studies m Natural Products Chemistry Volume 18 Stereoselective Synthesis (Part K)
Studies in Natural Products Chemistry edited by Atta-ur-Rahman
Vol. 1 Stereoselective Synthesis (Part A) Vol. 2 Structure Elucidation (Part A) Vol. 3 Stereoselective Synthesis (Part B) Vol. 4 Stereoselective Synthesis (Part C) Vol. 5 Structure Elucidation (Part B) Vol. 6 Stereoselective Synthesis (Part D) Vol. 7 Structure and Chemistry (Part A) Vol. 8 Stereoselective Synthesis (Part E) Vol. 9 Structure and Chemistry (Part B) Vol.10 Stereoselective Synthesis (Part F) Vol.11 Stereoselective Synthesis (Part G) Vol.12 Stereoselective Synthesis (Part H) Vol.13 Bioactive Natural Products (Part A) Vol.14 Stereoselective Synthesis (Part I) Vol.15 Structure and Chemistry (Part C) Vol.16 Stereoselective Synthesis (Part J) Vol.17 Structure and Chemistry (Part D) Vol.18 Stereoselective Synthesis (Part K)
studies in Natural Products Chemistry Volume 18 stereoselective Synthesis (Part K)
Edited by
Atta-ur-Rahman
H.EJ. Research Institute of Chemistry, University of Karachi, Karachi 75270, Pakistan
1996 ELSEVIER Amsterdam - Lausanne - New York - Oxford - Shannon - Tokyo
ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 211, 1000 AE Amsterdam, The Netherlands
Transferred to digital printing 2005
ISBN: 0-444-82458-8 © 1996 Elsevier Science B.V. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher, Elsevier Science B.V., Copyright & Permissions Department, P.O. Box 521,1000 AM Amsterdam, The Netherlands. Special regulations for readers in the USA - This publication has been registered with the Copyright Clearance Center Inc. (CCC), Salem, Massachusetts. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA. All other copyright questions, including photocopying outside of the USA, should be referred to the copyright owner, Elsevier Science B.V., unless otherwise specified. 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. This book is printed on acid-free paper. Printed and bound bv Anton\ Rowe Ltd, Eastboume
FOREWORD The present volume of this series should again provide highly interesting articles written by some of the most eminent organic chemists today. They range from stereocontrolled synthesis of complex natural products to structural studies on a variety of different types of natural products. It is hoped that this volume will be received with the same enthusiasm by the readers as the previous ones of the series. I wish to express my thanks to Miss Farzana Akhter and Syed Ejaz Ahmed Soofi for their assistance inthe preparation of the index. I am also grateful to Mr. Wasim Ahmad and Mr. Ahmed Ullah for the typing work and Mr. Mahmood Alam for secretarial assistance. Prof. Atta-ur-Rahman H.E.J. Research Institute of Chemistry University of Karachi
This Page Intentionally Left Blank
Vll
PREFACE Further developments in organic chemistry, natural products chemistry, and associated fields continue unabated. This high level of activity lies in sharp contrast to statements made during the past two decades by some prognosticators who had quite mistakenly predicted the rapidly approaching obsolescence of these fields of investigation. These predictions were based upon organic chemistry having reached a very mature level of development at a time when new areas of scientific inquiry were opening. Nevertheless, organic chemistry remains as vital and as active as ever in laboratories around the world. This continued activity may be attributed to many factors, including the development of new screening procedures for biologically active compounds, improvements in spectroscopic methods for determination of molecular structure, the availability of new, highly selective and often asymmetric methods for the synthesis of ever more complex, highly functionalized structures, and the applications of computer technology to chemistry. Another driving force for further work in organic chemistry continues to be the search for more effective pharmaceutical agents to treat many diseases such as cancer and other maladies that continue to plague humankind. In this same vein, continued searches are underway for new antibiotics to combat dangerous infectious bacterial strains that have become resistant to previously developed antibiotics. Organic chemistry has also been widely adopted as a tool for use in other areas of science, most notably in the biological realm wherein specially synthesized compounds can, for example, be used to probe the molecular details of cell function. In the most recent volume of this well-established series. Professor Atta-ur-Rahman again brings together the work of several of the world's leading authorities in organic chemistry. Their contributions demonstrate the rapid, ongoing development of this field by illustrating many of the latest advances in synthetic methods, total synthesis, structure determination, biosynthetic pathways, and biological activity. The opening chapter presents an overview of strategies for the synthesis of several classes of natural products with an emphasis on complex polycyclic systems. The next several chapters discuss the synthesis of specific classes of compounds, including morphine, polyketides, acetogenins, nonactic acid derivatives, complex spirocyclic ethers, 8-lactam and pyridone derivatives, inositol phosphates, sphingolipids, brassinosteroids, Hernandia lignans, and dimeric steroidal pyrazine alkaloids. Structure determination and biological function provide additional themes through many of these chapters. On the other hand, structure is discussed more exclusively in chapters on liverwort sesquiterpenoids, gymnemic acids, compounds of the Celastraceae plant family, fungal and protozoan glycolipids, and coumarins. Finally, the ever stronger links between chemistry and biology are reinforced by chapters on the origin and function of secondary metabolites, bioactive conformations of gastrin hormones, and immunochemistry. Professor Atta-ur-Rahman is to be congratulated for bringing together the present set of contributions as a continuation of this outstanding series. He has again met the goal of this series in demonstrating the strength, the vitality, and the diversity of organic chemistry as a central field of scientific investigation. Paul Helquist University of Notre Dame January 1996
This Page Intentionally Left Blank
CONTRIBUTORS
G. Adam
Department of Natural Products Chemistry, Institute Biochemistry, Weinberg 3, P.O. Box 250, D-06018 Halle/S.
of
Plant
N.L. Alvarenga
C.P.N.O. Antonio Gonzalez, Universidad de La Laguna, Carretera La Esperanza 2, La Laguna-Tenerife, Espana.
Masao Arimoto
Osaka University of Pharmaceutical Sciences, 10-65 Kawai 2-Chome, Matsubara 580, Japan
Nancy S. Barta
Department of Chemistry, University of Michigan, Ann Arbor, Michigan 48109-1055, U.S.A.
I.L. Bazzocchi
C.P.N.O. Antonio Gonzalez, Universidad de La Laguna, Carretera La Esperanza 2, La Laguna-Tenerife, Espana.
Eliana Barreto Bergter
Instituto de Microbiplogia da UFRJ, Centro de Ciencias da Saude-blocol 21.944-970-Cidade ijniversitaria, Rio de Janeiro-RJ
Maria Helena S. Villas Boas
Instituto de Microbiologia da UFRJ, Centro de Ciencias da Saude-blocol 21.944-970-Cidade Universitaria, Rio de Janeiro-RJ
Gabor Butora
Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061
Andre Cav6
Universite Paris-sud, Faculte de Pharmacie de Chatenay-Malabry, Laboratoire de Pharmacognosie, URA 1843 CNRS (BIOCIS)
Carsten Christophersen
Department of General and Organic Chemistry, The H.C. 0rsted Institute, K0benhavns Universitet, Universitetsparken 5, DK-2100 Copenhagen, Denmark
Helmut Duddeck
Institut fur Organische Chemie, Universitat Hannover, Schneiderberg IB, D-3000 Hannover 1, Germany
Bruno Figadere
Universite Paris-sud, Faculte de Pharmacie de Chatenay-Malabry, Laboratoire de Pharmacognosie, URA 1843 CNRS (BIOCIS)
Ian Fleming
Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 lEW, U.K.
Stephen P. Feamley
Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061, U.S.A.
A. Ganesan
Centre for Natural Products Research, Institute of Molecular and Cell Biology, National University of Singapore, 10 Kent Ridge Cresent, Singapore 0511
Manfred Gemeiner
Veterinar-Medizinische Universitat, Wien, Austria.
Sunil K. Ghosh
Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 lEW, U.K.
A.G. Gonzalez
C.P.N.O. Antonio Gonzalez, Universidad de La Laguna, Carretera La Esperanza 2, La Laguna-Tenerife Espana.
Andrew G. Gum
Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061, U.S.A.
Maria Helena
Institute de Microbiologia da UFRJ, Centro de Ciencias da Saude-blocol 21.944-970-Cidade Universitaria, Rio de Janeiro-RJ
GerdHiibener
Max-Planck-Institut fiir Biochemie, Am Klopferspitz 18a, D-82152 Martinsried, Germany
Tomas Hudlicky
Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061, U.S.A.
Akitami Ichihara
Faculty of Agriculture, Hokkaido University, Kitata 9, Nishi 9, KJTAKU, Sapporo 060, Japan
Tadao Kamikawa
Department of Chemistry, Kinki University, Faculty of Science & Technology, Kowakae, Higashi, Osaka 577, Japan
Jiirgen Lutz
Max-Planck-Institut fiir Biochemie, Am Klopferspitz 18a, D-82152 Martinsried, Germany
Shashi B. Mahato
Indian Institute of Chemical Biology, A Unit of C.S.I.R. Govt, of India, 4, Raja S.C. Mullick Road, Jadavpur, Calcutta-700-032, India
B. Mikhova
Institut fiir Organische Chemie, Universitat Hannover, Schneiderberg IB, D-3000 Hannover 1, Germany
Luis Moroder
Max-Planck-Institut fiir Biochemie, Am Klopferspitz 18a, D-82152 Martinsried, Germany
Johann Mulzer
Institut fur Organische Chemie der Freien Universitat Takustra^e 3, D14195, Berlin, Germany
O. Muhoz
Universidad de Chile, Facultad de Ciencias Casilla 653 Santiago, Chile
XI
S. Nishibe
Osaka University of Pharmaceutical Sciences, 10-65 Kawai 2-Chonie, Matsubara 580, Japan
Hideaki Oikawa
Department of Bioscience and Chemistry, Hokkaido University, Sapporo 060, Japan
Leo A. Paquette
Department of Chemistry, The Ohio State University, 120 West 18th Avenue, Columbus, OH 43210-1173, U.S.A.
A. Penaloza
Universidad de Chile, Facultad de Ciencias Casilla 653-Santiago, Chile
A. Porzel
Department of Natural Products Chemistry, Institute Biochemistry, Weinberg 3, P.O. Box 250, D-06018 Halle/S.
A.G. Ravelo
C.P.N.O. Antonio Gonzalez, Universidad de La Laguna, Carretera La Esperanza 2, La Laguna-Tenerife Espana.
J. Schmidt
Department of Natural Products Chemistry, Institute Biochemistry, Weinberg 3, P.O. Box 250, D-06018 Halle/S.
of
B. Schneider
Department of Natural Products Chemistry, Institute Biochemistry, Weinberg 3, P.O. Box 250, D-06018 Halle/S.
of Plant
Michele R. StabiV.
Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24062, U.S.A.
John R. Stille
Chemical Process Research and Development Eli Lilly and Company, Indianapolis, Indiana 46285-4813, U.S.A.
Motoo Tori
Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho, Tokushima 770, Japan
H. Toshima
Faculty of Agriculture, Hokkaido University, Kita 9, Nishi 9, KTTA-KU, Sapporo 060, Japan
B. Voigt
Department of Natural Products Chemistry, Institute Biochemistry, Weinberg 3, P.O. Box 250, D-06018 Halle/S.
Yutaka Watanabe
Faculty of Engineering, EHIME University, 3, Bunkyo-cho, Matsuyama 790, Japan
H. Yamaguchi
Osaka University of Pharmaceutical Sciences, 10-65 Kawai 2-Chome, Matsubara 580, Japan
of Plant
Plant
of Plant
This Page Intentionally Left Blank
Xlll
CONTENTS Foreword
v
Preface
vii
Contributors
ix
Strategies for the StereocontroUed De Novo Synthesis of Natural Products L.A. PAQUETTE
3
A Historical Perspective of Morphine Synthesis T. HUDLICKY, G. BUTORA, S.P. FEARNLEY, A.G. GUM AND M.R. STABILE
43
New Developments in the Synthesis of Polyketides and of Chiral Methyl Groups J. MULZER
155
Total Stereoselective Synthesis of Acetogenins of Annonaceae : A New Class of Bioactive Polyketides B.nGADERE AND A. CAVE
193
The Synthesis of Nonactic Acid. Its Derivatives and Nonactin itself I. FLEMING AND S.K. GHOSH
229
Total Synthesis of Bioactive Natural Spiroethers, Tautomycin and Oscillatoxin D A. ICHIHARA, H. OIKAWA AND H. TOSHIMA
269
Aza-Annulation of Enamine Related Substrates with a,p-Unsaturated Carboxylate Derivatives as a Route to the Selective Synthesis of 5-Lactarns and Pyridones J.R. STBLLE AND N.S. BARTA
315
Selective Reactions and Total Synthesis of Inositol Phosphates Y.WATANABE
391
Synthesis of Phytosphingolipids T. KAMIKAWA
457
New Developments in Brassinosteroid Research G. ADAM, A. PORZEL, J. SCHMIDT, B. SCHNEIDER AND B. VOIGT
495
Structure Elucidation and Synthesis of the Lignans from the Seeds of Hemandia M. ARIMOTO, H. YAMAGUCHI AND S. NISHIBE Studies on the Absolute Configuration of Some Liverwort Sesquiterpenoids M. TORI
ovigera L. 551
607
XIV
Bioactive Gymnemic Acids and Congeners from Gymnema sylvestre S.B. MAHATO
649
TM
Theory of the Origin, Function, and Evolution Secondary Metabolites C. CHRISTOPHERSEN
677
The Celastraceae from Latin America Chemistry and Biological Activity O. MUNOZ, A. PENALOZA, A.G. GONZALEZ, A.G. RAVELO, I.L. BAZZOCCHI AND N.L. ALVARENGA
739
Structural Chemistry of Glycolipids from Fungi and Protozoa E.B. BERGTER AND M.H.S.V. BOAS
785
Potential Bioactive Conformations of Hormones of the Gastrin Family L. MORODER AND J. LUTZ
819
When Two Steroids are Better than One : The Dimeric Steroid-Pyrazine Marine Alkaloids A. GANESAN
875
Human IgGl Hinge-Fragment as a Core Structure for Immunogens L. MORODER, G. HUBENER AND M. GEMEINER
907
^^C-NMR Spectroscopy of Coumarins and their Derivatives : A Comprehensive Review B. MIKHOVA AND H. DUDDECK
971
Subject Index
1081
Stereoselective Synthesis
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
Strategies for the Stereocontrolled De Novo Synthesis of Natural Products Leo A. Paquette
LEO A. PAQUETTE In no area of chemistry is stereoselectivity more often a necessary consideration than in the synthesis of structurally complex natural products. A practitioner in this field must be knowledgeable not only of many useful transformations and the mechanistic principles underlying their ability to bring about controlled chemical change, but also be capable of deploying the vast array of available reagents in that chemoselective, regioselective, and stereoselective manner appropriate to the target molecule under consideration. Although the achievements of the last three decades have in the minds of many caused these very important prerequisites to become highly developed, the demands placed on synthetic chemists are hardly exhausted. A recently pubhshed treatise entitled "Stereocontrolled Organic Synthesis" addresses many of the relevant issues from the viewpoint of how the field can expect to develop well into the 21st century [1]. There exists no doubt that the pace of progress has been breathtaking. Certainly, the fantastic advances in NMR spectroscopy and X-ray crystallography have greatly reduced the time needed to determine the structures of newly synthesized compounds. Notwithstanding, effective strategies remain the province of synthetic organic chemists, and it is in this arena where stereochemical elements are deployed with remarkable sophistication. In this chapter, we welcome the opportunity to provide an overview of some of the stereocontrolled syntheses successfully brought to completion in this laboratory in recent years. A. THE LYCOPODIUM ALKALOIDS MAGELLANINE AND MAGELLANINONE In a series of insightful papers, Castillo and MacLean established that the club mosses Lycopodium magellanicum and Lycopodium paniculatum produce alkaloids possessing structural features distinctively different from other metabolites known to arise from these and related sources. The three members of this small and unique subset were identified to share in common a central bicyclo[3.3.0]octane unit to which a functionalized cyclohexane and an Nmethylpiperidine ring were laterally fused. The occurrence of magcllaninc (1) [2],
magellaninone (2) [3], and paniculatine (3) [4] in nature has attracted significant attention [5-7], since all three represent challenging objectives for total synthesis.
HgC-N
H3C-N.
Our successful acquisition of both 1 and 2 gave particular attention to the requirement for strict stereochemical control at six of the eight carbons of the diquinane substructure by retrosynthetic disassembly of the two six-membered rings. The broadly defined goals were therefore to realize proper cyclohexannulation of enone 4 [8] in advance of a tandem vicinal difunctionalization process that would establish the heterocyclic ring. Disconnection of strategic bonds in this manner provided long term for the development of a new MichaelMichael ring-forming sequence as well as a novel means for incorporating the piperidinering[9]. The most expedient means for incorporating ring A involved the K2CO3promoted condensation of 4 with ethyl 5-ethoxy-3-oxo-4-pentenoate in tetrahydrofuran and ethanol containing alumina as a surface catalyst at room temperature. As a consequence of the somewhat folded conformation of 4, the face selectivity of the first conjugate addition proceeds syn to the angular hydrogen as in 5 for obvious steric reasons (Scheme I). Stereocontrol is not sacrificed in proceeding from 5 to 6 because the acceptor side chain is already positioned on the p surface and the diquinane segment possesses a latent thermodynamic preference for becoming cis- and not trans-fused. As a consequence, 7 is obtained in good yield. Following acid-catalyzed elimination of ethanol in 7, it proved possible to reduce the cyclopentanone carbonyl in 8 chemoselectively as expected. Noteworthy at this stage is the fact that borohydride attack occurs stereoselectively from the p face. Silyl protection of the resulting a alcohol afforded 9 and set the stage for unmasking of the second five-membered ring carbonyl. Recourse to thallium nitrate as the means for removing the dithiane moiety gave 10. The advantage of this strategy was that both ketone functional groups in 10 could be simultaneously modified now and at a later stage. Although the reduction of 10 with diisobutylaluminum hydride was not 100% stereocontroUed at -78 °C, the unwanted minor diastereomers could be separated chromatographically and reconverted quantitatively to 10 for recycling
O
o
o
COOEt
EtO
Qv
KgCOa.AlgOa,
>r^ Z^^^COOEt
COOEt
lb
THF. EtOH 25 «C
OH EtO,
COOEt
1. NaBH^, EtOH, CH2CI2,
(TsOH)
QOC
2. TBSOTf. imid, CH2CI2, n
TBSO
^
OH TI(N03)3
1. MOMCI. (/-POgNEt. CH2CI2
(/-Bu)2AIH. CH2CI2. -78 °C
TBSO
^
»
4
PCC/A^Oa, CHgCI^ rt
MeOH, THF
2. B L ^ N ^ F "
OH 11
OMOM
HMPA. 3A MS rt
OMOM
OMOM 1. LiN(SiMe3)2, THF; PhSeCI 2. H202.py
CH2CI2 OMOM 12
OMOM
OMOM 13
14
Scheme I purposes. This simple tactic raised the efficiency with which 11 was produced to the 76% level and permitted its ready conversion via 12 to 13. It is significant in the context of what is to follow that hydride delivery to both carbonyl groups in 10 once again operates with a dominant p-face kinetic preference. Once 13 was in hand, enone 14 was generated through adaptation of conventional organoselenium technology for the purpose of incorporating the piperidine ring properly. The recognized propensity of the anion of (trimethylsilyl)acetonitrile to exhibit 1,4-addition to conjugated enones [10] was applied to 14. To our satisfaction, the diastereofacial guidance available to this reagent was identical to that provided to the reducing agents utilized earlier. Furthermore, the enolate intermediate thus formed proved entirely amenable to stereoselective C-acylation with methyl cyanoformate [11] and fumished 15 in a single laboratory operation (Scheme II). As a direct consequence of the relatively high acidity of the proton
OMOM
/
1. LICH(CN)SIMe3 HMPA, THF
^.
2. KF.aqCHgCN, OMOM 3. LDA, NCCOOMe
1. NaBH4, MeOH, -20 "C
MeOOC
2. COC^.py, THF; PhSeH
I
MeOOC. -^
OMOM 16
15
14
OMOM
/
O It PhSeCO
OMOM
(MegSij^SIH. AIBN'. CBHC A
MeOOC \ OMOM
NaBH4. CoClg, MeOH; *KOH, MeOH; HgO*
/ O,
17
Scheme II positioned central to the p-keto ester subunit of 15, enolization is facile. It is therefore not known whether the a orientation of the carbomethoxy substituent is the result of kinetic or thermodynamic control. Suffice it to indicate, however, that this stereogenic center has been improperly set and requires subsequent inversion. Since utilization of the ketone carbonyl was now complete, its removal was implemented via an efficient three-step sequence involving reductive cleavage of the derived selenocarbonate with tris(trimethylsilyl)silane [12] under free radical conditions [13]. With the acquisition of 17 in this manner, the serviceability of the reagent produced by adding sodium borohydride to cobaltous chloride for chemoselective reduction of the nitrile group [14] was assessed. Indeed, treatment of 17 in this manner, followed directly by basification with potassium hydroxide in methanol, secured 18. In this step as well as in the subsequent progression to the N-methyl derivative 19, no epimerization was seen within ring A. To our mind, the enolate of 19 should exhibit a decided kinetic bias for kinetically controlled protonation on its a face because of the steric encumbrance associated with p proton delivery. In actual fact, rapid introduction of its lithium salt into a 1:4 mixture of water and tetrahydrofuran at -78 °C resulted in its quantitative conversion to 20 (Scheme HI). Once the MOM groups had been removed, controlled oxidation with manganese dioxide led to 21, a very pivotal intermediate. To arrive at magellaninone (2), 21 was treated with methyllithium and the resulting unprotected diol 22 was directly reduced with lithium aluminum hydride. Subsequent Jones oxidation proceeded with the customary allylic rearrangement. The plan now called for producing mageUanine (1) by standard borohydride reduction of 2. However, in contrast to the directionality observed earlier for a
carbonyl group in this locale, only the p alcohol 23 was obtained perhaps because of the presence of the fused piperidine ring on the convex surface. In any event, Mitsunobu inversion [15] was successful in delivering the targeted alkaloid and in demonstrating that these unusual Lycopodium alkaloids can indeed be prepared in stereocontrolled fashion by three-fold annulation of 2-cyclopentenone. OMOM
0
/ {
\ \ ^^^ H 20 HaC,
1. UAIH4. THF. A
1\. » •^ H
OMOM
1. HOI. HgO.THF 2. MnOg, CHCb
0^ J H^--
\
if
\ ^•^*H
OH
21
-OH
CH3M
THF ^ > -78 °C
)
2. Jones oxid.
NaBH4 EtOH
OH
22
1. PhgP. DEAD HCOOH, THF 2. 10%KOH, H2O
H3C' 23
Scheme HI B. THE MOST HIGHLY CONDENSED PENTALENOLACTONE ANTIBIOTIC Ecological concerns have prompted chemists to become increasingly "atomeconomic" in their synthetic pathways. The goals associated with this concept are near-perfectly realized in the course of efficient isomerization reactions. Accordingly, we have incorporated a number of stereocontrolled rearrangements into our synthetic undertakings. Illustrated here is proper application of the oxadiTC-methane rearrangement to a total synthesis of pentalenolactone P methyl ester (24b) [16], the stable esterified form of naturally occurring 24a. Pentalenolactone P is the only member of the pentalenone family of antibiotics to possess a fused three-membered ring, which notably resides on the highly congested concave luifaeo or me moidouio [17].
24a.R - H b. R - C h ^
The central stereochemical issue in any projected synthesis of 24b is the establishment of a trans relationship between the cyclopropane and lactone rings. This being the case, we set out to develop a convenient route to the (J^y-unsaturated ketone 34 in advance of its triplet-state photoisomerization, which was projected [18] to generate the tetracyclooctanone 35 (Scheme IV). The Diels-Alder reactions of 1-methylcycloheptatriene (25) [19] with fumaroyl chloride followed by indirect hydrolysis was capable of producing large amounts of the dicarboxyUc acid 26. Necessary chemoselective differentiation of the functional groups in 26 was made feasible by oxymercuration. By this means, the role exercised by the methyl substituent on the steric course of the [4+2] cycloaddition was capitalized upon to considerable advantage. Moreover, the strained nature of lactone 27 allowed for smooth conversion to diol 28 by reduction with sodium borohydride, thereby effectively accomplishing suitable oxygenation of the proximal carbon of the original etheno bridge in 26. Buoyed by the ease of this oxygen atom transfer, we proceeded to generate the acetonide 29 and to advance the synthesis by implementing conversion to a,p-unsaturated ester 30 through deployment of oxidative elimination involving the a-phenylseleno derivative. Evidently, the significant strain introduced upon installation of the double bond accelerates acetal hydrolysis. Attention was next directed to regioselective chain extension and this maneuver was accomplished by sequential exhaustive silylation, reduction with diisobutylaluminum hydride, perruthenate oxidation to the aldehyde, and Wittig olefination. Once the conjugated diene 32 had been produced, it proved an easy matter to effect its conversion to 33 by regiocontroUed hydroboration and selective pivaloylation of die primary hydroxyl groups. Perruthenate oxidation of 33 efficiently delivered 34 whose irradiation in acetone solution with 3000 A light proceeded with full retention of stereochemistry to introduce a second cyclopropane ring as in 35. The stmctural assignment to 35, initially deduced on spectroscopic groups, was corroborated by X-ray analysis of the highly crystalline diol 36 produced by saponification.
a
CICO CHa
' •
\
^ ^COCI toluene, A
1. Hg(OAc)2. MeOH. rt;
COOH
2. CH3OH. py 3. NaOH, MeOH, H2O
NaBH^. -78 °C 2. CH2N2
COOH
25
COOMe
»>
26
NaBH4
^I^^X^COOMe
^
MeOH. rt
OH
COOMG
TsOH, THF
l^
I
OH
29
1. TBSCI.
V^OH
YT
2. Dibal4 CH2CI2
1. TRAP. N M O
Y\
OTBS OTBS
30
1. 9-BBN; NaBOg 2. PvCI. EtgN
OTBS 31
32
OPv
OPv (n-Pr)4NRu04,
3000 A
^
^>
3. 48% HF, CH3CN; PvCI, Et^. DMAP
^.
2. MCPBA NaHCOg. CH2CI2
28
J\
1.LDA,THF; PhSeBr
(CH3)2C(OMe)2
OPv OH
NMO, 4A MS,
33
acetone
34
OH
PvO
\
OPv
^
OPv NaOH H2O, Eton
OH
p^^^'^^o 36
35
Scheme IV The developments described above were predicated upon the expectation based on less highly substituted examples that the dissolving metal reduction of 35 would likewise result in regioselective rupture of the central bond of the threemembered ring conjugated to the carbonyl. No guidance was available to insure that the second cyclopropane would be insulated from electron transfer chemistry or that the a-pivaloyloxymethyl group would survive intact. Once the experiment
10
was carried out, it was made clear that the stereoelectronic factors operative in 35 were adequate to limit reduction to the dihydro level. Of the two products fomied, 37 was produced to a somewhat greater extent than 38 in ratios varying from 2:1 to 1:1 (Scheme V). hi this setting, it was opportune to acetylate the mixture and to effect P-elimination within esterified 37 to give 39. Careful saponification of this intermediate produced 38 in a high state of purity. If 38 was left too long under these alkaline conditions or a stronger base was employed, intramolecular Michael addition to the exomethylene ketone occurred prematurely. In order to craft the lactone ring, 38 was oxidized to 40 under Swem conditions in a prelude to intramolecular 1,4-addition of the hemiacetal anion [20] formed via nucleophihc attack by methoxide ion at the aldehyde site. With the availability of acetal 41, it became necessary to consider carefully whether to elaborate the epoxy lactone segment in advance of, or subsequent to, introduction of the a,p-unsaturated ester subunit. Since the latter option was considered more workable, 41 was transformed into the enol triflate and subjected to palladium(n) catalyzed methoxycarbonylation [21]. This methodology allowed for proper homologation of 42 to 43, and subsequent conversion to 44, in totally regiocontrolled fashion. The sector where theremainingcarbon atom needed to be introduced in 44 proved to be so sterically crowded that a number of standard methods for achieving lactone a-methylenation fared very poorly or, more often, worked not at all. Following these probe experiments, we found it possible to engage the neopentyl carbon in the capture of monomeric formaldehyde [22] as electrophile. The 10:1 mixture of epimeric hydroxymethyl products was directly dehydrated via the mesylates to deliver 46. The final oxidation could be effected either directly with m-chloroperbenzoic acid or by way of a three-step sequence involving DibalH, /-BuOOH with V0(acac)2, and TPAP with NMO [23]. Thus, 32 steps were required to reach pentalenolactone P methyl ester. The relative stereochemical relationship of its cyclopropane and lactone rings was immediately secured by Diels-Alder cycloaddition and maintained during the photoisomerization and reductive cleavage steps that followed. C. (+).IKARUGAMYCIN, AN UNUSUAL MACROCYCLIC TETRAMIC ACID ANTIBIOTIC As early as 1972, the culture broths of Streptomyces phaeochromogenes wererecognizedto be capable of producing a powerful and specific antiprotozoal and antiamoebic agent [24]. This dextrorotatory substance was determined to be the architecturally uncommon macrocyclic compound 47 and called ikarugamycin. The incorporation within 47 of a trans,anti.cis-AtcdiiyAxO'aS' indacene subunit, a largeringlactam, and an enoyltetramic acid prompted us [25] and others [26-28] to undertake its constmction in the laboratory.
-^sb^, ^
^^^ 37
35
ACgO, EtgN DMAP, CH2CI2
MeOH, H2O
Swern
OAc
40 NaOMe. MeOH
1. Pd(0Ac)2, PhgP, EtgN
LDA;
COatm MeOH, OMF 2. CH2N2
COOMe
OTf
PhNTfg THF
43
41 1. 10%HCI, THF 2. (n-Pr)4NRu04. NMO. 4A MS CH2CI2 HO
O 1. CH3SO2CI, EtgN, CHgCb
LDA. THF; COOMe
CH2O
^.
COOMe
2. DBU.CeHe
COOMe 46
MCPBA CH2CI2.A
Scheme V We saw in 47 an opportunity to deploy a triply convergent and enantioselective strategy. The challenge of obtaining the western half of the molecule, which was addressed first, was met with a concise route to racemic tricyclic hydroxy ketone 56 in six short steps [29] from readily available 48 [30]
12
(Scheme VI). The desirability of producing a major segment of the target molecule in racemic condition may appear illogical and is therefore deserving of comment. In brief, we were highly attracted to the possible deployment of a subsequent kinetic resolution of 56 by suitable application of Koga's chiral a,punsaturated aldimine methodology [31]. The superb success realized in the course of this adaptation is presented subsequently. CHgOv^OCHa
CH3OV.0CH3
48
49 KH. THF 25 "C; H2O (immed. / work-up)/^
H Q C H
H H
CHgC^^GCHg
CeCl2
3 OCH3 ^
*^*^« ''"^^' H2O (30 mInV
HQCH
KgCCb
I
• CH3OH
" OCH3
H H
1. Diba)-H. CH2CI2 ^--=-* 2. 3N HOI, ether
HO
^
53
52
& H H 54
HO
HO. KgCOs CH3OH
y^\^
NH3
A V ^ 56
55
Scheme VI An appreciation of the ability of 48 to attain appreciable levels of double diastereoselection when reacted with chiral (racemic) vinyl organocerium reagents had earlier been gained in this laboratory [32]. Consequently, it occasioned no surprise to observe that 49 [33] adds to this bicyclic ketone with customary endo stereoselectivity to deliver 50 and 51 in a relative ratio of 92:8. The major product, easily purified by chromatographic means, was smoothly isomerized to 52 under anionic conditions at room temperature. For structural reasons, this sigmatropic change is required to proceed via a boat-like transition state. The all-
13
cis tricyclic isomer must therefore be formed. However, if the quenched reaction mixture is left at 20 °C for 30 min, the basic environment promotes wholesale epimerization to 53. Consequently, only two steps need to be expended for stereocontroUed elaboration of the targeted framework having four stereogenic centers properly set in trans A/B-locked fashion. In order to invert the stereochemistry of those two carbon atoms that unite rings B and C, the ketone carbonyl was reduced and deketalization effected to give 54. Double bond migration to the intracyclic site in enone 55 and dissolving metal reduction completed the conversion to 56. The strongly acidic character of tetramic acids and their usual low solubility prompted us to delay the assembly of this heterocychc unit until very late in the synthesis. Accordingly, the appropriate ornithine segment was constructed next (Scheme VII). The known amino acid 57 [26b] was transformed via the fully protected derivative 58 to 59 by chemoselective unmasking of the y-amino group with formic acid. The remaining two substituents on the a-amino group are to be removed at different times, with the allyl carbamate destined to precede the 2,4dimethoxybenzyl functionality. 1. ArCHO. NaBHaCN. MeOH
2- 0 1 - ^ 0 ' 57 1. HCOOH. 10«C,3h
3. CH2N2
^3^" ^ ^ f
2. HOAc
Ar
- — ^ / ^ OMe OMe
59
Scheme VII The time had now arrived to append properly to 56 those sidearm substituents needed for elaboration of the macrocyclic ring. Rapid advance was realized when the silyl protected derivative 60 was formylated and O-aUcylated in situ to produce 61. Hydride reduction and acidic hydrolysis of this intermediate made available the a,p-unsaturated aldehyde 62 needed for evaluation of the potential usefulness of Koga's chemistry (Scheme Vni). Condensation of 62 with enantiopure L-rerr-leucine rerr-butyl ester led to an inseparable 1:1 mixture of the diastereomeric aldimines 63 and 64. Our expectations regarding the subsequent addition of vinylmagnesium bromide to this mixture were based on the recognized bidendate chelating ability of divalent magnesium to fix the nitrogen and oxygen atoms in a manner which significantly enhances conformational rigidity
14 OTBDMS %, P
OTBDMS
OTBDMS
^ ^^,,^,,, , 1 . KN(SIMe3)2. THF; HCOgEt
(/>Bu)2AIH.
2. (CH3)2CHI, 62
61
60 C02^Bu
b
CHO
HP*
HMPA
H
OTBDMS
OTBDMS
/ N
^
(HOAc). MgS04
^N>^C02^Bu r^f-Bu H H
Nv^COg^Bu H
H
^f-Bu 64
63 1.CH2=CHMgBr, THF
1. CH2=CHMgBr, THF
2. H3O*
2. H P *
OTBDMS
OTBDMS
H.>-^y=.
Br
Mg: 1
^Bu
.3^
T f-Bu
66
65
OTBDMS
OTBDMS
H. V - ^ H
0 67
(87:13)
68
Scheme Vin [31,34]. The relevant complexes are depicted as 65 and 66. The 1,4-addition in 65 is consequently relegated to the less sterically congested jc-face and should occur without complication. In contrast, 66 is the "mismatched" diastereomer lacking the ability to deliver the vinyl nucleophile well from the much more crowded concave direction. This competing process is kinetically disadvantaged to an extent such that the ratio of 67 to 68 obtained after citric acid quench is 87:13. When proper allowance is made for the quantity of unreacted 62 recovered, the efficiency of the vinylation was determined to be 48%. The enantiomeric purity of 67 was defined by chemical conversion to 69 and Mosher ester analysis to be 91% ee (Scheme IX). Three recrystaUizations of 67 provided enantiopure material. All eight of the stereogenic centers present in the westem sector of ikarugamycin had now been set in their proper absolute configuration.
15 OTBDMS
1. HC(OMe)3, (TsOH)
.
OTBDMS
1. PCC. NaOAc, CHgClg
^»
2. Disiamyiborane, THF; HgOij. NaOH, HjO
2. CBr4. PPhg,' CH^Ig.py
OMe
OTBDMS I
J^COg. MeOH. HgO;
-^r-COOMe
70
OMe 71 OTBDMS I
-^=-CONH
COOMe
r
2,4,6-(CH3)3PhS02CI, THF:DMAP.59
OMe
OMe 1. (TsOH). acetone 2. KN{SiMe3)2,
r
OMe
o „V„ 72
{EtO)2h 2P
\A^p 73
OTBDMS PdCPPhg)^.
-CONH
"""i^^-ys^^
THF
0.^0 74 OTBDMS -CONH-
r\^-~.^y^ °x.°
OMe
75
OMe
Scheme IX With a bountiful supply of 69 at our disposal, the synthesis was continued by PCC oxidation to the aldehyde level and application of the Corey-Fuchs procedure [35] for chain homologation via dibromo olefin 70 to the acetylenic ester 71. Since amide bond construction next had to be implemented, this ester was saponified under mild conditions and the resulting carboxylic acid was activated by formation of a mixed anhydride with mesitylenesulfonyl chloride in advance of in situ condensation with 59. In order to preclude hydrolysis of the silyl ether functionality in 72, deacetalization had to be performed under anhydrous conditions in dry acetone containing a catalytic quantity of p-toluene-
16
sulfonic acid. This maneuver enabled condensation of the aldehyde so formed with phosphonate 73 [36] without encountering any detectable epimerization. The functional group array in 74 lent itself quite satisfactorily to chemoselective cleavage of the allyl carbamate residue by means of (tetrakistriphenylphosphine)palladium(O) [37], provided that acetic acid was present to inactivate the nucleophilic character of the liberated amine. The time had now arrived to effect the crucial macrocyclization. From the background experience gained by others [38], it was anticipated that the ketene 76 liberated by heating 75 in toluene for 4 h would be appropriately electrophihc. In addition, the extensive representation of diagonal and trigonal centers in 76 was expected to facihtate the desired intramolecular trapping. Indeed, the ring closure OTBDMS toluene 110*0 4h
OTBDMS
OTBDMS
CONH-
O
O
77 O^^NH1. Hg. 5% Pd-BaS04. quinoline
MeOGNSOgNEta.
2. 48% HF, CHgN O
O
COOMe
78 KOf-Bu (1 equjv)
TiuOH*^ O
O
COOMe
79
80
Q^NH CFgCOOH
/"~\
65 °G, lOmIn
-V^^ q J ^
^.
s^^
/ NH /
1
/ 1 1
H
Ar
.—^^OMe OMe
47
1
Scheme X
COOMe
17
proceeded smoothly to deliver 77 with 94% efficiency (Scheme X). Successive semisaturation of the acetylenic bond by means of the Lindlar method, desilylation to liberate to hydroxyl group, and dehydration of alcohol 78 with the Burgess reagent [39] led most satisfactorily to introduction of the B ring double bond. Arrival at ikarugamycin from this vantage point was predicated upon the successful Dieckmann cyclization of 79. As a result of our awareness of the disastrous potential for base-promoted racemization of the proximal stereogenic center, 79 was treated with only one equivalent of potassium tert-huioxide and reaction was allowed to proceed for only 10 min at room temperature. These conditions provided enantiomerically homogeneous 80 in 66% yield. The major complication of the entire synthesis materiaUzed during subsequent removal of the 2,4-dimethoxybenzyl protecting group. After an exhaustive experimental search for proper conditions, it was recognized that heating 80 in trifluroacetic acid at exactly 62 °C and for precisely 10 min was uniquely effective in delivering 47. This successful enantioselective route to ikarugamycin demonstrates the latent capability of the anionic oxy-Cope rearrangement for highly dependable chirality transfer [40] and the potential for absolute stereochemical control offered by Koga's 1,4-asymmetric conjugate addition process. D. A REPRESENTATIVE FURANOSESQUITERPENE: (+)-PALLESCENSINA Nature has found it possible to assemble a wide range of furanosesqui- and diterpenes. Although it is quite clear that these substances are not biosynthesized via any sigmatropic scheme, the atom economy of such isomerization reactions appeared to us to warrant appHcation to this field. A thrust in this direction would require, however, that a furan ring be willing to utilize its n electrons in a manner suitable to rebonding. Precedent for an adaptation of this type was scarce [41]. Nonetheless, we have succeeded in developing a relatively concise enantioselective synthesis of natural (+)-pallescensin A (81), a marine metabolite first isolated in 1975 [42] and prepared earlier on several occasions [43-48].
81
Retrosynthetic considerations suggested that the obvious inducement for us was the opportunity to transform the known optically pure ketone 82 [49] into 83 in advance of an anionic Cope rearrangement (Scheme XI) [50]. Although 1,2addition of the cerate prepared from 3-furyllithium proceeded with appropriately high facial selectivity, subsequent isomerization of the potassium salt of 83
18
'\y
^ c „ .
Ho.^i>
CeCb.THF -78 "C -* 0 X
KH 18-cr-6, diglyme, 100 "C
^
84
83
82
CH(OMe)2
CHO
V^o
MeOH, A
-O
85
LDA,
CH(OMe)2
NaHCO^ MeOH, H2O
/ K ^ O 87
86 O
CH(OMe)2
f-BuOOH
BF3'OEt2
LiAJHd.
NaOH. MeOH, A
CH2Cl2,25°C
AICI3, Et,0
••(p:.
89
88
90
H2. Pd-C. EtOAc. EtOH. Et2NH 91 \
H2. Pd-C,
X
EtOAc, EtOH, EtgNH
Scheme XI required elevated temperatures (100 °C) even when 18-crown-6 was present. Under these circumstances, the generation of enolate anion 84 was met with ensuing p-elimination of the alkoxide ion to give 85. This retro-Michael reaction is obviously facilitated by the resonance stabilization available to the leaving group. This development set the stage for chemoselective acetalization by heating 85 with ammonium chloride in methanol. Once 86 had been produced, it was possible to introduce further unsaturation as in 87, whose lone stereogenic center was to be the linchpin for establishing the proper absolute configuration of pallescensin A. In fact, the angular methyl group served as a stereocontrol element particularly well suited to introduction of the necessary trans ring fusion. Prior to that, the furan ring was concisely reconstructed by regioselective epoxidation of 87 to give 88 followed by exposure of this oxygenated intermediate to boron trifluoride etherate at room temperature [51]. In the presence of alkaline
19
tert'hutyl hydroperoxide, 87 experiences remarkably face-selective nucleophilic attack from the a direction at the more highly substituted enone double bond. These very accommodating steps were followed by reductive removal of the carbonyl group in 89 with alane [52]. NMR studies on 90 involving the use of Eu(dcm)3 as chiral shift reagent showed this advanced intermediate to be of 100% enantiomeric purity. Catalytic reduction of 90, necessarily performed in the presence of diethylamine to guard against the destructive effect of acid buildup, led via 91 to the targeted furanosesquiterpenoid. E. (-)-VULGAROLIDE, A HIGHLY OXYGENATED POLYCYCLIC METHYLENE LACTONE Our interest in developing the anionic oxy-Cope rearrangement into a powerful tool for the elaboration of structurally intricate natural products in a stereocontroUed manner has recently been successfully applied to the total synthesis of natural (-)-vulgarolide (92) [53]. Li addition to its highly rearranged isoprenoid framework, 92 features a central cyclooctanone ring to which tetrahydrofuran and y-lactone subunits are serially fused in trans-anti-trans fashion across the C-3 to C-6 positions [54]. The key elements of the stratagem designed to realize such twofold distal annulation involved initial addition of vinylmagnesium bromide to (+)-93 (100% ee), charge-accelerated [3,3] sigmatropy of the potassium salt of 94, and direct treatment of the enolate anion
produced regiospecifically witti ethyl iodoacetate (Scheme XII). Of particular relevance at this point was the fact that the new C-C bond in 95 had been installed from the p-surface. The next objective was to form lactone 96. Steric approach control operates during hydride reduction with the result that the configuration of the two cyclooctyl C-0 bonds are exactly opposite to those defined in the target. This feature was purposefully designed into the synthesis in anticipation that vulgarolide would be reached more concisely by double inversion. In fact, once the exo-methylene group had been introduced and the ahydroxyl substituent unmasked as in 97, the generation of a leaving group at the latter site and lactone hydrolysis was met by formation of oxirane 98. The bridgehead double bond was now selectively ozonolyzed. Spontaneous
20
cyclization occurred in a spectacularly facile manner to deliver vulgarolide (92) and its anomer 99 in a 1:1 ratio. Both hemiacetals converged to Omethylvulgarolide (100) during methylation, a maneuver that facilitated purification of the highly insoluble 92. Hydrolysis of 100 in turn produced predominantly vulgarolide.
^f-k-^O
CHg-CHMgBr,
4Y
THF OSEM
.78«C-4ft
^f-^^^ OH OSEM OSEI
• 4S^o
1. KN(SiMe3)3. THF, A 2. ICHjCOOEt,
94
93
1. UAIH4
KX
2. TRAP, CHsCfe
j ^
\EUO SEMO
O—^ 96
98
SEMO
O
COgEt
95
1. LDA, CH2O THF,-78 to-25 X
1. MsCI, EtaN, DMAP
2. MsCI. EtgN, DMAP 3. DBU.CeHe 4. 5%HF, CH3CN
2. LiOH, CH3OH
a
97 A&O, GH3I CH2CI2
O3. CH2CI2;
HO
HMPA -78 "C -^ ft
10%HCI, THF HO
MeO 92,p-OH 99.a-OH
100
Scheme XII F. (+)-CEROPLASTOL I, A DICYCLOPENTA[a,d]CYCLOOCTANE SESTERTERPENE The Claisen rearrangement, a heteroatomic variant of the Cope process, holds equal appeal as a scaffolding element that is totally atom-efficient. We have addressed and defined those stereocontrol elements associated with a two-carbon intercalation tactic [55] in several contexts as, for example, in the preparation of (+)-ceroplastolI(101)[56].
21
To this end, it was opportune in light of background information to prepare 105 by sequential epoxidation of 102 [57], heating of the epoxy ketone with sodium methoxide in methanol containing a small amount of water [58], and Shapiro degradation of the ketone [59] in advance of acidic hydrolysis (Scheme Xm). Subsequently, 105 was oxidized with MPCBA to the ring-expanded epoxy lactone, heating of which at 175-180 °C in benzene solution (sealed tube) dehvered 106. FoUowing Wittig olefmation to give 107, a second carbon atom OCH3 1. NaBH^ NaOCH3. 2. MCPBA 3. PDC
H g C o ^
HgCO^
102
103
1. TSNHNH2. CH3OH, rt
l
\
H,CO \ H3U O ^ 107
104
1. MCPBA. NaHCO^CHzClg.A
2. CHaLi.THF, EtgO; NH4CI, H2O 3. HOAc, H2O.A
Y
CH3OH, (HgO)
r...o
HaCO^ 105
1. Cp2TI(CI)(CH2)AI(CH3)2 THF.(py) 2. 200 oQ
Ph3P=CH2
2. 175-180 "C. GeHg. sealed tube 106
K2CO3.
^- \ _ / f Y ' 0
CH^H, A
H3C
O ^
109
was introduced by means of the Tebbe reagent [60]. Heating this product at 200 °C in sealed, KOH-coated glass tubes resulted in conversion to the cyclooctenone 108, which was easily epimerized to the thermodynamically more favorable trans fused isomer 109. With this bicyclic intermediate available in sizeable amounts, ready advance to 111 could be conveniently accomplished prior to annulation of the second fivemembered ring (Scheme XIV). 1,3-Carbonyl transposition was realized by complete eradication of the original carbonyl by Ireland's method [60] followed by ally lie oxidation. Application of the Piers cyclopentannulation protocol [61] to 111 made 113 conveniently available. Introduction of a methyl group into ring B was brought about by treatment of the kinetically derived enol triflate [62] with lithium dimethylcuprate [63]. Hydrolysis of 114 gave the dienone, which was directly transformed into 115 by oxidation of its silyl enol ether with palladium acetate in acetonitrile [64].
