50 values in the range of 0.33-3 |LiM and 4.5-10 |ULM, respectively. Since the power of cyclotheonamides as serine protease inhibitors is associated with the presence of the a-keto group in the k-Arg

1195 residue, it is not surprising that compounds 62-67, in which this group was modified or missing, showed moderate activity.

Pseudotheonamide B2 (64)

d^^

d

NH2 HN^

^1

^

0y \

<\

^NH

H^^°'J^^ » ^ ' NH

„V T CJ

Pseudotheonamide C (65)

H

^O

Pseudotheonamide D (66)

1196

NHCHO

NH

Dihydrocyclotheonamide A (67)

Macrolides and Polyketides Macrolides of marine origin are of current interest because of their structural novelty and biological activities [60-63]. Swinholide A (68) was first isolated by Carmely and Kashman as an antibiotic compound from the sponge Theonella swinhoei collected from the Red Sea [64]. These authors proposed its plain structure as a monomeric macroUde 0CH3

3P

Swinholide A (68): R^ = R2 = CH3 SwinhoUde B (69): Rj =H, R2 = CH3 ^ SwinhoUde C (70) :Ri = CH3,R2=H

mainly based on 2D-NMR studies. The structure of swinholide A (68) as a 44-membered dimeric macrolide, wasfinallyestablished by Kitagawa et al. when they isolated four cytotoxic dimeric macrolides, swinholides A

1197 (68) [65], B (69) [66], C (70) [66] and isoswinhoUde A (71) [66] from the Okinawan marine sponge Theonella swinhoei. The absolute stereostructure of swinholide A (68) was determined by a combination of X-ray crystallographic analysis and chemical degradations [67,68]. 0CH3

o, .0

.^OY"^^-^ 0CH3

Isoswinholide A (71) The absolute stereostructures of swinholides B (69), C (70) and 0CH3

0CH3

Misakinolide A (72)

1198 isoswinholide A (71) were elucidated by means of chemical correlations. In 1986 another highly active antitumor macroUde, misakinoUde A (72), was isolated from the sponge Theonella sp. and its structure reported to be a monomeric 20-membered lactone [69]. The isolation of bistheonellide 0CH3

HO

OH

OH

O

Preswinholide A (73) A (72) [70] from Theonella sp. led to the conclusion that both bistheonellide A and misakinoUde A had the same dimeric structure 72. The absolute stereochemistry of misakinoUde A (72) was determined by chemical correlations with swinhoUde A (68) [71]. Four new cytotoxic minor congeners, swinhoUdes D-G [72] and the monomeric carboxyUc acid of swinhoUde A, pre-swinhoUde A (73) [73] were then isolated from the Okinawan marine sponge Theonella sp. The high cytotoxic activity exhibited by swinhoUde A (68) [74], probably related to the spatial ring conformation [75], has been attributed to its abiUty to dimerize actin and to disrupt the actin cytoskeleton [76]. Bioassay-guided separation on the extract from an Okinawan species of Theonella led to the isolation of an active constituent, onnamide A (74)

M r^ OH MeO =

O' H

Xp"^

Onnamide A (74)

1199 [77], whose structure was elucidated on the basis of spectroscopic data. Since the relative configuration for the ring portion of onnamide A (74) are fully identical with those of the insect toxin pederin [78] and mycalamide A [79], the configurations at C-11 and C-21 of 74 were tentatively assigned by analogy to these compounds. The structures of eight cytotoxic metabolites [80] closely related to onnamide A (74) and of three onnamide congeners [81] were then determined by interpretation of NMR spectral data as well as by comparison of spectral data with those of 74. Theopederins A (75)-E [82], potent cytotoxic metabolites closely related to mycalamides A and B [83] and onnamide A (74), were isolated from a marine sponge Theonella sp. and their structures established

DMe O

O^ ^O

TheopederinA (75)

mainly by extensive 2D NMR analysis. Theopederins A (75) and B showed promising antitumor activity. Theonezeolides A (76)-C, isolated from the Okinawan marine sponge

Theonezolide A (76)

1200 Theonella sp. [84,85], belong to a new class of polyketide natural products consisting of two principal fatty acid chains with various functional groups. These novel cytotoxic 37-membered macrolides, having unique bioactivity of induction of rabbit platelet shape change and aggregation [86] contain 23 chiral centres among which only the absolute stereochemistries of the chiral centre at the terminal position and of the C4-C17 fragment have been determined [85]. Alkaloids Four novel pyridine alkaloids, theonelladins A-D (77-80), have been reported from the Okinawan Theonella swinhoei [87]. These long-chains ,NHR

,NHR

(CH2)io

^

N

Theonelladin C (79) R=H Theonelladin D (80) R=CH3

Theonelladin A (77) R=H Theonelladin B (78) R=CH3

substituted pyridines showed potent antineoplastic activity against murine lymphomas L1210 with IC50 values of 4.7, 1.0, 3.6 and 1.6 |ig/ml, respectively. Their potent antineoplastic activity as well as their powerful Ca -relasing activity from sarcoplasmic reticulum (20 times more potent than caffeine) stimulated the stereoselective total synthesis by Rao and co-workers [88]. Analysis of the extracts of different collections of Theonella sponge resulted in the isolation of the first example of marine occurring brominated benzyl tetrahydroprotoberberine alkaloid, theoneberine (81)

H3C0,

HO* Br "^ k^ ^^

Theoneberine (81)

^OH

Theonellin (82) X=NCS TheoneUin (83) X=NHCHO

OCH3

[89]. Theoneberine (81) exhibited antimicrobial activity against Gram positive bacteria and cytotoxicity against murine LI210 and human Kb cells (IC50 2.9 and 10 lig/ml, respectively). A class of novel bisabolene

1201 type sesquiterpenoids, theonellins, has been isolated from Okinawan collections of Theonella. Theonellins 82 and 83 [90] possess an isothiocyanate group or a formamide group at CI and a conjugated diene in the side chain. The other compounds, for example aminobisabolene 84, characteristically have an amino group at C8 in the side chain [91]. An unusual nucleoside, kumusine (85), has been reported from the Indonesian marine sponge Theonella sp. [92]. Kumusine (85) represents the first natural occurring halogenated purine nucleoside containing an alkylated sugar, 2-methyl-5-deoxy ribose, that also represents a new NH2

Aminobisabolene (84)

oHOH

*^"°^"^"^^ ^^^^

Structural feature in a nucleoside. This unusual metabolite showed moderate immunosuppressive activity (MLR IC50 0.195 ^ig/ml) and cytotoxicity against P388 and HT29 with IC50 values of 5.0 |ig/ml. Metabolites from Discodermia Cyclic peptides Fifteen years ago the Fusetani group first isolated the bioactive sponge peptides, discodermins A-D (86-89) from the sponge Discodermia kiiensis, collected off Atami in the Gulf of Sagami [93-95]. They were later found to be potent inhibitors of phospholipase A2 [96] and discodermin A (86) inhibited the tumour promotion activity of okadaic acid [97]. Subsequent papers reported the structures of discodermins E-H (90-93) and the structure revision of 86-89 [98,99]. Discodermin derivatives are cyclic depsipeptides, particularly rich in D-aminoacids, showing a negative ninhydrin reaction due to the presence of a formylated N-terminus. Discodermins E-H (90-93) were cytotoxic against P388 cells at ICso's of 0.02, 0.1, 0.4 and 0.1 |Lig/ml, and inhibited the development of starfish embryos of Asterina pectinifera at 5, 10, 20, and 10 |Lig/ml, respectively. In 1992, a depsipeptide, polydiscamide A (94), composed of 13 aminoacids, was isolated from a Caribbean Discodermia sp. [100]. Polydiscamide A (94) inhibits the in vitro proliferation of the cultured

1202 human lung cancer A549 cell lines with an IC50 value of 0.7 |LXg/nil. Compared to discodermins, 94 contained p-bromophenylalanine instead of phenylalanine in the side chain and the lactone ring consisting of five aminoacids instead of six as in discodermins.

-

O ^ ^

<^

\

OQ ^

HN

)-NH

. N ^

C

H2N

Discodermin A (86) Discodermin B (87) Discodermin C (88) Discodermin D (89) Discodermin E (90) Discodermin F (91) Discodermin G (92) Discodermin H (93)

Ri=R2=H, R3=R4=Me, R5=X ^ Ri=R2=R3=H, R4=Me, R5=X J^Y^ Ri=R2=R4=H, R3=Me, R5=X ^M A ^ Ri=R2=R3=R4=H, R5=X H ^ Ri=R2=H, R3=R4=Me, R5=Y Ri=R2=H, R3=Me,R4=Et, R5=X Ri=R3=R4=Me, R2=H, R5=X Ri=H, R2=0H, R3=R4=Me, R5=X

9 '^^MT^ u M^'^N^ H2N

H NH2

H o V - n O f^

NH

HT^NHcPYViiSrV^iV H HN .WgOgNa 00

>-
X feHJW HHN..,

O'^NHMe 0^0

Polydiscamide A (94)

OQi^H

H (^

'

The same group reported the structures of two active peptides, discobahamins A (95) and B (96), isolated from a new species of the Bahamian deep-water marine sponge Discodermia [101]. The

1203 discobahamins (95-96) are structurally related to the above cyclic peptides orbiculamide A [20] and keramamides B-D [21,22] isolated from two Theonella sp. The distinctive features of discobahamins are the substitution of a 5-hydroxytryptophan residue in place of a 2-bromo-5hydroxytryptophan residue in the cyclic core and the substitution of various aminoacids in the side chain. The N-terminus is protected by 2hydroxy-3-methylpentanoic acid as in keramamides B-E. However, the discobahamins possess one less aminoacid residue than the keramamides

1 H

X ^A^

Discobahamin A (95) R=H Discobahamin B (96) R=Me

in the side chain. Both 95 and 96 showed antifungal activity against the yeast form of C albicans (MIC of 25 |ig/ml). Cytotoxic cyclic depsipeptides, discokiolides A-C (97-99) were ,NH2

o= O

Discokiolide A (97) R=H, R^OMe DiscokioUde B (98) R=Me, R^OMe Discokiolide C (99) R=Me,Ri=H

1204 isolated from Discodermia kiiensis as methyl esters [102]. These metabolites contain the oxazole ring reminiscent of keramamide B-E [21,22]. Macrolides and Polyketides In the course of a search for antitumor substances from Japanese marine invertebrates, eight unique polyketides bearing nitrogen and phosphorous functions, calyculins A-H, were isolated from the sponge Discodermia calyx. They exhibit a variety of biological activities including cytotoxicity against several cell lines, in vivo antitumor activity, potent inhibition of protein phosphatases 1 and 2A [103] and insecticide activity. The structure and the relative stereochemistry of the major metaboUte, calyculin A (100), was determined by X-ray diffraction [104]. The absolute stereochemistry was elucidated only later (1991) [105], when Matsunaga and Fusetani isolated a y-aminoacid fragment (C33-C37 portion

Calyculin A (100): R = H Calyculin C (101): R = Me

of calyculins) from the acid hydrolysed of a mixture of calyculins A, B, E, and F. The absolute stereochemistry of this fragment was tentatively assigned by comparison of its CD spectrum to those of simple a-hydroxyl acids. Shioiri and co-workers unambiguously confirmed the assignment through the synthesis of the C33-C37 acid fragment [106]. Spectroscopic analysis including extensive NMR experiments and comparison of spectral data with those of calyculin A (100), showed that calyculins B, E, and F are geometrical isomers of calyculin A with respect to the C2-C3 and C6-C7 bonds while calyculins C (101), D, G and H are homologous

1205

series with a methyl group at C-32 [107-110]. It should be noted that the presence of an additional methyl group at C-32 in calyculins C (101), D, G and H as well as the changes of the geometry in the tetraene portion (calyculins B, E, and F) have a negligible effect on the biological activity of these molecules. Further examination of D. calyx led to the isolation of calyculin J (102), whose structure was confirmed by chemical transformation from calyculin A (100) [111] and four more calyculin derivatives, OH O '

"

>

^

Calyculin J (102)

calyculinamides

A and

F OH

HQN^^O

/

(103), des-N-methylcalyculin O

A and

=

\

OH

OH OMe

Calyculinamide F (103)

dephosphonocalycuhn A [112,113]. The combination of NMR spectroscopy and X-ray crystallographic analysis revealed a novel polypropionate structure for discodermolide (104) [114], a polyhydroxylated lactone isolated from Discodermia

1206 dissoluta. It inhibits purified murine T cell proliferation with an IC50 of 9 nM, inhibits the mixed leukocyte reaction, and suppresses graft-versushost disease in transplanted mice. Discodermolide (104) shared the same mechanism of cytotoxic action of paclitaxel: the stabilisation of microtubules leading to mitotic arrest [115]. Discodermia dissoluta also

OH

OCONH2

OH

Discodermolide (104)

Discodermide (105)

contained the antifungal and cytotoxic macrocyclic lactam discodermide (105) [116]. Metabolites from Jerecopsis During our search for new anti-tumour compounds from New Caledonian marine organisms, the first lithistid marine sponge investigated in our laboratories was Jereicopsis graphidiophora whose Upidic fraction showed to contain several 3P-methoxy steroids [117] whereas the

MeO'

MeO

106

JereisterolB(107)

conventional SP-hydroxysteroids were totally absent. Five of them were

1207 novel compounds containing an additional oxygenated function such as an epoxy, hydroxyl or ketone group. Compound 106 is the first example of a natural occurring steroid containing a A^^^^^-8,14 epoxide function. Jereisterol B (107), ^o isolated from Jereicopsis graphidiophora [118] is the first example of the unique 8,14 seco structure successively encountered in swinhosterols A and B [7]. The unusual structure of 107 was confirmed by synthesis of the model 3p-acethoxy-8,14secoergostane-8,14-dione by oxidation of the corresponding 3p-acethoxy-

MeO

H

Jereisterol A (108)

109

ergost-8(14)-ene. Jereisterol A, also isolated from Jereicopsis graphidiophora [118], has the unique structure 108, which is without precedent in the steroid literature. Metabolites from Neosiphonia and Reidispongia In a later stage we had the opportunity to investigate in our laboratories other two New Caledonian lithistid sponges, Neosiphonia superstes and Reidispongia coerulea which proved to be a rich source of biological active macrolides and cyclic peptide with unusual structures. The first paper on Neosiphonia superstes appeared in Hterature in 1991 and deals with the isolation of the steroidal compound 109, 25S-25Methyl-24-methylenecholest-4-en-3-one, with an unconventional sidechain [119]. Macrolides and Polyketides Along with sphinxolide (110), previously isolated from a pacific nudibranch [120], sphinxolides B-D (111-113) were first isolated from a 1993 analysis of Neosiphonia superstes [121]. At the same time, investigation of the sponge Reidispongia coerulea afforded two related macrolides, named reidispongiolide A (114) and B (115) [122]. Sphinxolides and reidispongioUdes represent a new class of cytotoxic macrolides characterised by a very similar 26-membered lactone ring with an N-formyl side chain. They differ in that sphinxolides are hydroxylated at CIO whereas reidispongioUdes are not. During a recent reinvestigation

1208 of Neosiphonia superstes and Reidispongia coerulea, with the aim to

OMe

SphinxoUde (110) Sphinxolide B (111) Sphinxolide C (112) SphinxoUde D (113) Reidispongiolide A (114) ReidispongioUde B (115)

R=OMe, Ri=OH, R2=H R=H, Ri=OH, R2=H R=OMe, Ri=OH, R2=Me R=H, Ri=OH, R2=Me R=H, Ri=H, R2=Me R=H, R^H, R2=H

obtain additional quantities of sphinxolides and reidispongiolides for

SphinxoUde E (116) further biological and stereochemical studies, we isolated four new related compounds named sphinxolides E-G (116-118) and reidispongiolide C (119) [123]. Complete structural assignments of 116119 were obtained through careful analysis of 2D-NMR spectra, such as COSY, HMQC and HMBC. In particular SphinxoUde E (116) represents a derivative of sphinxolide C (112) in which an epoxide fiinctionality replaces the 4,5 double bond. The presence of an epoxide at this position, in agreement with the MS data, is also reminiscent of the antifungal macrolide of bacterial origin, pimaricin [124]. Inspection of ^H and ^C NMR of sphinxoUdes F-G (117-118) and

HOOC OMe

SphinxoUde F (117) SphinxoUde G (118) ReidispongioUde C (119)

R=OMe, Ri=OH, R2=H R=H, Ri=OH, R2=Me R=H, Ri=H, R2=Me

1209 reidispongiolide C (119) clearly showed a single set of resonances, thus it was readily observed that all these compounds lacked the N-methylformyl end group, responsible for the doubling of selected NMR signads. The structures of 117-119 were determined by 2D-NMR and MS data as truncated derivatives of sphinxolide (110), sphinxoUde D (113) and reidispongioUde A (114), respectively, and confirmed by preparing a semisynthetic sample of sphinxolide F (117), by exposing sphinxolide (110) to Jones' reagent. All metabolites of the sphinxolide and reidispongiolide family were evaluated in the U.S. National Cancer Institute by disease-oriented in vitro primary screening on human tumour cells. In this test, compounds 110-119 proved to be very potent and selective [123]. The truncated derivatives 116-119 are less potent by 10-100 times when compared to other sphinxolides, thus indicating that the side chain terminus plays an active role in the mode of action of these molecules. Like scytophicins and tolytoxin, isolated from terrestrial algae [125], sphinxoUdes caused rapid loss of microfilaments in cultured cells without affecting microtubule organisation [126]. Importantly, they circumvent multidrug tumour resistance mediated by overexertion of P-glycoprotein. Thus sphinxolides may be useful models in the development of new anticancer agents for the treatment of drug-resistant human tumours. Bioassay guided fractionation of active extracts of Neosiphonia superstes has afforded two more unrelated highly cytotoxic macrohdes OCONH2

? H

Superstolide A (120)

superstoUdes A (120) [127] (IC50 value of 0.003 ^g/ml), and its 24,25dehydro derivative superstolide B [128]. The gross structure of superstoUde A was determined by extensive 2D NMR experiments on the lactone 120 and on its opened-ring-derived methyl esters. The relative stereochemistry of superstolide A (120) was determined on the basis of NOESY data, coupling constants analysis and molecular mechanics calculations performed on the C22-C25 acetonide derivative. The absolute stereochemistry was assigned by the application of Korean's and Mosher's methodologies.

1210 Cyclic peptide From the methanolic extract of Neosiphonia superstes, we isolated a new cyclodepsipeptide, neosiphoniamolide A (121) [129] closely related to

Neosiphoniamolide A (121) geodiamolides and jaspamide, previously isolated from Jaspis [130,131] and Geodia [132] sponges, respectively. These metabolites contain a common 12-carbon hydroxy acid and three aminoacid residues linked each other to form a IS-membered macrocycle. In particular 121 is a valine analogue of geodiamoilide D. NeosiphoniamoUde A (121) inhibits the growth of the fungi Piricularia oryzae and Helminthosporium gramineu with IC50 values of 5 ppm. Metabolites from Callipelta Macrolides Investigation on the dichlorometane extract obtained from several collections of the New Caledonia sponge Callipelta sp., led to the isolation of new cytotoxic glycosides, callipeltosides A-C (122-124) [133,134]. Structural assignment of callipeltosides was accomplished through extensive 2D NMR spectroscopy. The complete relative stereochemistry was proposed on the basis of the analysis of ROESY and NOE difference experiments. Callipeltoside A (122) is the first member of an unprecedented class of marine natural products with unusual structural features such as the previously unknown 4-amino-4,6-dideoxy2-0,3-C-dimethyl-a-talopyranosyl-3,4-urethane (callipeltose), linked through an 0-glycoside linkage to a hemiketal oxane ring, the latter being part of a 14-membered macrocyclic lactone with a dienyne chlorocyclopropane side chain. Branched chain sugars have been reported as constituents of antibiotic macroUdes produced by micro-organisms. The 6-deoxy-2-0,3-C-dimethyl-L-talose (vinelose) and the 4,6-dideoxy-

1211 3-C-methyl-4-(methylairiino)-L-mannose (sibirosamine), both similar to callipeltose, were found to be part of nucleosides isolated from cultures of Azobacter vinelandii [135] and as the saccharide moiety of sibiromycin, an antitumor antibiotic isolated from Streptosporangium sibiricum [136].

ft' ft'

NH

MeO/^

R= ^ ^ O ^ ' - / CaUipeltoside A (122) MeO

n

V

^

...^ ^^y MeO/,

.NHCHO

CaUipeltoside B (123) ^ .>PH MeO/,^ ^

CaUipeltoside C (124) This may suggest a symbiotic origin for 122. Callipeltosides B (123) and C (124) differ from 122 in the saccharide moieties, with callipeltoside B (123) possessing an A^-formyl group at C-4' that replaces the urethane functionality while the sugar portion in callipeltoside C (124) is a mannopyranose derivative. Callipeltosides A-C (122-124) resulted moderately active against the NSCLC-N6 cell Unes, with IC50 values of 11.3, 15.1 and 30.0 |ig/ml, respectively. However, cell cycle analysis by flow cytometric assays of NSCLC-N6 cells treated with 122 revealed a cell cycle-dependent effect, involving a dose dependent blockage of NSCLC-N6 cell proliferation in vitro at the level of the Gl phase, or by enzyme inhibition or by inducing terminal cell differentiation. In the latter case callipeltoside A (122) would be an interesting lead for mechanismbased studies. Cyclic peptides Callipeltins A-C are a new class of cyclic depsipeptides isolated from Callipelta sp. [137,138]. Callipeltin A (125) appears very intriguing because of some structural features such as the presence of four new aminoacid residues, not previously isolated from natural sources (p-MeOTyr, N-Me-Gln, di-Me-Glu and 4-amino-7-guanidino-2,3-dihydroxyeptanoic acid), the N-terminus blocked with a 3-hydroxy acid residue and

1212

the C"terminus lactonized with a threonine residue. The structural work on the decapeptide 125 required an extensive use of 2D-NMR on the intact molecule as well as a careful degradative work followed by chromatographic steps to isolate the unprecedented aminoacids [137]. Some of these unique structural features have been successively found in

NH

%

^

HN

,v

'N H

X

NH2

OH

CaUipeltin B (126)

papuamides A-B (56-57)^^ isolated from Theonella sponges, belonging to another family (Theonellidae) of the order Lithistida. CaUipeltin B (126) possesses the same cycUc depsipeptidal structure as in callipeltin A (125) and differs from 125 by having the N-terminal 2,3-

1213 dimethylpyroglutamic acid unit instead of the tripeptide chain with the Nterminus blocked by an hydroxyacid. Callipeltin C is simply the acycUc callipeltin A. Callipeltin A (125) was found to protect human cells infected by HIV virus with CD50 of 0.29 ^ig/ml and ED50 of 0.01 |Lig/ml giving a selective index (SI ratio CD50/ED50) of 29. Callipeltins A-C were also found to be cytotoxic and antifungal; importantly callipeltin C was less active then 125 so indicating the importance of the intact macrocycle for the bioactivity. All metabolites isolated from Callipelta sp. are suspected to be of microbial origin. In order to confirm this hypothesis and to find the actual producer of callipeltins and callipeltosides and urged by the need to have a constant supply of these bioactive compounds, we studied the bacterial strains isolated from the sponge tissue. Unfortunately, both callipeltins and callipeltosides appear to be absent in the bacterial extracts examined, which showed to contain known bacterial metabolites, namely diketopiperazines [139] and violaceine [140]. Metabolites from Corallistes Alkaloids The investigation of the New Caledonian Hthistid sponge Corallistes fulvodesmus led to the isolation of two polynitrogen compounds, 1methyl-pteridine-2,4-dione, previously known as a synthetic product o I

MGN

NMe

Corallistine (127) [141], and the new corallistine 127, whose structure was determined by

.C00CH3

CorallistinA methyl ester (129) H3COOC

1214

X-ray single crystal analysis of its 6'-isobutyloxycarbonyl derivative [142]. Corallistine 127 possesses the unusual 2-methyl-thioiniidazole moiety, never found before in a natural product. The lithistid sponge Corallistes undulatus has been shown to contain the related pteridine alkaloid 128 together with C28 and C29 steroids, typical of higher terrestrial plants [143], and halogenated indole derivatives. Recently Pietra and co-workers reported the structures of corallistins A-E, unusual free porphyrins isolated as methyl esters from the sponge Corallistes sp., belonging to the family Corallistidae of the order Lithistida [144, 145]. Corallistin A (129), whose structure was derived from NMR analysis and confirmed by total synthesis [146], proved to be active against KB cell Une. It is to be remarked that the geoporphyrins so far isolated from oil shale [147] and coal [148] lack 0-atoms, whereas corallistins A-E are oxygenated. Metabolites from Aciculites Cyclic Peptides The Uthistid sponge Aciculites orientalis contains three cyclic peptides, aciculitins A-C (130-132), that are identical except for homologous lipid residues [149]. The aciculitins consist of a bicycUc peptide that contains an unusual histidino-tyrosine bridge. Attached to the bicyclic portion are Ci3-Ci5-dihydroxy-4,6-dienoic acids bearing D-lyxose at the C-3 position. The paper also reports the structures of the two artifacts aciculitamides A and B obtained earlier from the same sponge. The acicuUtins inhibited the growth of Candida albicans and are cytotoxic toward the HCT-116 cell line. NH2

1.

H2N

lOH

WM

OH

\

N

OHX'

OH

HO/,, A , . 0 ,

/ \ R

Aciculitin A (130) R=C5Hi 1 Aciculitin B (131) R=QHi3 AcicuUtin C (132) R=C7Hi5

V NH

1215 Metabolites from Microscleroderma Analogously to J. graphidiophora, the Senegalese sponge Microscleroderma spirophora elaborates large amounts of 3-Pmethoxysteroids, including compound 133 that contains a tetrasubstituted oxirane ring located at position 8,9 [150]. A homologue of jereisterol A

MeO

133

MeO'

(104), compound 134, has been also isolated along with its isomer 135 [150]. The occurrence of 3P-methoxysteroids has an interesting chemotaxonomic significance. Among the sponge belonging to the lithistid order, the ability of producing 3p-methoxysteroids appears limited to the species Jereicopsis graphidiophora and Microscleroderma spirophora whereas the metabolites produced by other lithistid sponges lack a chemotaxonomic rationalisation because of the evident participation of symbionts in their production. Another example of 3p-0-

MeO'

alkyl steroid has been reported as a constituent of a Uthistid sponge: compound 136, a new methoxymethyl sterol ether isolated from a deep water marine sponge Scleritoderma sp. cf. paccardi [151]. This compound exhibited in vitro cytotoxicity against murine P388 tumour cell line with an IC50 of 2.3 |Lig/ml. Cyclic Peptides In 1994 Faulkner et al. reported the isolation of microsclerodermins A (137) and B (138) from a Microscleroderma sp. sponge, collected off the

1216

COOH A.

I

O Me

HN—^ ur^•

I

Microsclerodermin A (137) R=OH Microsclerodermin B (138) R=H

MeO

Norfolk Rise near New Caledonia [152]. Microsclerodermins are cyclic hexapeptides that inhibit the growth of C. Albicans at a concentration of 2.5 iLig/ml. Recently three more microsclerodermins (139-141) have been isolated o o

HN—^

^ < y ' ^ ^ N " " ' ^ - - - ^ 7 \ ^ Microsclerodermin C (139) R=C0NH2 L ^® HO 1 Microsclerodermin D (140) R=H

GOGH

NH

O

HN—^

i

Microsclerodennin E (141)

from two species of lithistid sponges from Visayan Island, including a

1217 Theonella sp. and a Microscleroderma sp. [153]. Microsclerodermin C (139) and E (140) are modified with respect to 137 in the P-aminoacid moieties, 3-amino-8-phenyl-2,4,5-trihydroxyoct-7-enoic acid and 3amino-I0-(p-ethoxyphenyl)-2,4,5-trihydroxydeca-7,9-dienoic acid, respectively. In addition microsclerodermin C (139) contains the unprecedented 6' -chloro-N' -formammidotryptophan residue. Microsclerodermin D (140) is identical to C except for the Trp unit that was no longer N-substituted. Microsclerodermins 139-141 were active against Candida albicans at a concentration of 5, 10 and 50 |Llg, respectively. TOTAL SYNTHESES The total synthesis of macrolides and peptides from lithistid sponges has attracted the interest of a huge number of research groups. This interest is motivated by the biological activities exhibited by these natural products and their potential as chemiotherapeutic agents. Since only small quantities are available from natural sources, often total synthesis represents the only tool for supplying sufficient quantities of the product for further biological and pharmaceutical studies. Total synthesis is also a valuable tool for confirming, and in some cases determining the stereochemistry of lithistid metabolites, that, due to the small quantities and the unusual structural features, can not be unambiguously determined on the basis of spectroscopic and crystallographic data. The very intriguing structures of lithistid metabolites with their array of substituents and functional groups, have become the yardsticks for measuring progress in the efficiency of stereoselective synthesis. Many new methods for controlling the sp^ and sp^ stereochemistry of the carbon skeleton have been developed and a clearer understanding of the factors affecting the stereochemical outcome of many established reactions, such as the aldol condensation, has also emerged from these works. Due to the impressive number of papers {ca. 200) dealing with the synthesis of Uthistid metabolites, the choice has been made to include in the following section only the works reporting total syntheses, disregarding, a part few representative exceptions, formal syntheses, semisyntheses and preparation of unnatural enantiomers, diastereoisomers or analogues of natural products. Motuporin (53) Motuporin and other members of serine/threonine phosphatase inhibitors such as calyculins, microcystins, nodularins have attracted considerable synthetic interest in the last years. Schreiber's group reported the first total synthesis of motuporin in 1995 [154], whereas very recently three alternative approaches appeared in the literature [155].

1218

Fig. (1) shows the retrosynthetic approch proposed by Schreiber. The 19-membered pentapeptide macrocycle of motuporin was initially disconnected at the (5)-valuie /^-o-D-^ry^rt^-P-methylaspartate amide bond. The (N-methylamino)-dehydrobutyrate residue (mdhb) was masked

Motuporin (53)

OCHo N-Boc-L-Val-Adda

CO2CH3 ^

OCH3

Linear peptide 142

CO2CH3 ^^ O I ' "^ Me ij

NHTs

143, N-Ts-Adda

/

CO2CH3 • ^ COaPac

Tripeptide 144

\

/ NHBoc > ^ XOoPac

N(Boc)Ts js. rn rw 146

\

\

NMeThr

Glu

. 1 4 7

D-Mandelic Acid

z>

rD-Thr|

^NHTs 148

Figure (1). Schreiber's retrosynthetic approach to motuporin (53).

as a threonine until late stages of the synthesis. Further disconnection of the linear peptide 142 afforded N-Boc-valine Adda fragment and the remaining tripeptide fragment 144. Schreiber and co-workers developed a highly convergent approach in which each of three unusual aminoacids in motuporin was synthesized starting from the same chiral building block: D-threonine. This latter was easily converted to y-lactone 148 through cyanide opening of the intermediate tosylilaziridine 149 (Scheme 1). The lactone 148 was converted to aldheyde 150, corresponding to the C1-C4 subunit of the Adda residue or, alternatively, to the differentially

1219

protected P-MeAsp derivative 147. The C5-C10 portion of the Adda Ts N / \

I.MS2O

QH

^I>,^^C02CH3 2- NaBH4 H3C 3. TBS-CI NHTs

/

LNaCN.

V ^ ^™o

2. HCI

CH2OTBS 149

N(Boc)Ts

N(Boc)Ts

D-Thr

^COgCHa

HOH2C'

^

^C02CH3

Swern

^ 150

146

NHBoc NHBoc

nr

H3CO2C

C02Pac

C02CHPh2 1. H2, Pd/C

HgCOgC^^^r^

2. PhCOCH2Br KF(H20)2

147

Scheme 1 - Schreiber synthesis of the C1-C4 subunit of the Adda residue (150) and of protected Me-Asp residue (147).

residue was obtained from benzyloxymethyl-protected mandelaldehyde through initial crotylstannane addition. The diene functionality was installed through sequential application of Wittig and Julia olefinations to give the protected Adda 143 (Scheme 2). 1.DIBAL-H

BOMO

MgBr20Et2 3. NaH,CH3l

C02Et

CH2TS

^

CO2CH3 1. BuLi, then 150 2. NaNaphthalene

Scheme 2. Schreiber synthesis of the Adda residue.

Both dipeptide N-Boc-valine-Adda and tripeptide 144 were obtained

1220

using corresponding protected amino acids and conventional coupling technology. The macrolactamization of the protected linear pentapeptide 142, Fig. (1), was achieved by a four-step sequence involving: a) reductive removal of the phenacyl group (Pac), b) activation of the Cterminus as pentafluorophenyl ester, c) N-terminal Boc group deprotection, d) neutralization and dilution. Treatment with Ba(0H)2 afforded concurrent methyl ester deprotection and N-methylthreonine dehydration. In their retrosynthetic plan Armstrong and co-workers proposed to effect the macrolactamization of the linear peptide at the Adda (iS')-valine

Motuporin (53) COgf-Bu'

y" 0^. .OMe

U H I r.Buo\^YT'

!f^?25,J^H OH NH

COoMe - '^ NHg

.. Me

O

OH

ZHN

'/OBn

OMe

152

153

fl

fl

COaMe

OH

°v^H„r^"°/r"™"-'V™' ™''-^' NHBoc

O^

OMe

^

156

o

o

155

OBn

157

Fig. (2). Armstrong's retrosynthesis of motuporin (53)

amide bond. Fig. (2). This is supposed to be the actual point of cyclization in the biosynthesis of the natural product based on the reported isolation of related linear peptides from cyanobacteria and subsequent feeding studies [156]. The linear peptide was further disconnected to give Bocprotected Adda 151 and the remaining tetrapeptide which was in turn disconnected into dipeptides 152 and 153. Installation of the diene moiety in Adda residue was anticipated to be performed via a Wittig or a related olefmation reaction between a nucleophile derived from bromide 154 and aldehyde 155. The p-methylaspartate residue may arise from the same

1221 aldehyde intermediate. The dipeptide 153 was envisioned to arise from an Ugi four-component condensation (4CC) of monoprotected glutamic acid 156, aldehyde 157, methylamine and cyclohexenyl isocyanide. The two contiguous stereocenters present in the aldehyde 155 were installed in a straightforward manner and with an extremely high stereoselectivity using the dihalomethyllithium insertion method of Matteson (Scheme 3) [157]. The (-h)-pinanediol boronic ester derivative

cte^

PMBOv^B,

158

CHaMgCI LiCHCl2 2. ZnCl2 P M B O ^ ^QV\^, 2. LDA, CHgBrg 01 159 3. ZnCl2

^•^160

LNaNg 2. BULC PMBO

CH2ICI

Scheme 3. Armstrong synthesis of aldehyde 155

158 which acts in the process as chiral director, was prepared according to Uterature [158] and subjected to the first insertion step with dichloromethyllithium followed by ZnCl2. The a-chloroborate ester 159, obtained as single diastereoisomer, was treated with methylmagnesium chloride to give the corresponding a-methyl borate ester subjected to the second insertion with dibromomethyllitium. After displacement of the bromide in 160 with sodium azide, a third insertion step was performed with chloroiodomethyllithium affording the azido borate 161, which was transformed in four steps with standard methods into the desired aldehyde 155. The remaining two stereocenters in the Adda residue were installed through Brown crotylboration method using phenylacetaldehyde and the reagent derived from (+)-5-(methoxy)diisopinocampheyl-borane and cis2-butene (Scheme 4). The obtained homoallylic alcohol was elaborated by using standard conditions to give the allylic bromide 154. The coupling of different Wittig or Homer-Emmonds reagents derived from bromide 154 with the aldehyde 155 was tested in different base and solvent conditions. Best results in terms of yield and stereoselectivity were obtained using the triphenylphosphonium yUde derived from bromide 154 and LDA as base. The diene 162 was easily converted in three steps to N-Boc protected Adda 151. The Y-(D)-glutamyl-N-Methreonine dipeptide 153 was prepared in a one-pot process which utilizes the Ugi technology (Scheme 5) [159]. This latter offers the advantage to form the synthetically challenging tertiary amide bond, usually obtained in low yields using standard peptide coupling techniques. Ugi reaction of the carboxylbenzyloxy (Z) protected glutamic acid derivative 156,

1222 aldehyde 157, methylamine and cyclohexenyl isocyanide afforded cyclohexenamide dipeptide 163, which was converted in the

ou.

1. a) (/Pc)2B\ = / BF30Et2

1.a)03;b)PPh3

b) H2O2, NaOH

'^^^^y^oEt

2. NaH, CH3I

2. a) LDA b)155 162 Scheme 4. Armstrong synthesis of Boc-Protected Adda residue (151).

corresponding free carboxylic acid 153 through acid hydrolysis. A 1:1

HCI 153 DBn Schemes. Armstrong synthesis of the Y-(D)-Glutamyl-N-MeThreonine residue (153).

mixture of diastereoisomers at the a-stereocenter of the threonyl residue was obtained, but this is inconsequential for purposes of synthesizing motuporin as this carbon becomes part of the dehydroamino residue. Panek and co-workers adopted a convergent approach in which the chiral aUylsilane bond construction methodology, elaborated by the same author [160] was extensively used for the introduction of the stereochemical relationships. Fig. (3). Disconnection of 53 at the

1223 indicated sites afforded two principal fragments: A^-Boc-valine-Adda dipeptide 164 and the remaining tripeptide fragment 165. The dipeptide 164 was further disconnected at the C5-C6 bond to give a vinyl metal species 166, and the (F)-vinyl iodide 167. In the synthetic Fragment Coupling

"

Macrocyclization CHaOaq BocHN

O^^^OH

O

HN

XX 164 AZ-Boc Valine-Adda

I

.OSIR3

HN^ ^ O

^O

Tripeptide fragment 165

N3 Tce02Cx^^^xV^

ZnCI

166 Left hand subunit

168

167 Right hand subunit

C02Me

C02Me Me2PhSi R=H, (S)-Silane 170 R=N3, (S,/?>SiIane 171

CO2MG

169 ^*g* (^)- Panek's retrosynthesis of motuporin (53)

1224

direction, a Pd(0)-catalyzed cross-coupling reaction of the C-6 vinyl zinc species and the C-5 vinyl iodide was planned to complete the carbone backbone of the dipeptide 164. Further analysis of the individual subunits produced two chiral silane synthons 170 and 171, of which the anf/-azido silane 171 is derived from 170 through diastereoselective azidation of the corresponding P-silyl enolate [161]. The 1,3-azido alcohol 169, a precursor of both dipeptide 167 and of the methyl aspartate derivative 168 was prepared in two steps from (»S',/?)-silane 171 via sequential diastereoselective crotylation and allyUc azide isomerization (Scheme 6) [162]. After protection of the primary alcohol as TBDPS ether, and reduction with SnCl2 of the azide group, the free amine was condensed with the pentafluorophenyl ester activated A^-Boc-L-valine to afford the dipeptide 172. The double bond was oxidatively cleaved to the corresponding aldehyde which was subjected to Takai's homologation protocol to give dipeptide 167. The C6-C10 subunit of the Adda residue was prepared in a high yielding short sequence involving a ^-yn-selective

C02Me Me

169

NHBoc

1.TBDPSCI 2. SnCl2

1.03,thenDMS

3. /V-Boc-(L)-Valine-OC6F5 aq. NaHCOa

TBDPSO

2. CrCl2/CHl3

-•167

COaMe Scheme 6. Panek synthesis of dipeptide 167.

crotylation reaction to install the correct C8-C9 stereochemical relationship (Scheme 7). Therefore, the phenyl acetaldehyde dimethyl acetal was condensed, under Lewis acid catalysis, with silane 170. Oxidative cleavage of trans double bond in 173 and direct dibromoolefmation of the resulting aldehyde gave the dibromoolefm 174. This latter was treated with n-BuLi to give the acetylene anion which was trapped with methyl iodide to afford the methyl substituted acetylene 175. The vinyl zinc intermediate 166 to be used in the Pd(0)-catalyzed cross coupling reaction with dipeptide 167 was obtained from acetylene 175 through hydrozirconation foUowed by in situ transmetalation with ZnCli. The one-pot cross coupling reaction between 166 and 167 afforded the configurationaUy pure trisubstituted (E,E)-ditne 176 in 80% yield. This latter was easily transformed into required A^-Boc-valine-Adda dipeptide

1225 164 through silyl deprotection of the primary alcohol and oxidation of the C02Me

1. O3, DMS

2. ZnCl2 166 OTBDPS 1.TBAF-SiOo HN^^O 176

2. TPAP/NMO

•^

Dipeptlde 164

- ^ , NHBoc

Scheme 7. Panek synthesis of the N-Boc Valine Adda dipeptide 164.

aforementioned functional group to a carboxyl group. The protected P-methyl aspartate fragment was prepared (Scheme 8) starting from the same azido alcohol 169 previously used in the synthesis of dipeptide 167 through a) Jones'oxidation of the primary alcohol, b) protection of the carboxyUc acid as trichloro-ethyl ester, c) oxidative cleavage of the double bond (RuCla, NaI04), d) protection of the resulting a-azido acid as methyl ester, e) reduction of the azido group in 1. Jones'Oxidation

_ • MeOgC

2. 2,2,2-trichloroethanol DCC, DMAP

(OTce

1.RuCl3(cat),Nal04 • 1168 I 2. CH2N2

Scheme 8. Panek synthesis of the p-methyl aspartate precursor 168.

168 with SnCl2. The obtained free amine was immediately used in standard peptide coupling conditions to eventually give the tripeptide 165. The approach to the key fragment iV-Boc-Adda methyl ester used by Toogood and co-workers in their total synthesis of motuporin (53) [155b]

1226 parallels that employed by other groups [163]. In fact, as already reported by Armstrong and co-workers, see Fig. (2), the diene functionality, present in this molecule, was envisioned to arise from the coupling of a nucleophile derived from bromide 154 (Scheme 9) and aldehyde 177

Ph' 178 1-Se02

,

,

179

180

„./-v^.xV^^^^/0'^ - • Ph J '

2.NaBH4. CeCl3

MeO

CBr4,Ph3P

•Ph

/s.

^^^

JL ^

MeO

^Br

^54

Scheme 9. Toogood synthesis of the C5-C10 fragment of Adda residue (154).

(Scheme 10). For controlling the chirality at C-8 and C-9 positions of Adda, the Evans' OH

9

1.LDA,-78°C

''//

2. CH3I, NaH 182

-

.

Q

BocHN

5 Steps BocHN^^A^

/

^

OCH3

183

CO2CH3

177

\l.Nal04,Rua3 2. DCC, DMAP

BocHN,^X^

'OCH3

^ ^

CeFsOH

^^2C6F5 184

Scheme 10. Toogood synthesis of the C1-C4 fragment of Adda (177) and of the a-methyl p-pentafluorophenyl-A^-Boc-methyl aspartate residue (184).

chiral oxazolidine template methodology was used, as already reported by Rinehart and Beatty [163]. The aldol 178 obtained from phenylacetaldehyde was easily converted to aldehyde 179 which was homologated under standard Wittig condition to alkene 180. This latter was regioselectively oxidated with Se02 to the corresponding £-allylic alcohol 181 which, in turn, was transformed to the corresponding allylic bromide 154 (Scheme 9). The C1-C4 fragment of Adda was prepared

1227 (Scheme 10) from (/?)-3-pentyn-2-ol through a highly efficient method previously reported by the same authors [164]. The /^-Boc-L-glycine ester 182, obtained in two steps from the above alkynol, was subjected to a Kazmaier's modification of the Ireland-Claisen rearrangement to give, after esterification, the methyl ester 183. Standard ftinctional groups elaboration afforded the aldehyde 177, which was used in the olefmation reaction to give A^-Boc-Adda methyl ester.The methyl ester 183 was also transformed in the a-methyl P-pentafluorophenyl-A^-Boc-methyl aspartate residue 184, ready for the subsequent fragment coupUng, through oxidative cleavage (RuCl3/NaI04) of the double bond and conversion of the resulting carboxylic acid to its pentafluorophenyl ester. After completion of the synthesis of the key monomeric fragments, several routes to motuporin were explored. They differ in the order of fragment assembly and in the timing of the dehydratation step to install the mdhb double bond. It was found that macrolactamization of linear peptides containing the A^-methyldehydrobutyrine residue proceeds primarily with epimerization, whereas, as reported previously by Schreiber, the same process was cleaner using the dehydroamino acid residue masked as a threonine. The synthesis of 5-[L-Ala]-motuporin, in which the dehydroamino acid residue was replaced by L-alanine, was also performed in order to explore the role of this dehydroamino acid to the biochemical activity of motuporin. Some epimerization was also observed during the cycUzation of the alanine-containing peptide. 5-[L-Ala]motuporin was found to inhibit PPl activity with IC50 values comparable to those found for motuporin itself so suggesting that the A^methyldehydrobutyrine residue is not essential for the inhibition of PPl [155b]. Swinholide A (68) and congeners The limited natural supply and the remarkable anticancer properties, together with the spectacular molecular architecture made the swinhoUde A (68) and its congeners popular targets for synthetic efforts which cuhninated with two excellent and flexible total syntheses of swinholide. The first total synthesis of swinholide was achieved by Paterson and its group at Cambrige in 1994 [165] shortly after the completion of the synthesis of the monomeric seco acid pre-swinholide A (73) [166] Nicolaou and co-workers reported an alternative total synthesis of swinholide A (68) in 1996 [167]. Paterson total synthesis was already reviewed in 1995 by Paterson itself in the review 'Total Synthesis of Bioactive Marine MacroUdes" [168], that also included detailed accounts of the alternative strategy and the synthesis of significant segments of swinholide A at that time reported by Nicolaou.Rather than duplicate this review, we will not report detailed analysis of the two synthetic planes but we will show only the general strategies, focusing our discussion on the

1228 key steps which allowed the stereospecific construction of a so complex molecule. As reported in Fig. (4), the synthesis by Paterson et al was based on the selective deprotection and regiocontroUed dimerization of 185, a fully

Fig. (4). Paterson's retrosynthesis of swinholide A (68)

protected version of the monomeric seco-acid preswinholide A. The carbon skeleton of this derivative was constructed by the union of the ClC15 aldehyde segment 186 [169] and the C19-C32 aldehyde segment 187 [170] using a suitable butanone synton as a linking unit [171]. Owing to the polypropionate nature of the skeleton of preswinholide A, the stereocontrolled synthesis of this derivative was achieved through a correct choice of different types of asymmetric aldol reactions which allowed the stereoselective formation of the C6-C7, C12-C13, C15-C16, C18-C19 and C22-C23 bonds. The final macrolactonization steps

1229 affording swinholide A from the monomelic secoacid were proved to be very regioselective and highly yielding (Scheme 11).

ester at C-23 (191) 190:191=2:1 r- 190 P^=H, P^=Me. P^=Steu2 U-192 P^=TBS, p2=P^=H P^O "O

OMe Protected Swinholide 193 a) 2,4.6-Cl3(C6H2)COCI EtaN, DMAP, Yamaguchi b) DCC, DMAP, DMAPHCI, Keck

+

1 1

1

Protected 1 isoswinholide 1 194

Scheme 11. Paterson swinholide A and isoswinhoUde A synthesis (final steps).

The resulting ring size (44- vs 46- membered) was simply controlled by the macrolactonization conditions employed without differential C21 and C23 hydroxyl groups protection. Hydrolysis of the terminal methyl ester in 185 gave the acid 188 which selectively esterified, under Yamaguchi conditions [172], the C21 hydroxyl group of the diol 189 obtained by removal of the silylene group in 185. A 2:1 mixture of the desired C21

1230 ester 190 and its C23 regioisomer 191 was obtained. This latter by-product could be recycled by methanolysis to give back diol 189 and the fully protected methyl ester 185. Then a silyl C23 hydroxyl protection was made before sUylene removal and selective hydrolysis of the terminal methyl ester to give the dimeric secoacid 192. This protection was needed in order to prevent cleavage of the C21 ester linkage. The regioselective macrolactonization of 192 to generate the desired 44-membered ring was facile and high yielding. Under Yamaguchi conditions [172] [a) conditions in Scheme 11], a 86:14 mixture of desired 44-membered ring 193 over the 46-membered ring 194 was obtained. This ring-size selectivity may be reversed by changing to the Keck DCC protocol [b) conditions in Scheme 11] [173]; in facts in this case a 9:1 mixture of 194 and 193 was obtained permitting selective formation of isoswinhoUde ring. Total deprotection of 193 and 194 completed the syntheses of swinholide A (68) and isoswinholide (71), respectively. Fig. (5) shows the retrosynthetic analysis on which is based the OMe Dithiane-cyclic sulphate coupling ^OK Macrolactonization; esterification

Ghosez ! cyclization r ^ o

Ghosez ^'•,^^^ cyclization OH 6 ^ ' OH

MeO,

Wadsworth-Horner Emmons reaction

OH Dithiane-cyclic sulphate coupling

OMe Fig. (5). Retrosynthetic dissection of swinholide A (68) according to Nicolaou

alternative synthetic strategy proposed by Nicolaou. As in Paterson's protocol the final key reactions affording the 44-membered macrodioUde are an esterification and eventually a macrolactonization of the resulting hydroxy acid. Also in this case, both reactions were performed in satisfactory yields by using the Yamaguchi protocol. Diastereoselective formation of the C2-C3 double bond was achieved through extension of the C"3 aldehyde via a Wadsworth-Homer-Emmons olefination reaction. The C3-C32 segment was constructed by coupling of the C3-C17 segment and the C18-C32 segment through the reaction of the C18-C19 cyclic sulphate with the Hthio derivative of the dithiane of the C-17

1231 aldehyde. The C3-C17 fragment was prepared from (5)-dimethyl malate in a synthetic sequence that features a Ghosez cyclization [174] for the assembly of the dihydropyran system. The C18-C32 segment was prepared from L-rhamnose which fiimished three of the nine stereogenic centers present in this fragment. The other stereogenic centers were installed using a combination of reagent-controlled, auxihary-controlled and substrate-controlled reactions. Onnamide A (74) The total synthesis of onnamide A (74) was achieved by Hong and Kishi in 1991 [175] after the completion of the synthesis of mycalamides A and B, reported by the same authors [176] and the synthesis of pederin, reported by Nakata and co-workers [177], Retrosynthetically Fig. (6), onnamide A can be divided in four Onnamide A (74)

MeO ?DMPM

OMe

195 "left hand subunit"

196 "right hand subunit" ?02

TMS-

-CHgCHaLi 197

o

T

AcN

198 Fig. (6). Kishi's retrosynthesis of onnamide A (74)

fragments: the C20-C26 "left hand subunit", 195, corresponding to the pederic acid fragment, the "right hand subunit" C10-C19, 196, containing the interesting acylaminalfimctionalgroup, that is presumed to play a key role in determining the biological activity, the lithium derivative 197, corresponding to the C6-C9 fragment, and the L-arginine-containing side chain 198. The synthesis of the right hand of onnamide (Scheme 12) was modified with respect to that proposed for the construction of the same subunit in mycalamides A and B. The starting material was the ketone

1232 199, prepared according to the literature [178] subjected to reduction and OMe

OMe H/J

P" x^o

1.NaBH(0Ac)3 CeCla/MeOH

H/,

2. Mel/NaH/THF

6.,^.6

OMe

OMe

_ ^ _^

DMe AcO

Ph 200

I.OSO4/ S,S-Corey Ijgand

CH2=CHCH2TMS I

TMSOTf/BF30Et2 MeCN

OBn 201

Q-

2. lm2CO DMe

DMe

DMe 205, X=N3 196,X=NH2 Scheme 12. Synthesis of the C10-C19 fragment of onnamide A (196).

methylation to give the desired methyl ether 200. Several hydride reductions were tested and only sodium triacetoxyborohydride in the presence of CeCls was proved to furnish the desired diastereomer. Protecting group manipulation afforded 201 which was subjected to a Cglycosidation [179] to give exclusively the expected, axially substituted product 202. Corey asymmetric osmylation of 202, followed by carbonate formation afforded 203 which was transformed in the dimethyl acetal 204, identical with the intermediate used in the synthesis of mycalamides A and B. Standard functional groups manipulation allowed for the transformation of 204 into the azide 205, obtained as a 2:1 C-18 diastereomeric mixture favouring the natural configuration. The corresponding amine 196 ("right hand subunit"), obtained by hydrogenation, was configurationaJiy unstable under acidic, basic and

1233 neutral conditions. The "left hand subunit" was prepared in two steps starting ft-om 206, an intermediate in the Nakata's total synthesis of pederin [177] and coupled with the amine 196 giving a separable C-18 diastereomeric mixture (Scheme 13). The unnatural C(18)-p-epimer may be recycled by equilibration in the presence of f-BuOK/THF at reflux. The obtained carbonate 207 was transformed in the corresponding C-

MeO q^

VT^

^OMe

TMS

206

"II

MeO

1^^ TMSCgCCHgCH^Li^^o,"

II

''bMe

'^,^Sn(/hBu)3

+ MeO

OAc r

H AcHN.^ ^N AcN

198 Pd(Ph3)^

I Onnamide A (73)

Scheme 13. Final steps toward onnamide A

10/C-ll epoxide 208, which was subjected to a mixed cuprate addition to give elonged derivative 209. A straightforward fiinctional group manipulation allowed for the transformation of 209 into the vinyltri-nbutyktannane 210, suitable for the synthesis of the C2/C-7 triene of onnamide. Pd(0)-catalyzed coupling of the stannane 210 with 5-iodo amide 198 gave the coupling product as a mixture of geometric isomers. This latter was transformed by using standard procedures to the synthetic onnamide A. The spectroscopic data of the synthetic material was found

1234 to be identical with those of the authentic samples from natural source, thus confirming the relative and absolute stereochemistry, previously assigned by analogy to mycalamides and pederin. CalycuBns (100-102) The calyculins have aroused considerable synthetic interest due not only to their unique polyfunctional structure but also to their intriguing biological activities[180]. However, initial synthetic efforts were addressed to (+)-calyculin (arbitrarily depicted in the 1986 original paper), which is enantiomeric to the natural occurring compound. The Evans group reported the total synthesis of ewf-calyculin A in 1992 [181], which allowed to establish the actual absolute configuration of the natural product. In 1994 Masamune et al published the total synthesis of the natural enantiomer [182]. Shoiri et al have also reported a formal total synthesis of calyculin A [183]. Very recently an alternative synthesis of (+)-calyculin A and (-)-calyculin B was reported by Smith [184] whereas the natural calyculin C was obtained by Armstrong and co-workers [185] Owing to the impressive worlc in this field, rather than being comprehensive, the following discussion will be limited to the description of the general synthetic strategies and of the key synthetic steps. Retrosynthetic analysis of calyculins backbone revealed several possibilities for disconnection involving carbon-carbon bonds. All the synthetic strategies were centred on the initial disconnection at C25-C26 double bond which divided the natural product into two halves of similar functional complexity. Model studies [180d,180m] indicated that Wittig olefination using stabilized ylides derived from C26-C37 phosphonium salt gave the best results in terms of (£^-selectivity and overall yield. The late introduction of the photolabile and easily isomerizable C1-C9 cyanotetraene moiety and of the Cn-phosphate group is another common feature of all synthetic plans. Main disconnections of the Evans' total synthesis of ^n/-calyculin A are depicted in Fig. (7). The key building blocks C1-C25 and C26-C37 were elaborated in homochiral form and eventually coupled in the final stages of the synthesis through a Wittig reaction of the C25 aldehyde with the C26 phosphonium salt. All the synthetic strategy is based on an extensive use of well known author's auxiliary-controlled asymmetric synthesis. 10 of the 15 stereogenic centers (Cio, C12, C13, Cn, C22, C23, C30, C34, C35, €35) were incorporated through auxiliary-based asymmetric aldol, alkylation, hydroxylation, and Michael reactions, whereas the remaining chirality arose from substrate-controlled induction. Truncation of the C1-C25 subunit, Fig. (8), at the three indicated sites revealed four smaller fragments which were coupled according to Eq. (a). Aldol coupling of the C13-C20 methyl ketone (fragment C, 213) with the C21-C25 aldehyde subunit (fragment D, 214) with complete Felkin-Anh stereocontrol of the

1235 ^—ammide coupling OH

MegN

o

/

OH

^Stille

J

OHXOH

OM©

Homer-Emmons ^ ^ Fig. (7). Evans' retrosynthetic analysis of enr-calyculin A O

1CN

213, C 214, D Fig. (8). Evans* retrosynthetic analysis of the C1-C25 subunit of enr-calyculin A

B C + D — • CD

A •

BCD

• ABCD

(a)

newly generated C21 stereocenter was achieved (Scheme 14) using the Mukaiyama-type aldol reaction between the trimethylsUyl enol ether 215 (corresponding to the fragment C) and aldehyde 216. Exposure of the aldol adduct 217 to HF/MeCN/H20 led the desired spiroketal 218, along

1236 with minor amounts of its epimer at the C19 stereocenter. The spiroketal OTES TESO

n 21 : 23

48% HF CH3CN, H2O

Piv0^3xv^i5>o OMe O P M B O T M S BFaOEta

215

MeO

OPMBO

Mukaijama aldol TBSO

O H OR

217

'OTBS

O

O

O

OH OMe OTBS

spiroketalization

O OTICI4 220 ^ aldol

OH 5 H 5Me

221

XP\x<S.X^vi>'^v^^^O* ^ d

OH OAc OMe

^OTBS

H^^^<>^^x^Nsi3^/\^,x^O 6

OPi OPi OMe

P^=TBS

222 223 Scheme 14. Evans synthesis of the C9-C25 fragment of en/-calyculin A.

218 was then trasformed into the C13 aldehyde 219, the appendage point for the C10-C13 dipropionate subunit (fragment B). Aldol addition of the titanium enolate 220 to the above aldehyde afforded the adduct 221 with wrong stereochemistry at C13. After reduction of the Cn ketone, the C13 hydroxy group was regioselectively inverted under Mitsunobu condition. Conventional elaboration of the acyl group at C9 in 222 afforded the aldehyde 223 to which the tetraene moiety was attached through the sequential application of Homer-Emmons and Stille reactions. Disconnection of the nitrogen rich C26-C37 subunit at the amide junction led the C33-C37 aminoacid and the C26-C32 y-amino azole subunit, Fig. (9). Further disconnection at the C34-C35 bond of the y aminoacid fragment through an anti glycolate aldol reaction fiimished the Dserinale derivative 224. The 2,4 disubstituted oxazole system was obtained by cyclodehydratation and oxidation of the amide alcohol

1237 derivative 225, The stereocenter at C-30 was installed using a chiral OR OH

O U

I

NMeg OR

OR

O

MeO

33^0H NMe2 OR cyclodehydration pxydative aromatization

anti-aldol reaction

MeO*

BocHN

H N. ^COaMe

30^ O

NH2

CH2OH

225 L-serine methyl ester

224

BocHN

OH

Michael addition Curtius rearrangment

Fig. (9). Evans' retrosynthetic approach to the C26-C37 subunit of enr-calyculin A

auxiliary-controlled Michael addition of the enolate derived from propionic acid to tert-butyl acrilate. Even if the retrosynthetic dissection of (-)-calyculin A (100) by Masamune, Fig (10), parallels that by Evans it differs for the timing of coupling of individual fragments (the order was B->BC-^A'BC->AA'BC) and for the approach to the key C9-C25 (228) and C26-C37 fragments (229). The spiroketal subunit (Scheme 15) was constructed starting from 230 through a) Claisen condensation with the litium enolate derived from methyl isobutirrate; b) stereoselective reduction of the Cn ketone, c) standard functional group elaboration to fiimish the silyl enol ether 231 d) Ti^^-mediated Mukaiyama-type aldol reaction with aldehyde 232 e) trasformation into aldehyde 233 f) aldol

1238 reaction of the above aldehyde with the E(0) enolate 234. The OH O

C

=

'^^nr*^K^^°) MegN

OH

OH

O ''/OR

229, C

C3HO

OR

OR

OMe

228, B

O

226, A 227, A' Fig. (10). Masamune's retrosynthetic analysis of calyculin A (100)

17 COgMe

V

f

Methyl Isobutyrate; V ^ ' I / Q ' ^ " 2. MPMBr,

^ 2 3 ^ ^ ^OBn OTrO^O

OTr

OMe-^p,0^^^ 236

21

^ 25 OTES

O ^

OMe g^p^O

238

237

Scheme 15. Masamune synthesis of the C9-C25 subunit of calyculin.

oH

OTES

1239 anthanthanti diastereoisomer 235 was obtained with a 12:1 diastereoselection thanks to the extraordinarily high diastereoeselectivity of the enolate 234 that overrode that (1:4) of the aldehyde 233. With all the carbon in place 235 was trasformed into methyl ketone 236 with a set of standard reactions. The final aldol reaction of cyclohexyl boron enolate of methyl ketone 236 with the aldehyde 237 furnished the acycUc intermediate 238 corresponding to the C9-C25 spiroketal unit. The C33C37 Y-aminoacid was obtained from D-galactose [180f], whereas the 2,4disubstituted oxazole was obtained through condensation of the amide 239 with bromopyruvate, followed by dehydratation (Scheme 16). Conventional functional group elaboration provided the amino-oxazole fragment in excellent overall yield. Smith's retrosynthetic approach, Fig. (11), to en^calyculin A and to ()-calyculin B (the C2 geometrical isomer of calycuUn A) begins with the disconnection at C(2) via a Peterson olefination which would provide both molecules from a common advanced intermediate. Further

^ 239

iX.

2. TFAA ,

'COgEt

S26

OMPM Scheme 16. Masamune synthesis of the C26-C32 subunit of calyculin A.

disconnection as indicated by dashed lines generated five fragments: the phosphonate A (240), the bromoolefin B (241), the spiroketal epoxide C (242), the oxazole D (243) and the lactam E (244). Fig. (12) shows the retrosynthetic approach to the pivotal spiroketal C (242) [180o]. Model synthetic studies [180rl indicated that the stereocontroUed formation of the C(14,15) a epoxide would be achieved via Payne epoxidation of the vinyl spiroketal 245. The corresponding open-chain precursor 246 was envisioned to arise via coupling of epoxide 247 with the sterically hindered ditiane 248. This latter, in turn, was obtained in a three step sequence from alcohol 249, readily available from f5j-dimethyl malate [186]. The epoxide 247 was elaborated (Scheme 17) via Brown asymmetric addition of (Z)-crotyldiisopinocamphenylborane to aldehyde 250, followed by a highly diastereoseiective IBr-induced iodocarbonate cycHzation of the homoallylic carbonate 251 [187]. Treatment of iodocarbonate 252 with K2CO3 (3 eq) in dry MeOH and protection fiimished the desired epoxide 247. The y amino acid segment E 244 was synthetized [180p] (Scheme 18) starting from the known lactone 253 through a) addition of the Weinreb reagent of p-methoxybenzylamine, b)

1240

oxidation to the hemiaminai 255, c) nucleophilic addition of the TMS enol ether of pinacolone to N-acyliminium cation derived from 255, d) conversion of the ketone 256 to a silyi enol ether, ozonolysis with reductive workup affording 257 e) 0-methylation of the primary alcohol =

amide coupling

OH O I

I

MeO* NMe^H WIttig—"HornerEmmons 1CN Peterson

A '

OH^OH\OMe B' w mixed-cuprate coupling

PhosphorusFG

13^ Br O

240, A

•Try

242, C

241, B

Boc I

MeOgC,

N3

OBn

N^.O

\_7

''/

X

243, D

244, E

Fig. (11). Smith's retrosynthesis of enr-calyculin A and (-)-calyculm B

f) N-acylation of the amide to provide the coupling precursor E 244. Recently Armstrong and co-workers reported the total synthesis of calyculin C (101) which differs from the parent compound for the presence of an additional methyl group at C-32. As in other approaches, the initial disconnection at C25-C26 double bond through a Wittig olefination afforded the fragments C1-C25 and C26-C35. The synthesis of the C1-C25 subunit (Scheme 19) involved the initial formation of the C15-C25 spiroketal fragment 259 to which extension of the polypropionate chain was achieved via three subsequent steroselective allylborations. So, the aldehyde 259 was subjected to an allylation reaction via addition of allylmagnesium bromide-ZnCl2 mixture. The addition proceeded with complete Felkin-Anh substrate control affording the wrong stereochemistry at C-15 (compound 260), which was inverted through an oxidation (Swem)-reduction (LiBH^) two-steps procedure. After methylation of the C-15 hydroxyl group, ozonolysis of double bond

1241

PMBO

OTBSOTBS

Fragment C c=:> '^^^^A-cT ^^ '^ = i > < ^ ' y X > J^^^^^^^^OBn 242 Payne spiroketalizatjon HO ^ ^ o>ic " epoxidation 245 246 OTBS aoV^^^^^^'^^OBn = 247 PMR. dithiane coupling / ^ Y I

«

;>

.

s

H^ ^"^^ ^OBn

J>

248

249

Fig. (12). Smith's retrosynthetic approach to the C14-C25 fragment of enr-calyculin A

O

1 \yr-B(lpc)2 250

2-^202 OTBS OBn

iodocarbonate cyclization 252

247

Scheme 17. Smith synthesis of the C20-C25 fragment of enr-calyculin A. PMB

i V

a

H ^ J^

2. AC2O

t \

OTMS PMB

BFaEtgO

/

^

253

254 255 . 256

.^N^ O N ,^ BQN. T H F , ' ' ^ ' v ^ O H N a H , MeJ Vjv/^OMedeprotectlon^ ^^^v/^OMe 2.03, NaBH4 d

|-C 0.^0

e

*^ r \ 0^0

Boc-protection i i ^ 0^0

257 258 Scheme 18. Smith synthesis of the C33-C37 fragment of enr-calyculin A.

in 261 afforded an aldehyde which was subjected to an asymmetric Brown crotylboration reaction which occurred with complete diastereoselectivity, affording, after benzoylation compound 262. More

1242 ^OTBS

^OTBS

MEMO/,

1. Swern 2. LIBH4 3. methylation

MEMO/,^

ZnClp

OBz

^^^ 260

^OTBS

^OTBS

I.O3

BzCI OBz

OBz

2.

OMe

OBz OMe 262

-B(lpc)2

261

derived from (+)-lpc2BOMe, BF3-Et20 .OTBS

1.00

C1-C25 fragment several steps -B(lpc)2

II

OH

derived from (+)-lpc2B0IVIe BF3-Et20

OBz OMe

263

Scheme 19. Annstrong synthesis of the C1-C25 fragment of calyculin C (101).

problematic was the introduction of the ClO-Cll propionate unit. In fact, the subsequent ozonolysis-Brown crotylboration sequence was found to proceed with modest diastereoselectivity to yield a 35:39 mixture of homoallylic alcohols favouring the undesired anti-syn-anti isomer. After ozonolysis of terminal double bond, the introduction of the C1-C8 tetraene moiety in 263 was achieved following the same procedure reported by Evans in its synthesis of ^n^calyculin A. As in other synthetic approaches to calycuHns, the C26-C37 fragment of calyculin C (101) was initially disconnected at C33 amide bond to give the y-aminoacid C33C37 and the C26-C32 aminooxazole units. The first one was obtained in a straightforward manner from Z>-lixose whereas the aminooxazole unit was synthesised starting from L-pyroglutamic acid. Scheme 20 outlines key steps of the synthesis of the aminooxazole moiety. The endomethylation of the bicyclic MO-acetal 264, derived from L-pyroglutamic 1. pTSA .0 264

CH3I

2. MsCI, NEt3 265

^

(Boc)^^^"^^^^^'^^'''^^^^ NH2

267

32%

3. BugSnH, Nal, DME

1,3-Dichloroacetone^

266

(Boc)HN^3N/\j-'0.

K2CO3

N ^ 268

Scheme 20. Armstrong synthesis of the C26-C32fragmentof calyculin C (101).

1243 acid, was obtained with a 60% de using LDA followed by CH3L The C32 methyl group arose from radical deoxygenation of an in i-Z^w-generated iodide obtained from 265. The dimethyl-pirrolidone 266 was converted to the open chain amide 267 via N-Boc-protection followed by Weinreb aluminium-amide opening. The required oxazole fragment 268 was obtained through condensation of amide 267 with 1,3-dichloroacetone under vigorous reflux. Discodermolide (104) Owing to its exceptional pharmacological potential and scarcity of natural material [0.002% (w/w) from frozen marine sponge], discodermolide have stimulated intensive synthetic efforts [188]. The total synthesis of both the unnatural [189] and natural antipodes [190] [(-)- and (+)-104, respectively] by Schreiber and co-workers allowed to establish the correct absolute configuration.

PhS

Me

TBSO

Me

Me

TBSO

OH 272

Me

Me

OH

TBSO

273

n

\

Me

OH 272

^

Me (+)-3-hydroxy-2-mGthylpropionate I^^^COgMe

"^^^'^y' ®s^®''

OH Figure (13)- Schreiber's retrosynthetic analysis of discodermolide (104).

1244 To date, three syntheses of (-)-disco(lermolide [189,191] and two of the natural antipode [190,192] have been described. In this review we will report only the syntheses of the natural product (+)-discodermoUde (104). Very recently an alternative total synthesis of the natural (+)discodermolide has been reported by Paterson [193]. Schreiber and co-workers used an highly convergent approach, depicted in Fig. (13), which allowed the straightforward preparation not only of both enantiomers of discodermolide but also of a radiolabeled variant as well as of several structural variants of the natural product that should be useful in further characterizing the interaction of discodermolide with its receptor. From the retrosynthetic perspective, dissection of the skeleton at indicated sites revealed three fragments (269, 270, 271) of roughly equal complexity The repeating stereochemical triad embedded in the discodermolide backbone suggested that these fragment could arise from the homoallyUc alcohols 272 and 273 that, in turn, originated from a common precursor through asymmetric crotylboration addition to (+)-3hydroxy-methypropionate methyl ester. The homoallyUc alcohol 272 was oxidatively cleaved (Scheme 21) and homologated to the £-enoate 274 via a Wittig olefination. The hemiacetal obtained through reaction with benzaldehyde furnished the Michael adduct 276 with complete Me Me

Me Me

- Dur^ur* M® Me

^^'-^sXX^ I.O3, DMS ^ r^-'^V^^^^'^^COgMe TBSO OH " 2. Ph3P=CHC02Me JBSO OH 272

base ^ |-^'^V^'S;^C02Me 2.HF OR O ^ O

275 OHO

LDess-Martin 3. CSA

^.^

^-ti

Me^S^'/Me^ OH 278

OK/I^

PH^ n^J

Ph 276,R=TBS 277, R=H

Me^V^'^Me OTBS 269

Scheme 21. Schreiber synthesis of the C1-C7 fragment of discodermoUde (269).

Stereoselectivity. After deprotection of the primary alcohol and DessMartin oxidation of compound 277 to the corresponding aldehyde, the acetal 278 was obtained as a mixture of anomers through spontaneous cycUzation of the dimethyl acetal of the C-1 aldehyde with the 5-hydroxy group obtained by cleavage of the benzylidene acetal. The lactone ftinctionality at C-1 of discodermoUde was masked in its reduced form as thiophenyl acetal 269 obtained by treatment of methyl acetal with PhSSiMes, Znl2 and BU4NI. Elaboration of the C8-C15 fragment was achieved (Scheme 22) from the homoallyUc alcohol 273 in a high yielding three steps procedure involving a) silyl protection b) oxidative cleavage of the terminal double bond and subsequent construction of the

1245 cis-trisubstituted olefin by the method of Still and Gennari [194] (condensation of the obtained aldehyde with the potassium salt of Me

TBSO

Me

OH

"

Me

Me

15V.0H

^OPiv

-i^pjvCI

TBSO

273 •*(OMe)2 Scheme 22. Schreiber synthesis of the C8-C15 fragment of discodermolide.

(CF3CH20)POCH(Me)C02Me) and c) reduction of the carbomethoxy group to a hydroxymethyl group, affording compound 279. After protection of the C-15 hydroxyl group and selective deprotection of the C-9 primary alcohol, the acetylene derivative 280 was obtained via Swem oxidation and one carbon homologation using Gilbert's reagent. The third fragment was elaborated (Scheme 23) starting from the /7-methoxybenzyl ether derivative (281) of the homoallyUc alcohol 272. Ozonolysis of the terminal double bond fiimished an intermediate aldehyde which was converted to the diene 283 via Wittig reaction with Me

Me

1.03;NaHMDS

1.Pd{PPh3)3 CH2=CHZnBr

2. Ph3P''CH2l r TBSO

TBSO

HO

OPMB

OPMB

2.TFA

OPMB

283 271 Scheme 23. Schreiber synthesis of the C16-C24 fragment of discodermolide.

(iodomethylene)triphenylphosphorane followed by a palladium-catalyzed coupling of the resulting c/^-vinyl iodide 282 with vinylzinc bromide. Conventional functional group elaboration afforded the methyl ketone 271. The construction of discodermolide backbone, starting from the previously synthesized fragments had required a carefiil setting of proper coupling strategy and of a correct sequence of events (alkylation, reductions, introduction of the C19 carbamoyl functionality that would avoid functional groups' incompatibility and would furnish the best substrate control in order to obtain the desired stereochemical outcome (Scheme 24). The formation of the C7-C8 bond involved the NiCyCrCt mediated coupling of aldehydes and iodoacetylenes developed by Nozaki

1246 and Kishi [195]. The iodoacetylene 284, derived from 280 by treatment with iodine and morpholine, was coupled with aldehyde 269 yielding the propargilic alcohols 285 as a 2:1 mixture of epimers at C-7 favouring the desired one. Conventional functional groups elaboration afforded the allylic bromide 286, which was subjected to an alkylation reaction with the Hthium enolate of the ketone 271, to form the C15-C16 bond. The CMe

Me

^OPiv

.OPIv Me

Me 15^

269, NiCl2/CrCl2

^Me

Scheme 24. Schreiber's coupling strategy toward discodermolide.

16 methyl group was introduced with correct stereochemistry through a second alkylation reaction with methyl iodide, thus affording the full carbon backbone of discodermoUde (104). The final elaboration to discodermolide proceeded by transformation of the thiophenyl acetal at C-1 to the corresponding lactone, DDQ deprotection of the PMB ether, carbamoylation of the C-19 alcohol and reduction of the C-17 ketone. It was observed that reduction occurred with excellent selectivity for the correct R configuration thanks to the neighbouring carbamate group which provides significant chelation control. The highly convergent approach adopted by Marshall and co-workers in the total synthesis of (-i-)-discodermoUde (104) involves elaboration of three stereotriads, Fig. (14), through addition of chiral allenyltin, -indium, and -zinc reagents (of general structure 297) to (5)-2-methyl-3-silyloxypropanal (296). The resulting antUsyn- or ^-yn^^yn-homopropargylic alcohols adducts (294 and 295) were reduced to the corresponding (F)homoallylic alcohols (292 and 293) which, in turn, were subjected to asymmetric Sharpless epoxidation. Addition of Red-Al or methyl cuprate reagents to the resultant epoxy alcohols (290 and 291), respectively, afforded the key precursors aldehyde 287 and alcohol (289), after

1247 conventional functional group elaboration. Alkyne 288 arose from direct addition of the chiral allenylzinc reagent 297 (R=H, M=EtZn) to aldehyde 296 (Ri=TES). Formation of the C7-C8 bond was obtained through addition of the alkyne 288 (as the lithio species) to the aldehyde 287 followed by Lindlar hydrogenation and installation of appropriate protecting groups.

Me

i , ^

Me

Me

OH

Me

OCONH2

discodermolide (104) Me

Me

Me

.OSiEta

OR2

OR2

291 llsharpless Me

OR2

Me

0R2

293

B

reduction

SI ^n

syn

M€\ ( UQ 15^^^

OR2

121

OR2

OPIv

295

294 R=CH20Ac

« < . RiO^ OHO

296

CHaOPiv

Ri=TBS R2=PMP

M

Me/,^

•H 297

Fig. (14). Marshall's synthetic plan for discodermolide (104, Ri= TBS, R2 = PMP)

1248 The 13-hydroxymethyl group in the CI-CI 3 coupled subunit was converted to an aldehyde which was subjected to a Wittig reaction with a-iodoethyUdene triphenylphosphorane to give a (Z)-vinylic iodide (formation of the C13-C14 double bond). A palladium catalyzed Suzuki coupling of this vinylic iodide with a boranate derived from alcohol 289 led to the formation of the C14-C15 bond, thus completing the backbone of discodermolide. Final elaboration to the natural product involves cleavage of cyclic PMP acetal at C-1, transformation of the primary hydroxy group to a methoxycarbonyl, installation of the C-19 carbamate function, cleavage of the remaining alcohol protecting groups and lactonization. ABBREVIATIONS Ac Adda

= =

Bn Boc BOM Bz Cp DCC DDQ DIBAL-H DMAP DMPM DMS HMDS Ipc k-Arg LDA Mdhb MeCtrp METrp MEM MPM MOM NAZA NMO

= = = = = = z=

= = = = = = = = = = = = = = =

1

Acetyl (25, 35, 85, 95)-3-Amino-9-methoxy-2,6,8 trimethyl- 10-phenyl-4,6-decadienoic acid Benzyl f^r^Butyloxycarbonyl Benzyloxymethyl Benzoyl Cyclopentadienyl MA^*-Dicyclohexylcarbodiimide 2,3-Dichloro-5,6-dicyano-1,4-benzoquinone Diisobutylaluminium hydride 4-Dimethylaminopyridine 3,4-Dimethoxybenzyl Dimethyl sulfide bis(trimethylsilyl)Amide Isopinocampheyl a-Ketohomoarginine Lithium diisopropylamide (A^-methylamino)dehydrobutyrate 6-Chloro-A^-methytryptophan Methyltryptophan 2-Methoxyethoxymethyl 4-methoxyphenylmethyl Methoxymethyl Nazumamide A 4-Methylmorpholine-A^-oxide

1249

OTf Piv PMP PPl Red-Al TBAF TBDPS TBS Tee TES TFA TFAA THF TMS TPAP Tr Ts pTSA V-Tyr Xc

z

Trifluoromethanesulfonate Pivaloyl 4-Methoxyphenyl Phosphatase 1 Sodium bis(2-methoxyethoxy)aluminium hydride Tetra-n-butylammonium fluoride ^^rf-Butyldiphenylsilyl ^^rf-Butyldimethylsilyl Trichloroethyl Triethylsilyl Trifluoroacetic acid Trifluoroacetic anhydride Tetrahydrofuran Trimethylsilyl Tetra-n-propylammonium pemithenate Trityl 4-Toluenesulfonyl 4-Toluensulfonic acid Vinylogous tyrosine Chiral auxiliary carbobenzoxy

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9]

Fusetani, N.; Matsunaga, S. Chem. Rev., 1993, 93, 1793-1809. Bewley, C. A.; Faulkner, D. J. Angew. Chem. Int. Ed. 1998, 57, 2162-2178. Kho, E.; Imagawa, D. K.; Rohmer, M.; Kashman, Y.; Djerassi, C. J. Org. Chem. 1981, 46, 1836-1839. Kobayashi, M.; Kawazoe, K.; Katori, T.; Kitagawa, I. Chem. Pharm. Bull. 1992, 40, 1773-1778. Inouye, Y.; Sugo, Y.; Kusumi, T.; Fusetani, N. Chemistry Letters 1994, 419420. Sugo, Y.; Inouye, Y.; Nakayama, N. Steroids 1995, 60, 738-742. Umeyama, A.; Shoji, N.; Enoki, M.; Arihara, S. / Nat. Prod 1997, 60, 296298. Fusetani, N.; Matsunaga, S.; Matsumoto, H.; Takebayashi, Y. J. Am. Chem. Soc. 1990,112, 7053-7054. Lee, A. Y.; Hagihara, M.; Karmacharya, R.; Albers, M. W.; Schreiber, S. L.; Clardy, J. J. Am. Chem. Soc. 1993, 775,12619-12620.

1250

[10] [11] [12] [13] [14] [15] [16]

[17] [18] [19] [20] [21]

[22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32]

Nakao, Y.; Matsunaga, S.; Fusetani, N. Bioorg. Med. Chem. 1995, 5, 11151122. Nakao, Y.; Oku, N.; Matsunaga, S.; Fusetani, N. J. Nat. Prod. 1998, 61, 667670. Lewis, S. D.; Ng, A. S.; Baldwin, J. J.; Fusetani, N.; Naylor, A. M.; Shafer, J. A. Tromb. Res. 1993, 70,170-190. Hagihara, M.; Schreiber, S. L. J. Am. Chem. Soc. 1992,114, 6570-6571. Wipf, P.; Kim, H. J. Org. Chem. 1993,58,5592-5594; Corrigendum: ibid.1994, 59,2914. Deng, J.; Hamada, Y.; Shoiri, T.; Matsunaga, S.; Fusetani, N. Angew. Chem. Int. Ed. Engl. 1994, 33, 1729-1731. (a) Maryanoff, B. E.; Qiu, X. Y.; Padmanabhan, K. P.; Tulinsky, A.; Almond, H. R.; Andrade-Gordon, P.; Greco, M. N.; Kauffinan, J. A.; Nicolau, K. C ; Liu, A. J.; Brungs, P. H.; Fusetani, N. Proc. Natl. Acad. Sci. U. S. A. 1993, 90, 8048. (b) Maryanoff, B. E.; Greco, M. N.; Zhang, H.-C.; Andrade-Gordon, P.; Kauf&nan, J. A.; Nicolau, K. C ; Liu, A. J.; Brungs, P. H.; J. Am. Chem. Soc. 1995,117, 9434-9453. Bastiaans, H. M. M.; van der Baan, J. L.; Ottenheijm, H. C. J. J. Org. Chem. 1997, 62, 3880-3889. Wipf, P. Chem. Rev. 1995, 95, 2115-2134. Humphrey, J. M.; Chamberlain, A. R. Chem. Rev. 1997, 97, 2243-2266. Fusetani, N.; Sugawara, T.; Matsunaga, S.; Hkota, H. J. Am. Chem. Soc. 1991, 775,7811-7812. Kabayashi, J.; Itagaki, F.; Shigemori, H.; Ishibashi, M.; Takahashi, K.; Ogura, M.; Nagasawa, S.; Nakamura, T.; Hirota, H.; Ohta, T.; Nozoe, S. J. Am. Chem. Soc. 1991,113, 7812-7813. Itagaki, F.; Shigemori, H.; Ishibashi, M.; Nakamura, T.; Sasaki, T.; Kabayashi, J.; J. Org. Chem. 1992, 57, 5540-5542. Kabayashi, J.; Itagaki, F.; Shigemori, H.; Takao, T.; Shimonishi, Y. Tetrahedron 1995, 51, 2525-2532. Sowinski, J. A.; Toogood, P. A. Chem. Commun. 1999, 981-982. Kabayashi, J.; Sato, M.; Ishibashi, M.; Shigemori, H.; Nakamura, T.; Ohizumi, Y. J. Chem. Soc, Perkin Trans. 11991, 2609-2611. Uemoto, H.; Yahiro, Y.; Shigemori, H.; Tsuda, M.; Takao, T.; Shimonishi, Y.; Kabayashi, J. Tetrahedron 1998, 54, 6719-6724. Kabayashi, J.; Sato, M.; Murayama, T.; Ishibashi, M.; Walchi, M. R.; Kanai, M.; Shoji, J.; Ohizumi, Y. J. Chem. Soc, Chem. Commun. 1991,1050-1052. Schmidt, E. W.; Harper, M. K.; Faulkner, D. J. J. Nat. Prod. 1997, 60, 779782. Chill, L.; Kashman, Y.; Schleyer, M. Tetrahedron, 1997, 53,16147-16152. Matsunaga, S.; Fusetani, N.; Hashimoto, K.; Walchli, M. J. Am. Chem. Soc 1989, 111, 2582-2588. Matsunaga, S.; Fusetani, N. J. Org. Chem. 1995, 60,1177-1181. Bewley, C. A.; Faulkner, D. J. J. Org. Chem. 1994, 59, 4849-4852.