22
Completion of the synthesis involved some adaptation of Boeckman's original route to 101 [65]. Introduction of the sidechain was accomphshed by copper-catalyzed conjugate addition of Grignard reagent 116 to 115. Nucleophilic attack occurred exclusively from the p-face with formation of a 3:2 mixture of 117 and its diastereomer. Once chromatographic separation had been accomplished, the carbonyl group in 117 was reduced by the action of sodium borohydride and zinc chloride on the tosylhydrazone [66]. Desilylation occurred during this step to deliver ceroplastol I (101) in a global overall yield of 0.13%. 1. LiAIH^ 2. n-BuU, HaCO^
CIP(0)(NMe2)2 3. Li, EtNHg, f-BuOH, ether
V-/f/c HaC
1. Se02,KH2P04, toluene, A 2. PDC
O ^
*"
V^T'O 111
110
109
1. KN(SiMe3)2, PhN(Tf)2 THF. -78 "C
C ' ~ \ /
HgC O ^
THF
H^O
A ^^
113
112
1. (TsOH), acetone, H2O Jif
7'"0 HgC O 1 ^^^ ^^^
MgBr
CH3 116
^ 2. UN(SiMe3)2, MegSiCI 3. Pd(0Ac)2, CH3CN
^"^^^
2. (CH3)2CuLi. THF. -20 ''C
115
1. TsNHNH, (COOH)2, Eton
^.
CuBr*Me2S HMPA, THF, M©3SiCI, -78 °C
H j c X ^ '^'^''Y^OHI CH3 1 »3C^y—\J
^
2. NaBHsCN, ZnClg, CHgOH. 90 °C 117
UjyHaC 101
Scheme XIV G. THE MARINE TOXIN (+)-ACETOXYCRENULIDE The crenulatan diterpenes, now believed to arise in the marine environment by solar-induced photoisomerization of dictyolactones [66], have been isolated from small brown seaweeds and the sea hares that feed on them. In light of the high survival rate of these species, the crenulatans were investigated and found to function as defensive agents [67,68]. The most bountiful of these toxins appears to be acetoxycrenulide (118), which features a central eight-membered ring and
23
fused cyclopropane and butenolide (or equivalent) subunits characteristic of this class [69-71]. H A H CeHil^^ OAc 118
In our initial studies aimed at the realization o^ an enantioselective, stereocontrolled synthesis of 118, the end game was to attach the methylheptenyl sidechain to position 3 quite late in the reaction sequence [72]. We saw in this undertaking an opportunity to again utilize the Claisen rearrangement as the
LDA
Vj^^-^
x'^VCHO 120
119
(1:1)
121
N-PSP. (TsOH), CH2CI2
H
1. Nal04. NaHCOg. MeOH, H2O
SePh H (1:1)
\
2. EtgNH. mesitylene, A
SePh
^-tS^ H
HC(0Me)3, (TsOH)^
2. {0^^)^r^, \
CH^2.CeHe
125
H
'^M H " O
123
124
.OH, A V ^ ^ - ^
O.
O ^ ' " ^ ] ' ^
W V - - A ^
^
1. KN(SiMe3)2. PhSeCI r^"^-'^-
2. Nal04. NaHCC^. MeOH. H^O
126
127 H A H
5% HOI
128
Scheme XV
^ ^
24
means for elaborating the cyclooctane core. To this end, the well known lactone 119 [73-75] was engaged in aldol condensation with crotonaldehyde and intramolecular selenonium ion-promoted cyclization with participation by the neighboring hydroxyl group (Scheme XV). As illustrated, the aldol process is fully trans-selective, and provides an easily separable 1:1 mixture of 120 and 121. By making recourse to N-(phenylseleno)phthalimide [76], we were successful in transforming 120 into 122 and 123, both of which underwent elimination via the derived selenoxides to introduce an enol ether double bond exocyclic to the pyran ring. This transient species entered into the Claisen rearrangement exclusively via its chair conformation 124, whose adoption guaranteed not only the location of the intracyclic double bond but, most importantly, the absolute configuration of the carbon atom carrying the methyl substituent. The key intermediate 125 so produced was then ketalized and subjected to Simmons-Smith cyclopropanation [77]. The three-membered ring is introduced smoothly from the sterically less congested n surface to deliver 126. With the northern sector now completely constructed, the butenolide double bond was introduced by organoselenium technology and ketal hydrolysis implemented. Under these conditions, 128 was formed without any evidence of double bond migration. The rigid conformation adopted by 128, corroborated by X-ray crystallography, proved inimical to fmitful enolization at the "doubly activated" site in order to incorporate the Cg sidechain, thereby requiring that this unit be present from the outset. This realization led us to probe the consequences of incorporating this rather bulky substituent at an early stage instead [78]. Several issues were considered to hold relevance: (a) would the Claisen rearrangement such as that defined by 124 continue to prove serviceable in providing the means for delivering stereodefined 4-cyclooctenones? (b) would the ponderal effect of the Cg sidechain impact negatively on utilization of the chair-like arrangement in light of the fact that the configuration at C-3 would require this moiety to be axially disposed? (c) would the need to fix C-3 stereochemistry first be a deterrent to proper installation of the remaining stereogenic centers? To answer some of these questions, (5)-citroneUic acid (129), whose methyl-substituted carbon is enantiomeric to that in 118, was transformed via oxazolidinone 130 to the hydroxy acetate 132 (Scheme XVI) using quite standard reactions. When access was subsequently gained to lactone 134, it is made clear that the chiral auxiliary was deployed to set the absolute configuration of the stereogenic ring carbon properly [79]. At this stage, introduction of a hydroxymethyl substituent was undertaken. The conversion to 136, mediated by 135 and involving 1,4-addition of iPrOMe2SiCH2MgCl in the presence of copper(I) iodide and chlorotrimethylsilane [80], proved quite superior to the altematives which were examined. Following oxidative desilylation, it proved an easy matter to convert 137 into the thermody-
25 /-O OH O
1.(CCX)I)2 2. n-BuLi,
LDA. ak.
HN^O.THF 11 O
129
^
1. LIAIH4
OAc
m
Br,
'^^
THF 130
131 OH 1. CrOg. H2SO4, acetone
9-BBN.THF; NaB03-4I^O.
2. AcCI, py. CH2CI2
2. KOH.THF 3. TsOH.CeHe
HP
132
u
133
V I ^O-SiCHgMgCI,
1. KN(SiMe3)2; N-PSP, THF 2. H2O2. py, CH2CI2
Cul, MeaSiCI. THF
134
136 0/a= 85:15)
135 TsOH,
KHF2^ H2O2, DMF 137
138
Scheme XVI namically favored lactone 138. In order to safeguard the structural integrity of this pivotal intermediate, its oxidation with PDC to aldehyde 139 was effected as soon as possible (Scheme XVII). Although a variety of attempts to achieve chemoselective 1,2-addition to the aldehyde carbonyl in 139 proved troublesome because of steric shielding, these difficulties could be circumvented by the introduction of (phenylseleno)methyllithium under high dilution conditions [81]. As revealed by the product structure 140, the generation of an alkoxide ion in this manner was followed by intramolecular attack at the lactonic center. Fortunately, reconversion to the ylactone could once again be easily realized by acid-catalyzed isomerization. Protection of the hydroxyl group made it possible to effect aldol condensation with crotonaldehyde and subsequent ring closure to afford the bicyclic selenolactones 142 and 143. The selenoxide derived from 143 underwent both 1,2-elimination and Claisen rearrangement when heated in mesitylene containing
26 PDC. _4A_MS^ CHgCl/
PhSeCHjjU,
HO,^>v.Js^SePh
THF 140
1. TsOH, CeHe 2. = < ° ^ " 3 CH3. (POCI3)
A?"-'
1. LDA,
^^^^^
2. (TsOH). CeHg.A
142
141
SePh
1. Nal04, NaHCO»
°tCLsePh
9.
H
H2O. MeOH 2. EtgNH. mesityiene, A
143
145
Scheme XVH diethylamine to give 145 in 76% overall yield. Consequently, the associated [3,3] sigmatropic change unquestionably proceeds via the chair arrangement shown in 144. Furthermore, no adventitious epimerization operates and the p,'y-unsaturated double bond does not migrate into conjugation. The findings detailed in Scheme XVII provided important guidance and insight into the requirements necessary to the actual adaptation of this pathway for the production of 118. It is obvious that (/?)-citronellol must serve as the building block for the sidechain. Beyond that, however, it is not just a matter of producing the RJi'isomcT of 132, for this course of action will ultimately provide only the enantiomer of the target. This is because the remaining stereocenters are introduced under the fuU control of those present in this acetate. As a consequence, this intrinsic bias must be overridden by involving a chiral auxihary that is capable of properly establishing stereogenic centers in an absolute sense totally independent of those preexisting in the substrate. The successful realization of these objectives is outlined in Scheme XVIH. Since probe experiments disclosed convincingly to us that the double bond in the intact sidechain of acetoxycrenulide is more reactive to Simmons-Smith cyclopropanation than cyclooctenyl double bonds, the decision was made to introduce the isopropenyl group in the final stages of the synthesis. Accordingly, ester 146, which is easily produced from (/?)-citronellol [81], was transformed into
27 Br
1. LDA, RO'^'^""-^'"V"^COOMe
KN(SiMe3)2.
2. 0 » CH2CI2; Pfyp 3. NaBH^.MeOH
146
I o RO
b
k^-N' V ^
149
HzC^. py
147
V
kJ-..^'-'
PhSeCI;
RO^^^^-^^'VlW : H o
1. Og.CHgClg.
a
MeOH; MegS
rhBuU; THF -78 °C
2. HC{0Me)3, (TsOH),MeOH
148
150
(MaO)2CH
9^^)^^ H
L.1^^
O
1. 5 % H C I , OH
2. Pr^P-CHg
151
THF
2. PDC. 4AMS. CH2CI2
f^O-
152
153
1. O ^ C H ^ I g ; PhaP SePh
2. n-BuLi, CH2(SePh)^
^SePh
THF, -78 '^C 3. CH3C(OMe)-CH2,
155
POCI3
1. Nal04. N a H C O j
O
H
^ /A
MeOH. H p 2. EtgN. CHg-CHOEt. MegNCOCHa, 220 °C
H
CH2I2,
7t'
(C2H5)Zn. CgHg
O
156
157 1. AC2O, DMAP
1. (^Bu)2AIH. CH2CI2. -78 ''C
2. P y H F , CH3CN. H2O
2. A g ^ O ^ - C e l l t e , CgH^A
3. PDC. 4A MS
B
3. KN{SiMe^2. PhSeCI; NalO^. NaHC03
4. Ph3P«C{CH3)2. THF, -78 "C ^ rt
^^
158
159
0 0
XHI J
OAc
118
Scheme XVIII
28
147 by sequential C-allylation, ozonolysis, and borohydride reduction. Once butenolide 148 had been accessed, 1,4-addition of enantiopure allylphosphonamide 149 [82] was carried out. As hoped for, 150 was formed uniquely. The configurations at sites a and b arise therefore from the chirality inherent in 149 and not elsewhere. Immediate removal of the chiral auxiliary followed by functional group manipulation gave rise to 153, which was transformed by means of chemistry developed above into 155. Once this advanced intermediate had been oxidized to the selenoxide, more elevated temperatures than usual were found necessary to bring about the conversion to 157. More heating is presumably required because the sidechain in 156 must be projected axially if a chair-like sigmatropic transition state is to be utilized. Despite this steric inhibition, 157 was isolated in 55% yield. Particularly satisfying was the ease with which 157 was homologated to 158 (92%). Reduction of the ketone carbonyl in a chemoselective manner was not possible because of the steric protection it benefits from. This potential complication was skirted when it was found that the hydroxy lactol produced by diisobutylaluminum hydridereductionrespondedto the Fetizon reagent only at the five-membered site to deliver 159. With the stereochemistry of 159 securely established by NOE analysis, no obstacles were encountered during acetylation and the subsequent completion of sidechain construction. Thus, although this enantioselective approach to (+)-118 is linear, the five stereocenters that adorn the eight-membered ring are conveniently set in a fashion which could well prove useful in a wide range of synthetic settings. H. (+).CLEMEOLIDE, THE STRUCTURALLY UNIQUE DITERPENE LACTONE CONSTITUENT OF CLEOME VISCOSA The herb Cleome viscosa (syn. cleome icosandra), which is widely distributed in India, has long been recognized by the native population to serve as a rubefaciant, vesicant, and anthelmintic agent. As a consequence of these reputed properties, three research groups undertook almost simultaneously in the late 1970's to determine the principal active constituent of this sticky, odoriferous plant [82,83]. On the basis of the NMR, X-ray, and CD data, the substance was determined to be the macrocyclic diterpene lactone 160 and named cleomeolide. The structural features of this macrolide are unusual in several respects: (a) the double bond positioned a,p to the lactone carbonyl resides at a bridgehead site, a
=
'OH
160
uH3C„
CH3H \ H
^^^^^'
29
property shared in common with taxol and other select natural products [84]; (b) the nine-carbon chain cis-fused to the methylenecyclohexane subunit is projected diaxially from the six-membered ring such that three of the four groups pendant on the cyclohexane are oriented in this fashion; and (c) this three-dimensional arrangement provides extensive steric screening to the exocyclic methylene carbon, such that its introduction at a modestly advanced stage of the total synthesis should be viewed as problematical. The correctness of this assumption has been assessed experimentally [85]. Our enantioselective approach to cleomeolide began by controlled dithioketalization [86] of optically pure Wieland-Miescher ketone [87] in order to distinguish between the two carbonyl groups. The best means uncovered for the homologation of 161 to the cis-dimethyl ketone 163 involved 162 as an intermediate (Scheme XIX). The action of (methoxymethylene)triphenylphosphorane on 161 afforded a 7:1 cis/trans mixture of isomers, which were easily separated after sodium borohydride reduction to the primary carbinols. The major component underwent reductive conversion to 163 very smoothly. O
(
'. /—V S ^ "^
)
1. NaBH4.MeOH 2. CHaSD^I,
"•• PhaPCHgOCI-^ Cl KN(SiMe3)2, THF
EtgN. CHjClg 3. LiBHEtg.THF 4. TI(N03)3«3H20, MeOH, THF
2. 10% HOI
^ 161
162
HC(OCH:^3, (TsOH),
OCH3
(
MeOH, DMF
OCH3
MCPBA
^
163
NaBH4
GeHe, hexanes, silica gel
MeOH
164
OCH3
1. TBSCI, imid, DMF 2. LiAIH4. THF
1. CH3SO2CI. EtgN. CH2CI2
OH
//
OTBS
OH 168
167
166
ON
2. KCN. 18-cr-6, DMF
H
ON
CI^-CHOCgHs.
xylene 200 °C
Hg(OCOCF3)2. EtgN 169
H3C
i^---^Sf? H
170
Scheme XIX
O
171
30
Since bond disconnection within 163 had to be implemented a to the carbonyl so as to maintain attachment of the methylene carbon, the dienol ether 164 was generated and treated with two equivalents of m-chloroperbenzoic acid. As a consequence of preferred electrophilic attack at the enol ether double bond and the high latent reactivity of the resulting oxygenated epoxide, ring cleavage occurred to deliver the aldehydo ester 165 [88]. Following the reduction of 165 to 167 via 166, the requisite additional carbon atom was introduced by cyanide ion displacement at a primary mesylate center. This transformation and the subsequent removal of the rerr-butyldimethylsilyl protecting group proceeded efficiently to provide 168. Transetherification of 168 with ethyl vinyl ether under strictly defined conditions [89] led to 169 and set the stage for the projected Claisen rearrangement. A reasonable rate of thermal isomerization was achieved at 200 °C in xylene under sealed tube conditions. The substitution plan in 169 led us to anticipate that transition state 172 would be kinetically favored ahead of 173 because of the development of destabiUzing 1,3-diaxial interaction in the latter. If 172 were indeed to be utilized, die new C-C bond would be appropriately installed in the desired cis manner as in 171. In actuality, the 170:171 ratio was found to be 2.7:1 and this product distribution was comparably found in analogues of 169 [90,91]. These findings suggest that a chair-like arrangement may not be strictly adopted during installation of the axial bond present in 170. H
v----Jr-CH3
172
Notwithstanding the adverse product distribution, contined use was made of this isomerization process because chromatographic separation of the epimers proved to be facile, and exceptional convergency could be subsequently reahzed during extension of the "lower chain". Thus, the two allylic alcohols 174 and 175 formed upon treatment of 171 with 2-propenylmagensium bromide were individually transformed into the identical aldehyde 176 following the implementation of a second Claisen rearrangement step (Scheme XX). Once this important finding was made clear, the three-step process could be streamlined by omitting all chromatography. Once acetal 177 had been produced [92], the cyano-substituted carbon was activated by conversion to 178 [93]. Cychzation of the unmasked aldehyde was efficiently realized with potassium carbonate and 18-crown-6 in toluene at room temperature [94], the ring closure occurring while both functionalized sidechains
31
HaC H,C.
o
H
THF, -78 "C
•
H3C
171
1. CH2.CHOC2H5. EtgN, Hg(OCOCF3)2 2. CBHe.170«'C
V >^V^^,
Z-^S^/-^v^5^^>...,x-s^H HgCT H
X O
pOSIMeg •-'OSIMeg
H^
TMSOTf, CHgCIs -78 "C
HgC' H 177
176
9 LDA, THF, -78 "C; aP(0)(OC2Kt)2
l?(OEt)2
H
1. TsOH.HgO, acetone 2. KjCOa. 18-crown-6, toluene, 20 °C
HscT
178
"1 H
H^
l-CN CH,
H3CI 179
Scheme XX
were equatorially disposed since the cyano triene clearly adopted the conformation depicted in 179 (NOE studies). Consequently, a substantive topographical change had to be implemented if cleomeolide was to be reached. Variable-temperature NMR studies performed on 179 indicated that this triene was not prone to reside in the alternative six-ring chair conformation related to 160. The issue of conformational dynamics was again specifically addressed with the methyl ester 181, available in three steps [95,95] from 179 as detailed in Scheme XXI. Although modest equilibration of 181 with 181* does occur at elevated temperatures, the geometrical modification occurs only within the larger ring and is not driven by a concomitant flexing within the cyclohexane subunit. These three-dimensional characteristics essentially guarantee that the direct capture by 181 of electrophilic reagents will occur on the wrong face of the trialkyl substituted double bond. The peracid oxidation leading to 182 is exemplary. Molecular models suggested to us that the proper conformational features might well be adopted if the oxirane ring were positioned instead on the more sterically congested inner surface of this unsaturated center. Indeed, the desired diastereofacial outcome was reahzed upon treatment of 181 with iodine and silver(I) oxide in aqueous dioxane [97,98]. The conclusion that 183 had indeed been formed was corroborated by NOE measurements and by the fact that the synthetic material was identical to an authentic sample produced by the saponifi-
32
'CH3
1. NaCI02. NaHgP04 2-methyl-2-butene H3C H2O. ^BuOH 2. CHgNg, ether
HaC
3.5% KOH. MeOH.AiHgO*
OOMe CH3
H^^^''-
183 5%HCI, V\3-5%*
""1'T/OCH
0
.^jXi^3c^ ^ • ^ 1 ^
-OH
C^b H ^ H
^^
H HJ-,
1
0
H3C-7—Yvj j
160 182
Scheme X X I
cation, acidification, and esterification of cleomeolide according to a published procedure [83]. Once the two new stereogenic centers had been introduced as in 183, it proved an easy matter to saponify the ester functionahty and to effect the ring closure of 184 under acidic conditions. Since spectroscopic comparison showed our product to be identical to natural cleomeolide, retrospective analysis must necessarily focus on the effective use of functional group deployment to achieve global conformational control as the key element to stereocontroUed macrocychc ring assembly. I. (.)-AUSTALIDE B, AN ORTHO ESTER MEROTERPENOID MYCOTOXIN PRODUCED BY ASPERGILLUS USTUS Austalides A-F (185-190) have gained prominence as a result of their discovery as toxigenic agents produced in dried fish by Aspergillus ustus [99-102]. The imique molecular architecture of the metabolites, which is common to all six
33
CX^Hg^ OCH, 185, A: R ' . Ac R2 - H 186,B:R'.H R^-H 187, C: R- - Ac R^ . OAc
188, D: R* « H R2 « OAc 189, E: R" - Ac R^. OH 190, F: R' « H R^ . OH
members of the family, includes an ortho ester subunit spanning rings A and B, a dense array of stereogenic centers around ring C, and a pyran/p-cresol/butenolide triad in the eastern sector. It appeared to us that a potentially serviceable disconnection of austalide B could be made across the two C-C bonds of ring E conjoined to the pyran ring as illustrated below. Advanced assembly could therefore be based upon the annealing of rings E and F onto a preexisting dihydropyran (viz., 191) or 5-lactone (viz., 192), a tactic highly accommodating of convergency. Accordingly, our first goals were to craft an optically active intermediate related to 192 [103,104] and to develop a unified strategy for elaborating ttie DEF tricychc subunit [105,106].
H3c-rTV^-^ H3C0 186
191
The first of these objectives was realized starting from the readily available diketone 193 (98% ee) [107]. Its conversion to 194 by regioselective ketahzation [108], dissolving metal reduction [109], and in situ methylation [110] provided 194 (Scheme XXII). This intermediate proved difficuh to engage in Robinson annulation and only an acid-catalyzed variant involving the use of 4-chloro-2butanone [111] proved feasible for delivering 195 stereoselectively. Regioselective dimethylation of this tricyclic enedione could be readily accomplished without carbonyl protection. The ensuing osmylative dihydroxylation of 196 proceeded exclusively from the p-face [112] to give 197, thereby setting the absolute configuration of the tertiary carbinol center correctly. The secondary hydroxyl was transformed into the SEM ether with the notion that the requisite a-configuration of this carbinol would be properly established later in the synthesis. Further oxidation of both carbonyl-containing rings was now required in order to incorporate an oxygen atom into both sites. Quite unexpectedly, peracid
34
H3C
9">
1. H,0^
o
TsOH
.
2. U.NHg;
^
TsOH.CgHe A
Ch^i. ether
195
194
193
Ka-Bu ^
^
•
^BuOH; CHgl
H2O, acetone; Na2S204
196 SEMOL
1. SEMCI. (/-PQgNEt
O8O4.NMO.
HgC
197 SEMQ,
p
^BFr
H3C, H3C
2. MCPBA. NaHCOg, CHgClg
MCPBA
^
NaHCOg. CH2CI2
iiC 0 . ^ 0
H3C, HgC
,co 198
199
Scheme XXn oxidation of SEM-protected 197 resulted in the operation of Baeyer-Villiger ring expansion uniquely within ring A. The possibility exists that this impressive regioselectivity may stem from conformational biases that sterically shield the cyclopentanone keto group or from the close proximity of the tertiary hydroxyl to the six-ring ketone. Whatever the case, we were led to adopt a protocol that would implement these ring expansions nonsimultaneously. Once 198 had been transformed into ortho lactone 199 by 0-methylation with trimethyloxonium tetrafluoroborate [113], the second oxidation was discovered to proceed at a convenient rate under the peracid conditions originally emloyed. The model study originally pursued to determine how to fuse rings E and F to 200 began with the lithiation of dihydropyran with /er/-butyllithium [114] and condensation of the vinyl anion so generated with N,N-dimethylformamide [115] (Scheme XIII). Condensation of aldehyde 201 with the multifunctional Wittig reagent 202 [116] gave rise exclusively to 203 in which the dihydropyran subunit is positioned cis to the acetic acid sidechain. The E configuration made possible the spontaneous ring closure of this intermediate upon heating with oxalyl chloride in dichloromethane solution. Phenol 204 so produced was 0-methylated and subjected to hydride reduction in steps preliminary to attachment of the final two aryl substituents. To this end, advantage was taken of the remarkably disparate extent to which the methoxyl and cyclic ether oxygen atoms in 205 are capable of control-
35 PhgP^^COOCjHg
0
a O^
^BuU; HCONMe.
crt'
COOH 202
XHO
COOC2H5
CeHe. 55 X
ccx><
201
_JCOCI)2, CBHB,„ 45 '^C
'^^^^ XX^SLX^ OH
203
1.NaK CH3I 2. L i A H , THF.A
THF; HCONMG2
205
CHO
OCH, 206
N2H4 K2CO3 diethylene giycoi, A
n-BuU (2 equlv),
OCH3
204
<^^Y^OH
OH
OH
OCH3
^^BuLi (2 equlv), EtgO. TMEDA; CO2 (solid)
CH,
Tp OCH3 208
207
Scheme XXIII ling the regiochemistry of aryl lithiation. Well aware of the greater directing power of the pyran oxygen, we proceeded to initially implement formylation and to subject 206 to Wolff-Kishner reduction. Finally, the lactone ring was incorporated into 207 by metalation and carboxylation with solid carbon dioxide. With arrival at 208, a direct means for realizing the regiocontrolled functionalization of austalide's DEF subunit had been developed. Unfortunately, later investigations involving the application of this chemistry to dihydropyran 209 [117] revealed that the requisite use of tert- and n-butyllithiums had untoward degradative consequences. An altemative pathway had therefore to be developed that would skirt such demanding conditions.
" ^ • ^ 6 3'CH3 0CH3
209
Following additional trial experiments which revealed the need to activate 200 by C-acylation [104], the anion of this lactone was condensed with methyl cyanoformate [118] to produce 210 (Scheme XXIV). In line with plans to achieve convergency by means of Stille coupling [119], it was necessary to activate 210 as
36 HgC
SEMO
SEM H3C
LDA NCCXDOMe. THF
^JaC Q
^OTf
KN(SiMe3)2; ^COjjCHa
0 HgCO'^
PhNTfg
0 HgCO
210
200
^COgCHa
ol 211
Scheme XXIV the enol triflate 211. Curiously, no precedent was found that could delineate unequivocally our expectations regarding the regioselectivity with which a plactonic ester would engage in O-triflation. To our delight, reaction of the enolate anion of 210 with N-phenyltriflimide gave rise in good yield to a single triflate confirmed by chemical means to be 211. The projected palladium-catalyzed cross-coupling required the availability of vinylstannane 216. As shown in Scheme XXV, the preparation of this lactone was initiated by copper-catalyzed 1,4-addition of l-(trimethylsilyl)vinylmagnesium bromide to 5(2//)-furanone (212). For this process to be successful, excess trimethylsilyl chloride had to be present from the outset in order to trap the enolate as it was formed and circumvent its polymerization. This modification gave rise to C-silylated lactone 213, which was chemoselectively desilylated and transformed via vinyl bromide 215 [120] into stannane 216. CH2-C
Q
MgBr
MeoSi
HF
o
MeoSi
CHaCN. H2O
GuBr-Me2S MesSiCl 213
212
Bro
(Me 380)2
BU4N+ F •
Pd(PPh3),
214
Me 380'
O 215
216
Scheme XXV As our investigation of the conjoining of 216 to 211 progressed, it was soon recognized that prevailing steric effects exerted substantial kinetic regardation when the usual coupling conditions, viz. Pd(PPh3)4 and LiCl in THF at 60 °C, were applied. However, this complication could be conveniently bypassed by carrying out the reaction in the presence of tris-2-furylphosphine and Pd2(dba)3 instead [121] (Scheme XXVI). Dienyl ester 217 was obtained as a 1:1 mixture of
37
216 Pd2(dba)3 ^
KN(SiMe3)2
^
(furyOgP. UCI
THF, -78 «'C
THF, 60 "C HgCO
H;^0
211
217
KN(SiM63)^ HMPA, Me2S04;
BU4N*F HMPA. 45 "C
HgCO
218
HO,
219
HgC
O.
H3C
TPAP CH,CU -.jV/lg. 0 "C 220
H3C
^.. NaBH4 MeOH. 0 "C
fcHj
HgC
1 HgC'T 0
rr CH3
0
-Sjx^
L
0CH3
H3C0
01 0 1
186
Scheme XXVI diastereomers. As a result, the product of Claisen cyclization was constituted of an equal distribution of both possible cis-fused 4-methylenecyclohexenones. Without separation, these advanced intermediates were 0-methylated in order to facilitate intemalization of the external double bond and aromatization. The [1,3] hydrogen sigmatropy was easily accomplished simply by heating the pair of ethers in benzene for 15 min while exposed to the atmosphere. Following arrival at 219, the SEM protecting group was removed to give palcohol 220, perruthenate oxidation of which [122] provided ketone 221, which had earher been generated during degradative studies on the austalides. Reduction of 221 with sodium borohydride proceeded with exclusive attack from the less sterically shielded p face to deliver austalide B (186).
38
ACKNOWLEDGMENTS I take distinct pleasure in recognizing the many significant experimental results of my co-workers and the extensive intellectual input of these individuals: Larry Anderson, Jesus Ezquerra, Dirk Friedrich, Wei He, Ho-Jung Kang, Dongsoo Koh, Ho-Shen Lin, Dwight Macdonald, Rob Maleczka, Christophe Philippo, Enmianuel Pinard, Jeremy Prodger, Choon Sup Ra, Brian Roden, Jeff Romine, Denis St. Laurent, Max Schulze, Matt Sivik, Nha Vo, Shaopeng Wang, TingZhong Wang, Xiaodong Wang, John Williams, and Jon Wright. Financial support from the National Institutes of Health has been greatly appreciated. REFERENCES 1 2 3 4
5 6 7 8 9
10 11 12 13
14
B. M. Trost, Ed. "StereocontroUed Organic Synthesis"; Blackwell Scientific Publications, Oxford, England, 1994. M. Castillo, L. A. Loyola, G. Morales, I. Singh, C. Calvo, H. L. Holland and D. B. MacLean, Can. J. Chem., 54 (1976) 2893. L. A. Loyola,G. Morales and M. Castillo, Phytochemistry, 18 (1979) 1721. (a) M. Castillo, G. Morales, L. A. Loyola, I. Singh, C. Calvo, H. L. Holland and D. B. MacLean, Can. J. Chem., 53 (1975) 2513. (b) M. Castillo, G. Morales, L. A. Loyola, I. Singh, C. Calvo, H. L. Holland and D. B. MacLean, Can. J. Chem., 54 (1976) 2900. (a) G. Mehta and K. S. Rao, J. Chem. Soc, Chem. Commun. (1987) 1578. (b) G. Mehta, G. and M. S. Reddy, Tetrahedron Lett., 31 (1990) 2039. (a) G. C. Hirst, P. N. Howard and L. E. Overman, J. Am. Chem. Soc, HI (1989) 1514. (b) G. C. Hirst, T. O. Johnson, Jr. and L. E. Overman, J. Am. Chem. Soc, 115(1993)2992. M. T. Crimmins and P. S. Watson, Tetrahedron Lett., 34 (1993) 199. D. R. St. Laurent and L. A. Paquette, J. Org. Chem., 51 (1986) 3861. (a) L. A. Paquette, D. Friedrich, E. Pinard, J. P. Williams, D. R. St. Laurent and B. A. Roden, J. Am. Chem. Soc, 115 (1993) 4377. (b) J. P. Williams, D. R. St. Laurent, D. Friedrich, E. Pinard, B. A. Roden and L. A. Paquette, J. Am. Chem. Soc, 116 (1994) 4689. K. Tomioka and K. Koga, Tetrahedron Lett., 25 (1984) 1599. L. N. Mander and S. P. Sethi, Tetrahedron Lett., 24 (1983) 5425. (a) C. Chatgilialoglu, D. Griller and M. Lesage, J. Org. Chem., 53 (1988) 3641. (b) B. Giese, B. Kopping and C. Chatgilialoglu, Tetrahedron Lett., 3 30, (1989)681. (c)K.J.KulickeandB.Giese,Synlett(1990)91. (a) J. Pfenninger, C. Heuberger and W. Graf, Helv. Chim. Acta, 63 (1980) 2328. (b)M.D.Bachi and E.Bosch, Tetrahedron Lett., 27 (1986) 641. (c) D. L. J. Clive, H. W. Manning and T. L. B. Boivin, J. Chem. Soc, Chem. Commun. (1990) 972. (d) D. Friedrich and L. A. Paquette, J. Chem. Soc, Perkin Trans. 1(1991)1621. (a) T. Satoh, S. Suzuki, Y. Suzuki, Y. Miyaji and Z. Imai, Tetrahedron Lett., 10 (1969) 4555. (b) T. Harayama, M. Ohtani, M. Oki and Y. Inubushi, Chem. Pharm. BuU., 23 (1975) 1511. (c) S. W. Heinzman and B. J.Ganem, J. Am. Chem. Soc, 104 (1982) 6801.
39
15 D. L. Hughes, Org. React.. 42 (1992) 335. 16 (a) H.-J. Kang, C. S. Ra and L. A. Paquette, J. Am. Chem. Soc, 113 (1991) 9384. (b) L. A. Paquette, H.- J. Kang and C. S. Ra, J. Am. Chem. Soc, 114 (1992) 7387. 17 H. Seto, T. Sasaki, H. Yonehara, S. Takahashi, M. Takeuchi, H. Kuwano and M. Arai, J. Antibiot.. 37 (1984) 1076. 18 M. Demuth and K. Schaffner, Angew. Chem., Int. Ed. Engl., 21 (1982) 820. 19 T. Asao, S. Kuroda and K. Kato, Chem. Lett. (1978) 41. 20 (a) L. A. Paquette, H. Schostarez and G. D. Annis, J. Am. Chem. Soc, 103 (1981) 6526. (b) L. A. Paquette, G. D. Annis and H. Schostarez, J. Am. Chem. Soc, 104 (1982) 6646. 21 S. Cacchi, E. Morera and G. Ortar, Tetrahedron Lett., 26 (1985) 1109. 22 M. Schlosser, T. Jenny and Y. Guggisberg, Synlett (1990) 704. 23 S. Danishefsky, M. Hirama, K. Gombatz, T. Harayama, E. Berman and P. F. Schuda, J. Am. Chem. Soc, 100 (1978) 6536; 101 (1979) 7020. 24 K. Jomon, Y.Kuroda, M. Ajisaka and H. Sakai, J. Antibiot., 25 (1972) 271. 25 (c) L. A. Paquette, D. Macdonald, L. G. Anderson and J. Wright, J. Am. Chem. Soc, 111 (1989) 8037. (b) L. A. Paquette, J. L. Romine, H.-S. Lin and J. Wright, J. Am. Chem. Soc, 112 (1990) 9284. (c) L. A. Paquette, D. Macdonald and L. G. Anderson, J. Am. Chem. Soc, 112 (1990) 9292. 26 (a) R. K. Boeckman, Jr., C. H. Weidner, R. B. Pemi and J. J. Napier, J. Am. Chem. Soc, 111 (1989) 8036. (b) R. K. Boeckman, Jr. and R. B. Pemi, J. Org. Chem., 51 (1986) 5486. (c) R. K. Boeckman, Jr., J. J. Napier, E. W. Thomas and R. L Sato, J. Org. Chem., 48 (1983) 4152. 27 M. J. Kurth, D. H. Bums and M. J. O'Brien, J. Org. Chem., 49 (1984) 731. 28 (a) J. K. Whitesell and M. Minton, J. Am. Chem. Soc, 109 (1987) 6403. (b) J. K. Whitesell, M. A. Minton and V. D. Tran, J. Am. Chem. Soc, 111 (1989) 1473. 29 L. A. Paquette, J. L. Romine and H.-S. Lin, Tetrahedron Lett., 28 (1987) 31. 30 (a) M. E. Jung and J. P. Hudspeth, J. Am. Chem. Soc, 100 (1978) 4309. (b) M. E. Jung and J. P. Hudspeth, J. Am. Chem. Soc, 99 (1977) 5508. 31 (a) S. Hashimoto, S. Yamada and K. Koga, J. Am. Chem. Soc, 98 (1976) 7450. (b) S. Hashimoto, H. Kogen, K. Tomioka and K. Koga, Tetrahedron Lett., 20 (1979) 3009. (c) K. Tomioka and K. Koga, in: J. D. Morrison (Ed.), "Asymmetric Synthesis Volume 2, Part A," Academic Press, New Yoik, 1983, Chapter 7. 32 (a) L. A. Paquette and K. S. Learn, J. Am. Chem. Soc, 108 (1986) 7873. (b) L. A. Paquette, K. S. Learn, J. L. Romine and H.-S. Lin, J. Am. Chem. Soc, 110(1988)879. 33 W. Barth and L A. Paquette, J. Org. Chem., 50 (1985) 2438. 34 (a) S. Yamada and S. Hashimoto, Chem. Lett. (1976) 921. (b) S. Hashimoto, N. Komeshima, S. Yamada and K. Koga, Tetrahedron Lett., 18 (1977) 2907. (c) S. Hashimoto, S. Yamada and K. Koga, Chem. Pharm. Bull. Jpn., 27 (1979) 771. 35 E. J. Corey and P. L. Fuchs, Tetrahedron Lett. (1972) 3769. 36 (a) R. K. Boeckman, Jr. and A. J. Thomas, J. Org. Chem., 47 (1982) 2823. (b) R. K. Boeckman, Jr., R. B. Pemi, J. E. McDonald and A. J. Thomas, Org. Synth., 66 (1987) 194. 37 P. D. Jeffrey and S. W. McCombie, J. Org. Chem., 47 (1982) 587.
40
38 (a) J. A. Hyatt, P. L. Feldman and R. J. Clemens, J. Org. Chem., 49 (1984) 5105. (b) R. J. Clemens and J. A. Hyatt, J. Org. Chem., 50 (1985) 2431. 39 E. M. Burgess, H. R. Penton, Jr. and E. A. Taylor, J. Org. Chem., 26 (1973) 38. 40 (a) L. A. Paquette, Synlett (1990) 67. (b) L. A. Paquette, Angew. Chem., 102 (1990) 642; Angew. Chem., Int. Ed. Engl., 29 (1990) 609. 41 M. E. Jung and J. P. Hudspeth, J. Am. Chem. Soc, 100 (1978) 609. 42 G. Cimino, S. De Stefano,A. Guerriero and L. Minale, Tetrahedron Lett. (1975)1417,1425. 43 S. P. Tanis and P. M. Herrington, J. Org. Chem., 48 (1983) 4572. 44 D. Nasipuri and G. Das, J. Chem. Soc., Perkin Trans. 1 (1979) 2276. 45 K. Shishido, K. Umimoto and M. Shibuya, Heterocycles, 31 (1990) 597. 46 A. B. Smith, m and R. Mewshaw, J. Org. Chem., 49 (1984) 3685. 47 T. Akita and T. Oishi, Chem. Phann. Bull., 29 (1981) 1580. 48 T. Matsumoto and S. Usui, Chem. Lett. (1978) 105. 49 R. E. Maleczka, Jr. and L. A. Paquette, J. Org. Chem., 56 (1991) 6538. 50 L. A. Paquette and R. E. Maleczka, Jr., J. Org. Chem., 57 (1992) 7118. 51 M. W. Creese and E. E. Smissman, J. Org. Chem., 41 (1976) 169. 52 (a) P. A. Jacobi and R. F. Frechette, Tetrahedron Lett., 28 (1987) 2937. (b) B. R. Brown and A. M. S. White, J. Chem. Soc. (1957) 739. 53 L. A. Paquette, D. Koh, X. Wang and J. C. Prodger, Tetrahedron Lett., 36 (1995) in press. 54 G. Appendino, P. Gariboldi and M. G. VaUe, Gazz. Chim. Ital., 118 (1988) 55. 55 (a) C. M. G. Philippo, N. H. Vo and L. A. Paquette, J. Am. Chem. Soc, 113 (1991) 2762. (b) L. A. Paquette, C. M. G. Philippo and N. H. Vo, Can. J. Chem., 70 (1992) 1356. 56 L. A. Paquette, T.-Z. Wang and N. H. Vo, J. Am. Chem. Soc, 115 (1993) 1676. 57 Z. G. Hajos and D. R. Parrish, Org. Synth., 63 (1985) 26. 58 K. M. Patel and W. Reusch, Synth. Commun., 5 (1975) 27. 59 (a) R. H. Shapiro, Org. React., 23 (1976) 405. (b) A. R. Chamberlin and S. H. Bloom, Org. React., 39 (1990) 1. 60 R. E. Ireland, D. C. Muchmore and U. Hengartner, J. Am. Chem. Soc, 94 (1972) 5098. 61 E. Piers and V. Karunaratiie, J. Chem. Soc, Chem. Commun. (1983) 935. 62 J. E. McMurry and W. J. Scott, Tetrahedron Lett., 24 (1983) 979. 63 J. E. McMurry and W. J. Scott, Tett-ahedron Lett., 21 (1980) 4313. 64 Y. Ito, T. Hirao and T. Saegusa, J. Org. Chem., 43 (1978) 1011. 65 R. K. Boeckman, Jr., A. Arvanitis and M. E. Voss, J. Am. Chem. Soc, 111 (1989) 2737. 66 G. Guella and F. Pietra, J. Chem. Soc, Chem. Commun. 1993,1539. 67 H. H. Sun, F. J. McEnroe and W. Fenical, J. Org. Chem., 48 (1983) 1903. 68 S. L. Midland, R. M. Wing and J. J. Sims, J. Org. Chem., 48 1983) 1906. 69 M. Ishitsuka, T. Kusumi, H. Kakisawa, Y. Kawakami, Y. Nagai and T. Sato, Teu-ahedron Lett., 24 (1983) 5117 70 T. Kusumi, D. Muanza-Nkongolo, M. Goya, M. Ishitsuka, T. Iwashita and H.Kakisawa, J. Org. Chem., 51 (1986) 384. 71 C. Tringah, G. Oriente, M. Piattelli, C. Geraci, G. Nicolosi and E. Breit-
41
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 97 98
maier, Can. J. Chem., 66 (1988) 2799. (a) J. Ezquerra, W. He and L. A. Paquette, Tetrahedron Lett., 31 (1990) 6979. (b) L. A. Paquette, J. Ezquerra and W. He, J. Org. Chem., 60 (1995) in press. (a) S. Takano, M. Yonaga, M. Morimoto and K. Ogasawara, J. Chem. Soc. Perkin Trans. 1 (1985) 305. (b) S. Takano, N. Tamura and K. Ogasawara, J. Chem. Soc, Chem. Commun. (1981) 1155. M. Taniguchi, K. Koga and S. Yamada, Tetrahedron, 30 (1974) 3547. S. Takano, M. Yonaga and K. Ogasawara, Synthesis (1981) 265. (a) K. C. Nicolaou, D. A. Claremon, W. E. Bamette and S. P. Seitz, J. Am. Chem. Soc, 101 (1979) 3704. (b) P. A. Grieco, J. Y. Jaw, D. A. Claremon and K. C. Nicolaou, J. Org. Chem., 46 (1981) 1215. S. Sawada and Y. Inoue, Bull. Chem. Soc Japan, 42 (1969) 2669. W. He, E. Pinard and L. A. Paquette, Helv. Oiim. Acta, in press. (a) D. A. Evans, M. D. Ennis and D. J. Mathre, J. Am. Chem. Soc, 104 (1982) 1737. (b) D. A. Evans and D. J. Mathre, J. Org. Chem., 50 (1985) 1830. (a) K. Tamao, N. Ishida and M. Kumada, J. Org. Chem., 48 (1983) 2120. (b) K. Tamao and N. Ishida, Tetrahedron Lett.. 25 (1984) 4245,4249. L. A. Paquette, T.-Z. Wang and E. Pinard, J. Am. Chem. Soc, 117 (1995) in press. S. B. Mahato, B. C. Pal, T. Kawasaki, K. Miyahara, O. Tanaka and K. Yamasaki, J. Am. Chem. Soc, 101 (1979) 4720. B. A. Burke, W. R. Chan, V. A. Honkan, J. F. Blount and P. S. Manchand, Tetrahedron, 36 (1980) 3489. L. A. Paquette, Chem. Soc. Rev., 24 (1995) in press. C. M. G. Philippo, Ph.D. Dissertation, The Ohio State University, 1991. P. M. Bosch, F. Camps, J. Coll, A. Guerrero, T. Tatsuoka and J. Meinwald, J. Org. Chem., 51 (1986) 773. N. Harada, T. Sugioka, H. Uda and T. Kuriki, Synthesis (1990), 53. D. N. Kirk and J. M. Miles, Chem. Commun. (1970) 1015. D. B. Tulshian, R. Tsang and B. Fraser-Reid, J. Org. Chem., 49 (1984) 2347. L. A. Paquette, T.-Z. Wang, S. Wang and C. M. G. Philippo, Tetrahedron Lett., 34 (1993) 3523. L. A. Paquette, T.-Z. Wang, C. M. G. Philippo and S. Wang, J. Am. Chem. Soc, 116(1994)3367. T. Tsunoda, M. Suzuki and R. Noyori, Tetrahedron Lett., 21 (1980) 1357. D. L. Comins, A. F. Jacobine, J. L. Marshall and M. M. Tumbull, Synthesis (1978) 309. W. C. Still and C. Gennari, Tetrahedron Lett., 24 (1983) 4405. B. S. Bal, W. E. Childers, Jr. and H. W. Pinnick, Tetrahedron, 37 (1981) 2091. J. Wright, G. J. Dritna, R. A. Roberts and L. A. Paquette, J. Am. Chem. Soc, 110(1988)5806. M. Parrilli, G. Berone, M. Adinolfi and L. Mangoni, Tetrahedron Lett. (1976) 207. R. Pohiiaszek and R. V. Stevens, J. Org. Chem., 51 (1986) 3023.
42
99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122
R. M. Horak, P. S. Steyn, P. H. Van Rooyen, R. Vleggaar and C. J. Rabie, J. Chem. Soc, Chem. Commun. (1981) 1265. R. M. Horak, P. S. Steyn, R. Vleggaar and C. J. Rabie, J. Chem. Soc, Peikin Trans. 1(1985)345. R. M. Horak, P. S. Steyn and R. Vleggaar, J. Chem. Soc, Peikin Trans. 1 (1985)357. R. M. Horak, P. S. Steyn, R. Vleggaar and C. J. Rabie, J. Chem. Soc, Peikin Trans. 1(1985)363. L. A. Paquette, T.-Z. Wang and M. R. Sivik, J. Am. Chem. Soc, 116 (1994) 2665. L. A. Paquette, T.-Z. Wang and M. R. Sivik, J. Am. Chem. Soc, 116 (1994) 11323. L. A. Paquette and M. M. Schulze, Heterocycles, 35 (1993) 585. L. A. Paquette, M. M. Schulze and D. G. Bolin, J. Org. Chem., 59 (1994) 2043. Z. Hajos and D. R. Parrish, Organic Syntheses; Wiley: New York, 1990, Collect. Vol. Vn,p 363. G. Bauduin and Y. Pietrasanta, Tetrahedron, 29 (1973) 4225. G. Stork, P. Rosen, N. Goldman, R. V. Coombs and J. Tsuji, J. Am. Chem. Soc, 87 (1965) 275. H. A. Smith, B. J. L. Huff, W. J. Powers, HI and D. Caine, J. Org. Chem., 32(1967)2851. P. A. Zoretic, B. Branchaud and T. Maestrone, Tetrahedron Lett. (1975) 527. M. R. Sivik, J. C. GaUucci and L. A. Paquette, J. Org. Chem., 55 (1990) 391. H. Meerwein, P. Bomer, O. Fuchs, H. J. Sasse, H. Schrodt and J. Spille, Chem. Ber., 89 (1956) 2060. R. K. Boeckman, Jr. and K. J. Bmza, Tetrahedron Lett. (1977) 4187. R. K. Boeckman, Jr. and K. J. Bmza, TeU-ahedron, 37 (1981) 3997. E. Roder and H. Krause, Liebigs Ann. Chem. (1992) 177. M. R. Sivik, Ph.D. Dissertation, The Ohio State University, 1991. L. N. Mander and S. P. Sethi, Tetrahedron Lett., 24 (1983) 5425. W. J. Scott and J. K. Stille, J. Am. Chem. Soc, 108 (1986) 3033. J. K. Stille and B. L. Groh, J. Am. Chem. Soc, 109 (1987) 813. V. Farina, S. R. Baker, D. A. Benigni, S. I. Hauck and C. Sapino, Jr., J. Org. Chem., 55 (1990) 5833. W. P. Griffith and S. V. Ley, Aldrichim. Acta, 23 (1990) 13.
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
A Historical Perspective of Morphine Syntheses Tomas Hudlicky,* Gabor Butora, Stephen P. Feamley, Andrew G. Gum, and Michele R. Stabile Virginia Polytechnic Institute & State University, Blacksburg, Virginia 24061. Table of Contents 1. Introduction 2. History of Morphine and Its Chemistry a. First Documented Use b. Legal and Illicit Use c. Isolation of Pure Compound d. Structure Elucidation e. Biosynthesis 3. Total and Formal Syntheses a. Gates b. Ginsburg c. Barton d. Morrison, Waite, Shavel; Grewe e. Kametani f. Schwartz g. Beyerman h. Rice i. Evans j . Rapoport k. White 1. Schafer m. Fuchs n. Tins o. Parker p. Overman 4. Synopsis of Approaches to the Ring Systems of Morphine 5. Conclusions 6. Acknowledgements 7. References 8. Appendix I. List of dissertation titles concerning morphine syntheses n. List of references connected to synthetic transformations of morphine and derivatives and biological testing
43
44
1.
Introduction Morphine 1, named after the Greek god of dreams, Morpheus, is one of the oldest
drugs on record and has been on the drawing boards of synthetic chemists longer than any other natural product. Not as complex as, for example, the dimeric indole alkaloids or taxol, among other targets, morphine remains unique in its resistance to elhcit a truly practical synthesis which would also please the artistically demanding beholder. Despite several ingenious approaches, the synthesis fails to compete favorably with the isolation in terms of cost. That morphine remains on the list of goals of synthetic chemists is supported by the nearly forty year span from the first total synthesis by Gates to the latest report by Overman.
There is of course a sound reason for the synthetic challenge: morphine contains nearly completely dissonant^ arrangement of carbon atoms, a benzylic quaternary center at Ci3, and its skeleton is prone to a number of fascinating rearrangements thus hmiting the methods by which it might become accessible. It may be desirable in the long run to produce morphine synthetically given the uncertainty over the political and economic stability of the countries that produce it (Turkey, India, Thailand, among others), as well as the fact that the increasing demand for its use may soon exceed its supply from natural sources. Perhaps advances in plant tissue culture research will yield a tentative preparation of 1 and its congeners once the
45
details of the complex enzymatic pathways are elucidated. However, until there appears a synthesis of morphine in less than ten steps, this molecule will remain an elusive target of exercise for the synthetic practitioner. In this chapter, a brief overview of the state of the art of morphine research is provided along with a detailed review of the several total syntheses that have materialized to date. Some approaches to the general skeleton of morphine alkaloids are also presented in order that such attempts may be put in perspective with the completed studies. Omitted from the review are the syntheses of derivatives and pharmaceuticals resulting from manipulation of the natural product itself. The authors would hope that the comparison of total synthetic approaches will serve to inspire the creative reader and lead to an efficient and artistic attainment of the title compound in the future. The Uterature in this area has been reviewed through the end of 1994.
2. History of Morphine and Its Chemistry 2a.
First Documented Use The unripened seed of the opium poppy, Papaver somniferum, contains a milky
white fluid. If the seed is lacerated and the fluid exposed to air, it dries and darkens to a thick black paste which has been known for centuries as opium. No other single plant shares the rich history of the opium poppy. Since the ancient empires of Mesopotamia through today, the opium alkaloids have been at the center of medicine, pohtical turmoil, and economic growth, and this influence has been described in an excellent contemporary re view. 2 A collection of some interesting opium facts is summarized in Table I. Since opium's first documented use in the Smith and Ebers Papyri, circa 1500 BC,3 its impact on society has been extremely significant. This significance rests on the potency of the components of opium, notably, the major one, morphine, which gives it its analgesic, euphoric, and potentially addictive properties. In ancient times, as described in Homer's Odyssey, opium alkaloids were used "to quiet all pain and strife, and bring
46 Table I. Historical Facts Concerning Opium^^'*' Event
Date 1500 BC
First documented use of opium - Smith and Ebers Papyri, Cyprus
circa 30 AD
"gall" (opiated wine) offered to Christ on the cross - Matthew 27:34
1st century
Dioscorides describes opium as medication - De Materia Medica
7th century
Knowledge of the opium poppy appears in India and China
15th century
Japan starts opium cultivation
1530 17th century
Paracelsus dissolves opium in alcohol making laudanum Opium smoking reported as serious problem in China
1806
F.W.A. Sertumer isolates principal opium alkaloid - morphine
1822
T. DeQuincey describes his addiction to laudanum in "Confessions of an Enghsh Opium Eater"
1839-42
First "Opium War"; trade dispute between China and Britain
1856-60
Second "Opium War"
1925
J.M. GuUand and Sir R. Robinson present formulation for the molecular structure of morphine
1952
First total synthesis of morphine by M. Gates and G. Tschudi
1980
First practical synthesis of morphine by K.C. Rice
1984
Global production quotas for morphine set at 56,540 kg (55,490 kg converted and 1,050 kg for sales)
1994
Global production quotas for morphine increase to 82,148 kg (75,668 kg converted and 6,480 kg for sales)
|
|
' i
47 forgetfulness of every ill". As apparent from the progression of use in Table I, morphine demand continues to increase in modem times. To date, no synthetic drug or naturally occurring compound has been found which can rival the broad spectrum analgesic properties of the opium alkaloids. The primary medical use of morphine is for relief of pain, the activity believed to be caused by the agonist binding to the |i receptors in the central nervous system. Currently, morphine is being used in open-heart surgery because of its cardiac inactivity. As in an overdose, the large dosage leads to depression of respiration while the heart remains unaffected - a situation which is deemed ideal in a procedure where the patient's blood is artificially oxygenated. Derivatives of morphine, such as codeine 2, (Figure 1), are used as cough
nalmcfcnc
noroxymorphonc
Figure 1. Derivatives of Morphine in Pharmaceutical Use suppressants. Morphine antagonists such as naloxone 3, naltrexone 4, and nalmefene 5 are effective in the treatment of accidental overdose and have also been shown to be
48 active in the treatment of alcohol abuse and eating disorders. These compounds are manufactured from morphine or thebaine via noroxymorphone 6.
2b.
Legal and Illicit Use
According to a United Nation's treaty,^^ which was signed by 116 nations, India is recognized as the largest legal producer of opium. All of the opium cultivated in India by licensed growers must be sold to the Indian government. Inspection of the 1994 aggregate production quotas of morphine for conversion and sales, which were set at 75,668 kg and 6,480 kg respectively,'^ reveals the immensity of this legahzed cultivation. While the medicinal benefits of morphine and other opium alkaloids are extremely significant, it is their unfortunate addictive qualities which can burden society. Both the desire to satisfy addiction and human greed for profit foster the iUicit production and trade of opium alkaloids between nations. Globally in 1993, an estimated 3,699 metric tons of opium was ilHcitly cultivated. The majority of this ilUcitly cultivated opium came from the nation of Burma, and it was almost exclusively trafficked in the form of heroin 7. AcO.,
NMe AcO'
In that same year, drug enforcement agencies seized only 23 metric tons of heroin worldwide.^ Heroin, formed by simple acetylation of natural morphine,^ was synthesized in an attempt to alleviate morphine addiction, but ironically, it proved to be many times more addictive, and therefore, more profitable on the illicit market. Controlling ilUcit
49 cultivation, conversion, and trade is difficult because of the variety of international philosophies and interests involved.
2c.
Isolation of Pure Compound
Morphine is the major component of opium representing 10% - 20% of its weight.'7 Following initial isolation of a crystalline opium constituent by Derosne in 1803^ and a description of the isolation of morphine from opium by Seguin at a presentation to the Institute of France in 1804,^ Sertiimer was fmally credited with the first isolation of crystalline morphine in 1806.^^
Subsequently, a number of other
alkaloids have been isolated from opium, and their structures and biological activities elucidated, Figure 2.
NMe
NMe HO^^
NMe MeO'
8
1.
(T)- thcbainc
(-)- codeine
(-)- morphine
MeO.
MeO.
MeO. NMe
NMe MeO
10 (-)- neopine
papaverine Figure 2. Some Naturally Occuring Opium Alkaloids
50 2d. Structure Elucidation With Sertiimer's initial isolation of crystalline material,^^ the formidable task of structural elucidation arose. Why did this prove so important a labor? The answer of course lies in a consummate desire to separate morphine's equally powerful addictive and analgesic properties. With an established structure in hand, surely in rationalized chemical manipulation of the molecule could one curb its less desirable effects. As a result, morphine may not only be responsible for the birth of alkaloid chemistry, but also the study of structure-activity relationship. Much in keeping with the opiate's medicinal properties, this seemingly innocent molecule held great sway over many a noble mind; Liebig, Knorr, Wieland, Pschorr, Gadamer, Freund, Schdpf, all extensively contributed to an effective understanding of the chemical behavior of morphine. Finally, in 1926,^^ Robinson conjured forth a structural suggestion that has stood fast against all tests and trials, being ultimately crowned by Gates' total synthesis.^^ ^ ^lost excellent treatise of these structural studies, presented in a historically realistic revelationary style, has already appeared. ^^ Starting only with the empirical formula, this account presents itself as a marvelous educational tool, particularly in these times when high field N.M.R. and X-ray instrumentation is so abundant! Herein we only attempt to summarize the major features of the investigations that climaxed in Robinson's proposed structure. - First correct elemental analysis of morphine, C17H19NO3, by Laurent, 1847.^^ - The nature of the oxygens was readily apparent. Diacetylation to heroin by Wright^^ proved the presence of two hydroxyl groups. Grimaux' facile monomethylation to codeine ^^ established the phenohc nature of the first of them. (A reasonable stability to acid ruled out the possibility of a methyl enol ether.) On a more practical note, most subsequent studies in fact proceeded with codeine, in order to avoid potential oxidative and zwitterionic complications inherent in morphine's phenolic nature. The oxidation of
51 codeine to ketonic codeinone showed the presence of a secondary alcohol. The third oxygen, due to its lack of reactivity, was correctly assumed to be an ether linkage. - Hydrogenation, with one molar equivalent of hydrogen over palladium, estabhshed the presence of an isolated double bond. This procedure when combined with monobromination further vahdated aromatic nature of some type. - The compound was quite obviously an amine due to its basic character. Von Braun degradation (NMe -^ NCN) and Hofmann degradation (sequential methylation elimination) impUed a cycMc tertiary amine. Further exhaustive Hofmann degradation led to an unusual net elimination of both trimethylamine and ethylene. (This reaction was driven by aromatization of the C-ring after elimination of the C15-C16-N17 ethylamine bridge). - Rather harsh distillation conditions (from zinc dust) of several derivatives led repeatedly to oxygenated phenanthrenes, thus suggesting a reduced phenanthrene core of some type. Comparison with ingeniously designed synthetic materials allowed assignment of the oxygenation pattern. ^^ - Having established the gross framework of the alkaloid, only the points of attatchment of the ethylamine bridge remained unsolved. Detailed analysis of previous work coupled with further deductive studies allowed Robinson to propose his ultimate structure, which explained and accounted for every prior and subsequent chemical investigation, including total synthesis. - In 1955, Hodgkin^^ pubHshed a crystallographic study of morphine hydroiodide dihydrate, thus revealing important conformational information while verifying the structure. The absolute configuration was later estabhshed by Kartha,^^ via a study of codeine hydrobromide dihydrate. 2e.
Biosynthesis Early biosynthetic studies presented experimental evidence that radio labeled 1-
benzyhsoquinoline systems can be converted by the opium poppy into morphine.^^
52
Through the examination of diverse examples,^^ it is well established that the benzylisoquinoline skeleton originates in two molecules of tyrosine. The first steps of the morphine biosynthetic pathway, outlined in Scheme 1,
HO, HO*
I
1
NH,
— HO' 15 dopamine
13 tryaniine
.CO2H
NHo
CO.
12 L-tyrosinc
^ Q ^ ^'-<^
HO'
14 4-hyroxyphenylpyruvate
16 4-hydroxyphcnylaccialdchydc
MeO. NH
h
HO'
HO'
HO' 17 S-norcoclaurine
19 S-N-mcthylcoclaurine
18 S-coclaurine MeO,
MeO. NMe
HO*^ ^ ^
MeO
20 S-3-hydroxy-N-methylcoclaurine
21 S-rcnculinc
a. L-tyrosine decarboxylase; b. phenolase; c. L-tyrosine transaminase; d. p-hydroxyphenylpyruvate decarboxylase; e. S-norcoclaurine
synthase; f. norcoclaurine-6-0-
methyltransferase; g. tetrahydrobenzylisoquinoline-N-methyltransferase h. phenolase; i. S3'-hydroxy-N-methylcoclaurine-4'-0-methyltransferase Scheme 1
53
involve the condensation of dopamine 15, which is formed directly from tyrosine, with 4hydroxyphenylacetaldehyde to give S-norcoclaurine 17. The enzymes implicated in the formation of S-norcoclaurine have been detected in Berber is cell cultures.22 in particular, five enzymes which are involved in this primary portion of the biosynthetic pathway from L-tyrosine to S-norcoclaurine have been isolated and characterized.23 Following the formation of S-norcoclaurine, hydroxylation at C3 with subsequent methylation yields S-reticuline.24 Although the R form corresponds to the correct stereochemistry of morphine at C9, both R and S isomers of reticuline have been found to act as precursors for thebaine, codeine and morphine and it is therefore apparent that Sreticuline is con-verted to its R form.25 Inversion is best explained by formation of an intermediate 1,2-dehydroreticulinium ion 22, followed by stereospecific reduction to yield R-reticuline, Figure 3. An NADPH-dependent enzyme, 1,2-dehydroreticuline
MeO,
MeO.
MeO.
MeO"
MeO
MeO'
21 S-redculinc
22
23 R-redcuIinc
Figure 3. Isomerization of Reticuline reductase, recently discovered in seedlings of the opium poppy,26 is believed to be responsible. The next step involves an intramolecular ortho-para phenolic coupling of Rreticuline to form the crucial C12-C13 bond of morphine, a process initially proposed on theoretical grounds by Barton and Cohen.27 Some eight years later, experimental proof was obtained for this regioselective oxidative coupling from in vivo studies.2^^ Further
54 experiments have demonstrated without a doubt that microsomal preparations containing cytochrome P-450 catalyze the tranforaiation to salutaridine 25, shown in Figure 4.28.29
MeO,
MeO
25 salutaridine Figure 4. Regioselective Oxidative Coupling The final step for the completion of the pentacyclic ring system of the morphine alkaloids involves the closure of the C4-C5 ether bridge. Synthetic salutaridine has been transformed in vivo to thebaine, codeine, and morphine in Papaver
sominiferum.
Reduction of the dienone was proposed,^^ and indeed, hydride reduction of salutaridine produces a mixture of two epimeric alcohols. However, upon feeding 7-^H labelled salutaridinol epimers to Papaver somniferum, only the S-isomer was converted into thebaine; an indication that the reaction is enzyme mediated.^^'^2 xhis finding contradicted earlier work by Barton in which the configuration of the alcohol was assigned as R (See description of Barton's synthesis, ref. 50). MeO.
pH«9
NMe MeO'
25 salutaridine Figure 5. Reduction of Salutaridine
55 Salutaridinol 26, as drawn in Figure 5, contains the correct configuration for an allylic syn displacement to undergo ring closure and form thebaine. Thebaine, after enol ether cleavage, yields neopinone 27 which has been shown to exist in a chemical equilibrium with codeinone 28 in aqueous solution as well as under the conditions of plant metabolism.3^
Codeinone:
NADP oxidoreductase in the presence of NADPH
MeO,
MeO.
NMe
NMe
MeO'
27 neopinone
8 thebaine
MeO.
MeO.
NMe
NMe
28 codeinone
2 codeine
1 morphine
Figure 6. Codeinone Reductase/NADPH Pathway
drives this equilibrium towards codeine. Figure 6. Seemingly, the transformation of neopinone 27 to codeinone 28 occurs spontaneously. The fmal step of the biosynthetic pathway simply involves o-demethylation to afford morphine.^"^
56 3.
Total and Formal Syntheses
As a result of the extensive semisynthetic studies carried out during the structural and biosynthetic elucidation, the subtle chemistry of the morphine skeleton is well understood and documented. Despite this wealth of information, the total synthesis of morphine proved to be an arduous task, occupying the attention of several formidable research groups for many years, often with no reward. The chief difficulties remain in the efficient formation of the C4-C5 secondary ether linkage and the C13 quaternary center, not withstanding the inclusion of the elements of absolute stereochemistry. Hence, since Gates' landmark total synthesis of 1952,^2 most successful approaches have either intercepted his route or culminated with known compounds previously converted to the alkaloid. This unfortunately has resulted in little change in endgame strategy, which perhaps would benefit greatly from fresh input. To date, possibly only one synthesis, the work of Rice,2^'^^ would be amenable to large scale production of the drug on a commercially viable basis. However, the sum of knowledge thus generated has proven a worthy tribute to both the chemists involved and to the captivating allure of morphine. In this section, an overview of the total syntheses thus far reported is presented in chronological fashion. The main features of each successful approach are summarized in Table 11. As can readily be seen, the true total synthesis, culminating in morphine itself, is rare; more popular is the formal total synthesis, where some key intermediate, previously converted to morphine, is intercepted. At this point it is necessary to clarify what the authors of this review will regard as a total synthesis of morphine. For example, one could in fact consider the successful demethylation of codeine as a synthesis of morphine, but synthetically "trivial" problems such as these, although indeed important and challenging in terms of drug production and/or interconversion, will be ignored here. At the opposite end of the synthetic pathway, total syntheses of, say, reticuline, although
57
converted to morphine on several occasions, will similarly be overlooked, as a crucial question, that of C13 stereocontrol, is obviously not addressed.
Indeed, several
monographs pertaining to the synthesis of these simpler alkaloid systems already abound.^ So, with regards to this review, and for our own purposes, we will consider the following as evidence of a "total synthesis": the synthesis of a naturally occuring opiate alkaloid which possesses the complete stereochemically correct morphine azacarbon skeleton and includes the C4-C5 ether bridge; either morphine itself, or some other natural opiate which has previously been converted to morphine by standard, well documented procedures, e.g. codeine, thebaine, and others. We will consider the synthesis of naturally occurring morphinans without the C4-C5 ether bridge (e.g. salutaridine, thebainone) although previously converted to higher order opiates, as formal syntheses, as indeed we uphold and insist that the successful closure of the C4-C5 ether linkage is a crucial element of any truly satisfactory synthetic design. Finally, as expected, interception of any intermediate (natural or unnatural), utilized in previous total and formal syntheses, will also be considered a fomal synthesis, if so claimed. On closer comparison of these successful routes, several distinct styles are apparent. The early approaches of Gates and Ginsburg are truly "synthetic", effectively working blindfolded due to their reliance upon knowledge gained solely from the extensive, but essentially crude, degradative structural studies; in fact, as a result, morphine synthesis has probably offered as much to phenanthrene, and in turn to steroid chemistry, as to actual production of the alkaloid itself! Eventual biosynthetic postulations and revelations have led several groups over the years to approach the synthesis of morphine as does Nature, choosing to adopt a biomimetic phenoUc oxidative coupling as the key step. Unfortunately, the in vitro oxidation of reticuline derivatives to salutaridines tends to proceed with poor regiocontrol and in low yield, despite employment of a variety of novel reagents. This of course
58 directly reflects the marvelous specificity of Nature's enzyme mediated processes. The four possible phenolic oxidative coupling products are shown in Figure 7. MeO,
MeO.
MeO'
MeO'
29 corytubcrinc (onho-ortho)
30 isoboldine (ortho-para)
OH MeO^
^^
MeO^
NMe MeO*^ >f
0 31 isosalutaridine (para-para)
NMe MeO
d 25 salutaridine (para-ortho)
Figure 7. Possible Products from Oxidative Coupling
More interesting in the concept of this review is the utiUzation of "bioanalogous" routes to opiates, in which the reactivity and regioselectivity of the systems involved are tempered and enhanced by judicious inclusion of alternative functionalities. Investigations of this type were inaugurated by the initial discoveries of Grewe, in his original and brilliant construction of the parent morphinan system,^^ summarized in Figure 8.
5,6,7,8-Tetrahydroisoquinoline 32 was N-methylated and the resulting
pyridinium salt 33 treated with benzyl Grignard to yield dienamine 34. Reduction of the less substituted double bond gave precursor 35, which cyclized under acid catalysis (the Grewe cyclization) to simple morphinan 37. As can be seen from the conformational
59 representation
3 5 , the preexisting stereocenter yields the required product
stereochemistry.
^
McI N
BnMgCI "
32
^
N* Mel'
NMe
33
NMe
H3PO4
r-^^^P^NMe
NMe
35 Figure 8. Grewe Cyclization
This now classic name reaction has over the decades proved amenable to both synthetic improvements and introduction of increased substitution,^'^ resulting in a formal synthesis of morphine itself, via dihydrothebainone, as reported in 1967.^^ The same approach was almost simultaneously reported by Morrison, Waite, and ShaveP^ (vide infra) and adapted in a later practical synthesis by Rice.^^ Further related examples of Grewe-type approaches are summarized in Section 4 of this review. In more recent years, a return to original and imaginative syntheses became apparent, yet one still cannot fail to notice, even at this stage, vague similarities in overall synthetic design; echoes of Fuchs' approach in that of Parker for instance, or Overman's essentially Grewe-type strategy. Most of these recent approaches employ "state-of-theart" methodologies, such as vinyl sulphone additions, tandem radical cycHzations, or palladium mediated intramolecular annulation, respectively. In addition, many are
60 potentially amenable to enantiocontrol. However, the expensive, sensitive technologies involved most likely preclude their use in truly practical, large scale routes. Table II offers a glimpse at the actual yields of the total syntheses listed in this review. The yields are calculated from the numerical values provided in the publication. Where no yield for a particular step is mentioned an assumption of 50% has been made.
Table 11. Summary of Reported "Total" Syntheses Steps:
Starting Materials/Final Product
Author Year
Yield(%)
Gates
29; NM«
1952
0.00112%
HO*
Ginsburg
:x)
22;
1954
NMe
0.0137%
M«0.
Barton
3;
1963
< 0.012% M©0'
MeO
by radiodilution
61
Morrison, Waite &
"•°xi MeO'
COjH
Shavel MeO
1967
NMe
NH,
"^
(for cyclization step)
Grewe 1967
::A
6;
CO2H
MoO.
NMo
NH,
3% (for cyclization step)
MeO,
Kametani 1969
MeO,
BnO*
3; NMe
V
MeO'
2.8 X 10-5
MeO'
OMe MeO,
Schwartz
4; CO2H
1975
HoA^
^
NMe
7.6%
COjMe
OBn MeO,
5;
Beyerman 1979
BnO'
X5^""
MeO
NMe
28.3%
62
MeO,
Rice 1980
MeO,
MeO CO.H MeO,
^=>
9;
o' NMfl
NH,
Evans
29%
11;
1982
NMe
17.5^
MeO.
Rapoport
MoO, HO'
1983
MeO,
1983
1.13%
l^^A^NMe
CHO
White
18;
MeO'
MeO.
6;
HO* NMe
X^'
MeO*
2%
HO*'
MeO,
Schafer
15;
1986
NMe
^1 NBu
MeO
3.0%
63
MGO.
MeO,
Fuchs
Br
HO'
16;
Br
1987
OH TBSO.
NMa
3.29%
^^-""^SOjPh
Tius
18;
1992
r^'^NHCOoMe
NMe
2.66%
MeO,
Parker
13; 9.42%
1993 NMo
MeO"
Overman
14;
1993
6.47% NMe
64 3a.
Gates, 195212,41,45 Although adequately summarized in several other critical texts,^^ only after a
detailed examination of Gates' pioneering work can one fully appreciate this monumental acheivement in the field of synthetic organic chemistry. To convert a common dyestuff intermediate^ 1 to one of the most complex and significant medicinal alkaloids, the structure of which was still a subject of debate at the time, was and still remains a truly amazing accomplishment and compares favorably with such achievements as Woodward's later reserpine synthesis.'^^ x^at so many subsequent synthetic ventures have intercepted Gates' route is ample testament to the depth of understanding apparent in this approach. Despite morphine's phenanthrene core. Gates' approach started with manipulation of a naphthalene diol template 37, which after monoprotection, expected a-nitrosylation at the less hindered site, and reduction, gave amine 39, Scheme 2,^^ An iron (III) HO.
OBz MeO,L
^/>„^^
•^JL MeO.
COsEt
Conditions: (a) i. BzCl, pyridine; ii. NaN02/AcOH; (b) H2, Pd/C; (c) FeCls; (d) i. SO2; ii. (MeO)2S02, K2CO3; (e) i. KOH; ii. NaN02/AcOH; iii. H2, Pd/C; iv. FeCls; (f) i. Et02CCH2CN, NEt3; ii. K3Fe(CN)6; (g) KOH; (h) butadiene, A. Scheme 2
65 mediated oxidation to an orthoquinone was followed by reduction and methylation to give dimethoxynaphthalene derivative 41. Saponification and repetition of this series of functionalizations afforded orthoquinone 42.
A Michael-type addition of ethyl
cyanoacetate, followed by reoxidation, hydrolysis, and decarboyxlation gave the necessary dienophilic component 43 for a simple intermolecular Diels-Alder reaction with butadiene to complete construction of ring C. An unusual catalytic hydrogenation, developed somewhat unexpectedly during model studies,^^ bestowed morphine's azacarbocyclic skeleton, although stereochemically incorrect at C14. The following mechanism to explain this fortuitous cychzation has been advanced. Figure 9M MeO.
MeO^
MeO. MeO
NH
Figure 9. Reductive Construction of Ring D
A modified Huang-Minion procedure, followed by N-methylation and amide reduction, yielded (±)-[3-A6-dihydrodesoxycodeine methyl ether 52, which after resolution proved identical to material derived from natural sources^^^'^ via manipulation of P-thebainone, and consequently served as a synthetic relay. (It is within an obscure portion of this series of rigorous structural studies and proofs that Evans later intercepted Gates' route and thus formaUzed his synthesis).
66 Having established the identity of synthetic material, Gates turned to the introduction of a Ce oxygen functionality. A series of unsuccessful studies culminated in a low yielding acid-catalyzed hydroxylation. Selective demethylation and oxidation gave P-dihydrothebainone 54, Scheme 3, and most other subsequent formal syntheses intercept MeO.
y^ ii ^^%^ II
^
45
MeO'
1 1\
^^
» i
^
J 'H
k>^ MeO^
C
52
f^\llII
V
HO"^
1 'H 0*^
54
Conditions: (a) H2, copper chromite; (b) i. KOH, N2H4; ii .NaH, Mel; iii .LiAlH4; (c) i. dibenzoyl tartrate resolution; ii. H2SO4/H2O; (d) i. KOH, ethylene glycol, 225 °C; ii. KO^Bu, Ph2CO. Scheme 3
at this point. Now, in Gates' own words, "... inversion at C14 to produce the cis fusion of rings II [B] and III [C] loomed large." Bromination followed by conversion to the 2,4-DNP hydrazone resulted in a cascade of events which eventually led to the correct Ci4 stereochemistry. However, it is in the proof of this epimerization, by thorough and extensive comparison with several other opiate alkaloids, that the truly complex character of this task comes to the fore. Having arrived at the final carbocyclic core of the molecule, an efficient closure of the C4-C5 ether bridge was required, but this proved impossible with several advanced systems, presumably because of unfavorable
67 stereochemical and conformational factors. In the end, it was necessary "...with some reluctance..." to reduce the enone moiety 57 to its corresponding ketone, Scheme 4.
MeO,
MeO.
C
NMe
'H
NMe
HO' NMe
NMe
56 Br
NMe
NMe
NMe
NMe
HO* Conditions: (a) Br2; (b) 2,4-DNPH, A; (c) HCl, aq. acetone; (d) i. H2, Pt02; ii. Br2 (2.0 eq.); iii. 2,4-DNPH; (e) HCl, aq. acetone; (f) LiAlH4; (g) pyHCl, 220 °C. Scheme 4
68 Treatment of the ketone with two equivalents of bromine, followed by hydrazone formation, finally effected closure of the ether bridge with concomitant elimination to hydrazone 58.^^^ Hydrolysis to the enone, followed by reduction, removed both the aryl bromide and introduced the alcohol stereoselectively to give natural codeine 2,^^'^*^ which was demethylated under established conditions'^ to yield morphine 1.^2,45 ^ summary of these studies can be divided into synthesis of 50,"*^ Cu-chromite reductive cyclization,"^^^ isomerization of C\4,^^^ reduction to codeine,"*^^ and a full paper sumary."^^^ 3b.
Ginsburg, 1954^7 A close second to Gates' momentous acheivement, Ginsburg set a lasting trend by
intercepting dihydrothebainone 75, thus rendering his synthesis formal. Despite an elegant stepwise construction of the ABC ring system, the closure of the C15-C16-N bridge proceeded with some element of luck. Initial condensation of cyclohexanone with ortho-hihiaitd veratrole, followed by elimination and an ingenious ally lie oxidation yielded enone 64, Scheme 5. Michael
X)
MeO. MeO'
60 62
61 MeO.
MeO.
MeO'
MeO'
0=N
HO-N
CI 63a
63b (continued)
69 (continued) MeO,
MeO^
70 (continued)
MeO,
^QQ
NMe
NH
74
75
Conditions: (a) i. "BuLi; ii. oxalic acid, toluene, A, Dean-Stark; (b) NOCl (from namylnitrite and HCl); (c) i. py, A; ii. aq. H2SO4, A; (d) dibenzyl malonate, KO^Bu; (e) i. H2, Pd; ii. A; (f) HF; (g) ethylene glycol, pTsOH, A, Dean-Stark; (h) i. AmONO, NaOEt; ii. H2, Pd, HCl; (i) acetoxyacetyl chloride, 2 eq. py , CHCI3; (j) ethylene glycol, pTsOH, A, Dean-Stark; (k) i. AmONO, NaOEt; ii. aq. acid; (1) hydrazine, ethylene glycol, A; (m) i. aq. acid, A; ii. LiAlH4; (n) i. CH2O, HCO2H, A; ii. benzophenone, KO^Bu; iii. resolution (d-tartaric acid). Scheme 5 addition, decarboxylation and Friedel-Crafts annulation led to the ABC tricycle 67. A selective protection-deprotection sequence allowed introduction of the amino group exclusively at C9, followed by a smooth acid chloride condensation which afforded acetoxyamide 70. Prior to reduction of these functionalities, an attempt to block C5 as its ketal gave unexpected results. Selective cleavage of the C4 methyl ether, not uncommon in higher morphine analogues, was accompanied by a spontaneous formation of the ethylamine bridge with the correct C13 quaternary center, accompanied by an undesired Cio ketahzation resulted in 71, containing the complete morphine carbon connectivity. Introduction of the required C6 ketone while removing those at C5 and Cio, reduction of the lactam followed by a necessary N-methylation and oxidation gave (±)dihydrothebainone 75. This was resolved to the 1-form with d-tartaric acid, and thus the synthesis was rendered formal."^^
71 3c.
Barton, 196349 Barton's approach was truly biomimetic, as it pursued an initially described
conversion of labelled reticuline 76 to radioactive morphine 77, thus supporting the proposed biosynthetic route, Figure 10. MeO^
papaver N^^Me
somniferum
N^*Me
MeO'
Figure 10. Transformation of Labelled Reticuline to Morphine An attempt to emulate this pathway in vitro followed, unfortunately hampered by a dismal Mn02 promoted oxidative coupling. However, a radioisotope dilution study did indeed suggest a 0.012% conversion to radioactive salutaridine 79, Scheme 6.
NMe
NMe MeO'
MeO'
NMe
NMe
Conditions: (a) Mn02, 0.012%; (b) NaBH4; (c) < pH 4, rt. Scheme 6
72
Reduction to a mixture of the epimeric diols, followed by acid catalyzed allylic displacement by the phenolic group, yielded radioactive thebaine 81 and thus formahzed "...a third and long sought synthesis of the morphine alkaloids...".^^ 3d.
Morrison, Wake, Shavel,
196739and Grewe, 196738
Pubhshing almost simultaneously with Grewe, this group adopted a bioanalogous synthesis starting from a substituted benzyltetrahydroisoquinoline 84, which underwent selective Birch reduction and monodeprotection to 85, Scheme 7. Treatment with acid
NMe
88 para coupled Conditions: (a) i. amidation; ii. Bischler-Napieralski; iii. reduction; methylation; (b) Na/^BuOH/NHs; (c) 10% aq. HCl, A, ^-3% andp-37%. Scheme 7
73
(refluxing 10% hydrochloric acid as reported by Morrison, Waite, and Shavel; phosphoric acid in the work of Grewe) causes enol ether hydrolysis and effects cychzation to both the para-coupltd flavinantine derivative 88 (37%) and dihydrothebainone 75 (3%), thus intercepting Gates' synthesis. The low regioselectivity of this process directly contrasts the effectiveness of the enzyme mediated biosynthesis.
3e.
Kametani, 196951 In this approach, a Pschorr-type cyclization was adopted in order to maximize
ortho-para
selectivty.
Diazotization of 2-aminobenzyltetrahydroisoquinoline 89
followed by thermal decomposition yielded racemic salutaridine 25 in a meager yield of 1.1%, Scheme 8. Other products isolated from the cychzation included benzaldehyde and
MeO.
MeO.
BnO NMe
NMe MeO'
MeO'
O
OMe MeO,
MeO,
NMe MeO'
25
NMe MeO
Conditions: (a) i. NaN02, H2SO4/ACOH; ii. 70 °C, (b) NaBH4, MeOH; (c) IM HCl. Scheme 8 laudanine 11 (3.5%), but no products of ortho-ortho couphng, which have been observed when zinc powder was employed.^^ Reduction followed by acid catalyzed ring closure
74
in the method of Barton^^^ gave racemic thebaine 8, and thus constituted a formal total synthesis of morphine. Resolution of 89 with di-p-toluoyl tartaric acid afforded entry to both enantiomeric series, allowing comparison of ORD/CD data.^^'^^
3f.
Schwartz, 197554 Schwartz successfully emulated the in vivo para-ortho coupling of N-
acylnorreticuline derivatives by the use of thallium tristrifluoroacetate (TTFA), Scheme 9. Treatment of ethoxycarbonyl derivative 91 with one equivalent of the salt afforded a
MeO^
MeO'
Conditions: (a) 1 eq. T1(TFA)3, CH2CI2, -78 to -20 "C; (b) LiAlH4; (c) IM HCl. Scheme 9
23% yield of the corresponding salutaridine 92. Once again, reduction followed by acid treatment yielded racemic thebaine and thus formalized syntheses of codeinone,^^ codeine,45b and morphine.46
75 This approach was later extended to an enantioselective synthesis,^"^^ Scheme 10. Unfortunately, some racemization occurred during preparation of substrate 94, but not during cyclization to 9 5 , this time mediated by iodosobenzene diacetate. Reduction/closure, followed by Barton radical decarboxylation yielded the known thebaine analogue 97.^6
MeO. .COoMe
MeO'
'^l"'''^
94
OH MeO,
MeO,
NCOjMe MeO
MeO
Conditions: (a) i. 3-benzyloxy-4-methoxyphenyl acetic acid, 1,1-carbonyldiimidazole, THF; ii. H2, Pd/C, EtOAc; iii. Me02CCl, CH2CI2, EtsN; iv. POCI3, MeCN; v. NaBH4, MeOH; vi. Me02CCl, Na2C03, MeOH; vii. Na2C03,H20; (b) PhI(OAc)2, TFA; (c) i. NaBfLt; ii. N,N-DMF dineopentenyl acetal; (d) Barton decarboxylation.^^^ Scheme 10 3g.
Beyerman, 1979^7 Beyerman has adopted a classic Grewe type bioanalogous approach, neatly
avoiding the problems of ortho-para temporary element of symmetry.
regioselectivity by introduction of a subtle
76 In the culmination^'^ of a series of related studies,^^ by now standard procedures yielded the chiral mono-Birch reduced benzyltetrahydroisoquinoline 100, Scheme 11.
OBn
OBn
MeO,
MeO.
BnO'
BnO'
NMe MeO'
MeO
r-"^P^NMe
NMe
M e O - ^ ' ^ ^ -'^
NMe NMe 75 Conditions: (a) CH2O, H2, Pt/C; (b) U/NH3, ^BuOH; (c) HCl, Et20,; (d) 5-chloro-lphenyltetrazole, K2CO3, DMF, 70 °C; (e) H2, Pd/C, 50-55 °C. Scheme 11
This substance underwent a smooth cycloalkylation, in aknost quantitative yield, to give morphinan 101. All that remained was the selective removal of the unwanted C2
77
hydroxyl. Presumed steric shielding indeed allowed a selective etherification to yield tetrazole derivative 102; a simple catalytic hydrogenolysis^^ afforded (-)dihydrothebainone 75, and thus a formal synthesis.^^^ The use of elevated temperature (55 °C) during fmal hydrogenolysis of the tetrazole seems to be critical since de Graw, arriving at the N-carbomethoxy derivative of 102 virtually via the same synthetic sequence, was not able to remove the C2 oxygen when performing the cleavage at room temperature.^^
3h.
Rice, 19802b,35 Rice employed bromo derivative 108, Scheme 12, to avoid undesired para
Ti
a
MeO.
HO'
iT'
NH,
MeO'
104
MeO.
MeO.
NCHO
MeO-^^^^jjjg
VO
^eOv^^^^jv^Br
^eO\^;^v^Br
^ i—\ ^^^^^>f^^NCHO \,0
106
•^
^•^
j ~ \
['''^'''^i^^
107
108 (continued)
78 (continued) MeO,
MeO. g
h
HO
HO' NMe
NCHO
MeO,
NMe
111
Conditions: (a) i. 200 T ; ii. MeCN, POCI3, reflux; iii. NaCNBHa, 65%-MeOH, pH 4-5, A; (b) Li/NHs, ^BuOH; (c) i. 1.5 eq. PhOCHO, A; ii. cat. CH3SO3H, ethylene glycol, THF, quant; (d) N-bromoacetamide, 0 T ; (e) HCO2H/H2O; (f) 14% NH4F-HF, CF3SO3H, 0 "C; (g) 10:1 MeOH, aq. HCl, A; (h) H2, Pd/C, 2N AcOH, HCHO, NaOAc, quant; (i) i. Br2, AcOH; ii. CHCI3, IN NaOH; iii. as (h). Scheme 12 coupling. The idea of blocking C2 was poineered by Beyerman^^^ (C2-Me), who proposed^^g blocking with C2-halogen. Readily available amine 104 was subjected to Birch reduction and protection to give 106, prior to treatment with N-bromoacetamide at 0 °C. Deprotection afforded ketone 108 which underwent cyclization smoothly to give 109 in 60% yield. Hydrolysis and a one-pot hydrogenation to effect both reductive Nmethylation (using formaldehyde) and debromination led to dihydrothebainone 75, intercepting Gates' route. Alternatively, bromination, ether ring closure, and the same
79
hydrogenation protocol yielded dihydrocodeinone 111. This whole synthesis required isolation of only six intermediates, obtained sufficiently pure for immediate further use, and proceeded in 29% overall yield. It remains as the most practical preparation of morphine to date.^^'^^'^^ 3i.
Evans, 1982^2 Evans' initial steps are reminiscent of Ginsburg's, as orr/io-lithiated veratrole 115
was coupled with piperidone 116 to give after dehydration alkene 118, Scheme 13. COjH
^ :
CO2H
112 MeO^
M e O ^ ' ^ f ^ 115
^-^
Me 116
N^Me CIO4-
122 (continued)
CHO
80 (continued) MeO.
MeO.
MeO'
MeO' NMe
NMe
MeO. MeO' NMe
NMe
Conditions: (a) i. B2H6»THF, 25 "C; ii. PBrs, HBr (48%), CH2CI2; (b) ZnBri, C6H6, A; (c) Et20, 0 °C; (d)/?TSOH, toluene, A,; (e) i. "BuLi, THF, -10 T ; 114; ii. Nal, K2CO3, MeCN, A, (f) HCIO4, Et20, MeOH, ii, MeOH, 50 "C; (g) CH2N2, CH2CI2; (h) DMSO; (i) BF3»Et20, toluene, -10° C, (j) i. MsCl, NEts; ii. LiEtsBH; iii. OSO4, NaI04, THF, aq. AcOH. Scheme 13
Deprotonation to the enamine anion, selective coupling with the allylic terminus of dibromide 114, followed by an intramolecular enamine alkylation, afforded reduced isoquinoHne 119. A rather elegant conversion to aminoaldehyde 122 ensued. Immonium ion formation in 119 via protonation with perchloric acid at first yielded the kinetic trans isomer, which underwent equilibration upon reflux in methanol to give the corresponding crystalhne cis product 120. Diazomethane treatment led to aziridinium salt 121, which upon exposure to DMSO, ring opened with concomitant oxidation in a Komblum fashion to the aldehyde 122.^^ Treatment with Lewis acid effected B-ring closure, thus
81 completing the carbon framework. Reduction of the benzylic hydroxyl and LemieuxJohnson cleavage yielded Gates' ketone 124, thus formalizing the synthesis. A C14 epimerization procedure allowed verification by comparison with authentic epimeric 126, although conditions for the actual transformation of 124 to 125 and 126 are not given. 3j. Rapoport, 19836^ For a number of years prior to Evans' revelations, Rapoport had been involved in the development of a general methodology for the synthesis of several morphine structural analogues. These included both cis and trans 4a-aryldecahydroisoquinolines 127,^^ octahydro-lH-benzofuro-[3,2-e]-isoquinolines 128,^^ and novel octahydro-lH[l]-benzopyrano-[4,3,2-e,f|-isoquinoUnes 129:^'^ MeO^
MeO'
NMe
NMe
NMe
127
128
129
Although armed with a wealth of experience in the field,^^ several stereochemical problems proved unavoidable, and Rapoport finally resorted to interception of Evans' route, thus doubly formalizing his synthesis. However, the construction of the key intermediate, via an effective a-methylene lactam rearrangement, is markedly different. Starting from 2-hydroxy-3-methoxybenzaldehyde 130, Scheme 14, standard MeO.
MeO. MeO
EtOsC
(continued)
82 (continued)
"'°V^ J U -1^ MeO' Y
V "Tf^
r T 133
e MeO'
MeC^
A^CO^Et
EtOjC
MeO,
MeO^
.C02Et
COjEl
0^"^ N ^ H 134
CN
MeO,
MeO.
MeO.
MeO'
MeO
MeO
COjEt
S
MeO.
MeO. j, k
MeO'
MeO'
MeO. MeO'
i/-C02H
HOjC
O2CH
NMe
O
Me
140
MeO,
:MeO » O,
'BUO2C
m NMe
141 (continued)
NMe
'BuOgC
0
142
83 (continued)
MeO Y ^ 0
1
1
^-ir.
D
NM^
0 143
MeO'» 1—0, ^—O" k A^
^NMe
^ ^ 144
Conditions: (a) Me2S04, K2CO3, MeOH, A; (b) Et02CCH2C02H, piperidine, py, A; (c) Et02CCH2CN, NaOEt, EtOH, A; (d) i. H2, Pt, EtOH, ii. toluene, A; (e) i. Me30+BF4". CH2CI2; ii. NaBH4, EtOH; (f) CH2O, H2, Pd/C, EtOH; (g) i. NaOH, H2O, MeOH, reflux; ii. glac. AcOH; iii. AC2O, distill; (h) Se02, PhCl, 100 °C; (i) HCO2H; (j) K2CO3, MeOH; (k) i. CH3C(OMe)3; ii. hydrolysis; (1) i. (imid)2C0, CH2CI2; ii. tBu02CCH2C02^Bu, isopropylmagnesium bromide; (m) NaOMe, MeOH; (o) TFA, CH2CI2, 0 °C; ii. toluene, A; (p) pTsOH, HOCH2CH2OH, benzene, A. Scheme 14
procedures led to cinnamate 132, which underwent smooth Michael addition of ethyl cyanoacetate. A subsequent reduction and cyclization furnished lactam ester 134. Selective amide reduction and N-methylation gave nipecotate 136 which underwent rearrangement to a-methylene lactam 137 in quantitative yield. A benzylic oxidationally lie shift, followed by Claisen orthoester rearrangement, led, after hydrolysis, to acid 140.^^ Homologation to the p-keto ester, an intramolecular Michael addition, and finally decarboxylation proceeded without incident to give ketone 143, ^^ protected as its ketal 144.
84 Rapoport then embarked on several attempts to effect B-ring closure via readily available enamine 145, Scheme 15. Direct cyanation gave aminonitrile 146, containing
MeO' -O,
^^^cp NMe
H n
c
r
MeO' O.
MeO* O,
«—O'
NMe
'IN
145
144
MeO
NMe
'''
OMe
147
148
Conditions: (a) DffiAL, 0 'C; (b) KCN, aq.MeOH; (c) i. MeLi, THF 0°C; ii. 0.5 M H2SO4, (d)anh.HF. Scheme 15
the BC trans ring junction but, unfortunately, also the incorrect Cg stereochemistry. Several attempts to redress this proved unsuccessful.
Acid-catalysed closure with
concomitant C9 epimerization failed, but treatment with methyl lithium followed by hydrolysis furnished diketone 147 in good yield, although still epimeric at C9. Attempted acid-catalysed annulation under several conditions surprisingly afforded products of type 148, which underwent the undesired addition to the €5 ketone. The use of more effective protecting groups, N-demethylation, diazotization of amines derived from cyanide reduction, Grignard displacement of cyanide, all proved fruitless.
Rapoport then returned to enamine 145 and, in keeping with Evans'
observations, found that both kinetic and thermodynamic protonation, to trans and cis immonium salts respectively, was possible (see also Scheme 14). However, attempted
85
closure of the derived aziridinium salts, both under thermal conditions and after attempted nucleophilic opening with chloride ion, failed with either ring junction isomer. Finally, after considerable frustration, Rapoport backtracked^"^ to ketone 143, Scheme 16,
MeO, MeO
MeO,
^
-^
MeO.
-^
MeO
^
MeO
L^^A^NMe H \
143
H \
149
119
MeO,
NMe
NMe
Conditions: (a) CH2PPh3, THF, A; (b) DIBAL, THF 0 T . Scheme 16
introduced the required methylene group, and reduced 149 to enamine 119, and in so doing intercepted Evans' successful route. As to why the DMSO-aziridinium salt oxidative technique was not applied to advanced substrates such as 145 or 150 has not been divulged.
3k.
White, 1983^9 In this synthesis, the use of aryhodoso complexes to promote oxidative couphng
was featured, with bromine substitution preventing any undesired para-para coupling.
86 Scheme 17. Protection and bromination of resolved^^ (-)-norreticuline 151,^^ followed
NCOCF3 MeO' MeO.
Br
HO NMe MeO
MeO 156 MeO,
NMe
NMe
Conditions: (a) i. Bri, AcOH; ii TFAA, pyridine; (b) PhI(TFA)2, CH2CI2, -40 "C; (c) K2CO3, aq. MeOH; (d) aq. CH2O (37%), NaBH4; (e) N,N-DMF dineopentenyl acetal, CH2CI2; (f) i. Hg(OAc)2, HCO2H, H2O; ii. HCl then NaOH; (g) LiAlH4. Scheme 17
87 by treatment with a variety of aryliodoso complexes at -40 °C, produced salutaridine analogue 153 in yields up to 21% (typically -10%). Hydrolysis of the trifluoroacetate group, N-methylation and reduction, followed by a mild dehydration, afforded (-)bromothebaine 156. Hydrolysis, double bond migration into conjugation and lithium aluminium hydride reduction of both enone and aryl bromide gave (-)-codeine 2.
31.
Schafer, 1986^0 Although Schafer's approach at first appears equivalent to a straight forward
oxidative coupling, he successfully avoided his previously encountered problems^^ of ortho-para selectivity by utilizing a judiciously functionalized, non-aromatic A-ring moiety. Scheme 18. a-Lithiation of formamidine 160 followed by alkylation with the ^^
MeO
MeO.