1251

[33] [34] [35]

[36] [37] [38] [39] [40] [41]

[42]

[43] [44] [45] [46] [47] [48]

[49] [50] [51]

Schmidt, E. W.; Bewley, C. A.; Faulkner, D. J. J. Org. Chem. 1998, 63, 12541258 Gulavita, N. K.; Pomponi, S. A.; Wright, A. E.; Yarwood, D.; Sills, M. A. Tetrahedron Lett. 1994, 35, 6815-6818. a) Kitagawa, I.; Kobayashi, M.; Lee N. K.; Shibuya, H.; Kawata, Y.; Sakiyama, F. Chem. Pharm. Bull. 1986, 34, 2664-2667. b) Kitagawa, I.; Lee N. K.; Kobayashi, M.; Shibuya, H.; Chem. Pharm. Bull. 1987, 35, 2129-2132. c) Kobayashi, M.; Lee N. K.; Shibuya, H.; Momose, T.; Kitagawa, I. Chem. Pharm. Bull. 1991, 39, 1177-1184. Kobayashi, M.; Kanzaki, K.; Katayama, S.; Ohashi, K.; Okada, H.; Ikegami, S.; Kitagawa, I Chem. Pharm. Bull. 1994, 42, 1410-1415. Tsuda, M.; Shimbo, K.; Kubota, T.; Mikami, Y.; Kobayashi, J. Tetrahedron 1999,55,10305-10314. Li, Y.; Dumdei, E. J.; Blunt, J. W.; Munro, M. H. G.; Robinson, W. T.; Pannel, L. K. J. Nat. Prod 1998, 61, IIA'11%. Clark, D. P.; Carroll, J.; Naylor, S.; Crews, P. J. Org. Chem. 1998, 63, 87578764. de Silva, E. D.; Williams, D. E.; Anderser, R. J.; Klix, H.; Hohnes, C. F. B.; Allen, T. M. Tetrahedron Lett. 1992, 33, 1561-1564. Rinehart, K. L.; Harada, K.; Namikoshi, M.; Chen, C ; Harvis, C. A.; Munro, M. H. G.; Blunt, J. W.; Mulligan, P. E.; Beasley, V. R.; Dahlem, A. M.; Carmichael, W. W. J. Am. Chem. Soc. 1988,110, 8557-8558. Namikoshi, M.; Rinehart, K. L.; Sakai, R.; Stotts, R. R.; Dahlem, A. M.; Beasley, V. R.; Carmicheal, W. W.; Evans, W. R. J. Org. Chem. 1992,57, 866872 and references therein. Goldberg, J.; Huang, H. B.; Kwon, Y. G.; Greengard, P.; Nairn, A. C ; Kuriyan, J. Nature, 1995, 376, 745-753. Samy, R.; Kim, H. Y.; Brady M.; Toogood, P. L. J. Org. Chem. 1999, 64, 27112728 and references therein. Sheppeck, II, J. E.; Gauss, C. M.; Chamberlin, A. R. Bioorg. Med. Chem. 1997, 5, 1739-1750. Gupta, v.; Ogawa, A. K.; Du, X.; Houk, K. N.; Armstrong, R. W. J. Med. Chem. 1997, 40, 3199-3106. Yoshizawa, S.; Matsushima, R.; Watanabe, M. F.; Harada, K. L; Ichihara, A.; Carmicheal, W. W.; Fujiki, H.; J. Cancer Res. Clin. Oncol. 1990,116,709-714. Nishiwaki-Matsushima, R.; Otham, T.; Nishiwaki, S.; Suganuma, M.; Kohyama, K.; Ishikawa, T.; Carmicheal, W. W.; Fujiki, H. J. Cancer Res. Clin. Oncol. 1992,118, 420-424. Bonnington, L. S.; Tanaka, J.; Higa, T.;Kimura, J.; Yoshimura, Y.; Nakao, Y.; Yoshida, Y. W.; Scheuer, P. J. J. Org. Chem. 1997, 62, 7765-7767. Mauger, A. B. J. Nat. Prod 1996, 59, 1205-1211. Gerwich, W. H.; Jiang, Z. D.; Agarwal, S. K.; Fanner, B. T. Tetrahedron 1992, 48, 2313-2324.

1252

[52]

[53] [54] [55] [56] [57] [58]

[59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75]

Ford, P.W.; Gustafson, K. R.; McKee, T. C ; Shigematsu, N.; Maurizi, L. K.; Pannell, L. K.; Williams D. E.; de Silva, E. D.; Lassota, P.; Allen, T. M.; Van Soest, R.; Andersen, R. J.; Boyd, M. R. 7. Am. Chem. Soc, 1999, 121, 58995909. Fusetani, N.; Nakao, Y.; Matsunaga, S. Tetrahedron Lett. 1991,52,7073-7074. Nienaber, V. L.; Amparo, E. C. J. Am. Chem. Soc. 1996,118, 6907-6810. Hayashi, K.; Hamada, Y.; Shioiri, T. Tetrahedron Lett. 1992, 33, 5075-5076. Kundu, B.; Bauser, M.; Betschinger, J.; Kraas, W.; Jung, G. Bioorg. Med. Chem. Lett. 1998, 8, 1669-1672. Fusetani, N.; Warabi, K.; Nogota, Y.; Nakao, Y.; Matsunaga, S. Tetrahedron Lett. 1999, 40, 4687-4690. (a) Hamada, T.; Sugawara, T.; Matsunaga, S.; Fusetani, N. Tetrahedron Lett. 1994, 35, 719-720. (b) Hamada, T.; Sugawara, T.; Matsunaga, S.; Fusetani, N. Tetrahedron Lett. 1994, 35, 609-612. Nakao, Y.; Masuda, A.; Matsunaga, S.; Fusetani, N. J. Am. Chem. Soc. 1999, 121, 2425-2431. Faulkner, D.J. Nat. Prod. Rep. 1984,1, 251-551; ibidem 1986, 3,1. Matsunaga, S.; Fusetani, N.; Hashimoto, K.; Koseki, K.; Noma, M. J. Am. Chem. Soc. 1986,108, 847-849. Ishibashi, M.; Moore, R.E.; Patterson, G.M.L.; Xu, C ; Clardy, J. J. Org. Chem. 1986, 57, 5300-5306. Hirata, Y.; Uemura, D. Pure Appl. Chem. 1986, 58, 701-710. Carmely S.; Kashman, Y.Tetrahedron Lett. 1985,26, 511-514. Kobayashi, M.; Tanaka, J.; Katori, T.; Matsuura, M.; Kitagawa, I. Tetrahedron Lett. 1989, 30, 2963-2966. Kobayashi, M.; Tanaka, J.; Katori, T.; Kitagawa, I. Chem. Pharm. Bull. 1990, 38, 2960-2966. Kitagawa, I.; Kobayashi, M.; Katori, T.; Yamashita, M.; Tanaka, J.; Doi, M.; Ishida, T. /. Am. Chem. Soc. 1990,112, 3710-3712. Kobayashi, M.; Tanaka, J.; Katori, T.; Matsuura, M.; Yamashita, M.; Kitagawa, I. Chem. Pharm. Bull. 1990, 38, 2409-2418. Sakai, R.; Higa, T.; Kashman, Y. Chem. Lett. 1986,1499-1502. Kato, Y.; Fusetani, N.; Matsunaga, S.; Hashimoto, K.; Sakai R.; Higa, T.; Kashman Y. Tetrahedron Lett. 1987, 28, 6225-6228. Tanaka, J.; Higa, T.; Kobayashi, M.; Kitagawa, I. Chem. Pharm. Bull. 1990,38, 2967-2970. Tsukamoto, S.; Ishibashi, M.; Sasaki, T.; Kobayashi, J. J. Chem. Soc. Perkin Tran^ 71991, 3185-3188. Todd, J.S.; Alvi, K.A.; Crews, P. Tetrahedron Lett. 1992, 33,441-445. Kobayashi, M.; Kawazoe, K.; Okamoto, T.; Sasaki, T.; Kitagawa, I. Chem. Pharm. Bull. 1994,42,19-26. Doi, M.; Ishida, T.; Kobayashi, M.; Kitagawa, I. J. Org. Chem. 1991, 56, 36293632.

1253

[76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97]

[98] [99] [100]

Bubb, M.R.; Spector, I.; Bershadsky, A.D.; Korn, E.D. J. Biol Chem. 1995, 270, 3463. Sakemi, S.; Ichiba, T.; Kohmoto, S.; Saucy, G; Higa, T. /. Am, Chem. Soc, 1988,110, 4851-4853. Furusaki, A.; Watanabe, T.; Matsumoto, T.; Yanagiya, M. Tetrahedron Lett. 1968, 6301-6304. Perry, N.B.; Blunt, J.W.; Munro, M.H.G.; Pannel, L.K. J. Am. Chem. Soc.\9SS, 770,4850-4851. Matsunaga, S.; Fusetani, N.; Nakao, Y. Tetrahedron 1992, 48, 8369-8376. Kobayashi, J.; Itagaki, F.; Shigemori, H.; Sasaki, T. /. Nat. Prod., 1993, 56, 976-981. Fusetani, N.; Sugawara, T.; Matsunaga, S. J. Org. Chem. 1992,57, 3828-3832. Perry, N.B.; Blunt, J.W.; Munro, M.H.G.; Thompson, A.M. J. Org. Chem. 1990, 55, Tli'lll. Kobayashi, J.; Kondo, K.; Ishibashi, M.; Walchii, M.R.; Nakamura, T. J. Am. Chem. Soc. 1993, 775, 6661-6665. Kondo, K.; Ishibashi, M.; Kobayashi, J. Tetrahedron 1994, 50, 8355-8362. Rho, M. C ; Park, Y. H.; Sasaki, M.; Ishibashi, M.; Kondo, K.; Kobayashi, J.; Ohizumi, Y. Can. J. Physiol Pharmacol. 1996, 74, 193-199. Kobayashi, J.; Murayama, T.; Ohizumi, Y.; Sasaki, T.; Ohta, T.; Nozoe, S. Tetrahedron Lett. 1989, 30,4833-4836. Rao, R. A. V.; Reddy, G. R.; Rao, B. V. J. Org. Chem. 1991, 56, 4545-4547. Kobayashi, J.; Kondo, K.; Shigemori, H.; Ishibashi, M.; Sasaki, T.; Mikami, Y. J. Org. Chem. 1992, 57, 6680-6682. Nakamura, H.; Kobayashi, J.; Ohizumi, M. K.; Hirata, Y. Tetrahedron Lett. 1984, 25, 5401-5404. Kitagawa, I.; Yoshioka, N.; Kamba, C ; Yoshikawa, M.; Hamamoto, Y. Chem. Pharm. Bull. 1987, 35, 928-931. Ichiba, T.; Nakao, Y.; Scheuer, P. J.; Sata, N. U.; Kelly-Borges, M. Tetrahedron Lett. 1995, 36, 3977-3980. Matsunaga, S.; Fusetani, N.; Konosu, S. J. Nat. Prod. 1985, 48, 236-241. Matsunaga, S.; Fusetani, N.; Konosu, S. Tetrahedron Lett. 1984, 25, 51655168. Matsunaga, S.; Fusetani, N.; Konosu, S. Tetrahedron Lett. 1985, 26, 855-856. Fusetani, N. New J. Chem. 1990,14,111-11%. Yatsunami, J.; Fujiki, H.; Komori, A.; Suganuma, M.; Nishiwaki, S.; Okabe, S.; Nishiwaki, R.; Ohta, T.; Matsunaga, S.; Fusetani, N.; Sugimura, T. Proceedings of the 1st International Congress on Vitamins and Biofactors in Life Sciences. J. Nutritional ScL, Vitaminol. 1992, 333 (special issue). Ryu, G.; Matsunaga, S.; Fuselani, N. Tetrahedron Lett. 1994, 35, 8251-8254. Ryu, G.; Matsunaga, S.; Fusetani, N. Tetrahedron 1994, 50,13409-13416. Gulavita, N. K.; Gunasekera, S. P.; Pomponi, S. A.; Robinson, E. V. J. Org. Chem. 1992, 57, 1767-1772.

1254

[101] 102] 103]

104] 105] 106] 107] 108] 109] 110] Ill] 112] 113] 114] 115] 116] 117] 118] 119] 120] 121] 122] 123]

Gunasekera, S. P.; Pomponi, S, A.; McCarthy, P. J. J, Nat. Prod 1994, 57, 7983. Tada, H.; Tozyo, T.; Terui, Y.; Hayashi, F. Chem. Lett. 1992, 431-434. Ishihara, H.; Martin, B.L.; Brautigan, D.L.; Karaki, H.; Ozaki, H.; Kato, Y.; Fusetani, N.; Watabe, S.; Hashimoto, K.; Uemura, D.; Harshome, D. /. Biochem. Biophys. Res. Commun. 1989,159, Sll-Sll. Kato, Y.; Fusetani, N.; Matsunaga, S.; Hashimoto, K.; Fujita, S.; Furuya, T. J. Am. Chem. Soc. 1986,108, 2780-2781 Matsunaga, S.; Fusetani, N. Tetrahedron Lett. 1991, 32, 5605-5606. Hamada, Y.; Tanada, Y.; Yokokawa, F.; Shioiri, T. Tetrahedron Lett. 1991,32, 5983-5986. Kato, Y.; Fusetani, N,; Matsunaga, S.; Hashimoto,K.; Koseki, K. J.Org. Chem. 1988, 53, 3930-3932. Matsunaga, S.; Fujiki, H.; Sukata, D.; Fusetani, N. Tetrahedron 1991,47,29993006. Matsunaga, S.; Fusetani, N. Tetrahedron Lett. 1991, 32, 5605-5606. Okada, A.; Watanabe, K.; Umeda, K.; Miyakado, M. Agr. Biol. Chem. 1991, 55, 2765-2771. Barlett, P. A.; Richardson, D. P.; Myerson, J. Tetrahedron 1984, 40, 23172327. Matsunaga, S.; Wakimoto, T. and Fusetani, N. J.Org. Chem. 1997, 62, 26402642. Matsunaga, S.; Wakimoto, T. and Fusetani, N. Tetrahedron Lett. 1997, 38, 3763-3764. Gunasekera, S.P.; Gunasekera, M.; Longley, R.E.; Schulte, G.K. J. Org. Chem. 1990, 55,4912-4915; (erratum) ibidem 1991, 56,1346. ter Haar, E.; Kowalsky, R.J.; Hamel, E.; Lin, CM.; Longley, R.E.; Gunasekera, S.P.; Rosenkranz, H.S.; Day, B.W. 1996, Biochemistry, 35, 243-250. Gunasekera, S.P.; Gunasekera, M.; McCarthy, P. J. Org. Chem. 1991,56,48304833. D'Auria, M.V.; Gomez Paloma, L.; Minale, Riccio, R. J. Nat. Prod. 1992, 55, 311-320. D'Auria, M.V.; Gomez Paloma, L.; Minale, Riccio, R.; Debitus, C. Tetrahedron Lett. 1991, 32, 2149-2152. Oger, J. M.; Richonune, P.; Bruneton, P. R.; Guinaudeau, H.; Sevenet, T.; Debitus, C. J. Nat. Prod 1991, 54, 273-275. Guella, G.; Mancini, I.; Chiasera, G.; Pietra, F.; Helv. Chim. Acta 1989, 72, 237-246. D'Auria, M.V.; Gomez Paloma, L.; Minale, L.; Zampella, A. Verbist, J.F.; Roussakis, C ; Debitus, C. Tetrahedron 1993, 49, 8657-8664. D'Auria, M.V.; Gomez Paloma, L.; Minale, L.; Zampella, A. Verbist, J.F.; Roussakis, C ; Debitus, C; Patissou, J. Tetrahedron 1994, 50,4829-4834. Carbonelli, S.; Zampella, A.; Randazzo, A.; Debitus, C ; Gomez Paloma, L. Tetrahedron 1999, 55, 14665-14674.

1255

[124] [125]

Lancelin, J. N.; Beau, J. M.;. J. Am. Chem. Soc, 1990, 772, 4060-4061. (a) Ishibashi, M.; Moore, R. E.; Patterson, G. M. L.; Xu, C ; Clardy, J. J. Org. Chem. 1986, 57, 5300-5306. (b) Carmeli, S.; Moore, R. E.; Patterson, G. M. L. J. Nat. Prod. 1990, 55, 1533-1542. [126] Zhang, X.; Minale, L.; Zampella, A.; Smith, C. D. Cancer Research, 1997, 57, 3751-3758. [127] D'Auria, M.V.; Debitus, C; Gomez Paloma, L..; Minale, L.; Zampella, A. J. Am. Chem. Soc. 1994, 776, 6658-6663. [128] D'Auria, M.V.; Gomez Paloma, L.; Minale, L.; Zampella, A.; Debitus, C. J. Nat. Prod.,1994, 57, 1595-1597. [129] D' Auria M. V.; Gomez Paloma, L.; Minale, L.; Zampella, A.; Debitus, C ; Perez, J. J. Nat. Prod. 1995, 58, 121-123. [130] Zabriskie, T. M.; Klocke, J. A.; Ireland, C. M.; Marcus, A. H.; Molinski, T. R; Faulkner, D. J.; Xu, C ; Clardy, J. C. J. Am. Chem. Soc. 1986,108, 3123-3124 [131] Crews, P.; Manes, L. V.; Boehler, M. Tetrahedron Lett 1986, 27, 2797-2800 [132] Chan, W. R.; Tinto, W. P.; Manchand, P. S.; Todaro, L. J. J. Org. Chem. 1987, 52, 3091-3093 [133] Zampella, A.; D'Auria, M.V.; Minale, L.; Debitus, C ; Roussakis, C. J. Am. Chem. Soc. 1996,118, 11085-11088. [134] Zampella, A.; D'Auria, M.V.; Minale, L.; Debitus, C. Tetrahedron 1997, 53, 3243-3248. [135] Brimacombe, S. J.; Mahmood, S.; Rollis, A. J. J. Chem. Soc, Perkin Trans. 1 1975, 1292-1297. [136] Parker, K. A.; Babine, R. E.; J. Am. Chem. Soc. 1982,104, 7330-7331. [137]. Zampella, A.; D' Auria M. V.; Gomez Paloma, L.; Casapullo, A.; Minale, L.; Debitus, C ; Henin, Y. J. Am. Chem. Soc. 1996,118, 6202-6209 [138] D' Auria M. V.; Zampella, A.; Gomez Paloma, L.; Minale, L.; Debitus, C ; Roussakis, C; Le Bert, V. Tetrahedron 1996, 52, 9589-9596 [139] Stierle, A.; Cardellina II, J.H.; Singleton, F.L. Experientia 1988, 44,1021. [140] Ballantine, J.A.; Barrett, C.B.; Beer, R.J.S.; Eardley, S.; Robertson, A.; Shaw, B.L.; Simpson,T.H. J. Chem. Soc. 1958, 755-760. [141] Pfleiderer, W. Chem. Ber. 1957, 90, 2582-2587 [142] Debitus, C ; Cesario, M.; Guilhem, J.; Pascard, C; Pais, M. Tetrahedron Lett. 1989, 30, 1535-1538. [143] Guerriero, A.; D'Ambrosio, M.; Pietra, F.; Debitus, C; Ribes, O. J. Nat. Prod. 1993, 56, 1962-1970. [144] D'Ambrosio, M.; Guerriero, A.; Debitus, C ; Ribes, O.; de Forges, B. R.; Pietra, F. Helv. Chem. Acta 1989, 72, 1451-1454. [145] D'Ambrosio, M.; Guerriero, A.; Debitus, C ; Ribes, O.; Pietra, F. Helv. Chem. Acta 1993, 76, 1489-1496. [146] Yon-Hin, P.; Scott, A. I. Tetrahedron Lett., 1991, 32, 4231-4234. [147] Chicarelli, M. I.; Maxwell, J. R. Tetrahedron Lett. 1986, 27,4653-4654. [148] Bonnet, R.; Gechowski, F. J. Chem. Soc. Perkin Trans. 11984,125-131.

1256

[149] [150] [151] [152] [153] [154] [155]

[156] [157] [158] [159] [160] [161] [162] [163]

[164] [165] [166]

[167] [168] [169] [170] [171]

Bewley A. C; He H.; Williams, D. H.; Faulkner, D. J. J. Am. Chem. Soc. 1996, 775,4314-4321. Costantino, V.; Fattorusso, E.; Mangoni, A.; Aknin, M.; Gaydou, E. M. Steroids 1994, 59, 181-184. Gunasekera, S. P.; Kelly-Borges, M.; Longley, R. E. J. Nat. Prod. 1996, 59, 161-162. Bewley, C. A.; Debitus, C ; Faulkner, D. J. J. Am. Chem. Soc. 1994,116, 76317636. Schmidt, E. W.; Faulkner, D. J. Tetrahedron 1998, 54, 3043-3056. Valentekovich, R. J.; Schreiber, S. L. J. Am. Chem. Soc. 1995,117, 9069-9070. (a) Bauer, S. M.; Armstrong, R. W. J. Am. Chem. Soc. 1999,121, 6355-6366. (b) Samy, R.; Kim, H. Y., Brady, M.; Toogood, P. L. J. Org. Chem. 1999, 64, 2711-2728. (c) Hu, T.; Panek, J. S. J. Org. Chem. 1999, 64, 3000-3001. Choi, B. W.; Namikoshi, F. S.; Rinehart, K. L.; Carmichael, W. W.; Kaup, A. M.; Evans, W. R.; Beasley, V. R. Tetrahedron Lett. 1993, 34, 7881-7884. Matteson, D. S. Pure Appl. Chem. 1991, 63, 339-344. Matteson, D. S.; Kandil, A. A.; Soundararajan, R. J. Am. Chem. Soc. 1990,112, 3964-3969. Gokel, G.; Liidke, G.; Ugi, I. In Isonitrile Chemistry: Ugi, I., Ed.: Academic Press: New York, 1971; pp 145-199. Masse, C. E.; Panek, J. S. Chem. Rev. 1995, 95,1293-1316. Panek, J.S. Beresis, R.; Xu, F.; Yang, M. J. Org. Chem. 1991, 56, 7342-7344. Panek, J. S.; Hu, T. J. Org. Chem. 1997, 62, 4914-4915. (a) Namikoshi, M.; Rinehart, K. L.; Dahlem, A.; Beasley, V. R.; Cannicheal, W. W. Tetrahedron Lett. 1989, 30, 4349-4352. (b) Beatty, M. F.; JenningsWhite, C ; Avery, M. A. J. Chem. Soc, Perkin Trans. 1 1992,1637-1641. Kim, H. Y.; Toogood, P. L. Tetrahedron Lett. 1996, 37, 2349-2352. Paterson, I.; Yeung, K.-S.; Ward, R. A.; Cununing, J. G.; Smith, J. D. J. Am. Chem. Soc. 1994,116, 9391-9392. (a) Paterson, I.; Smith, J. D.; Ward, R. A.; Gumming, J. G.; J. Am. Chem. Soc. 1994,116, 2615-2616. (b) Paterson, I.; Cmnming, J. G.; Smith, J. D.; Ward, R. A.; Yeung, K.-S. Tetrahedron Lett. 1994, 35, 3405-3408. (c) Paterson, I.; Yeung, K.-S.; Ward, R. A.; Smith, J. D.; Cunmiing, J. G.; Lamboney, S. Tetrahedron 1995, 57, 9467- 9486. Nicolaou, K. C.; Ajito, K.; Khatuya, P. H.; Richter, P. K.; Bertinato, P. J. Am. Chem. Soc. 1996,118, 3059-3060. Norcross, R. D.; Paterson, I. Chem. Rev. 1995, 95, 2041-2114. (a) Paterson, I.; Smith, J. D. J. Org. Chem. 1992, 57, 3261-3264. (b) Paterson, I.; Smith, J. D.; Ward, R. A. Tetrahedron 1995,57, 9413- 9436. Paterson, L; Cununing, J. G.; Ward, R. W.; Lamboley, S. Tetrahedron 1995, 57, 9393- 9412. Paterson, I.; Cmnming, J. G.; Smith, J. D.; Ward, R. A. Tetrahedron Lett. 1994, 55, 441-444.

1257

[172] [173] [174] [175] [176] [177] [178] [179] [180]

[181]

lanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamagushi, M. Bull Chem. Soc. Jpn. 1979, 52,1989-1994. Keck, G. E.; Boden, E. P. 7. Org. Chem. 1985, 50, 2394-2395. Carretero, J.; Ghosez, L. Tetrahedron Lett. 1988, 29, 2059-2062. Hong, C. Y.; Kishi, Y. J. Am. Chem. Soc. 1991,113, 9693-9694. Hong, C. Y.; Kishi, Y. J. Org. Chem. 1990, 55, 4242-4245. (a) Nakata, T.; Nagao, S.; Mori, N.; Oishi, T. Tetrahedron Lett. 1985,26,64616464. (b) ibidem 6465-6468. RocheroUe, V.; Lopez, J. C; Olesker, A.; Lukacs, G. J. Chem. Soc , Chem. Commun. 1988, 513-514. Lewis, M. D.; Cha, J. K.; Kishi, Y. J. Am. Chem. Soc. 1982,104, 4976-4978. For leading references to synthetic efforts to the calyculins see: (a) Evans, D. A.; Gage J. R. Tetrahedron Lett. 1990, 31, 6129-6132. (b) Evans, D. A.; Gage J. R. J. Org. Chem. 1992, 57, 1958-1961. (c) Evans, D. A.; Gage J. R.; Leighton, J. L. ; Kim A. S. J. Org. Chem. 1992, 57, 1961-1963. (d) Evans, D. A.; Gage J. R.; Leighton, J. L. J. Org. Chem. 1992, 57, 1964-1966.(e) Duplanter, A. J.; Nantz, M. H.; Roberts, J. C.; Short, R. P.; Somfai, P.; Masamune, S. Tetrahedron Lett. 1989, 30, 7357-7360. (f) Vaccaro, H. A.; Levy, D. E.; Sawabe, A.; Jaetsch, T,; Masamune, S. Tetrahedron Lett. 1992,33, 1937-1940. (g) Hara, O.; Hamada, Y.; Shioiri, T. Synlett. 1991, 283-284. (h) Yokokawa, F.; Hamada, Y.; Shioiri, T. Synlett. 1992, 149-150. (i) Matsubara, J.; Nalcao, K.; Hamada, Y.; Shioiri, T. Tetrahedron Lett. 1992, 33, 4187-4190. (j) Takebushi, K.; Hamada, Y.; Shioiri, T. Tetrahedron Lett. 1994, 35, 52395242. (k) Hamada, Y.; Yokokawa, F.; Kabeya, M.; Hatano, K.; Kurono, Y.; Shioiri, T. Tetrahedron 1996, 52, 8297-8306.(1) Kabeya, M.; Hamada, Y.; Shioiri, T. Tetrahedron 1997, 53, 9769-9777. (m) Zhao, Z.; Scarlato, G. R.; Armstrong, R. W. Tetrahedron Lett. 1991, 32, 1609-1612. (n) Armstrong, R. W.; De Mattei, J. A. Tetrahedron Lett. 1991, 32, 5749-9752. (o) Salvatore, A. B.; Smith, A. B., III. Tetrahedron Lett. 1991, 32, 4855-4858. (p) Smith, A. B., Ill; Salvatore, A. B.; Hull, K. G.; Duan, J. J.-W. Tetrahedron Lett. 1991, 32, 4859-4862. (q) Smith, A. B. Ill; Duan, J. J.-W.; Hull, K. G.; Salvatore, A. B. Tetrahedron Lett. 1994, 32, 1329-1330. (r) Smith, A. B. Ill; Iwashima, M. Tetrahedron Lett. 1994, 35, 6051-6052. (s) Koskinen, A. M. P.; Chen, J. Tetrahedron Lett. 1991, 32, 6977-6980. (t) Koskinen, A. M. P.; Pihko, P. M. Tetrahedron Lett. 1994, 35, 7417-7420. (u) Pihko, P. M.; Koskinen, A. M. P. /. Org. Chem. 1998, 63, 92-98. (v) Barrett, A. G. M.; Edmunds, J. J.; Horita, K.; Parkinson, C. J. J. Chem. Soc, Chem. Commun. 1992, 1236-1238. (w) Barrett, A. G. M.; Edmunds, J. J.; Hendrix, J. A.; Horita, K.; Parkinson, C. J. J. Chem. Soc, Chem. Commun. 1992, 1238-1240. (y) Barrett, A. G. M,; Edmunds, J. J.; Hendrix, J. A.; Malecha, J. W.; Parkinson, C. J. J. Chem. Soc, Chem. Commun. 1992, 1240-1242. (z) Trost, B. M.; Flygare, J. A. Tetrahedron Lett. 1994, 35, 4059-4062. Evans, D. A.; Gage, J. R.; Leighton, J. L. J. Am. Chem. Soc 1992,114, 94349453.

1258

[182] [183] [184]

[185]

[186] [187] [188]

[189] [190] [191]

[192] [193] [194] [195]

Tanimoto, N.; Gerritz, S. W.; Sawabe, A.; Noda, T.; Filla, S. A.; Masamune, S. Angew. Chem., Int. Ed. Engl. 1994, 55, 673-675. Yokokawa, R; Hamada, Y.; Shioiri, T. Chem Commun. 1996, 871-872. Smith, A. B., Ill; Friestad, G. K.; Duan, J. J.-W.; Barbosa, J.; Hull, K. G.; Iwashima, M.; Qiu, Y.; Spoors, P. G.; Bertounesque, E.; Salvatore, B. A. 7. Org. Chem. 1998, 55, 7596-7597. (a) Scarlato, G. R.; DeMattei, J. A.; Chong, L. S.; Ogawa, A. K.; Lin, M. R.; Armstrong, R. W. J. Org. Chem. 1996, 61, 6139-6152. (b) Ogawa, A. K.; DeMattei, J. A.; Scarlato, G. R.; Tellew, J. E.; Chong, L. S.; Armstrong, R. W. J. Org. Chem. 1996,67, 6153-6161. (c) Ogawa, A. K.; Armstrong, R. W J. Am. Chem. Soc. 1998,120, 12435-12442. Lavall6e, P.; Ruel, R.; Grenier, L.; Bissonnette, M. Tetrahedron Lett. 1986, 27, 679-682. Duan, J. J.-W.; Smith, A. B., Ill J. Org. Chem. 1993, 58, 3703-3711. (a) Paterson, I.; Wren, S.P. J. Chem. Soc, Chem. Commun. 1993, 1790-1792. (b) Clark, D. L.; Heathcock, C. H. J. Org. Chem. 1993, 58, 5878-5879. (c) Golec, J. M. C; Jones, S. D. Tetrahedron Lett. 1993,34, 8159-8163. (d) Evans, P. L.; Golec, J. M. C ; Gilliespie, R. J. Tetrahedron Lett. 1993, 34, 8163-8166. (e) Golec, J. M. C; Gilliespie, R. J. Tetrahedron Lett. 1993, 34, 8167-8169. (f) Yang, G.; Myles, D. C. Tetrahedron Lett. 1994, 35, 1313-?. (g) Yang, G. Myles, D. C. Tetrahedron Lett. 1994, 35, 2503-2504. (h) Paterson, I. Schlapbach. A. Synlett 1995, 498-500. (i) Filla, S. A.; Song, J. J.; Chen, L. Masamune, S. Tetrahedron Lett. 1999, 40, 5449-5453. Nerenberg, J. B.; Hung, D. T.; Somers, P. K.; Schreiber, S. L. J. Am. Chem. Soc. 1993, 775, 12621-12622. Hung, D. T.; Nerenberg, J. B.; Schreiber, S. L. J. Am. Chem. Soc. 1996, 118, 11054-11080. (a) Smith, A. B., Ill; Qui, Y.; Jones, D. R.; Kobayashi, K. /. Am. Chem. Soc. 1995, 777, 12011-12012. (b) Harried, S. R ; Yang, G.; Strawn, M. A.; Myles, D. C. J. Org. Chem. 1997, 62, 6098-6099. Marshall, J. A.; Johns, B. A. J. Org. Chem. 1998, 63, 7855-7892. Paterson, I.; Florence, G. J.; Gerlach, K.; Scott, J. P. Angew. Chem. Int. Ed. 2000, 39, 377-380. Still, W. C ; Gennari, C. Tetrahedron Lett. 1983, 24, 4405-4408. (a) Aicher, T. D.; Kishi, Y. Tetrahedron Lett. 1987, 28, 3463-3466. (b) Takai, K.; Kuroda, T.; Nakatsukasa, S.; Oshima, K.; Nozaki, H Tetrahedron Lett. 1985, 26, 5585-5588.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 26 © 2002 Elsevier Science B.V. All rights reserved.

1259

COPPER/TOPA QUINONE-CONTAINING AMINE OXIDASES - RECENT RESEARCH DEVELOPMENTS MAREK SEBELA, IVO FREBORT, MAREK PETRIVALSKY AND PAVEL PEC Department of Biochemistry, Faculty of Science, Palacky University, Slechtitelii 11, 783 71 Olomouc, Czech Republic ABSTRACT: Amine oxidases (EC 1.4.3.6) that contain copper/topa quinone cofactor belong to a new protein group of quinoproteins emerging in recent years. This review brings together information on the general properties of the enzymes and their physiological functions. In plants, these enzymes are involved in processes of development and senescence, they reduce the concentration of toxic amines produced during exposure to stress conditions, provide hydrogen peroxide for wall stiffening and lignification and precursor compounds for biosynthesis of some alkaloids. Major attention is currently being paid to the structure of the active site of the enzymes that contains copper ions and a posttranslationally modified tyrosyl residue, topa quinone. Three-dimensional structures recently obtained for several amine oxidases by X-ray diffraction analysis of the respective crystals provide important structural information about the unique protein folding of the native enzyme and molecular arrangement of the active site. Biogenesis of the quinone cofactor is another issue that is addressed frequently at present. Differences between copper/topa quinone-containing amine oxidases and flavoprotein polyamine oxidases are outlined. Finally, future research directions on amine oxidases and the possibilities of their practical application are discussed.

INTRODUCTION The history of investigation into amine oxidases represents a fascinating interplay of several disciplines using different techniques and approaches. Study on the metabolism of biogenic amines takes its origin from the end of the last century, when organic chemists isolated various substances from biological material, among which were amino acids and a number of diamines and polyamines [1]. For example, the diamines, putrescine and cadaverine were isolated from putrefying animal organs during the years 1885-1887 and named accurately in accordance with their origin. After identifying the polyamine spermine as a phosphate salt in semen, it was isolated and crystallized in 1888. As these findings were then interpreted,

1260 the problem of the metabolic origin of these amines arose. During the period 1890-1910 decarboxylating conversions of the amino acids, lysine, ornithine and arginine were described, resulting in the respective diamines, cadaverine, putrescine and agmatine. The presence of putrescine and cadaverine in putrefied soybean extract was confirmed in 1910. Later, in 1927 the polyamine spermidine was discovered. The metabolism of biogenic amines then became the subject of interest of a newly developing discipline - biochemistry [1]. Currently it has been shown, that there is a new group of enzymes detectable in animal and plant tissues, which decrease the concentration level of amines by the oxidative deamination. These enzymes have been declared as amine oxidases [2]. The first in vitro study on amine oxidases was the demonstration of the presence of tyramine oxidase in liver preparations in 1928. A couple of years later, histaminase, an oxidase sensitive to carbonyl reagents was discovered. During the 30s and 40s, histaminase was reported in lung and later a diamine oxidase that oxidizes putrescine and cadaverine in porcine kidney. The activity of diamine oxidase was also found in plant extracts, especially those prepared from legumes. The enzyme from porcine kidney was partially purified and its identity with histaminase in this tissue was confirmed in 1951. Crude preparation of diamine oxidase from pea seedlings was achieved in 1956. The period 1960-1980 saw an intensive research on the pea and porcine kidney enzymes. Many of their substrates and inhibitors were described. These enzymes were also purified to homogeneity and the kidney enzyme was crystallized. The amine oxidases from bovine and porcine plasma were also characterized at this time. The classification of amine oxidases according the enzyme cofactor subsequently led to distinctions between flavin-containing amine oxidases and copper-containing amine oxidases, the latter including the above mentioned enzymes from pea and porcine kidney. A number of other amine oxidases belonging to these groups were isolated and their properties studied in detail: mitochondrial monoamine oxidase and polyamine oxidase from cereals (both flavin-containing amine oxidases), and a number of plant diamine oxidases (all copper-containing amine oxidases). Various proofs for the presence of a second (organic) cofactor in copper-containing amine oxidases began to appear. The cofactor was primarily thought to be pyridoxal phosphate due to several spectral characteristics. Nevertheless, it was never directly confirmed until the middle of 80s.

1261 In 1984, some indirect evidence was published identifying the organic cofactor of copper-containing amine oxidases with pyrroloquinoline quinone (PQQ). This substance had been known as a cofactor of some microbial oxidoreductases for several years [3]. Its identification as the cofactor PQQ was then indirectly confirmed using instrumental methods in a number of copper amine oxidases, including plant enzymes from pea and lentil. A major breakthrough in the study of amine oxidases as well as in the whole field of protein biochemistry came in 1990 when the existence of a novel cofactor was shown in bovine serum amine oxidase. HPLC isolation of cofactor peptides and their analyses using modem techniques such as amino acid sequencing and mass spectrometry, led in 1990 to the unambiguous disclosure of topa quinone (TPQ) in coppercontaining amine oxidase from bovine serum [4]. Study of amine oxidases then shifted to a new dimension. The occurrence of copper amine oxidase was demonstrated in a number of animal tissues showing high levels of metabolic activity, and also in plants, fungi, yeast and bacteria. The presence of topa quinone was shown in a number of enzymes showing the typical consensus sequence of the active site. Evidence was obtained for the mechanism of the catalytic reaction involving radical intermediate and autocatalytic biogenesis of topa quinone from a tyrosyl precursor, both in the presence of protein bound cupric ions. To date, the complete amino acid sequence is available for several amine oxidases, and some enzymes have been crystallized and analyzed by X-ray diffraction. All these achievements have facilitated better understanding of the protein structure of the enzyme and coordination geometry of the active site, thus providing direct insight into the mechanisms of enzyme action. When other quinonoid cofactors, tryptophan tryptophylquinone (TTQ) and lysyl-tyrosyl quinone (LTQ) were found in bacterial amine dehydrogenase and mammalian lysyl oxidase, respectively [5,6], a completely new class of quinoproteins (or quinoenzymes) in protein classification was declared. At the present time, research on copper amine oxidases revolves more and more around the elucidation of their physiological role. The enzymes of plant origin especially are subjected to intensive study on the interface between biochemistry and plant physiology since they are suspected of playing a key role in essential metabolic pathways in plants [1,7]. At the molecular level, the enzymes affect the concentration of biogenic amines in plant cell. It has been shown that the amines and the products of their catabolism participate in such delicate processes as cell division and

1262 proliferation, apoptosis, senescence, wound healing, and the mechanisms of defense against molecular stress and pathogen infection. Plant copper amine oxidase also takes part in the biosynthesis of several types of alkaloids. During recent years, a lot of work has been done on elucidation of the enzyme action in the lignification process. It has been confirmed that the hydrogen peroxide released by the amine oxidase reaction serves as an input compound for the formation of lignin and suberin. However, research on changes of the activity of copper amine oxidases during development and understanding of the genetic basis of its distribution in different plant organs, e.g. roots, stems and leaves, still represents a big challenge for plant physiologists. Copper amine oxidases Copper amine oxidases (EC 1.4.3.6) [CAOs, amine: O2 oxidoreductase (deaminating)] catalyze the oxidative deamination of biogenic amines to corresponding aldehydes and ammonia, accompanied by a two-electron reduction of molecular oxygen to hydrogen peroxide [7]: RCH2NH3'^ + O2 + H2O -^ RCHO + H2O2 + NH4"^

The enzymatic reaction follows a bi-ter ping-pong mechanism with participation of the quinone cofactor and Cu(II) ions bound in the active site [8]. The enzymes are irreversibly inhibited by carbonyl reagents and reversibly by copper chelating agents (non competitive inhibitors), substrate analogues (competitive inhibitors) and some alkaloids [9,10]. A former classification used the terms monoamine oxidase, diamine oxidase and polyamine oxidase [11]. However, the substrate specificity of the above enzymes is usually quite broad [12], and such a classification was far from accurate. Novel classification of copper amine oxidases comes from both the origin and substrate specificity of the enzymes [13] and includes four groups: (a) mammalian blood plasma amine oxidases showing high affinity for benzylamine and the polyamines spermine and spermidine, (b) mammalian diamine oxidases from kidney and placenta preferentially oxidizing the diamines putrescine and cadaverine, and also histamine, (c) plant diamine oxidases showing very high affinity for putrescine and cadaverine, also oxidizing spermidine at a high rate, (d)

1263 microbial amine oxidases preferentially converting arylalkylamines such as benzylamine, phenethylamine, tyramine and histamine (with the exception of methylamine oxidase). Representatives of all groups were studied in detail [13]. Best known enzymes of the group (a) are the amine oxidases from bovine and porcine blood plasma or serum; The enzymes from rabbit, sheep and equine plasma were also isolated. Mammalian diamine oxidases, the group (b), is represented by the enzymes from porcine kidney, human kidney and placenta, and rat colon. Group (c), the plant diamine oxidases, includes enzymes from legumes (Fabaceae). Several enzymes were purified to homogeneity and characterized in this regard. The starting material for purification is mostly etiolated plant seedlings, less often leaves. Best studied are the enzymes from seedlings of pea Pisum sativum, lentil Lens culinaris [10,12] and from latex of the Mediterranean shrub Euphorbia characias [14]. Finally, the group (d), microbial amine oxidases, includes the enzymes from the filamentous fungus Aspergillus niger, yeasts Hansenula polymorpha and Pichia pastoris. Gram-negative bacteria Escherichia coli, Klebsiella aerogenes and Gram-positive bacteria Arthrobacter globiformis and Arthrobacter strain PI [12,13]. All sources of amine oxidases have been comprehensively reviewed [12,13] and since the enzymes are widely distributed in many organisms and species, it is considered and generally accepted that they are ubiquitous.

Flavin polyamine oxidases Polyamines (namely spermidine) are oxidatively deaminated not only by the action of copper amine oxidases. Flavin-containing polyamine oxidases (PAOs, EC 1.5.3.-) are also involved in their catabolism. An alternative grouping divides both above mentioned catabolic enzymes into those that act on the primary amino group of di- and polyamines (diamine oxidases) and those that act on the secondary amino group of polyamines (polyamine oxidases) [15]. As suggested by Morgan, the latter should be further subdivided according to whether propane-1,3-diamine or 3aminopropanal is the reaction product [16]. The exact nature of the reaction products of PAOs depends on the source of the enzyme [1]. Mammalian PAOs transform spermidine and spermine into putrescine and spermidine, respectively, plus 3aminopropanal. By contrast, plant and bacterial polyamine oxidases

1264 catalyze the conversion of spermidine and spermine to 4-aminobutanal and l-(3-aminopropyl)-4-aminobutanal, respectively, plus propane-1,3diamine. The aminoaldehydes produced in the reaction spontaneously cycle to 1-pyrroline and l-(3-aminopropyl)pyrrolinium, respectively [17], the latter compound occurring mainly in the bicyclic form of 1,5diazabicyclo[4.3.0]nonane in the leaves of various cereals [17]. As typically observed in flavin-dependent oxidases, the overall reaction catalyzed by PAO can be divided into a reductive half reaction, in which the flavin is reduced upon polyamine oxidation, and an oxidative half reaction, in which the reduced flavin is deoxidized by molecular oxygen with the release of hydrogen peroxide [18]. Polyamine oxidation is likely to result in the formation of an imino compound, which is hydrolyzed to produce the final products. In contrast to CAOs [10], the polyamine oxidation by PAOs results in the production of aldehyde and hydrogen peroxide but not in ammonia [15]. In higher plants the wellknown pea seedling copper amine oxidase has been shown to oxidize spermine only very slowly, while spermidine is quickly converted to l-(3aminopropyl)pyrrolinium that undergoes spontaneous cyclization to 1,5diazabicyclo[4.3.0]nonane along with the formation of H2O2 and ammonia [10]. Spermine is oxidized by lentil seedling amine oxidase at the terminal (primary) amino groups producing a dialdehyde [19,20] Physiological aspects of copper amine oxidases Concerning prokaryotes, copper amine oxidases simply allow the organism to grow using the amine as a carbon and energy source. However, the general function of copper amine oxidases in higher organisms has not yet been well understood. The most significant role of the enzymes in plants and animals is considered to be catabolic regulation at the cellular level of biogenic amines, especially the diamine putrescine and poly amines spermidine and spermine. These amines play important roles in fundamental cellular processes such as tissue differentiation, cellular growth and proliferation, transformation in cellular cultures and tumor growth, wound healing, defense activity against parasites and probably also in programmed cell death, apoptosis [7,12,21]. Diamine oxidases found in animal tissues also participate in the metabolism of histamine [12,22]. Primary products of the enzymatic oxidation of mono-.

1265 di- and polyamines, aldehydes and aminoaldehydes, themselves have a regulatory effects on nucleic acids, but are further metabolized to other biologically active compounds [7]. In plants, the polyamines show growth effects, stabilize cellular membranes and delay senescence. Moreover, plant amine oxidases participate in the biosynthesis of some alkaloids and probably also provide hydrogen peroxide for the lignification and stiffening of the cell wall [23,24]. It has been also suggested that amine oxidase may show similarity to or even identity with cytokinin oxidase that cleaves the A^-alkylated adenine derivatives, known as the plant hormones cytokinins, on the secondary amino group [25]. However, CAOs are not known to oxidize secondary amines [12] and our recent results confirm that cytokinins are in fact weak competitive inhibitors of plant amine oxidase [26]. Molecular properties of copper amine oxidases With few exceptions [12,27], copper amine oxidases are homodimers in the native conformation. The molecular mass of the single subunit lies in the range 70-100 kDa. Most of the eukaryotic copper amine oxidases are glycoproteins. Their isoelectric point is usually slightly below pi 7.0, but some enzymes have pi > 7.0 such as pea seedling amine oxidase [12]. The amino acid composition is known for many amine oxidases and in the last decade, the complete amino acid sequence has been determined for a number of them. Some enzymes have been prepared in a crystalline form [12], four of them have been analyzed by X-ray diffraction with complete resolution of their three dimensional structure including detailed spatial conformation of the active site [28-31]. The quinone cofactor of copper amine oxidases Even in the early days of amine oxidase research, it became evident that the organic cofactor showing absorption at 500 nm and conferring a pink color to the enzyme contains a carbonyl functional group [32]. Until the mid eighties, it was believed that the cofactor was pyridoxal phosphate, but nobody succeeded in proving its presence experimentally [33,34]. In 1984, indirect evidence was shown of a covalently bound

1266 pyrroloquinoline quinone (methoxatin, PQQ) in the amine oxidase from bovine blood serum [35,36]. The PQQ had already been known for some time as a dissociable coenzyme of some prokaryotic oxidoreductases [3,37]. Covalently bound PQQ was then shown to be present in a number of other copper amine oxidases, but always indirectly [38-44]. O

O

?

««««—HN—CH—C —NH—CH —C—NH—CH—CCH, CONH,

(CHJ„_-1,2 COO,©

- Asn - TPQ - Asp/GIu -

Fig. (1). Topa quinone as a part of the cofactor consensus amino acid sequence in the active site of copper amine oxidases [4].

The beginning of a new era in the study of copper amine oxidases dates to 1990, when a group of scientists lead by Prof Judith Klinman from the University of California in Berkeley unambiguously eradicated the speculations about PQQ as the cofactor of copper amine oxidases [4]. A ^^c.iabeled phenylhydrazine derivative of bovine plasma amine oxidase was digested with protease thermolysine and the resulting peptides were separated by HPLC. Following the radioactive label, the pentapeptide Leu-Asn-X-Asp-Tyr was isolated in high yield allowing subsequent analysis by mass spectrometry. The unknown amino acid X was identified as a phenylhydrazone of the quinone form of 2,4,5trihydroxyphenylalanine (6-hydroxydopa quinone, named topa quinone, TPQ), Fig. (1). The structure was confirmed by synthesizing a stable model compound 5-(2,4,5-trihydroxybenzyl)hydantoin undergoing rapid autooxidation to the quinone, which was then derivatized with

1267 phenylhydrazine. The unique identity of the spectral properties of the model combined with those of the active site pentapeptide isolated from the amine oxidase from bovine plasma produced firm evidence that topa quinone is in fact the organic cofactor of the enzyme [4]. The results were further confirmed by resonance Raman spectrometry on comparing the spectra of phenylhydrazone and /?nitrophenylhydrazone of bovine plasma amine oxidase with the derivatized pentapeptides of the active site and the model compound. All these spectra showed great similarity in position (wavenumber) and spectral band intensity, while the spectrum of a PQQ model compound differed markedly [45]. Similar experiments confirmed the presence of topa quinone in porcine kidney, pea seedling and Arthrobacter PI amine oxidases. Moreover, the experimental data obtained for intact enzymes excluded the possibility of an artificial topa quinone formation during the proteolysis and peptide isolation [45]. Topa quinone is not a typical oxidoreductase cofactor, since it is an integrated part of the peptide backbone and must be formed in the active site by a posttranslational modification of a tyrosyl precursor. Search for this precursor exploited the data obtained from cloning and sequencing of the gene for peroxisomal copper amine oxidase from methylotrophic yeast Hansenula polymorpha [46]. The yeast is able to form two distinct amine oxidases depending on the inducing amine used in the cultivation medium, methylamine oxidase [46] and benzylamine oxidase [47]. The enzyme derivatized with phenylhydrazine gave, after proteolysis and peptide separation, the cofactor peptide Val-Ala-Asn-X-Glu-Tyr-Val showing the typical resonance Raman spectrum of the topa quinone derivative. Comparing the amino acid sequence obtained by translating the nucleotide sequence of the cloned gene with the peptide sequence obtained, it was found that the precursor of topa is a tyrosyl residue in a typical consensus sequence [47]. Later, the presence of topa quinone was accordingly confirmed in the amine oxidases from porcine serum and kidney and pea seedling by resonance Raman spectrometry of active-site labeled peptides [48]. Comparison of amino acid sequences of these peptides with the sequences of those from bovine plasma and H. polymorpha amine oxidases demonstrated the presence of a consensus sequence Asp-TPQ-Asp/Glu as shown in Fig. (1). Using the pH-dependent shift of the absorption maximum of the enzyme /^-nitrophenylhydrazone, which is considered to be a reliable indirect proof, the presence of topa quinone was also shown

1268 in the amine oxidase from seedlings of chick pea (Cicer arietinum) [48], fenugreek {Thgonella foenum-graecum) [49], sainfoin [50], grass pea (Lathyrus sativus) and sweet pea (Lathyrus odoratus) [51], etc. Recently, the cofactor peptides have also been isolated from semicarbazide-sensitive amine oxidases purified from bovine and porcine aortas [52], sequenced and confirmed to contain the topa quinone. The same topa quinone consensus sequence was also found in the primary structures of amine oxidases from human kidney [53], human retina [54] and rat colon [55], so called "amiloride-binding proteins", and amine oxidase from human placenta [56] that shows 81% identity with bovine plasma amine oxidase [57], bovine lung amine oxidase [58], and amine oxidases from pea and lentil seedlings [59,60], chick pea seedlings [61], and Arabidopsis thaliana [62] obtained by the molecular cloning of respective DNAs. The case of the two amine oxidase forms isolated from the filamentous fungus Aspergillus niger AKU 3302 after induction with w-butylamine is interesting. The dimeric AO-I and monomeric AO-II are encoded by a single gene, which was cloned and sequenced. The primary structure deduced then showed the cofactor consensus sequence Asn-Tyr-Glu, where the tyrosyl Tyr404 is the topa quinone precursor [27]. Recent results show that the AO-II is probably an unfolded inactive precursor of AO-I [63]. The enzymes were derivatized with /^-nitrophenylhydrazine and proteolyzed. Labeled peptides were then purified by HPLC and analyzed by ' H - N M R , MS, and resonance Raman spectrometry, which confirmed the presence of topa quinone in both enzymes. The fact that topa quinone in the derivatized peptides from AO-I was linked by an ester bond to the distant glutamyl residue Glul45 was surprising [64]. However, recent crystallographic data for the amine oxidase from Arthrobacter globiformis [30] and the alignment of its sequence with AOI from Aspergillus niger show that the Glul45 lies in the substrate channel leading to the active site, hence such a link may be created artificially in the derivatized enzyme. Quite recently, model compounds related to the proposed structure of AO-I cofactor have been studied. On the basis of their redox and spectral properties, the authors concluded that the cofactor has been misidentified. The above mentioned carboxylate ester may be considered an unlikely candidate for a biologically functional quino-cofactor [65]. The case of lysyl oxidase from bovine aorta is different. Here the native cofactor is formed by linked lysyl and tyrosyl side-chains (lysyl-tyrosyl quinone, LTQ) [6]. On the other hand.

1269 the lysyl oxidase from the yeast Pichia pastoris contains unmodified topa quinone in the active site and should be classified as a copper amine oxidase [66]. In the amine oxidase from Escherichia coli, the topa quinone was confirmed by a detailed analysis of the cofactor dipeptide X-Asp [67] and the resonance Raman spectrometry of the enzyme and its derivatives[68,69]. The primary structure of the enzyme also contains the cofactor consensus sequence [70]. More bacterial genes were shown to encode proteins containing the topa quinone consensus sequence, such as amine oxidase from Klebsiella aerogenes [71], phenethylamine oxidase and histamine oxidase from Arthrobacter globiformis [72,73], and methylamine oxidase from Arthrobacter strain PI [74]. Amino acid sequences around the position of the cofactor for selected amine oxidases from various sources are given in Table 1. The lengths of the complete amino acid sequences vary from 638 amino acids for A. globiformis phenethylamine oxidase to 762 residues for the enzyme from bovine plasma and the homology is in the broad range of 20-99%, but the important residues determining the enzyme action are conserved throughout all sequences [13]. The homology is high for the amine oxidases having the same substrate specificity obtained from similar organisms or organs, e.g., 92.2% for the amine oxidases from lentil and pea seedlings and 83.0% for the amine oxidases from E. coli and K. aerogenes, but becomes very low when comparing enzymes from bacteria and mammals (around 25%). There are 33 strictly conserved amino acid residues located mostly in the central part of the sequence (in the vicinity of topa quinone) and in the C-terminal region [13,60]. High homology of the primary structure at the C-terminus suggests its structural and functional importance. There are three conserved histidyl residues that are the copper ligands; two of which form the His-X-His motif and one is located near the C-terminus of the protein. Highly conserved asparagine and glutamate (or aspartate) residues in the consensus active site sequences of all known CAOs have been shown to be crucial in maintaining the balance of cofactor mobility versus rigidity expected to be necessary during the dual processes of biogenesis and catalysis, respectively, that all CAOs must accomplish. In addition, a structural linkage between these two highly conserved residues is proposed which spans both subunits of the dimeric CAOs, and may have implications for intersubunit communication [75,76].

1270 Table 1. Alignment of amino acid sequences of several copper amine oxidase around the position of topa quinone. The sequences were obtained by translation the corresponding cDNAs except for the enzymes from porcine kidney and porcine serum and the benzylamine oxidase from Hansenula polymorpha where they were determined by automated Edman degradation of peptides. Homologous consensus sequence around the cofactor is underlined, the tyrosyl precursor of topa quinone is shown as y.

Source of amine oxidase

Amino acid sequence

Reference

Bovine serum (monoamine oxidase)

SVSTMLNyDYVWDMVFYPNGAIE

[57]

Porcine serum (monoamine oxidase)

SVSTMLNxDYVXDMIFHP*

[48]

Porcine kidney (diamine oxidase)

DTSTVYNxDYIXDFIFYYN*

[48]

Porcine aorta (semicarbazide sensitive amine oxidase)

NxDYY*

[52]

Human kidney (diamine oxidase)

TTSTVYNyPYIWDFIFYPNGVME

[53]

Human placenta (diamine oxidase)

TTSTVYNyPYIWDFIFYPNGVME

[56]

Rat colon (diamine oxidase)

TTSTVYNyPYIWDFIFYSNGVME

[55]

Pea seedlings (diamine oxidase)

TIVTVGNyPNVIDWEFKASGSIK

[60]

Lentil seedlings (diamine oxidase)

TVVTVGNyDNVLDWEFKTSGWMK

[59]

Chick pea seedlings (diamine oxidase)

TVVTVGNyPNVLDWEFKTSGWSI

[61]

Arabidopsis thaliana (ataol)

MVATLGNyPYIVDWEFKKSGAIR

[62]

Aspergillus niger (monoamine oxidase, AO-I)

FIITLANyEYIFAYKFPQSGGIT

[27]

Hansenula polymorpha (methylamine oxidase)

QIFTAANyEYCLYWVFMQPGAIR

[46]

Hansenula polymorpha (benzylamine oxidase)

VANxEYV*

[47]

Escherichia coli (monoamine oxidase)

WISTVGNyPYIFDWVFHPNGTIG

[70]

Klebsiella aerogenes (tyramine oxidase)

WISTVGNyPYIFDWVFHPNGTIG

[71]

Arthrobacter globiformis (phenethylamine oxidase)

FFTTIGNyPYGFYWYLYLPGTIE

[72]

Arthrobacter globiformis (histamine oxidase)

FFTTVGNyPYGFYWYLYLPGTIE

[73]

Arthrobacter sixdAU PI (methylamine oxidase)

FIATVANyEYAFYWHLFLDGSIE

[74]

*t[ie unknown amino acid x in the consensus sequence represents the derivatized topa quinone in the sequenced peptide

1

1271 Catalytic mechanism of copper amine oxidases Each subunit of the copper amine oxidase binds one Cu(II) ion. The two cupric ions are indistinguishable in the EPR spectrum. Spectroscopic properties of the copper bound in the enzyme are similar to those of low molecular tetragonal, square-pyramidal or planar complexes of Cu(II) of so called type 2 "nonblue" copper [12,77,78]. Parameters of the EPR spectrum for the enzyme-bound copper correspond to tetragonal complex of Cu(II) with N- and O- ligands [79]. Changes in the EPR spectrum after the reaction of the enzyme with phenylhydrazine or after anaerobic addition of substrate suggest possible interaction of Cu(II) and the organic cofactor [80]. A model of the Cu(II) complex in copper amine oxidases was constructed on the basis of several spectroscopic studies [81-84]. The copper is coordinated in square-pyramidal complex by four equatorial ligands, three nitrogen atoms of histidyl residues, one water molecule, and one axial water ligand. The accuracy of the model was recently confirmed by resolving the crystal structure of the amine oxidase from Escherichia coli [28], as shown in Fig. (2). The distance of the Cu(II) and quinone group of the cofactor favors direct electron transfer between both prosthetic groups [28,29,85]. However recently it has been found that the electron transfer proceeds rather via an integrated water network [31]. The copper can be reversibly removed from the active site by reaction with diethyldithiocarbamate under non-reducing conditions [86-88], or by cyanide after reduction by dithionate to Cu(I) [78]. The catalytic activity of the enzyme can be restored with high yield by addition of free Cu(II) ions to the apoenzyme [78,86]. In the case of pea seedling amine oxidase, addition of other bivalent metal ions does not lead to reactivation [86]. However the activity can be partially restored (from 15%) for the amine oxidase from bovine plasma by adding Co(II) [89]. Addition of Co(II) and Ni(II) can restore the original spectrum of the native enzyme with bovine serum amine oxidase reduced by dithionate [90]. The main role of copper in the active site is to keep the essential amino acid residues in a geometry favorable for substrate binding [78,91] and participate in the reoxidation of the substrate reduced organic cofactor [86,88]. Hypotheses that the copper acts as a Lewis acid [92] or only indirectly participates in the catalysis [78] were based on unsuccessful attempts to detect changes in its redox state by EPR spectroscopy [93].

1272

Fig. (2). Ribbon diagram of the three-dimensional crystal structure of copper amine oxidase from Escherichia coli [28]. Similar structures of amine oxidases from pea seedlings [29], Arthrobacter globiformis [30] and Hansenulapolymorpha [31] lack the domain Dl.

It has been found only recently by EPR spectroscopy that a transition state of Cu(I)-semiquinone radical that might be the looked-for catalytic intermediate directly reacting with oxygen, is generated during anaerobic reduction of the enzyme with a substrate at a laboratory temperature [94]. EPR spectrum of this radical is identical to that of substrate reduced amine oxidases measured in the presence of Cu(I)-stabilizing ligands [84,94,95]. This observation lead to the model of the mechanisms of catalytic reaction as is shown in Fig. (3). The catalytic cycle is composed of two phases [21,96-98]. In the anaerobic phase, the interaction of topa quinone with the primary amino group of a substrate leads to the formation of a quinoketimine. Abstraction of the a-proton of the substrate by a catalytic base (amino acid residue of the active site) then forms a carbanion that is quickly converted to colorless quinoaldimine. The quinoaldimine is then hydrolyzed and releases the product aldehyde, forming a reduced form of the cofactor, aminoresorcinol [99,100]. The Cu(II)-aminoresorcinol formed then exists in equilibrium with the radical

1273

Fig. (3). Mechanism of the substrate oxidation by copper amine oxidases [29]. The scheme shows the roles of copper, topa quinone cofactor and proton abstracting base (Asp) in the catalytic cycle. The oxidized enzyme (a) reacts with an amine substrate giving a Schiff base formation at C-5 of the TPQ (b-c), followed by proton abstraction (d). After hydrolysis and release of the aldehyde, an aminoresorcinol species is formed (e), and the reduced cofactor is reoxidized by molecular oxygen via Cu(I)-semiquinone intermediate (/).

form of Cu(I)-semiquinone, semiquinolamine. Its reaction with molecular oxygen in the aerobic phase leads to release of hydrogen peroxide and ammonia, and reoxidation of the cofactor that completes the catalytic cycle [97-100]. The equilibrium between the intermediates Cu(I)semiquinone and Cu(II)-aminoresorcinol is maintained by rapid electron transfer between copper and the organic cofactor [100,101]. At low temperature, the equilibrium is shifted to the Cu(II)-aminoresorcinol state unless cyanide or other stabilizer is added [84]. Copper-depleted bovine serum amine oxidase has been recently reconstituted with Co(II) ions. The benzylamine oxidase activity of the enzyme was increased to 20% on incorporation of cobalt. Furthermore, Co(II) restored to nearly native level the intensity of the absorption spectrum and the reactions with phenylhydrazine or benzylhydrazine, which had been slowed down or abolished, respectively, in Cu(II)depleted samples. The amine oxidase activity of the Co(II)-derivative, which cannot form a semiquinone radical as an intermediate of the

1274 catalytic reaction, strongly suggests that the Cu(I)-semiquinone is not an obHgatory intermediate in the catalytic cycle of bovine serum amine oxidase [102]. Recently, evidence has been obtained by X-ray absorption spectroscopy that the Cu(I) complex formed after substrate reduction of the enzyme can directly react with dioxygene [103]. This observation is further supported by steady state-kinetics exploring isotopic effects on the oxidative half-reaction of amine oxidase [104]. On the basis of a recently determined X-ray structure for Hansenula polymorpha amine oxidase the authors predicted the existence of a hydrophobic oxygen-binding pocket that is located near the 0-2 position of the reduced cofactor [31], Fig. (1). X-ray crystal structures of three species related to the oxidative half of the reaction of the copper amine oxidase from Escherichia coli have now been determined. Crystals were freeze-trapped either anaerobically or aerobically after exposure to substrate, and then the respective 3-D structures were determined. The oxidation state of the quinone cofactor was investigated by single-crystal spectrophotometry. The structures revealed the site of bound dioxygen and the proton transfer pathways involved in oxygen reduction [105]. The quinone cofactor is regenerated from the iminoquinone intermediate by hydrolysis involving Asp383, the catalytic base in the reductive half-reaction. Product aldehyde inhibits the hydrolysis, making release of product the rate-determining step of the reaction in the crystal. Spectral properties of copper amine oxidases The absorption spectrum of the copper amine oxidases shows a characteristic broad band at around 500 nm (460-510 nm) that confers a typical pink or yellow-pink color to highly purified enzyme preparations [10,12,32]. Absorption in the visible region is caused by the presence of the quinone cofactor and thus is not affected by removal of copper from the enzyme under non-reducing conditions [12,32]. The quinone cofactor shows emission of fluorescence when excited at 280 and 365 nm [41,106]. Electron transition of Cu(II) is seen in circular dichroism spectra at 600-800 nm [12,32,107]. Under aerobic conditions, addition of the substrate leads to a temporary bleaching of the pink color of the enzyme, which is restored after complete consumption of the substrate. However, anaerobic conditions lead to the formation of a stable yellow intermediate

1275 with the absorption maxima at 350, 435 and 465 nm [86,88]. Addition of oxygen then causes rapid reoxidation of the enzyme and restoration of the absorption at 500 nm [99]. The process of reoxidation requires the presence of enzyme bound copper. It does not proceed with a copper depleted and substrate reduced apoenzyme [86,88]. Characteristic absorption spectra of intermediates formed during the catalytic turnover of copper amine oxidases are given in Fig. (4).

O

a cd

o <

300

400

500

600

Wavelength (nm) Fig. (4). Characteristic absorption spectra of intermediates formed during the catalytic turnover of copper amine oxidases. Resting oxidized enzyme (a), Cu(I)-semiquinoIamine radical (b), and Cu(II)-aminoresorcinol (c) [98].

The enzymes form intensively colored derivatives of phenylhydrazine that were often used for titration of the carbonyl group of the organic cofactor [32]. There is one functional active site containing the cofactor and Cu(II) per subunit [108-110], thus giving the stoichiometry of the titration 2 moles of phenylhydrazine reagent per mole of the enzyme

1276 dimer. The 1:1 stoichiometry shown in older papers [80,111] is now explained by lower purity of the enzyme preparations used, differences in determination of molecular mass and protein concentration [10,112]. Some copper amine oxidases show so called "half-site'' reactivity when the binding of phenylhydrazine to one of the active sites changes the conformation of the second site in such a way that considerably slows down the reaction rate or even completely blocks binding to the second site [113,114]. The stoichiometry of the titration is also affected by defective formation of the quinone cofactor from its precursor [112]. Mechanism based inactivators of copper amine oxidases Any inhibitor that requires enzyme processing before it inhibits the same enzyme is a mechanism-based inhibitor [115]. Such inhibitor is usually a substrate or product analogue, which binds to the active site being converted to an intermediate that eventually inactivates the target enzyme upon turnover. In most cases, the inhibitor can be turned over many times before the inactivation occurs. Mechanism-based enzyme inactivation must fulfil the following criteria [116]: time-dependent loss of enzyme activity with pseudo-first order and saturation kinetics, kinetic protection by normal substrate, a rate of inactivation proportional to the concentration of inactivator and irreversibility of the inactivation. Studies on mechanism-based inhibitors of copper amine oxidases have shown high dynamics during recent years [116]. Indeed, disclosure of molecular mode of enzyme inactivation could bring new insights into the active site structure [117] Early studies performed with bovine plasma and porcine kidney amine oxidases have shown that the enzymes undergo irreversible inactivation upon reaction with several acetylenic substrates (propargylamine, 2chloroallylamine and 2-butyne-l,4-diamine), which was diminished by substrate protection [118]. Other types of mechanism-based inactivators of bovine plasma amine oxidase are some glycine esters with relatively acidic a-protons. These esters are converted to ketenes, which may acylate the active site and inactivate the enzyme [119]. Of the group of plant copper amine oxidases, the enzymes from pea and lentil have been studied thoroughly as regards their substrate and inhibitory properties [116]. An acetylenic analogue of the substrate putrescine, 2-butyne-l,4-diamine (DABI), has been described as a typical

1277 mechanism-based inhibitor of plant CAOs [120]. The aminoaldehyde product of DABI oxidation is in equilibrium with its reactive aminoallenic form that attacks an essential nucleophile at the enzyme active site. Covalently bound pyrrole moiety (UV absorption maximum at 310 nm) is then formed and the enzyme is consequently inactivated by blocking the substrate channel entrance [117,121]. Other unsaturated putrescine analogues, (£")- and (Z)-isomer of 2-butene-l,4-diamine have been originally reported as good substrates of pea seedling amine oxidase [122,123]. The respective position isomers of 4-amino-2-butenal formed as the oxidation products can further undergo transformation to intramolecular dehydratation at a higher temperature. The pyrrole formed can be easily trapped by Ehrlich's reagent in a colored reaction. Recently, 2-butene-l,4-diamines have been quite surprisingly mentioned as mechanism-based inactivators of porcine kidney amine oxidase [124]. The haloamines 2-bromoethylamine and 2-chlorethylamine together with the substrate analogue ethane-1,2-diamine were found to be both poor substrates and irreversible inhibitors of the lentil seedling enzyme where their inhibition mechanism has been demonstrated [125]. The irreversible inactivation is caused by the aldehydes produced during the reaction which attack a highly reactive species of the quinone cofactorderived free radical catalytic intermediate. When the aldehydes react with the free radical formed in the holoenzyme at anaerobiosis, a covalent modification of the enzyme occurs [125]. By contrast, the corresponding propylamine compounds 3-bromopropylamine, 3-chloropropylamine and propane-1,3-diamine are only reversible inhibitors of lentil amine oxidase [125]. Propane-1,3-diamine had been found not to be substrate of plant CAOs in earlier studies [86]. However, recently published interpretations of kinetic results showed the compound as an inactivator of pea amine oxidase [126]. The oxidation of several indoleamines by lentil amine oxidase leads to irreversible loss of the enzymatic activity only in the absence of oxygen due to formation of a stable adduct between indoleacetaldehydes and the semiquinone form of the topa quinone cofactor [127]. Plant amine oxidase from Euphorbia characias is not sensitive to hydrogen peroxide in the absence of substrate [128]. In the presence of substrate and in absence of catalase, however, either hydrogen peroxide formed by the reaction or more rapidly the hydrogen peroxide added before initiating the enzymatic reaction causes total inactivation. The inactivation is considered to be related to the sulfhydryl group

1278 modification [128]. Bovine serum amine oxidase, reduced by an excess of substrate amine under limited turnover conditions, was over 80% inactivated by hydrogen peroxide upon oxygen exhaustion. The UV-Vis spectrum and the reduced reactivity with carbonyl reagents showed that the cofactor topa quinone was stabilized in reduced form. The inactivation reaction appears to be a general feature of copper-containing amine oxidases, It may be part of an autoregulatory process in vivo, possibly relevant to cell adhesion and redox signaling [129]. Crystal structures of copper amine oxidases Several amine oxidases have been crystallized including the enzymes from bovine and porcine blood plasma [130,131], porcine kidney [132], pea seedling [133], Aspergillus niger [63], Hansenula polymorpha [31], Escherichia coli [134,135], and Arthrobacter globiformis [136]. Crystals diffracting X-rays to high resolution (< 2.5 A) were first obtained for pea seedling amine oxidase [133] and shortly thereafter for the enzyme from E. coli [135]. Crystals of both enzymes belong to the orthorhombic crystallographic group. The three-dimensional protein structure was first solved for the enzyme from E. coli [28], with a resolution of 2.0 A. Each subunit of the mushroom-shaped dimer is comprised of four domains, a large C-terminal P-sandwich domain D4, which contains the active site and provides the dimer interface, two smaller peripheral a/p domains D2 and D3 and the "stalk" Dl. The active sites are buried in the protein and lie some 35 A apart connected by a pair of p-hairpin arms. The protein structure of the amine oxidase from E. coli is shown in Fig. (2). Crystals of active and inactive forms of the enzyme were prepared using different crystallization media. The differences between the two forms were only in the vicinity of the active site. Both forms contained cupric ion bound in complex with three histidyls, His524, His526 and His689. The inactive form crystallized from ammonium sulfate had a tetragonal copper coordination where the fourth ligand was the C-4 oxygen of topa quinone, while the active form crystallized from a citrate buffer showed a square-pyramidal copper coordination with three histidyls and one water as equatorial ligands and another water molecule as an axial ligand. Topa quinone was turned away from the copper, being located close to Asp383, which is the probable catalytic base abstracting the proton in the early phase of the

1279 catalytic cycle. Surprisingly there were also other metal sites of unknown catalytic importance found in the structure, located close to the protein surface. The structure interpretation of the E. coli enzyme was further enhanced by obtaining the coordinates for the enzyme derivatized with an irreversible inhibitor 2-hydrazinopyridine that mimics the substrate binding [137]. The data confirmed the role of Asp383 as the catalytic base and showed point of entry of the substrate/inhibitor to the active site. Other roles of Asp383 have been recently explored in more depth [138]. The results showed that the aspartate residue acts not only as the active site base at different stages of the catalytic cycle but also in regulating the mobility of the TPQ that is essential to catalysis. The crystal structure of native pea seedling amine oxidase was solved one year later [29] at a resolution of 2.2 A from crystals grown in lithium sulfate. The structure is similar to that ofE. coli except for the absence of a "stalk" domain Dl, so both subunits of the dimer are composed of three domains. As in the E. coli amine oxidase, the largest domain consisting mostly of p-sheets contains the active site that is inaccessible directly to the solvent. Both subunits are connected by a pair of hairpin arms, tightly embracing each other. The copper is coordinated by three histidyls (His442, His444 and His603) and two water molecules. Topa quinone does not take part in the coordination, its C-2 oxygen lies 6 A apart from the copper. All oxygen atoms of topa quinone make hydrogen bonding to other amino acid residues of the active site as shown in Fig. (5). There are two disulfide bridges per subunit and some potential glycosylation sites. Also another metal site, probably occupied by Mn(II) was found at a distance of 33 A from copper. Signals of Mn(II) were earlier detected in EPR spectrum, but were thought to be an impurity [111]. Recently it has been confirmed for other plant amine oxidases that there is only a low content of Mn(II) in the enzyme having no catalytic importance [49,51]. The recent structure of the recombinant phenethylamine oxidase from Arthrobacter globiformis [30] showed the conformation for active and inactive forms of the holoenzyme and for the copper/topa quinone free apoenzyme. Basic structural parameters are in agreement with previous structural studies. The data provided further evidence for the proposed biogenesis of the cofactor and substrate entry into the active site, localizing precisely the substrate channel and its residues. A novel feature that was not described previously is the solvent filled cavity at the major interface between the two subunits of the dimer. The location of the

1280 substrate channel was further explored by chemical modificatin and sitespecific mutagenesis [139]. Recently, carbon monoxide complexes have been generated for amine oxidases from A. globiformis and Aspergillus niger and characterized by various spectroscopic measurements [140]. The results obtained indicate that the coordination structure of the Cu(I) ion in the Cu(I)/semiquinone state may be modulated by the chemical and redox states of the TPQ cofactor.

Asn386

Lys296

Tyr286

His444

Fig. (5). Sterical conformation of the active site of pea amine oxidase [29].

The most recent crystallographic study discloses the structure of the methylamine oxidase from the yeast Hansenula polymorpha [31], which shows an integrated network of water molecules providing electron transfer from topa quinone to copper and other important features such as the channel for oxygen entry and hydrogen peroxide release. The role of the active site aspartate base (Asp319) in the aminotransferase mechanism of the copper amine oxidase from H. polymorpha has been probed by sitedirected mutagenesis [141]. It has been demonstrated by several

1281 instrumental experiments that the residue has a central role in positioning the free cofactor and several enzyme intermediates for optimal activity. Different molecular features of polyamine oxidases Contrary to the copper-containing enzymes, flavin polyamine oxidases show very restricted substrate specificity oxidizing only the polyamines spermine and spermidine by attacking secondary amino group [1,15]. As a member of mammalian PAOs, the enzyme from rat liver can be mentioned. The enzymes from Penicillium chrysogenum and Aspergillus tereus represent microbial PAOs [16]. During recent years, interest has been growing especially on the field of plant flavin polyamine oxidases. Plant PAOs have been isolated from monocotyledonous plants, namely from cereals (maize, oat, barley) belonging to the Gramineae [15]. The enzyme from maize has been subjected to intensive biochemical and physiological studies. Maize PAO is a monomeric glycosylated (3%) protein with a molecular mass of 60 kDa. The primary structure of maize PAO does not share any similarity with that of plant copper amine oxidases [142]. Plant PAOs have been found to be localized in the cell wall. They participate on the production of hydrogen peroxide utilized in peroxidase-mediated cross-linking reaction leading to lignification and suberization of cell wall [143]. These processes are enhanced under stress conditions caused by wounding or pathogen infection. The only polyamine oxidase that has been crystallized and whose three-dimensional structure has been solved up to date is the enzyme of maize seedlings [18]. The X-ray crystal structure of native PAO from maize has been determined at 1.9 A resolution. The enzyme consists of 13 a helices and 19 p strands, which fold to form two well-defined domains. The FAD-binding domain comprises three fragments, whose main structural elements are a central parallel p sheet flanked by a P meander and three a helices. The substrate-binding domain is composed of two fragments and is characterized by a six-stranded mixed p sheet flanked by five a helices. The two domains create a tunnel that defines the enzyme active site at their interface as shown in Fig. (6). The folding topology of maize PAO resembles that of several other flavoenzymes such as glucose oxidase or D-amino acid oxidase [18].

1282 The glycosylation site in maize PAO molecule has been identified to be Asn77 [18]. The FAD cofactor is non-covalently bound to the protein and is deeply buried within the structure. The isoalloxazine ring of FAD is located at the interface of the two domains [18]. With the exception of the flavin C5a, N5 and C4a atom that line the active site, all FAD atoms are solvent-inaccessible. The conformation of the oxidized flavin is nonplanar - the orientation might be important in precisely aligning the cofactor with respect to the polyamine substrate. The PAO active center consists of a remarkable "U-shaped" tunnel, which passes through the protein structure at the interface between the two above-mentioned domains [18]. The tunnel extends to a length of about 30 A. The U-shape brings its two openings onto the same side of the protein surface. The turning point, around which the tunnel sharply bends and reverses its direction, represents the core of the catalytic center, where the flavin ring is located.

Fig. (6). The 3-D structure of monomeric FAD-containing polyamine oxidase from maize seedlings [18] completely differs from that of copper amine oxidases.

There is a marked contrast in the chemical nature of the two arms of the U-shaped catalytic center [18]. One arm is lined mainly by aromatic residues and it opens to the outside like a funnel with several acidic side-

1283 chains (Asp, Glu) on its rim. In contrast, the other arm contains mostly oxygen atoms on its surface and displays a narrow entrance. In this respect, the ring of Asp and Glu residues seems to be suited to fulfil the role of guiding the polyamine substrate into the tunnel. Thus the substrate might be admitted into its binding site preferentially through only one of the two tunnel openings. Unravelling the configuration of maize PAO active site allowed elucidation of the difference between substrate specificity of copper amine oxidases and flavin polyamine oxidases. Polyamine substrates in PAO active site are bound and bent at an orientation that facilitates attack of the substrate secondary amino group by the flavin cofactor [18].Biogenesis of the quinone cofactor in copper amine oxidases Biogenesis of the quinone cofactor of copper amine oxidases Topa quinone is a ubiquitous cofactor of copper amine oxidases, since it is present in mammalian as well as in plant and microbial enzymes [13]. Considerable attention has been given to the process of topa quinone formation from the structural gene-encoded tyrosyl precursor. Recent studies suggest that it probably happens via an autocatalytic mechanism in the presence of protein-bound Cu(II) [47,144]. Recombinant enzyme obtained by the expression of the yeast Hansenula polymorpha amine oxidase structural gene in the yeast Saccharomyces cerevisiae was enzymatically active and the substrate specificity was the same as for the wild type enzyme. Since S. cerevisiae lacks the ability to form endogenous amine oxidase, it has no specific enzymatic system for the conversion of tyrosyl precursor to topa quinone [144]. Mutant H456N prepared by site directed mutagenesis of His456 (a possible copper ligand) to Asn is not enzymatically active and shows a very low ability to bind Cu(II) compared to the wild type enzyme. Although mutation did not affect enzyme expression in S. cerevisiae and purification yield, mutant H456N contained no detectable amount of topa quinone. Exchange of Glu for Asn in the consensus sequence did not affect topa quinone production in vivo. Enzymatic activity of the E406N mutant was not much altered compared to the normal enzyme [145]. Autocatalytic formation of topa quinone in prokaryotic copper amine oxidases has already been demonstrated several times in vitro. Escherichia coli K-12 produced almost inactive amine oxidase after

1284 mutating the signal sequence of the gene encoding the E. coli amine oxidase. After incubating the enzyme without any effector for 30-60 minutes at 30°C, it was several fold activated. While topa quinone was not detected in the inactive enzyme, the redox-cycling reaction showing its presence was positive after the enzyme activation [146]. In the absence of copper in the cultivation media, E. coli produced an inactive form of recombinant phenethylamine oxidase of Arthrobacter globiformis, which did not contain copper and the quinone cofactor. Activation of this enzyme form was possible by its aerobic incubation with Cu(II) that led to simultaneous topa quinone generation [147-149]. The oxidation of the tyrosyl precursor to topa quinone was effected only by the active sitebound Cu(II) [149]. It was assumed from a preliminary spectroscopic study of the activation of the recombinant enzyme under anaerobic conditions that the copper redox state is changing during the process of precursor conversion (the electron donor should be a hypothetical tyrosyl radical) [147]. Measurement of the EPR and CD spectra, however, did not confirm the formation of Cu(I) [149]. The proposed route of topa quinone formation from Tyr466 in amine oxidase from E. coli [21] is shown in Fig. (7). In the first step, the reduction of Cu(II) to Cu(I) mediated by tyrosyl precursor is assumed, although there is so far no experimental evidence for the transient existence of such a radical. Copper can then bind the molecular oxygen and generate an intermediate analogous to a superoxide radical; reaction 2, Fig. (7). Its reaction with a tyrosyl radical leads to a peroxide intermediate bound to copper; reaction 3, Fig. (7). Such a reaction between tyrosyl and superoxide radicals has been described recently [150]. In the next step, the 0 - 0 bond is spliced homolytically with the oxygen transfer to the position 2 on the aromatic ring of tyrosyl, see Fig. (1). Topa quinone arises by the consequent autooxidation, reactions 4-6 in Fig. (7). The proposed mechanism requires the rotation of the side chain, but it appears more feasible than the previously suggested process, which includes dopa and dopa quinone as intermediates [47], reactions 8 and 9, in Fig. (7). Under that hypothesis, the topa quinone is formed by a nucleophilic attack of a hydroxyl bound to Cu(II) to dopa quinone.