OMe b
COjEt 15'^
OMe 0TBDMS
COjEt 1^^ MeO MeO-
"'°X»
c
TBDMSO'''''^^^-'^^^
BnO'
160
^tBu
..^ny'^^^ MeO"*
161
OBn MeO.
MeO. TBDMSO' NMe
NMe
g ^ MeO'
MeO'
0
O
162 (continued)
20
(continued) Conditions: (a) i. H2, Ra/Ni, 110 bar EtOH, 160 T ; ii. Na2Cr207, H2S04/Et20/H20; iii. HC(0Me)3, MeOH, catpTsOH; iv. mCPBA, MeOH, 8-10 °C; v. Cr03-2py, CH2CI2, 20 "C. (b) i. TMSCH2C02Et, cat BU4NF, THF, 20 °C; ii. LAH, Et20, 15 °C; iii. (Me3S)2NH, TMSCl, hexanes; iv. LDA, TBSCl, THF, -78 T , H2O, 0 "C; v. MsCl, EtsN, Et20; vi. LiBr, acetone; 15% overaU; (c) i. LDA, -IS^ C; 1; ii. NaBH4, MeOH; (d) 1.5 eq. SnCU, CH2CI2, 20° C; (e) DDQ, PhH, A. Scheme 18
laboriously synthesized bromide 159 (11 steps) and reductive cleavage yielded precursor 161.
Several cyclization conditions were investigated, the best proving to be 1.5
equivalents of tin tetrachloride at room temperature. DDQ mediated aromatization afforded salutaridine 25 in 3% overall yield over 15 steps.
3m.
Fuchs, 198772,73 Fuchs' elegant approach to the morphinan skeleton utilized an intramolecular
conjugate addition/alkylation sequence, in which connections C12-C13 and C9-C14 were formed as a result of one tandem process.'73,74 Mitsunobu condensation between alcohol 164 (from 2-allylcyclohexene-l,3-dione in five steps and 43% overall yield) and phenol 163 (from isovanilline in 6 steps and overall yield of 40%)'73c gave the ether 165, Scheme 19. TBDMS deprotection followed by an oxidation/reduction sequence set the aryl ether
OH
HO"
.
^'
^
163
,
^ ^SOjPh
164
(continued)
89 (continued)
MeO,
MeO,
Q
165
HO*'''
Br SOsPh
HO>-^C^
TBDMSO^
MeO.
Br
166 MeO.
HO*'
MeO.
MeO,
HO*'
MeO
MeO.
MeO,
MeO*
(continued)
90 (continued)
MeO,
MeO,
NMe
MeO.
NMe
NMe
HO*^
Conditions: (a) BU3P, DEAD, THF; (b) i. (48%) HF, MeCN; ii. CrOs, H2SO4, aq. acetone, 0 T ; iii. DffiAL, THF, -78 to 25 °C; (c) "BuLi, THF, -78 °C; (d) OSO4, NMO, aq. acetone; ii. Pb(0Ac)4, CHCI3; (e) i. MeNHi-HCl, MeOH, NaCNBHs; ii. Me3SiCH2CH20COCl, aq. NaHCOs, (f) i. DMSO, TFAA, CH2CI2 then EtsN, -78 to 20 T ; u. (MeO)3CH, MeOH, pTsOH, 65 "C iii. TEOCCl, aq. NaHCOa; (g) KO^Bu, THF, (h) DDQ, pTsOH, CHCI3, H2O; (i) i. TFA; ii. CHCI3, aq. NaHC03; (j) HCl, Et20, CH2CI2 then 0.2 N HaOH, CHCI3; (k) NaBH4, MeOH; (1) BBr3, CHCI3. Scheme 19
at C5 and alcohol C6 cis., yielding the cycHzation precursor 166 (Attempted cyclization of the trans alcohol derived from 165 resulted in an "inseparable mixture"). Selective metal-halogen exchange at the aromatic ring induced an intramolecular conjugate
91 addition forming the C12-C13 bond, followed by alkylative closure at C14 to complete the A,B and C ring system, 167. Subsequent manipulation of the allyl moiety via oxidative cleavage, reductive amination and protection yielded the trimethylsilyethoxycarbonyl, ester, 169. Swem oxidation was followed by methyl enolether formation, 170, and base elimination of the sulfonyl moiety afforded the diene, 171. Subsequent DDQ oxidation yielded dienone 172 which upon TEOC-deprotection gave, via 1,6-addition, a mixture of racemic codeinone 27 and neopinone 28. Isomerization of the double bond as described by Rapoport and Barber,'^^ followed by reduction afforded (±)-codeine 2. Finally, this material was 0-demethylated following the conditions of Rice^^ to afford racemic morphine 1.
3n.
Tius, 19927 7 In this original and imaginitive approach, a rapid assembly of the phenanthrene
core of morphine, containing a novel non-aromatic A ring, was achieved in an intermolecular Diels-Alder reaction between quinone 173 (prepared from 3-methoxy-2hydroxy benzaldehyde in 7 steps and an overall yield of 35%) and diene 174 (from 1,4cyclohexanedione monoethylene ketal in 2 steps with an overall yield of 30%), Scheme 20. In one of several unsuccesful attempts to aromatize ring A, an unexpected tandem
O
92 (continued)
MeOoCN
MeOoCN
183
182
(continued)
93 (continued)
Conditions: (a) toluene, 100 °C; (b) i. PhSeCl, MeOH; ii. H2O2, THF; (c) aq. HCl, THF; (d) KN(TMS)2, THF, -78 °C, then 2-(p-toluenesulfonyl)-3-phenyloxaziridine, -78 °C; (e) H2, Pd, THF; (f) TFAA, DMSO, CH2CI2, -78 °C; (g) i. BF3»OEt2, -30 °C; ii. Mel, K2CO3, acetone; (h) i. PhSeCl, EtOAc; ii. H2O2, THF; (i) NaBH4, MeOH; (j) i- MeLi, THF, 0 °C; ii. aq. H2CO, NaCNBHs, MeCN; (k) Dess-Martin, CH2CI2; (1) Zn, NH4CI, aq. EtOH; (m) i. DIBAL, THF; ii. aq. HCl; iii. glac. AcOH, 100 °C. Scheme 20
selenocyclization and subsequent oxidative elimination gave urethane-aminal 176. Deprotection and kinetic enolization of the resultant ketone, followed by oxidation with Davis reagent, introduced the C4 oxygen and provided 178. Hydrogenation of the double
94 bond, followed by Swem oxidation of the C4 hydroxyl yielded acyloin 180. In a "...fortunate turn of events...", boron trifluoride-mediated rearrangement induced aromatization with simultaneous closure of the C4-C5 ether bridge. Methylation of the phenol to 181 followed by selenoxide elimination protocol produced enone 182, which was reduced to 183. This allowed cleavage of the carbamate with methyl lithium and reductive amination of the secondary amine afforded 184. The Cg hydroxyl was reoxidized under Dess-Martin conditions to give enone 185 which upon exposure to zinc dust and ammonium chloride underwent reductive cleavage of the aminal with concomitant closure at Cg to yield morphinan 186. Reduction of the Cg carbonyl, followed by acid catalyzed hydrolysis produced p-thebainone which was isomerized at Ci4 under acidic conditions yielding thebainone, 187, thus intercepting Gates' synthesis.
3o. Parker, 1993^8 80 The elegant formal total synthesis of morphine, accomplished by Parker, shows some similarities to that of Fuchs through analogous disconnections. In both syntheses, the core of the molecule was formed as a result of a tandem process; in this case as a result of a radical cascade.^^'80 The inmiediate cyclization precursor 191 was prepared via a Mitsunobu reaction between monoprotected cis-dio\ 189 (prepared in 8 steps from 2-((3-methoxyphenyl)ethylamine) in 47% overall yield) and phenol 188, followed by cleavage of the silyl ether. Scheme 21. The key step, homolytic cleavage of the Ci2-Br
« Me
MeO^ ^.^^
a 189
(continued)
95 (continued)
MeO,
MeO, SPh Br
Q NTs Me
TBDMSO^
H O ^ ^ ^
191
190 MeO,
MeO
SPh
Br
NTs Me MeO,
NMe
NMe
111
Conditions: (a) BU3P, DEAD, THF, 0 T ; (b) 10% HF, MeCN; (c) BusSnH, AIBN, benzene, 130 °C, sealed tube; (d) Li/NHs^BuOH, THF, -78 "C; (e) (COCl)2, DMSO, CH2CI2, -78 "C to 0 °C. Scheme 21
bond in 191, was mediated by BusSnH, and AIBN under sealed tube conditions. The aryl radical closed at C13 to form the dihydrofuran ring, yielding a new radical at C14 which in turn was trapped by the p-carbon of the styrene to give a resonance-stabilized intermediate. EUmination of the phenylthiol group afforded advanced intermediate 192, containing the tetracycHc carbon skeleton, with correct stereochemistry at C5, C13, and C14. Finally, a nitrogen radical anion, generated during cleavage of the tosyl group, was trapped by the C9-C10 double bond "....in an unprecedented closure....", completing ring D and setting the C9 absolute stereochemistry correctly, as in 193. Swem oxidation of
96 alcohol at C^ yielded racemic dihydrocodeinone 111, establishing the formal total synthesis of codeine'^^*^ and morphine 7^
3p.
Overman, 199381 The crucial step of Overman's approach is essentially a Grewe-type disconnection
but involves an intramolecular Heck reaction to complete ring B. An enantioselective reduction of 2-allyl-cyclohexeneone 195 introduced a chiral element. Condensation of the resultant S-alcohol, (196) with phenyhsocyanate, oxidation of the sidechain olefm with osmium tetroxide and acetonide protection afforded 198, Scheme 22. A copper^ «
Ph
N^g/ 194
^ ^
OH
195
196 OCONHPh
OCONHPh
197 SiMejPh
6^
MeO.
NHDBS
199
OBn OMe
200
(continued)
BnO
97 (continued)
MeO.
MeO^
BnO' NDBS
NMe
Conditions: ; (a) 194, catecholborane; (b) PhNCO; (c) i. Os04, R3NO; ii. acetone, acid catalysis; (d) i. THF, -30 "C; ii. "BuLi, Cu(Ph3P)2, 0 °C; iii. PiiMe2SiLi, 0 "C; iv. /7TsOH, MeOH; v. NaI04; vi. DBS-NH2, NaBHsCN; (e) Znl2, EtOH, 60 °C; (f) 10% Pd(OCOCF3)2(PPh3)2, 1,2,2,6,6,-pentamethylpiperidine, toluene, A; (g) i. BF3.0Et2, EtSH; ii. (as camphorsulfonate), 3,5-dinitrophenylperbenzoic acid, CH2CI2, 0 °C; (h) NMO, TPAP; (i) H2, Pd(0H)2, HCHO. Scheme 22
catalyzed suprafacial S N 2 ' displacement of the ally lie carbamate with lithium dimethylphenyl silane, deprotection and diol cleavage furnished the intermediate aldehyde, whose reductive amination with dibenzosuberyl amine
afforded 199.
Condensation of 200 (prepared in 7 steps from isovanillin in an overall yield of 62%) with allylsilane 199 at 60 °C in the presence of Znl2 was followed by iminium ion allylsilane cyclization to yield the advanced isoquinoline intermediate 201.^^ Palladium-
98 mediated coupling connected C12-C13 and afforded morphinan 202 with the correct stereochemistry at C9, C13, and C14. In the final steps, the phenolic oxygen was liberated, the double bond at C6-C7 was epoxidized on the P face, and intramolecular attack of the phenolic hydroxyl completed the dihydrofuran ring.^^ Oxidation, followed by reductive DBS cleavage in the presence of formaldehyde yielded (-)-dihydrocodeinone 111, established the latest reported formal total synthesis of (-)-morphine.
Using (S)-
oxazaborolidine catalyst for reduction of 195 establishes the formal total synthesis of (+)morphine.
4. SvnoDs of Apwoaches to the Ring Svstems of
Authormate
BIZ
Kcv Step
Starting Material
M rphine Final Product
MeO
d Angelo,
1990 Ref 84
M
lhcn 1 4 2 0 __c
65%
OMe
0-OBn
0-OBn
>95% e.e. from chiral Robinson
8 steps
Boger,
1982 Ref 85
Broka, 1988 Ref 86
0
Ho%
13 steps
&d 0
NCOfls
Ms
Ciganek, 1981 Ref 87,88
NMe
~
HO
Ciganek, 1981 Ref 87,88
0 ~
Ciufolini, 1993 Ref 89
T"ao4 steps
OAc
OAc
Hudlicky, I992 Ref 90
4-
to1ucne,
4 steps
*,'
THSO'"' THSO"" homochiral from 9 3 9 0 biooxidation
Jones, 1985
Ref 91
2
Kametani, 1986
OMe
OMe
Ref 92 NC
M;igtiiis,
I00 I
Ref 93 79%
Mamuno.
I993 Ref 94
OMe
I
McMuny.
I984 Ref 95.96
b
iPr0
8z I
W% c.c. from chiral formmidinc
I I
Monkovi ' c , 1973
Rcf 98.99
Monkovi-c, 1975
Ref 1 0 0
8. 0
I
Me0
Monkovi'c, 1978 Ref 101
Me0
4045' C __t
MeO
Noyori, 1987 Ref 102
__c
Me0
MeoB... '.;
"H NMs
Parsons,
1984 Ref 103
Schultz, 1976 Ref 104, 105
@NMe '"OH
PC5
Me0
OH
I Schultz,
1985 ief 106, 107
Shenvi, 1984
Ref 108
Stella, 1977
Ref 109, 110
Stella, 1983
Ref 1 1 1
($
yields connilio;. unknown
,..'
,o-NMeCl
NMe
X = CI. O h . 011
X
= CI. OAc. 011
Bz-y& 3. loo' C. 30 min
Tius,
19 steps
1986 Ref 112
e
Weller, 6 steps
1983 Ref 113, 114
0
e N' I
Me. I
Dalton I Costanzo 1988 Ref 1.8 (Appendix)
Hudlicky 1992 Ref 115
HO
13 stcps
Several steps
C02E1
I
Hudlicky 1993 Ref 115
Hudlicky 1994 Ref 115
I
107 5.
Conclusions This review attempted to collate all of the design elements inherent in the various
existing approaches to morphine. The authors hope that this information presented together in one document will make it easier for potential future contributors to this field to review the hterature and augment the existing approaches with their own. After nearly forty years of serious effort, it is evident that the field of morphine synthesis is still wide open. The pioneering synthesis of Gates and the most efficient one by Rice are accompanied by other ingenious approaches. What remains before the organic chemical community is the design and implementation of a truly practical approach.
6.
Acknowledgments The authors are grateful to Mallinckrodt Speciahty Chemicals for support of the
research work regarding their own approaches to morphine. We thank Kenner Rice (NIH) for reading the manuscript and for providing information connected to the use of morphine and derivatives and Professor David R. Dalton of Temple University for sharing with us a copy of a recent dissertation. Scott Richardson of Mallinckrodt Specialty Chemicals is acknowledged for sharing with us a review of morphine synthesis.
108 7.
1.
References
For a discussion of chemical dissonance and consonance see H.N.C. Wong, M.-Y. Hon, C.-W. Tse, Y.-C. Yip, J. Tanko, and T. HudUcky, Chem. Rev., 89, (1989), 165-198, and reference 19 therein, pertaining to an unpublished review on this topic by D.A. Evans.
2.
a. P.T. White and S. Raymer "The Poppy," National Geographic, Feb. 1985, vol. 167, pp. 142-188. b. See also K.C. Rice in The Chemistry and Biology of Isoquinoline Alkaloids, Phillipson et al, ed.; Springer-Verlag: Berlin, 1985, pp 191-203.
3.
C.E. Terry, and M.Pellens, The Opium Problem, Ch. 2, Bur. Soc. Hyg., New York, 1928.
4.
U.S. Drug Enforcement Agency, "Controlled Substance Aggregate Production Quota History," Federal Register, July 7, 1994.
5.
National Narcotics Intelligence Consumers Committee Report 1993, "The Supplies of Illicit Drugs to the United States," DEA-94066, August 1994.
6. 7.
O. Hesse, Ann., 220, (1883), 203. a. F. Santavy, Alkaloids, New York, 17, (1979), 385; b. For annual updates on chemistry of morphine, see Specialist Periodical Reports, Chem. Soc. London, "The Alkaloids", vol 1-13, (1970-1983).
8.
J.F. Deronse, Ann. Chim., 45, (1803), 257.
9.
M.A. Seguin, Ann. Chim., 92, (1806), 225.
10.
F.W.A. Sertiimer, Trommsdorjfs J. Pharm., 14, (1806), 47.
11.
J.M. GuUand and R. Robinson, Proc. Mem. Manchester Lit. Phil. Soc, 69 (1925), 79.
12.
M. Gates and G. Tschudi, J. Am. Chem. Soc, 74, (1952), 1109.
13.
D. Ginsburg, The Opium Alkaloids, Interscience Publishers: New York, 1962.
14.
A. Laurent, Ann. Chim. Phys., 19, (1847), 359.
15.
C.R.A. Wright, J. Chem. Soc, (1874), 1031.
16.
E. Grimaux, C. R. Acad. ScL, 93, (1881), 591.
17.
For an excellent overview of Pschorr phenanthrene synthesis see D.F. DeTar, Organic Reactions, 9, (1957), 409.
18.
M. Mackay and D.C. Hodgkin, J. Chem. Soc, (1955), 3261.
19.
G. Kartha, F.R. Ahmed, and W.H. Barnes, Acta Cryst., 15, (1962), 326.
109 20.
a. A.R. Battersby, R.C.F. Jones, and R. Kazlauskas, Tetrahedron Lett., 22+23, (1975), 1873, and references therein, b. M.L. Wilson and C.J. Coscia, J. Am. Chem. Soc.,97, (1975), 431.
21.
a. A.R. Battersby, R. Binks, R.J. Francis, D.J. McCaldin, and H. Rauz, / . Chem. Soc, (1964), 3600. b. D.H.R. Barton, G.W. Kirby, W. Steglich, G.M. Thomas, A.R. Battersby, T.A. Dobson, and H. Ramuz, J. Chem Soc, (1965), 2423, and references therein.
22.
M. Rueffer and M.H. Zenk, Z Naturforsch, 42c (1987), 319.
23.
a. T. Frenzel and M.H. Zenk, Phytochem., 29, (1990), 3505. b. R. Stadler and
24.
R.B. Herbert, The Biosynthesis of Secondary Metabolites, Chapman and Hall:
M.H. Zenk, Liebigs Ann. Chem., (1990), 555. London, (1989), 144. 25.
a. A.R. Battersby, D.M. Foulkes, and R. Binks, J. Chem. Soc, (1965), 3323. b.
P.R. Borkowski, J.S. Horn, and H. Rapoport, J. Am. Chem. Soc, 100,
(1978), 276. 26
W. De-Eknamkul and M.H. Zenk, Phytochem., 31, (1992), 813.
27.
D.H.R. Barton, and T. Cohen, Festschrift A., StoU; Birkhauser Verlag: Basel, (1957), 117.
28.
a. R. Gerardy and M.H. Zenk, Phytochem., 32, (1993), 79. b. For a related study, see: R. Stadler and M.H. Zenk, J. Biol. Chem., 268, (1993), 823.
29.
a. T. Amann and M.H. Zenk, Tetrahedron Lett., 32, (1991), 3675. b. M.H. Zenk, R. Gerardy, and R. Stadler, J. Chem. Soc. Chem. Comm., (1989), 1725.
30.
a. A.R. Battersby, Proc. Chem. Soc, (1963), 189. b. G. Stork, The Alkaloids, Vol 6; R. Manske and H. Holmes EDs.; Academic Press, Inc.: New York, (1960), 219.
31.
R. Gerardy and M.H. Zenk, Phytochem., 34, (1993), 125.
32.
H. Lotter, J. Gollwitzer, and M.H. Zenk, Tetrahedron Lett., 33, (1992), 2443.
33.
J. Gollwitzer, R. Lenz, N. Hampp, and M.H. Zenk, Tetrahedron Lett., 34, (1993), 5703.
34.
G.W. Kirby, S.R. Massey, and P. Steinreich, /. Chem. Soc, Perkin Trans. 1., (1972), 1642.
35.
K.C. Rice, J. Org. Chem., 45, (1980), 3135.
36.
R. Grewe and A. Mondon, Chem. Ber., 81, (1948), 279.
37.
For a review see: D.C. Palmer and M.J. Strauss, Chem. Rev., 11, (1977), 1.
[10 38.
a.
R. Grewe and W. Friedrichsen, Chem. Ber., 100, (1967), 1550. b. R.
Grewe, H. Fischer, and W. Friedrichsen, Chem. Ber., 100, (1967), I . e . R. Grewe and H. Fischer, Chem. Ber., 96, (1963), 1520. 39.
G.C. Morrison, R.O. Waite, and J. Shavel, Jr., Tetrahedron Lett., (1967), 4055.
40.
I. Fleming, Selected Organic Syntheses, Wiley Interscience: London, (1973).
41.
M. Gates, J. Am. Chem. Soc, 72, (1950), 228.
42.
R.B. Woodward, F.E. Bedar, H. Bickel, A.J. Frey, and R.W. Kierstead, Tetrahedron, 2, il95S), 1.
43.
M. Gates, R.B. Woodward, W.F. Newhall, and, R. Kunzli, J. Am. Chem. Soc, 72, (1950), 1141.
44.
G.M. Badger, J.W. Cook, and G.M.S. Donald, J. Chem. Soc, (1951), 1392.
45.
For a more complete account of these studies, see also: a. M. Gates, G. Tschudi, J. Am. Chem. Soc, 78, (1956), 1380. b. M. Gates, J. Am. Chem. Soc, 75, (1953), 4340. c. M. Gates and R. Helg, J. Am. Chem. Soc, 75, (1953), 379. d. M. Gates, and G. Tschudi, J. Am. Chem. Soc, 72, (1950), 4839. For information concerning the preparation of dihydrodesoxycodeine methyl ether as a synthetic relay, see: e. L. Small and G.L. Browning, J. Org. Chem., 3, (1939), 618. f. G. Stork, J. Am. Chem. Soc, 74, (1952), 768.
46.
H. Rapoport, C.H. Lovell, and B.M. Tolbert, J. Am. Chem. Soc, 73, (1951), 5900.
47.
D. Elad and D. Ginsburg, /. Am. Chem. Soc, 76, (1954), 312.
48.
For a fuller account of these studies, see also: a. R.O. Duthaler and D. Ginsburg, Helv. Chim. Acta, 69, (1986), 1559. b. D. Elad and D. Ginsburg, /. Chem. Soc, (1954), 3052. c. D. Elad and D. Ginsburg, J. Chem. Soc, (1953), 2664. d. D. Ginsburg and R. Pappo, J. Chem. Soc, (1953), 1524. e. D. Ginsburg and R. Pappo, /. Chem. Soc, (1951), 516. f. D. Ginsburg and R. Pappo, J. Chem. Soc, (1951), 939. g. E.D. Bergmann, R. Pappo, and D. Ginsburg, J. Chem. Soc, (1950), 1369.
49.
D.H.R. Barton, G.W. Kirby, W. Steglich, and G.M. Thomas, Proc Chem. Soc, (1963), 203.
50.
For related work, see: D.H.R. Barton, D.S. Bhakuni, R. James, and G.W. Kirby, J. Chem. Soc. (C), (1967), 128; see also reference 21b.
51.
T. Kametani, M. Ihara, K. Fukumoto, and H. Yagi, J. Chem. Soc. (C), (1969), 2030.
52.
I. Kikkawa, J. Pharm. Soc. Japan, 78, (1958), 1006.
53.
For applications of this methodology to the synthesis of numerous alkaloids, see: a. T. Kametani, A. Kozuka, and K. Fukumoto, J. Chem. Soc. fCj, (1971), 1021. b. T. Kametani, C. Seino, K. Yamaki, S. Shibuya, K. Fukumoto, K. Kigasawa, F. Satoh, M. Hiiragi, and T. Hayasaka, J. Chem. Soc. (C), (1971), 1043. c. T. Kametani, K. Shishido, E. Hayashi, C. Seino, T. Kohno, S. Shibuya, and K. Fukumoto, J. Org. Chem., 36, (1971), 1295. d. T. Kametani, T. Sugahara, H. Yagi, and K. Fukumoto, J. Chem. Soc. (C), (1969), 1063. e. T. Kametani, K. Fukumoto, A. Kozuka, H. Yagi, and M. Koizumi, J. Chem. Soc. (C), (1969), 2034. f. T. Kametani, K. Yamaki, H. Yagi, and K. Fukumoto, J. Chem. Soc. (C), (1969), 2602. g. T. Kametani, K. Fukumoto, and T. Sugahara, Tetrahedron Lett., 52, (1968), 5459. h. T. Kametani, K. Fukumoto, F. Satoh, and H. Yagi., J. Chem. Soc. Chem. Commun., (1968), 1398.
54.
a. M.A. Schwartz, and I.S. Mami, / Am. Chem. Soc, 97, (1975), 1239. b. M.A. Schwartz, and P.T.K. Pham, /. Org. Chem., 53, (1988), 2318.
55.
H. Conroy, /. Am. Chem. Soc, 77, (1955), 5960.
56.
For a number of ingenious related studies, see: a. D.G. Vanderlaan and M.A. Schwartz, J. Org. Chem., 50, (1985), 743. b. M.A Schwartz and M.F. Zoda, J. Org. Chem., 46, (1981), 4623. c. M.A. Schwartz and R.A. Wallace, Tetrahedron Lett., 35, (1979), 3257. d. D. H. R. Barton, Y. Herve, P. Potier, and J. Theirry, J. Chem. Soc Chem. Commun., (1984), 1298.
57.
T.S. Lie, L. Maat, and H.C. Beyerman, Reel. Trav. Chim. Pays-Bas, 98, (1979), 419.
112 58.
For a series of related studies, see also: a. H.C. Beyerman, J. van Berkel, T.S. Lie, L. Maat, J.C.M. Wessels, H.H. Bosman, E. Buurman, E.J.M. Bijsterveld, and HJ.M. Sinnige, Reel Trav. Chim. Pays-Bas, 97, (1978), 127. b. H.C. Beyerman, L. van Bommel, L. Maat, C. Olieman, Red Trav. Chim. Pays-Bas, 95, (1976), 312. c. H.C. Beyerman, T.S. Lie, L. Maat, H.H. Bosman, E. Buurman, E.J.M. Bijsterveld, and HJ.M. Sinnige, Reel Trav. Chim. Pays-Bas, 95, (1976), 24. d. H.C. Beyerman, E. Buurman, L. Maat, and C. Olieman, Reel Trav. Chim. Pays-Bas, 95, (1976), 184. e. C. Olieman, L. Maat, and H.C. Beyerman, Reel Trav. Chim. Pays-Bas, 95, (1976), 189. f. H.C. Beyerman, E. Buurman, T.S. Lie, and L. Maat, Reel Trav. Chim. Pays-Bas, 95, (1976), 43. g. H.C. Beyerman, E. Buurman, and L. Maat, J. Chem. Soe. Chem. Commun., (1972), 918. h. For a continuation of these studies, see also: C. Olieman, L. Maat, and H.C. Beyerman, Reel Trav. Chim.Pays-Bas., 99, (1980), 169.
59.
W.J. Musliner and J.W. Gates, Jr., /. Am. Chem. Soe., 88, (1966), 4271.
60.
J.I. DeGraw, J.C. Christensen, V.H. Brown, and M.J. Cory, J. Heteroeycl C/iem., 11, (1974), 363.
61.
K.C. Rice, and A. Brossi, J. Org. Chem., 45, (1980), 592.
62.
D.A. Evans, and C.H. Mitch, Tetrahedron Lett., 23, (1982), 285.
63.
An analogous opening with chloride followed by intramolecular Friedel-Crafts had served as a model study: D.A. Evans, C.H. Mitch, R.C. Thomas, D.M. Zimmerman, and R.L. Robey, J. Am. Chem. Soe., 102, (1980), 5955.
64.
W.H. Moos, R.D. Gless, and H. Rapoport, J. Org. Chem., 48, (1983), 227.
65.
D.D. Weller, R.D. Gless, and H. Rapoport, /. Org. Chem., 42, (1977), 1485.
66.
W.H. Moos, R.D. Gless, and H. Rapoport, J. Org. Chem., 46, (1981), 5064.
67.
W.H. Moos, R.D. Gless, and H. Rapoport, J. Org. Chem., 47, (1982), 1831.
113 68.
For some related studies see also: a. R.D. Gless, and H. Rapoport, J. Org. Chem., 44, (1979), 1324. b. D.D. Weller and H. Rapoport, /. Am. Chem. Soc, 98, (1976), 6650.
69.
J.D. White, G. Caravatti, T.B. Kline, E. Edstrom, K.C. Rice, and A. Brossi, Tetrahedron, 39, (1983), 2393.
70.
W. Ludwig, and H. J. Schafer, Angew., Chem., Int. Ed. Engl, IS, (1986), 1025.
71.
For an earlier, unsuccesful approach via anodic coupUng, see: H. Kliinenberg, C. Schaffer, and H. J. Schafer, Tetrahedron Lett., 23, (1982), 4581.
72.
J.E. Toth, and P.L. Fuchs, /. Org. Chem., 52, (1987) 473.
73.
Several model studies strengthened this approach: a. J.E. Toth, and P.L. Fuchs, /. Org. Chem., 51, (1986), 2594. b. P.R. Hamann, J.E. Toth, and P.L. Fuchs, J. Org. Chem., 49, (1984), 3865. For a full account, see: c. J.E. Toth, P.R. Hamann, and P.L. Fuchs, J. Org. Chem., 53, (1988), 4694.
74.
For an untethered variant of this strategy, see: J. Ponton, P. Helquist, P.C. Conrad, and P.L. Fuchs, J. Org. Chem., 46, (1981), 118.
75.
R.B. Barber and H. Rapoport, J. Med Chem., 19, (1976), 1175.
76.
K.C. Rice, J. Med Chem., 20, (1977), 164.
77.
M.A. Tins and M.A. Kerr, /. Am. Chem. Soc, 114, (1992), 5959.
78.
K.A. Parker and D. Fokas, J. Am. Chem. Soc, 114, (1992), 9688.
79.
Prehminary investigations of this radical approach were presented in: a. K.A. Parker, D.M. Spero, and J. Van Epp, / Org. Chem., 53, (1988), 4628. b. K.A. Parker, D.M. Spero, and in part K.C. Inman, Tetrahedron Lett., 27, (1986), 2833.
80.
A full account of this work has recently appeared: a. K.A. Parker, and D. Fokas, J. Org. Chem., 59, (1994), 3927. b. K.A. Parker and D. Fokas, J. Org. Chem., 59, (1994), 3933.
114 81.
C.Y. Hong, N. Kado, and L.E. Overman, J. Am. Chem. Soc, 115, (1993), 11028.
82.
For further details regarding this Mannich type cyclization, see: D.A. Heerding, C.Y. Hong, N. Kado, G.C. Look, and L.E. Overman, J. Org. Chem., 58, (1993), 6947.
83.
This methodology has recentiy been extended to a palladium catalyzed biscyclization in which the furan ring is formed directly: C.Y. Hong and L.E. Overman, Tetrahedron Lett., 35, (1994), 3453.
84.
H. Sdassi, G. Revial, M. Pfau, and J. d'Angelo, Tetrahedron Lett., 31, (1990), 875.
85.
D.L. Boger, M. Patel, and M.D. MulUcan, Tetrahedron Lett., 23, (1982), 4559.
86.
C.A. Broka and J.F. GerUts, J. Org. Chem., 53, (1988), 2144.
87.
E. Ciganek, J. Am. Chem. Soc, 103, (1981), 6261.
88.
E. Ciganek, U.S. Patent 4,243,668, (1981); Eur. Pat. Appl. 9780, (1980); Chem. Abstr. (1980), 93, 220720.
89.
M.A. CiufoHni, M.A. Rivera-Fortin, and N.E. Byrne, Tetrahedron Lett., 34, (1993), 3505.
90.
T. HudHcky, C.H. Boros, and E.E. Boros, Synthesis, (1992), 174.
91.
S. Handa, K. Jones, C.G. Newton, and D.J. Williams, J. Chem. Soc. Chem. Commun., (1985), 1362.
92.
T. Kametani, Y. Suzuki, and T. Honda, J. Chem. Soc. Perkin Trans. 1, (1986), 1373.
93.
P. Magnus and I. Coldham, / Am. Chem. Soc, 113, (1991), 672.
94.
Y. Genisson, C. Marazano, and B.C. Das, J. Org. Chem., 58, (1993), 2052.
95.
J.E. McMurry, V. Farina, W.J. Scott, A.H. Davidson, D.R. Summers, and A. Shenvi, J. Org. Chem., 49, (1984), 3803.
115
96.
For a successful application of this methodology to a total synthesis of 0-methylpallidinine, see: J.E. McMurry and V. Farina, Tetrahedron Lett., 24, (1983), 4653.
97.
A.I. Meyers and T.R. Bailey, J. Org. Chem., 51, (1986), 872.
98.
I. Monkovic, T.T. Conway, H. Wong, Y.G. Perron, I.J. Pachter, and B. Belleau, J. Am. Chem. Soc, 95, (1973), 7910.
99.
For the utilization of this methodology in the synthesis of several morphinan, isomorphinan, normorphinan and hasubanan derivatives of potential pharmacological interest, see also: a. I. Monkovic and H. Wong, Can. J. Chem., 54, (1976), 883. b. I. Monkovic, H. Wong, A.W. Pircio, Y.G. Perron, I.J. Pachter, and B. Belleau, Can. J. Chem., 53, (1975), 3094. c. I. Monkovic, H. Wong, B. Belleau, I. J. Pachter, and Y.G. Perron, Can. J. Chem., 53, (1975), 2515. d. T.T. Conway, T.W. Doyle, Y.G. Perron, J. Chapuis, and B. Belleau, Can. J. Chem., 53, (1975), 245. e. B. Belleau, H.Wong, I. Monkovic, and Y.G. Perron, J. Chem. Soc. Chem. Commun., (1974), 603. f. M. Saucier and I. Monkovic, Can. J. Chem., 52, (1974), 2736.
•
100.
I. Monkovic, Can. J. Chem., 53, (1975), 1189.
101.
I. Monkovic, C. Bachand, and H. Wong, J. Am. Chem. Soc, 100, (1978), 4609.
102.
M. Kitamura, Y. Hsiao, R. Noyori, and H. Takaya, Tetrahedron Lett., 28, (1987), 4829.
103.
M. Chandler and P.J. Parsons, J. Chem. Soc. Chem. Commun., (1984), 322.
104.
A.G. Schultz and R.D. Lucci, J. Chem. Soc. Chem. Commun, (1976), 925.
105.
For a full account see: A.G. Schultz, R.D. Lucci, W.Y. Fu, M.H. Berger, J. Erhardt, and W.K. Hagmann, J. Am. Chem. Soc, 100, (1978), 2150.
106.
A.G. Schultz, R.D. Lucci, J.J. Napier, H. Kinoshita, R. Ravichandran, P. Shannon, and Y.K. Yee, J. Org. Chem., 50, (1985), 217.
116 107.
For further developments see: A.G. Schultz, and PJ. Shannon, /. Org. Chem., 50, (1985), 4421.
108.
A.B. Shenvi, and E. Ciganek, J. Org. Chem., 49, (1984), 2942.
109.
L. Stella, B. Raynier, and J.-M. Surzur, Tetrahedron Lett., 31, (1977), 2721.
110.
For a full account, see: L. Stella, B. Raynier, and J.M. Surzur, Tetrahedron, 37, (1981), 2843.
111.
L. Stella, Angew. Chem. Intl. Ed. Engl., 22, (1983), 337, citing the thesis of J.L. Bourgeois, Marseille, 1980.
112.
M.A. Tius, and A. Thurkauf, Tetrahedron Lett., 11, (1986), 4541.
113.
D.D. Weller, E.P. Stirchak, and D.L. Weller, /. Org. Chem., 48, (1983), 4597.
114.
For preliminary results in this series, see: a. D.D. Weller, D.L. Weller and G.R. Luellen, J. Org. Chem., 48, (1983), 3061. b. D.D. WeUer and D.L. WeUer, Tetrahedron Lett., 23, (1982), 5239. c. D.D. Weller and G.R. Luellen, Tetrahedron Lett., 22, (1981), 4381.
115.
T. Hudlicky, G. Butora, S.P. Feamley, A.G. Gum, and M.R. Stabile, (pending release of publication from sponsors of research)
117 8. Appendix I. List of Dissertation Titles Concerning Morphine Synthesis 1.
The tandem radical cychzation synthesis of morphine alkaloids Fokas, Demosthenes (1993) 234 pp. Avail.: Univ. Microfihns Int., Order No. DA9406935 From: Diss. Abstr. Int. B 1994, 54(10), 5150
2.
Synthetic approaches to novel morphine analogs Turner, Stephen Michael (1991) 202 pp. Avail.: Univ. Microfihns Int., Order No. BRDX94736 From: Diss. Abstr. Int. B 1992, 52(10), 5283-4
3.
Part I. The total synthesis of two human urinary metaboHtes of delta-9-THC. Part n. The total synthesis of (d,l)- morphine Kerr, Michael Andre (1991) 387 pp. Avail.: Univ. Microfihns Int., Order No. DA9205864 From: Diss. Abstr. Int. B 1992, 52(9), 4733
4.
Intramolecular Diels-Alder cychzations in an approach to the morphine skeleton Wu, Chengde (1990) 254 pp. Avail: Univ. Microfihns Int., Order No. DA9128007 From: Diss. Abstr. Int. B 1991, 52(4), 2044
5.
Organic synthesis via palladium coupling reactions Pyatt, D. (1990) 167 pp. Avail.: Univ. Microfihns Int., Order No. BRD-92664 From: Diss. Abstt. Int. B 1991, 52(3), 1445
6.
A synthetic approach to morphine Ellwood, Charles Walter (1989) 141 pp. Avail.: Univ. Microfihns Int., Order No. BRDX91246 From: Diss. Abstr. Int. B 1991, 51(9), 4343
7.
New synthetic approaches towards the synthesis of morphine Spoors, Paul Grant (1989) 195 pp. Avail.: Univ. Microfihns Int., Order No. BRDX89587 From: Diss. Abstr. Int. B 1990, 51(4), 1836-7
8.
A radical cyclization approach to the synthesis of morphine and synthetic approaches to trialkoxyphthalic acid derivatives Spero, Denice Mary (1988) 143 pp. Avail.: Univ. Microfihns Int., Order No. DA8825202 From: Diss. Abstr. Int. B 1989, 50(3), 970
9.
A study directed at the total synthesis of (-)-codeine and (-)- morphine: synthesis via a novel asymmetric intramolecular Diels-Alder reaction Costanzo, Michael John (1988) 276 pp. Avail.: Univ. Microfilms Int., Order No. DA8818767 From: Diss. Abstr. Int. B 1989, 49(7), 2647
10.
Approaches to the synthesis of morphine. Wan, Barbara Yu Fong (1987) 112 pp. Avail.: Univ. Microfilms Int., Order No. DA8715769 From: Diss. Abstr. Int. B 1987, 48(6), 1692
11.
The total synthesis of (.+-.)- morphine Toth, John Eldon (1986) 682 pp. Avail.: Univ. Microfihns Int., Order No. DA8709865 From: Diss. Abstr. Int. B 1987, 48(1), 143
118 12.
A novel approach to the synthesis of morphine. Hinton, Michael (1987) 153 pp. Avail.: Univ. Microfihns Int., Order No. DA8711350 From: Diss. Abstr. Int. B 1987, 48(2), 441
13.
A contribution toward the synthesis of morphine. Rodriguez, Cesar (1986) 155 pp. Avail.: Univ. Microfihns Int., Order No. DA8617023 From: Diss. Abstr. Int. B 1987, 47(7), 2922
14.
A study of the phenoHc oxidative coupling reaction in the synthesis of morphine alkaloids. An approach to the asymmetric synthesis of codeine Pham Phuong Thi Kim (1985) 157 pp. Avail: Univ. Microfihns Int., Order No. DA8529558 From: Diss. Abstr. Int. B 1986, 46(11), 3851
15.
New aromatic annulation methods: total syntheses of juncusol, sendaverine, and morphine -related analgesics Mullican, Michael David (1984) 163 pp. Avail: Univ. Microfihns Int., Order No. DA8513829 From: Diss. Abstr. Int. B 1985, 46(4), 1175-6
16.
A study of the phenolic oxidative couphng reaction in the synthesis of morphine alkaloids Vanderlaan, Douglas George (1984) 105 pp. Avail: Univ. Microfihns Int., Order No. DA8428711 From: Diss. Abstr. Int. B 1985, 45(11), 3512
17.
An approach to the morphine alkaloids: synthesis of 9-methoxy-3-methyl2,3,4,4a,5,6-hexahydro-lH-benzofuro[3,2-e]i>oquinohne-7(7aH)-ones Weller, Doreen L. (1984) 109 pp. Avail: Univ. Microfihns Int., Order No. DA8402152 From: Diss. Abstr. Int. B 1984, 44(11), 3412
18.
Studies directed toward the total synthesis of morphine. Hamann, Phihp Ross (1983) 684 pp. Avail: Univ. Microfihns Int., Order No. DA8407547 From: Diss. Abstr. Int. B 1984, 45(3), 875
19.
New methods in organic synthesis. Part I. Regioselective conversion of ketones into olefins via vinyl triflates. Part U. An approach to the total synthesis of morphine Scott, Wilham Johnston (1983) 188 pp. Avail: Univ. Microfilms Int., Order No. DA8321902 From: Diss. Abstr. Int. B 1983, 44(6), 1832
20.
Approaches to the synthesis of morphine. Harris, David Jude (1982) 159 pp. Avail: Univ. Microfihns Int., Order No. DA8220104 From: Diss. Abstr. Int. B 1982, 43(4), 1102
21.
The apphcation of metalated enamines to the synthesis of morphine alkaloids Mitch, Charles Howard (1982) 159 pp. Avail: Univ. Microfilms Int., Order No. DA8218846 From: Diss. Abstr. Int. B 1982, 43(3), 731
22.
Heteroatom directed photoarylation. Approaches to the synthesis of morphine and the study of a stereospecific benzodihydrofuran photorearrangement Napier, James Joseph (1981) 309 pp. Avail: Univ. Microfihns Int., Order No. 8119452 From: Diss. Abstr. Int. B 1981, 42(4), 1458-9
119 23.
Approaches to the synthesis of morphine McGowan, Cynthia Baker (1981) 96 pp. Avail.: Univ. Microfilms Int., Order No. 8116395 From: Diss. Abstr. Int. B 1981, 42(2), 636
24.
Biomimetic syntheses of several morphine alkaloid analogs Zoda, Michael Francis (1981) 96 pp. Avail.: Univ. Microfihns Int., Order No. 8113273 From: Diss. Abstr. Int. B 1981, 42(1), 225
25.
Synthetic approaches to morphine and colchicine alkaloid analogs Wallace, Rebecca Abemathy (1979) 147 pp. Avail.: Univ. Microfihns Int., Order No. 7926834 From: Diss. Abstr. Int. B 1980, 40(7), 3179
26.
A biogenetically patterned synthesis of the morphine alkaloids Mami, Ismail Sadeg (1978) 78 pp. Avail.: Univ. Microfilms Int., Order No. 7917053 From: Diss. Abstr. Int. B 1979, 40(2), 755-6
27.
Heteroatom directed photoarylation. AppUcation toward the synthesis of morphine Lucci, Robert Dominick (1977) 189 pp. Avail: Univ. Microfihns Int., Order No. 7807790 From: Diss. Abstr. Int. B 1978, 38(12, Pt. 1), 5942
28.
Cell division and macromolecular synthesis in Tetrahymena pyriformis. Action of tetrahydrocannabinol, morphine, levorphanol and levalloiphan McClean, Daniel K. (1972) No pp. given Avail.: Natl. Libr. Canada, Ottawa Ont From: Diss. Abstr. Int. B 1974, 34(9), 4258-9
29.
Effect of cycloheximide, an inhibitor of protein synthesis on the development of tolerance to morphine Feinberg, Michael P. (1973) 140 pp. Avail.: Univ. Microfilms, Ann Arbor, Mich., Order No. 7323,480 From: Diss. Abstr. Int. B 1973, 34(4), 16
30.
Excitatory actions of morphine and synthetic surrogates Brister, Calvin Cotten (1972) 146 pp. Avail.: Univ. Microfihns, Ann Arbor, Mich., Order No. 72-20,226 From: Diss. Abstr. Int. B 1972, 33(1), 351-2
31.
Synthesis of morphine isomers Chang, Jaw-Kang (1969) No pp. given Avail.: Natl. Libr. Canada, Ottawa, Ont From: Diss. Abstr. Int. B 1970, 31(4), 2157-8
32.
Synthesis of some derivatives of (-)-3-hydroxy-6- oxomorphinan structurally related to known analgesics and analgesic antagonists of the morphine type. Neubert, Mary E. (1968) 211 pp. Avail.: 68-15,850 From: Diss. Abstr. B 1968, 29(5), 1612-13
120 8. Appendix n. List of references connected to synthetic transformations of morphine and derivatives and biological testing 1.
Synthesis and analytical characterization of dansyl derivatives of morphine-like substances Hosztafi, Sandor; Repasi, Janos Acta Pharm. Hung. (1994), 64(1), 22-5
2.
The NMDA receptor antagonists, LY274614 and MK-801, and the nitric oxide synthase inhibitor, NG-nitro-L-arginine, attenuate analgesic tolerance to the muopioid morphinebut not to kappa opioids Elliott, Kathryn; Minami, Nobuko; Kolesnikov, Yuri A.; Pasternak, Gavril W.; Inturrisi, Charles E. Pain (1994), 56(1), 69-75
3.
Nitric oxide (NO) synthase inhibitors attenuated naloxone-precipitated withdrawal Dzoljic, M. R.; Cappendijk, S. L. T.; de Vries, R. Regul. Pept. (1994), (Suppl. 1), S285-S286
4.
Inhibition of nitric oxide synthase attenuates the development of morphine tolerance and dependence in mice Majeed, N. H.; Przewlocka, B.; Machelska, H.; Przewlocki, R. Neuropharmacology (1994), 33(2), 189-92
5.
Involvement of the nitric oxide pathway in nociceptive processes in the central nervous system in rats Przewlocka, B.; Machelska, H.; Przewlocki, R. Regul. Pept. (1994), (Suppl. 1), S75-S76
6.
Synthesis of N,C 10-bridged morphine derivatives: 5H-10,13 iminoethanophenanthro[4,5-bcd]furan. I Fleischhacker, W.; Richter, B.; Voellenkle, H. Monatsh. Chem. (1993), 124(8-9), 909-22
7.
Synthesis and analgetic activity of nicotinic esters of morphine derivatives Hosztafi, S.; Kohegyi, I.; Simon, C ; Furst, Z. Arzneim.-Forsch. (1993), 43(11), 1200-3
8.
Biochemical characterization of a synthetic NPFF agonist with high affinity and resistance to brain peptidase inactivation Devillers, J. P.; Reeve, A.; Mazarguil, H.; AUard, M.; Zajac, J M.; Dickenson, A. H.; Simonnet, G. Regul. Pept. (1994), (Suppl. 1), S123-S124
9.
Structure activity relationships of synthetic and semisynthetic opioid agonists and antagonists Hosztafi, Sandor; Friedmann, Tamas; Furst, Zsuzsanna Acta Pharm. Hung. (1993), 63(6), 335-49
10.
Inhibitory effect of nitric oxide (NO) synthase inhibitors on naloxone-precipitated withdrawal syndrome in morphine -dependent mice Cappendijk, Susan L. T.; de Vries, Rene; Dzoljic, Michailo R. Neurosci. Lett. (1993), 162(1-2), 97-100
121 11.