1285

Cu2*

OH

Cu*

O'

Cu-0

Cu-OH HO

OH

Dopa

O2

Cu^*

CUTOH

V HO—/

V-OH

c /

)

OH

,0,

/ -

^

0

V

Topa quinone

Dopa quinone

Fig. (7). Probable mechanisms of topa quinone biogenesis as proposed for the enzyme from Escherichia coii [21],

Recent spectroscopic studies suggest, however, that this could happen only with the assistance of an active site lysyl residue [148]. Such a lysyl residue is found in the crystal structure of the active site of pea seedling amine oxidase [29], but is absent from E. coli amine oxidase [28]. The process of redox-active cofactor formation in phenethylamine oxidase and histamine oxidase of A. globiformis was recently analyzed by Raman spectroscopy using isotopic exchange. It was found that the oxygen on the

1286 C-2 of topa quinone is provided by a solvent (water) during the autocatalytic oxidation of the cofactor precursor. The origin of the C-5 oxygen was not reliably shown, but it is likely that by analogy with other enzyme-catalyzed hydroxylation reactions it comes from the air [151]. Resonance Raman spectroscopic studies also imply that only C-5 oxygen has a real carbonyl character; while the C-2 and C-4 oxygen bear a delocalized negative charge [152]. It has been also discovered by a sitedirected mutagenesis of the histamine oxidase from A. globiformis, that the presence of unchanged consensus sequence is necessary for the optimal formation of topa quinone and thus activation of the enzyme [153]. Recent crystallographic study on the structure of the recombinant amine oxidase from A. globiformis provided further evidence for the proposed autocatalytic mechanisms [30]. The autocatalytic formation of topa quinone has been also demonstrated in vitro with the eukaryotic copper amine oxidase [154]. Inactive apoenzyme of the recombinant amine oxidase from Hansenula polymorpha (purified after expressing the gene in E. coli) was activated by an incubation with Cu(II). Further work [75,76] explains the importance of amino acid residues adjacent to topa quinone for the integrity of the active site. When either Asn or Glu (or Asp) from the cofactor consensus sequence Asn-TPQ-Asp/Glu were mutated, the recombinant enzymes contained topa quinone in a nonproductive orientation that resulted in accumulation of an inactive Schiff base complex of the cofactor with substrate. Using the crystallographic coordinates for this enzyme [31], the amino acid residues adjacent to topa quinone were found to participate in a structural loop bound together by hydrogen bonding. The loop includes four other amino acid residues, two of them from the other enzyme subunit. The integrity of the loop is essential to keep the proper structure of the active site as well as the connectivity between the subunits that may possibly play the key role in intersubunit communication resulting in "half-site" reactivity with some inhibitors [113,114]. The effect of Cu(II) ions on amine oxidase production in lentil seedlings has been described previously [155]. While the amount of specific mRNA and the protein was not affected by exogenous addition of Cu(II), addition of cupric ions to the homogenized seedlings grown in its absence led to an essential increase in the enzymatic activity. On the other hand, the addition of copper into cultivating media increased the enzyme expression in the case of monoamine oxidase of E. coli [156]. Several 4-

1287 tert-butyl-derived models for the putative intermediates of topa quinone generation and studied the effect of Cu(II) ions on each autoxidative step from dopa- to topa quinone-like compounds at physiological pH [157]. The results obtained indirectly confirm the formation of dopa and dopa quinone during topa quinone biosynthesis. Following the formation of dopa, the role of the active-site copper ion in topa quinone biogenesis would be limited to the catalysis of the two subsequent quinonization steps (i.e. from dopa to dopa quinone and from topa to topa quinone), thus disfavoring the possibility of a direct intervention of the metal ion in the hydroxylation of dopa quinone. In particular, Cu(II) was shown to influence deeply the autoxidation of l,2,5-trihydroxy-4-tert-butylbenzene, used as model of topa, both increasing the reaction rate and changing its mechanism. The function of a strictly conserved tyrosine located within hydrogenbonding distance to TPQ has been explored by employing site-directed mutagenesis on the enzyme from H. polymorpha to form the mutants Y305A, Y305C, and Y305F. The relative effects of mutations at Y305 on catalytic turnover are dependent on the nature of the amino acid which substitutes for tyrosine and the substrate used in amine oxidase assays. Despite the strict conservation of this residue in all CAOs, neither biogenesis nor catalytic turnover were abolished upon mutation of this residue. An important, but nonessential, role for Tyr305 in the positioning of the TPQ precursor for biogenesis, and in the maintenance of the correct conformation for TPQ-derived intermediates during catalytic turnover has been postulated [158]. The biogenesis of TPQ has been also characterized in amine oxidase from H. polymorpha expressed as the apo-enzyme in E. coli. The results obtained suggest that the tyrosine precursor is activated for the reaction with oxygen by liganding to Cu(II) [159]. Molecular oxygen is consumed in a single, exponential phase, the rate of which responds linearly to dissolved oxygen concentration. Binding of oxygen appears to occur faster than its consumption and to result in displacement of the precursor tyrosine onto copper to form a charge-transfer species. Reaction between this intermediate and oxygen is proposed to occur in a rate-limiting step, and to proceed more rapidly when the tyrosine is deprotonated. This rate-limiting step in cofactor biogenesis does not display a solvent isotope effect and is, thus, uncoupled from proton transfer [160]. Quite recently, the cDNA coding for pea seedlings amine oxidase has been cloned and a heterologous expression system for the cloned enzyme

1288 was constructed with the yeast Pichia pastoris. Adding copper to the culture medium increased the secretion of an active, quinone-containing enzyme. Furthermore, the inactive enzyme produced in a copper-deficient medium was activated considerably by subsequent incubation with excess cupric ions. These resuhs strongly suggest that the topa quinone cofactor is produced in the plant enzyme by post-translational modification that proceeds through the copper-dependent, self-processing mechanism, as in the enzymes from bacteria and yeast [161]. Physiological studies on the role of copper amine oxidases in plants The metabolic importance and exact physiological function of the amine oxidase in plants has been a question of long standing debate, but only recent reports provided closer insight into the problem. Early papers in 70s show that the biosynthesis of the enzyme in pea {Pisum sativum) cotyledons is several fold depressed by light. The activity reached a maximum at 5-8 days of development and in etiolated plants retains its higher activity even if later exposed to light, which means that only the initial phase of the enzyme biosynthesis is light modulated [162,163]. The role of the enzyme in plant senescence, wound healing and defense has been the subject of speculation [7]. In the early 90s, evidence was obtained by combining biochemical and histochemical techniques that the same spatial and fimctional correlation occurs between locations of amine oxidase and peroxidase activities, the key enzyme of lignification, in chick pea {Cicer arietinum) stems upon de-etiolation and wounding [164,165]. A similar pattern was later observed for chick pea exposed to the fungal pathogen Ascochyta rabiei [166]. Pathogenic attack or mechanical wounding induces the level of specific mRNA and subsequently specific activity of copper amine oxidase in plants [61]. It was shown for chick pea epicotyls injured with a blade that increase in the respective mRNA is detectable 3 h after the wounding, reaches a maximum at 9 h and then declines to reach similar level to unwounded plant. The specific activity has a similar trend during the first phases after the wounding and continues to increase up to 48 h, which suggests that the newly synthesized enzyme has a very slow turnover. It seems that transcriptional modulation may be the major event governing the phase of rapid increase in enzyme activity, whereas pottranscriptional control mechanisms may become important in the later

1289 phase of the response, when the level of the amine oxidase transcript decreases. The kinetics of the amine oxidase accumulation matches closely that observed for elicitors, pathogens and wound-induced mRNAs encoding phytoalexin biosynthesis enzymes and enzymes of phenylpropanoid pathway. Therefore, the data suggest that the enzyme has a protective role in defense mechanisms. Involvement of the amine oxidase in the synthesis of the lignin-suberin protective barrier during wound healing was shown in experiment with plants, where the amine oxidase was irreversibly inhibited by aminoguanidine which strongly reduced the lignin-suberin deposition along the wound compared to control plants [61]. Histochemical determination of the amine oxidase showed parallel effects on the amine oxidase activity. Amine oxidase activity was very high along the lesion, as well as in the xylem. This may be due to oxidation of endogenous polyamines released during senescence [164]. Aminoguanidine did not affect the peroxidase activity. The data strongly support the hypothesis that amine oxidase serves as a hydrogen peroxide delivering agent in the peroxidase mediated synthesis of the lignin-suberin protective barrier during wound healing. The possible role of the amine oxidase in plant development was recently unravelled by making transgenic Arabidopsis thaliana plants, in which the promoter region of the amine oxidase gene was transcriptionally fused with the reporter genes encoding P-glucuronidase and modified green fluorescent protein [62]. Analysis of the trangenic Arabidopsis together with in situ hybridization of wild-type plants revealed temporally and spatially discrete patterns of gene expression in lateral root cap cells, vascular tissue of roots, developing leaves, the hypocotyl and in the style/stigmatal tissue. Histochemical analysis confirmed that the enzyme expression overlaps with lignification of vascular tissues. In both vascular tissue and the root cap, the enzyme expression occurs in cells destined to undergo programmed cell death, the apoptosis. The amine oxidase seems to serve also in defense and vascular redifferentiation of cells after attack by nematodes. Recently it was observed that challenging the above mentioned transgenic Arabidopsis with Meloidogyme incognita induces amine oxidase expression around the feeding site [167]. Polyamine catabolism in plants, which is mediated by CAOs and PAOs, results in the formation of aminoaldehydes along with hydrogen

1290 peroxide and ammonia. In plants, the aminoaldehydes formed are further metabolized by the effect of NAD-dependent aminoaldehyde dehydrogenases (EC 1.2.1.19, 1.2.1.54). Their specific role in plant physiology is still unknown. Nevertheless, recent investigations have indicated possible homology of the enzymes with plant betaine aldehyde dehydrogenases (EC 1.2.1.8), which participate in betaine aldehyde accumulation in response to water deficit and increased salinity [168]. The dehydrogenases have been found in legumes and grasses [169]. However, the enzymes have not yet been purified to homogeneity and extensively characterized. The best substrates of aminoaldehyde dehydrogenases, 3-aminopropanal, 4-aminobutanal and 4-guanidinobutanal, are oxidized to the respective co-amino acids[170-172]. Conclusions and future trends In enzymology, physiology and protein structure studies, the main research interest in the field of copper amine oxidases is now focused on explaining the topology of the active site after substrate binding. For this reason, the study of mechanism-based inhibitors has major importance, since their conversion in the active site gives products that bind covalently to essential groups causing the inactivation of the enzyme. The exact role of copper in the reaction mechanism and autocatalytic formation of topa quinone cofactor from tyrosyl radical is being fiirther investigated. Also carbohydrate moieties in amine oxidases have not yet been reasonably characterized and nor has their function elucidated. Recently, a study on the structure of the N-linked glycans in the enzyme from porcine kidney has appeared that utilized various analytical techniques, including matrix-assisted laser desorption/ionization time-of flight spectrometry in conjunction with specific exoglycosidases, highperformance capillary zone electrophoresis and ionex chromatography with pulsed amperometric detection. The structures found in the glycoprotein are primarily linear, di-, or-tribranched fucosylated complex type. Other analyses revealed the presence of several di- and trisialylated structures [173]. Due to experimental difficulties concerning the purification and crystallization of mammalian amine oxidases, we are still waiting for the crystal structure of such an enzyme. Another issue, which is under current investigation, is the cellular localization of the enzymes and study of the factors affecting their gene expression in the host

1291 organisms. Regarding the overproduction of recombinant amine oxidase, the important findings are that the catalytic activity and even substrate specificity of the enzyme can be aUered by a site-directed mutagenesis of some amino acid residues in the vicinity of the topa quinone cofactor. The most important issue remaining to be solved is the exact physiological role of copper amine oxidases in higher organisms. Further studies can eventually lead to practical outputs, exploitation of the enzymes, e.g., in clinical testing and food analysis or agriculture. Promising are the pioneer studies on the application of the amine oxidase in the construction of biosensors for amine assays and enzyme reactors in the food industry [174-177]. In human medicine, the plasma amine oxidase level is related to pregnancy, enhanced levels in specific tissues then signal tumor growth. The HIV virus is in vitro rapidly inactivated by the peroxidase system, where the hydrogen peroxide supply is maintained by the amine oxidase converting polyamine spermidine and spermine [178]. From fiiture directions of amine oxidase research as outlined above, one can expect further findings that will have direct implications not only in protein chemistry, but also in understanding their physiological function. ABBREVIATIONS CAO D ABI EPR FAD HPLC LTQ MS NMR PAO PQQ TPQ TTQ

= copper-containing amine oxidase = 2-butyne-1,4-diamine = electron paramagnetic resonance = flavin adenine dinucleotide = high performance liquid chromatography = lysyl-tyrosyl quinone = mass spectrometry = nuclear magnetic resonance = flavin-containing polyamine oxidase = pyrroloquinoline quinone = topa quinone = trptophan tryptophylquinone

1292

ACKNOWLEDGEMENTS This paper is dedicated to Professor-emeritus Lumir Macholan from the Department of Biochemistry, Facuhy of Science, Masaryk University in Brno, Czech RepubHc. The authors would like to thank him for his sincere encouragement, helpful discussions and supply of synthesized substrates and inhibitors throughout the studies on copper amine oxidases in Olomouc, the Hana region of Moravia, Czech Republic. Mr. Alexander Oulton is thanked for critical reading of the manuscript. REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15] [16] [17]

Cohen, S.S. A Guide to the Polyamines, Oxford University Press: New York, 1998. Tabor, H. In Structure and Functions of Amine Oxidases', Mondovi, B., Ed.; CRC Press: Boca Raton, 1985; pp. 1-4. Salisbury, S.A.; Forest, H.A.; Cruse, W.B.T.; Kennard, O. Nature, 1979, 280, 843-844. Janes, S.M.; Mu, D.; Wemmer, D.; Smith, A.J.; Kaur, S.; Maltby, D.; Burlingame, A.L.; Klinman, J.P. Science, 1990, 248, 981-987. Mclntire, W.S.; Wemmer, D.E.; Christoserdov, A.; Lidstrom, M. Science, 1991, 252,817-824. Wang, S.X.; Mure, M.; Medzihradszky, K.F.; Burlingame, A.L.; Brown, D.E.; Dooley, D.M.; Smith, A.J.; Kagan, H.M.; Klinman, J.P. Science, 1996, 273, 1078-1084. Bachrach, U. In Structure and Functions of Amine Oxidases', Mondovi, B., Ed.; CRC Press: Boca Raton, 1985, pp. 5-20. Hartmann, C; Klinman, J.P. Biochemistry, 1991, 30,4605-4611. Bardsley, W.G. In Structure and Functions of Amine Oxidases; Mondovi, B., Ed.; CRC Press: Boca Raton, 1985; pp. 135-152. Medda, R.; Padiglia, A.; Floris, G. Phytochemistry, 1995, 39, 1-9. Malmstrom, B.C.; Andreasson, L.E.; Reinhammar, B. In The Enzymes', Boyer, P.D., Ed.; Academic Press: New York, 1975, Vol. 12, pp. 507-527. Mclntire, W.S.; Hartmann, C. In Principles and Applications of Quinoproteins', Davidson, V.L., Ed.; Marcel Dekker: New York, 1992, pp. 97-171. Frebort, I.; Adachi, O. J. Ferment. Bioeng., 1995, 80, 625-632. Padiglia, A.; Medda, R.; Lorrai, A.; Murgia, B.; Pedersen, J.Z.; Finazzi-Agro, A.; Floris, G. Plant Physiol, 1998, 117, 1363-1671. Federico, R.; Angelini, R. In Biochemistry and Physiology of Poly amines in Plants', Slocum, R.D.; Flores, H.E., Eds.; CRC Press: Boca Raton, 1991, pp.4156. Morgan, D.L. Biochem. Soc. Trans., 1985, 13, 322-326.. Smith, T.A.; Croker, S.J.; Loeffler, R.S.T. Phytochemistry, 1986, 25, 683-690.

1293 [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42]

Binda, C; Coda, A.; Angelini, R.; Federico, C; Ascenzi, P.; Mattevi, A. Structure, 1999, 7, 265-276. Cogoni, A.; Padiglia, A.; Medda, R.; Segni, P.; Floris, G. Plant Physiol., 1991, 95, 477-479. Padiglia, A.; Medda, R.; Paci, M.; Sette, M.; Lorrai, A.; Floris, G. Biochem. Mol. Biol. Int., 1997,41,407-413. Fontecave, M.; Eklund, H. Structure, 1995, 3, 1127-1129. Mondovi, B.; Riccio, P. In Structure and Functions of Amine Oxidases; Mondovi, B., Ed.; CRC Press: Boca Raton, 1985; pp 63-76. Slocum, R.D.; Furey, M.J. Planta, 1991, 183, 443-450. McDougall, G.J.; Stewart, D.; Morrison I.M. Planta, 1994, 194, 9-14. Hare, P.D.; van Staden, J. Physiologia Plantarum, 1994, 91, 128-136. Galuszka, P.; Sebela, M.; Luhova, L.; Zajoncova, L.; Frebort, I.; Stmad, M.; Pe^, P. J. Enz. Inhib., 1998, 6, 457-463. Frebort, I.; Tamaki, H.; Ishida, H.; Pee, P.; Luhova, L.; Tsuno, H.; Halata, M.; Asano, Y.; Kato, Y.; Matsushita, K.; Toyama, H.; Kumagai, H.; Adachi, O. Eur. J. Biochem., 1996, 237, 255-265. Parsons, M.R.; Convery, M.A.; Wilmot, CM.; Yadav, K.D.S.; Blakeley, V.; Comer, A.S.; Phillips, S.E.V.; McPherson, M.J.; Knowles, P.F. Structure, 1995, 3, 1171-1184. Kumar, V.; Dooley, D.M.; Freeman, H.C.; Guss, J.M.; Harvey, I.; McGuirl, M.A.; Wilce, M.C.J.; Zubak, V.M. Structure, 1996, 4, 943-955. Wilce, M.C.; Dooley, D.M.; Freeman, H.C.; Guss, J.M.; Matsunami, H.; Mclntire, W.S.; Ruggiero, C.E.; Tanizawa, K.; Yamaguchi, H. Biochemistry, 1997,36, 16116-16133. Li, R.; Klinman, J.P.; Mathews, F.S. Structure, 1998, 6, 293-307. Finazzi-Agro, A. In Structure and Functions of Amine Oxidases; Mondovi, B., Ed.; CRC Press: Boca Raton, 1985; pp 121-125. Mondovi, B.; Costa, M.T.; Finazzi-Agro, A.; Rotilio, G. Arch. Biochem. Biophys., 1967, 119,373-381. Finazzi-Agro, A.; Guerrieri, P.; Costa, M.T.; Mondovi, B. Eur. J. Biochem., 1977,74,435-440. Lobenstein-Verbeek, C.L.; Jongejan, J.A.; Frank, J.; Duine, J.A. FEBS Lett., 1984, 170,305-309. Ameyama, M.; Hayashi, M.; Matsushita, K.; Shinagawa, E.; Adachi, O. Agric. Biol. Chem., 1984, 48, 561-565. Duine, J.A. Eur. J. Biochem., 1991, 200, 271-284. van der Meer, R.A.; Jongejan, J.A.; Frank, J.; Duine, J.A. FEBS Lett., 1986, 206, 111-114. Williamson, P.R.; Moog, R.S.; Dooley, D.M.; Kagan, H.M. J. Biol. Chem., 1986, 261, 16302-16305. Moog, R.S.; McGuirl, M.A.; Cote, C.E.; Dooley, D.M. Proc. Natl. Acad Sci. a^".^., 1986,83, 8435-8439. Glatz, Z.; Kovaf, J.; Macholan, L.; Pee, P. Biochem. J. , 1987, 242, 603-606. Knowles, P.F.; Pandeya, K.B.; Rius, F.X.; Spencer, CM.; Moog, R.S.; McGuirl, M.A.; Dooley, D.M. Biochem. J., 1987, 241, 603-608.

1294

[43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64]

[65] [66]

van der Meer, R.A.; van Wassenaar, P.D.; van Brouwershaven, J.H.; Duine, J.A. Biochem. Biophys. Res. Commun., 1989, 159, 726-733. Citro, G.; Verdina, A.; Galati, R.; Floris, C ; Sabatini, S.; Finazzi-Agro, A. FEBS L^r/., 1989,247,201-204. Brown, D.E.; McGuirl, M.; Dooley, D.M.; Janes, S.M.; Mu, D.; Klinman, J.P. J. BioL Chem., 1991, 266, 4049-4051. Bruinenberg, P.G.; Evers, M.; Waterham, H.R.; Kuipers, J.; Amberg, A.C.; Ab, G. Biochim. Biophys. Acta, 1989, 1008, 157-167, Mu, D.; Janes, S.M.; Smith, A.J.; Brown, D.E.; Dooley, D.M.; Klinman, J.P. J. Biol. Chem., 1992, 267, 7979-7982. Janes, S.M.; Palcic, M.M.; Seaman, C.H.; Smith, A.J.; Brown, D.E.; Dooley, D.M.; Mure, M.; Klinman, J.P. Biochemistry, 1992, 31, 12147-12154. Sebela, M.; Luhova, L.; Frebort, I.; Hirota, S.; Faulhammer, H.G.; Stuzka, V., Pea, P. J. Exp. Bot., 1997, 48, 1897-1907. Zajoncova, L.; Sebela, M.; Frebort, 1.; Faulhammer, H.G.; Navratil, M.; Ped, P. Phytochemistry, 1997, 45, 239-242. Sebela, M.; Luhova, L.; Frebort, I.; Faulhammer, H.G.; Hirota, S.; Zajoncova, L.; Stuzka, v.; Pee, P. Phytochem. Anal., 1998, 9, 211-222. Holt, A.; Alton, G.; Seaman, C.H.; Loppnow, G.R.; Szpacenko, A.; Svendsen, I.; Palcic, M.M. Biochemistry, 1998, 37,4946-4957. Barbry, P.; Champe, M.; Chassande, O.; Munemitsu, S.; Champigny, G.; Lingueglia, E.; Maes, P.; Frelin, C; Tartar, A.; Ullrich, A.; Lazdunski, M. Proc. Natl. Acad. Sci. U.S.A., 1990, 87, 7347-7351. Imamura, Y.; Noda, S.; Mashima, Y.; Kudoh, J.; Oguchi, Y.; Shimizu, N. Genomics, 1998, 51, 293-298. Lingueglia, E.; Renard, S.; Voiley, N.; Waldmann, R.; Chassande, O.; Lazdunski, M.; Barbry, P. Eur. J. Biochem,, 1993, 216, 679-687. Zhang, X.; Mclntire, W.S. Gene, 1996, 179, 279-286. Mu, D.; Medzihradsky, K.F.; Adams, G.W.; Mayer, P.; Hines, W.M.; Burlingame, A.L.; Smith, A.J.; Cai, D.; Klinman, J.P. J. Biol. Chem., 1994, 269, 9926-9932. Hogdall, E.V.; Houen, G.; Borre, M.; Bundgaard, J.R.; Larsson, L.I,; Vuust, J. Eur. J. Biochem., 1998, 251, 320-328. Rossi, A.; Petruzzelli, R.; Finazzi-Agro, A. FEBS Lett., 1992, 301,253-257. Tipping, A.J.; McPherson, M.J.J. Biol. Chem., 1995, 270, 16939-16943. Rea, G.; Laurenzi, M.; Tranquilli, E.; D'Ovidio, R,; Federico, R.; Angelini, R. FEBS Letters, 1998,437, 177-182. Moller, S.G.; McPherson, M.J. Plant J., 1998, 13, 781-791, Frebort, I.:In Ph.D. Thesis, Yamaguchi University, Japan, 1997. Frebort, I.; Pec, P.; Luhova, L.; Toyama, H.; Matsushita, K.; Hirota, S,; Kitagawa, T,; Ueno, T.; Asano, Y.; Kato, Y.; Adachi, 0. Biochim. Biophys. Acta, 1996, 1295, 59-72. Melville, C.R.; Green, E.L.; Sanders-Loehr, J.; Klinman, J.P. Biochemistry, 2000, 39, 7589-7594. Dove, J.E.; Smith, A.J.; Kuchar, J.; Brown, D.E,; Dooley, D,M.; Klinman, J.P. FEBS Lett., 1996, 398, 231-234.

1295

[67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92]

Steinebach, V.; Groen, B.W.; Wijmenga, S.S.; Niessen, W.M.A.; Jongejan, J.A.; Duine, J.A. Anal. Biochem., 1995, 230, 159-166. Cooper, R.A.; Knowles, P.F.; Brown, D.E.; McGuirl, M.; Dooley, D.M. Biochem. J., 1992,288,337-340. Moenne-Loccoz, P.; Nakamura, N.; Steinebach, V.; Duine, J.A.; Mure, M.; Klinman, J.P.; Sanders-Loehr, J. Biochemistry, 1995, 34, 7020-7026. Azakami, H.; Yamashita, M.; Roh, J.H.; Suzuki, H.; Kumagai, H.; Murooka, Y. J. Ferment. Bioeng., 1994, 77, 315-319. Sugino, H.; Sasaki, M.; Azakami, H.; Yamashita, M.; Murooka, Y. J. BacterioL, 1992, 174,2485-2492. Tanizawa, K.; Matsuzaki, R.; Shimizu, E.; Yorifiiji, T; Fukui, T. Biochem. Biophys. Res. Commun., 1994, 199, 1096-1102. Choi, Y.-H.; Matsuzaki, R.; Fukui, T.; Shimizu, E.; Yorifuji, T.; Sato, H.; Ozaki, Y.; Tanizawa, K. J. Biol. Chem., 1995, 270, 4712-4720. Zhang, X.; Fuller, J.H.; Mclntire, W.S. J. BacterioL, 1993, 175, 5617-5627. Cai, D.Y.; Dove, J.; Nakamura, N.; Sanders-Loehr, J.; Klinman, J. P. Biochemistry, 1997, 36, 11472-11478. Schwartz, B.; Green, E.L.; Sanders-Loehr, J.; Klinman, J.P. Biochemistry, 1998, 37, 16591-16600. Rotilio, G. In Structure and Functions of Amine Oxidases', Mondovi, B., Ed.; CRC Press: Boca Raton, 1985; pp 127-134. Suzuki, S.; Sakurai, T.; Nakahara, A. Biochemistry, 1986, 25, 339-341. Dooley, D.M.; McGuirl, M.A. Inorg. Chim. Acta, 1986, 123, 231-236. Rinaldi, A.; Floris, G.; Sabatini, S.; Finazzi-Agro, A.; Giartosio, A.; Rotilio, G.; Mondovi, B. Biochem. Biophys. Res. Commun., 1983, 115, 841-848. Scott, R.A.; Dooley, D.M. J. Am. Chem. Soc, 1985, 107, 4348-4350. Baker, G.J.; Knowles, P.F.; Pandeya, K.B.; Rayner, J.B. Biochem. J., 1986, 237, 609-612. McCracken, J.; Peisach, J.; Dooley, D.M. J. Am. Chem. Soc.,\9^1, 109, 40644072. Dooley, D.M.; McGuirl, M.A.; Cote, C.E.; Knowles, P.F.; Singh, I.; Spiller, M.; Brown, R.D.; Koenig, S.H. J. Am. Chem. Soc, 1991, 113, 754-761. Steinebach, V.; de Jong, G.A.H.; Wijmenga, S.S.; de Vries, S.; Duine, J. Eur J. Biochem., 1996, 238, 683-689. Hill, J.M.; Mann, P.J.G. Biochem. J., 1964, 91, 171-182. Mondovi, B.; Rotilio, G.; Costa, M.T.; Finazzi-Agro, A.; Chiancone, E.; Hansen, R.E.; Beinert, H. J. Biol. Chem., 1967, 242, 1160-1167. Rinaldi, A.; Giartosio, A.; Floris, G.; Medda, R.; Finazzi-Agro, A. Biochem. Biophys. Res. Commun., 1984, 120,242-249. Agostinelli, E.; Morpurgo, L.; Wang, C; Giartosio, A.; Mondovi, B. Eur J. Biochem., 1994, 222, 727-732. Suzuki, S.; Sakurai, T.; Nakahara, A.; Manabe, T.; Okuyama, T. Biochemistry, 1983,22, 1630-1635. Giartosio, A.; Agostinelli, E.; Mondovi, B. Biochem. Biophys. Res. Commun., 1988, 154,66-72. Yadav, K.P.S.; Knowles, P.F. Eur J. Biochem., 1981, 114, 139-144.

1296

93]

Grant, J.; Kelly, I.; Knowles, P.F.; Olsson, J.; Petersson, G. Biochem. Biophys. Res. Commun., 1978, 83, 1216-1224. 94] Dooley, D.M.; McGuirl, M.A.; Brown, D.E.; Turowski, P.N.; Mclntire, W.; Knowles, P.P. Nature, 1991, 349, 262-264. 95] Finazzi-Agro, A.; Rinaldi, A.; Floris, G.; Rotilio, G. FEBS Lett., 1984, 176, 378380. 96] Steinebach, V.; de Vries, S.; Duine, J. J. Biol. Chem., 1996, 271, 5580-5588. 97] Medda, R.; Padiglia, A.; Bellelli, A.; Sarti, P.; Santanche, S.; Finazzi Agro, A.; Floris, G. Biochem. J., 1998, 332, 431-437. 98] Medda, R.; Padiglia, A.; Bellelli, A.; Pedersen, J.Z.; Finazzi Agro, A.; Floris, G. FEBS Lett., 1999,453, 1-5. 99] Bellelli, A.; Finazzi-Agro, A.; Floris, G.; Brunori, M. J. Biol. Chem., 1991, 266, 20654-20657. 100] Turowski, P.N.; McGuirl, M.A.; Dooley, D.M. J. Biol. Chem., 1993, 268, 1768017682. 101 Williams, T.J.; Falk, M.C. J. Biol. Chem,, 1986, 261, 15949-15954. 102 Agostinelli, E.; De Matteis, G; Mondovi, B; Morpurgo, L. Biochem. J., 1998, 103 104 105 106

107 108 109 110 111 112 113 114 115 116 117 118

330,383-387. Dooley, D.M.; Scott, R.A.; Knowles, P.F.; Colangelo, C.M.; McGuirl, M.A.; Brown, D.E. J. Am. Chem. Soc, 1998, 120, 2599-2605. Su, Q.; Klinman, J.P. Biochemistry, 1998, 37, 12513-12525. Wilmot, C. M.; Hajdu, J.; McPherson, M. J.; Knowles, P. F.; Phillips, S. E. V. Science, 1999,286, 1724-1728. Frebort, I.; Toyama, H.; Matsushita, K.; Kumagai, H.; Pec, P., Luhova, L., Adachi, O. In Biochemistry of Vitamin B^ and PQQ, Marino, G.; Sannia, G.; Bossa, F., Eds.; Birkhauser Verlag: Basel, 1994, pp. 241-245. Rinaldi, A.; Floris, G.; Giartosio, A. In Structure and Functions of Amine Oxidases; Mondovi, B., Ed.; CRC Press: Boca Raton, 1985; pp. 51-62. Janes, S.M.; Klinman, J.P. Biochemistry, 1991, 30, 4599-4605. Padiglia, A.; Medda, R.; Floris G. Biochem. Int., 1992, 28, 1097-1107. He, Z.; Zou, Y.; Greenaway, F. Arch. Biochem. Biophys., 1995, 319, 185-195. Nylen, U.; Szybek, P. Acta Chem. Scand B, 1974, 28, 1153-1160. Steinebach, V.; Benen, J.A.E.; Bader, R.; Postma, P.W.; de Vries, S.; Duine, J.A. Eur. J. Biochem., 1996, 237, 584-591. Morpurgo, L.; Agostinelli, E.; Mondovi, B.; Avigliano, L.; Silvestri, R.; Stefancich, G.; Artico, M. Biochemistry, 1992, 31, 2615-2621. Frebort, 1.; Toyama, H.; Matsushita, K.; Adachi, O. Biochem. Mol. Biol. Int., 1995,36, 1207-1216. Waley, S.G. Biochem. J., 1985, 227, 843-849. Padiglia, A.; Medda, R.; Pedersen, J.Z.; Lorrai, A.; PeC, P.; Frebort, 1.; Floris, G, J. £«Z./A7/7/^., 1998, 13,311-325. Frebort, I.; Sebela, M.; Svendsen, I.; Hirota, S.; Endo, M.; Yamauchi, O.; Bellelli, A.; Lemr, K., Pee, P. Eur. J. Biochem, 2000, 267, 1423-1433. Hevey, R.C., Babson, J., Maycock, A.I., Abeles, R.H. J. Am. Chem. Soc, 1973, 95,6125-6127.

1297

[119] Maycock, A. L.; Suva, R. H.; Abeles, R. H. J. Am. Chem. Soc, 1975, 97, 56135614. [120] Pee, P.; Frebort, I. Eur. J. Biochem., 1992, 209, 661-665. [121] Frebort, 1.; Pee, P.; Luhova, L.; Matsushita, K.; Toyama, H.; Adachi, O. Eur. J. Biochem., 1994, 225, 959-965. [122] Macholan, L.; Hubalek, F.; Subova, H. Collect. Czech. Chem. Commun., 1975, 40, 1247-1265. [123] Pee, P.; Chudy, J.; Macholan, L. Biologia (Bratislava), 1991, 46, 665-672. [124] He, Z.; Nadkarni, D.V.; Sayre, L.M.; Greenaway, F.T. Biochim. Biophys. Acta, 1995, 1253, 117-127. [125] Medda, R.; Padiglia, A.; Pedersen, J.Z.; Agro, A.F.; Rotilio, G.; Floris, F. Biochemistry, 1997, 36, 2595-2602. [126] Awal, H.M.A.; Hirasawa, E. Phytochemistry, 1995, 39, 489-490. [127] Medda, R.; Padiglia, A.; Finazzi Agro, A.; Pedersen, J.Z.; Lorrai, A.; Floris, G. Eur. J. Biochem, 1997, 250, 377-382. [128] Floris, G.; Giartosio, A.; Rinaldi, A. Arch. Biochem. Biophys., 1983, 220, 623627. [129] Pietrangeli, P.; Nocera, S.; Fattibene, P.; Wang, X. T.; Mondovi, B.; Morpurgo, L. Biochem. Biophys. Res. Commun., 2000, 267, 174-178. [130] Yamada, H.; Yasunobu, K. J. Biol. Chem., 1962, 237, 1511-1516. [131] Buffoni, F.; Blaschko, H. Proc. Roy Soc. B., 1964, 161, 153-164. [132] Yamada, H.; Kumagai, H.; Kawasaki, H.; Matsui, H.; Ogata, K. Biochem. Biophys. Res. Commun., 1967, 29, 723-727. [133] Vignevich, V.; Dooley, D.M.; Guss, J.M.; Harvey, I.; McGuirl, M.A.; Freeman, H.C. J. Mol. Biol., 1993, 229, 243-245. [134] Roh, J.H.; Suzuki, H.; Azakami, H.; Yamashita, M.; Murooka, Y.; Kumagai, H. Biosci. Biotech. Biochem., 1994, 58, 1652-1656. [135] Roh, J.H.; Suzuki, H.; Kumagai, H.; Yamashita, M.; Azakami, H.; Murooka, Y.; Mikami, B. J. Mol. Biol., 1994, 238, 635-637. [136] Freeman, H.C; Guss, J.M.; Kumar, V.; Mclntire, W.S.; Zubak, V.M. Acta Cryst. D52, 1996, 197-198. [137] Wilmot, C.A.; Murray, J.M.; Alton, G.; Parsons, M.R.; Convery, M.A.; Blakeley, v.; Comer, A.S.; Palcic, M.M.; Knowles, P.F.; McPherson, M.J.; Phillips, S.E.V. Biochemistry, 1997, 36, 1608-1620. [138] Murray, J.M.; Saysell, C.G.; Wilmot, C. M.; Tambyrajah, W.S.; Jaeger, J.; Knowles, P.F.; Phillips, S.E.V.; McPherson, M.J. Biochemistry, 1999, 38, 82178227. [139] Matsuzaki, R.; Tanizawa, K. Biochemistry, 1998, 37, 13947-13957. [140] Hirota, S.; Iwamoto, T.; Tanizawa, K.; Adachi, O.; Yamauchi, O. Biochemistry, 1999,38, 14256-14263. [141] Plastino, J.; Green, E.L.; Sanders-Loehr, J.; Klinman, J.P. Biochemistry , 1999, 38,8204-8216. [142] Tavladoraki, P.; Schinina, M.E.; Cecconi, F; Di Agostino, S; Manera, F; Rea, G.; Mariottini, P.; Federico, R.; Angelini, R. FEBSLett., 1998,426, 62-66. [143] Laurenzi, M.; Rea, G.; Federico, G.; Tavladoraki, P.; Angelini, R. Planta, 1999, 208, 146-154.

1298

[144] Cai, D.; Klinman, J.P. Biochemistry, 1994, 33, 7647-7653. [145] Cai, D.; Klinman, J.P. J. BioL Chem., 1994, 269, 32039-32048. [146] Hanlon, S.P.; Carpenter, K.; Hassan, A.; Cooper, R.A. Biochem. J., 1995, 306, 627-630. [147] Matsuzaki, R.; Fukui, T.; Sato, H.; Ozaki, Y.; Tanizawa, K. FEBS Lett., 1994, 351,360-364. [148] Matsuzaki, R.; Suzuki, S.; Yamaguchi, K.; Fukui, T.; Tanizawa, K. Biochemistry, 1995,34,4524-4530. [149] Ruggiero, C.E.; Smith, J.A.; Tanizawa, K.; Dooley, D.M. Biochemistry, 1997, 36, 1953-1959. [150] Jin, F.; Leitich, J.; von Sonntag, C. J. Chem. Soc. Perkin Trans., 1993, II, 15831588. [151] Nakamura, N.; Matsuzaki, R,; Choi, Y.-H.; Tanizawa, K. J. Biol Chem., 1996, 271,4718-4727. [152] Nakamura, N.; Moenne-Loccoz, P.; Tanizawa, K.; Mure, M.; Suzuki, S.; Klinman, J. P.; Sanders-Loehr, J. Biochemistry, 1997, 36, 11472-11486. [153] Choi, Y.-H.; Matsuzaki, R.; Suzuki, S.; Tanizawa, K. J. BioL Chem., 1996, 271, 22598-22603. [154] Cai, D.; Williams, N.K.; Klinman, J.P. J. Biol. Chem., 1997, 272, 19277-19281. [155] Rossi, A.; Maccarrone, M.; Buganza, M.; Finazzi-Agro, A. Gene, 1995, 154, 167170. [156] Takenaka, Y.; Roh, J.H.; Suzuki, H.; Yamamoto, K.; Kumagai, H. J. Ferment. Bioeng.,\991,S3, 194-196. [157] Rinaldi, A.C.; Porcu, CM.; Oliva, S.; Curreli, N; Rescigno, A; Sollai, F; Rinaldi, A; Finazzi-Agro, A; Sanjust, E. Eur. J. Biochem., 1998, 251, 91-97. [158] Hevel, J.M.; Mills, S.A.; Klinmann, J.P. Biochemistry, 1999, 38, 3683-3693. [159] Dove, J.E.; Schwartz, B.; Williams, N.K.; Klinman, J.P. Biochemistry, 2000, 39, 3690-3698. [160] Schwartz, B.; Dove, J.E.; Klinman, J.P. Biochemistry, 2000, 39, 3699-3707. [161] Koyanagi, T; Matsumura, K; Kuroda, S; Tanizawa, K. Biosci. Biotech. Biochem., 2000,64,717-722. [162] Macholan, L.; Minaf, J. Biologia Plantarum, 1974, 16, 86-93. [163] Suresh, M.R.; Ramakrishna, S.; Adiga, P.R. Phytochemistry, 1976, 15,483-485. [164] Angelini, R.; Manes, F.; Federico, R. Planta, 1990, 182, 89-96. [165] Scalet, M.; Federico, R.; Angelini, R. J. Plant Physiol., 1991, 137, 571-575. [166] Angelini, R.; Bragaloni, M.; Federico, R.; Infantino, A.; Portapuglia, A. J. Plant Physiol, 1993, 142,704-709. [167] Moller, S.G.; Urwin, P.E.; Atkinson, H.J.; McPherson, M.J. Physiol Mol Plant Pathol., \99S, 53,13-79. [168] Trossat, C; Rathinasabapathi, B.; Hanson, A. Plant Physiol., 1997, 113, 14571461. [169] Flores, H.E.; Filner, P. Plant Growth Reg., 1985, 3, 277-291. [170] Matsuda, H.; Suzuki, Y. Plant Physiol., 1984, 76, 654-657. [171] Awal, H.M.A.; Yoshida, I.; Doe, M.; Hirasawa, E. Phytochemistry, 1995, 40, 393-396.

1299

[172] Awal, H.M.A.; Kinoshita, T.; Yoshida, I.; Doe, M.; Hirasawa, E. Phytochemistry, 1997,44,997-1000. [173] Huang, Y; Mechref, Y.; Novotny, M.V. Carbohydrate Res., 2000, 323, 111-125. [174] Wimmerova, M.; Macholan, L. Biosens. Bioelectron., 1999, 14, 695-702. [175] Frebort, 1.; Skoupa, L.; Pee, P. Food Control, 2000, 11, 13-18. [176] Niculescu, M.; Nistor, C; Frebort, I.; Pec, P. Matiasson, B.; Csoregi, E. Anal. Chem., 2000,12, 1591-1597. [177] Niculescu, M.; Frebort, I.; PeC, P.; Galuszka, P.; Mattiasson, B.; Csoregi, E. Electroanalysis, 2000, 12, 369-375. [178] Klebanoff, S.J.; Kazazi, F.J. Clin. Microbiol., 1995, 33, 2054-2057.