Attenuation of some signs of opioid withdrawal by inhibitors of nitric oxide synthase Kimes, Alane S.; Vaupel, D. Bruce; London, Edythe D. Psychopharmacology (BerUn) (1993), 112(4), 521-4
12.
Manufacture of multilayered controlled-release transdermal patches Wick, John; Weimann, Ludwig J.; Pollock, Wayne C. Eur. Pat. AppL, 35 pp. HP 92-850190 920813 PRAI US 92-861534 920401
13.
Method for synthesizing glucuronides of 4,5-epoxymorphinans Mertz, Alfred Adophe Henri PCX Int. AppL, 25 pp. WO 9305057 Al 930318 WO 92-FR846 920904 PRAI FR 91-10927 910904
14.
Morphine alkaloids. 120. Synthesis of N-demethyl-N-substituted 14.beta.hydroxy-isomorphine and dihydroisomorphine derivatives Hosztafi, Sandor; Simon, Csaba; Makleit, Sandor Heterocycles (1993), 36(7), 1509-19
15.
Morphine suppresses DNA synthesis in cultured murine astrocytes from cortex, hippocampus and striatum Stiene-Martin, Anne; Hauser, Kurt F. Neurosci. Lett. (1993), 157(1), 1-3
16.
Biological synthesis of the analgesic hydromorphone, an intermediate in the metaboUsm of morphine, by Pseudomonas putida MIO Hailes, Anne M.; Bruce, Neil C. AppL Environ. Microbiol. (1993), 59(7), 2166-70
17.
Effect of genetic obesity and phenobarbital treatment on the hepatic conjugation pathways Chaudhary, Inder P.; Tuntaterdtum, Somsong; McNamara, Patrick J.; Robertson, Larry W.; Blouin, Robert A. J. Pharmacol. Exp. Ther. (1993), 265(3), 1333-8
18.
Synthetic and natural opiates interact with P-glycoprotein in multidrug-resistant cells Callaghan, Richard; Riordan, John R. J. Biol. Chem. (1993), 268(21), 16059-64
19.
Inhibition of nitric oxide synthase enhances morphine antinociception in the rat spinal cord Przewlocki, Ryszard; Machelska, HaUna; Przewlocka, Barbara Life Sci. (1993), 53(1), PL1-PL5
20.
Inhibition of the morphine withdrawal syndrome by a nitric oxide synthase inhibitor, NG-nitro-L-arginine methyl ester Adams, Michael L.; Kahcki, Joelle M.; Meyer, Edward R.; Cicero, Theodore J. Life Sci. (1993), 52(22), PL245-PL249
122 21.
Blockade of tolerance to morphine but not to .kappa, opioids by a nitric oxide synthase inhibitor Kolesnikov, Yuri A.; Pick, Chaim G.; Ciszewska, Grazyna; Pasternak, Gavril W. Proc. Natl. Acad. Sci. U. S. A. (1993), 90(11), 5162-6
22.
Enzymatic hydroxylation of arene and synunetry considerations in efficient synthetic design of oxygenated natur^d products HudUcky, Tomas; Fan, Rulin; Luna, Hector; Olivo, Horacio; Price, John Indian J. Chem., Sect. B (1993), 32B(1), 154-8
23.
Morphine regulates DNA synthesis in rat cerebellar neuroblasts in vitro Hauser, Kurt F. Dev. Brain Res. (1992), 70(2), 291-7
24.
Morphine alkaloids. 119. A new efficient method for the preparation of 2-fluoroN-propylnorapomorphine Berenyi, Sandor; Hosztafi, Sandor; Makleit, Sandor J. Chem. Soc, Perkin Trans. 1 (1992), (20), 2693-4
25.
Synthesis of a new morphine derivative with anorexogenic activity Berenyi, Sandor; Makleit, Sandor; Hosztafi, Sandor; Furst, Susanna; Friedmann, Tamas; Knoll, Jozsef Med. Chem. Res. (1991), 1(3), 185-90
26.
Synthesis of N-demethyl-N-substituted-14-hydroxycodeine and morphine derivatives Hosztafi, Sandor; Simon, Csaba; Makleit, Sandor Synth. Commun. (1992), 22(17), 2527-41
27.
NG-Nitro-L-arginine prevents morphine tolerance Kolesnikov, Yuri A.; Pick, Chaim G.; Pasternak, Gavril W Eur. J. Pharmacol. (1992), 221(2-3), 399-400
28.
An improved synthesis of noroxymorphone Ninan, Aleyamma; Sainsbury, Malcolm Tetrahedron (1992), 48(32), 6709-16
29.
Inhibition of estradiol-induced DNA synthesis by opioid peptides in the rat uterus Ordog, Tamas; Vertes, Zsuzsanna; Vertes, Marietta Life Sci. (1992), 51(15), 1187-96
30.
Morphine-induced downregulation of .mu.-opioid receptors and peptide synthesis in neonatal rat brain Tempel, Ann; Espinoza, Kathryn Ann. N. Y. Acad. Sci. (1992), 654(Neurobiol. Drug Alcohol Addict.), 529-30
31.
Design and synthesis of an opioid receptor probe: mode of binding of S-activated(-)-6.beta.-sulfhydryldihydromorphine with the sulfhydryl group in the .mu.-opioid receptors Kanematsu, Ken; Kaya, Tetsudo; Shimohigashi, Yasuyuki; Yagi, Kunio; Ogasawara, Tomio Med. Chem. Res. (1991), 1(3), 191-4
123 32.
Synthesis of N-substituted C-normorphinans and their pharmacological properties Takeda, Mikio; Inoue, Hirozumi; Noguchi, Katsuyuki; Honma, Yasushi; Okamura, Kimio; Date, Tadamasa; Nurimoto, Seiichi; Yamamura, Michio; Saito, Seiichi Chem. Pharm. Bull. (1992), 40(5), 1186-90
33.
Lx>ng term effects of morphine on mesangial cell proliferation and matrix synthesis Singhal, Pravin C ; Gibbons, Nora; Abramovici, Mirel Kidney Int. (1992), 41(6), 1560-70
34.
Facile syntheses of aporphine derivatives Hedberg, Martin H.; Johansson, Anette M.; Hacksell, Uh J. Chem. Soc, Chem. Conmiun. (1992), (11), 845-6
35.
Morphine alkaloids. Part 116. Synthesis of N-demethyl-N-substituted dihydroisomorphine and dihydroisocodeine derivatives Hosztafi, Sandor; Simon, Csaba; Makleit, Sandor Synth. Commun. (1992), 22(12), 1673-82
36.
Morphine alkaloids. Part 114. A stereohomogeneous synthesis of N-demethyl-N-substituted-14hydroxydihydromorphines Hosztafi, Sandor; Berenyi, Sandor; Toth, Geza; Makleit, Sandor Monatsh. Chem. (1992), 123(5), 435-41
37.
Structure-activity smdies of morphine fragments. El. Synthesis, opiate receptor binding, analgetic activity and conformational studies of spiro-[tetralin-l,4'-piperidines] Lawson, J. A.; Toll, L.; Polgar, W.; Uyeno, E. T.; Loew, G. H. Eur. J. Med. Chem. (1991), 26(8), 775-85
38.
Morphine alkaloids. 113. Synthesis of C-3 halogen-substituted apocodeines and apomorphines Simon, Csaba; Hosztafi, Sandor; Makleit, Sandor; Berenyi, Sandor Synth. Commun. (1991), 21(22), 2309-16
39.
Structure-activity studies of morphine fragments, n. Synthesis, opiate receptor binding, analgetic activity and conformational studies of 2-R-2(hydroxybenzyl)piperidines Loew, G. H.; Lawson, J. A.; Toll, L.; Polgar, W.; Uyeno, E. T. Eur. J. Med. Chem. (1991), 26(8), 763-73
40.
Controlled release pharmaceutical preparation and process for preparing same Zsuga, Miklos; Kelen, Tibor; Nagy, Jozsef; Barkanyi, Judit; Bene, Magdolna; Ondi, Sandor; Gulyas, Imre; Gyoeker, Istvan; Repasi, Janos; Repasi, Agota Eur. Pat. AppL, 6 pp. EP 463833A2 920102 AI EP 91-305669 910624 PRAI HU 90-4007 900627
41.
Coordination compounds as precursors for materials synthesis Langfelderova, H.; Papankova, B.; Makanova, D.; Gersi, P.; Kozisek,
124 J. Proc. Conf. Coord. Chem. (1991), 13th, 149-54 42.
Sustained-release pharmaceutical mucosal patches Scholz, Matthew T.; Scherrer, Robert A.; Marecki, Nelda M.; Barkhaus, Joan K.; Chen, Yen Lane PCT Int. AppL, 48 pp. PI WO 9106290 Al 910516 DS W: AU, BR, CA, JP, KR RW: AT, BE, CH, DE, DK, ES, FR, GB, GR, IT, LU, NL, SE AI WO90-US6505 901102 PRAI US 89-431664 891103
43.
Effect of opioids on the activity of some key enzymes involved in milk synthesis in manmiary gland of lactating rabbit Hossain, M. A.; GanguLL, N. C. Indian Vet. J. (1991), 68(7), 630-5
44.
Design and synthesis of HTV protease inhibitors. Variations of the carboxyterminus of the HTV protease inhibitor L-682,679 DeSolms, S. Jane; Giuhani, Elizabeth A.; Guare, James P.; Vacca, Joseph P.; Sanders, William M.; Graham, Samuel L.; Wiggins, J. Mark; Darke, Paul L.; Sigal, Irving S.; et al. J. Med. Chem. (1991), 34(9), 2852-7
45.
Inhibition of cell growth and DNA, RNA, and protein synthesis in vitro by fentanyl, sufentanil, and opiate analgesics Nassiri, M. Reza; Flynn, Gordon L.; Shipman, Charles, Jr. Pharmacol. Toxicol. (Copenhagen) (1991), 69(1), 17-21
46.
Opioid involvement in the control of melatonin synthesis and release Stankov, B.; Esposti, D.; Esposti, G.; Lucini, V.; Mariani, M.; Cozzi, B.; Scaglione, F.; Fraschini, F. Adv. Pineal Res. (1990), 4, 45-8
47.
Preparation of racemic and optically-active fatty amino acids, their homo- and hetero-oligomers and conjugates, as pharmaceuticals Gibbons, WilHam A. Brit. UK Pat. AppL, 55 pp. PI GB 2217319 Al 891025 AI GB 88-9162 880419
48.
New method for synthesis of tricyclic morphine analog Zhang, Yongmin; Zhang, Lihe; Liu, Weiqin; Thai, C ; Labidalle, S. Huaxue Xuebao (1990), 48(10), 1030-5
49.
Assay of semisynthetic codeine base with simultaneous determination of.alpha.-codeimethine and 06-codeine methyl ether as by-product impurities by high-performance Uquid chromatography Ayyangar, N. R.; Bhide, S. R.; Kalkote, U. R. J. Chromatogr. (1990), 519(1), 250-5
50.
Effects of morphine in arachidonic acid metaboHsm, of calcium-uptake and on cAMP synthesis in uterine strips from spayed rats Faletti, A.; Bassi, D.; Franchi, A. M.; Gimeno, A. L.; Gimeno, M. A.F.
125 Prostaglandins, Leukotrienes Essent. Fatty Acids (1990), 41(3), 151-5 51.
Design and synthesis of an opioid receptor probe: mode of binding of S-activated (-)-6.beta.-sulfhydryldihydromorphine with the sulfhydryl group in the .mu.-opioid receptor Kanematsu, Ken; Naito, Ryo; Shimohigashi, Yasuyuki; Ohno, Motonori; Ogasawara, Tomio; Kurono, Masayasu; Yagi, Kunio Chem. Pharm. BuU. (1990), 38(5), 1438-40
52.
Synthesis and analgesic activity of sulfur-containing morphinans and related comf)ounds Hori, Mikio; Iwamura, Tatsunori; Imai, Eiji; Shimizu, Hiroshi; Kataoka, Tadashi; Nozaki, Masakatsu; Niwa, Masayuki; Fujimura, Hajime Chem. Pharm. Bull. (1989), 37(5), 1245-8
53.
A novel synthesis of pyrimidobenzodiazepines Dlugosz, Anna Arch. Pharm. (Weinheim, Ger.) (1990), 323(1), 59-60
54.
Oxidative coupling of cis-3,N-bis(methoxycarbonyl)-N-norreticuline: an approach to the asymmetric synthesis of morphine alkaloids Schwartz, Martin A.; Pham, Phuong T. K. Adv. Biosci. (Oxford) (1989), 75(Prog. Opioid Res.), 121-4
55.
Synthetic opioids compared with morphine and ketamine: catalepsy, cross-tolerance and interactions in the rat Benthuysen, J. L.; Hance, A. J.; Quam, D. D.; Winters, W. D. Neuropharmacology (1989), 28(10), 1011-15
5 6.
Preparation and use of monomeric phthalocyanine reagents Stanton, Thomas H.; Schindele, Deborah C ; Renzoni, George E.; Pepich, Barry V.; Anderson, Neils H.; Clagett, James A.; Opheim, Kent E. PCT Int. Appl., 54 pp. PI WO 8804777 Al 880630 AI W 0 87-US3226 871211 PRAI US 86-941619 861215 US 86-946475 861224 US 87-61937 870612
57.
Synthesis and antinociceptive activity of thiohydantoin derivatives Xu, Guoyou; Yu, Zhengwei; Peng, Sixun Zhongguo Yaoke Daxue Xuebao (1988), 19(4), 245-8
58.
Enantioselective inhibition: strategy for improving the enantioselectivity of biocatalytic systems Guo, Zhi Wei; Sih, Charles J. J. Am. Chem. Soc. (1989), 111(17), 6836-41
60.
Attenuation of morphine withdrawal syndrome by macromolecular synthesis inhibitors in rats Copeland, Robert L., Jr.; Pradhan, S. N. Drug Dev. Res. (1989), 17(2), 169-74
60
Morphine alkaloids. 104. Synthesis and conversions of new epoxy derivatives
126 Gulyas, Gyongyi; Berenyi, Sandor; Makleit, Sandor Acta Chim. Hung. (1988), 125(2), 255-65 61.
Enzymic synthesis and immunocheniical characteristics of antigens of morphine -protein conjugates Kovalev, I. E.; Tomilin, V, A. Farmakol. Toksikol. (Moscow) (1989), 52(3), 62-6
62.
Synthesis of l,8-diaryl-2,7-di(hexamethyleneiniinomethyl)1,8-octanedione dihydrochlorides and their analgesic properties Petrosyan, L. M.; Gevorgyan, G. A.; Durgaryan, L. K.; Azhvyan, A.S.; Vlasenko, E. V.; Mndzhoyan, O. L. Khim.-Farm. Zh. (1988), 22(9), 1073-6
63.
8-Phenylmorphinones and 8-phenylcodeinones: synthesis and analgesic activity KaUnin, V. N.; Kazantseva, S. A.; Petrovskii, P. V.; Kobel'kova, N.I.; Ignatov, Yu. D.; Zvartau, E. E.; Dorokhova, M. I. Khim.-Fann. Zh. (1989), 23(1), 48-50
64.
Design and synthesis of suLfur-containing morphine and an opioid receptor probe Fujii, Dcuo; Togame, Hiroko; Yamamoto, Mayumi; Kanematsu, Ken; Takayanagi, Issei; Konno, Fukio Chem. Pharm. Bull. (1988), 36(6), 2282-5
65.
Peptides related to leucine-Zmethionine-enkephalinamides: synthesis and biological activities Sivanandaiah, K. M.; Gurusiddappa, S.; Suresh Babu, V. V. Indian J. Chem., Sect. B (1988), 27B(7), 645-8
66.
Synthesis of 8-aryldihydrocodeinones and -morphinones via palladium-catalyzed reactions Kalinin, V. N.; Kazantseva, S. A.; Petrovskii, P. V.; Kobel'kova, N.I.; Polyakov, A. v.; Yanovskii, A. I.; Struchkov, Yu. T. Dokl. Akad. Nauk SSSR (1988), 298(1), 119-22 [Chem.]
67.
Electrosynthesis of morphine-derivatives, isoquinolines, and cyclopentanoids Schaefer, H. J.; Schlegel, C ; Eilenberg, W.; Mueller, U.; Huhtasaari, M.; Becking, L. F.E.C.S. Int. Conf. Chem. Biotechnol. Biol. Act. Nat. Prod., [Proc.], 3rd (1987), Meeting Date 1985, Volume 1, 96-116 Publisher: VCH, Weinheim, Fed. Rep. Ger.
68.
Morphine analogs derived from tetra- and hex^ydrobenzofurans. II. Synthesis and stereoselective functionahzation of dibenzofurans Labidalle, Serge; Min, Zhang Yong; Reynet, Annick; Moskowitz, Henri; Vierfond, Jean Michel; Miocque, Marcel; Bucourt, Robert; Thal,Claude Tetrahedron (1988), 44(4), 1171-86
69.
Morphine analogs derived from tetra-and hexahydrobenzofurans. I. Synthesis and stereospecific epoxidation of arylalkylcyclohexane intermediates Labidalle, Serge; Min, Zhang Yong; Reynet, Annick; Moskowitz, Henri; Vierfond, Jean Michel; Miocque, Marcel; Bucourt, Robert; Thal,Claude
127 Tetrahedron (1988), 44(4), 1159-69 70.
Some pharmacological properties of a newly synthesized morphine derivative, (-)-6.beta.-acetylthiomorphine Takayanagi, I.; Konno, F.; Goromaru, N.; Koike, K.; Kanematsu, K.; Fujii, L; Togame, H. Arch. Int. Pharmacodyn. Ther. (1988), 294, 71-84
71.
Narcotic alkylating agents: synthesis, structure and biological activities Frigola, J.; Colombo, A.; Mas, J.; Pares, J. Farmaco, Ed. Sci. (1988), 43(4), 347-62
7 2.
Preparation of new echibohne derivatives having opioid properties Robinson, Brian; Rees, John Michael Hugh; Cox, Brian PCX Int. Appl., 40 pp. PI WO 8800193 Al 880114 WO 87-GB457 870630 GB 86-16089 860702
73.
Synthesis and carbon-13 NMR study of some podocarpic acid derivatives Ortellado, Maria AmeUa A. C ; MarsaioH, Anita J. J. Chem. Res., Synop. (1987), (10), 324-5
74.
Oxidative coupling of cis-3,N-bis(methoxycarbonyl)-N-norreticuline. An approach to the asymmetric synthesis of morphine alkaloids Schwartz, Martin A.; Pham Phuong Thi Kim J. Org. Chem. (1988), 53(10), 2318-22
75.
General asymmetric synthesis of morphine and its analogs via enantioselective hydrogenation Kitamura, Masahito; Hsiao, Yi; Noyori, Ryoji; Takaya, Hidemasa Tennen Yuki Kagobutsu Toronkai Koen Yoshishu (1987), 29, 385-92
76.
Alcohol and morphine affect protein synthesis of astrogUa-enriched primary cultures from various brain regions Roennbaeck, L.; Hansson, E. Neurol. Neurobiol. (1988), 39(Biochem. Pathol. Astrocytes), 215-17
77.
Synthesis of carbon-11 labeled diprenorphine: a radioUgand for positron emission tomographic studies of opiate receptors Lever, John R.; Dannals, Robert F.; Wilson, Alan A.; Ravert, Hayden T.; Wagner, Henry N., Jr. Tetrahedron Lett. (1987), 28(35), 4015-18
78.
Synthesis and conversion of new epoxy derivatives of morphine alkaloids Gulyas, Gyongyi; Berenyi, Sandor; Makleit, Sandor Magy. Kem. Foly. (1986), 92(11-12), 508-13
7 9.
Synthesis of the skeleton of the morphine molecule by mammalian hver Weitz, Charles J.; FauU, Kym F.; Goldstein, Avram Nature (London) (1987), 330(6149), 674-7
80.
Improved synthesis of 3-substituted 7-methoxybenzofurans. Useful intermediates for the preparation of morphine analogs of organic chemistry
128 Jung, Michael E.; Abrecht, Stefan J. Org. Chem. (1988), 53(2), 423-5 81.
Pharmacological evidence of morphine-induced inhibition of gastric mucus synthesis in rats Del Tacca, M.; Bemardini, C ; Corsano, E.; Bertelli, A.; Roze, C. Int. J. Tissue React. (1987), 9(5), 413-18
82.
The relative stabilities of substituted cis- and trans-l,2,3,4,4a,9,l0,10aoctahydrophenanthrenes, including configurational corrections in the EladGinsburg morphine synthesis Duthaler, Rudolf O.; Ginsburg, David Helv. Chim. Acta (1986), 69(7), 1559-66
8 3.
Synthesis of N- [methyl-11 C]hydromorphone by using multivariate strategies for optimization of radiochemical yields Rimland, Annika; Bergson, Goeran; Obenius, Ulf; Sjoeberg, Stefan; Laangstroem, Bengt Appl. Radiat. Isot. (1987), 38(8), 651-4
84.
Synthesis of metabolites of 3-O-tert-butylmorphine Mohacsi, E. J. Heterocycl. Chem. (1987), 24(2), 471-2
85.
Synthetic studies on morphine. Racemization of biaryl intermediates Weller, Dwight D.; Runyan, Mark T. Tetrahedron Lett. (1986), 27(40), 4829-32
86.
Synthesis, purification and biological evaluation of porcine corticotropin-releasing factor Guoth, J.; Olsen, D. B.; Kovacs, M.; Schally, A. V. Life Sci. (1987), 41(8), 1003-10
87.
Synthesis and pharmacological properties of morphine congeners having pendant crown ether as an opioid receptor probe Fujii, Ecuo; Togame, Hiroko; Kanematsu, Ken Chem. Pharm. Bull. (1986), 34(10), 4439-42
88.
In vivo tracer studies of glucose metabolism, cerebral blood flow, and protein synthesis in naloxone precipitated morphine withdrawal Geary, W. A., II; Wooten, G. F. Neurochem. Res. (1987), 12(7), 573-80
89.
Syntheses of 3-0- and 6-O-propanoylmorphine. A reinvestigation and correction Sy, Wing Wah; By, Arnold W.; NeviUe, George A.; Wilson, Wilham L. J. Pharm. Sci. (1986), 75(8), 787-9
90.
Further syntheses with nitroxide .alpha.,.beta.-unsaturated aldehydes and allyhc bromides Hideg, Kalman; Cseko, Jozsef; Hankovszky, H. Olga; Sohar, Pal Can. J. Chem. (1986), 64(8), 1482-90
91.
Synthesis of morphine analogs with a tetrahydrodibenzofuran skeleton
129 Labidalle, S.; Min, Zhang Yong; Reynet, A.; Thai, C; Moskowitz, H. Tetrahedron Lett. (1986), 27(25), 2861-2 9 2.
Synthesis of labelled morphine compounds Toth, Geza; Sirokman, Ferenc; Hosztafi, Sandor Izotoptechnika (1985), 28(4), 226-36
93.
Aporphines 65: Chemical, microbial synthesis and characterization by gas chromatography/mass spectrometry of (R)-(-)-10-hydroxy 11-methoxy-N-npropylnoraporphine, a potential metabolite of N-n-propylnorapomorphine Neumeyer, John L.; Abdel-Maksoud, Hamdy M.; Trainor, Thomas M.; Vouros, Paul; Davis, Patrick J. Biomed. Environ. Mass Spectrom. (1986), 13(5), 223-9
94.
Effects of morphine on DNA synthesis in neonatal rat brain Komblum, Harley I.; Loughhn, Sandra E.; Leshe, Frances M. Dev. Brain Res. (1987), 31(1), 45-52
95.
Long-lasting inhibitory effects of a synthetic enkephalin (FK 33-824) on evoked neuronal thalamic firing in the rat Braga, Pier Carlo; BieUa, Gabriele; Rigoh, Milena; Tiengo, Mario; Fraschini, Franco; Netti, Carmela; Guidobono, Francesca IRCS Med. Sci. (1986), 14(12), 1181-2
96.
Total synthesis of rac-salutaridine and sinoacutine [(-)-salutaridine] - a new method to prepare morphine Ludwig, Wolfgang; Schaefer, Hans J. Angew. Chem. (1986), 98(11), 1032-3
97.
Stimulation of brain-stem protein synthesis.by morphine Roennbaeck, Lars; Hansson, EHsabeth Biochem. Pharmacol. (1986), 35(21), 3685-92
98.
Precursors of the mammahan synthesis.of morphine: (-i-)-salutaridine and (-)thebaine from (-i-)-6-demethylsalutaridine, and (-)-N-13CH3-thebaine from (-)northebaine Dumont, Raymond; Newman, Amy Hauck; Rice, Kenner C; Brossi, A.; Toome, Voldemar; Wegrzynski, Bogda FEBS Lett. (1986), 206(1), 125-9
99.
Synthesis of N-p-azidophenylethyl-7,8-dihydronormorphine and its 7,8-ditritio analog. Potential opiate receptor photoaffmity labels Cooper, Geoffrey K.; Rapoport, Henry J. LabeUed Compd. Radiopharm. (1985), 22(11), 1201-7
100.
Status of opioids in bovine milk and its influence on enzymes in milk synthesis Hossain, M. Afzal; Kumar, Amalendra; Ganguh, N. C. Indian J. Anim. Sci. (1986), 56(5), 588-92
101.
Purifying codeine synthesized from morphine Cemy, Jozef; Proksa, Bohumil; Brezovsky, Zdenko Czech., 2 pp. PI CS 223100 B 860315 AI CS 82-2895 820423
130 102.
Effects of synthetic.analogs of enkephalins, morphine and their antagonists on the course of experimental traumatic shock Motin, V. G.; Yasnetsov, V. V. Farmakol. Toksikol. (Moscow) (1986), 49(3), 103-7
103.
Morphine enhances gastric mucus synthesis in rats Ho, Mai M.; Ogle, Clive W.; Dai, Soter Eur. J. Pharmacol. (1986), 122(1), 81-6
104.
A short, stereospecific synthesis of a morphine fragment via an intramolecular Diels-Alder reaction Handa, Sheetal; Jones, Keith; Newton, Christopher G.; WiUiams, David J. J. Chem. Soc, Chem. Commun. (1985), (19), 1362-3
105.
Amino acid incorporation during morphine intoxication: I: Dose and time effects of morphine on protein synthesis in specific regions of the rat brain and in astrogUa-enriched primary cultures Roennbaeck, L.; Hansson, E. J. Neurosci. Res. (1985), 14(4), 461-77
106.
Analysis of the analgesic and anti-inflammatory effects of rimazolium, a pyridopyrimidine derivative, compared with that of prostaglandin synthesis inhibitors and morphine Gyires, K.; Furst, S.; Miklya, I.; Budavari, L; Knoll, J. Drugs Exp. CHn. Res. (1985), 11(8), 493-500
107.
Studies directed at a synthesis of the morphine alkaloids. Regiocontrol in Robinson-type annulations of 2-(hydroxymethyl)-4-oxo-3-piperidinecarboxyUc acid .gamma.-lactones Schultz, Arthur G.; Shannon, Paul J. J. Org. Chem. (1985), 50(23), 4421-5
108.
Isolation, sequencing, synthesis, and pharmacological characterization of two brain neuropeptides that modulate the action of morphine Yang, H. Y. T.; Fratta, W.; Majane, E. A.; Costa, E. Proc. Natl. Acad. Sci. U. S. A. (1985), 82(22), 7757-61
109.
Morphine-like structures and semi-rigid side chains as molecular probes for the bioactive conformation of enkephalin. Cardinaux, Francis; Pless, Janos Pept., Proc. Eur. Pept. Symp., 18th (1984), 321-4. Editor(s): Ragnarsson, Ulf. PubHsher: Almqvist & Wiksell, Stockholm, Swed.
110.
Opiate dependence and withdrawal - a new synthesis Hendrie, Colin A. Pharmacol., Biochem. Behav. (1985), 23(5), 863-70
111.
A direct synthesis of 06-monoacetylmorphine from morphine Sy, Wing Wah; By, Arnold W.; Neville, George A.; Wilson, WilUam L. J. - Can. Soc. Forensic Sci. (1985), 18(2), 86-91
112.
1 -Substituted isoquinoUnes Hendrickson, James B.; Rodriguez, Cesar
131 U.S., 7 pp. PI US 4514569 A 850430 AI US 82-343407 820128 113.
Chemistry of natural products, LXXXXV. Synthesis of 4,5.alpha.-epoxy17,6',6'-trimethyl-6'H-morphin-7-eno[3,2-b]pyran- 6.beta.-ol via PTC reaction Reisch, Johannes; Dharmaratne, H. Ranjith W. Arch. Pharm. (Weinheim, Ger.) (1985), 318(4), 382-4
114.
Recent progress on the total synthesis of morphine Szantay, Csaba; Blasko, Gabor; Barczai-Beke, Marietta; Domyei,Gabor; Pechy, Peter Magy. Kem. Lapja (1985), 40(1), 37-42
115.
Biochemical correlates of morphine withdrawal. 1. Characterization in the adrenal medulla and locus ceruleus DiStefano, Peter S.; Brown, Ohver M. J. Pharmacol. Exp. Ther. (1985), 233(2), 333-8
116.
Preparation of optically active, functionalized cis-.DELTA.6-1 -octalones Boger, Dale L.; Mullican, Michael D. J. Org. Chem. (1985), 50(11), 1904-11
117.
Synthesis and determination of 10-oxomorphine, a decomposition product of morphine Proksa, B. Pharmazie (1984), 39(10), 687-8
118.
Metorphamide - a morphine-like peptide Barchas, Jack D.; Weber, Eckard; Evans, Christopher J. U.S., 5 pp. PI U S 4490363 A 841225 AI U S 83-508140 830624
119.
Comparative studies on morphine- and stress-induced analgesia and the development of tolerance to the effects: implication of protein synthesis mechanism in the process Takahashi, Masakatsu; Izumi, Reiko; Kaneto, Hiroshi Jpn. J. Pharmacol. (1985), 37(2), 197-202
120.
Studies directed at a synthesis of the morphine alkaloids. A photochemical approach Schultz, Arthur G.; Lucci, Robert D.; Napier, James J.; Kinoshita, Hideki; Ravichandran, Ramanathan; Shannon, Paul; Yee, Ying K. J. Org. Chem. (1985), 50(2), 217-31
121.
Synthesis of the new morphine partial structure- 15,16-secomorphinan Helv. Chim. Acta (1984), 67(6), 1598-602
122.
In vivo alteration in hypothalamic amino acid synthesis during perfusion of ethanol and morphine in unrestrained rat Noto, T.; Myers, R. D. Neurochem. Res. (1984), 9(11), 1653-65
132 123.
Profile of amino acid synthesis in rat hippocampus during push-pull perfusion of ethanol or morphine Noto, T.; Hepler, J. R.; Myers, R. D. Neuroscience (Oxford) (1984), 13(2), 367-76
124.
Participation of opioid receptors and protein- synthetic mechanisms in the development of tolerance to morphine Hirota, Noritake; YamazaM, Akira; Takahashi, Masakatsu; Kaneto, Hiroshi Yakubutsu, Seishin, Kodo (1984), 4(1), 85-6
125.
Anew synthesis of morphine-based analgesics Chandler, Malcohn; Parson, Philip J. J. Chem. Soc, Chem. Conmiun. (1984), (5), 322-3
126.
Morphine diesters. I. Synthesis and action on guinea pig ileum Owen, J. A.; Elliott, J.; Jhamandas, K.; Nakatsu, K. Can. J. Physiol. Pharmacol. (1984), 62(4), 446-51
127.
Diastereoisomeric quaternary morphinium salts: synthesis stereochemistry and analgesic properties lorio, Maria A.; DisciuUo, Anna; Mazzeo-Farina, Anna; Frigeni, Viviana Eur. J. Med. Chem. - Chim. Ther. (1984), 19(1), 11-16
128.
Increase in 5-HT synthesis in the dorsal part of the spinal cord, induced by a nociceptive stimulus: blockade by morphine Weil-Fugazza, Jeanne; Godefroy, Francoise; Le Bars, Daniel J. Brain Res. (1984), 297(2), 247-64
129.
Studies aimed at the synthesis of morphine. 7. Biomimetic total synthesis of (.-i-.)-pallidine Blasko, Gabor; Domyei, Gabor; Barczai-Beke, Marietta; Pechy, Peter; Szantay, J. Org. Chem. (1984), 49(8), 1439-41
130.
Identification and determination of by-products of the codeine synthesis Proksa, B.; Cemy, J. Chem. Zvesti (1983), 37(6), 837-42
131.
Dynorphin amide analogs Lee, Nancy M.; Loh, Horace H.; Chang, Jaw Kang Eur. Pat. Appl., 24 pp. PI EP96592A1 831221 AI EP 83-303313 830608 PRAI US 82-387005 820610
132.
Studies on the synthesis of morphine. VI. Recent progress on the biomimetic synthesis of morphine Szantay, Csaba; Blasko, G.; Barczai-Beke, Marietta; Domyei, G.; Pechy, P. Planta Med. (1983), 48(4), 207-11
133.
A new synthetic metiiod for tiie preparation of tert-butyl aryl ethers Mohacsi, Emo Synth. Commun. (1983), 13(10), 827-32
133 134.
DELTA. 16,17-Dehydroheroinium chloride: synthesis and characterization of a novel impurity detected in illicit heroin Allen, Andrew C ; Moore, James M.; Cooper, Donald A. J. Org. Chem. (1983), 48(22), 3951-4
135.
The effect of morphine on 5 -hydroxytryptamine synthesis and metabolism in the striatum, and several discrete hypothalamic regions of the rat brain Johnston, C. A.; Moore, K. E. J. Neural Transm. (1983), 57(1), 65-73
136.
Effect of prostaglandin synthetase inhibitors on non-analgesic actions of morphine Wallenstein, Martin C. Eur. J. Pharmacol. (1983), 90(1), 65-73
137.
Mechanism of inhibition of protein synthesis by morphine in rat brain and hver Retz, Konrad C ; Steele, WiUiam J. Mol. Pharmacol. (1982), 22(3), 706-14
138.
Alkaloids. PartXLV. Partial synthesis of (-)-butorphanol Mouralova, J.; Hajicek, J.; Trojanek, J. Cesk. Farm. (1983), 32(1), 23-6
139.
An inhibitory role for morphine on the release of dopamine into hypophysial portal blood and on the synthesis of dopamine in tuberoinfundibular neurons Reymond, Marianne J.; Kaur, Charanjit; Porter, John C. Brain Res. (1983), 262(2), 253-8
140.
Blockade of morphine dependence-related enhancement of secretory protein synthesis in the pons-medulla and striatum-septum by naltrexone Retz, K. C ; Steele, W. J. Neuropharmacology (1983), 22(2), 183-9
141.
Total synthesis of (.+-.)-3-deoxy-7,8-dihydromorphine, (.-i-.)-4-methoxy-Nmethylmorphinan-6-one and 2,4-dioxygenated (.+-.)-congeners Hsu, Fu Lian; Rice, Kenner C ; Brossi, Arnold Helv. Chim. Acta (1982), 65(5), 1576-89
142.
Stereospecific synthesis of a new morphine partial structure Haefliger, Walter; Kloeppner, Edgar Helv. Chim. Acta (1982), 65(6), 1837-52
143.
Tryptophan and tryptophan pyrrolase in heme regulation. The role of hpolysis and direct displacement of serum protein-bound tryptophan in the opposite effects of administration of endotoxin, morphine, palmitate, salicylate, and theophylline on rat hver 5-aminolevulinate synthase activity and the heme saturation of tryptophan pyrrolase Badaway, Abdulla A. B.; Morgan, Christopher J. Biochem. J. (1982), 206(3), 451-60
144.
How are morphine-Hke peptides synthesized?
134 Imura, Hiroo Gendai Kagaku (1982), 140, 28-31 145.
Myocardial protein synthesis in magnesium deficiency: effects of morphine Essman, Walter B.; Essman, Eric J. Monogr. Am. Coll. Nutr. (1982), 5(Nutr. Heart Dis.), 309-14
146.
Synthesis of some carbon-11 -labeled alkaloids Laangstroem, B.; Antoni, G.; Halldin, C.; Svaerd, H.; Bergson, G. Chem. Scr. (1982), 20(1-2), 46-8
147.
Biological activity of some synthesized compounds of the pyridine series Poddubnaya, L. V.; Olekhnovich, L. B. Izv. Sev.-Kavk. Nauchn. Tsentra Vyssh. Shk., Estestv. Nauki (1982), (1), 87-9
148.
Aporphines. 42. Synthesis of (R)-(-)-ll-hydroxyaporphines Ram, Vishnu J.; Neumeyer, John L. J. Org. Chem. (1982), 47(22), 4372-4
149.
Studies aimed at the synthesis of morphine. V. An economy approach to (.+-.)-reticuline from 3,4-dihydropapaverine Domyei, Gabor; Barczai-Beke, Marietta; Balsko, Gabor; Pechy, Peter; Szantay, Tetrahedron Lett. (1982), 23(28), 2913-16
150.
Synthesis and pharmacology of metabolically stable tert-butyl ethers of morphine and levorphanol Mohacsi, Emo; Leimgruber, Willy; Baruth, Herman J. Med. Chem. (1982), 25(10), 1264-6
151.
.beta.-Endorphin. Interaction of synthetic analogs having different chain lengths with morplune and enkephalin receptors in rat brain membranes Ferrara, Pascual; Li, Choh Hao Int. J. Pept. Protein Res. (1982), 19(3), 259-62
152.
Evaluation of the analgesic activity of morphine, enkephalins and their synthetic analogs Aloisi, P.; Scotti de Carolis, A.; Longo, V. G. Isr. J. Med. Sci. (1982), 18(1), 183-6
153.
Studies aimed at the synthesis of morphine. FV. A new approach to Nnorreticuline derivatives from homoveratronitrile Szantay, Csaba; Domyei, Gabor; Blasko, Gabor; Barczai-Beke, Marietta; Pechy, Peter Arch. Pharm. (Weinheim, Ger.) (1981), 314(12), 983-91
154.
Synthesis and biological properties of enkephalin-like peptides containing adamantylalanine in position 4 and 5 Kim Quang Do; Schwyzer, Robert Helv. Chim. Acta (1981), 64(7), 2084-9
155.
Studies aimed at the synthesis of morphine. 3. Synthesis of (.+-.)-salutaridine via phenoUc oxidative coupling of (.-i-.)-reticuline Szantay, Csaba; Barczai-Beke, Marietta; Pechy, Peter; Blasko, Gabor;
135 Domyei, Gabor J. Org. Chem. (1982), 47(3), 594-6 156.
Synthesis of serotonin at the spinal level in the rat: modifications induced by nociceptive somatic stimulation, associated or not with administration of morphine Weil-Fugazza, Jeanne; Godefroy, Francoise; Chitour, Djamel; Le Bars, Daniel C. R. Seances Acad. Sci., Ser. 3 (1981), 293(1), 89-92
157.
(-)-4-Hydroxymorphinanones: their synthesis and analgesic activity Manmade, Awinash; Dalzell, Haldean C ; Howes, John P.; Razdan, Raj K. J. Med. Chem. (1981), 24(12), 1437-40
158.
Restoration of pituitary prolactin synthesis and release by the administration of morphine to rats bearing a transplanted prolactin-secreting tumor Login, Ivan S.; Nagy, Ivan; MacLeod, Robert M. Neuroendocrinology (1981), 33(2), 101-4
159.
Studies related to the synthesis of morphine. n. Conversion of 06-demethylsalutaridine into salutaridine Horvath, G.; Makleit, S. Acta Chim. Acad. Sci. Hung. (1981), 106(1), 37-42
160.
Short total synthesis of dihydrothebainone, dihydrocodeinone and nordihydrocodeinone Rice, Kenneth C. U. S. Pat. AppL, 15 pp. Avail. NTIS. Order No. PAT-APPL-6-165 690 PI US 165690 801219 AI US 80-165690 800703
161.
Approaches towards a practical synthesis of morphine Brossi, Arnold Proc. Asian Symp. Med. Plants Spices, 4th (1981), Meeting Date 1980, Volume 1, 261-74. Editor(s): Kusamran, Kosan; Pohmakotr, Manat; Reutrakul, Vichai. Publisher: Aksom Charoen - Tat Publ. House, Bangkok, Thailand.
162.
Paper electrophoresis for the determination of some synthetic morphine substitutes in chemicotoxicological analysis Mikhno, V. V. Farm. Zh. (Kiev) (1981), (2), 49-51
163.
Restoration of pituitary prolactin synthesis and release by the administration of morphine to rats bearing a transplanted prolactin-secreting tumor Login, Ivan S.; Nagy, Ivan; MacLeod, Robert M. Neuroendocrinology (1981), 33(2), 101-4
164.
Morphiceptin (NH4-Tyr-Pro-Phe-Pro-CONH2): a potent and specific agonist for morphine (.mu.) receptors Chang, Kwen-Jen; Killian, Anthony; Hazum, Eli; Cuatrecasas, Pedro; Chang, Jaw-Kang Science (Washington, D. C , 1883-) (1981), 212(4490), 75-7
136 165.
The synthesis and agonist activity of some 14.beta.-substituted morphine and codeine derivatives Osei-Gyimah, Peter; Archer, Sydney Endog. Exog. Opiate Agonists Antagonists, Proc. Int. Narc. Res. Club Conf. (1980), Meeting Date 1979, 13-16. Editor(s): Way, E. Leong. Publisher: Pergamon, Elmsford, N. Y.
166.
Morphine analgesia and newly synthesized 5-hydroxytryptamine in the dorsal and the ventral halves of the spinal cord of the rat Weil-Fugazza, J.; Godefroy, F.; Coudert, D.; Besson, J. M. Brain Res. (1981), 214(2), 440-4
167.
Stimulation of dopamine synthesis and release by morphine and D-Ala2-D-Leu5enkephalin in the mouse striatus in vivo Urwyler, Stephan; Tabakoff, Boris Life Sci. (1981), 28(20), 2277-86
168.
Chemicalstudiesondrugmetabolism.lv. Synthesis and analgesic activity of morphine-7,8-oxide and heroin-7,8-oxide Miyata, Naoki; Uba, Kiyoko; Watanabe, Keizo; Hirobe, Masaaki Chem. Pharm. Bull. (1980), 28(12), 3722-4
169.
Morphine-like activities of synthetic enkephalin analogs Moritoki, H.; Kiso, Y.; Kageyama, T.; Matsumoto, K. J. Pharm. Pharmacol. (1981), 33(1), 54-5
170.
The effects of melanocyte-stimulating hormone release inhibiting factor and synthetic analogs on tolerance to and physical dependence on morphine Bhargava, Hemendra N.; Ritzmann, Ronald F.; Walter, Roderich Endog. Exog. Opiate Agonists Antagonists, Proc. Int. Narc. Res. Club Conf. (1980), Meeting Date 1979, 545-8. Editor(s): Way, E. Leong. Publisher: Pergamon, Elmsford, N. Y.
171.
Comparison of the analgetic action of inhibitors of PG synthesis, Chinoin-127 and morphine Gyires, Klara; Knoll, J. Adv. Pharmacol. Res. Pract., Proc. Congr. Hung. Pharmacol. Soc, 3rd (1980), Meeting Date 1979, Volume 5, Issue Opiate Recept. Neurochem. Correl. Pain, 227-33. Editor(s): Furst, Susanna. PubUsher: Pergamon, Oxford, Engl.
172.
Total synthesis of (.+-.)-3-deoxy-7,8-dihydromorphine. Preliminary communication Hsu, Fu Lian; Rice, Kenner C ; Brossi, Arnold Helv. Chim. Acta (1980), 63(7), 2042-5
173.
Studies aiming at the synthesis of morphine. n. Studies on phenoUc couphng of N-norreticuhne derivatives Szantay, Csaba; Blasko, Gabor; Barczai-Beke, Marietta; Pechy, Peter; Domyei, Gabor Tetrahedron Lett. (1980), 21(36), 3509-12
137 174.
Morphine-like peptides: their regulation in the neuroendocrine system and the effect of guanyl nucleotides and divalent ions on opiate receptor binding Simantov, R. Prog. Biochem. Pharmacol. (1980), 16(Endog. Pept. Cent. Acting Drugs), 22-31
175.
Effects of endogenous opioid peptides and their synthetic analogs on anterior pituitary function Halasz, B.; Marton, J.; Koves, K.; Nagy, G.; Vizi, E.; Bajusz, S.; Molnar, J.; Lukats, O. Recent Results Pept. Horm. Androg. Steroid Res., Proc. Congr. Hung. Soc. Endocrinol. Metab., 9th (1979), 67-73. Editor(s): Laszlo, F. A. PubUsher: Akad. Kiado, Budapest, Hung.
176.
Studies on the synthesis of morphine, n. Conversion of 06demethylsalutarichne into salutaridme Horvath, Geza; Makleit, Sandor Magy. Kem. Foly. (1980), 86(6), 260-3
177.
Studies aiming at the synthesis of morphine. I. Separation and characterization of the amide rotamers of 6'-halogeno-N-formylnorreticulines Szantay, Csaba; Blasko, Gabor; Barczai-Beke, Marietta; Domyei, Gabor; Radics, Lajos Heterocycles (1980), 14(8), 1127-30
178.
Synthetic and pharmacological studies with enkephalin analogs in relation to structural features of morphine DiMaio, J.; Schiller, P. W.; Belleau, B. Pept., Struct. Biol. Funct., Proc. Am. Pept. Symp., 6th (1979), 889-92. Editor(s): Gross, Erhard; Meienhofer, Johannes. PubUsher: Pierce Chem. Co., Rockford, 111.
179.
Apphcation of metalated enamines to alkaloid synthesis. An expedient approach to the synthesis of morphine based analgesics Evans, D. A.; Mitch, C. H.; Thomas, R. C ; Zimmerman, D. M.; Robey, R. L. J. Am. Chem. Soc. (1980), 102(18), 5955-6
180.
Some remarks on the development of synthetic analgesics with morphine-like effect Halbach, H. Wien. Z. Suchtforsch. (1980), 3(3), 27-8
181.
Effect of acute administration of morphine on newly synthesized 5-hydroxytryptamine in spinal cord of the rat Godefroy, Francoise; Weil-Fugazza, Jeanne; Coudert, Danielle; Besson, Jean Marie Brain Res. (1980), 199(2), 415-24
182.
Morphine differentially alters synthesis and turnover of dopamine in central neuronal systems Alper, R. H.; Demarest, K. T.; Moore, K. E. J. Neural Transm. (1980), 48(3), 157-65
183.
Influence of prostaglandin synthetase inhibitors on the analgesic activity of morphine in the rat
138 Poggioli, R.; Castelli, M.; Genedani, S.; Bertolini, A. Riv. FarmacoL Ter. (1980), 11(1), 11-14 184.
Synthesis of 10(S)-methylcodeine and 10(S)-methylmorphine Arzeno, Humberto B.; Barton, Derek H. R.; Davies, Stephen G.; Lusinchi, Xavier; Meunier, Bernard; Pascard, Claudine Nouv. J. Chim. (1980), 4(6), 369-75
185.
Modulation of ganglioside synthesis by enkephalins, opiates, and prostaglandins. Role of cychc AMP in glycosylation Dawson, Glyn; McLawhon, Ronald W.; Schoon, Gwendolyn; Miller, R. J. ACS Symp. Ser. (1980), 128(Cell Surf. GlycoUpids), 359-72
186.
Comparative effects of synthetic enkephalinamides and morphine on abstinence responses in morphine-dependent mice Bhargava, Hemendra N. Pharmacol., Biochem. Behav. (1980), 12(5), 645-9
187.
Activities of hepatic carbamoyl phosphate synthetase and argininosuccinate synthetase after morphinization Wong, S. C ; Yeung, D. IRCS Med. Sci.: Libr. Compend. (1980), 8(3), 151
188.
Synthesis of 4a-ary 1-decahydroisoquinolines Rapoport, Henry; Weller, Dwight D.; Gless, Richard D. U. S. Pat. AppL, 46 pp. Avail. NTIS. US 31749 791221 US 79-31749 790420
189.
Benzomorphans - clinically used synthetic analogues of morphine. A short review for nonpharmacologists Palmer, David C ; Strauss, Michael J. Ind. Eng. Chem. Prod. Res. Dev. (1980), 19(2), 172-4
190.
Syntiiesisof morphine alkaloid analogs. Hasubanans and 9,17-secomorphinans Schwartz, Martin A.; Wallace, Rebecca A. Tetrahedron Lett. (1979), (35), 3257-60
191.
Synthesis and kinetic study of the hydrolysis of 4-cyano-4-phenylpiperidines: access to morphine analogs Gervais, Christian; Anker, Daniel; Chareire, Martine; Pacheco, Henri Bull. Soc. Chim. Fr. (1979), (5-6, Pt. 2), 241-8
192.
Synthesis of morphine analogs. Part 2 Substituted pyrrolo[3,4-h]isoquinoUnes Lindsay Smith, John R.; Norman, Richard O. C ; Rose, Malcolm E.; Curran, Adrian C. w.; Lewis, John W. J. Chem. Soc, Perkin Trans. 1 (1979), (11), 2868-72
193.
Syntiiesisof morphine analogs. Parti. Syntiiesis of some 5-benzyl-2-methyloctahydroisoquinoUnes Lindsay Smitii, John R.; Norman, Richard O. C ; Rose, Malcohn E.; Curran, Adrian C. W. J. Chem. Soc, Perkin Trans. 1 (1979), (11), 2863-7
139 194.
Chemical studies on drug metabolism. Part 1. Stereospecific synthesis of codeine 7,8-oxide and codeinone 7,8-oxide Uba, Kiyoko; Miyata, Naoki; Watanabe, Keizo; Hirobe, Masaaki Chem. Pharm. Bull (1979), 27(9), 2257-8
195.
Analgesic narcotic antagonists. 1. 8.beta.-Alkyl-, 8.beta.-acyl-, and 8.beta.(tertiary alcohol)dihydrocodeinones and-dihydromorphinones Kotick, Michael P.; Leland, David L.; Polazzi, Joseph O.; Schut, Robert N. J. Med. Chem. (1980), 23(2), 166-74
196.
Synthesis and analgesic activity of somel4.beta.-substituted analogs of morphine Osei-Gyimah, Peter; Archer, Sydney J. Med. Chem. (1980), 23(2), 162-6
197.
Oxidative phenol coupling in the synthesis of morphine alkaloids and related compounds Schwartz, Martin A. Symp. Pap. - lUPAC Int. Symp. Chem. Nat. Prod., 11th (1978), Volume 4, Issue Part 2, 274-9. Editor(s): Marekov, N.; Ognyanov, I.; Orahovats, A. Publisher: Izd. BAN, Sofia, Bulg.
198.
Synthesis of 2-aminomorphine and 2-aminocodeine. Reduction of aromatic nitro groups with formamidinesulfmic acid Chatterjie, Nithiananda; Minar, Arlene; Clarke, Donald D. Synth. Commun. (1979), 9(7), 647-57
199.
Synthesis, chromatography and tissue distribution of methyl-11C- morphine and methyl-11 C-heroin Kloster, G.; Roeder, E.; MachuUa, H. J. J. LabeUed Compd. Radiopharm. (1979), 16(3), 441-8
200.
Presynaptic effect of morphine and haloperidol on dopamine synthesis Shyu, Bai-Chuang; Chiang, Yi; Wang, Wei-Kung BuU. Inst. ZooL, Acad. Sin. (1978), 17(2), 125-30
201.
Dual action of methadone on 5-HT synthesis and metabohsm Ahtee, Liisa; Carlsson, Arvid Naunyn-Schmiedeberg's Arch. Pharmacol. (1979), 307(1), 51-6
202.
Physiologically active nitrogeneous organic compounds. Synthesis of morphinetype analgesics Kametani, Tetsuji; Fukumoto, Keiichiro Kagaku, Zokan (Kyoto) (1979), (79), 77-97
203.
Morphine-like analgesia by a new dipeptide, L-tyrosyl-L-arginine (kyotorphin) and its analog Takagi, Hiroshi; Shiomi, Hirohito; Ueda, Hiroshi; Amano, Hiro Eur. J. Pharmacol. (1979), 55(1), 109-11
204.
Influence of morphine, .beta.-endorphin and naloxone on the synthesis of phosphoinositides in the rat midbrain Natsuki, R.; Hitzemann, R. J.; Loh, H. H. Res. Commun. Chem. Pathol. Pharmacol. (1979), 24(2), 233-50
140 205.
The synthesis rate and turnover time of S-hydroxytryptamine in brains of rats treated chronically with morphine Bhargava, Hemendra N. Br. J. Pharmacol. (1979), 65(2), 311-17
206.
Opiate-receptor mediated changes in monoamine synthesis in rat brain Garcia-Sevilla, J. A.; Ahtee, Liisa; Magnusson, T.; Carlsson, A. J. Pharm. Pharmacol. (1978), 30(10), 613-21
207.
Effects of morphine, naloxone, naltrexone and .beta.-endorphin on monoamine synthesis in rat brain Ahtee, Liisa; Garcia-Sevilla, J. A.; Magnusson, T.; Carlsson, A. Dev. Neurosci. (Amsterdam) (1978), 4(Charact. Funct. Opioids), 345-6
208.
The effects of synthetic and natural opioid peptides in isolated organs Ronai, A. Z.; Berzetei, L; Szekely, J. I.; Bajusz, S. Dev. Neurosci. (Amsterdam) (1978), 4(Charact. Funct. Opioids), 493-4
209.
In vivo antagonism by naloxone of morphine, .beta.-endorphin and a synthetic enkephalin analog Szekely, J. I.; Dunai-Kovacs, Zsuzs; Miglecz, Erzsebet; Ronai, A. Z.; Bajusz, S. J. Pharmacol. Exp. Ther. (1978), 207(3), 878-83
210.
The effect of morphine tolerance and dependence on cell free protein synthesis Craves, Frederick B.; Loh, Horace H.; Meyerhoff, James L. J. Neurochem. (1978), 31(5), 1309-16
211.
Pentapeptide with morphine-like activity Sarantakis, Dimitrios; Grant, Norman H. US 4128541 781205 US 76-754794 761227
212.
Effect of some prostaglandin synthesis inhibitors on the antinociceptive action of morphine in albino rats Srivastava, D. N.; Bhattacharya, S. K.; Sanval, A. K. Clin. Exp. Pharmacol. Physiol. (1978), 5(5), 503-9
213.
Polypeptides with morphine-like activity Sarantakis, Dimitrios US 4098781 780704 US 77-777181 770314
214.
Synthesis of opioid pentapeptides, Leu-enkephahn and Met-enkephalin by solution methods and their biological activities Wang, Kung-Tsung; Chen, Shui-Tien; Huang, Min-Che; Chang, Chuan-Chiung Proc. Natl. Sci. Counc, Repub. China (1978), 2(3), 232-7
215.
Morphine and related compounds - inducers of specific antibody synthesis(literature review) Kovalev, I. E. Khim.-Farm. Zh. (1978), 12(9), 3-14
216.
Polypeptides with morphine-like activity
141 Sarantakis, Dimitrios; Stein, Larry US 4097471 780627 US 77-800678 770526 217.
Improved synthesis of (+)- morphine, (+)-cocleine and (+)-heroine Rice, Kenner C ; lijima, Ikuo; Brossi, Arnold Symp. HeterocycL, [Pap.] (1977), 49. Editor(s): Kametani, Tetsuji. Pubhsher: Sendai Inst. Heterocycl. Chem., Sendai, Japan.
218.
Inhibition of protein synthesis and tolerance to opiates Delia Bella, D.; Frigeni, V. Factors Affecting Action Narc, [Proc. Meet.] (1978), Meeting Date 1976,403-12. Editor(s): Adler, Martin W.; Manara, Luciano; Samanin, Rosario. Publisher: Raven, New York, N. Y.
219.
Importance of in vitro measurements of adrenocortical steroid synthesis in assessing locus of morphine effects Essman, Walter B.; Rosenthal, Richard Factors Affecting Action Narc, [Proc. Meet.] (1978), Meeting Date 1976,125-31. Editor(s): Adler, Martin W.; Manara, Luciano; Samanin, Rosario. Publisher: Raven, New York, N. Y.
220.
Morphine and .beta.-endorphin on RNA synthesis Lee, Nancy M.; Loh, Horace H.; Li, Choh Hao Adv. Biochem. Psychopharmacol. (1978), 18(Endorphins), 278-88
221.
Alteration of brain chromatin and nuclear synthetic activity in morphine-tolerant rats Sprague, G. L.; Fong, D. W.; Castles, T. R. Res. Commun. Chem. Pathol. Pharmacol. (1978), 19(3), 553-6
222.
Pentapeptide with morphine-like activity Sarantakis, Dimitrios US 4075190 780221 US 77-777211 770314
223.
Studies on morphine and related compounds: synthesis of 2,5,9-trimethyl- and 2ethyl-5,9-dimethyl- 3,4:6,7-dibenzomorphans Sharma, Rajendra; Goyal, V. K.; Tyagi, R. P.; Joshi, B. C. J. Indian Chem. Soc. (1977), 54(7), 753-5
224.
Effects of inhibitors of catecholamine synthesisand catecholamine agonists on morphine- and hypoglycemia-induced release of growth hormone in the rat Bluet-Pajot, M. T.; Schaub, C. J. Endocrinol. (1978), 76(2), 365-6.
225.
Conversions of tosyl and mesyl derivatives of the morphine group, XX. Synthesis of mesyl esters Makleit, Sandor; Berenyi, Sandor; Bognar, Rezso; Elek, Sandor Acta Chim. Acad. Sci. Hung. (1977), 94(2), 161-3
226.
Studies in the (-i-)-morphinan series. 4. A markedly improved synthesis of {+)- morphine
142 lijima, Ikuo; Minamikawa, Junichi; Jacobson, Arthur E.; Jacobson, Arthur E.; Rice, Kenner C. J. Org. Chem. (1978), 43(7), 1462-3 227.
Effect of antiphlogistics, adrenaline, histone, morphine and other substances on prostaglandin synthesis by rabbit renal medulla sUces Bekemeier, H.; Giessler, A. J.; Hirschelmann, R. Congr. Hung. Pharmacol. Soc, [Proc] (1976), Volume Date 1974, 2(2, Symp. Prostaglandins), 183-91
228.
Synthesis and analgetic activity of 1,2,3,4,5,6-hexahydro1,6-methano-3-benzazocines Mazzocchi, Paul H.; Harrison, Aline M. J. Med. Chem. (1978), 21(2), 238-40
229.
.beta.-Endorphin: synthesis and morphine-like activity of analogs with D-amino acid residues in positions 1,2,4 and 5 Yamashiro, Donald; Tseng, Liang-Fu; Doneen, Byron A.; Loh, Horace H.; Li, Choh Hao Int. J. Pept. Protein Res. (1977), 10(2), 159-66
230.
Isolation, structure, synthesis and morphine -like activity of .beta.-endorphin from human pituitary glands Li, Choh Hao; Yamashiro, Donald; Chung, David; Doneen, Byron A.; Loh, Horace H.; Tseng, Liang-Fu Ann. N. Y. Acad. Sci. (1977), 297, 158-66
231.
A synthetic enkephalin analog with prolonged parenteral and oral analgesic activity Roemer, Dietmar; Buescher, Heinz H.; HiU, Ronald C ; Pless, Janos; Bauer, Wilfried; Cardinaux, Francis; Closse, Annemarie; Hauser, Daniel; Huguenin, Rene Nature (London) (1977), 268(5620), 547-9
232.
Synthesizing codeinone from thebaine Calvo, Fernando 4052402 771004 AI US 76-666663 760315
23 3.
New N-2-hydroxyalkyl-6,7-benzomorphan derivatives: Synthesis and preliminary pharmacology Rahtz, Dieter; Paschelke, Gert; Schroeder, Eberhard Eur. J. Med. Chem. - Chim. Ther. (1977), 12(3), 271-8
234.
Synthesis and antitussive activity of 3- azabicyclo[3.2.2]nonane derivatives Arya, V. P.; Kaul, C. L.; Grewal, R. S. Arzneim.-Forsch. (1977), 27(9), 1648-52
235.
Value of the vinyloxycarbonyl unit in hydroxyl protection: application to the synthesis of nalorphine Olofson, R. A.; Schnur, Rodney C. Tetrahedron Lett. (1977), (18), 1571-4
236.
Converting allyhc ethers to dienol ethers Rapoport, Henry; Barber, Randy B.
143 U. S. Pat. Appl., 17 pp. Avail. NTIS. US 724018 760916 US 76-724018 760916 237.
Influence of morphine on protein synthesis in synaptic plasma membranes of the rat brain Hitzemann, Robert J.; Loh, Horace H. Res. Commun. Chem. Pathol. Pharmacol. (1977), 17(1), 15-28
238.
A comparison of the effects of morphine and other psychoactive drugs on brain phosphohpid synthesis Natsuki, Reiko; Hitzemann, Robert; litzemann, Barbara; Loh, Horace Opiates Endog. Opioid Pept., Proc. Int. Narc. Res. Club Meet. (1976), 451-4. Editor(s): KosterUtz, H. W. PubUsher: North-Holland, Amsterdam, Neth.
239.
Differential effects of morphine on rates of translation in vivo in free and bound polysome compartments of rat brain regions Ramsey, James C ; Steele, William J. Opiates Endog. Opioid Pept., Proc. Int. Narc. Res. Club Meet. (1976), 435-8. Editor(s): Kosterlitz, H. W. Publisher: North-Holland, Amsterdam, Neth.
240.
Control of tolerance reversal by protein synthesis inhibitors DeUa Bella, Davide; Frigeni, Viviana; Sassi, Alessandro Opiates Endog. Opioid Pept., Proc. Int. Narc. Res. Club Meet. (1976), 353-60. Editor(s): Kosterhtz, H. W. Publisher: North-Holland, Amsterdam, Neth.
241.
Some morphine-like properties of a potent antinociceptive synthetic pentapeptide in relation to physical dependence in rodents Baxter, M. G.; Follenfant, R. L.; Miller, A. A.; Sethna, D. M. Br. J. Pharmacol. (1977), 59(3), 523P
242.
Alteration of RNA biosynthesis induced by chronic treatments of opiate Lee, N. M.; Loh, H. H. Opiates Endog. Opioid Pept., Proc. Int. Narc. Res. Club Meet. (1976), 423-6. Editor(s): Kosterhtz, H. W. PubUsher: North-Holland, Amsterdam, Neth.
243.
3H-Lysine accumulation into brain and plasma proteins in male rats treated with morphine, naloxone, or naloxone plus morphine in relation to controls Ford, D. H.; Rhines, R. K.; Joshi, M.; Cheslak, H.; Simmons, H.; Toth, J. Tissue Responses Addict. Drugs, [Proc. Workshop Sess. Int. Soc. Neuroendocrinol.] (1976), Meeting Date 1975, 489-507. Editor(s): Ford, Donald Herbert; Clouet, Doris H. PubUsher: Spectrum Publ., Inc., HolUswood, N. Y.
244.
The effects of morphine-like drugs and chlorpromazine on the synthesis of glyceroUpids by homogenates of rat Uver Sanderson, Robert F.; Dodds, Peter F.; Brindley, David N. Biochem. Soc. Trans. (1977), 5(1), 295-6
144 245.
The effect of morphine on the metabolic transport of 3H-lysine and incorporation into protein Levi, M. A.; Rhines, R. K.; Ford, D. H. Tissue Responses Addict. Drugs, [Proc. Workshop Sess. Int. Soc. Neuroendocrinol.] (1976), Meeting Date 1975, 471-88. Editor(s): Ford, Donald Herbert; Clouet, Doris H. Publisher: Spectrum PubL, Inc., Holliswood, N. Y.
246.
Effect of a potent synthetic opioid pentapeptide in some antinociceptive and behavioral tests in mice and rats Baxter, M. G.; Goff, D.; Miller, A. A.; Saunders, I. A. Br. J. Pharmacol. (1977), 59(3), 455P-456P
247.
Electrooxidation and biosynthesis of natural products Tobinaga, Seisho Kagaku To Seibutsu (1977), 15(4), 240-2
248.
Stereochemical studies on medicinal agents. 23. Synthesis and biological evaluation of 6-amino derivatives of naloxone and naltrexone Jiang, Jack B.; Hanson, Robert N.; Portoghese, Philip S.; Takemori, A. E. J. Med. Chem. (1977), 20(8), 1100-2
249.
Synthesis of 4a-aryldecahydroisoquinohnes. FunctionaUty in the carbocychc ring Weller, Dwight D.; Gless, Richard D.; Rapoport, Henry J. Org. Chem. (1977), 42(9), 1485-95
250.
Heteroatom directed photoarylation: synthesis of a tetracycUc morphine structural analog Schultz, Arthur G.; Lucci, Robert D. J. Chem. Soc, Chem. Commun. (1976), (22), 925
251.
N-Acyl-N-norsalutaridines Schwartz, Martin Alan U.S., 9 pp. US 4003903 770118 US 75-549491 750212
252.
Studies on the syntheses of heterocyclic compounds. Part DCLXIX. Studies on the syntheses of analgesics. XLI. Optical resolution of (.-i-.)-N-cyclopropylme3iyl-3-hydroxy-9azamorphinan Kametani, Tetsuji; Kigasawa, Kazuo; Hiiragi, Mineharu; Wagatsuma, Nagatoshi; Kusama, Osamu; Uryu, Tsuneo Chem. Pharm. Bull. (1976), 24(10), 2563-6
253.
Morphine effects on RNA synthesis in purified ohgodendroglial nuclei Stokes, K. B.; Lee, N. M. Proc. West. Pharmacol. Soc. (1976), 19, 48-54
254.
The inhibition of the effects of morphine by synthetic substance P
145 Stern, P.; Hukovic, S.; Radivojevic, M. Experientia (1976), 32(10), 1326-7 255.
Synthesis of morphine-cl5 and codeine-d8 Lawson, John A.; DeGraw, Joseph I.; Anbar, Michael J. Heterocycl. Chem. (1976), 13(3), 593-5
256.
Screening of new synthesized compounds for analgesic effect according to kiioll's method Nikolova, M.; Stefanova, D. Farmatsiya (Sofia) (1975), 25(4), 47-53
257.
Influence of morphine on protein synthesis in discrete subcellular fractions of the rat brain Hitzemann, Robert J.; Loh, Horace H. Res. Commun. Chem. Pathol. Pharmacol. (1976), 14(2), 237-48
258.
Response of division-synchronized protozoa to morphine and levorphanol McClean, Daniel K.; Zimmerman, Arthur M. Gen. Pharmacol. (1975), 6(2-3), 171-9
259.
Perinatal narcotic addiction in mice. Sensitization to morphine stimulation Shuster, L.; Webster, G. W.; Yu, G. Perinat. Addict., Proc. Conf. (1975), Meeting Date 1974, 277-91. Editor(s): Harbison, Raymond D. Pubhsher: Spectrum Publ., Inc., Holliswood, N. Y.
260.
Synthesis of new morphine derivatives Bognar, R.; Gaal, G.; Horvath, G.; Kerekes, P. Izv. Khim. (1975), 8(1), 194-202
261.
A synthetic peptide with morphine-like pharmacologic action Goldstein, Avram; Goldstein, Joshua S.; Cox, B. M. Life Sci. (1975), 17(11), 1643-54
262.
The preparation and synthetic utility of 0-substituted normethylmorphines Borowitz, Irving J.; Diakiw, Vladimir J. Heterocycl. Chem. (1975), 12(6), 1123-6
263.
Tolerance to morphine-induced calcium depletion in regional brain areas. Characterization with reserpine and protein syntiiesis inhibitors Ross, D. H. Br. J. Pharmacol. (1975), 55(3), 431-7
264.
Conversion of tosyl and mesyl derivatives of the morphine group. XV. Synthesis of pseudocodeine tosylate and the study of its nucleophilic reaction Makleit, Sandor; Somogyi, Gabor; Bognar, Rezso Magy. Kem. Foly. (1975), 81(11), 517-19
265.
Protein synthesis in mouse brain during development of acute morphine tolerance Nakajima, Tohru; Sasano, Hiroshi; Koida, Masao; Kaneto, Hiroshi Jpn. J. Pharmacol. (1975), 25(4), 367-74
266.
Influence of peptides on reduced response of rats to electric footshock after acute administration of morphine Gispen, Willem H.; Van Wimersma Greidanus, Tjeerd B.; Waters-Ezrin, Cheryl; Zimmermann, Emery; Krivoy, William A.; De Wied, David Eur. J. Pharmacol. (1975), 33(1), 99-105
267.
Mechanisms of appearance of morphine dependency about the synthesis of a morphine artificial antigen Aoki, Masatada; Kusama, Tadashi; Kuboyama, Noboru; Murakoshi, Yoshie Nippon Daigaku Yakugaku Kenkyu Hokoku (1974), 14, 8-14
268.
In vivo protein synthesis in morphinetolerant monkey brain Hollinger, Mannfred A.; KiUam, Keith F.; Deneau, Gerald A. Psychopharmacol. Commun. (1975), 1(3), 319-25
269.
Morphine. Synthesis of 2'-hydroxy-2,5-dimethyl3,4:6,7-dibenzomorphan. EI Pandey, R. K.; Pandey, P.; Joshi, B. C. Bull. Acad. Pol. Sci., Ser. Sci. Chim. (1975), 23(6), 469-72
270.
Synthesis of thebaine and oripavine from codeine and morphine Barber, Randy B.; Rapoport, Henry J. Med. Chem. (1975), 18(11), 1074-7
271.
Acute effects of heroin and morphine on newly synthesized serotonin in rat brain Perez-Cruet, Jorge; Nguyen Bich Thoa; Ng, Larry K. Y. Life Sci. (1975), 17(3), 349-62
272.
Synthesis of morphine-3-glucuronide Berrang, Bertold; Twine, Charles E.; Hennessee, G. L.; Carroll, F.I. Synth. Commun. (1975), 5(3), 231-6
273.
Syntheses of analgesics. DI. Synthesis of l,2,3,4,5,6-hexahydro-3-benzazocine derivatives. 2 Sawa, Yoichi; Kato, Takeshi; Morimoto, Akira; Tom, Masuda; Hori, Mikio; Fujimura, Hajime Yakugaku Zasshi (1975), 95(3), 261-8
274.
Elimination of phenolic hydroxyl groups in morphine alkaloids. Synthesis of 3-deoxydihydromorphine Bognar, Rezso; Gaal, Gyorgy; Kerekes, Peter; Horvath, Geza; Kovacs, Maria T. Magy. Kem. Foly. (1975), 81(2), 51-3
275.
Apomorphine and morphine stimulate prostaglandin biosynthesis ColHer, H. O. J.; McDonald-Gibson, Wendy J.; Saeed, S. A. Nature (London) (1974), 252(5478), 56-8
276.
Biogenetically patterned synthesis of the morphine alkaloids Schwartz, Martin A.; Mami, Ismail S. J. Am. Chem. Soc. (1975), 97(5), 1239-40
277.
Effect of morphine on the turnover and synthesis
147 3H-leucine-labeled protein and 14C-choline-iabeled phosphatidylcholine in discrete regions of the rat brain Loh, Horace H.; Hitzeman, Robert J. Biochem. Pharmacol. (1974), 23(12), 1753-65 278.
Synthetic compounds, pentazocine and Darvon, which have moiphine-like activity, and their effects on the pituitary-adrenocortical system. Morphine-like inhibitory effect on histamine ACTH-releasing action Chang, Wung Hua; Hiraga, Kogo Tokyo Jikeikai Dca Daigaku Zasshi (1973), 88(2), 314-21
279.
Qiolinergic mechanism of action of morphine and of some of its synthetic analogs Hadzovic, Jelena; Hadzovic, Safet Veterinaria (Sarajevo) (1973), 22(3), 347-51
280.
Synthesis of l,3,4,5,6,7,8,8a-octahydro-2-methyl-4a- phenylisoquinolin-6-ols. Novel fragments of the morphine molecule Finch, Neville; Blanchard, Louis; Puckett, R. T.; Werner, L. H. J. Org. Chem. (1974), 39(8), 1118-24
281.
Synthetic antigens Gross, Stanley J. Ger. Offen., 38 pp. PI DE 2324544 731129 PRAI US 72-253632 720515
282.
Morphine lethality in rats. Effects of inhibitors of brain catechol amine synthesis and methylation Davis, W. M.; Khalsa, J. H. Res. Commun. Chem. Pathol. Pharmacol. (1973), 6(3), 867-72
283.
Synthesis and pharmacological effect of morphine phosphate esters Mori, Masaaki; Oguri, Kazuta; Yoshimura, Hidetoshi; Kamata, Osamu Yakugaku Zasshi (1973), 93(10), 1302-7
284.
Late effects of perinatal morphine administration on pituitary-thyroidal and gonadal function Bakke, John L.; Lawrence, Nancy L.; Bennett, Jane Biol. Neonate (1973), 23(1-2), 59-77
285.
Thin-layer chromatography of drugs. V. Analysis of drugs which are synthetic substitutes of morphine Botev, B. Farmatsiya (Sofia) (1973), 23(3), 19-25
286.
Synthesisof B/C trans-fused morphine structures. VI. Mass spectrum, optical rotatory dispersion, and circular dichroism of B/C trans-morphine derivatives Inoue, Hirozumi; Takeda, Mikio; Kugita, Hiroshi Chem. Pharm. Bull. (1973), 21(9), 2004-13
287.
Acute effects of morphine on dopamine synthesis
148 and release and tyrosine metabolism in the rat striatum Gauchy, C ; Agid, Y.; Glowinski, J.; Cheramy, A. Eur. J. Pharmacol. (1973), 22(3), 311-19 288.
Effects of narcotic analgesic drugs on brain noradrenergic mechanisms Smith, Charles Bruce; Sheldon, M. I. Agonist Antagonist Actions Narcotic Analg. Drugs, Proc. Int. Symp. (1973), Meeting Date 1971, 164-75. Editor(s): KosterUtz, H. W. Publisher: Univ. Park Press, Baltimore, Md.
289.
Effect of chronic administration of morphine on mouse brain aminoacyl-tRNA synthetase and tRNA-amino acid binding Datta, Ranajit Kumar; Antopol, William Brain Res. (1973), 53(2), 373-86
290.
Action of levallorphan. Macromolecular synthesis and cell division Stephens, Ralph E.; Zimmerman, Arthur M. Mol. Pharmacol. (1973), 9(2), 163-71
291.
Isolation, x-ray analysis, and synthesis of a metabolite of (-)-3-hydroxy-N-allylmorphinan Blount, John P.; Mohacsi, Emo; Vane, Floie M. J. Med. Chem. (1973), 16(4), 352-5
292.
Fortral. Synthetic morphine derivative with morphine-antagonistic components Kubicki, St.; Haas, J.; Stoelzel, R. Verh. Deut. Ges. Inn. Med. (1972), 78, 1577-82
293.
Effects of inhibitors of protein synthesis in morphine tolerance and dependence Cox, Brian Martin Agonist Antagonist Actions Narcotic Analg. Drugs, Proc. Int. Symp. (1973), Meeting Date 1971, 219-31. Editor(s): Kosterlitz, H. W. PubUsher: Univ. Park Press, Baltimore, Md.
294.
Pharmacologic properties of synthetic.DELTA.9 tetrahydrocannabinol (THC) Kaymakcalan, Sulaii; Deneau, G. A. Acta Med. Turc, Suppl. (1972), No. 1, 27 pp.
295.
Structure related to morphine. Synthesis of.alpha.-2-N-heptyl-2'-hydroxy-5,9dimethyl-6,7-benzomorphan from 3,4-lutidine. II Ramachandran, Mrs. R.; Kishore, D.; Joshi, B. C. Def. Sci. J. (1972), 22(3), 201-3
296.
Brain chromatin activity of morphine-treated rats Hodgson, John R.; Lee, Cheng-Chun; Castles, Thomas R. Proc. Soc. Exp. Biol. Med. (1972), 141(3), 790-3
297.
[14C]-catechol amine synthesis in mouse brain during morphine withdrawal Rosenman, S. J.; Smith, C. B. Nature (London) (1972), 240(5377), 153-5
298.
Inhibition of development of tolerance to morphine by cycloheximide Feinberg, Michael P.; Cochin, Joseph
149 Biochem. Pharmacol. (1972), 21(22), 3082-5 299.
Inhibitory effects of chronic administration of morphine on uridine and thymidine incorporating abilities of mouse liver and brain subcellular fractions Datta, R. K.; Antopol, W. Toxicol. Appl. Pharmacol. (1972), 23(1), 75-81
300.
Incorporation of amino acids into proteins of synaptosomal membranes during morphine treatment Franklin, G. I.; Cox, B. M. J. Neurochem. (1972), 19(7), 1821-3
301.
Improvement in the technology of codeine synthesisfrom morphine Smimov, D. M.; Sigal, E. L.; Marechek, K. Ya.; Zakharov, V. P. Khim.-Farm. Zh. (1972), 6(5), 31-6
302.
Chemical synthesis and analgesic effect of morphine ethereal sulfates Mori, Masaaki; Oguri, Kazuta; Yoshimura, Hidetoshi; Shimomura, Kyoichi; Kamata, Osamu; Ueki, Showa Life Sci. (1972), ll(ll)(Pt. 1), 525-33
303.
Nucleic acid synthesis in brains from rats tolerant to morphine analgesia Castles, T. R.; Campbell, S.; Gouge, R.; Lee, C. C. J. Pharmacol. Exp. Ther. (1972), 181(3), 399-406
304.
Effect of morphine on cerebral glycogen content, glycogen synthetase, and incorporation of glucose into brain glycogen of mice Estler, C. J.; Mitznegg, P. Pharmacol. Res. Commun. (1971), 3(4), 363-7
305.
Dissociation of morphine tolerance and dependence from brain serotonin synthesis rate in mice Schechter, Paul J.; Lovenberg, Walter; Sjoerdsma, Albert Biochem. Pharmacol. (1972), 21(5), 751-3
306.
Effect of levorphanol tartrate on ribonucleic acid synthesis in normal and regenerating rat liver Becker, Frederick F.; Rossman, Toby; Reiss, Betti; Simon, Eric J. Res. Commun. Chem. Patiiol. Pharmacol. (1972), 3(1), 105-16
307.
Correlations between protein and serotonin synthesis during various activities of the central nervous system (slow and desynchronized sleep, learning and memory, sexual activity morphine tolerance, aggressiveness, and pharmacological action of sodium gamma-hydroxybutyrate) Laborit, H. Res. Commun. Chem. Patiiol. Pharmacol. (1972), 3(1), 51-81
308.
Minor alkaloids of morphine. VII. Synthesis of gnoscopine (dl-narcotine) Kerekes, Peter; Bognar, Rezso Magy. Kern. Foly. (1971), 77(12), 655-9
150 309.
Alkaloids associated with morphine. Vn. Synthesis of gnoscopines (DLnarcotine) Kerekes, P.; Bognar, R. J. Prakt. Chem. (1971), 313(5), 923-8
310.
Effect of morphine on protein synthesis in synaptosomes and mitochondria of mouse brain in vivo Kuschinsky, K. Naunyn-Schmiedebergs Arch. Pharmakol. (1971), 271(3), 294-300
311.
Amounts and turnover rates of brain proteins in morphinetolerant mice Hahn, D. L.; Goldstein, A. J. Neurochem. (1971), 18(10), 1887-93
312.
Biochemical pharmacology of tolerance to opioid analgesics Ginsburg, M. Sci. Basis Med. (1971) 305-19
313.
Morphine-associated alkaloids. 5. Synthesis and structure of narcotoline ethers Gaal, Gy.; Kerekes, P.; Gorecki, P.; Bognar, R. Pharmazie (1971), 26(7), 431-4
314.
Effect of p-chlorophenylalanine on the cardiorespiratory reflex response to morphine and serotonin in the rat Aldunate, Jorge; Prieto, Rafael Arch. Biol. Med. Exp. (1970), 7(1-2-3), 45-7
315.
Increase of brain tryptophan caused by drugs which stimulate serotonin synthesis Taghamonte, Alessandro; Tagliamonte, Paola; Perez-Cruet, Jorge; Gessa, Gian L. Nature (London), New Biol. (1971), 229(4), 125-6
316.
Correlations between protein synthesis and serotonin in various central nervous system activities. Slow and desynchronized sleep, memory training, sexual activity, morphine tolerance, aggressiveness, and sodium .gamma.hydroxybutyrate pharmacology Laborit, H. Agressologie (1971), 12(1), 9-24
317.
Structures related to morphine. Synthesisof .alpha.-2'-hydroxy-2-methyl-5-propyl-9-ethyl-6,7-benzomorphan. I Ramachandran, Reena; Joshi, Bhuwan C. Def. Sci. J. (1970), 20(4), 233-6
318.
Conversions of tosyl and mesyl derivatives of the morphine group. VIII. Synthesis and investigation of 6-deoxy-6-fluoroisocodeine Bognar, Rezso; Makleit, Sandor; Radics, Lajos Acta Chim. (Budapest) (1971), 67(1), 63-9
319.
Morphine alkaloids and related compounds. XX. Syntheses and pharmacology of some demethylated compounds related to 14-hydroxydihydro-6.beta.-thebainol4methyl ether (oxymethebanol), a new potent antitussive Seki, Isao; Takagi, Hiromu
151 Chem. Pharm. Bull. (1971), 19(1), 1-5 320.
Unchanged rate of brain serotonin synthesis during chronic morphine treatment and failure of p-chlorophenylalanine to attenuate withdrawal syndrome in mice Marshall, Ian G.; Grahame-Smith, D. G. Nature (London) (1970), 228(5277), 1206-8
321.
Tolerance to morphine-induced increases in [14C]-catechol amine synthesis in mouse brain Smith, Charles Bruce; Villarreal, Juhan E.; Bednarczyk, Janet H.; Sheldon, Murray I. Science (1970), 170(3962), 1106-8
322.
Is there a relation between protein synthesis and tolerance to analgesic drugs? Cox, Brian M.; Ginsburg, M. Sci. Basis Drug Depend., Symp. (1969), Meeting Date 1968, Volume Development of new p 77-92. Editor(s): Steinberg, Hannah. PubUsher: J. and A. Churchill Ltd., London, Engl.
323.
New neurotropic agents among synthesized compounds of the pyridine series Poddubnaya, L. V.; Olekhnovich, L. B.; Dorofeenko, G. N. Farmakol. Tsent. Khohnolitikov Drugikh Neirotropnykh Sredstv (1969), 317-24. Editor(s): Denisenko, P. P. Publisher: Leningrad. Sanit.-Gig. Med. Inst., Leningrad, USSR.
324.
Effects of morphine and pentobarbitone on acetylcholine synthesis by rat cerebral cortex Sharkawi, Mahmoud Brit. J. Pharmacol. (1970), 40(1), 86-91
325.
Synthesis of B/C transfused morphine structures. FV. Synthesis of B/C transisomorphine Inoue, Hirosumi; Takeda, Mikio; Kugita, Hiroshi Chem. Pharm. Bull. (1970), 18(8), 1569-75
326.
Synthesis of B/C trans-fused morphine structures. V. Pharmacological summary of trans-morphine derivatives and an improved synthesis of trans-codeine Kugita, Hiroshi; Takeda, Mikio; Inoue, Hirosumi J. Med. Chem. (1970), 13(5), 973-5
327.
Morphine alkaloids and its related compounds. XVUI. Syntiieses of N-substituted-morphinan dihydronormethines and 0-alkyhsoureas related to morphinan, norpethidine, or phenethylamine Seki, Isao; Takagi, Hiromu Chem. Pharm. Bull. (1970), 18(6), 1104-11
328.
Conversions of tosyl and mesyl derivatives of the morphine group. VI. Synthesis of acetylthio and mercapto derivatives Bognar, Rezso; Makleit, Sandor; Mile, Terez; Radics, Lajos Acta Chim. (Budapest) (1970), 64(3), 273-9
152 329.
Effect of morphine, nalorphine, naloxone, pentazocine, cyclazocine, and oxotremorine on the synthesis and release of acetylchohne by mouse cerebral cortex sUces in vitro Howes, John F.; Harris, Louis Selig; Dewey, William L. Arch. Int. Pharmacodyn. Ther. (1970), 184(2), 267-76
330.
Inhibition of morphine tolerance and physical dependence development and brain serotonin synfiiesis by cycloheximide Loh, Horace H.; Shen, Fu-Hsiung; Way, E. Leong Biochem. Pharmacol. (1969), 18(12), 2711-21
331.
Inhibition of the development of tolerance to morphine in rats by drugs which inhibit ribonucleic acid and protein synthesis Cox, Brian Martyn; Osman, O. H. Brit. J. Pharmacol. (1970), 38(1), 157-70
332.
Lack of a direct effect of morphine on the synthesis of pineal carbon-14 labeled indoles in organ culture Shein, Harvey M.; Larin, Frances; Wurtman, Richard J. Life Sci. (1970), 9(1), 29-33
333.
Conversion of tosyl and mesyl derivatives of the morphine group. V. of isocodeine and dihydroisocodeine Makleit, Sandor; Bognar, Rezso Magy. Kem. Foly. (1969), 75(5), 235
334.
Morphine alkaloids and its related compounds. XVI. Synthesis of 14hydroxyallopseudocodeine 8-ethers and its derivatives Seki, Isao; Takagi, Hiromu Chem. Pharm. Bull. (1969), 17(8), 1555-9
335.
Morphine derivatives, n. Stereochemistry of the by-products in the synthesis of 3-methoxy-N- methyhnorphinan Kawasaki, Kazuhiko; Matsumura, Hiromu Chem. Pharm. Bull. (1969), 17(6), 1158-74
336.
Metabohsm of drugs. LX. Synthesis of codeine and morphine glucuronides Yoshimura, Hidetoshi; Oguri, Kazuta; Tsukamoto, Hisao Chem. Pharm. Bull. (Tokyo) (1968), 16(11), 2114-19
337.
Synthesisof B/C trans-fused morphine structures. III. Synthesis of B/C trans-morphine Kugita, Hiroshi; Takeda, Mikio; Inoue, Hirosumi Tetrahedron (1969), 25(9), 1851-62
338.
Synthesis of B/C trans-fused morphine structures. II. Hydroboration of isoneopine, neopine, neopinone and thebaine Takeda, Mikio; Inoue, Hirosumi; Kugita, Hiroshi Tetrahedron (1969), 25(9), 1939-49
339.
Comparative action of morphine and a synthetic substance on behavior and pain in the baboon. Tolerance study Baetz, Pierre; Bourgoin, P.; Giono, Paulette; Giono, H.
Synthesis
153 Bull. Mem. Fac. Mixte Med. Pharm. Dakar (1967), 15, 270-4 340.
Conversions of tosyl and mesyl derivatives of the morphine series. Synthesis of acetylthio and mercapto derivatives Bognar, Rezso; Makleit, Sandor; Mile, Terez Acta Chim. (Budapest) (1969), 59(1), 161-4
341.
Synthesis of new morphine derivatives. 11. The preparation 0-benzoylmorphines with analgesic action and an O-benzylmorphine with a morphine-potentiating effect Selmeci, Gyorgy; Szlavik, Laszlo; Kaskoto, Zoltan; Lepenyene, Jilek Maria; Tothne, Aranyos Iren Khim.-Farm. Zh. (1968), 2(7), 19-23
342.
Morphine tolerance, physical dependence, and synthesis of brain 5-hydroxytryptamine Way, E. Leong; Loh, Horace H.; Shen, Fu-Hsiung Science (1968), 162(3859), 1290-2
343.
Elimination of the 4-hydroxyl group of the alkaloids related to morphine. XI. Syntiiesis of (-)-14-hydroxy-3metiioxy-N-methylmorphinan derivatives Sawa, Y. K.; Tada, H. Shionogi Res. Lab., Shionogi and Co., Ltd., Osaka, Japan Tetrahedron (1968), 24(20), 6185-96
344.
Synthesis of new morphine derivatives. I. Morphine derivatives substituted at the nitrogen and in position 3 Selmeci, G.; Szlavik, L.; Kaskoto, Z.; Jilek, L. M.; Maczko, I. Khim.-Farm. Zh. (1968), 2(6), 12-17
345.
Effect of morphine on acetylcholine release from rabbit brain tissue Sugano, Tsukasa; Takeno, Kazu; Yanagiya, Iwao Nippon Yakurigaku Zasshi (1967), 63(6), 494-500
346.
The synthesis of codeine and morphine D-glucuronides Yoshimura, H.; Oguri, K.; Tsukamoto, H. Tetrahedron Lett. (1968), (4), 483-6
347.
Synthesis and biological properties of l-dimethyl-amino-3-methyl-3-(3hydroxyphenyl)butane, a potential analgetic Pecherer, Benjamin; Sunbury, R. C ; Randall, Lowell O.; Brossi, Arnold J. Med. Chem. (1968), 11(2), 340-2
348.
Effect of morphine administration on the incorporation of leucine-14C into protein in cell-free systems from rat brain and hver Clouet, Doris H.; Ratner, Milton J. Neurochem. (1968), 15(1), 17-23
349.
Thin-layer-chromatographic distribution of opium alkaloids and some partially synthetic analogs Paris, R. R.; Sarsunova, Magda Pharmazie (1967), 22(9), 483-4
154 350.
Elimination of the 4-hydroxyl group of the alkaloids related to morphine. IX. Synthesis of 3-methoxy-N- methylisomorphin and derivatives Sawa, Yoshiro K.; Horiuchi, Masahiko; Tanaka, Katsura Tetrahedron (1968), 24(1), 255-60
351.
Alternate Route in the Synthesis of Morphine Morrison, Glenn Curtis; Waite, Ronald, O.; Shavel, John, Jr. Tetrahedron Lett. (1967), (41), 4055-6
352.
Synthesis of B/C-trans-morphine Kugita, Hiroshi; Takeda, Mikio; Inoue, Hirosumi Tetrahedron Lett. (1967), (14), 1277-81
353.
Biochemistry of poppy alkaloid synthesis. Phenolase complex in Papaver somniferum Kovacs, Peter; Jindra, Antonin; Psenak, Mikulas Abh. Dtsch. Akad. Wiss. BerHn, Kl. Chem., Geol. Biol. (1966), (3), 335-40
354.
Preparation of formaldehyde by photochemical condensation of carbon monoxide and tritiated hydrogen, and the synthesis of a tritiated morphine derivative Lane, A. C ; McCoubrey, Arthur; Peaker, R. J. Labelled Compd. (1966), 2(3), 284-8
355.
Analysis of medicine mixtures. VI. Microchemical detection of some alkaloids and synthetics with potassium-iron or potassium-copper-chloroiodide reagents Reisch, Johannes; Tittel, G. L.; Perlick, J. Dtsch. Apoth.-Ztg. (1965), 105, 575-6 From: CZ 1966, (22), Abstr. No. 1652
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
155
New Developments in the Synthesis of Polyketides and of Chiral Methyl Groups
Johann Mulzer Institut ftir Organische Chemie der Freien Universitat TakustraBe 3, D-14195, Berlin, Germany
Abstract: This review deals with recent advances in the synthesis of polypropionate structures. It focuses on the total synthesis of natural products (citreoviral, ACRL toxin IIIB) as well as on new synthetic methodology (chiral methyl branching, base induced 1,3-Hshift and chiral methyl groups).
Chiral Methyl Branching in Carbon Chains In the biosynthesis of polyketides the problem of chiral methyl branching is solved via enantiomer selective reduction of prostereogenic carbonyl groups (Scheme 1). The biosynthesis starts with the Claisen type condensation of the activated propionate 1 and methyl malonyl CoA 2 to give under elimination of carbon dioxide ^-ketoester 3 which has already a chiral methyl branching in 2-position. However, this center normally tends to racemization and is not configurationally stable. It may be safely assumed that it is the NADH mediated reduction of 3 to 4, which eventually defines the configurations of both the hydroxy and the methyl bearing stereogenic centers, the first one by direct chiral induction, the second one by kinetic resolution of a mobile equilibrium of the enantiomers of keto ester 3. The same process is then repeated for each new propionate subunit in the growing polyketide chain, e.g. from 4 via 5 to 6.
156
In this way hydroxymethyl (HM) or hydroxy-methyl-hydroxy-methyl (HMHM)subunits are generated in a stereodefined way (1).
CH3 CH3-CH2-C~SCoA
0
e.DoC
--C02
SCoA
CH,
OH
Reductase
O
II * // C113"" C112'"'C>"~ C H "" C
SCoA
+2
CH3-CH2-CH-CH-C' CH3 SCoA
-CO2
4 (stereodefined) OH 0 ^ 0 o ^ * I * II *2 /y Reductase GH3-CH2-CH-CH-C-CH-C ^ *
i
?
CH3 5
i
CH3
\
OH , OH 2 0 1 *4 1 *2 ^^ CH3-CH2-CH-CH~CH-CH-C *
SCoA
«
^
i
*
?
i
\
CH3 CH3 SCoA 6 (stereodefined)
Scheme 1: Polypropionate Biosynthesis
How far can this bioprocess be transferred to by in-vitro synthesis? An answer to this question has been given by R. W. Hoffmann et al. (2), who reduced ketoester 7, configurationally labile as discussed, with baker's yeast, hi fact the carbonyl reduction proceeded (5)-selectively, however, the kinetic resolution with respect to the enantiomers of 7 only led to an enantiomeric excess of 72% with respect to C-2 in 8/9.
Me-
OEt Me 7
baker's yeast (59%)
OH
0
Me-^ V ^
OH OEt
Me^ > ^
Me 8
0
Me 6.4: 1
9
"OEt
157
Apparently this method works unsatisfactorily in vitro so that purely chemical approaches appear advisable to generate HM- and HMHM-units in an enantiopure manner. As illustrated in Scheme 2, two possibilities (among others) may be envisaged to place a chiral methyl branching upon a pre-existing carbon chain: either by opening of a configurationally defined epoxide with a methyl cuprate reagent (equ. 1) or by alkylating a chirally substituted propionate type enolate (equ. 2). Alternatively, a suitably functionalized carbon chain may be methylated via the corresponding enolate.
Epoxide Opening:
f j y\yy^
1 j
0 CH3
•
M® i l j y"^^^^
l \
equ. 1
OH Enolate Alkylation of a Chiral Propionic Acid Derivative:
0
%
o Me
Scheme 2: How to Generate Chiral Methyl Branching in a Carbon Chain to Form an 1,2Hydroxy-Methyl (1,2-HM) Subunit
We describe some new methodology for both realizing equ. 1 and 2. Regarding the epoxide opening we reasoned that the v/c diol present in inexpensive carbohydrates such as Z)-mannitol would be highly suitable for the construction of HM- and HMHM-subunits. Specifically (Scheme 3) the D-mannitol diacetonide (10) gives the c/5'-epoxide 12 on treatment of tosylate 11 with base. With methyl cuprate regioselective ring opening occurs in favor of the HM-intermediate 13, which can easily be separated from the diastereomer 14. This opens a simple access to 5'j^/7-HM-structures.
158
Q K2CO3
Me2CuLi
6: 1
Scheme 3: syn-HM-Subunits from Epoxides
The a^2//-diastereomers 19 and 23 are available via the inversion of the 3- or 4-OHfunctions of 10 (Scheme 4). This is achieved via monoprotection of 10 to benzoate 15, oxidation to ketone 16, stereoselective reduction to 17 and formation of epoxide 18, which is C2-symmetrical and thus can only give the anti-UM derivative 19 on cuprate addition. Alternatively 17 is converted into 20 which gives the C2-symmetrical epoxide 22 via 21. Again only one HM-derivative, namely 23, can be formed on reaction with the dimethyl cuprate (3).