This Page Intentionally Left Blank

1301 SUBJECT INDEX AAPM (2,2'-azobis (2-amidino propane) dihydrochloride) 763 Aargesin 241 ABTS(2,2'-azino-Z?/5-(3-ethylthiazoline6-suIfonic acid) 1014,1015,1020 oxidation of 1016 Acacia auriculoformis 467 Acacia nilotica 358 AcaciasideAB 467 Acanthella cavernosa 465 A canthe Ila klethra 810 Acanthophyllum gyps ophihides 3 Acanthophyllum squarrosum 31,50,55 squarroside A from 31 Acer sp. 905 Acer nikoense 889,895,899 Acerogenin A-L 899 AcerosideXI 895 Aceroside IV-V 899 2-Acetamidofluorene 1003 Acetaminophen 257 (+)-l-Acetoxypinoresinol 200 (+)-1 -Acetoxypinoresinol-jJ-D-glu 200 1 -O-Acetyl-4-methoxy lycorine 100,615 acetylation of secologanin 100,615 15-Acetylcardiopetamine 869 (9-Acetyl-isopseudolycorine 614 1-O-Acetyllycorine 615 Acicidites oriental is 1214 Aciculites sp. 1175 Aciculitins A-C 1214 trans-kcomixQ acid 793,794 Aconitine 868,869,870,871 Aconitum 868 Acorus caninum 471 Acorus graminous 362,470 A ctinidia deliciosa 1145,1158 Actinocin 1009 Actinomadiira 451 Actinomycin D (Dactinomycin) 1009 Actinomycin 1009 Actinomycin synthetase I, II and III 1010

Acuminatin 199 Acute rheumatism 398,400 Acyclovir 226 Acylfulvenes 713 Adenocarcinoma 712 ^'-Adenosylmethionin-decarboxylase inhibitor 830 Adjuvant activity 55 of Gypsophila 55 of Qiiillaia 55 of Saponaria 55 Aedes aegy^ptii 439 Aegilops ovata (Gramineae) 187 Anatoxin B-1 271,304 African trypanosomiasis (HAT,sleeping sickness) 784 Agelas oroides 819 Aglaia elliptica 218 aglaiastatin from 220 4' -demethoxy-3' ,4' -methy lenedioxymethyl rocaglate from 218 1 -oxo-4'-demethoxy-3',4'-methylenedioxyrocaglaol from 218 4'-demethoxy-3',4'-methylenedioxyrocaglaol from 218 l-0-formyl-4'-methylenedioxymethyl rocaglaol from 218 Aglaia odorata 220 pyrimidinone from 220 rocaglaol from 220 Aglaiastatin 220 Agmenellum qiiadruplicatiim 358 Agrobacteriiim rhizogenes 640,641 Agropyron repens 696 Ailanthinone 815,816 Ajoenes 795,796,797 Alantolactone 463 Alaria sp. 1121 Albizia sp. 1157 Aldobiuronic acid 1157 Algal and cyanobacterial toxicity 376 testing of 376 Algicidal activity 358 Alginicacid 1150,1151,1152

1302 Alizarin 635-640,649,656-658,670-673, 875 1-methyletherof 658,673 2-methylether of 659 Alkaline phosphatase activity 378,382 3-[(a-Alkoxy)-a-methoxy]-2-thioxopyrrolidines 1097 3-[(a-Alkoxy)-a-methylthio]-methyl-2thioxopyrrolidines 1097 Allelopathic agent 187 Allescheria boydii 227 Allicin 493 Allium sativum 493,796,1135 (+)-Allokainic acid 478 Allopurinol 793 Alloxan 759 Alnus 905 Alnus maximoviczii 883 Alnus rubra 596 Alnusdiol 895 Alnusone 895 Alnusonol 895 Alpinia 905 Alpinia blepharocalyx 886,887,902 Alpinia conchigera 883 Alpinia officinarum (Zingiberaceae) 883 Alsidium corallinum 479 Alsidiiim helminthocorton 478 Alstonine 1060 Alternaria alternata 451 Alternaria helianthi 1091 Amalan 140 Amarathine 305 Amarogentin 806,807 Amaryllidaceae alkaloids 608-613,615-620 Ambrosia 437 Ambrox 394 American trypanosomiasis (Chagas disease) 784,798 Americanin A 270 Americanol A 270 Amiloride-binding proteins 1268 a-Amino-(J-carboxymuconic-£-semialdehyde 970 2-Amino-l,4-benzoquinone-3-carboxylic

acid 995 5'-([(Z)-4-Amino-2-butenyl]-methylamino) -5'deoxyadennosine 830 2-Amino-3,6-dihydroxy-benzoic acid 979 1 -(4' - Am ino-3' -butenal-2' -y l)-p-carboline 1087 2-Amino-3-hydroxy-benzoic acid 972 2-Amino-3-phenoxazionone 986 2-Amino-4,4-a-dihydro-4a,7-dimethyl3//-phenoxazin-3-one 986 2-Amino-5-[(2'-carboxy-6'-hydroxy pheny l)-amino]-1,4-benzoquinone-3carboxylic acid 977 Aminobisabolene 1201 1 -Aminocyclopropane-1 -carboxylic acid oxidase 935,950 4-Aminoquinoline (chloroquine) 781 Aminoresorcinol 1272,1273 Cu(II)-Aminoresorcinol 1273 Aminosidine (paromomycin) 831 Amoebicidal activity 458 Amoxicillin 366 c'-AMP phosphodiesterase inhibitory activity 38 of dianchinenoside E and F 38 (+)-Ampelopsin A&B 551,566 Ampelopsin C&D 558,566 c/5-Ampelopsin E 558,566 /raws-Ampelopsin E 566 Ampelopsin E-H 558,567 Ampelopsis brevipedunculata 551 Amphimedon sp. 454 Amphimedon complanata 90 Amphimedon sponge 74,88 Amphimic acid A 88,89 Amphotericin B 794,795 Anabaena circinalis 367 Anabaena doliolum 366 Anabaena sp. 352,354,358,359,364,365, 378,379 Anabaena variabilis 366 (-)-Anagyrine 487 Analgesic activity 32 of barbatosides A and B 56 of dianosides A-1 32

1303 Analgesic effects 401,403,404,406,416,418 Anaplastic carcinoma 712 Ancistrocladus heyneanus 813 Ancistrocladus konipensis 822 Ancylostoma 481 Ancylostoma duodenale 425,429 Aneuploidogen diethylstilbestrol- 271 Angelica pubescens 436 Angeloylgomisin H 197,247 Angiogenesis 757 Angiostrongylus 484 Angiotensin-II 267 Anguillula aceti 450,452,459,464,482 Anguina tritici 456 Anigopreissin A 566 Anisakis 475 larvae of 457,474,475 Annoba montana 212,243 syringaresinol from 212 (-)-syringaresinol from 243 Annonacin 441 tetraacetate of 441 Annonacinone 441 Annonin 441 Anogeissus 1157 Anogeissus acuminata var. lanceolata 224 anolignan A & B from 224 Anogeissus latifolius 1157 Anolignan B 228,231 Anopheles mosqyiWo 781,782 Antagonist effect ofdioxin 756 Antagonistic activity 460,1065 Anthelmintic activity 425,446,450,451, 456,466,476,478,481,482,487,491 Anthelmintics 426,431,444,465,478,479, 493 (+)-Anthocerotonic acid 185 Anthocyanidin 743,744,747,748,770,771 Anthocyanins 305 Anthragallol 657,659 2,3-dimethyletherof659 Anthragallol 3-methylether 659 Anthranilic acid 966,967,969,970-972, 1005,1008

Anthraquinone (alizarin) 629 Anthricus sylvestris 215 deoxypodophyllotoxin from 215 (-)-deoxypodorhizone from 215 (-)-hinokinin from 215 nemerosin from 215 morelensin from 215 Anti- inflammatory activity 884 of Curcuma comosa 884 Antiallergenic effects 264,265 Antiallergenic drug 759 Antiallergic activity 759 Antiamoebin I 482 Antiangiogenic agents 757 Anti-anxiety agent 413 Antibacterial activity 226 of dihydrodiisoeugenol 226 ofisoeugenol 227 of 5'-methoxydehydrodiisoeugenol 226 Antibacterial activity of HA 1002 Anti-bleeding activity 56 of Dianthus super bus var. longicalycinus 56 Anticancer activity 741,760,1060 of strychnos bisindole 1062 of 18,19-dihydrousambarine 1062 of 5',6'-dihydrousambarensine 1062 of strychnopentamine 1062 of usambarensine 1062 of usambarine 1062 Anti-catarrhal activity 398 Anticomplementary activity 340 Anticonvulsant agent 413 Anticonvulsive action 399 Antiestrogenic activity 221 Antiestrogenic effects 209 Antifeedant activity 461,851 Persea indica 851 Antifeedant effect 853,861,863,871,874,876 Antifilarial activity 228,443,467,485,759 Antifungal activity 227,228,553,1098 of 4'-0-demethldehdropodophy-

1304

llotoxin 227 of 3,4,3',4'-tetramethoxylignan-7-ol 228 of 2-thioxopyrrolidines 1098 Antifungal agent 368 Antifungal peptides 1184 theonellamide F 1184 Antigenic activity 1126 Antigenotoxic agent 271 Antiglycosuric activity 398 Antihepatotoxic activity 256 Antihepatotoxic agents 3-deoxysilychristin 255 silandrin 255 silchristin 255 silybin 255 silydianin 255 silymonin 255 Antihistaminic drugs 262 Antihyaluronic activity of 556 Antihypertensive effects 235,236 Antiinflammatory activity 340,341,343, 391,398,400,401,403,404,406,415-418, 482,556,558,741,759,1035 of Salvia desoleana 416 Anti-inflammatory agents 32 barbatosides A and B 56 from Dianthus barbatus 32,56 Anti-inflammatory effects 262,263,264 Antileishmanial activity 789,790,791, 800,805,807,810,811,821,828,830 Antileishmanial activity 229,790,887,813 oflignans 229,790,887,813 Antimalarial activity of 779,1036 of Cinchona succiruba 779 Antimalarial drugs 1035 Antimicrobial action 1073,1092 ofHMTP 1073,1092 Antimicrobial activity 64,330,331-334, 402,403,404, 406,417,418,443, 470, 485,1091,1092,1189 of Fraxinus ornus 334 ofhydroxycoumarins 330 of 13-methylpentadecanoic acid 64 ofphenylethanoids 330

of secoiridoids 330 of theonellapeptolides 1189 of Salvia officinalis 402 of Salvia sclarea oil 401 ofMTBI 1091,1092 ofHMPT 1091,1092 ofMMTP 1091,1092 of dithiocarbonates 1091,1092 Antimicrosporidial activity 458 Antimutagenic activity 1099,1100 ofHMTP 1099 ofMMTP 1099 ofMTBl 1100 Antimutagenic effects 510 Antinematodal activity 425,433,456,457, 464,472, 483,484,488,494 Antinematodal agents 456 Antinematodal effect 481 Antineoplastic drugs 158,169,214,799,825 Antioxidant activity 250,255,509,556, 672,763,768,887,1005,1006 of alizarin 672 of a-tocopherol 1006 of flavonoids 765 Antiparasitic Activity 800,817,819 with chalcones 800 Antiperoxidative effects 249-255 oflignans 249-255 Antiperspirant 398 Antiplamsodial naphthylisoquinoloine 822 Antiplasmodial activity 788-800,805, 806,808,810,813,815,824,1035 of dioncopeltine A & B 822,823 Antiplasmodial flavonoids 801 Antiplasmodial phloroglucinol 797,798 Antiproliferative activity 485 Antiproliferative effects 752 Antiprotozaol labdanes 810 Antiprotozoal activity 452,780,795,797, 798, 810,825,1060 of sesquiterpenes 807 Antiprotozoal anthraquinones 805 Antiprotozoal drugs 793 Antiprotozoal isoquinloines 821 from Annonaceae 821

1305

from Berberidaceae 821 from Hemandiaceae 821 from Menispaermaceae 821 Antipyretic activity 1035 Antiseptic-disinfectant 398 Antispasmogenic activity 239 Antitrypanosomal activity (Trypansosma cruzi) 792 Antitumor activity 56,482,706 of Dianthus superbus var. longicalycinus 56 of podophyllum 210 Antitumor lignans 211 austrobailignans-1 211 burseran 211 (+)-dimethylisolariciresinol-2axyloside 211 (-)-/m«^-2-(3",4"dimethyoxybenzyl)-3-(3',4'-methylene dioxybenzyl) butyrolactone 211 diphyllin211 diphyllin acetate 211 justicin-D 211 peltatins 211 podophyllotoxins 211 stegnangin 211 stegnanol211 Antitumoral immune responses 923,924,929 Antiviral activity 169,222,224,226,344, 452,741,759 of podophyllum extracts 169 ofverrucosin 228 Aphanizomenon flos-aquae 353 Aphanizomenon sp. 352,358 Aphelencoides besseyi 434,436,439, 465,469 Aphelencoides tritici 438,469 Apicomplexan 804 Apigenin 746,750,763,802 Aplysina fistularis 86 (J-Apoicropodophyllin 193 Apollonias barbujana 850 p-Apopicoropodophyllin 214 Apoptosis 752,922,925,993,1262,1264, 1289

Arabidopsis 1289 Arabidopsis cultivar 933 Arabidopsis thaliana 478,945,948,1268, 1289 Arabinogalactan 1156 Arachidonic acid 758,759 metabolism of 265-268 Arachnoideum 685 Arbortristosides A&C 807 Arctigenin 192,217,218,233,241,269 of methyl ether 192,241 Arctiin 192,237 Arctium lappa 217,233,241 arctiin from 217 arctigenin from 217 Ardisia crenata 8 Arenaria filicaulis 52 Argemone mexicana 435 Arnica montana 808 Arteether 833 Artelinate 833 Artemether 832,833 Artemia salina 703 Artemisia annua 363,783,801,831 Artemisinin 363,446,779,783,807,809, 811,825,832-834,840 Artesunate 833 Arthrobacter globiformis 1263,12681270,1272,1278-1280 Arthrobacter sp. 1122,1124 Arthrobacter strain 1263,1267,1270 ofPl 1263,1267,1270 Arthrobotrys oligospora 431,432,449 /ra^5-Aryltetral in lactone type lignans 154 Arylnaphthalene lignans 194 Aryltetrahydronaphthalene lignans 193 Aryltetralin lignans 195,196 Aryltetralins 155, 169, 170, 172 selectivity indices against HIV-1 170 Asarinin 226,231 (-)-Asarinin 260 (+)-Asarinin 267 a-Asarone 470 (J-Asarone 470 Ascaridia gain 457,484

1306 Ascaridole 431 Avermectins 425,426,442,453,494 A scar is 463,478 Avicularin 359 Ascaris lumbricoides 425,429 Axinellida 464 Ascaris suilla 443 Axinyssafenestratus A63,A6A Ascaris suum 443,477,450 Axisonitrile 810 Ascochyta visiae 797 Axisonitrile-3 812 Ascofuranone 797 Azadirachta indica 466,814 Ascorbic acid 770,993,1003,1006 azadirachtin from 466 oxidation of 742 Azaphilones 450 Asiasarum hetrotropoides var. Azotobacter vimlandii 1151,1152,1211 madshuricum 226 Azulen 305 asarininfrom 212,226 xanthoxylol from 212,226 Babesias^. 694,835 Asimicin 441 Bacillus cereus 1091 Aslchyta rabiei 1288 Bacillus subtilis 79,333,334,1093 Asparagine 1131 Bacitracin 366,374 Asparagus africanus 230,813 Bactericidal effect 1095 Asparagus officinalis All of HMTP on the strain Sacchromyces Asparagusic acid 477 epidermidis 1095 Aspergillimide A 491 Bafilodides 495 Aspergillimides 491 Bafilomycin Al 453 Aspergillus sp. 450,491 Bafilomycin B2 453 Aspergillus cladosporoides 228 Bafilomycin CI 453 Aspergillus flavus 228 Baicalin 672 Aspergillus fumigatus 228,458,673, Bakanosine 101,103,105-107,111,112,136 1093,1128 Balanocarpol 556,561 Aspergillus melleus 449 Barbatosides A and B 56 Aspergillus niger 228,307,971,1141,1144, from Dianthus barbatus 56 1263,1268,1278,1280,1283,1286,1287 flavonoids 743 Aspergillus oryzae 1007 Bemangovitra 103 7,103 8 Aspergillus tereus 1281 BengamideA&B 485 Asperula 631 BengazoleA 488 Asperuloside 666 Benzofuran lignans 253 Asperulosidic acid 666 Benzofuran neolignans 202 Aspiculamycin 493 Benzothiadiazole 846 Aspicularis tetraptera 493 Benzoylgomisin O 197 Aspidospermidine 140 Benzoylisogomisin O 247 Aspyrone 449 Benzylhistamine 119 Atisine 868 M-Benzyloxotryptamine 118 Atovaquone 840 Benzylstrictosidine 117 Atta sexdens rubropilosa 186 Benzylvincoside 117,138 Aurantoside A&B 1176,1179 Berberine 305,821,822,823 Aurones 744,801 p-Bergamotene 459 Austrobailignan-1 193 Betanin 305 Auxins 581

1307 Bisdesmoside 41 Betula sp. 894,905 from Sterllaria dichotoma 41 Betula davurica 895 Bisnordihydrotoxiferine 1050,1051 Betula ermanii 895 Bistheonellasterone 1176,1177 Betula maximowicziana 895,899 Blood thrombocytopenia 265 Betula ovalifolia 899 Bloom-forming cyanobacteria 352 Betula platyphlla var.Japomca 255 Anabaena 352 Betula platyphylla var. latifolia 895 Aphanizomenon 352 Betulatetraol 895 Cylindrospermopsis 352 BetulifolA&B 568 Lyngbya 352 Betulinic acid 813,814 Microcystis 352 Bicuculline 868 Nodularia 352 Bidesmosides 813 Nostoc 352 Bilharzia-mdiwcQdi bladder cancer 1003 Oscillatoria (Planktothrix) 352 Bioautographic TLC assays 255 Umezakia 352 Biogenesis 1054 Bocconia cordata 488 for Nb-C(21) secocuran alkaloids Bocconine 488 1054 Boerhaavia diffusa 233 ofsclareol 405 liriodendrin from 233 ofditerpenes 465 Bombyx mori 991 of Fraxinus ornus 313,314 Bomtrager reaction 644 of secologanin type alkaloids 142 Borrelia burgdorferi 695 of some iridoids 141 Botrytis cinerea 404,405 of the resveratrol oligomers 552 Botrytis cinerea 508,553,555 Biosynthesis 95,154,187,306,477,554, BPV-4 (Bovine papilloma Virus type 4) 637,703,909,1015,1151,1113,1139, 761 ofHA 1015 Brabeium 1157 of alginic acid 1151 Bracken carcinogens 705 ofanthraquinones 637 Bracken fern (Pteridium sp.) 685-692, of asparagusic acid 477 694-702,704,720,721,723,724,726 of bracken illudanes 703 Bradyrhizobium sp. 1154,1155 of C2o-carotenoids 306 Brassica campestris 187,481 of D-mannose 1113 Bras sic a oleraceae 748 of galactomannan 1139 Braxin A2 704 of iridoids 95 BraxinsAl 704 oflignans 154,187 Brefeldin 452 of lipoic acid 477 Brevilin A 809 of NO 909 Brine shrimp nauplii 720 of resveratrol oligomers 554 Brochamin 745 of the stilbenoids 508 (/?)-l-Bromo-3-methylnonanol 66 2,10-Bisaboladie-l-one 861 4-Bromobenzyltryptamine 125 Bisabolanes 457 2-(2-Bromoethyl)-l,3-dioxolane 80 Bisabolene 861 Bisbenzylisoquinolines 824,825,826,827 2'-Bromopodophyllotoxin 150,155 from Pedophyllum peltatum 151 Bisconicasterone 1177 Bromoxynil 374

1308 Brown crotylboration reaction 1241 Bniceajavamca 211 guaiacylglycerol-P-0-6'-(2-methoxy) cinnamyl alcohol from 211 Brucine 1030,1049 Brucine N-oxide 1049 Brusatol 815,816 Bryophytes 185 Btula platyphylla var. japonica 887,899 Bulgaria inquinam 450 Buparvaquone 804 Bupleurum salicifolium 229 Burasaia madagascariensis 1038 Burkholderia 582,605 Burkholderia caryophylli 594,599,601, 602,605 Burkholderia cepacia 605 Burkitt's lymphoma 926 Bursaphelenchus lignicolus 429,435,493 Bursaphelenchus xylophylus 436,442, 466,470,471,474,485,487 Bursaphelocide A&B 485 Bursehemin 192,229 Bursen 191 Bursera permollis 216 deoxypodophyllotoxin from 216 p-peltatin methyl ether from 216 Bursera permollis 216 Bursera permollis lignans 216 cholchicine 216 4'-demethoxy-3',4'-methylenedioxymethyl rocaglate 218 1 -0X0-4' -demethoxy-3' ,4' -methy 1enedioxyrocaglaol 218 4'-episteganangin 216 (-)-deoxypodophyllotoxin 218 l-(9-formyl-4'-methylenedioxymethylrocaglaol 218 methyl rocagalate 218 obavatifol 219 obovaten 219 obovatinal 219 perseal A 219 persealB 219 perseal Cfrm 219

perseal D from 219 steganangin from 216 stegancin from 216 steganoate B from 216 steganolide A from 216 (-)-yatein 218 Butea frondosa 457 Byrsocarpus coccineus 69

Cadaverine 1259,1260 Caenorhabditis elegans 427,429,432434,437,439,441,447-449,451,453,455, 457,460-462,466,468-471,473,474, 481,487,488,490,494,495,884 Caesalpinia pulcherima 1142 Caesalpinia vesicaria 1142 Cage-culture turbidostats (CCT) 378,382 Calceolariode A&B 327,328 Calcinosis cutis 675 Callipelta sp. 1210,1211,1213,1175 Callipeltins A-C 1211,1212,1213 Callipeltosides A-C 1210,1211,1213 Calmodulin 917,919,920 Ca'-Calmodulin complex 920 Calmodulin-pathways 250 Caloglossa liprieurii 478 Calothrix^A 96/S 366 Calyculins B&C 1240 Calyculins A-H 1204,1205,1240,1234, 1235 Calyx podatypa 64,65 Calyxin F 902 Calyxin G 902 Camptothecin 825,827 Camptothecine 97,144 from Camptotheca acuminata 144 Canaliculatol 562 Canavanine 932 Cancer chemopreventive activity 507,508, 510,511 Cancerostatic activity 452 Candida albicans 228,402,403,673,1093, 1124,1126,1134,1187,1203,1216,1217

1309 Candida glabrata 1126 Candida guilliermondii 1126 Candida sp. 330,331,1087,1126,1127, 1134 Candida valida 1091 Candidissimus 457 Cantharidin 457 Caraganaphenol A 535,563 Carbachol 267 Carcinogenesis 965 Carcinogenicity 677 of madder root Cardamon {Elettaria cardamomum) 474 Cardiopetamine 868,869 Cardiovascular effects 232 Cardiovascular hypertrophy 236 Cardiovascular modulating effects 507-509 Cardol (47) 443 Carex kobomugi 530 Carex pumila 551 (-)-£-viniferin from 551 Cargeenan-induced edema 341,344 Carob galactomannan 1139,1149 a-Carotene 299,300 p-Carotene 299,300,305 Carthamus tinctorius 436 Caryophyllan 601,604 P-Caryophyllene 409,413 Caryophyllene oxide 397,402 Caryophyllose 600,600,601 Caryose 603,604 Caspase-3 913 Cassia corymbosa 1140 Casticin 803 Catechin 743,744,748,757 (+)-Catechin 359 Catechin-3-gallate 359 Catecholamines 1019 Catecholic berberines 822 pessione 822 Catharanthine 139 Cathespin 913 Cattails {Typha latifolia) 358,361 Caudatoside 719,720 Caulerpa mcQvnosa. 68

Cefotaxime 333 Celastraceae 813 Cell surface glycoprotein 950 Centipedia minima (Asteraceae) 808 Centrocerus clavulatum 478 Centrolobium sclerophyllum 902 Cephaelis ipecuanhana 836 Cephalotaxaceae 185 Ceratocystis sp. 1127 Ceratonia siliqiia 1142 Ceratozamia spinosa 1158 Cercis siliquastrum 1145 Ceruloplasmin 1012,1016 //-Cetyl-cysteine 993 cGMP-radioimmunoassay 93 8 a-Chaconine 817,868,467 Chagas's disease 673 Chagiiyu 692 £-Chalcone 471 Chamigrenal 244 Chaparrinone 816,817 Cheilantes 722 Cheimonophyllal 457 Cheimonophyllon A&D 457 Cheimonophyllum candidissimus 456,457 Chelerythrine 488 Chemical composition 395,396 of Saliva sclarea Chemical correlation 117 of strictosidine and vincoside Chemical methods 552 of structural analysis Chemical transformation 101,102 ofbakankosine of secologanin 97,99 ofsweroside 103,104 Chemopreventive agents 219 genkwanin 219 (+)-lariciresinol 219 sitoindoside II 219 Chemoprophylaxis 674 Chemotaxonomic markers 1128 Chemotherapy 781,782,783 of malaria Chenopodium ambrosioides 431

1310

Cherimolin 441,442 Chick pea (Cicer arietinum ) 1268,1288 ChimanineB&D 820,821 Chinensinaphthol methyl ether 194 Chlamydia 1134 Chlamydomonas reinhardtii 364 Chlonostachydiol 451 Choleretic activity 399,400,414 Chlonostachys cylindrospora 451 Chi ore Ila enters onii 358 Chlorella vulgaris 358,372 Chlorhexidine 402 Chlorogenic acid 695 Chloroquine 1030,1035,1037,1038, 1063,1065 Chloroquine-enhancing activity 1065 Chlorphenamine 401 Chlorproguanil-dapsone 783 Chondria sp. 488 Chondria armata 479 Chondria baileyana 479 Chondriamide A&B 488 Chonistoneura fumiferana 876 Chorda 1121 Choriomeningitis virus 618 Chorisia speciosa 1157 Christofm 659 Chronic leukemia 925 Chrysin 746,762 Chrysoeriol 746 Chrysophanol 637 Chrysospermins 482 Cichorin 314 Cinachyrella alloclada sp. 67 Cinchona 835,836 Cinchona succiruba (Rubiaceae) 780,820 Cinchonine 836 Cinnabarin 1012 Cinnabarinic acid 967,1020 Cinnamomum camphor a 1039 Cinnamomum philippinense 240,249 cinnamophilin from 240,249 m^^o-dihydroguaiaretic acid from 240 Cinnamophilin 240,266,267 Cinnzeylanine 853

Cinnzeylanol 850 Cinnzeylanone 850,853 Circular dichroism 550,551 Cirsium japonicum 436 Cisplatin 213 Citrobacter freundii 402 Citronellic acid 86 L-Citrulline 931 Citrus flavonoids 744 Cladobotrin 449 Cladobotryum rubrobrunnescens (Hyphomycetes) 449 Cladobotryum varium 449 Cladosporium cladosporioides 553 Cladosporium colocasiae 1091,11093 Cladosporium cucumerinum 333,673 Cladosporium sphaerospermum 228 Cladosporium s\i. 1128 Claisen condensation 1237 Claisen orthoester rearrangement 72 Cleistanthus collinus (Euphorbiaceaae) 204 cleistanthin A&B from 204 Clitocybe acromelalga 479 Clivia miniata 619,620 Clivonine 620 Clonal expansion 921 of antigen-specific T 921 Closantel 440 Co-artemether (lumefantrine/ benflumetolartemether 783 Cochliobolus miyabeanus 460 Cochlioquinone A 460 Codium latum 1138 Coleoptera 851 Combretum 1157 Concanavalin A 268,1135 Condyloma acuminatum lignans 222 Condylomata acuminata 151 Conessine 837 Conicasterol 1176,1177 Conicasterone 1176,1177 Coniferyl alcohol 156,177 Conocarpan 202 Conoduramine 829 Conodurine 829

1311

Conservula cinisigna 698 Consolida 868 Cooperia curticei 490 Copalliferol A&B 561,563 Copper amine oxidases 1262,1271,1272, 1275,1283 Coptis japonica var. dissecta 268 Coral lis tes fulvodesm us 1213 Corallistes sp. 1214 Corall is tes undulatus 1214 Corallistins A-E 1213,1214 Corbicula sandai 1123 Corey asymmetric osmylation 1232 Coriolic acid 461 Coriolus sanguineus 1012 Corossolin 441,442 Corylus 894,905 Corynebacterium fescians 618 Corythuca 443 Cosculine 824,826 Costaricensis 484 COSY 16,29,34,37,41 ofbisdesmoside41 of dianchinenoside E 34 of dianchinenoside F 37 ofvaccarosideD29 Cotyledon bioassay 583 Coumarin glucoside esculin 318 Coumestrol 472 •H-'^N Coupling 1047 as a structural probe 1047 Coupling reactions 113,114,119 of secologanin 113,114,119 Covalent interactions of HA with 995-1001 Cowaxanthone 805,806 CPH82 170 for the treatment of rheumatoid arthritis 170 Crinum americanum 616 Crocetin 293-297,299,301-303,305,307 Crocetin glucosylation 307 Crocin 294,295,299,303,305 Crocin-l 299,300,301 Crocin-2 301

Crocin-3 301 Crocin-4 301 Crocin-5 301 Crocus sp. 297,299 carotenoids 297,299 derivatives 299 Crocus sativiis (Iridaceae) 293,294,296,302,303,304,306,307 Crynine-type alkaloids 615 Cryptococcus dimennae 618 Cryptococcus magnus 618 Cryptococcus neoformans 228,1127 Cryptococcus terreus 618 Cryptolepine 1060 Cryptolepis sanguinolenta 828,1060 cryptolepine from 828 Crystal structure 150,155 of 2'-bromopodophyllotoxin 150 Cu (I)-Semiquinone 1273 Cucurbitacin B 868 CupolamideA 1190 Curcuma coinosa 883 Curcuma longa 457,474,882,887 curcumin from 887 Curcuma sp. 905 Curcuma xanthorrhiza 883,884 Curcumin 305 Curcumin l-III 474 Curtius rearrangement 1237 Curvularia lunata 227 Cxenorhabditis remanei 471 Cyanobacteria 362,368 Cyanobacterial inhibitory compounds 357-364 Cyanobactericides 382 Cyanoepithio alkanes 1074 Cycleanine 824 Cycleatjehenine 824 Cycleatjehine 824 /3-Cyclocitral 354 Cyclocurcumin 902 Cyclolithistide A 1190 (lZ,5Z)-l,5-Cyclooctadiene 81,88 Cyclooxygease inhibitory activity 556 Cyclooxygenase 950

1312 Cyclooxygenase inhibitory activity 46 212 of jenisseensosides A-D 46 of (-)-nymphone 212 Cyclooxygenase-2 (COX-2) genes 926 of(-)-yatein213 D-Cycloserine 366 of theonellamides 1185 Cyclosporin A 486 Cyclotheonamide A-E 1179,1180 Cyclotheonamide E2 1180 Daizein 751 Cylindrocarpon olidum 471 Daldinia concentrica 470 Cylindrospermopsis raciborskii 352 Dalton's lymphoma ascites 303 Cyneromycins 444,445 Damnacanthal 673,806 Cyphostemmin A&B 567 Danshensuan B 235 Cyproteptadine 1063 Daphnandrine 824,826 Cysteine proteases 800 Daphne odorata 465 Cystic fibrosis 1151 Daunomycin 802 Cystoseira 1121 Davidiol B&D 463,564 (-)-Cytisine 487 3,4-Deacetylisoipecoside 115 Cytochrome P450 266,875 Deacetyllutessine 610 Cytokine production 921 A^„-Deacetyl-M-methylstrychnobrasiline Cytokine tumour necrosis factor a 1064,1066 (TNF-a) 916 M-Deacetylstrychnobrasiline 1064-1066 Cytokinins 581,583,584 Deglucosy lation 110,112 Cytoplasmic inhibitor protein 926 of 7-substituted bakankosine derivative Cytoprotective activity 224 112 Cytostatic effect 924 of secologanin 110 Cytotoxic activity 216-218,452,470,485 (+)-Dehydrocaffeic acid dilactone 238 Cytotoxic agents 760 Dehydroconiferyl alcohol 202 Cytotoxic T lymphocytes 950 Dehydroerucin 481 Cytotoxicity of 212 12-Dehydroporson 895 of (3-apopicropodophyllin 213 Dehydrosalviarin 868 of dehydropodphylltoxin 212 2,3-Dehydrosalvipisone 402 of dehydro-p-peltatin methyl ether Dehydroschisandrol A 245 212 6,7-Dehydroschisandrol A 247 of 4' -demethy Ideoxypodophy 1 lotoxin Dehydrozaluzanin C 371,808,809 213 Dehydro- p-peltatin methyl ether 194 of deoxydehydropodophyllotoxin 213 Delphinium 868,869,871 of deoxypicropodophyllin 213 Delphinium cardiopetalum 868 of dysodanthin A 212 I l-Demethoxy-12-hydroxymyrtoidine 1053 of dysodanthin B 212 4'-Demethoxy-3',4'-methylenedioxyofEtoposide213 methylrocaglate 203 of(-)-hemone212 4'-Demethoxy-3',4'-methylenedioxy of (-)-6'-hydroxyyatein 212 rocaglaol 203 of isodeoxypodophyllotoxin 213 II -Demethoxymyrtoidine 1053 of megaphone acetate 212 4'-Demethyepipodophyllotoxin 193 of 5-methoxydehydropodophyllotoxin 4'-Demethyldeoxypodophyllotoxin 193

1313

4'-(9-Demethyldohydropodophyllotoxin 194 9-0-Demethylhomolycorine 619 4'-Demethylpodophyllotoxin 156,157, 192,210,214 Demethylyatein 157 4'-Demethylyatein 192 Dennstaedtia hirsuta 722 Dennstaedtia scabra 722 Denudatin 240 DenudatinB 233 Deoxdehydropodophyllotoxin 194 7-Deoxy-D-glycero-D-glucoheptose 3 73 Deoxygomisin A 197,256 Deoxyhemoglobin 986 Deoxypodophyllin 193 Deoxypodophyllotoxin 167,214,215,216 (-)-Deoxypodophyllotoxin 218 Deoxypodophyllotoxin (anthricin) 193 (-)-Deoxypodorhizone 192,215,216 3-Deoxysilychristin 255 Derris 487 15-Desacaetylundulatone 816 A^-Desacety Ispermostrychine 1052,1053 Desferrioxamine 992 Desipramine 1063 Desmarestia menziesii 471 Desmopsamma anchorjkata 72 Desoxyribonucleotides synthesis 924 Dess-Martin oxidation 1244 Detoxification 1012 of xenobiotics 1012 Diabrotica virgifera Dithiins-dipsaci 867,868 Diabrotica virgifera 866,871 Dianchinenoside A-H 33,34 from Dianthus chinensis 33, 34 Dianchinenoside E 37 Dianchinenoside F 37 Dianthus barbatus 32,55 Dianthus caryophyllus 32,594 Dianthus chinensis 3,32,33,34 Dianthus deltoides 32 Dianthus super bus L. var. longicalycinus 30,32,33,56

1,7-Diarylheptanoids 881 Dibalanocarpol 556,562 Dibenzlbutyrolactone enerolactone 186 Dibenzocyclooctadiene lignans 196,197 Dibenzylbutane enterodiol 186 Dibenzylbutane lignans 191 Dibenzylbutane nordihydroguairetic acid 187 2,3-Dibenzylbutane-1, 4-diol 191,234, 249,250 inhibition of Ca'Mnflux 234 4,5-Dibromopyrole-2-carboxylic acid 819 methyl-ester of 819 Dicer ocaryum zanguebricum 1158 2,6-Dichloroisonicotineic acid 946 2,4-Dichlorophenoxy acetic acid 306 Diclofop 374 Dicty^oloma incanescens 820 Dicty^oloma peruviana 820 Dicty^oloma vandellianum 820 Dictyolomide 821 Dicty>ostelium discoideum 80 1,2-Didehydroaspidospermidine 1050 re/ro-Diels-Alder reaction 111,138 Diels-Alder reaction 172 Diethyldithiocarbamate 1271 Digenea simplex 41S Digitalis 174 Digitolutein 806 Dihydroangustine 103 6,1037 Dihydroartemisinin 832,833,834 16,17-Dihydrobrachycalyxolide 811,812 Dihydrocubebin 191 Dihydrocyclotheonamide A 1194,1195,1196 5,6-Dihydroflavopereirine 1060 we^o-Dihydroguaiaretic acid 191,240268, 271 Dihydroguaiaretic acid 219,249,473 DihydropenicilHc acid 449 Dihydrophenoxazine 976 1,2-Dihydrosantonin 463 20,21 -Dihydrostrychnobrasiline 1065 Dihydrosweroside 104 8,10-Dihydrosweroside aglucone 141

1314

18,19-Dihydrousambarine 1063 5',6'-Dihydrousambarensine 1063 Dihydroxanthommatin 988,989 3',4'-Dihydroxiflavone 757 2',5'-Dihydroxy-3'-nitrophenylacetic acid 592 Dihydroxybenzoic acid 337 2-Dihydroxymintlactone 456 Dihydroyashabushiketol 883 3,8-Dihyroxy-glucoside-2-hy droxymethy lanthraquinone 657 1,4-Diisothiocyanatobenzene 481 Dimerization ofresveratrol 532 Dimers 544,545,546,547,548,549 ampelopsin D&F 544 anigopreissin A 544 betulifolA&B 545 gnetinA&C 546 gnetinF&G 546,547 leachianol G 547 malibatol A 547 parthenocissin A 548 restrytisol A-C 548,458 tricuspidatol A 549 viniferifuran 549 (+)-e-viniferin 545 l,2-Dimethoxy-4-(£-3'-methyloxiranyl) benzene 362 3,4-dimethoxybenzoic acid 1013 10,13-Dimethyl-9-tetradecen-1 -ol 66 Dimethylallyl-haempferide 802 3',4-O-Dimethylcedrusin 202,269 Dimethylcrocetin 303 (+)-Dimethylisoariciesinol-2-a-xylose 195 7,11-Dimethyloleoside 334,343 9,13-Dimethyltetradecanoic acid 65,66 10,13-Dimethyltetradecanoic acid 65-67 Diminazene (berenil) 786 Dinocophylline B&C 822,823 Dioncopeltine A 822,823 Diosmetin 746,756 Diosmin 752,755-757,759,766 Diospyrin 804 Diospyros montana (Ebeaceae) 804

Dioxapyrrolomycin 488 Dipetalolema viteae 487 Diphylleia 172 Diphyllin 226 Diphyllin acetate 194 Diphyllin acetyl apioside 263 Diphyllin apioside-5-acetate 226 Diphyllin asioside 226 Diphyllin natural products 166 Diphyllinin 194 Diphyllinin crotonate 194 Diphyllinin monoacetate 194 Diphyllin-O-apioside-5-acetate 194 Diphyllin-(9-apisode 194 Diquat 374 Dirca occidentalis 219 genkwanin from 219 (+)-lariciresinol from 219 sitoindoside II from 219 Discobahamins A&B 1202 Discodermia 1175 Discoderm ia calyx 1204 Discodermia dissoluta 1205 Discoderm ia kiiensis 1201,1204 Discodermia sp. 1201,1202 Discodermide 1206 Discodermins A-D 1201,1202 Discodermins E-H 1201,1202 Discodermolide 1205,1206,1243,12451248 Discokiolide A-C 1203 Distichol 562 D ithi ins dipsaci 437 Ditylenchus destructor 460 Ditylenchus dipsaci 493 (-)-3, 4-Divanillyltetrahydroftiran 191 Divarine 1056 L-Djenkolic acid 481 DNA damage effect 984 DNA intercalating activity 1060 DNA repair 925 DNA topoisomerase II 1060 DNA-intercalators 805 Doliocarpus dentatus 229 (-)-liriorinol B from 229

1315 (+)-medioresinol from 229 (+)-pinoresinol from 229 Domoic acid 479 Drechslera rostrata 227 Dreschlera spp. 468 Drosera capemis 1158 Dryopteris filixmas 431 Dysidea herbacea 463 Dysosma 172

a-Ecdysone 700 Ecdysteroid antagonists 555 Ecdysteroids 699,700 Ecdysterone 700 Echimidine 874,876 Echium sp. 872 Echium wildpretti 874 Ectyoplasiaferox 66 Edman degradation 1270 Effective insect growth inhibitors 858 calopeptin 858 dehydrodiisoeugenol 858 galbacin 858 galbegin 858 licarin 858 licarinB 858 sesamin 858 sesamolin 858 veragensin 858 Eflomithine 787,792 Ehrlich ascites carcinoma 303 Ehrlich's reagent 1277 Eichhornia crassipes 359 (5Z,9Z)-5,9-Eicosadienoic acid 80 (11Z, 157)' 11,15-Eicosadienoic acid 89 (1OZ, 15Z)-10,15-Eicosadienoic acid 90 9-Eicosene 361 Eicosyl ferulate 666 Eimeria 804 Elaiophylin 453 Electrophile-Responsive Element 951 Eleocharis microcarpa 361 Ellagic acid 359

Ellipticine 828,829 Emaciation 690 Emanthidine 614 Emetic effect 239 Emetine 144 from Cephaelis ipecacuanha 144 Emodin 637 Encephalartos 1157 Encephalartos friderici-guilielm i 1158 Encephalartos longifolius 1157 Encephalitizoon intestinal is 788 Endlicheria dysodantha 211 dysodanthin A&B from 212 megaphone acetate from 212 Endopolygalacturoase 1133 Endothelin antagonism 237 Endothelin-1 267 Endothelium-Derived Relaxing Factor 951 EnniatinA&B 482 Entamoeba histolyca 836,1063 Enterobacter cloacae 402,1091 Enterodiol 191 Enterofuran 191 Entero lactone 192 Enterolactone lignans 209 Ent-kaur-16a-o]-19-oic acid 811 Ent-kauran-16-en-19-oic acid 811 Enzyme-linked immuosorbent assay 1125 Ephedra spp. (Ephadraceae) 235 ephedradine B from 235 Ephedradine B 235 (-)-Epiasarinin 260 (+)-Epiaschantin 201 (+)Epiashantin 267 10-Epibengazole A 488 Epicalyxin G 902 Epicatechin-3-gallate 359 Epicatechins 748,751 Epicatechins gallates 770 Epicinnzeylanol 850,853 2'-Epicycloisobrachycoumarinone epoxide 798 Epidermophytonfloccosum 227,228 14-Epidihydrocochlioquinone B 460 (-)-Epieudesmin 201

1316

Epigallocatechins 748 Epigallocathechin-3-galIate 755 (+)-Epimagnolin 201 Epiphyas postvittana 231 (+)-Epipinoresinol 201 dimethyl ether 201,241 Epipodophyllotoxin 155,155,210 4-Epipodophylltoxin 193 Epipodorhizol 192 (+)-Epiresinol 268 13-Episclareol 397 (-)-Episesamin 201 (+)-Episesamin 260 Episesmin 246 Episteganangin 196,216 Epithelizing effect 335 (+)-Epiyangambin 201 Epiyangambin 242 Epothilione 811 EpothiloneA 810 10,13-Epoxy-l 1-methyloctadeca-lO, 12dienoic acid 69 Eragrostis curvula 469 pyrocathecol from 469 Erigeron philadelphicus 436 Eriocitrin 744 Eriodyctiol 757 Erylus formosus 80,85 Escherichia coli 330,331,333,402,556, 605,802,1092,1093,1135,1263,1269, 1271,1272, 1274, 1278 Esculetin 314,315,317,318,330-334, 336-340, 342,344 Esculin 317,318,330-342 Espintanol 797,798,807 Estradiol 222 Estrogen agonist/antagonist properties 210 of secoisolariciresinol diglycoside 210 Estrogenic activity 760 3-[a-Ethoxy(a-methyl-thio)]methyl-2thionopyrolidine 1083 Etopophosphate 160 Etoposide 160-163,171,172,212,213 use in combination therapy 171 Etoposide phosphate 160

of lymphoid cancer 171 of myeloid leukemia 171 of refractory testicular cancer 171 of stomach, of ovarian, of brain, of breast, of pancreatic, small and non-small lung cancer 171 Eucalyptus grandis 445 Eucalyptus haemastoma 696 Eucommia ulmoides 235 (±)-pinoresinol diglucoside from 235 (-)-Eudesmin 200 Eudesmin 268 Euglena gracilis 363 Eumelanins 1018 Eunicea Sliceinea 83 Euphorabia characias 1263,1277 Eupomatenoid-5 202 Eupomatenoid-6 232 Europine 874 Eurotium chevalieri 1091,1093,1094, 1097-1099 Evans chiral oxazolidine 1226 Evodia rutaecarpa 487 Exiguaflavanone 801,803 Exoerythrocytic cycle 781 Fagraea racemosa 268 (+)-epiresinol from 268 (+)-isolariciresinol from 268 (+)-pinresinol from 268 (+)-Fargesin 201,267 Fargesin 233 Fargesone A-C 233 Farnesol 331,332 Fenton rection 975,995 Fenugreek (Trigonellafoenum-graecum) 1139,1141,1268 Ferrihemoglobin 985,986,991 mediated oxidation 991 Ficin flavonoids 744 Filospadin flavonoids 744