159
10
\. 2. OH-
o"
y^
^^^
^^
Me2CuLi 0
18 (C2-symmetricaI)
TBDPS-CI 17
22 (C2-symmetrJcal) Scheme 4: aA7//-HM-Subunits from Epoxides
HMHM-Subunits may be generated via bis-epoxides as demonstrated in Scheme 5. D-Mannitol is converted into the di-tosylate 24 which is cycHzed to 25 with base.
160
OTs OH
3 steps
D-Mannitol
f
•
:
HO""'^
TBDPS-CI
"^
p.^_^ K2CO3
^^^
.OTBDPS
TBDPSO'
"6 26 Me
TBDPso
i^r
Me OTBDPS
TBDPSO
3^^5^
OH 27 Scheme 5: HMHM-Subunits from Epoxides
Bis-protection of the terminal OH-functions of 25 furnished 26, to which Lipshutz' cuprate is added. The bulky silyl group directs the nucleophilic attack to the 3position to form 27, whereas smaller protective groups such as MOM lead to mixtures of 2,3-regioisomers. Monoepoxide 27 is not isolated but adds excess cuprate insitu to furnish 28 regioselectively. Once introduced the 3-methyl branching in 27 directs the cuprate attack towards the less hindered position at C-5.
OTs OH
TsO OH
K2CO3
OTs ^
'
29
2 Me2Cu(CN)Li2
Scheme 6: HMHM-Subunits from Epoxides
—
0
161
An alternative way to HMHM-structures such as 32 is shown in Scheme 6. The C2symmetrical ditosylate 24 is converted into the mono-acetonide 29 which forms the epoxide 30 on treatment with base. The cuprate attack on the epoxide generates a second epoxide 31 via tosylate ehmination. With excess cuprate 31 is opened to give 32 directly (4). Alkylation of Chiral Enolates As demonstrated in Scheme 7 the alkylation of chiral propionamide enolates has become a standard operation, since D. A. Evans introduced his unusually efficient oxazolidinone auxiliaries ®.
(1)
D. A. Evans O
SI
J
iPr
0
0
H
re
3 ) W. Oppolzer
1985
O
•••• Me"^ . . MPh
(4)' J. Rebekjr.
1984
T. Katsuki
0
^^
CH2OMOM
CH2OMOM
ds^ 100: 1
ds^ 90: 10
cfs « 95 : 5
O2
(2)
1981
1990
5)
K. Kimura
0
ds ^ 96 : 4
ds > 99 : 1 (one example!) ds > 500 : 1
Scheme 7: Alkylation of Chiral Propionamide Enolates
1992
162
This principle has found many variations and improvements, some of which are shown as (2) - (D in Scheme 7 (5).
0
NH2
Ph,^.OH
.L 1NHo
NaBH4
Cr U
33
1.
34
H ^ fBu
OTMS 2. HoC=C NTMS H
p
3: 1 Chromatographic Separation
35
3. E t - C ' CI (50%)
0
NHBoG
H
J^
MgBr
1. H + O
^^^
X
2. H ^ ^ f B u OTMS 3. H2C=C NTMS H 4.
Et-C CI (50%)
0 N'^fBu
r "0
'0 35a
3: 1 Chromatographic Separation
Scheme 8: Chiral Dihydroxooxazines as Amide Auxiliaries
36a
163
In all cases except (2) the nitrogen is part of an imide system which forms a chelate complex after deprotonation to the enolate. We used a different approach by using the A^,(9-acetals 35/36. This was first performed in racemic form to test the diastereoselectivity of the enolate alkylation (Scheme 8). The diastereomers 35 and 36 are readily separable by chromatography. Analogously, amides 35a and 36a are prepared from 0-jV-Boc-amino benzaldehyde. Derivative 36a is crystalline and was submitted to an X-ray crystal structure analysis (Fig. 1).
\j.^'
V\tX
Figure 1:
^^^
Crystal structure of 36a
The amide function (Nl-Cl 1-0-11) is planar and exerts an allylic 1,3-strain effect on the adjacent stereogenic center (C-2), which forces the bulky /-butyl substituent into a pseudo-axial position with respect to the boat conformation of the acetal ring. The isopropenyl group at C-4 adopts another pseudo-axial position. Despite its axial rearrangement, the shielding effect of the /-butyl group on the enolate carbon C12 is low. Indeed, deprotonation of 35 and subsequent allylation proceeded with a stereoselectivity of 4:1 in 89% yield. If the phenyl and /-butyl group are on opposite ring faces as in 36, the deprotonation is severely hindered . Both yield and stereoselectivity of the allylation are low.
164
35
INaHDMS
(89%)
36
1. IDA, DMPU^
(30%)
We reasoned that the /-butyl group is still too small for an efficient chiral induction; therefore, the optically pure 0-TBDPS lactaldehyde was chosen for the formation of the A^,0-acetal. But only amides 39 and 41 are now non-racemic; 39 exhibits a satisfactory chiral induction on allylation, because the enolate carbon is shielded by the adjacent axial bulky substituent. In 41, both sidechains at C-2 and C-4 are equatorial and the stereocontrol drops significantly.The use of a chiral aldehyde for acetal formation even allows the use of the achiral 0-aminobenzylalcohol (43) as a template. Acetals 44 and 45 are formed and separated; due to the allylic 1,3-strain of the amide moiety both derivatives have axial sidechains (as detectable in the crystal structure of alkylation product 46d) (Fig. 2) and the chiral induction is similarly high in both cases. The chiral auxiliary is removed with lithium aluminium hydride without any racemization of the newly created stereocenter (6).
165 Ph
Ph 1. KHDMS
I
^0
Br
2.
N^2
OTBDPS
"^^^^'""^O
39
^^^^^^
40 (cfs> 95:5)
E>h
Ph 1. KHDMS ,Br
OTBDPS
^
^
V
'0
OTBDPS
42 (ds = 80:20)
41
?
O ^ ^
.>o Gi
a
51^
Figure 2: Crystal Structure of 46d
IDi
166
, „V '"• OTBDPS
CC°"
OTMS
NHo
2. HoC=C
43 3.
Et-C
NHTMS
CI
(62%)
x^
N
OTBDPS
'0
0
8: 1 Chromatographic Separation
44
45 R-X
R-X
46,47
x^
R
R,, ^ x ^ a b c d
ally! Et nPr -v-/^^^
OTBDPS
0
92:8 - 94:6
OTBDPS ds
ds
46
OTBDPS
47
88:12-92:8
Scheme 9: Achiral Dihydrooxazine Template
1,3-H-Shift. By serendipity we found a novel base induced stereocontrolled sigmatropic 1,3-Hmigration. Benzylation of the readily available alcohol 48 to benzyl ether 49 occurs under standard conditions at 25°C. By mere accident the student (G. Funk) raised the temperature of the mixture to 80 °C and left the reaction at that temperature for 14 h.
167
After the usual workup the rearrangement product 50 was isolated in quantitative yield as the pure (>98% ) ^-olefin (7).
OBn
BnCI / NaH DMF / 25°C
NaH DMF 80°C BnCI / NaH DMF / 80°C
On closer investigation J. Bilow found that the rearrangement requires sodium or potassium hydride as the base and DMF or tetramethylurea as the solvent (8). DMSO is also suitable, but inferior with respect to the yield. In the absence of benzyl chloride the rearrangement does not occur. However, benzylether 49 can be prepared first and then submitted to the rearrangement by treatment with NaH in DMF. Analogous isomerizations could not be found in the literature; it was only by personal communication that we came across a similar example from Prof. W. Kreiser's group (University of Dortmund, Germany), namely the rearrangement of steroid 51 into 52, although a different base and solvent were applied. The scope of "our" rearrangement is reasonably large. Me^
Ma,
MeT ^V^
MeT
^"V
/ v ^
t^A^
Ha^
Me
Li / Ethylendiamine
MeT
^^/\
6min/120°C HO^^^^
H 51
f52
168
Scheme 10 shows some examples; the yields are uniformly higher than 85%. The conversion of 57 into 58 is noteworthy with respect to the presence of the n-butyl group. Allylic alcohol 61 does undergo the rearrangement, but the stereocontrol with respect to the ^-configuration is much lower than in the examples above (> 98%).
OBn
OH BnCI, NaH DMF, 80°C
BnCI, NaH OBn
DMSO, 60°C
BnO OBn
55
56
BnCI, NaH OH
DMF, 80°C 57 BnCI. NaH
"
DMF, 80°C
? 7 'Y'®
BnO^A/>^Me 60
BnCI, NaH DMF, 80°C
OH 61
OBn
62 (E)/(Z) = 3:1 -8:1
Scheme 10: Further Examples of the Double Bond Migration
Detailed mechanistic investigations (Scheme 11) revealed that the formation of the Eolefin is the resuh of a kinetically and not of a thermodynamically controlled reaction.
169
This was shown by preparing the Z-olefm 63 independently and submitting it to the conditions. No isomerization to 50 was observed. We suspected that the rearrangement proceeded via an allylic anion as an intermediate which could possibly be trapped with deuterium. However, to our surprise, no H-D-incorporation was observed, when the reaction mixture was quenched with D2O . Similarly, no deuterium was exchanged on performing the reaction in dy-DMF. This means that the concentration of the anionic species, if present at all, must be very low throughout the reaction. We next turned to the question whether the reaction proceeds inter- or intramolecularly. To this purpose the monodeuteriated alcohols 68 and 69 were prepared as shown in Scheme 11 and submitted to the rearrangement.
NaH / DMF 7^
50
•
80°C
LLiAIH.
H.C-
^
OH
64
H3C
H
D
2. D2O
PBh
H3C
^OH 65
66
CrCl2 NaH/DMF BnCI OBn
Scheme 11: Mechanistic Investigations
OBn
170
The deuterium was quantitatively transferred into the expected position to give 70 and 71 as the reaction products. No H-D-exchange of the substrate with the reaction medium was observed. As a final confirmation of the strictly intramolecular process a cross-over experiment of non- and dideuteriated material was performed (Scheme 12). MS-analysis clearly demonstrated that only do and d2-product (i.e. 75 and 76) was formedfi-om49 and 73, whereas in case of an intermolecular rearrangement also dj-material 77 and 78 should have been generated.
64
2. D2O
^
OBn D
D
1.LiAIH4
H,c\^
3. PBr3
Br
1
D
OBn D
+ 67 ^ CrCl2
72 OBn H
OBn D NaH
73
+
0'
^
6
H Me
DMF
0^
y^
-Ar-6
" T ^ "C:H2D +
Me
O'' > ^
—Vo
"T^
"CH3
Me
49
analogous result with 74
78 (MS-Analysis)
Scheme 12: Final Confirmation of Intramolecular Rearrangement: Negative Cross-Over Experiment with Dideuteriated and Undeuteriated Material
It was also interesting to know whether the intramolecular 1,3-H-shift follows a suprafacial (79) or an antarafacial (80). To distinguish between these pathways both deuterium and tritium had to be introduced into the C-6-position of olefin 49 in a stereodefined manner so that after the rearrangement a chiral methyl group could be
171
created. After oxidative removal from the rest of the molecule the chiral methyl group can be analyzed in form of the chiral acetic acids (R) or (6)-81 according to Arigoni's enzymatic method (9).
OBn
OBn H
.JV-6
(H)
OBn
B
or
Me H
Me
49
H
80 antarafacial
H
CO2H
H
(S)-81
CO2H
(R)-81
Thus, 49 was converted into dibromide 82 (Scheme 13) and submitted to a FritschButtenberg-Wiechell rearrangement to give the acetylide which was quenched with T2O (activity lOOmCi/ml) to give 83. Lindlar deuteriation furnished olefin 82 which was then rearranged to 85 under standard conditions. Lemieuix oxidation proceeded without racemization to furnish (5)-81 with an ee of 45% (10). Analogously, isomer 53 gave (i^)-81with 44% ee. This means that the suprafacial and the antarafacial 1,3shift compete with each other in a ratio of 73:27. Although the enantiomeric access is comparatively low it is sufficient for most labelling studies and; in view of the simple overall access to intermediate 85, may find application. Moreover, by functional group manipulation of 85 the chiral methyl group can be directly incorporated into polyketide structures and related natural products.
172 OBn 49
1.2BuLi
I.O3
^ 2. CBr4 / Zn / PPhg
OBn D KMn04
HO2C..D
Nal04 (S)-81
by enzymatic analysi (D. Arigoni)
Scheme 13: Synthesis of Chiral Acettic Acid (S)-81
For example, the method potentially opens an access to compounds with a doubly chiral isopropyl unit (Scheme 14). Li the pro-iS-selective enzymatic hydroxylation of isobutyric acid (88) to (5)-^-hydroxyisobutyric acid (89) the stereochemistry of the hydroxylation at C-3 is not known. It could be studied by preparing 88 in a doubly chiral form via stereocontrolled anti-S^l' reaction of dimethyl cuprate with the tosylate 90 to give 91 which is then degraded by Lemieux- and then Baeyer-Villiger oxidation to 88. In a final overview (Scheme 15) "our" 1,3-H-shift is compared with the one described by Cram some thirty years ago (11). It may be concluded that there are certain similarities, however Cram used a protic system and a C-H-acidic hydrocarbon as a substrate and observed a reversible rearrangment. Quite interestingly, he formulated a very similar transition state (94) which was termed a "guided tour mechanism".
173 . CH3 R-CH
. CHDT R-CH
CDs
CDs
86 (simply chiral isopropyl group)
87 (doubly chiral isopropyl gruop)
Application CHs
pseudomonas
HsC ' ^ C O s H
putida
CHs
88 "
(S)-89
Pros
Hydroxylation under Retention or Inversion at C-3 ?
Possible Synthesis: OTs D
Me D Me2CuLi
MeD
^
HO2C
SN2'
anti
Me*
double chiral 88
Scheme 14: Application to the Synthesis of Doubly Chiral Isopropyl Units
Me
KOt-Bu HOt-Bu
PK^
Me
Ph" 93
92 Me ^ / 0 ^ - B u .®
via
Characteristics of Cram's System: 1) Reversibility 2) Ar necessary 3) 6 - 56 % intramolecularly
94
4) Racemisation at C^ 5) Stereochemistry at C^ not tested
Scheme 15: Anionic Olefinisomerisation by Cram (1964)
174
Synthetic Applications of the Key Intermediate 50 The rearranged olefin 50 may be used in a variety of synthetic appHcations. For instance it can be converted into the novel di-bis-tetrahydroftiran-acetal 95 in a one-pot operation using trimethylsilyl iodide in dichloromethane at 22 °C (Scheme 16) (12). The mechanism involves the formation of an oxonium intermediate 96 which undergoes a Prins cyclization to form the cation 97. Subsequent pinacol rearrangement generates 98 which cyclizes to 99. This acetal dimerizes under elimination of trimethylsilylbenzyl ether and benzyliodide. The structure of 95 has been elucidated by X-ray analysis (Fig. 3) which nicely shows the C2-symmetry of the dimeric structure.
q
^
Figure 3:
Ci
013
Crystal Structure of 95
Another apphcation of 50 is the synthesis of ^,y-unsaturated amino acids such as 102 (Scheme 17). To this end, 50 was debenzylated with sodium in ammonia and then submitted to a Mitsunobu reaction. Clean SN2'-reaction with a«^/-stereochemistry occurred to furnish phthalimide 100 which was converted into the acid 102 by standard modifications (13).
175 OBn TMS-i, CH2CI2 RT, 5 min
TMS-I TMSOBn/-Bn
OBn
OBn
TMSO 0 99
tTMS.;; k 0
TMS-I ®
^-/OBn V-^O
,©. 98
Scheme 16:Tandem-Rearrangement-Dinnerization of 50
NHPht 1. Na/NH3
I.HgO^
0 ^
50 2. Phthaiimid, Azoester, PPha
2. Pb (0Ac)4 3. CrOa
100 NHPht
NHo
NoH 2^4 H02C
HO2C 101
102 (ca. 50% overall yield
Scheme 17: Synthesis of a,^-unsaturated y-Amino-acids
176
Another application of intermediate 50 lies in the synthesis of citrovireal 103 which is a metabolite of citreoviridine (104). This is an interesting polyene-pyrone toxin which has been isolated from penicillium citreoviride cultures. Citreoviridine has been shown to cause the Beri Beri disease which is acquired from eating infected rice. Li effect the toxin 104 acts as an inhibitor of the enzyme ATPase. So far, several syntheses have been reported for optically active 103 which has thus served as a goal for developing new synthetic methodology (14).
HQ
OH
Our retrosynthetic analysis is shown in Scheme 18. Retro-Wittig reaction leads to aldehyde 105 which is generated from alcohol 106 by Swem oxidation. This tetrahydrofuran system might be generated by ring closure of epoxy alcohol 107 although this would involve an SN2 type attack of the hydroxyl function at the more hindered position of the epoxide. The diol unit in 107 was to be created by osmylation of an allylic alcohol as represented by precursor 108 (8). However, the stereochemistry of this osmylation would be opposite to Kishi's model which predicts an anti attack of the osmium tetroxide with respect to the 1-ORfunction as shown by the conversion of 109 into 110. To circumvent this problem it was necessaiy to introduce the 1-OR group first in the wrong configuration in order to
177
exert the desired a/7//-controlling stereodirection on the osmylation to form 112. Subsequently the configuration at C-1 has to be inverted to get the overall correct arrangement at C-1,2 and 3 in intermediate 113.
HQ
OH
HQ
OH
RQ
OH
O H C ^ Q - ^ 105 \
OH =>
OR
OH 107
R^O OR 108
Scheme 18: Retrosynthesis of 103
The actual synthesis (Scheme 19) starts with 50 which is converted into the labile aldehyde 115 \ia diol 114. Aldehyde 115 adds methylmagnesium iodide with high (>95:5) chelate Cram selectivity to form 116, which is submitted to osmylation. Not surprisingly the stereocontrol in the sense of Kishi's model is high, as now two hydroxyl groups cooperate in the same direction. After ketalization intermediate 117 is obtained which after Swem oxidation and Wittig methylenation furnishes olefin 118. Debenzylation with sodium leads to allylic alcohol 119 which is epoxidized with high stereocontrol to form 120. Cyclization with acid generates 121 which fails to undergo selective oxidation at the primary or secondary position, hstead the keto aldehyde 122 is formed which turned out to be a dead end in the synthetic sequence.
178
TFA
Pb(0Ac)4
MeOH, RT (95%)
CH2CI2, 0°C
H
51
OBn 115
1. 1%0s04, NMO, MeMgl
OH -,
CH3CN, 60«C — •
1^
Et20, RT (72%)
OBn
2. DMP, CSA, CH2CI2, RT (85%)
116 (additional OH!)
fBuOOH, Na/NH3
1. Swern-Oxid.
^^ 2. PhaPCHs^Br", NaH,DMSO, RT
PhH, reflux
(95%)
(84%)
HO
TFA
^0
^
120 (> 98: 2)
^.
THF, -30°C
(41%)
MeOH, 50°C (78%)
\/0(acac)2
0^
OH
OH
HO
f^ 0 121
Oxid.
O H C - ^ o " ^ 122
Scheme 19: Synthesis of CItroviral from 51 (Part 1)
Therefore, despite lower stereoselectivity (3:1) in the epoxidation step the benzyl ether 118 (Scheme 20) was converted into 123 and then converted into tetrahydrofuran 124. After Swern oxidation a mixture of the aldehydes is generated; the isomer with the correct stereochemistry at C-2 cyclizes to the hemiacetal 125 whereas the second C-2 epimer did not cyclize and was thus easily removed by chromatography. By Wittig reaction 125 was transformed into 126 which was smoothly debenzydated under Hanesssian's conditions (15) to give alcohol 127. Inversion of configuration at C-2 was achieved by an oxidation reduction sequence with complete stereocontrol.
179
DEAL reduction of the ester to the allylic alcohol and oxidation to the aldehyde delivered citreoviral 103 eventually.
118
mCPBA Na2C03
TFA
CH2CI2, RT (78%)
MeOH. RT (78%)
PhMe, reflux (77%)
HO 0
BnQ
PH 0
SwernOxid.
EtOsC
PhSSiMeg Znl2, BU4NI
OH
0
»•
(CH2CI)2, 60°C (80%)
126 HO
Et02C
PH
124
Ph3PC(Me)C02Et
125
BnO
ISwern-Ox. OH2. ZnBH4, THF,-50°C ' Inversion 3. DIBAH
OHC
127 Scheme 20: Synthesis of Citroviral from 51 (Part 2)
First Total Synthesis of ACRL Toxin III B (128) (15) ACRL toxins form a family of metabolites of the microorganism altemaria citri rough lemon which is reponsible for the brown spot disease of citrus fruits. All these toxins are polyene pyrone polyketides in different oxidation levels. When we started the project only one synthesis of an ACRL toxin was known, namely that one of ACRL toxin I by Lichtenthaler et al. (16). Later two additional syntheses of 128 were reported (17). Our retrosynthetic disconnection of 128 is shown in Scheme 2L It results in the formation of three fragments 129-131. The first one contains the trisubstituted olefin unit which is accessible by the above-mentioned base induced 1,3-H-
180
shift. Fragment 130 can be preparedfi-omacid 89 with different patterns of protective groups and 131 is commercially available.
4" OMe ACRL Toxin III B (128)
OPG
0
OPG'
X 4 >r^2
129
130
131
Scheme 21: Retrosynthesis of ACRL Toxin II B
OTs
o-r^H ^
0
1. Crotylation
I.HgO^
2. TsCI
2. OMe"
132 1. Me2CuLi 2. TrCI 134
Scheme 22: General Synthesis of Triad Fragments
QH TrO^^^V^^'^V^^ Me
Me
135 (- 50% Yield overall)
181
So we concentrated on the synthesis of fragment 129. The two stereogenic centers at C-7 and C-8 were established from (i^)-2,3-isopropylidene glyceraldehyde 132 as shown in Scheme 22 via a sequence already employed in the total synthesis of erythronolide B (18). Stereotriad 135 is available in multigram quantities on this route via 133 and 134 (Scheme 22). After protection of the secondary OH as a pmethoxybenzyl (PMB) ether the base induced 1,3-rearrangement was achieved under standard conditions to furnish the desired olefin 137 (Scheme 23).
135
OPMB
PMB-CI
NaH / DMF
OPMB
•
OTr 136
80°C (85%)
OTr 137
Scheme 23:1,3-H-Shift of 136 to 137
An alternative route (Scheme 24) involved hydromagnesation of 2-butyne. Addition of the Grignard derivative 139 to aldehyde 140 resulted in the formation of a 1:1mixture of 141 and 142, which was oxidized to the ketone 143 under Swem conditions. With superhydride Felkin Anh controlled reduction occurred which led to alcohol 141 under high stereocontrol. After 0-protection compound 137 was formed, indistinguishable from the product obtained via the first route.
Superhydride
141 Felkin-Anh-Product
22: 1
142 af?^/-Felkin-Anh-Product
182 0
138
"'^^^^^^^^T^^^^Y^"^OTr 141
(140) 139 (E) - selective OH
OH +
f^^OTr
-^^^
1:1
+ H^^^OTr
MgCI
/ BuMgCI, EfeO -—— ^> 1,5mol%Cp2TiCl2
>wern-
C
0
-
OTr
Oxid 142
143
Scheme 24: Alternative Synthesis of 137
Intermediate 143 having secured, the synthesis was carried on by deprotection to form the primary alcohol 144 (Scheme 25) which was oxidized to the aldehyde 145 and converted into the envisaged alkyne (corresponding to 139) via a Corey Fuchs chain elongation via dibromide 146. Deprotonation with /?-butyllithium and addition of aldehyde 148 generated alcohol 149 as a 2:l-diastereomeric mixture. Again the stereochemistry at the newly created center was corrected by an oxidation reduction sequence via ketone 151. This time the chiral reduction had to be performed with using Corey's oxazaborolidine catalysts (19). In this way both the (3i?)- and (35)-diastereomer of alcohol were available. LAH-reduction of (35)-149 led to the £-alkene 150 which was eventually oxidized to aldehyde 154 after protection-deprotection via 152 and 153. Addition of the potassium salt of pyrone 131 gave 155 as a 4:l-epimeric mixture. Removal of the PMB protective group led to selective destruction of the minor diastereomer, so that a 95:5mixture in favor of the desired stereoisomer 156 was obtained (Scheme 26).
183
141
PMBCI
137
•
OPMB
ZnBr2
SwernOxid.
OPMB
NaH, DMF (95%)
CBr4, PPha, ^ Zn, CH2CI2 (90%)
THF, -80°C (91%)
146
. H'^^v-'^OTHP
A?BuLi
^
OTHP
(148) ,
THF, -80°C (65%)
OPMB
147
LAM, Et20
(95%) i49 OH
(3S;3R = 2:1) OPMB
SwernOTHP
OTHP
Oxid.
150 OPMB
chiral
OH
•
reduction according to Corey (20)
OTHP (3R)-150 with (S)-Oxa2aborolidine
""
OTHP (3S)-150 with (R)-Oxazaborolidine
Scheme 25: Synthesis of ACRL Toxin III B
Desilylation furnished the target compound which was crystalline in contrast to the natural product and was characterized by a X-ray crystal structure analysis (Fig. 4). The polyketide backbone forms an all anti zig-zag chain. Tlie 4-,8-, and 10-hydroxyl functions all point upward whereas the 3- and 7-methyl groups point downward and are on the opposite face of the chain, which in this manner shows a hydrophylic top face and a hydrophobic bottom face . Possibly this property is essential to the biological activity of the toxin. Additional interesting structural features are the allylic
184
1,3-strain relationships between C-17/C-5 C-6/C-8 and C-7/C-9, which help to rigidify the observed zig-zag-conformation.
OPMB
TBDPSCl Imidazol »
OH OTHP
OPMB
OTBDPS
•-
OTBDPS OTHP
CH2CI2 (84%)
(3S)-150
Eton
OPMB
152
^w®''"Oxid.^
OPMB
OTBDPS
'^OH
PPTS (95%)
(85%) 153 O' ^
154
(131) PMBO
^ OMe
OH
0
DDQ OMe
KHMDS, THF -100°C; (55%)
HO
TBDPSO
155
TBDPSO
OH
TBAF
0 OMe
156
»-
CH2CI2 (59%)
128
AcOH.THF (95%)
(95:5)
Scheme 26: Synthesis of ACRL Toxin IN B: the End Game
A final comment has to be made on the reduction of ketone 150 with Corey's catalyst 157 (19). The mechanism (20) involves the formation of transition state complexes such as 158 in which, by interaction with the rest of the molecule the small substituent (Rs) of the ketone points upward and the large substuent (RL) downward. Remarkable, for a,^-unsaturated ketones the vinyl group is the large one and this is indeed confirmed by our case. The reduction is reagent controlled, but the substrate in-
185
fluence is still rather high. So we obtained a ratio of (3i?):(35)-149 of 96:4 in the reduction with (5)-158, whereas for (7?)-158 the ratio was 15:85. Obviously the first combination was ,^atched" and the second one „mismatched".
Figure 4:
Crystal Structure of ACRL Toxin ill B (128)
^
I^Ph
'^^B
OH
BH3
Keton
Me (S)-OxazaborolidJne (157)
HV^i Rs
158 RL
159
186
Synthesis of the C-26-C-32-Tetrahydropyran Moiety of Swinholide A (161) Swinholide A is an interesting physiologically highly active marine metabolite with a macrocyclic diolide structure and a polyketide carbon skeleton. Recently the first total synthesis of 161 was reported by I. Paterson et al. (21). We focused on the synthesis of the tetrahydropyran part of the molecule as represented by compound 162. The particular feature of this ring is that it bears the largest substituent (at C-27) in an axial arrangement, as shown by the X-ray crystal structure of 161.
OMe
OMe
Me' ^^^^LST^'H MeO 162
Swinholide A (161)
X-ray strucute of 161: axial position of 27-substituent
187
This means that this substituent has to be arranged by a kinetically controlled stereo selective method, which in our case was a Hetero-Diels Alder reaction between a diene 164 and a glyoxylate 165. Ketone 166 is the precursor of 164 and tartaric ester 167 that of 165. The methyl ether 164 could not been made by deprotonation/methylation of 166 (Scheme 27).
.^^k^v-OH
MeO
COOR*
MeO
162
163
H ^ ^ ^ R * 0 165
V
0
OH OR*
R*0 OH 166
O
167
Scheme 27: Retrosynthesic Plan: Hetero-Diels-Alder-Reaction
Instead the silylether 168 was prepared and treated with 169 under thermal and Lewis acid mediated conditions. The stereochemical consequences were enormous. Under thermal conditions a 2:1-mixture of cis- and trans-isomcrs were formed, whereas the MgBr2-induced reaction furnished /ra/i^-cycloadduct 170 exclusively. In asymmetric
versions using menthyl-or 8-phenylmenthylglyoxalates 171 and 173, the cycloadducts 172 (l:l-diastereomeric mixture with respect to the absolute configurations around the tetrahydropyran ring) and 174 (> 97 % diastereomerically pure) were obtained (Scheme 28)
OEt
II 0
TMSO
0^^-^^C02Et
19
168
11
toluene / A >>
170 (cis:trans 2:1)
\^MgBr2/0°C^
170 (only trans\)
symmetric Versions:
168
^
"VY^ 0
Y
MgBr2 THF
0
^^
'C02Menthyl
172(1:1)
171 Ph 168
+
"YxJ) 173
MgBr2 THF
174 (97:3)
Scheme 28: Hetero-Diels-Alder Addition
The effect of the magnesium bromide is interpreted in terms of transition states 175 (small steric interaction of the bromide with the diene) and 176 (severe interaction).
189 exo
Br OTHF
endo
^ ^Mg-OTHF . Br 0
175 favorable H <=> Br
^\ '
OTHF
^Mg-OTHF " "Br :0
176 unfavorable HC <=> Br
172a was submitted to a X-ray crystal structure analysis (Fig. 5). As in 161 the largest substituent adopts an axial position which in this case may be due to an electrostatic repulsion between the lone pairs of the endocyclic oxygen and the ester oxygen atoms. Carbonyl reduction proceeds stereoselectively from the axial side. This is obviously a consequence of Cieplak's model (22). For nucleophilic additions to cyclohexanone with electron withdrawing substituents in the ^-position to the carbonyl an enhanced tendency towards axial attack is postulated.
Figure 5:
Crystal Structure of 172a
190 Although this model has so far been discussed for cyclohexanones only it apparently is also applicable to of tetrahydropyran ketones such as 172. For attachment to the acyclic polyketide chain 175 was converted into the bromide 176 (Scheme 29). The further synthesis of 161 is continued in our laboratories.
i
)'"'^^-''''^C02Menthyl 172b
NaBH4 / MeOH
axial attack contra sterically!
=
HO'' ^^•^^'^C02Menthyl 175 1. MeOTf 2. LAH
^^ . .^.^^^ V / ^ O H HO" 176
Scheme 29: Carbonal-Reduction
Acknowledgement. The results reported in this review have been achieved by a number of unusually capable and active young scientists: Dr. Jom Bilow, Dr. Catarina Pietschmann, Dr. Bemd Schollhom, Dipl.-Chem. Martin Hiersemann, Dr. Barry Bunn, Dr. Giinter Funk, Dr. Stefan Greifenberg and Dr. Susanne Dupre and Dipl.-Chem. Frank Meyer. The X-ray crystal structure analysis has been performed by Dr. Jlirgen Buschmann and Prof Dr. Peter Luger, FU Berlin. I thank all collaborators for their enthusiasm and experimental shill. 1 am also very indebted to Prof D. Arigoni and Dr. Martinoni, both ETH Zurich, for their splendid analysis of chiral acetic acids. Financial support from the Schering AG, Berlin and from the Deutsche Forschungsgemeinschaft is also greatfuUy acknowledged.
References See for example: J.D. Bu'Lock in: D.H.R. Barton and W.D. OlUs (Eds), Comprehensive Organic Chemistry, Vol. 5, Pergamon, Oxford, 1979, pp.927. S. Omura andY. Tanaka, m: S. Omura (Ed), Macrolide Antibiotics, Academic Press, Orlando, 1984, p. 199. J. Staunton, Angew. Chem. Int. Ed Eng., 30 (1991) 1302.
191 2
R.W. Hoffmann, W. Ladner, K. Steinbach, W. Massa, R. Schmidt and G. Snatzke, Chem. Ber., 114(1981)2786.
3
J. Mulzer, C. Pietschmann, B. Scholhom, J. Buschmann and P. Luger, Liebigs Ann. (1995) in press.
4
J. Mulzer and B. Schollhom, Angew. Chem. Int. Ed Eng.,29 (1990) 1476.
5
D.A. Evans and J.M. Takacs, Tetrahedron Lett., 21 (1980) 4233. P.E. Sonnet and R.R. Heath, J. Org Chem., 45 (1980) 3137. D.A. Evans, M.D. Ennis and D.J. Mathre, J. Am. Chem. Soc, 104 (1982) 1737. Y. Kawanami, Y. Ito, T. Kitagawa, T. Taniguchi, T. Katsuki and M Yamaguchi, Tetrahedron Lett., 25 (1984) 857. W. Oppolzer, P. Dudfield, T. Stevenson and T. Godel, Helv. Chim. Acta, 68 (1985) 212. W. Oppolzer, R. Moretti and S. Thomi, Tetrahedron Lett., 30 (1989) 5603. K.-S. Jeong, K. Parris, P. Ballester and J. Rebek, h., Angew. Chem., 102 (1990) 550. T.-H. Yan, V.-V. Chu, T.-C. Lin, C.-H. Wu and L.H.Liu, Tetrahedron Lett., 32 (1991)4959.
6
M. Hiersemann and B. Bunn, unpublished results, Freie Universitat Berlin, 1993-1995.
7
Gunter Funk, PhD Thesis, Freie Universitat Berlin, 1991.
8
J. Bilow, PhD Thesis, Freie Universitat Berlin, 1994.
9
J. Liithi, J. Retey and D.Arigoni, Nature, 221 (1969) 1213. J.W. Comforth, J.W. Redmond, H. Eggerer, W. Buckel and C. Gutschow, Nature, 221 (1969) 1212. Review: H.G. Floss and S. Lee, Ace. Chem. Res., 26 (1993) 116.
10
We thank Dr. Martinoni and Prof.Arigoni, both ETH Ziirich for the determination of the enantiomeric excess.
11
D.J. Cram and R.T. Uyeda, J. Am. Chem. Soc, 84 (1962) 4358. Review: A.J. Huber and H. Reimlinger, Synthesis, 1969, 97.
12
J. Mulzer, S. Greifenberg, J. Buschmaim and P. Luger, Angew. Chem. Int. Ed. Eng., 32(1993) 1173.
13
J. Mulzer and G. Funk, Synthesis, 1995, 101.
14
Y. Shizuri, S. Nishiyama, H. Shigemori and S. Yamamura, J. Chem. Soc, Chem. Commun. (1985)292. D.R. Williams and F.H. White, Tetrahedron Letters 26 (1985) 2529. M.C. Bowden, P. Patel and G. Pattenden, Tetrahedron Letters 26 (1986) 4793. M.J. Begley, M.C. Bowden, P. Patel and G.Pattenden, J. Chem. Soc, Perkin Trans. I (1991)1951 S. Nishiyama, Y. Shizuri, S. Yamamura, Tetrahedron Letters 26 (1985) 231.
192 B.M. Trost, J.K. Lynch and S.R. Angle, Tetrahedron Letters 28 (1987) 375. S. Hatakeyama, Y. Matsui, M. Suzuki, K. Sakurai and S. Takano, Tetrahedron Letters 26 (1985) 6485. H. Suh and C.S. Wilcox, 1 Am. Chem. Soc. 110 (1988) 470. S. Hatakeyama, K. Sakurai, H. Numata, N. Ochi and S Takano, J. Am. Chem. Soc. 110 (1988)5201. K. Wang, H. Venkataraman, Y.G. Kim and J.K. Cha, J. Org Chem. 56 (1991) 7174. 15
S. Hanessian and Y. Guindon, Tetrahedron Lett, 21 (1989) 2305.
16
J. Mulzer, S. Dupre, J. Buschmann and P. Luger, Angew.Chem.InlEd.Eng., 32 (1993) 1452.
17
F.W. Lichtenthaler, J. Dinges and Y. Fukuda, Angew. Chem. Int. Ed Eng, 30 (1991) 1339.
18
I. Paterson and D.J. Wallace, Tetrahedron Lett., 35 (1994) 9477. M.J. Munchhof and C.H. Heathcock, J. Org Chem., 59 (1994) 7566.
19
J. Mulzer, H. Kir stein, J. Buschmann and Ch. Lehmann, J. Am. Chem. Soc, 113 (1991)910.
20
E.J. Corey, Pure & ApplChem., 62 (1990) 1209.
21
T.K. Jones, D.C. Liotta, I. Shinkai and D.J. Mathre, J. Org Chem., 58 (1993) 799.
22
I. Paterson, K.S. Yeung, R.A. Ward, J.G. Gumming and J.D. Smith, J. Am. Chem. Soc, 116(1994)9391.
23
A.S. Cieplak, B.D. Tait and C.R. Johnson, J. Am. Chem. Soc, 111 (1989) 8447.
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
193
Total Stereoselective Synthesis of Acetogenins of Annonaceae : A New Class of Bioactive Polyketides Bruno Figadere and Andre Cave 1.
INTRODUCTION 1.1 Cl^$$ifigation Annonaceae, a family of tropical and subtropical trees, are known by populations of South
America either for their edible fruits (for species of the Armona genus) or for their uses in traditional medicine as pesticide, antiparasite, etc.... Until 1980, the chemical studies concerned mainly isoquinoline alkaloids and secondarily neutral compounds such as terpenes, fatty acids, flavonoids (1). In 1982, from Uvaria acuminata was isolated a compound with an original structure, uvaricin, displaying an antitumoral activity, which belongs to a new class of natural products, bistetrahydrofuranoid fatty acid lactones (2). The biogenesis of this product was discussed and the polyketide origin through acetyl-coenzymeA elongation process was admitted. In 1984, isolation of roUinicin, a related compound with an interesting cytotoxic activity, has been described (3) and the name of linear acetogenins was proposed for this type of natural product. Due to their specificity and natural source, the name of Annonaceous acetogenin is now systematically used. To date, about 100 related compounds have been isolated and characterized exclusively from the Annonaceae (4a-b).
TETRAHYDROFURAN MOIErY (l,2or3THF)
Y-LACTONE
R = O, OH, OAc
Acetogenins of Annonaceae These cytotoxic molecules possess 35 to 37 carbon atoms, in a long alkyl chain, bearing oxygenated functions (e.g. hydroxyl, acetoxyl, ketone), and/or double bonds, one to three tetrahydrofuran rings (THF), with a y-butyrolactone at the end. Because of the presence of these functionalized groups, acetogenins possess many stereogenic centres. These compounds have been classified in four main types A-D as a function of the number and position of the THF rings. Type A is characterized by the presence of one THF ring, a,a'-dihydroxylated, as for solamin 1 (5), type B by two adjacent THF rings, a,a'-dihydroxylated, as for isomolvizarin 2 (6), and type C by two THF rings separated by 4 carbon atoms as for otivarin 3 (7). Very recently a new acetogenin, goniocin 4 (8), representing the fourth type D has been characterized, bearing three contiguous THF rings, a-
194 hydroxylated. These compounds are further subdivided into three subtypes 1-3 as a function of the nature of the y-butyrolactone. Subtype 1 is characterized by the presence of an a,p-unsaturated ymethyl-y-lactone. Subtype 2 is characterized by an a-acetonyl-y-butyrolactone, and subtype 3 by an p-hydroxy-y-methyl-y-lactone.
f TYPE ) HO ^^ '
r OH
\ SUBTYPE
.O^
)
O
hr^ HO
^^^^
Ai 1 Bi 1
H
OH
\
SUBTYPE-1
hr
TYPEB HO
OH
r
B2 1 C2 1
SUBTYPE-2
HO
TYPEC .O^
OH
s
«°
^O^
HO-Q^
A3 1 Ba 1
\ SUBTYPE-3
Acetogenins of Annonaceae OH
16\
OH
/19
goniocin(Dl)4
wmJk
195 In addition to these acetogenins, some new products have been recently isolated, which bear in place of THF rings, epoxy groups and/or double bonds as for corepoxylone 5 (9), and sometimes only oxygenated function (e.g. hydroxyl, ketone) as for reticulatamol 6 and reticulatamone 7 (10). These related compounds belong to the rapidly growing group of the biogenetic precursors and metabolites, as for muricatacin 19 (11), of annonaceous acetogenins .
Acetogenins precursors (and metabolites) 10
corepoxylone 5
X= OH,H : reticulatamol 6 X= O: reticulatamone 7
muricatacin 19
1.2 Isolation Extraction and isolation of acetogenins of Annonaceae from the seeds, bark, leaves, or roots are guided by bioassays and TLC (12). The methanolic extract is partitioned with solvents (e.g. hexane, H2O, CH2CI2) and several chromatographic steps are necessary in order to separate from the complex mixture, compounds with very close polarity. The use of HPLC is very helpful (normal or reverse phase), since Light Scanning Detection (LSD) allows one to trace acetogenins even if they lack a chromophore (13). From the analytical point of view, it is worth noting that gpc can be used with acetogenins which have been previously treated with TMSCl, in order to prepare the corresponding silyl ethers (14). Recentiy, it was shown that extraction conditions are crucial in order to avoid re-arrangements occurring with 4-hydroxy acetogenins, leading to the isolation of artefacts (e.g. acetogenins of subtype 2). Indeed, it is now admitted that acetogenins of subtype 2 are formed by the fran^-lactonization of 4-hydroxy acetogenins (15). Dosage measurements of acetogenins in a crude extract have been studied by mass spectrometry. The close examination of FAB spectra with mnitrobenzyl alcohol doped with LiCI performed on the crude extract, allows one to know the relative composition in acetogenins of the mixture (16).
1.3 Structural Elucidation The elucidation of acetogenin structures is rather difficult and requires, besides classical methods such as UV, IR, proton and carbon NMR and mass spectrometry, some innovative mass strategies such as mass-tandem or colhsion-induced-dissociation (CID) of [M-hLi]"*" ions using linked
196 scan analysis at constant B/E (17). Concerning the determination of the configuration of the many stereogenic centres, the problem is complex because of the waxy nature of these compounds. Comparison of the NMR spectra with those obtained for models with known configurations, allows determination of the relative configuration of the THF skeletons (18). For the absolute configuration it has been proposed to apply Yamaguchi's method (19), which consists in analysing the NMR spectra of the Mosher's esters of acetogenins at high fields and deducing absolute configurations of carbon atoms bearing the hydroxyl groups. However, even though this method has been used with success for several natural products, some exceptions have been observed (20). Therefore, the determination of absolute configurations made so far by this method (21) have to be confirmed by stereoselective synthesis. Another point to stress is that the absolute configuration of an isolated carbon atom bearing a hydroxyl group cannot be determined by this method because of the intrinsic limitations. Recently, the determination of configurations of isolated hydroxyl groups has been made possible through NMR analyses of formaldehyde acetal derivatives coupled with Mosher esters methodology (22). For the stereogenic centre of the y-methyl-y-lactone, it has recently been proposed that the absolute configuration is (S), because of the oxidative degradation studies made for uvaricin which have shown the presence of (5)-lactic acid (23). In fact it is unclear for most acetogenins if this configuration is correct or not, because of the lack of any degradation studies. However, circular dichroism (CD) spectra have been used in order to deteimine the absolute configuration of this stereogenic centre, and a negative Cotton effect is in accord with the {S) proposed configuration (24).
1.4 Biological Agtivife It is now evident, that all acetogenins isolated so far, possess, to varying degrees, in vitro cytotoxicity against a large variety of carcinogenic cell lines (25). These cytotoxicities, measured at ED50, range among 10"! to 10" 1^ ^lg/mL according to the nature of acetogenins and the cell line. Some acetogenins exhibit an antiparasitic activity, and preliminary studies have shown some structure-activity relationships, leading to compounds with good therapeutic index, which have been patended (26). Pesticidal activity has also been described for several acetogenins, confirming traditional uses in South America (25). Recently, an interesting immunosuppressive activity was shown on mixed lymphocytes reaction in mouse cell system (27). For example, annonacin afforded CI50 = 3nM on this model (compared to cyclosporin with 10 nM on the same model). The mechanism of action of these new compounds is unknown. It has been shown that annonacin improved extrusion of K+ from lymphocytes (28) through a possible mechanism similar to antibiotic ionophores. The strong activity recently observed against complex I in mitochondrias could explain the high cytotoxicity found for such compounds (29). It is to answer so many questions that different groups around the world are studying the total stereoselective synthesis of acetogenins of Annonaceae.
197 2.
SYNTHESIS OF ACETOGENINS OF ANNONACEAE OF TYPE A (MONO-THF) 2. 1 Introduction Total asymmetric syntheses of natural and un-natural acetogenins of type A and type B have
been recently reported in the literature. Most of them are dealing firstly with the preparation of the THF fragment bearing the right relative and/or absolute configurations of the stereogenic centres, secondly with the preparation of the lactone moiety and finally with the coupling of the two synthons. The asymmetric syntheses are based on two different approaches, namely : (i) stereospecific strategies using as starting material a compound from the chiral pool (a-amino acids, sugars) and (ii) asymmetric induction using homochiral catalysts (Sharpless' epoxidation, Sharpless' asymmetric dihydroxylation). Besides these pathways, numerous approaches have also been reported dealing with the preparation of models which can be used as building blocks in the total synthesis of natural acetogenins (e.g. 2,5-disubstituted tetrahydrofurans (30-35), contiguous THF rings (36-38), ymethyl y-lactones (39, 40), a-acetonyl-y-lactones (41), ...). However, these approaches will not be discussed in this presentation.
2. 2 Stereospecific synthesis from chiral pool 2. 2. 1 From q-amino agids 2. 2. 1. 1 Svnthesis of gwr-4-oxo-2.33-dihvdrosolamin 8 (42) a-Amino acids are very convenient starting materials for the stereospecific syntheses of natural products (43). Glutamic acid, one of the most inexpensive a-amino acids is commercially available as its (5) and (R) form, allowing access to both parts of the molecule (the THF moiety and the y-methyl-y-lactone) in either (R) or (5) series. When the total syntheses of solamin 1 and murisolin were undertaken the relative configurations of contiguous stereogenic centers were known but the absolute configurations were unknown. Therefore arbitrarily the (155, 165, 195, 205, 34R) isomer of solamin 1 and (45, 155, 165, 195, 205, 34/?) isomer of murisolin, which appeared in 1993 to be the unnatural enantiomers of both compounds (21), were synthesized. The retrosynthetic pathway used was based on a disconnection of the carbon-carbon bond, between C-6 and C-7, which could be formed by a radical coupling of an alkyl iodide and an enone. The carbonyl so obtained could then be either completely reduced to afford solamin 1 after introduction of the unsaturation, or partially reduced to afford murisolin. This required the preparation of the enone 16 bearing the requisite configuration at C-34, and the alkyl iodide 27 bearing the THF moiety with the desired relative and absolute configurations for the four contiguous stereocentres. The synthesis of the enone, summarized on figure 1, starts from pure (R) or (5)-y-methyl-ylactone 12 which can be prepared in 4 steps from L- or D- glutamic acid. Deamination of glutamic acid by NaN02 in acidic medium gave rise to the carboxylic lactone 9 with complete retention at the stereogenic centre. Reduction of the carboxylic acid 9 by BH3.SMe2 then afforded the corresponding
198 alcohol 10 which was tosylated in a straightforward manner (TsCl pyridine). Reduction of the tosylate 11 was then performed in THF under reflux in the presence of 1 eq. of sodium iodide and 1 eq. of tributyltin hydride and a catalytic amount of AIBN. The y-valerolactone 12 was obtained in 80 % yield for the last step (46 % overall yield from glutamic acid in 4 steps and > 99% ee). Alkylation of 12, by treatment with 1 eq. of LDA and allyl bromide, led to a diastereomeric mixture of cis and trans alkylated products 13. Oxidative cleavage of the double bond by a catalytic amount of osmium tetraoxide in the presence of sodium periodate in dioxan, gave the desired aldehyde 14 in 70 % yield which upon addition of vinylmagnesium bromide, followed by a Swem oxidation, led to the desired enone 16 in 49 % yield for the last two steps.