1317 Fischer ella ambigua 364 Fischerella JAVA 94/20 366 Fischerella muscicola 366,364 F/5c/zere//aNEP95/l 366 Fischerellin 370,374 Fischerellin A 364,368 Fisetin 757,758 Flavin adenine dinucleotide 917,918,951 Flavin mononucleotide 917,918,951 Flavonol quercetin 753 Flavonols 744 Flavonone 743,744 Flexuosol A 569 Fluorenone 655 Fluorescence spectrometry 378 FMN-binding domain 920 Foeniculoside I-IV 556,560 Formononetin 745 l-0-Formyl-4'-Demethoxy-3',4'-methylenedioxymethylrocaglate 203 A^'-Formyl-kynurenine 966,998 8-Formyloct-5-enoate 86 Forsythia 327 Forsythia intermedia 156,174,177, 223 (-)-arctigeninfrom223 Forsythia spp. 156,157,235 pinoresinol monoglucoside from 235 Forsythia suspensa 156,241 Framoside 327,334,343 Fraxetin 314,315,318,331-334,336-340, 342 Fraxidin 315 Fraxin 314,317,318,331-334,336,338-340, 342 Fraxinol 315 Fraxinus 327,333,343 Fraxinus bungeana 317 Fraxinus chinensis 317 Fraxinus formosana 327 Fraxinus insular is 321 insularoside from 321 Fraxinus ornus 313-319,327,328,330, 332,334-340,341,343-345 Fraxinus stylosa 317 cichorin from 314

esculetin from 314 fraxetinfrom314 fraxin from 314 hydroxycoumarin esculin from 314 quercetin from 314 quercetin-3,7-digalactoside from 314 quercetin-3-galactoside from 314 quercetin-3-glucoside from 314 rhamnetin from 314 rutin from 314 Fraxinus uhdei 321 uhoside from 321 Fructus schisandraei 233 Fucus garderi 1152 Fumagillin 458,459 Fumagillol 459 Fumarate dehydrogenase 800 Fungal polysaccharides 1124-1129 Furano flavonoids 744 Furoftiran lignans 200,201 Furanoheliangolides 808 Fusarium solani 404,405 Fusarium spp. 464,482 Futoenone 245 Futoquinol 245 (+)-Galeon 899 Galactomannan 1148 Galacturonorhamnans 1156 Galangin 752 Galanthamine 614 Galanthine 614 Galinum sinaicum 220 bis-glucoside derivatives from 220 Galiosin 657 Galium 631 Gallic acid 359,798 Gallic-/?-coumaric acid 359 Gallic-sinapic acid 359 Gallocatechins 748 Ganoderma lucidum 449 Garbanzol 745 Garcinia cowa 805

1318

Garuga sp. 899 Garuga pinnata 899 Garuganinlll 899 Geduin 815 Geissoschyzine 97 General skeleton 190 oflignans 190 Genistein 750,751,754,757,760,802,803 Genkwanin 219 Genotoxicity assays 675 Geodia 488,1210 Geosmin 353,354 Geotrichum candidum 673 Geotricum louberi 673 GermacreneD 409,413 Germanocrenolide 808 Ghosez cyclization 1230,1231 Giardia intestinalis 1063 Gilbert's reagent 1245 Gingerol 475 Ginko biloba L. 929 GL-311162 used for anticancer treatment 162 Glaucarubin 837 Gleditsia macracantha 1140 Gleditsia thacanthos 1115,1116,1139, 1143 Globodera 429,432,467,471 Globodera pallida 229,474 Globodea rostochiensis 229,430,461, 474,477,486,487 Globodera spp. 428 Glossina spp. 785 Glucantime 828 Glucosinolates 1074 enzymatic breakdown of 1074 Glucotropolin 481 Glucuronomannan 1156 Glunapin 481 Glutamate synthetase (GS) inhibitor 372 y-Glutamylcysteine synthetase 913,951 Glutathinonylspermidine 808 Glutathion reductase inhibition 1036 Glutathione reductase 796 Glycanoxylans 1156

Glyceollin 473 Glyceollin Mil 473 Glyceraldehyde-3-phosphate dehydrogenase 913,951 Glycine max L. 930 Glycinoeclepin A 466 Glycofomis 1131 Glycolysis (GADPH inhibition) 924 Glycoproteins 1129-1137 Glycosidation-induced shift method 8 Glycosphingolipids 1123 Glycyrrhiza glabra 800 GnetinA-C 535,565 GnetinE-I 565 Gnetum parvifolium 530 Gomisin A-C 197,247,256-258,263,265 Gomisin G 223 Gomisin J 197,224,247,263 Gomisin K 197,247 Gomisin L 197,247 Gomisin M 197,240 Gomisin N 197,247,249 Gossypium hirsutum 463 GraminoneB 233,235 Grandisin 198,230,799 Granulocyte phagocytosis 47 of jenisseensosides C&D 47, 55 Gvdiss \iQdL {Lathyrus sativus) 1268 Grevillea robusta 1157 Guaiacylglycerol-p-(9-6'-(2-methoxy) cinnamyl alcohol 211 Guar (Cyamopsis tetragonolobus) 1139 Guayacasin 249 Guiaflavine 1055,1056 Guianensine 1055,1056 Guppy fish (Lebistes retculatus) 689 Gypsogenic acid 29,36 Gypsophila arrostii 38,39,40 Gypsophila bermejoi 44,52 Gypsophila capillar is 42,43 Gypsophila capitata 54 Gypsophila fastigiata 3 8 Gypsophila oldhamiana 41 Gypsophila paniculata 38-40,42,53 Gypsophila paniculata/muralis (baby's

1319 breath) 3 Gypsophila perfoliata 38 Gypsophila saponins 54 Gypsophiloides 3 Gyrocarpine 824,826 >H and ^^C spectra 8,9,22,30,33 of aglycone of dianchinenoside G & H 36 of bisdesmoside41 of dianchinenoside E 33,34 of dianchinenoside F 37 of dianchinenoside G and H 35 ofSAPO30 41 ofSAPO50 41 of saponarioside A 22 ofvaccaroside F30 HA 1002-1004 as a carcinogen 1002-1004 Haber-Weiss reaction 975 (+)-Haedoxan A&B 231,232 (-)-Haedoxan D 231 (+)-Haedoxan D 231 Haemonchus 425 Haemonchus contortus 429,431,432,440, 446, 451,453,457,470,482,484,488,490, 491 Hakea 1157 Haminaea templadoi 90 Hamycin 794,795 Hannoa quassia undulata 466 Hansenula polymorpha 1263,1267,1270, 1272,1274 Hansenula sp. 1087 Hapalindole A 363 Hapalosiphon intricatus 363 Hapalosiphorn fontinalis 363 Haplophyllum hispanicum 263 diphyllin acetyl apioside from 263 tuberculatin from 263 Harmaline 828,829 Harman 828,829 Hattalin 216,269 ouabain receptor 216

Hedera helix 813 Hederagin 814 a-Hederin 814 P-Hederin 814 Helenium 436 Helicoverpa zea 697 Helicteres isora 226 helisorin from 226 helisterculin A&B from 226 Heliothis armigera 850,851 Heliothis virescens 439,485 Heliotropium bovei 874 Heliotr opium floridum 874 Heliotropium megalantum 874 Heliotropium sinuatum 874 Heliotropium sp. 872 Helioxanthin 194,232 Helisorin 226 Helisterculin A&B 226 Helmindiol 451 Helm inthosporium gram ieu 1210 Helycobacter pylori 693 Hemangiomas 693 Hematopoietic cancer cells apoptosis 925 Hemicenttrotus pulcherrimus 1188 Hemoglobin / HA interactions 985-987 Hemolytic activity 41, 42 of Gypsophila paniculata 41 ofSAPO30 41 ofSAPO50 41 (Z)-4-Heneicosenal 81 Hepatic activity 253 Hepatocarcinogenesis 257 Hepatoprotective acitivity 255,257 Hepatospienomegaly 785 Heptamethoxyflavone 752 Heptanoic acid 361 15P-Heptylachaparrinone 815 Herbivorous crustacean zooplankton 355 Hernandia nymphaeifolia 218 (-)-deoxypodophyllotoxin from 218 epiaschantin from 218 epimagnolin from 218 epiyangambin from 218 (-)-yatein from 218

1320

Hernandia nymphaeifolia (Hemandiaceae) 212 (-)-hernone from 212 (-)-6'-hydroxyyatein from 212 (-)-nymphone from 212 Herniaria fontanesii 52,54 Herniaria glabra 50,51 Herniaria hirsuta treatment in arrthritis 50 in infections of urinary & respiratory tracts 50 in kidney disorders 50 in neural catarrh 50 in neuritis 50 in purifying the blood 50 in rheumatism 50 Herniaria saponins A-D 50-52 (-)-Hemone 199 Herpes simplex virus (HSV) 223 Herpetomonas samuelpessoai 1138 Hesperetin 757,762 Hesperidin 744,745,752,759,766,768 HETCOR 11,13,34,37 of bisdesmoside41 of dianchinenoside E&F 34 ofSAPO 30,41,50 Heterakis spumosa 482 Heteroclitin F&G 259 Heterodera 429,432 Heterodera cajani 469 Heterodera glycines 429,454,456,466, 467, 473,477 Heterodera rostochiensis 437 Heterodera schachtii 481 Heterorhabditis UQmditodQS 471 re^ro-Hetero-Diels Alder reaction 113 HeyneanolA 568 Hikizimycin 493 (-)-Hinokinin 192,215,216 Hinokinin 216,231 Hippamine 610,614 Hippeastrine 614 Hippeastrum sp. 614 His2i5 family of inhibitors 368

Histamine 759 Histamine-induced hind paw edema 401 Histiopteris incisa 722 HIV-1 RT inhibition by anolignan A&B 224 HK-mediated neurotoxicity 992 HMBC correlations 9,13,15,21,25,29,34 of bisdesmoside 41 of dianchinenoside E 34 ofSAPO 30,41,50 of saponarioside C 15,25 of saponaside A 21 of vaccaroside D 29 HMQC/HSQC experiment 11,34,37 of bisdesmoside 41 of dianchinenoside E&F 34 ofSAPO 30,41,50 'H and '^C NMR data of 321,324,326,329 of hydroxyframoside A & B 326 of hydroxyomoside 321 of insularoside 321 of isolugrandoside 329 of isolugrandoside acetate 329 iR NMR data 538,539,540,541,542,543 of ampelopsin A 538 of ampelopsin D 538 of ampelopsin F 538 of anigopreissin A 538 of balanocarpol 539 ofbetulifolA539 of betulifol B 539 of cyphostemmin A 539 of cyphostemmin B 540 of gnetin A 540 ofgnetinBV 541 of gnetin C 541 of gnetin F 541 of gnetin G 541 ofleachinoIG 542 of malibatol A 542 ofparthenocissin A 542 of restrytisol A 541 of restrytisol B 543 of restrytisol C 543 of tricuspidatol A 543

1321

of viniferifuran 543 of (+)-^-viniferin 540 of e-viniferin diol 540 HOHAHA spectrum 10,21,29,34 of bisdesmoside 41 of dianchinenosideE 34 of saponariosides A&C 10,21 ofvaccarosideD 29 Holarrhena sp. 837 Holothuria mexicana 72 Holstiine 1048 Homidium 786 Homolycorine 618 (-)-Hopeaphenol 561 (+)-Hopeaphenol 561 Honokiol 217,227,240,249,254 Hopeaphenol 561 Hormosira banksii 440 Homer-Emmons reaction 1221,123 6 Horseradish peroxidase 1133 HRFAB-MS 20,33 of bisdesmoside 41 of dianchinenosideE 33 ofSAPO 30,50,41 of saponarioside A 20 Hugonia 172 Human immunodeficiency virus (HIV) 223 Human papilloma virus (HPV) 223 (-)P-Hydrastine 868 (+)|3-Hydrastine 868 Hydrocholeretic effect 391,399,415 Hydrogenation 66 of 10,13-dimethyl-9-tetradecen-l-oI 66 Hydrolapachol(2-hydroxy-1,4-naphtho quinone) 803 5-Hydroperoxy-eicosatetraenoic acid 265 Hydrothermolysis 40 3-(Hydroxy)methylene-2-thioxopyrrolidine (HMTP) 1073,1077-1079,1083,1084, 1087,1092,1093,1095,1096,1099,1100, 1102,1103 2-Hydroxy-1,4-naphthoquione 804 12-Hydroxy-11 -methoxy-A'^-acetyl-nor-Cfluorocurarimine 1052,1053 12-Hydroxy-11-methoxy-nor-C-

flourocurarine 1052 8-Hydroxy-2'-deoxyguanosine 253 l-Hydroxy-2-hydroxymethylanthraquinone 660 1 -Hydroxy-2-methyl-anthraquinone 660 7-Hydroxy-2-methyl-anthrquinone 676 Hydroxy-2-methylantraquinone 676 2-Hydroxy-3-methoxy strychnine 1030,1052 3-Hydroxyanthranilic acid 965,966,978, 1015 5-Hydroxyanthranilic acid 971 2-Hydroxyanthraquinone 659 5-Hy droxy-benzoquinone-1 -(2' -hydroxy6'-carboxyanil)-4-imide 1003 /?-Hydroxybenzyl isothiocyanate 1076 6-Hydroxycalyxin F 902 Hydroxycoumarin esculin 314 Hydroxycoumarins 315-318 from Fraxinus spp. 315-318 Hydroxyeicosatrienoic acids 759 Hydroxyethylrutosides 756 3-Hydroxyflavone 757,761-764, 768,769 Hydroxyframoside A&B 325,326,333 3-Hydroxykynurenine 965-967,988,997, 999 10-Hydroxyligstroside 334,343 12-Hydroxymalagashanine 1053 2-Hydroxymethlanthraquinone-3-0-^Dglucoside 660 2-Hydroxymethyl-8-hydroxanthraquinone3-0-/?-D-glucoside 660 2-Hydroxymethylanthraquinone 3glucoside 657 2-Hydroxymethylquinizarin 661 12-Hydroxy-myricanone 895 Hydroxy naphthoquinone atovaquone 783 O-Hydroxynitropapuline 593 Hydroxynomuciferine 825 Hydroxyomoside 318,321,322,325,333, 343 0-Hydroxypapuline 592,593 1-Hydroxyphenazine 365 p-Hydroxyphenyl-I-bromoheptadecane 73 (+)-l-Hydroxypinoresinol 200,270 8-Hydroxyquinaldic acid 991

1322

7-Hydroxytectoquinone 661 Hydroxyverniladin 808 11-Hydroxyvittatine 610 (-)-6'-Hydroxyyatein 192 Hymenolepis nana 484 Hyophyllanthin 233 Hypericin 375 Hypericum 371 Hypericum calycinum 797 phloroglucinol from 797 Hyperoside 666 Hypocholesterolemic agent 246 Hypocholesterolemic atherosclerosis 248, 249 Hypocholesterolemic effect 44,54,205 of Gypsophila capitata 44 Hypocholesterolic effects 233 Hypoglycemic 398 Hypoglycin A 476 Hypolepis puncatata 722 Hypolipidemic action 887 Hypolipidemic activity 247 Hypoloside A&B 722 Hypophyllanthin 195,237 Hypoplasia of bone marrow 688 Hypotriglyceridemic effect 246,261 Hypselodory nana 463 Hyp t is verticillata 111 dehydro-p-peltatin methyl ether from 212 5-methoxypodophyllotoxin from 212 Hyriopsis schlegelii 1123 Hyssopus sp. 746 Ibogan 139 Ichthytoxic activity 485 Ictalurus punctatus 353 Idicol 850 Illudane-dienone 716 Illudanes 705 Illudin-M 706

Immunofluorescence detection 935 of the putative NOS 935 Immunofluorescence microscopy 932 Immunohistopathology 792 Immunomodulatory actions 1036 Immunomodulatory effect 50,55 of Acanthophyllum squarrosum 55 of Gypsophila saponins 55 of Vaccaria segetalis 50 Immunopathological reactions 787 Immunosuppressive activity 214 of p-apopicoropodophyllin 214 Immunosuppressive activity 482 Immunosuppressive effects 486 Immunstimulatory activity 797 Imperata cylindrica (Gramineae) 233 graminone Bfrom233 Indanones 700,701,702 by l-acetoxy-(+)-pinoresinol 238 by 1 -acetoxy-(+)-pinoresinol-p-Dglucoside 238 by Anemarrhena asphodeloides (Anthericaceacy) 237 by arctigenin 237 by didemethoxymatairesinol 208 by c/5-hinokiresinol 237 by Forsythia koreana 237 by Forsythia viridissima 237 by Forsythia suspensa 237 by (-)-mataresinol 237 by (+)-pinoresinol 237 by (+)-pinoresinol monomethylether 238 by (+)-pinoresinol-di-j3-D-glucoside 238 by (+)-pinoresinol-p-D-glucoside 237 by podophyllotoxin derivatives 159 by (+)-syringaresinol-di-P-D-glucoside 238 of acetylcholinesterase 472 ofaconitase 912 of cAMP-phosphodiesterase (PDE) 237-240 of cholesterol acyltransferase 247 of cytochrome C oxidase 912

1323

of cytochrome P450s 912 of DNA-topoisomerase II 160-163,166, 167,169,211 of human estrogen sythetase 208 by 3-demethoxy-3-(9-demethymatairesinol 208 of indoleamine-2,3-dioxygenase 912 ofmicrosomalacyl-coenzyme A 247 of photosynthesis 554 of the kinase activity of growth factor 924 of the proto-oncogene 924 of type I DNA- topoisomerase 804,806 of 5-lipoxygenase 887 of cyclooxygenase-1 and - 2 887 Inhibitory activity blastogensis 217 honokiol 217 machilin A 217 matairesinol 217 of mitogen-induced arctigenin 217 Innzeylanine 850 Insecticidal activity 231,232, 426,471, 479,485 Insecticidal effects 852,1136 of ryanodol/isoryanodol-type diterpenes 852 Insularoside 321-323,325,327,333,343 Insulin 222 Inter alia 481 ofentrodiol 208 Interferon-y 951 Interiotherin A 197,223,224 Interiotherin B 197 Interleukin 951 Intramolecular Aldol reaction 111 Intramolecular Claisen-Tishchenko reaction 102 3p-H-Ipecoside 97 Ipomoea cairica 223 Iris japonica 433 Iron regulatory Protein 940 Iron responsive Element 951 Irpex lacteus 439,468 Is at is tinctoria 632

Isoamericanol A 270 Isoampelopsin F 568 Isobrucine 1051,1052 Isobrucine-N-oxide 1051,1052 Isodeoxypodophyllotoxin 193 Isodomoic acid G&H 479 (+)-Isogmelinol 200,241 Isoguattouredigine 825,827 (-)-Isohopeaphenol 561 3a-H-Isoipecoside 97 /raw^-Isokielcorin B 259 Isolariciresinol 195,205 (+)-Isolariciresinol 268 Isolugrandoside 328,329 acetate of 329 Isomagnolin 201,241 Isomagnolol 227 Isonaringin 744 Isophorone 297,298 Isopicropodophyllone 193 Isopicrostegane 196 Isopregnomisin 249 Isoptaquiloside 719,722 Isosakuranetin 744 Isoscopoletin 315,332,333,334,342 Isostrychinine 1051,1052 Isostrychnine-N-oxide 1051 Isoswinholide A 1197,1198 Isothiocyanato-N-(4-nitrophenyl)benzeneamine (amoscanate) 481 Isovelleral 461 Isoxaben 370 Isoxazopodophyllic acid 214,215 Isoyangambin 242 Ivermectin 441,447,453,457,468 Ixodes ricinus 695 Jaspamide 485 J asp is 1210 Jaspis spp. 485,488 Jasplakinolide 485 Jatrogrossidione 809,811 Jatropha grossidentata 809,810 Jatropha isabelli 810

1324

Jatrophone 809 Jeitacin A&B 442 Jenisseensosides A-D 46,47,55 from Silenejenisseensis 46 Jereicopsis 1175 Jereicopsis graphidiophora 1206,1207, 1214 JereisterolB 1206,1207 Jones' oxidation 1209,1225 Juglans 905 Juglans mandshurica 899 Juglone 363,370,470,638,868 Julia olefmation 1219 Juniperus sabina 214 Justicia procmbens var. leucantha 226 diphyllin apioside-5-acetate from 226 diphyllin asioside from 226 diphyllin from 226 justicidin A&B from 226 Justiciaprocumbens 235,226,243,265 justicidin B from 243 justicidin E from 243 neojusticin A from 243 taiwanin E from 243 taiwanin E methyl ether from 243 diphyllin apioside-5-acetate from 226 diphyllin asioside from 226 diphyllin from 226 justicidin A&B from 226 Justicidin A 194,266 Justicidin B 243 Justicidin D (neojusticin A) 194 Justicidin E 243,265 Kadsura heterocollita 240,249,259 burchellin from 249 gomisin M from 240 heteroclitin F from 259 heteroclitin G from 259 kadsurenone from 249 Kadsura interior lll^ gomisin-G from 223 interiotherin A from 223 schisantherin D from 223

Kadsura japonica 255 Kadsuranin 197,224 KadsureninH 240 Kadsurenone 240,241,249 Kaempferol 747,748,750,763,764,769 (-)-a-Kainic acid 478 Kalihinane diterpenes 810 cycloamphilectanes 810 7,20-diisocanoisocycloamphilectane 810 epiamphilectadiene 810 7-isocyano-ll(20),14 7-isocyanocyclo-amphilect-10-ene 810 isocycloamphilectanes 810 KalihinolA 466,810,812 Kappa-carrageenan 1148 Kazmaier's modification 1227 of Ireland-Claisen rearrangement 1227 Keck DCC protocol 1230 Keramamide K 1183 KeramamideL 1183 Keramamides B-J 1181-1183 a-Ketogluturate 638 /m^5-Kielcorin B 259 Kinetin 640 Kishi's retrosynthesis ofonnamide 1231 Klebsiella 1155 Klebsiella aerogenes 1263,1269 Klebsiella pneumoniae 402 Kluyveromyces fragilis 338 Kluyveromyces lactis 1127 Klyne's rule of molecular rotation 8 Kobophenol A&B 535,558,560,561 Kobusin 200,231 Kombamide 1184 Konjac glucomannan 1149 Korundamide A 822,823 Koshikamide A, 1193 Krameria triandra 204 scavenging activity of 204 a-Kratin 667 Kreb's cycle 924,940

1325

Kumusine 1201 Kynurenic acid 965-968,970 Kynurenine 966,968-971,990,998 Lachnella 462 Lachnum papyraceum 446 Lachnumol A 446,447 Lactarius mitissimus 461 Lactuca sativa 187 Lamellomorpha strongylata 1189 Laminaran 1121 from brown algae 1121 Laminar ia 1121 Laminariales 478 Laminine 478 Lamium 392 Lampteromyces japonicus 106 Lapachol methylether 66 Lapdap 783 Lapinone 803 (+)-Lariciresinol 177,188,205,219 Lariciresinol 199 75',8/?,8'7?-(-)-Lariciresmol-4,4'-bis-0-pD-glucopyranoside 199 Lariciresmol-4p-D-glucoside 199 Larix leptolepis 185 Larrea tridentata 187,205,249 Laspeyresia pomonella 231 Lauraceae 849,850 Laurencia implicata 808,809 Laurilia tsugicola 703 Laurus azorica 850 Lawsone 638 Lawsonia alba 837 Leachianol A-G 535,555,563,564 NO-Leghemoglobin complex 937 Leguminosae 637,871,871,931 Leishmania 779,780,783,786,789,790, 793,801,803,805,807,808,810,821,822, 824,828,830 Leishmania amazonensis 229,887 amastigotes of 828 promastigotes of 828 Leishmania decemlineata 851 -859,860

863-866 Leishmania donovani 791,794,797,800, 804,805,806,819,821,828 Leishmania infantum 810 Leishmania major 230,805,784,791,800, 810,811 Leishmania panamensis 821,822 Leishmania tropica 802 Leishmaniasis 781,783,784,785,793 Leishmanicidal activity 793 Lens ciilinaris 1263 Leprae 1155 Leptinotarsa decemlineata 850,854,855, 869-872,874,876 Lethaloxin 451 Leucaena leucocephala 1142,1148 galactomannans from 1148 Leucaena spp. 1139 Leucoanthocyanidins from 747 Leucopaxillus albissimus 493 Leukocyte activity of 510 Leukocytopenia 243 Levamisole 440,454,485 Libocedrus bidwilli 231 Licarin A 247 Licarin B 858 Licarin C 202 Licarin D 247 Licochalcone A 800,801 Licorice 745 Ligistrum 327 Lignans 206,207 as phytoestrogens 206,207 Lignin 1262 Ligstroside 318,333,334,343 Limnocorrals 381 D-seco-Limonoids 815 Linalool 396,397 epoxide of 397 Linoleicacid 361,433,434,461 6-Linolenic acid 361 Linum 172,173 Linum album 173 Linum flavum 173,212 5-methoxypodophyIlotoxin from 212

1326 Linum usitatissimum (Linaceae) 206 Lipid peroxidation activity 253 Lipid-diphosphooligo-saccharide 1131 Lipomannan 1155 Lipopolysaccharide 597,951,1152, Liquiritigenin 745 Liriodendrin 233 Lirioresinol glycoside 268 p-dimethyl ether of 233,268 (-)-Liriorinol B 229 Lister a ovata 1135 Litalin flavonoids 744 Lithistid sponges 1175,1176 Litholytic properties 52,54 of Herniaria fontanesii 52, 54 3,4-Lobatin 808 Lobatin A&B 808,809 Locust bean (Ceratonis siliqud) 1139 A^-Lodosuccinimide 1162 Lolium perenne 696 Lonchocarpus 487 Longicaudatine 1051 Lossen rearrangement 1074 Lucidin 640,657,661,674,675,670 Lucidin glucoside 661 Lucidin primeveroside 642,662,676 Lucidin co-ethylether 661,676 Lucidin co-methyl ether 662,676 Lugrandoside 328 Lupinus Albus 931 Lupus nephritis 269 Luteolin 746,750,757,762,763,769 glycoside of 757 Lutessine 610,612 Lutzomyia 783 Lycoctonine 868 Lycopene 299,300 Lycopsamine 874 Lycorine 581,609-614,617,618 Lycoris radiata 614 Lymphocytic leukemia 925 Lymphokine-activated killer 923,951 Lyngbya majuscula 78 Lyngbya wollei 355 (+)-Lyoniresinol3-a-0-[3-D-glucopyra-

noside 195 Lysine malonate 366 Lysy 1-tyrosy 1 quinone (LTQ)

1261,1268

Macaronesia fortunata 850 Mechilin racemosus 1093 MechilinA 191,217,268 Mechilin G 242,247 Machilus japonica 857 bioactive lignans from 857,858 Machilus thiimbergii 247 Macrocyclic biarylheptanoids (Type III) 894-898 Macrocyclic diaryl ether heptanoids (Type IV) 899-901,905 Macrocyclic trichothecenes 811 Macrocystis alterniflorum 359 Macrocystis spicatum 359 Macrocystis verticilatum 359 Macronesiafortunata 850,851 Macrophage-mediated cytotoxicity 910 Macrozamia communis 3 66 UdiddigdiS^cm Strychnos 1030-1032,1035 Maduramicin 451 Magnolia denudata 247 Magnolia fargesa 233 denudatin B from 233 fargesin from 233 fargesone A-C from 233 lirioresinol-B dimethyl ether from 233 magnolin from 233 pinoresinol dimethyl ether from 233 Magnolia far ges a 245 magnone A&B from 245 schisandrol A from 245 Magnolia far ges a 268 Magnolia obovata 226,227,254 Magnolia officinalis 226 honokiol from 249 magnolol from 249 Magnolia salicifolia 233 Magnolia spp. 240,270 honokiol 240,270 magnolol 240,270

1327

Magnolia virginiana 203,232 4,4'-diallyl-2,3'-dihydroxybiphenyl ether from 232 magnolo from 232 4-methoxyhonokiol from 232 Magnoliae cortex 204 honokiol from 204 magnolol from 204 (+)-Magnolin 200 Magnolin 233,268 Magnolol 227,232,240,249,254 MagnoneA&B 199,245 Magnosalicin 264 Magnosalin 262,263 Magnoshinin 262,263 Major histocompatibility complex 921,951 Malagashanine 1053,1054,1063,1064 Malagashanol 1053,1064,1065 Malate dehydrogenase 800 MALDI mass spectroscopy 11 Malekulatine 824,825,827 MalibatolA&B 562 Manducasexta 439 Mannich reaction 125 Manno-oligomers 1126 Manool 402 Marasmic acid 462 Marasmius 462 MarcfortineA 490,491 MdiVigoXd {Tagetes) 430 Marine fatty acids 63 Matadine 1060 (-)-Matairesinol 156, 157, 192 Matairesinol 177,205,206,217,229,233 (-)-Matairesinol dimethyl ether 192 (-)-Matairesinol-p-D-glucoside 238 Matteson method 1221 for dihalomethylithiation insertion 1221 Matthiola fruticulosa 1075 (-)-Maximowicziol A 899 Mayapple 149,152,153,171,175,176 Measles virus 223 Medicago sativa 696 Medicarpin 473

(+)-Medioresinol 200,229 Mefloquine 782,783,838,839 (+)-Megacerotonic acid 185 Meglumine antimonate (glucantime) 822 Melaconis thelebola 596 Melanin 1019 biosynthesis of 1019 Melanins 1018-1020 Melanogenesis 965 Melanoproteins 1019 Melarsoprol 791,787 Melatonin 965 Melia azedarach 187 Melia cultivar 466 Melinonine F 1060 Mellein 449 Meloidogyne arenaria 456 Meloidogyne incognita 434,435,437,439, 447-50,460,462,463,466,467,469,471, 473,476,483-493,1289 Meloidogyne javanica 429,430,456,464, 466,473,478,482 Meloidogyne spp. 428,430,432,437,481, 487 Menoctone 803 Menoquinones 638 Mentha sp. 746 4-Mepacrine 836 Metastatic secondary tumor 221 in animals 221 Methanolysis 65 of (Z)-9-hexadecenoic acid 70 of 10-hydroxydecanoic acid 65 5-Methdehyropodophyllotoxin 194 Methoxatin 1266 5-Methoxy podophyllotoxin-4-|3-Dglucoside 212 11 -Methoxy-12-hydroxyspermostrychnine 1064-1066 1 -Methoxy-2-methylanthraquinone 662 2-Methoxyanthraquinone 662 7-Methoxyaromadendrin 802 4-Methoxyhonokiol 232 75',8/?,8'/?-(-)-5-Methoxylariciresinol-4,4'bis-O-p-D-glu 199

1328

5-Methoxypodophyllotoxin 193 3 -Methoxy spermostry chnine 1064-1066 Methyl (9-methoxycarbonylnonyl) triphenyl phosphonium bromide 65 Methyl 11-methylpentadecanoic acid 64,65 Methyl 4-(methylthio) butyldithiocarbamate (MBDC) 1081,1082 (Z)-Methyl4-(methylthio)-3-butenyl dithio-carbamate 1081 Methyl 9-deoxy-9-oxo-a-apopicoropo dophyllate 195 TV-Methyl cytisine 487 2-Methylanthraquinone 363,374 Methylated fatty acids 63,64 (-)-De-O-Methyl-centrolobin 902 (-)-N-Methylcytisine 487 (Z)-9-Methyldec-4-enal 83 3-Methyldodec-l-en-3-ol 72 19-Methyleicos-4-en-l-ol 84 9-Methyleicos-4-enal 84 2-(3,4-Methylenedioxyphenyl)-quinoline 820,821 7-Methylesculin 317,333,334,342 7-0-Methylesculin 332-334,342 O'-Methylguanine-DNA-methyl-transferase 913 A^b-Methylharmalane 1060 (4Z,9Z)-24-Methylhexacosa-5,9-dienoic acid 83,84 2-Methylisobomeol 353,365-367 Methyllycaconitine 871 3-Methyloxacylotridecan-2-one 69 4-O-Methylpicropodophyllotoxin 193 Methylpiperbetol 243,244 (-)-Methylpluviatolide 799 Methylpsedolycorine 615 Methylrocaglate 203 (£)-5-Methyltetradec-4-enal 72 4-(Methylthio)-(£,Z)-3-butenyl glucosinolate 1085 4-(Methylthio)-(£,Z)-3-butenyl isothiocyanate (MTBI) 1073,10751079,1082-1084,1090,1092,1093,10981100,1102,1103,1105 3-[4-(Methylthio)-3-buenyl]-5-benzyl-2-

thiohydantion 1089 3-[4-(Methylthio)-3-butenyl]-5-[2-(methyl -thio)-ethyl]-2-thiohydantoin 1089 3-[4-(Methylthio)-3-butenyl]-5-isobutyl-2thiohydantoin 1089 3-(Methylthio)methylene-2-thioxopyrrolidine (MMTP) 1073,1081-1084,1092, 1093,1095,1096,1099,1100,1102,1103 (/?)-17-methyltricos-4-en-l-ol 87 Methylumbeliferyl-phosphate 378 3-[a-Methyoxy(a-methylthio)]methyl-2thioxopyrrolidine(raphantin) 1083 Methyoxy-2-phenylquinoline 820 2,3-Bis(Methyoxycarbonyl)-l-(3,4dimethoxy phenyl)-4-hydroxy-6,7,8 trimethoxynaphthalene 196 Michael addition 1237 re/ro-Michael reaction 110 Miconidin 797,798 Microbacterium lacticum 1122 Microbial transformation ofisophorone 307 of xanthurenic acid 991 of 8-hydroxyquinaldic acid 991 Microbiostatic inhibitory activity 402 Micrococcus lysodeikticus 1122,1124 Micrococcus spp. 1155 Microcystis aeruginosa 359,365,366,367 Microcystis sp. 352,354,258 Microcystis viridis 366 Micromelum tephrocarpum 800 Microschlerodermin A 1215,1216 Microscleroderma sp. 1175,1215,1217, 1218 Microsclerodermins C-E 1216,1217 Microsclerodermin B 1215,1216 Microspectrofluorometry 469 Microsporum canis 227,228 Microsporum gypseum 228 MicrOSproum sp. 1127 Mikania obtusata 810 Milbemycins 425,426,432,453 Mimosa pudica 481 Misakinolide A 1197,1198 Mithramycin 220

1329

Mitochondrial monoamine oxidase 1260 Mitogen-activated protein kinase 753,951 Mitsunobu reaction 1236 MiyabenolA-C 552,556,558,560,561 c/5-Miyabenol C 556,560 possessing o-quinone 164 possessing spiroketal 164 Molinema dessetae 441 Mollugin 666 N(G)-Monomethyl-L-arginine 951 Monomycin 831 Monotropein 666 Morelensin 193,215,216 Morinda lucida 673,805 Mortierella alpina 259 Motuporin 1189,1217-1220, 1222-1227 Mozamides A 1184 Mozamides B 1184 Mucoproteins 1019 Mucuna aterrima 435 Mucuna hassjoo 932 Mukaiyama-type aldol reaction 1235,1237 Munjistin 657,663 methylester of 663 Munnozia maronii 808 Murine cytomegalovirus (MCMV) 223 Murine lymphomas L 1200,1210 Murisolin 441,442 Musa acuminata 474 Musacin C 444 Musca domestica 231 Mutagenic activity 675 of Rubia tinctorum 675 Muzanzagenin 813,814 Myasthenia gravis 486 Mycalamides A & B 1231,1234 Mycelia sterilia 482 Mycobacterium arium 1160 Mycobacterium fallax 69 Mycobacterium luteus 1122 Mycobacterium smegmatis 227,1123,1155 Mycobacterium sp. 1122 Mycobacterium tuberculosis 1122,1155 Mycobacterium malmoense 1122

Mycobacterium tuberculosis var. hominis 673 Mycosphaerella lethalis 451 MycotoxinMT81 805,806 Myocardial ischemia 233 Mvo-inositol 1155 Myrica 905 Myrica gale var. tomentosa 895,899 Myrica rubra 895 Myricanone 895 Myricatomentoside I&II 899 Myricetin 742,746,747,750,757,758, 763,770 Myriophyllum 359 Myriophyllum spicatum 378 Myristica fragrans 226,85 8 dihydrodiisoeugenol from 226 5 '-methoxydehydrodiisoeugenol from 226 Myrothecium hapla 464,414 Myrtoidine 1053,1064,1065 Myzus persicae 850,851,859,861,862 Najas sp. 361 1,2-Naphthoquinone 470 4-Naphthoquinone 470 Naphtoquinones 638 Napththylisoquinoline alkaloids 822 Narciclasine 610,612,613,616 Narciclasine-4-O-D-glucopyranoside 610, 613 Narcissus pseudonarcissus 608 Narcissus tazetta 610,618 Naringenin 745,759,762,768 Natriuretic effect 269 Natsudain 752 Natural algicides in freshwater ecosystems 380,381 Natural phytotoxins 369 Nauclea latifolia 1036 Navelbine 1047,1048,1049 Naviculla pelliculosa 358 Necator americanus 425,481,494 Necrosis 993

1330

NectandrinB 198 Neisseria gonorrhoeae 613 Nemalion vermiculare 1137 Nematocidal activity 439,443,451,454, 461,466, 469,470,474,482,485,887 of fatty acids 434 Nematodes 428,429 Chaenorhabdtis 429 control of 430-432 Nematostatic activity 429,430 Nemerosin 215 Neoglycoconjugates 1159 Neohesperidin 744 Neoipecoside 129 Neoisopecoside 129 Neojusticin A 243 NeojusticinB 194,235 Neolignans 244,551 7-S, 5-5Neolignan 571 7-R, (9-/? Neolignan 571 Neolignans 184 Neoolivil 198 Neoplasic cell lines 752 Neoplastic activity 452 Neosiphonia 1175 Neosiphonia superstes 1207-1210 Neosiphoniamolide A 1210 Neothalibrine 825,827 Nepeta 172 Nerve growth factor 951 Nervous polyarthritis 400 Neuroblastoma 924,927 cell differentiation of 927 Neuroblastoma development 754 Neuroexcitatory activity 479 Neuroleaena lobata 808 2,3-Neurolenin 808 Neurolenin A&B 808,809 Neuroleptic drugs 270 Neurospora crassa 1016,228,1128 Neurotoxicity 965,1004 HK-Neurotoxicity 993 Niacinic acid 968 Nicotiana plumbaginifolia 1146 Nicotine 487

Nicotinic acid 965 Nifurtimox 792 Nigrospora oryzae 227 Nimbinin 815 Nimbolide 815 Nippostrongyliis brasiliensis 465 kalihinol Y from 465 Nirtetralin 195,237 Nitrate reductase-dependent NO production 930 Nitric oxide 909,911-914 chemistry and bioactivity of 909, 911-914 as signal in plant defense response 937-949 Nitroquinoline-A^-oxide 271 Nitrosoglutathion (GSNO) 948,951 S-Nitrosylation 914,924,925,939 Nitzschia pungens 479 'CNMRdata 1052 for monoindoles 1052 NMR studies in structure elucidation 1040 Nitric oxide as mediator of the antitumor immune response 921-923 as plant growth regulator 934,935 cell apoptosis 925,926 cell differenciation 926 classification of 915 in plants 929 synthases 915 synthesis in plants 930 Nobiletin 746,752 Nobilisitine A&B 610,619 Nocardia 451 Nodularia spumigena 367 NO -mediated intramolecular disulfide formation 914 NO-mediated apoptosis 943 in animals 943 NO-modulated transcription factors 946 Non-specific immune response 923 Nonblue 1271 Non-phenolic linear diarylheptanoids (type I) 883,884,905

1331

Non-specific damage on DNA 1004 Non-volatile constituents 297-301 of saffron 297-301 Norambrienolide ether 394 Nordamnacanthal 641,657,663,673 Nordihydroguairaretic acid (NDGA) 191 Normelionine F 1060 Nortrachelogenin 192,233 (+)-Nortrachelogenin 270 NOS-dependent NO production in plant cells 931-934 NOS toe commune 362 Nostoe l^SW95/\0 367 Nostoe sp. 352,366,372 Nothenia anomala 440 Nothenia brasiliensis 446,463-465, 471,483-485,487,488,492,493 Nothenia dubius 432,463,471 Nothogeia fastigiata 1138 Novobiocin 366 Nudiposide 195 (+)-Nyasol 230,231 antiprotozoal agent 230 Nyctanthes arbortristis 807 (-)-Nymphone 199 Obaberine 824,826 Obavatifol 202,219,220 Obovaten 202,219 Obovatinal 219,220 Ocotea foetens Aiton 850 Octadecyl ferulate 666 Octadecanoic acid 361 Odoracin 465 Oleic acid 434 Oleiferin-B 228,230 Oleiferin-F&G 228,230 Oleoside 319 Oleuropein 324,334,343 Oligophagous chrysomelid beetle 871 Oligosporon 449 Olivacine 828,829 from Aspidosper ma olivaceum 144 Ommatins 965,991,992

Ommins 990 Ommidinis 990 Ommochrome 998,1018 Omphalotins A-D 483-486 Omphalotus illudens 706 Omphalotus olearius 483 Onitin 706 Onnamide A 1198,1231,1232,1233,1234 Onychium aiiratum 706 Ophiobolin 468 Ophiobolin K 468 Orbiculamide A 1181 Orchis morio 1144 Oriamide 1184 Origanum sp. 746 Ornithine decarboxylase inhibition 924 Ornoside 318,319,322 Ornosol 319,333 Oroidin 819 Orophea enneandra 204,255 (-)-epieudesmin 204 (-)-eudesmin 204 (-)-phylligenin 204 Oscillatoria 352,358 Oscillatoria agardhii 352,365 Oscillatoria cf chalybea 362,373,379 Oscillatoria latevirens 364 Oscillatoria perornata f attenuata 373, 374,379 Osmanthus 327 Ostertagia circumcinta 453,490 Ostertagia ostertagi 484 Ostreopsis lentcularis 69 Ostrinia nubilalis 186 Otivarin AA\M2 Otoba parvifolia 186 Otoseine 875 Ovalicin 459 Oxandra espinata (Annonaceae) 807 Oxazolidine-2-thiones (goitrogens) 1074 Oxidative coupling of lysine and cysteine 996 2-Oxo-2,3-dihydrotryptamine 117 l-Oxo-4'-Demethoxy-3',4'-methylenedioxyrocaglaol 203

1332

2-Oxo-isocaproate 968 7-Oxoroyleanone 402 Oxotryptamine 118 Oxychlororaphine 365 Oxyhemoglobin 986 Ozonation 355 Ozonolysis 81,86 Paccar di 1215 Paecilomyces lilacinus 432 PAF-induced inhibitors 240 (+)-acetoxypinoresinol dimethyl ether 241 arctigenin 241 arctigenin methyl ether 241 chamigrenal 244 cinnamophilin 240 dehydroschisandrol A 245 denudatin 240 (+)-epipinoresinol dimethyl ether 241 epiyangambin 242 fargesin 241 gomisin M 240 honokiol 240 (+)-isogmelinol 241 isomagnolin 241 isoyangambin 242 justicidin B 243 justicidin E 243 kadsurenin H 240 kadsurenone 240 L-653,150 242 machililin G 242 magnolol 240 magnoneA 245 magnone B 245schisandrin A 244 we^o-dihydroguaiaretic acid 240 methylpiperbetol 243 neojusticin A 243 phillygenin 241 (+)-pinoresinol dimethyl ether 241 piperbetol 243 piperol A 243

piperol B 243 pregomisin 244 puberulin A 243,244 puberulin C 243,244 schisandrol A 245 (-)-syringaresinol 243 taiwanin E 243 taiwanin E methyl ether 243 (+)-yangambin 242 PAF-induced thrombocytopenia 242,243 Palasonin 457 Pallidol 555,560 Palmitic acid 434 Palythoa sp. 69 Panagrellus redivivus 429,432,434,435, 438, 450,455,467,473 Panagrolaimus 488 Pancratium maritimum 610,613,615-617 Pandanus odoratissimus 255 3,4-bis(4-hydroxy-3-methoxy-benzyl) tetrahydrofliran 255 Panek's synthesis 1224,1225 of dipeptide 1224 of p-methylaspartate 1225 Panganensine R 1055,1057 Panganensine X&Y 1055,1057 Panotima angularis 698 Papilio 990 Papillary carcinoma 712 Papillomas 693 Papulin 592,593 Papulinone 592 Paraherquamide A 490,491 Paraherquamides C-G 490 Parasiticidal activity oflignans 229 Paratrichodorus 456 Paratylenchus hamatus 429 Parbendazole 444 Paromomycin 831 Paromomycin (aminosidine) 785 Parthenin 808,809 Parthenocissin A&B 567,568

1333 NO-Participation 936,937 in antioxidant cellular systems 936, 937 Parvaquone 804 Paterson's synthesis 1229 of swinholide A 1229 of iso-swinholide A 1229 Pathogen-induced oxygenase 947 Patulin 449 Pauridianthalyalii 1036 Pear {Pyrus communis) 592 Pederin 1233,1234 Pediastrum simplex 379 Peganum harmala 828 Pelargonidin 748 a-Peltain 193 p-Peltatin 193 methyl ether of 193 Peltatins 222 Penicillic acid 449 Penicillin G 366,374 Penicillium sp. 450,482 Penicillium carneum 490 Penicillium charlesii 490,1128 Penicillium chrysogenum 1281 Penicillium expansum 228 Penicillium nigricans 805 Penicillium paraherquei 490 Penicillium roqueforti 490 Peniophora 462 5-Pentadecen-l-ol 75 Pentamidine (pentacarinat(R)) 791,793 2-«-Pentylquinoline 820 Peppermint 745 Peripheral analgesic properties 400,415 Peripheric antinociceptive effect 416 (f Periplaneta 991 Periplaneta americana 991 Persea indica (lauraceae) 850,851,852 cinnzeylanol from 850 cinnzeylanone from 850 ^/7/-ciniizeylanol from 850 idicol from 850 innzeylanine from 850

perseanol from 850 ryanodanes from 850 ryanodol from 850 ryanodol-14-monoacetatefrom 850 vignaticol from 850 Persea obovatifolia 219 obavatifol from 219 obovatenfrom 219 obovatinal from 219 perseal A&D from 219 Perseal A-D 202,219 Perseanol 850 PerthamideB 1187 (-)-Pessione 823 Peterson olefmation 1239 Phag-inhibitory activity 187 of Melia azedarach 187 Phanerochaete chrysosporium 186,1014 Pharmacological properties 556-559 of resveratrol oligomers 556-559 Phaseolus lunatus All Phaseolus tribe 931 Phaseolus vulgaris 1136 Phase-sensitive NOESY spectrum 12-14, 29,34 of bisdesmoside 41 of dianchinenosideE 34 ofmimusopin 14 ofSAPO30 41 ofSAPO50 41 of saponariosides C 12 of vaccaroside D 29 Phenalenone 474 9-Phenanthrenemethanolhalofantrine chloroqiune 782,820,825,836 Phenolic linear diaryIheptanoids 885-894, 905 Phenoxazinone 965,1014,1009 Phenoxazone 493 Phenylalanin ammonia-lyase 938 p-Phenyllactic acid 592 Phenylpropanoid pathway 154 A^-Phenylquinoneimine 976 vV-Phenyl-a-naphthylamine 359 A^-Phenyl-(3-naphthylamine 359

1334

Pheochromocytoma 927 Pheomelanins 1018 Phillygenin 237,241 Phillyrin 201 Phlebiaradiata 1013 Phlebotomus 783 (-)-Phlligenin 201 9'-(9-(-)-Phlligenin 255 Phllirin 237 Phlorizidin 800,801 Phorn is 172 Phom ops is foeniculi 1128 Phormidium autumnale 358 Phormidium bohneri 365 Phormidium sp. 364 Phosphomannomutase 1140 Phryma leptostachya 231 (+)-haedoxan A from 231 (+)-haedoxan D from 231 (-)-haedoxan D from 231 Phthalimides 1161 Phycobiliproteins 376 PhyllamycinB 196 Phyllamyricin E 194 Phyllamyricoside A 196 Phyllanthin 237 Phyllanthostatin A 196 Phyllanthus acuminatus 211 phyllanthostatin A from 211 Phyllanthus amarus 217 hypophyllanthin from 217 phyllanthin from 217 Phyllanthus myrtifolius 225 phyllamyricin E from 225 phyllamycin B from 225 phyllamyricoside A from 225 retrojusticidin B from 225 Phyllanthus niruri 233,237 hyophyllanthin from 233 hypophy llanthin from 237 nirtetralin from 237 phyllanthin from 237 Phyllantin 191 Physiological effects of phytoestrogens 207

Physostigmine 838 Phytoalexins 468, 472,473,494,745,1289 Phytoecdysteroids 699 Phytoene 3299,300 Phytofluene 299,300 Phytoparasitic nematodes 427,428 Phytopathogenic fungi 418 Botrytis cinerae 418 Fusarium solani 418 Phytophthora nicotianae var. parasitica 418 Rhizoctonia solani 418 Sclerotinia sclerotiorum 418 Sclerotium rolfsii 418 Phytophthora infestans 936 Phytophthora nicotianae var. parasitica 404 Phytotoxic activity 452 Piceid 530,571 Pichia pastoris 1263,1269,1288 Picrocrocin 293,295,296,298-300,303, 306,307 Picropodophyllone 193 Picropodophyllotoxin 166,193 Picrotoxinin 867,868 Pictet-Spengler reaction 128 Pieris brassicae 702 Pier is rapae 439 (+>Pinoresinol 157,177,188, 200, 229,270,474 Pinoresinol 249 (±)-Pinoresinol diglucoside from 235 Pinoresinol dimethyl ether 233 neojusticin B 235 (+)-Pinoresinol dimethyl ether 241 (+)-l-Pinoresinol dimethylether 200 (-)-Pinoresinol di-P-D-glucoside 200 Pinoresinol monoglucoside 235 (+)-Pinoresinol monomethylether 200 (+)-Pinoresinol-p-D-glucoside 200,270 (-)-Pinoresinol-P-D-glucoside 200 Pinosylvin monomethyl ether 472 (±)-Pinresinol 267,268,269 Pinus 905 Pinus canariensis 859

1335 Pinus densiflora 471 Pinus flexilis 887 Pinus massoniana 472,474 Pinus thunbergii 471 Piper betle 243 methylpiperbetol from 243 piperbetol from 243 piperol A&B from 243 Piper betle 469 Piper decurrens 232 Piper futokadsura 240 kadsurenin from 240 kadsurenin H from 240 Piper kadsura 240 denudatin 240 Piper nigrum 819 piperine from 819 Piper puberulum 243 puberulin A&C from 243 Piperbetol 243,244 Piperol A&B 243,244 Piquerol 807 Pirano flavonoids 744 Piricularia oryzae 1210 Pisum sativum L. 832,1263 Plakinastrella sp. 73 Plakortis 492 Planinin 267 Plant growth regulator 585 from phytopathogenic bacteria 585 as 2'-deoxyzeatin riboside 585 as dihydrozeatin 585 as dihydrozeatin riboside 585 as O-hydroxynitropapuIine 585 as indolacetic acid (lAA) 585 as indolaldehyde 585 as isopentenyladenine 585 as methyl ester of IA A 585 as I'-methylzeatin 585 as r'-methylzeatin riboside 585 Plasmodium 779-781,797,798,800,802804,815,825,1030,1035,1036,1063 Plasmodium berghei 1063 Plasmodium chabaudi 789

Plasmodiim7 falciparum 230,231,456, 466,781-783,788,789,797,800,801,804, 805,808,811,813,815,816,819,820, 822-824,828,837-839,1037,1063 Plasmodium lophurae 803 Plasmodium malariae 781,837 Plasmodium ovale 781,789 Plasmodium trophozoite 781 Plasmodium vivax 781,783,789,803,837 Plasmodium yoelii 789,828 Plastoquinone 368 Platelet activating factor (PAF) antagonism and coagulation 240-245 Platelet aggregation 756,758 Platycarya 905 Platycarya strobilacea 899 Platycaryol 899 P38 leukemia Pleurotus ostreatus 227 Plumbagin 470,804 Plumeran 140 Pneumocystis carinii 804,835 Podophyllotoxin 149,154-159,166,168178, 184,193,205,210,220,231,272,799, 858 derivatives of 162,163,164 4-aminoaryl 162 1-o-quinone 162 Podophyllum hexanrum 227 4'-0-demethldehdropodophyllotoxin from 227 picropodophyylone from 227 podophyllum kexandrun 227 picropodophyylone from 227 Podophylum 149,151,153,154,156-158, 169-174,177,178,183,183,188,205 Podophylum bankakri 149 Podophylum emodi 149,153,154,156, 158,171,173,178 Podophylum hexandrum 149,156 Podophylum peltatum L. 149-152,158, 171-178 Podopyllotoxin-based anticancer drugs 160 epipodophylltoxins 160 etopophosphate 160

1336 etoposide 160 etoposide phosphate 160 teniposide 160 Podopyllotoxin-P-D-glucoside 193,210 Podorhizol 192 Polyarthrtis 398 Polydiscamide A 1201 Polygodial 461 Polygyala gazensis 204 eudesmin from 204 kobusin from 204 magnolin from 204 yangambin from 204 Polyphagous lepidopteran 871 Polytheonamides 1193 Ponasterone-A 700 Ponasteroside 700 Populus tremuloides 358 Porlieria chilensis 249 dihydroguaiaretic acid from 249 guayacasin from 249 isopregnomisin from 249 Porson 895 Portiera hornemannii 809 Potam ogeton sp. 361 Pratylenchus 432 Pratylenchus coffeae 436 Pratylenchus curvitus All Pratylenchus neglectus 464 Pratylenchus penetrans 429,436-438,449, 473,477,481 Pratylensus scribneri 472 Precocene II 471 Pregomisin 244,247 (+)-Prennylpiperitol 267 (-)-Prenylpluviatilol 267 Prestegane B 192,249,269 Primaquine 836 (-)-Prinsepiol 200,270 Pristimerin 813,814 Proanthocyanidins 672 Procyanidin 697 Prodelphinidin 697 Proguanil 783 2-/7-Propylquinoline 820,821

Proserpinaca palustris 359 Protease inhibitors 800 Protein S-nitrosylation 913 Proteus m irab His 402,1155 Prunin 745 Prunus sp. 1157 Pseudocolelomic cavity 469 Pseudomanas syringae 583,590- 592, 594,595,598,599,607 Pseiidomonas 969,1134 Pseudomonas aeruginosa 330,331,365, 402,1151 Pseiidomonas alginate 1151 Pseudomonas amygdali 583,587,590, 599 Pseudomonas caryophylli 594 Pseudomonas ciccaronei bacteriocins 595,599 Pseudomonas diminuta 1155 Pseudomonasfluorescens 78,333,334 Pseudomonas sp. {Alteromonas) 69,582, 584,588,597,617,1124 Pseudomonas syringae pv. maculicola 593, 933 Pseudopuropurin 657,663,666,675 Pseudotheonamides A„A„B„C, & D 1194 Pseudoxanthoma elasticum 675 PSIl inhibitors 368,371 Psychodelic drugs 820 cocaine 820 morphine 820 semisynthetic LSD 820 Psychotria campoutans 804 Ptaquiloside 703,705,706,707,709-712, 714,716,719,722-725,727 three dimensional view of 710 Ptaquiloside Z 719,720 Ptaquilosin 723 Pteidanoside's 703,722 toxicity of 703,722 Pteridanoside 722 Pteridioside 704 Pteridium aquilinum var. caudatum 687, 690-692,695,696,698,699,702,703,720, 722,724,725,727

1337

Pteridium arachnoideum 699 Pteridium esculentum 699 Pteridium latiusculum 690,702 Pteris aquilina 691 Pteris cretica 722 Pterocarpanes 745 PterosinA&B 702 PterosinV 700,701,702 Pterosterone 700 papuline 585 papulinone 585 /-zeatin 585 /-zeatin riboside 585 PuberulinA&C 243,244 Puccinia podophylli 176 (/?)-(+)-Pulegone 86,88 Purpurin 640,649,657,663,666,674 Purpuroxanthin 657,666 Putrescine 1259,1260,1263,1276,1277 Pycnamine 824,826 Pycnanthus angolensis 219,473 Pycnoporus 1013 Pycnoporus cinnabarinus 1010-1016 Pycnoporus coccineus 1012 Pycnoporus sanguineus 1012 Pyrimethamine 805 Pyrimidinone 220 Pyrogallic acid 359 L-Pyroglutamic acid 1242 Pyrole-2-carboxylic acid 819 Quadrangularin A 567 Quassin 816,817 Quercetin 314,359,742,746-748,750, 757,758,696,697,714,1008 Quercetin-3,7-digalactoside 314 Quercetin-3-glucoside 314,763 Quercitrin 359 Quilic acid [3-0-P-D-galacttopyranosyl(l->2)[|i-D-xylopyranosyl-(1^3)]-p-Dgiucurono-pyranoside] 21 Quillaia saponaria 54 Quinidine 144,836 from Cinchona ledgeriana 144

Quinimax 836 Quinine 97,780,781,800,835,838,888 Quinizarin-2-carboxylic acid 664 Quinoaldimine 1272 Quinoketimine 1272 Quinolinic acid 965,966,968,970 o-Quinoneimine 976,986,988,992,998, 1000,1002 Quinoneimine 979 Quinonization 1287 Quinoproteins 1259,1270,1273,1277, 1278,1279,1283,1285,1286,1290 Quisqualic acid 476 Quisqualis indica 476 Radicicol (monorden) 452 Radopholus similis 474 Ralstonia solanacearum 933 Raman spectrometry 1267 Raphanus sativus L. 1075 (3R*, 6S*)-Raphanusanins 1084 3R*, 6R*Raphanusanins 1084 Rauwolfia serpentina 1060 Reduction 69,71,123 of 3-methyloxacylotridecan-2-one 69 of methyl 10(rer/-butyldimethyl silyloxy)-2-methyldecanoate 71 relative Van der Waals distances 123, 127 Reidispogia coerulea 1207 Reidispogiolide C 1209 Reidispongia 1175 Reidispongia coerulea 1207,1208 Reidispongiolide A-C 1207,1208 Repandine 824,826 Resda luteola 632 Reserpine 143 from Rauwolfia serpentina 143 Restrytisol A-C 553,558,570,571 Resveratol 753 Resveratrol oligomers 533,534 isolation of 533,534 cw-Resveratrol 555,571 /rara-Resveratrol 560

1338

Resveratrol oligomers 507-511,513-529, 532,556 sources of 513-529 Resveratrol trans-dohydrodlrnQv 555,565 Resveratrolside 530,571 Retrojusticin B 194 Rhabditissp. 432,488 Rhamnaceae 637 Rhamnetin 314 Rheumatoid arthritis 170,486 Rhinacanthin E&F 191,226 Rhinacanthus nasutus 226 rhinacanthin E&F from 226 Rhinoclaviella sp. 673 Rhizobium lipoplysaccharides 931 Rhizoctonia solani 404 Rhizopus oligosporus 1007 Rhodnius prolixus 187,230 Rhodotorula rubra 61?> Rhoiptelea 905 Rhoiptelea chiliantha 895,902 Rhoiptelol A&B 895,902 Rhyncholacis penicillata 471 Riboflavin 1007 Ricinoleate 361 Rishitin 460 Ristocetin 366 Rocaglaol 220 Rodophilia bifida 615 Rollidecin 793 Rollinia emarginta 793 Rolliniastatin-1 793 RoridinE 811,812 Rosmarinus sp. 746 Rotenone 473 Ruberythric acid 629,640,642,657,674 Rubia teep 675 Rubia tinctorum (madder) 629-632, 638,639,641,643,644,647,648,650,653, 656,657,658,666,675-677 anthragallol from 657 lucidin from 657 munjistin from 657 nordamnacanthal from 657 pseudopurpurin from 657

purpurin from 657 rubiadin from 657 xanthopurpurin from 657 Rubiadin 657,664,666,676 Rubiadin 3-(i-primeveroside 657 Rubiadin glucoside 657 Rubiadin-1 -methyl ether 806 Rubianin 664 Rubicunosides A-D 45,46 from Silene rubicunda 45 Russulaceae 461 Rutin 314,666,747,762,763,768,769 Ryanodanes 850 Ryanodine/spiganthine ryanoids 852,853 Ryanodine-type compounds 850,852 Ryanodol 850,856 Ryanodol-14-monoacetate 850 Saccharomyces cerevisiae 227,228,618, 673,1123,1125,1126,1129,1283 Saccharomyces kluyveri 1127 Sacchromyces sp. 1087 Saffron 296,297,299,301-305 Safranal 293,295,296,298,300,303,306, 307 Sainfoin 1268 Sakuranetin 802 Saliva desoleana 407,409,410,411 Composition of 407,409,410,411 Saliva officinalis 391,403,404 Saliva sclarea 404 Composition of 404 Salmonella 1161 Salmonella enteridis 613 Salmonella newington 1160 Salmonella paratyphi A-C 673 Salmonella typhi 673 Salmonella typhimurium 673,675,676, 707,712,722,1002,1093,1099-1101,1104, 1105 Salvia 172 Salvia desoleana 391,406,407,413-419 Salvia miltiorrhiza 235 danshensuan B from 235

1339

Salvia sclarea 391-395,397,398,400-406, 412,413,415-419 Sanguinarine 488 Sanshool 450 Santonin 462,463,466,477,478 SAPO30 41,42 from Gypsophila paniculata 41 SAPO 50 41,42 from Gypsophila paniculata 41 Saponaria officinalis 3,4,10,13,15,18,27, 30,42,54,57 antiinflammatory activity of 27 gypsogenic acid from 5,6,18 gyspogenin from 5,6,18 3(i-hydroxy-oleana-l 1, 13(18)-dien23, 28-dioic acid from 5,6 saponarioside from saponariosides C-M from 10,12, 13,17,18 saponaside A from 17,20-23,42 3, 4-secogypsogenic acid from 5,6 triterpenoid saponins from 17,20 Saponarioside C 26 HMBC correlations of 15, 25 HOHAHA spectrum of 10 MALDI-TOFMSof26 phase-sensitive NOESY spectrum of 12 SAR of podophyllotoxins 167,168 Sarcoma-180 303 Sargassum muticum 1152 SarothalenB 797 Savinin 231 Scenedesmus obliquus 373 Scheriber's synthesis of discodermolide 1244-1246 Schisandra 183 affect on cardiovascular system 183 decrease by heart rate 183 display platelet activating factor antagonist activity 183 for the treatment of hepatitis 183 Schisandra chinensis 203,204,244,247, 249-251,255,258,265

chamigrenal from 244 gomisin N from 249 pregomisin from 244 schisandrin A from 203,244 schisandrol A&B 203 schisandrin C 203 Schisandra rubriflora 205 Schisandrin 238,270 Schisandrin A 197,244 Schisandrin C 224 Schisandrol A 197,245 Schisanhenol 250,251 Schisantherin D 197,223,224,247,256 Schistocerca gregaria 700 Schizandrin 197 5-(-)-Schizandrin C&D 250 Schizonepeta tenuifolia 251, 252 schizotenuin A from 251,252 schizotenuin C, from 251,252 schizotenuin Q from 251,252 Schizonticidal drug 837 Schizosaccharomyces pombe 1093,1097, 1098,1128 Schizotenuin A 251 Schizotenuin C&C, 251 Sclareol 394,402 Sclareolide 394 Sclelerotinia sclerotiorum 405 Scleritoderma sp. 1215 Sclerotium rolfsii 405 Scoparone 315,332-334,342 Scopoletin 315,333,666 Scytonema hofmanni 363 3,4-Secogypsogenic acid 29 Secoilidoids and phenylethanoids 318-329 from Fraxinus ornus 318-329 Secoirridoids 95 (-)-Secoisolariciresinol 156,157,186,188, 191 Secoisolariciresinol 204,205,206 (-)-Secoisolariciresinol diglucoside 191 Secoisolariciresinol diglycoside 248 Secologanin 95-98,100,105,110,111,114118,127,132,137,138,140,143,144

1340 Secoxyloganic acid 100 Segetalic acid 30 Selenastrum capricornutum 373,374, 377-379 Semiquinolamine 1273 Senecio 872 Senecio erraticus 874 Senecio glaber 874 Senecio litura 851,853-855,858-860, 863-866 Senecio microphyllus 874 Senecio palmensis 859 Senecionine 876 Senna occidentalis 1143 Senna s^^. 1139 Ser264 inhibitors 368 Sergolide 815 Serotonin 267 Serotonine 965,965 Serpentine 1060 Serratia marcescens 402 (+)-Sesamin 200,267 Sesamin 231,259,260,264 (-)-Sesamin 268 Sesaminol 249,260,264 Sesamol 249,253 Sesamolin 253,260,231 Sesamolinol 249,253 Sesamum indicumi 249 pinoresinol from 249 sesaminol from 249 sesamol from 249 Sesmin 246,264 Setaria cervi 467 Setaria italica 476 Sharpless asymmetric epoxidation 89 of (E)-2-nondecen-l-ol 89 Shewanella putrefaciens 69 Shigella ambigua 61?> Shigella boidii 673 Shigella flexneri 673,1160 Shigella largei-sachsii 673 Shigella sonnei 673 Shogaol 475 Silandrin 255

Silchristin 255 Silene fortunei 46,47,55 Silenejenisseensis 45,46,47 Silene latifolia 45 Silene species 45 Silene siicculenta 45 Silene villosa 45 Silene vidgaris 47 Silenosides A-C 47,48 from Silene vulgaris 47,48 Silphinene 864,865,866,867,871 Silybin 255 Silybum marianum 255 Silydianin 255 Silymarin 754 Silymonin 255 Simaba guianensis 815 Simalikalactone D 815,816 Simarouba 837 Sinapis alba L. 1074 1 '-0-Sinapoyl-6'-0-galloyl-p-Dglucopyranose 359 Sindbis virus 223 Sinefrmgin 830 Sinensetin 746 Sinigrin 481 Sitoindoside II 219 P-Sitosterol 358 Smith periodate degradation 1156,1157 Snatzkeins A-E 52 from Arenaria fdicaulis 52 Solanaceous steroidal glycoalkaloids 467 a-Solanine (solatriose) 818,868 Solanum tuberosum 936,941 Solarmargine 817 a-Solarmargine (chacotriose) 819 Solavetivone 461 Solricin 135 362 Sophoraflavescens 487 Sophora Japonica 1142 Sorgoleone 370 South American trypanosomiasis 787,788 Spathulenol 402 Spectral data 1078 forHMTP 1078

1341

Spermicidal activity 53 Gypsophila paniculata 53 Spermidine 1263,932 Spermine 1263,932 Spermostrychnine 1064,1065,1066 Sphaerotrichia 1121 (-)-Spinosine 823 (+)-Spinescin 267 Sphinxolides B-G 1207,1208,1209 Spigelia anthelmia 852 Spiro-indolopirrolidine 126 of ^p/ro-indolopirrolidine Spodoptera eridania 871 Sp adopter a litura 850 Spodopterra littoralis 850,852-854,869, 870-872,874-876 Spontaneous apoptosis 926 Sporangium cellulosum 450 Sporotrichum pulverulentum (anamorph of Phanerochaete chrysosporium) 1013 Squamocin 793,794 Squarroside A 31,50 coyncentration dependent immunomodulatory effect of 31 from Vaccaria segetalis 50 Staphylococcus aureus 77,79,83,227, 330-333,402,673 Staphylococcus epidermidis 402,1093, 1095,1097,1098 Staphylococcus spp. 556 Stearic acid 461 Steganacin 196,216 Steganangin 216 Stegangin 196,216 SteganoateB 216,217 SteganolideA 216,217 Steganotaenia araliacea 216 episteganangin from 216 steganangin from 216 stegancin from 216 steganolide A from 216 steganoate B from 216 Stemmadenin 139 Stemonoporol 562 Stenbergine 610,614

Stenophyllol A&B 564 Stephanitis AA3 Stereum purpureum 1133 Stichococcus bacillaris 358 Stille reaction 1236 Stiretrus anchorago 66 Stoichactis helianthus 72 Stomoxys calcitrans 691 Str. chrysomallus 1010 Str lividans 1010 Streblus asper 467 Strenbergia lutea 609-612,614,615 Streptococcus faecal is 71,S3 Streptococcus haemolyticus 613 Streptococcus mutans 64,226,227 Streptococcus pneumoniae 1154 Streptococcus salivarius 402 Streptococcus sanguis 402 Streptomyces ^'mdixmdimsQ 1145 Streptomyces antibioticus 1010 Streptomyces griseolus 830 Streptomyces griseoviridis 444 Streptomyces griseus var autrophicus 454, 1123,1155 Streptomyces incarnatus 830 Streptomyces purpeofuscus 373 Streptomyces sp. 426,442,443,450,453, 488, 1009 Streptomyces toyocaensis 493 Streptosporagium sibiricum 1211 Strictosamide 115,127 Strictosidine 114-116,127,128,130,131, 135 3a-H Strictosidine 97 Strongyloides ratti 486 Strongyloides stercolaris 486 Structure elucidation 535-551 of reseveratrol oligomers 535-551 of ptaquiloside 708 Strychinine 1030,1049,143 from Strychnos nux vomica L.143 Strychnine TV-oxide 1049 Strychnine 868 Strychnobrasiline 1050,1054,1055, 1058,1063

1342

Strychnofendlerine 1064,1065,1063 Strychnos alkaloids 1029-1031, 1040, 1049, 1053,1054,1060,1063-1065,1067 Strychnos atlantica 1050 Strychnos divaricans 1055 Strychnos guianensis 1055 Strychnos monoindolQ 1061 anticancer/protozoal 1061 istonine 1061 cryptolepine 1061 5,6-dihydroflavopereine 1061 harmane 1061 matadine 1061 melinonine 1061 normelinonine 1061 serpentine 1061 strychnoxanthine 1061 Strychnos panganensis 1052,1055 Strychnos sp. 1032-1034 Alkaloid of 1032-1034 Strychnos vom ica 1052 Strychnospermine 1064,1065,1066 Strychnoxanthine 1060 Strychnos alkaloids 1063 chemosensitizing activity of 1063 Strychonos bifurcata 1034,1039 Strychonos camptoneura 1060 Strychonos colubrina 1029 Strychonos decussata 1032,1035,1036 Strychonos diplotricha 1034,1038 Strychonos floribunda 1032 Strychonos grossweilleri 1060 Strychonos henningsii 1032,1039,103 5 Strychonos madagascariensis 1033,1035 Strychonos mitis 1033 Strychonos mostueoides 1034,1035 Strychonos myrtoides 1034,1035,1038, 1039,1050,1053,1065 Strychonos nuxvomica 1029 Strychonos panganensis 103 3 Strychonos pentantha 1034,103 9 Strychonos potatorum 1031,1032,1036 Strychonos spinosa 1033 Strychonos trichoneura 1034 Strychonos usambarensis 1060,1063

Suberin 1262 Succinate dehydrogenase 800 Succinate ubiquinone oxydoreductase inhibitions 924 Suffruticosol A-C 555,566 Sulfinemycin 443 Superoxide dismutase (SOD) 974 Superstolide A&B 1209 Supinine 874 Suramin 342,791 Suzuki coupling 1248 Suzuki-Miyaura cross coupling 74 of 4-bromo-1 -buty 1-9-borabicy clononane 74,75 Sweet pea {Lathyrus odoratus) 1268 Swem oxidation 1245 Sweroside 103,104,105 Swertia chirata (Loganiaceae) 806 Swiholides D-G 1198 Swinholide A&B 1196,1227 Swinhonsterols A-C 1178 Sylvaticin 793 Symphacia obvelata 463,471,478,493 Syndrome type-lA 1130 adipose tissue distribution 1130 cerebellar dy sfunciton 1130 liver insufficiency 1130 peripheral neuropathy 1130 psychomotor retardation 1130 Synechococcus cultures 363 Synechococcus leopoliensis 359,362-364 Synechocystis PCC 6803 366 Synthesis 635-637,642 of alizarin 635-637,642 of mannans and oligomers 1158 Synthetic derivatives of etoposide 162 acetyl salicylic acid 162 4P-aminoaniline 162 2-aminobenzothiazole 162 2-mercaptobenzothiazole 162 Syringa 327 (-)-Syringaresinol 200 (+)-Syringaresinol 200 (±)-Syringaresinol 267,270 (-)-Syringaresinol diacetate 200

1343

Thannilignan 224,225 Theileria 804 Theoezeolide A 1199 Theoneberine 1200 Theonegramide 1186 Tabtoxinine 372 Theonella conica 1189 Tageretin 746,752,758 Theonella cupola 1190 Tagetes sp. 437 Theonella mirabHis 1191 Taiwania cryptomeriodes 232 Theonella sp. 1175,1176,1179,1181, helioxanthin from 232 1184,1188,1193,1198-1200,1212,1217 TaiwanianE 194,243 Theonella swinhowe 1176-1178,1180, methyl ether of 194,243 1186-1191,1194,1196,1197,1200 Takai's homologation 1224 Theonelladins A-D 1200 Tamine 965 Theonellamides A-F 1185 Tamoxifen 222 Theonellapeptolide la-Ie 1187-1189 Taverniera abyssinica 473 Theonellasterol 1176,1177 Taxaceae 185 Theonellasterone 1176,1177 Taxodiaceae 185 Theopalauamide 1186 Taxol 753,811,812,828,830 Theopederins A 1199 Taxus brevifoUa 212 Thermoplasma acidophilum 1155 brevitaxin from 212 Thiarubrine A&C 437 Tazettine 614,618 Thiohydroximate-(9-sulphonate 480 Tecoma stans 1010 2-Thioxopyrrolidines 1096,1098,1099 Tectoquinone 665 Thopsentia sp 464,466 Tellimagrandin II (eugeniin) 359,372 Three dimensional NMR spectroscopy Temuconine 824,825,827 1044-1046 Teniposide 160,163,212,213 Thromboxane A^ 266 Termilignan 224,225,228,231 Thuja plicata 188 Term inalia hellerica 228,231 Thujopsis 172 anolignan Bfrom231 Thymus s^. 172,746 anolignan B from 224 Tigloylgomisin P 197 termilignan from 224, 231 Tingenone 813,814 thannilignan from 224 Toad {Bufo regularis) 689 Terpenic indanones 701 Tobacco {Nicotiana tabacum) 1145 O, V, O \ O -Tetraacetyl-4-(4' '-brom a-Tocopherol 33 7,993,1006 benzyl) strictosidine 127 (±)-a-Tocopherol 251,258,261 a 'aO',0-Tetraacetyl-4-(4"(±)-Y-Tocopherol 261 brombenzyl) vincoside 116 Tolypocladium niveus 486 O, 'O, O', O -Tetraacetylsecologanin 116, Tomatine 818 118, 125 a-Tomatine 467 Tetrahydrobiopterin 917,950 TOP-53 163 Tetrahydrofuran lignans 198,199 used for anticancer treatment 163 3,4,3',4'-Tetramethoxylignan-7-ol 228,230 Topa quinone 1259,1266-1269,1284,1285 2,6,8,10-Tetra-O-methyl NDGA 191 Topoisomerase I inhibitor 825 Teucrium 172 Topoisomerase II expression 924 (+)-Syringaresinol dimethyl ether 200 (-)-Syringaresinol di-P-D-glucoside 200 (+)-Syringaresinol di-P-D-glucoside 200

1344

Total synthesis 66,67,69,71,72,75,77,78, Tribulus terrestris 259 87,90 tribulusamide A from 259 of (1OZ, 15Z)-10,15-eicosadienoic tribulusamide B from 259 acid 90 Tribulusamide A&B 259 Trichinella spiralis 483 of (/?)-22-methy 1-5,9-octacosadienoic Trichoderma spp. 482,1128 acid 87 of 9, 13-dimethyltetradecanoicacid 66 Trichoderma viride cellulase 1145 Trichophyto rubrum 1159 of 10, 13-dimethyltetradecanoic acid Trichophyton mentagrophytes 673 67 Trichophyton spp. 1127,1128 of (5Z,9Z)-5,9-hexacosadienoic 82 Trichormus var. P-9 359 of (Z)-2-methoxy-5-hexadecanoic acid in agar-diffusion assays 359 77 Trichostrongtlus colubriformis 429,440, of (4£,75)-(-)-7-methoxy-4-tetra 443,453,461, 472,484,490,491 decenoic acid 78 of (/?)-2-methoxyhexadecanoic acid 78 Trichothecium 464 of 11-methyl-12-octadecenoic acid 69 Trichphyton mentagrophytes 227 isomagnolol 227 of 7-methyl-6(£)-hexadecenoic acid 71 magnolol 227 of7-methyl-6-octadecenoicacid 72 Triclisia sp/ 824 of methyl (Z)-2-methoxy-6-hexa Tricophton rubrum 228 decenoate 75 A239,\2A(rricophytonmentagro of ent-calyculin A \23A,\235 Tricuspidatol A 568 Toxicity assays Trifluoromethanesulphonic acid 1162 for extracellular promastigote 790 2,2,2-Trifluroethylhydrazonemethyl9Toxiferine-I 143 deoxy-9-oxo-a-apopicropodophyllate 195 from Strychnos toxifera \ 43 Trifolium repens 696 Toxocara canis 470,471,474,887 2,4,5-Trihydroxyphenylalanine 1266 Toxocara mystax 471 Trihydroxystilbene reseratrol 507,511 Toxoplasma 783,804 {-)-trans-2-{2>" ,4 " ,5' '-Trimethoxy Toxoplasma gondii 835 benzyl)-3-(3',4'-ethylenedioxybenzyl) Trachelogenin 192,233 butyrolactone 192 (-)-A^or-Trachelogenin 474 6,7,8-Trimethoxycoumarin 317,332Trachelosperum jasminoides 233 334,342 arctigenin from 233 2-Trimethylsilyloxy-6-hexadecenonitrile76 matairesinol from 233 Triphophyllum peltatum 813,822 nortrachelogenin from 233 Triterpenoid saponins trachelogenin from 233 from Agrostemma githago var. (-)-Trachyloban-19-oic 812 githago 49 Tramesanguin 1012 from Arenaria filicaulis 52 Trametes cinnabarina 1012 from Caryophyllaceae 6,48 Treatment from Dianthus barbatus 32 of Chagas disease 787 from Dianthus superbus var. of cryptosporidiosis 788 longicalycinus 32,33 of malaria 804 from Gypsophila bermejoi 44 from Gypsophila capillaris 43

1345

from Gypsophila perfoliata 38 from Herniaria glabra 51 from Saponaria officinalis 17,20 from Silene fortunei 47 from Silene villosa 45 from Spergularia ramosa 49 from Vaccaria segetalis 27 Trolox 672 Trychostrongylus 425 Trypanocidal activity 800 Trypanosoma berghei brucei 791,797 Trypanosoma berbhei gambiense 785787,791,792,828 Trypanosoma brucei rhodesiense 785787,791, 792,825 Trypanosoma congolense 785 Trypanosoma cruzi 230,673,787,788, 792, 793, 797,799, 801,802, 813,819, 822,828 Trypanosoma spp. 786,779,780,785, 786, 797,800,803,806-808,810,822, 824,828,838 Trypanosoma vivax 785 Trypanosomiasis 781,785,786,787,840 Trypanosomicidal activity 230 of veraguensin 230 of grandisin 230 Trypanothione synthesis 808 Tryptamine 966 Tryptase inhibitor 1129 (-+-)-Tsugacetal 195 Tuberculatin 194,263 Tubulin 753,828 Turbatrix aceti 481 (+)-ar-Turmerone 457 Tylenchulus semipenetrans 456,469 Tyrosin nitrosylation 926 Tyrosine phosphorylation 753 Tyrosol 333 Ubiquinol-10 1006 Ubiquinone oxydoreductase 924 Ugi reaction 1221 Uhdoside 321,322

Ungeremine 610,616,617,618 Ungernia minor 616 Unonpsine 827 Uronicacid 1154,1156 Urospora penicilliformis 1137 Usambarensine 1063 Usambarine 1063 Vaccaria segetalis 27,28,29,50,56 vaccarosides A-H from 27-30,56 VaccarosideD 29,30 enzymatic hydrolysis of 30 COSY of 29 HMBC/HSQC correlations of 29 HOHAHA spectrum of 29 phase-sensitive NOESY spectrum of 29 Vaccaroside E 30 Vaccaroside F 29,30 •H and 'C spectra of 30 acid induced rearrangement of 31 alkaline and enzymatic hydrolysis of 30 FAB- MS of 30 HMBC/HSQC correlations of 29 IR spectrum of 30 Valanimycin 443 Valeriana officinalis (Valerianaceae) 270 (+)-l-hydroxypinoresinol from 270 (+)-pinoresinol-P-D-glucosidefrom 270 (+)-pinoresinol from 270 (-)-prinsepiol from 270 Vanaliculatol 553 Vancomycin 366 Vatdiospyroidol 563 Vaticaffmol 553,562 Vaticaphenol A 563 Veatchine 868 Velbanamine 1047,1048 Velutinal 461 Veraguensin 230,231,799 Verapamil 1063 Veratraldehyde 1013

1346

Veratryl alcohol 1013 Verbascoside 327,328 Verbesina 437 Vemodalin 808,809 Vernonia amygdalina 808 Vernonia brachycalyx 798,811 Verolide 808 Verrucarin A 464 Verrucosin 228 Vertebrate-parasitic nematodes 428 Vesicular stomatitis virus (VSV) 223 Vezical hematuria 689 Vibrio alginolyticus 70 Vibsanol 255 Viburnum awabuki (Caprifoliaceae) 255 9'-0-(-)-phlligenin from 255 vibsanol from 255 Viciafaba 748 Vignaticol 850 Viguirea aspillioides 810 (-)-trachyloban-19-oic- from 810 Vinblastine from catharanthus roseus 828,829, 217,753 Vincristine 144 from Catharanthus roseus 144 Vincamine 144 from Vinca minor 144 Vincosamide 115,116,117,127 Vincoside 127,128,130,135,136 3p-HVincoside 97 Vindoline 1047,1048 (-)-Viniferal 570 Viniferiftiran 570 (-)-a-Viniferin 535,560 2-Viniferin 552,570 (-)-e-Viniferin 530 (+)-£-Viniferin 530,532,551 Viniferin 552,569 £:-Viniferin 553,560 a-Viniferin 553,560 (+)-a-Viniferin 558,560 (-)-^ra^5-e-Viniferin 560 (+)-a-/m«5-e-Viniferin 560 c/^-e-Viniferin 567

£-Viniferin diol 569 Vinorelbine 1048,1049 Virgilia 1157 Virola oleifera 228 3,4, 3' ,4'-tetramethoxylignan-7-ol from 228 oleiferin-B from 228 oleiferin-F from 228 oleiferin-G from 228 Virola seb if era 186 Virola surinamensis 230 grandisin from 230 veraguensin from 230 Vitamin K 638 Vitis vinifera 530 (+)-Vitisftiran A 570 (-)-Vitisfriran B 570 trans-Wimn A 558,568 VitisinA 568 (+)-c/5-Vitisin A 568 l-ytrans-Vitisin B 559,569 (-)-c/5-Vitisin B 569 (+)-Vitisin C-E 559,569 Volatile components 298 of saffron 298 Volatile isothiocyanates 1075 from Raphanus sativus 1075 Wadsworth-Homer-Emmons olefmation 1230 Warabi 691 Wedelia paludosa 810 e/7/-kaura-16-en-l 9-oic acid from 810 Weinerb reaction 1239,1243 Wikstroemia indica 270 (+)-nortrachelogenin from 270 Wikstromol 192 Wittig olefmation 1234 Wittig reaction 65,66,70,72,73,75,81, 83-91 of bromodecanal 83 of 8-formyloct-5-enoate 86 of (iE:)-5-methyltetradec-4-enal 72 of 12-hydroxy-2-methyldodecanal 70

1347

of 6-methyl-2-heptanone 65 of 7-methy 1-2-heptanone 65 of 5-methyl-2-hexanone 66 of 5-pentadecenal 75 Woorenoside-V 268 Wounding-Induced Protein Kinase 946 Wuweizisu B&C 247,251,256,258,263 Xanthan 1155 Xanthommatins 967,988-992,995 Xanthomonas 582,605 Xanthomonas c. pv. armoraciae 596 Xanthomonas c. pv. carotae 596 Xanthomonas c. pv. orizae 596 Xanthomonas c. pv. raphani 596 Xanthomonas campesths 596,1154 Xanthomonas campestris pv. vitians phytotoxins 596 Xanthomonas fragahae 596 Xanthomonas hortorum campestris pv. vitians 596,606 Xanthones 805,806 Xanthopurpurin 657,665,676 Xanthopurpurin 3-methylether 665 Xanthopurpurin dimethylether 665 Xanthoxylol 226 Xanthurenic acid 966 Xenobiotics 755 Xestospongia sp 487 Xiestoaminol A 487 Xiphimena diversicaudatum 460 Xylans 1156 Yamaguchi conditions 1229,1230 (+)-Yangambin 242,243,264 Yatein 157,192 (-)-Yatein 218 YinghaosuA 807 y-Ylidene-y-butyrolactone structure 363 Yohimbine 143 from Corynanthe johimbe K. 143 Zanthoxyllum naranjillo

799

Zanthoxylum ailanthoides 188 Zanthoxylum armatum 267 Zanthoxylum liebmannianum 450 Zeaxnthin 299,300,306 Zefbetaine 610,616,617 /^o-Zefbetaine 616 Zephyranthes flava 616 Zingiber 905 Zingiber officinale 902,474 Zymosan 342

This Page Intentionally Left Blank