D-giutamic acid
13V
1 4\
15V
16
\
Reapents! 1) NaN02, H2SO4, 70 %; 2) BH3.SMe2, THF. 98 %; 3) TsCl, pyridine, 87 %; 4) Nal, n-BusSnH, AIBN cat., THF, 80 %; 5) (i) LDA, TMSCl, (ii) allyl bromide, THF, 90 %; 6) OSO4 cat., NaI04, dioxan, 70 %; 7) vinylmagnesium bromide, THF, 0 °C, 51 %; 8) (C0C1)2, DMSO, Et3N, 96 %.
Preparation of the alkyl iodide 27 also started from L-glutamic acid, through a deamination process as described above. Treatment of the carboxylic acid 9 with oxalyl chloride in dichloromethane with a catalytic amount of DMF, gave the desired carboxylic acid chloride 17 in 92 % yield. Acylation of dodecylmagnesium bromide at low temperature and concentration with the acid chloride 17 afforded the corresponding ketone 18 in 85 % yield. Reduction of this ketone with LSelectride^M gave rise to the syn compound 19, namely (+)-muricatacin (44), as the major product (syn/anti = 98:2). It is worth noting that the use of tri-n-butyltin hydride with silica gel in dichloromethane allowed the and compound to be prepared as the major product with a
IMlsynlanti
ratio (45). Muricatacin 19 was then protected as a silyl ether 20 in 92 % yield by treatment with tertbutyldimethylsilyl chloride in DMF in the presence of imidazole. Reduction of the latter by DIB AL in toluene at -78 °C afforded the desired hemiacetal, which upon addition of acetic anhydride led to a 1:1 mixture of the anomeric acetates 21. This mixture when treated with trimethylsilyl cyanide in Et20 in the presence of a catalytic amount of either trityl- or scandium perchlorate gave rise to a 1:1 mixture of cis and trans nitriles which were separated by flash chromatography. Treatment of the cis nitrile with sodium r^rr-butoxide at room temperature in r^rr-butanol for 24 h led to the trans product 22 in quantitative yield. DIBAL reduction of 22 then afforded the corresponding aldehyde 23 (46),
199 whereas direct treatment with a functionalized Grignard reagent in the presence of trimethylsilyl chloride gave rise to the expected ketone 24 in 87 % yield. Reduction of the latter with LSelectride""^ yielded the syn-trans-syn compound 25 (98:2 d.e. determined by NMR) with the (5) absolute configuration for all stereogenic centres. Deprotection of the silyl ethers with tetrabutylamonium fluoride (TBAF) led to the triol 26 which upon treatment with 1 eq. of tosyl chloride at 0 °C gave rise to the monotosylated compound in 61 % yield. Displacement of the tosyl group by sodium iodide afforded the desired iodo compound 27 in quantitative yield. The cross coupling of the enone 16 with the iodo derivative 27 was performed under radical conditions by treatment of a stoichiometric mixture of 16 and 27 with 2 eq. of tributyltin hydride and a catalytic amount of AIBN in toluene under reflux. The desired coupled compound 8, namely 4-oxo-2,33dihydrosolamin, was then obtained in 55 % yield (Fig. 2). The synthesis was therefore achieved in 14 steps and 6.4 % yield from L-glutamic acid. Two more steps, i.e. reduction of the carbonyl group and introduction of the unsaturation would lead to either solamin 1 or murisolin.
L-glutamic acid 1 8
COCizHzs
1 9>—C12H25
<^
HO C(0)C8H,60TBDMS
^>— C12H25
8 Reagents: 1) NaN02, H2SO4, 70 %; 2) (C0Cl)2, DMF cat., CH2CI2, 92 %; 3) dodecylmagnesium bromide, -78 °C, THE. 85 %; 4) L-SelectrideTM, .78 ^c, THE, 88 % {syn/anti= 98/2); 5) TBDMSCl, imidazole, DMF, 99 %; 6) (i) DIBAL, -78 X , toluene, 99 %; (ii) (Ac)20, Et3N, DMAP, 20 °C, 96 %; 7) (i) TMSCN, SCCIO4 cat., Et20, 0° °C, 96 %, (x/p= 1:1; (ii) tert-BuOK, tert-BuOH, 20 °C, 24 h, 100 %, a/p= 100:1; 8) rfrr-BuMe2SiO(CH2)8MgBr, toluene, TMSCl, -78 °C, 75 %; 9) L-SelectrideTM, -78 °C, THE, 71 %, isyn/anti=^ 98:2); 10) TBAE, 20 °C, THE, 91 %; 11) (i) TsCl, pyridine; (ii) Nal, acetone, 61 % (for the last two steps); 12) n-Bu3SnH, AIBN cat., 16, toluene, 55 %.
200 2. 2. 1. 2 Synthesis of gp/-corrossolin 41 (47) Wu reported the synthesis of an epimeric mixture of natural corrossolin, starting from L-glutamic acid, which does not bear either the correct relative or the correct absolute configurations of the stereogenic centres in the molecule. Corrossolin (48) was described having a threo-trans-threo configuration across the THF ring, which means that the absolute configuration must be either (155, 165, 195, 205) or (15i?, 16/?, 19/?, 20/?). The strategy used was based on the enantiocontrolled preparation of both parts of the molecule and coupling of the two synthons by addition of a lithium acetylide on an epoxide. The disconnection was envisaged between carbon atoms C-12 and C-U. The lactone fragment was synthesized from methyl undecenoate which upon treatment with 1 eq. of LDA and then (/?)-0-tetrahydropyranyl lactal gave the aldol type product which was protected as its methoxymethyl ether 28 (in 55 % overall yield) and then hydrolyzed by H2SO4 10 % in THF to yield the lactone 29 quantitatively. Epoxidation of the double bond by MCPBA led to the desired epoxide 30 as an epimeric mixture of (34 R) diastereomers at C-10, C-2 and C-33 (unsaturation will suppress the stereogenic centres at C-2 and C-33) (Fig. 3). Figure 3
Methyl undecenoate
2 9
MOMO
3 0
Rfiassma: 1) (i) LDA; (ii) (/?)-0-THPlactal, 65 % (for the last two steps); (iii) MOMCI, /-Pr2NEt, 85 %; 2) 10 % H2SO4, THE, 100 %; 3) MCPBA, 64 %.
The THF fragment was prepared from L-glutamic acid which upon deamination with NaN02 in acidic medium followed by reduction of the so formed carboxylic acid 9, led to the desired alcohol 10 which was protected as a benzyl ether 31 by the usual method. Reduction of the lactone by DIBAL-H at -78 °C then led to the corresponding hemiacetal which upon Wittig homologation with methylenetriphenylphosphorane gave the desired alkene 32. lodo-etherification of this y-hydroxy alkene led to a 5:1 trans/cis mixture of 2,5-disustituted THF 33. Iodine displacement by ammonium acetate followed by saponification, and subsequent oxidation of the resulting free alcohol 34, led to the desired aldehyde. Addition of dodecylmagnesium bromide to this aldehyde afforded a 3:1 mixture of the syn/anti alcohols, 35 and 36, respectively, in 67 % yield which was separated by flash chromatography. It is worth noting that the undesired anti isomer 36 can be oxidized into the corresponding ketone under Swem conditions and the latter reduced by L-Selecu-ide-"^ to give rise to the syn alcohol 35 in 51 % yield for the last two steps. Acetylation of the free hydroxyl group of 35
201 and hydrogenolysis of the benzyl ether function led to alcohol 37, which under Swem oxidation conditions led to the the expected aldehyde. Addition of propargylzinc bromide then gave the anti homopropargyl alcohol 38 as the major compound with a 8.3:1 anti/syn ratio. It is worth pointing out that the anti product is the undesired epimer (since in corrossolin, the relative configuration is syntrans-syn), but the synthesis was carried out with this compound. Therefore, after saponification of acetate 38 and protection of the free hydroxyl groups as tetrahydropyranyl acetal 39, n-butyl lithium was added followed by BF3.0Et2 and epoxide 30 to afford the coupled product 40 in 58 % yield. Hydrogenation of the triple bond, deprotection of hydroxyl groups with PPTS in methanol and dehydration of p-hydroxyl-y-methyl-y-lactone by treatment with DBU in THF at room temperature, then led to the title compound 41. The synthesis was achieved in 20 steps and in 1.14 % overall yield from L-glutamic acid (Fig. 4).
L-glutamic acid
C12H25
Reagents: 1) NaNOi, H2SO4, 70 %; 2) BH3.SMe2. 98 %; 3) Ag20, BnBr, 83 %; 4) (i) DIBAL; (ii) CH2PPh3, 49 % (for the last two steps); 5) I2, NaHCOs, 50 % {trans/cis^ 5:1); 6) (i) Et4N0Ac, 67 %; (ii) K2CO3, 100 %; 7) (i) (C0C1)2, DMSO, Et3N; (ii) dodecylmagnesium bromide, -20 °C, THF, 67 % (for the last two steps) {syn/anti= 3:1); 8) Jones* oxidation; 9) L-SelecUide^M, 51 % (for the last two steps); 10) (i) (Ac)20, pyridine, 98 %; (ii) H2, Pd-C, 96 %; 11) (i) (C0C1)2, DMSO, Et3N; (ii) Zn, propargyl bromide, DMF/Et20 (1:1), 65 % (for the last two steps) {anti/syn= 8.3:1); 12) (i) K2CO3, 100 %; (ii) DHP, PPTS, 98 %; 13) n-BuLi, BF3.0Et2, 30, -78 °C, 58 %; 14) (i) H2, Pd-C, 100 %; (ii) PPTS, MeOH, 76 %; (iii) 4 eq. DBU, THF, R.T., 4 h, 68 %.
202
2. 2. 2 Synthesis of aldehyde 23 Even though glycosides seem to be the starting materials of choice for the synthesis of monoTHF acetogenins of Annonaceae, yery few examples are known in the literature. Indeed only one approach has been proposed by Gesson (49, 50) starting with D-glucofuranose which has been protected as a bis-acetonide and a benzyl ether before oxidation at C-6 to giye the corresponding aldehyde 42. Addition of tetradec-3-ynyimagnesium bromide on 42 afforded a 4:1 mixture of syn/anti alcohols 43, 44 which could be separated by flash chromatography. Lindlar hydrogenation of the major compound 43 followed by oxidation with MCPBA gave rise to a 1:1 mixture of trans/cis THF compounds 46, 47, which were separated by HPLC. Treatment of the trans product 46 by acetic acid and then sodium periodate led to the desired trans aldehyde 23 (Fig. 5). The use of this aldehyde as a building block for the total synthesis of acetogenins has not been reported yet by these authors. Figure 5 C10H21
H
R2H
D-glucose RiO
H
H H
C10H2
47 (Ri= Bz) Bno^ +
BnO 4 2
'-.x
BJ
HO.
>
R10 = H H H - O I 5 .0
46 (R,= Bz)
BnO
43 : Ri= OH. R2= H 44 : Ri= H, R2= OH 5
ot
CioHj'
45 R,0 = H
'••X_ CioHat
^^^
H i f "
^
23 : Ri= H
Rgaggnts: 1) ref 51; 2) l-bromo-3-tetradecyne, Mg, THF, syn/anti= 4:1; 3) Lindlar hydrogenation; 4) (i) MCPBA. CH2CI2, cis/trans^ 1:1; (ii) BzCl; 5) (i) H2, Pd-C; (ii) ACOH/H2O (1:1). 50 °C; (iii) 3 eq. NaI04. H2O. 20 °C.
2. 3 Asymmetric synthesis of Solamin 1 2. 3. 1 Keinan's synthesis (52) Keinan prepared separately the two fragments (the THF moiety and the y-methyl-ylactone) and used, as a key step of his sequence, the asymmetric dihydroxylation (AD-mix.-p), the yery efficient Sharpless' procedure for the formation of a.p-diols. Then, the cross-coupling was performed by addition of an alkyne and a yinyl halide in the presence of palladium and copper catalysts (Fig. 6). Treatment of the unsaturated ester 48 (prepared in 4 steps from commercially ayailable starting material, and 65 % oyerall yield) with AD-mix.-p in ^err-butanol/water (1:1) with methanesulfonamide for 16 h at 0 T afforded the lactone 49 which possessed 3 carbon atoms out of the 4 with the desired absolute configuration. Inversion of the fourth stereocentre after acetonide
203 formation of the vicinal diol (2 steps: tosylation, and epoxidation) afforded the lactone-THF 53. DIBAL reduction of the latter, followed by Wittig homologation with dibromomethylene triphenylphosphorane gave rise to the bromo alkene 54. Alternatively the lactone fragment was prepared in a straightforward manner as an alkyne derivative 55 (in one step and 70 % yield from(2/?, 4S) and (25, 45)-4-methyl-2-phenylthio-Y-butyrolactone (53)), which upon reaction with the bromo-alkene 54 in the presence of palladium triphenyl tetrakis, copper iodide, EtsN in THF at 50 °C gave rise to the enyne 56 in 70 % yield. Hydrogenation of the enyne 56 afforded 57, which after oxidation-thermal elimination of the phenylsulfoxide led to the desired solamin 1. In conclusion, this synthesis was achieved in 14 steps and 7.7 % yield from commercially available starting material. It is worth noting that the total synthesis of reticulatacin 154 (54) (which differs from solamin only by the length of the alkyl chain which bears two extra carbon atoms) has also been realized by the same authors. Figure 6
O ^COOB
f^^
OH
1
C12H25
4 8
k^COOMe
°
51
°-^
^y. -,^^%^ 56
i^ o
solamin 1 : n= 1 reticulatacin 154: n= 3
.^
o k.^ \ uo
57
C12H25
Reagents: 1) AD-mix-P, 66 %; 2) DMP, acetone, TsOH, 98 %; 3) TsCl, EtsN, DMAP, CH2CI2, 97 %; 4) K2CO3, MeOH, 88 %; 5) BF3.0Et2, CH2CI2, 75 %; 6) (i) DIBAL, -50 °C, THF; (ii) BrCH2PPh3-^Br-, rerr-BuOK, THF, 60 % (for the last two steps); 7) 55, Pd(PPh3)4, Et3N, Cul, 70 %; 8) H2, RhCl(PPh3)3, 95 %; 9) (i) MCPBA; (ii) toluene reflux, 72 % (for the last two steps).
204 2. 3. 2 Tanaka's synthesis (55) The key steps in Tanaka's synthesis are on the one hand the very efficient asymmetric epoxidation of an allylic alcohol, known as Sharpless' epoxidation, and on the other hand crosscoupling of an alkyne with a vinyl halide catalyzed by palladium and copper. Alkylation of propargyl alcohol 59 with dodecyl bromide 58 in liq. ammonia by lithium amide, followed by Lindlar hydrogenation gave rise to the (Z) allylic alcohol 61. Asymmetric epoxidation by the improved Sharpless procedure afforded the epoxy alcohol 62 with 84 % ee. Tosylation of the free alcohol (TsCl, pyridine) and then displacement with iodine, gave the iodo epoxide 63 which was then reacted with lithium enolate of tert-huty\ acetate to yield the alkylated product 64. Acidic hydrolysis afforded the hydroxy lactone 19 (muricatacin) (56). The latter was protected as its methoxymethyl ether 65 before reduction with DIBAL, yielding the hemiacetal 66 which upon reaction with pent-4ynylidenetriphenylphosphorane gave the acyclic compound 67. Epoxidation with MCPBA, followed by acidic cyclization led to a 3:2 mixture of trans, cis products (detemiined later in the synthesis) with predominantly the desired trans compound, which was separated by thin-layer chromatography. Protection of the free hydroxyl as a benzoyl ester, and deprotection of both the hydroxyl groups led to the THF moiety 68 (Fig. 7).
Figure 7
n-C-j2H25Br •
5 8
^ ' 59
C12H25
3 0
C12H25
- \ / -
3 X * \ / V , ' 5 = ^ Ci2H25
C12H25
6 6
65
OMOM
C12H25,
C12H2S,
^^^ 69 : Ri=R2= H
solamin 1 : n= 1 reticulatacin 154 : n= 3
Rgaggnt?: 1) LiNH2, Et20, DMSO, 71 %; 2) Lindlar hydrogenation, 91 %; 3) rerr-BuOOH, Ti(0/-Pr)4, L-(+)-diethyl tartrate, M. S., CH2CI2, 76 %; 4) (i) TsCl, DMAP, Et3N; (ii) Nal, acetone. 97 % (for the last two steps); 5) tertbutylacetate, cyclohexylisopropylamine, n-BuLi, HMPA, 81 %; 6) camphosulfonic acid, CH2CI2, 70 %; 7) MOMCl, /-Pr2NEt, 94 %; 8) DIBAL, -78 T , CH2CI2; 9) Pent-4-yn-l-yliriphenylphosphoniuin iodide, NaOEt, 0 °C, DMF; 10) (i) MCPBA, CH2CI2, 56 % (for the last three steps); (ii) BzCl, pyridine, 0 °C; (iii) NaOH, MeOH, 79 % (for the last two steps); 11) 78, Pd(PPh3)4, Cul, Et3N, 61 %; 12) (i) H2, RhCl(PPh3)3, 60 %; (ii) MCPBA, (iii) toluene, reflux, 40 % (for the last two steps).
205
The lactone fragment 78 was prepared from ethyl lactate, which in a few steps gave the lactone 77, and from propargyl alcohol 71. Alkylation of the lactone 77 with the diiodo compound 76 (prepared in 5 steps from 71) gave the desired furanones 78 (Fig. 8). The cross-coupling reaction is based on the same palladium catalyzed reaction of a vinyl halide with an alkyne used by Keinan, but herein the alkyne bears the THF skeleton and not the lactone part, as in Keinan's strategy. Therefore the reaction of the two synthons 68 and 78 with Pd(PPh3)4 in the presence of Cul and Et3N gave the desired coupled product 79, which after hydrogenation followed by a two steps sequence (oxidation-thermal elimination) afforded the desired solamin 1 (Fig. 7). In conclusion, the synthesis was achieved in 16 steps and 1.5 % overall yield, using as key steps the Sharpless epoxidation and the palladium catalyzed cross-coupling of an alkyne with a vinyl halide. It is worth noting that the same authors succeeded in the total synthesis of reticulatacin 154.
Figure 8
70 ^^""^^^
59 OTBDMS
*^73o
^'^"
^ ^ *^
71
72
' ' ' ^ ^ ^ OTBDM9-
'^*
_,
75
^'n'
76
-DsPh
78 Reagents: 1) n-BuLi, 58 %; 2) KAPA, H2N(CH2)3NH2, 71 %; 3) TBDMSCl, imidazole, DMF, 92 %; 4) (i) nBu3SnH, AIBN cat.; (ii) I2, 70 % (for the last two steps) (£/2= 3:1); 5) TBAF, THF, 85 %; 6) (i) TsCl, pyridine; (ii) Nal, acetone, 81 % (for the last two steps) {E/Z=^ 3:1); 7) 77, NaHMDS, 51 %.
2. 3. 3 Trost's synthesis (57) The very elegant and original su*ategy used by Trost relied on key steps such as (i) asymmetric epoxidation of allylic alcohol, (ii) a new synthesis of 2,5-disubstituted THF via a Ramberg-Backlund olefination and (iii) a ruthenium catalyzed butenolide annelation to form a direct precursor of solamin (Fig. 9). The synthesis starts by treatment of propargyl alcohol 59 with n-butyl lithium in THF/HMPA, followed by addition of bromododecene 80 to afford the alkylated product 81, which upon Lindlar hydrogenation gave the (Z) allylic alcohol 82. Asymmetric epoxidation with tertBuOOH, Ti(0i-Pr)4 and L-(+)-tartrate gave the desired epoxide 83 in 82 % ee which after recrystallization gave >99 % ee. At this point, this intermediate was used to prepare the two halves of the THF skeleton. Firstly the hydroxy-epoxide 83 was convened into the corresponding iodide 84 in 93 % yield (I2, PPh3, Et3N, THF). Secondly hydrogenation of 83 led to the alkane 85 which after a Payne rearrangement, on treatment withy re/t-BuSH, afforded the expected sulfide. Removal of the
206 f^rr-butyl group was performed with Hg(0Ac)2, PhOMe and CF3COOH at 0 °C, to give the thiol 86. Coupling of the two halves was performed under basic conditions to give the 1,4-oxathiane 87 in good yield. The best protocol for the Ramberg-Backlund olefmation was then performed on the corresponding sulfone (MCPBA oxidation of 87) with the hydroxyl groups protected as their silyl ethers, by treatment with tert-BuOK in tert-BuOH in the presence of CCI4 at room temperature to afford the dihydrofuran 88 in 65 % yield. Ruthenium catalyzed butenolide annelation of diol 88 with the ynoate occured chemoselectively at the less sterically demanding double bond to give the bisdehydrosolamin 89. Chemoselective hydrogenation of the isolated double bonds was then performed with (Ph3)3PRhCl and H2 to yield solamin 1 in 95 % yield. In conclusion, the synthesis was achieved in 14 steps and 11.7 % overall yield using a new and very efficient method for the synthesis of 2,5-disubstituted THF rings as well as the atu*active ruthenium catalyzed butenolide annelation.
Figure 9
Esamils: 1) n-BuLi. -78 °C, THF, HMPA, 76 %; 2) Lindlar hydrogenation; 3) terhBuOOH, Ti(0/-Pr)4, L-(+)-diethyl tartrate, M. S. -20 '^C, CH2CI2, 90 %; 4) I2, PPhs, C3H4N2, Et3N, THF, 0 °C, 93 %; 5) H2, Pd-C, 98 %; 6) (i) tertBuSH, NaOH. tert-BuOK H2O, 81 %; (ii) Hg(0Ac)2, PhOMe, CF3COOH, 0 °C, 92 %; 7) CS2CO3, DMF, R.T., 92 %; (ii) KOH, H2O, tert-BuOH, 65 %; (iii) m-CPBA, PhH,-hexane, 0 T , 95 %; 8) (i) TMSCl, Et3N, CH2CI2, 0 °C, R.T., 94 %; (U) tert-BuOK, r^rr-BuOH, CCI4. R.T., 65 %; (iii) TsOH, H2O, EtOH, R.T., 95 %; 9) CPRU(C6D)CI! MeOH, EtOOCC<:CH*(OH)CH3, 65 %; 10) H2, RliCl(PPh3)3, PhH, EtOH, R.T., 95 %.
207 3.
SYNTHESIS OF ACETOGENINS OF ANNONACEAE OF TYPE B 3. 1 Synthesis of/^g^r^p^-yivarigin 90 (58) In 1991, Hoye was the first to report the total synthesis of an acetogenin hexepi-uvancin 90,
which has in common with natural uvaricin the right relative configuration across the THF skeleton but is of the opposite absolute configuration (2). The strategy used as a key step the asymmetric epoxidation of a homochiral bis-allylic alcohol, which allowed 5 out of 6 stereocentres to be built up at once with the desired relative configuration. A few more steps were necessary for the inversion of the last stereocenter (Fig. 10). The diiodide 91, derived from L-(+)-diethyl tartrate, was converted to E-E-bis-allyl alcohol 93 through the following sequence : Weiler dianion alkylation with f^rr-butyl acetoacetate, followed by ketone reduction to yield the corresponding diol, dehydration via the bis-mesylate and finally reduction of the ester functions by DIBAL. Asymmetric epoxidation of the bis-allylic alcohol 93 so obtained with tert-BuOOH, Ti(0i-Pr)4, D-(-)-diisopropyl tartrate afforded the bis-epoxy diol 94 in 60 % yield which possess a C2 symetry axis. Therefore monotosylation (TsCl, EtsN, CH2CI2, 0 °C) followed by acidic treatment (Amberlyst H-15, MeOH, room temperature) led to the bis-THF skeleton 96 where one out of the two primary hydroxyl groups was tosylated. Displacement of the tosyl group with excess of lithium dionylcuprate followed by acetonide formation of the vicinal diol, acetylation of the last hydroxyl led to the diol 97. After acetonide removal, three steps are required for inversion of the configuration at C-15 (numbering of acetogenin series). This was performed by protection of the primary hydroxyl group as a silyl ether (TBDPSCl, DMAP, CH2CI2, EtsN), tosylation of the secondary hydroxyl (TsCl, DMAP, pyridine), and deprotection-epoxidation with TBAF, leading to 98. This sequence was achieved in 78 % overall yield. Epoxide-opening with lithium trimethylsilylacetylide/BF3.0Et2, and removal of the TMS group by acidic treatment, then led to the terminal alkyne 99 in 58 % overall yield. Preparation of the lactone fragment started with a mixture of (2i?,45) and (25,45)-4-methyl2-phenylsulfenyl-Y-butyrolacione (53) which was alkylated with (£')-l,9-diiodo-l-nonene. The corresponding iodo compound 100 so obtained was then coupled with the alkyne 99 through the efficient palladium catalyzed reaction (Pd(PPh3)4, Cul, EtsN, room temperature) in 86 % yield. Enyne reduction of 101 with Wilkinson's catalyst, then oxidation of the sulfide into sulfoxide and subsequent thermal elimination gave rise to the title compound 90. The synthesis was achieved in 20 steps and in 0.36 % yield.
208 Figure 10
Reagents: 1) (i) NaBH4, MeOH, 0°C; (ii) MsCl, EtsN, CH2CI2, then DBU, 4 0 T ; (iii) DIBAL-H, Et20, O T ; 2) Sharpless epoxidation, (D-(-)-diisopropyl tartrate); 3) TsCl (1.0 eq.), Et^N, DMAP, CH2CI2, 0°C; 4) Amberlist-15, MeOH, 20°C; 5) (i) (n-C9Hi9)2CuLi, THE, - l O T ; (ii) H^, Me2C0, 2d°C; (iii) AC2O, pyridine, 20°C; 6) (i) MeOH, p-TsOH, 20°C; (ii) r-BuPh2SiCl, DMAP, EtsN, CH2CI2, 20°C; (iii) TsCl, DMAP, pyridine, CH2CI2, 20^C; (iv) n - B u 4 N T , 20°C; 7) (i) LiC^CTMS, BFc,:0Et2, THE, -78°C; (ii) AI-BU4N-'F, THE, 20°C; 8) (100 ), Pd(PPh3)4, Cul, EtsN, 20°C; 9) (i) Rh(PPhci)3Cl, H2, benzene; (ii) OXONE^, MeOH.HiO, 0°C; toluene,A.
3. 1 Synthesis of gnr-rolliniastatin-2 (=g/7f-bullatacin) 102 (59) Rolliniastatin-2 was first isolated from Rollinia mucosa (60), before appearing also with the name of bullatacin few years later (4a-b). Hoye reported the total syntheis of ^/7t-rolliniastatin-2 102 (=^nr-bullatacin) (Fig. 11) using a strategy very similar to that employed for hexepi-uwmcin 90. The
209 bis-THF skeleton was prepared as already described (Fig. 10). The lactone fragment was prepared from (.S)-(-)-malic acid, which was reduced with BH3.SMe2 to afford the corresponding trioL Acetonide formation of the vicinal diol and treatment with iodine, PPh3, ImH afforded the iodo compound 103. Alkylation of 1-lithio-l-pentyne with (5')-iodide 103, followed by acetonide removal and "Zipper" reaction gave the terminal alkyne 104. The latter was then converted in 4 steps into the epoxide 105 (trimethylsilylation, hydrolysis, tosylation, epoxide formation) in 20 % yield for the last 7 steps. Treatment of this epoxide by lithium enolate of a-phenylthioacetic, followed by acidification and protection of the free hydroxyl group as its silyl ether, gave the acyclic compound 106 in 52 % yield for the last 3 steps. Enolization of the carboxylic acid 106 with 2 eq. of LDA gave the corresponding dianion, which was alkylated with (/?)-propylene oxide to give, after acidic catalyzed lactonization, the expected product 107. Selective removal of the TMS group by treatment with K2CO3 gave the terminal alkyne which was iodinated (I2, morpholine) to give the iodoalkyne 108. Palladium catalyzed cross-coupling of the alkyne 99' (58) with the iodoalkyne 108 provided the diyne 109 in 30-45 % yield. Hydrogenation of the diyne 109, oxidation-thermal elimination and deprotection led to the title compound 102. The synthesis was achieved in 20 steps and in 0.8 % overall yield.
S-malic acid TMS,
>r^ 5
SPh
107 : R= H; 108 : R=l
Reagents: 1) (i) BH3.SMe2; (ii) acetone, TsOH; (iii) I2, PPh3, ImH, Et20/MeCN; 2) (i) 1-lithio-l-pentyne; (ii) CSA, MeOH; (iii) KAPA, DAP; 3) (i) EtMgBr; (ii) TMSCl; (iii) 10 % HCi; (iii) TsCl, pyridine, -10 °C; (iv) NaH, THF; 4) (i) PhSCH=C02Li; (ii) 10 % citric acid; (iii) TBSCl, ImH; (iv) MeOH; 5) (i) LDA; (ii) (/?)-propylene oxide; (iii) 10 % citric acid; (iv) CSA, PhH, reflux; (v) K2CO3, MeOH; (vi) I2, morpholine; 6) 99', Pd(PPh3)4, Cul, i-Pr2NH, THF; 7) (i) H2, RJiCI(PPh3)3; (ii) oxone^M, MeOH, H2O; (iii) PliMe, reflux; (iv) 5 % HF, MeCN, THF.
210 3. 1 Synthesis of gyI^rQlliniastatin-1 110 (61) Rolliniastatin-1 was isolated in 1987 from Rollinia mucosa (62), and since then has been found in several species of Annonaceae (4a-b). Koert reported the total synthesis of ^wf-rolliniastatin1 110 from L-glutamic acid. The strategy is a sequential synthesis, using as a key step the diastereoselective copper catalyzed Grignard addition on aldehydes (Fig. 12). Nitrile 113 (prepared from L-glutamic acid in 5 steps and 61 % overall yield) was converted into the corresponding methyl ester, which was reduced to the corresponding alcohol. Benzylation and desilylation of the latter afforded 114, which under Swern oxidation conditions, led to the corresponding aldehyde. Copper catalyzed addition of the Grignard reagent afforded the threo acetonide alcohol 115 with a 97:3 threo/erythro ratio. Removal of the acetonide, mesylation of the primary hydroxyl and basic treatment gave rise to the epoxide 116. Acidic treatment of the latter led to the bis-THF compound 117, through an intramolecular epoxide opening. Swern oxidation of the free primary hydroxyl group and subsequent addition of the Grignard reagent, followed by another Swern oxidation led to the expected ketone 118. The latter was reduced with Zn(BH4)2 to yield the erythro alcohol as the major compound {erythro/threo ratio = 82:18). After protection of the free hydroxyl
as a silyl ether (by treatment with TBDPSCl in the presence of imidazole) and
debenzylation of the primary hydroxyl, the alcohol so obtained was oxidized under Swern conditions to afford the corresponding aldehyde. Upon addition of the Grignard reagent in the presence of CuBr, the threo alcohol 119 was obtained with a 95:5 threo/erythro ratio. Deprotection of the hydroxyl groups followed by reprotection as r^rr-butyldimethylsilyl ether, yielded the expected protected tetraol. Debenzylation of the primary hydroxyl and oxidation, via the non isolated corresponding aldehyde, led to the desired carboxylic acid 120. The latter was treated with 2 eq. of LDA, followed by reaction with PhSSPh, and then with (5')-propylene oxide to give after acidic treatment the desired lactone. Oxidation and thermal elimination of the sulfoxide allowed the introduction of the double bond. Deprotection of the silyl ether with 5 % HP in CH3CN/THF afforded the title compound 110. The synthesis was achieved in 30 steps and in 2.1 % overall yield.
211 Figure 12
L-glutamic acid CHjOTBDPS
C^j20TBDPS
CH2OH
1 17
TBDMSO*
C10H2
HO*
CioHg,
Reagents: 1) NaN02. H2SO4, 70 %; 2) BH3.SMe2, THF. 98 %; 3) TBDPSCl, imidazole, DMF, 92 %; 4) (i) DIBAL; (ii) Ac20, Et3N; 5) TMSCN, BF3.OE12. Et20, 0 °C (d.e. = 50:50); 6) (i) NaOMe, MeOH; (ii) LiAlH4, THF; (iii) BnBr, NaH; (iv) TBAF, THF; 7) (i) (C0CI)2, DMSO, EtsN, CH2CI2, -78 °C; (ii) BrMoCH2CH2CH*(OR)CH20R, CuBr.SMe2, -78 °C, Et20; 8) (i) HOAc, THF/H2O; (ii) MesitylS02Cl, pyridine, 0 °C; (iii) K2CO3, MeOH; 9) HOAc, CH2CI2, R.T.; 10) (i) (C0C1)2, DMSO, Et3N; (ii) CioH2lMgBr; (iii) (C0C1)2, DMSO, Et3N; 11) (i) Zn(BH4)2, Et20, -78 °C; (ii) TBDMSCI; (iii) H2, Pd-C, THF; (iv) (C0CI)2, DMSO, Et3N; (v) CuBr.SMe2, Et20, -78 °C, 14-benzyloxy-ll-ren-butyldimethylsilyioxy-l-yimagnesium bromide; 12) (i) TBAF, THF; (ii) TBDMSCI; (iii) H2. Pd-C, THF; (iv) (C0C1)2, DMSO, Et3N; (v) KMn04, NaHP04, r^rr-BuOH, H2O; 13) (i) LDA; (ii) (5")-propylene oxide, thenp-TsOH; (iii) MMPP, THF, MeOH, then toluene; (iv) HF, MeCN, THF, R.T.
212 SYNTHESIS OF RELATED ACETOGENINS OF ANNONACEAE 4. 1 Reticulatamol 6 and reticulatamone 7 Recently, several acetogenin derivatives having in common the lack of THE rings have been isolated from Annonaceae. Reticulatamol 6 (10) and reticulatamone 7 are probably the first natural precursors which after chemical transformations could lead to unsaturated fatty acids, giving rise after oxidation to epomuricenin A 122 and then to diepomuricanin A 121 (63) which, after oxiranesopening followed by ring closures, could afford the natural acetogenin solamin 1 (Fig. 13). Chemical transformations of epomuricenin A 122 to diepomuricanin A 121 and then to solamin 1 have confirmed this proposed biogenetic pathway (see section 5. 1). Figure 13
epomuricenin A 122
32
mticulatamol 6
Total syntheses of natural reticulatamone 7 and reticulatamol 6 in 5 and 6 steps, respectively, have been recently reported in the literature (10). The strategies used required the preparation of the enone 123 in two steps from octadecanal 124 through Grignard addition of vinylmagnesium bromide followed by oxidation with manganese dioxide. The iodo compound 125 was prepared from methyl 12-bromododecanoate in 3 steps; enolization by LDA, alkylation with phenylselenium chloride followed by another enolization and alkylation with (5)-propylene oxide, which afforded after acidic treatment the desired lactone. Oxidation by H2O2 in acidic conditions led to the unsaturated lactone which after bromine displacement with Nal in acetone led to the iodo compound 125. Treatment of a stoichiometric mixture of 123 and 125 with 2 eq. of/7-Bu3SnH and a catalytic
213 amount of AIBN in toluene under reflux led, after purification by flash chromatography, to reticulatamone 7 in 57 % yield. Reduction of 7 with the couple /2-Bu3SnH.Si02 in CH2CI2 at room temperature for 48 h led to reticulatamol 6 as an epimeric mixture at C-15 in 92 % yield (10, 45). It is noteworthy that the reduction step is highly chemoselective, since neither the carbonyl of the lactone nor the conjugated double bond was affected by the reaction conditions (Fig. 14).
Figure 14
X= O : reticulatamone 7 X= OH: reticulatamol 6 Reagenis: D (i) LDA, -78 °C; (ii) PhSeCl; 2) (i) LDA. -78 °C; (ii) (5)-propylene oxide; 3) H2O2, AcOH, 55 % for Uie last 3 steps; 4) Nal, acetone, 100 %; 5) vinylMgBr; 6) Mn02, 85 % for tlie last 2 steps; 7) n-Bu3SnH, AIBN cat., reflux toluene, 56 %; 8) n-Bu3SnH/Si02, CH2CI2, r.t., 48 h, 92 %.
4. 2 (+) and/or (-)-muricatacin 19 Muricatacin 19 (5-hydroxyheptadecan-4-olide) was isolated from Annona muricata (11) as a mixture of (+) and (-) muricatacin with a slight excess of the (-)-enantiomer {4R, 5R) and it is supposed to be a natural (or unnatural) metabolite of acetogenins. It has been shown that this hydroxy-butyrolactone presents some in vitro cytotoxicity (11). Since its isolation in 1991, many syntheses have been reported in the literature and are discussed herein. 4. 2. 1 Figadere's synthesis (44) As shown in section 2.1.1 and figure 15, (+)-muricatacin 19 has been synthesized in 4 steps and 48 % overall yield from a very inexpensive starting material, L-glutamic acid. The key steps of the synthesis are a nitrous deamination of an a-amino acid widi retention of the configuration of the stereogenic centre, and a very diastereoselective reduction of a-butyrolactonic ketone 18 with L-Selectride"^. The use of D-glutamic acid allowed the preparation of (-)-muricatacin 19 as well.
214 Figure 15
L-glutamic acid
==Qs^oH _2^o=rXj^c, j - ^ o ^ r j ^ 9 O^
17
0
Ci2H25
18 o
0
o^ OH
(+)- muricatacin
19
1
Reagents: 1) NaN02, HCl, 70 %; 2) (C0C1)2, DMF cat., CH2CI2, 92 %; 3) dodecylmagnesium bromide, -78 X, THF, 85 %; 4) L-SelectrideTM, .78 ^c THF, 88 % {syn/anti= 98/2). 4. 2. 2 Tochtermann's synthesis (64, 65) Tochtermann's strategy was based on a diastereoselective reduction of a racemic abutyrolactonic ketone 124, followed by a chemical resolution in order to obtain a homochiral synthon 129 which after further chemical transformations led to the desired compound, (-)-muricatacin 19 (Fig. 16). The synthesis has been achieved in 17 steps and each transformation ranging from 77 to 97 % yield. Figure 16
CH2OH
Rfiasenls: D H2, Pt; 2) O3; 3) O2; 4) 2N NaOH, r.t.; 5) AgN03, H2O, r.t.; 6) CH3I, Et20; 7) (i) (-)-cainphanoyl acid chloride, pyridine; (ii) separation; 8) L-selecuide^M, -78 °C, THF; 9) Amberiyst H 15, CH2CI2; 10) (i) NaC)CH3, CH3OH, reOux, 2 h; (ii) NaOH, H2O, CH3OH, reflux, 5 h; (iii) HCl; 11) BH3.SMe2, THE, r.t.; 12) 2,2dimethoxypropane, CH3OH, Amberiyst H 15, 60 h, r.t.; 13) PCC, CH2CI2, 2.5 h, r.t.; 14) Ph3P+C5Hii-Br, KOrBu, Et20, reflux, 1 h; 15) H2, Pt; 16) THF, H2O, HCl, reflux, 1 h; 17) id. 16).
4. 2. 3 Marshall's synthesis (66, 67) Marshall separately prepared {+) and (-)-muricatacin 19 through 1,2-addition of nonracemic sililoxy allylic stannanes 135 on a,p-unsaturated aldehydes. The acylstannane 133 was reduced with either (R) or (S) BINAL-H to the hydroxy stannanes 134a or 134b with 95 % ee,
215 respectively. Protection as a silyl ether followed by treatment with BF3.0Et2 at low temperature led, after rearrangement, to allylic stannanes 135a and 135b. Separately treatment of the latter two compounds with the appropriate enal in the presence of BF3.0Et2 at -78 °C led to adducts 136a and 136b. Hydrogenation of 136a and 136b, followed by HF cleavage led to (+) and (-)-muricatacin 19. The syntheses were achieved in 8 steps and in 27.6 % overall yield (Fig. 17).
Figure 17 o
1,2
C10H21
CR
133
OR
132
3,4
CioH;
C10H2
SnBus
SnBua
or
5 134a BuaSn
5 134b
QR
BusSn
C10H21
OR
C10H21
135a
135b
6
CioH2i
C10H21
136a
'^^ OR
136b
7
Reagents: 1) n-Bu3SnLi; 2) DIAD; 3) (S) or (R) BINAL-H, 95 % ee; 4) TBDMSCl, imidazole, CH2CI2, 48 % for the last 4 steps; 5) BF3.0Et2, -78 °C, 95 % ee; 6) BF3.0Et2, CH2CI2, CH0CH=CHC02Et, 75-80 %; 7) H2, Pd/C, 8688 %; 8) HF, THF, H2O, 83-87 %.
4. 2. 4 Sharpless' svnthesis (68) Sharpless' strategy is based on the very effective asymmetric dihydroxylation (ADmix.-a or p) of the Y,S-unsaturated ester 140 to afford directly the desired (+) or (-)-muricatacin 19 with 95 and 96 % ee, respectively, and in 74 % overall yield from tridecanal 138 (Fig. 18). The ester 140 is obtained through a Claisen rearrangement of the allylic alcohol 139.
216 Figure 18
C12H25CH0
138
Reagents: 1) vinylMgBr, 96 %; 2) CH3C(OEt)3, EtC02H cat., reflux, 92 %; 3a) AD-mix-a., 82 %, 95 % e.e.; 3b) AD-mix-p., 84 %, 96 % e.e.
4. 2. 5 Kang's synthesis (69) Kang chose D-glucose as the starting material which was converted to the known aldehyde 141. Wittig homologation, followed by hydrogenation gave the product 142. Protection of the free hydroxyl as a benzyl ether, removal of the acetonide followed by oxidative cleavage gave rise to the acyclic aldehyde 143. Reaction of the latter with the anion of triethylphosphonoacetate provided the unsaturated ester 144. Hydrogenation and treatment with aqueous trifluoroacetic acid afforded the desired title compound 19. The synthesis was achieved in 8 steps from the known aldehyde 141 and in 15.8 % overall yield (Fig. 19).
Figure 19
Reagents: 1) CH3(CH2)l0PPh3Br, /i>BuLi, THF, r.i., 12 h, 86 %; 2) H2, Pd/C, EtOAc, r.t., 10 h, 91 %; 3) NaH, BnCl, THF, r.t., 5 h; 4) 2N HCl, DME, r.t., 48 h; 5) NaI04, MeOH, r.t., 1 h, 53 % for the last 3 steps; 6) NaH, (EtO)2POCH2C02Et, THF, r.t., 3 h, 67 %; 7) H2, Pd/C, EiOAc, r.t., 24 h, 71 %; 8) TEA, H2O (4/1), r.t., 3 h, 80%.
217 4. 2. 6 Bessodes' synthesis (70) For the preparation of (+)-muricatacin 19, Bessodes used as a key step, the kinetic resolution of Sharpless asymmetric epoxidation of the allylic alcohol 139. The alcohol 139 obtained by Grignard reaction of dodecylmagnesium bromide 58 with acrolein 146, was subjected to oxidation with tert-BuOOH in the presence of Ti(/-PrO)4 and D-diisopropryl tartrate under kinetic conditions to give the epoxy alcohol 147a. After a Mitsunobu reaction (DEAD, PPhs, CICH2COOH), the chloroacetate 148 was obtained with inversion of configuration of the stereogenic centre bearing the hydroxyl group. Subsequent deacylation by treatment with sodium methanolate, followed by treatment with lithium acetate and then acidic work-up, led to (+)-muricatacin 19. The synthesis was achieved in 5 steps from dodecyl bromide 58 and in 26.7 % overall yield (Fig. 20). The same year, Liu et al. (71) reported the total synthesis of racemic muricatacin 19, using a similar strategy, but performing the epoxidation step without asymmetric induction, and using the lithium enolate of acetonitrile for opening epoxide the 147b, before acidic treatment, leading to (-1-/-)19. Figure 20
146
139
°
147a
0COCH2C1
-^ r^^c,2H25 — • r^^c,2H25 148
147b
Reagents: 1) Ci2H25MgBr; 2) Ti(0-/Pr)4, D-diisopropyl tartrate, /-Bu02H, CH2CI2, r.t., -20 °C, 95 % (resolution yield); 3) DEAD, PPh3, CICH2CO2H, benzene, reflux, 90 %; 4) MeONa 1%, MeOH, 100 %; 5) (i) LiCH2C02Li, THE, reflux; (ii) H3O+, 65 %.
4. 2. 7 Tanaka's synthesis (56) Tanaka also prepared (-)-muricatacin 19 from dodecylbromide 58 in 7 steps and in 27 % overall yield, using as a key step the Shaipless asymmetric epoxidation of the allylic alcohol 61, as already described in section 2.3.2 and depicted on Fig. 21.
218 Figure 21
n-Ci2H25Br +
^
5 8
59
OH-v
C,2H25
^ '
6 0
C12H25
0
O 62
63
: 1) LiNH2, Et20, DMSO, 71 %; 2) Lindlar hydrogenation, 91 %; 3) /^r?-BuOOH, Ti(0/-Pr)4, L-(+)-diethy] tartrate, M. S., CH2CI2, 76 %; 4) (i) TsCl, DMAP, Et3N; (ii) Nal, acetone, 97 % (for the last two steps); 5) leru butylacetate, cyclohexylisopropylamine, /i-BuLi, HMPA, 81 %; 6) camphosulfonic acid, CH2CI2, 70 %;
4. 2. 8 Depezay's synthesis (72) The synthesis starts from (2/?, 3/?)-3,4-epoxy-l,2-0-methylidene butane 1,2-diol 150 prepared from D-isoascorbic acid 149 in 6 steps and 40 % overall yield (73). Nucleophilic opening of the epoxide with undecylmagnesium bromide in the presence of Li2CuCl4 led to the corresponding alcohol, which was protected as its 4-methoxybenzyl ether 151. Acetonide removal (AcOH, H2O), followed by intramolecular Mitsunobu reaction (PPh3, DEAD, 125 °C), gave the epoxide 152. Addition of diethyl malonate anion on this epoxide 152 led to a mixture of a-carbetoxy-ybutyrolactones 153, which upon treatment with magnesium chloride hexahydrate in dimethylacetamide, followed by deprotection of the alcohol through a DDQ oxidation led to the desired (-)-muricatacin 19. The synthesis was achieved in 12 steps from D-isoascorbic acid 149 and in 8.8 % overall yield (Fig. 22).
Figure 22 HO-I
HO/,.[ CizHzs
HO ^ 149
10
OH
OH
150
C12H25
151
E102C
0=^QXX^C,2H25
11,12
Reagents: 1-6) ref. 73; 7) CiiH23MgBr. Li2CuCl4, THF, -35 °C, 80 %; 8) NaH, DMF, imidazole, 20 °C, MPMCl, n-Bu4Nl, 20 °C, 93 %; 9) (i) ACOH/H2O 4/1, 20 °C; (ii) PhsP, DEAD, 125 X , 70-80 % for the last 2 steps; 10) CH2(C02Et)2, EtONa. EtOH, 60 °C; 11) Mga2.6H20, CH3CON(CH3)2; 12) DDQ, CH2CI2/H2O 18/1, 20 °C, 37 % for the last 3 steps.
219 5.
HEMI-SYNTHESIS OF ACETOGENINS OF ANNONACEAE 5. 1 Hemi-synthesis of solamin 1 (63) and reticulatacin 154 (14) Roblot et al. (63) were the first to report the hemi-synthesis of a natural acetogenin of
Annonaceae. They have shown that treatment of a natural precursor, diepomuricanin-A 121, by NaOH followed by acetylation with Ac20-pyridine afforded solamin diacetate 155 in 60 % yield. Unfortunately the composition of the expected diastereomeric mixture was not discussed. Indeed opening of the bis-epoxide system at either C-15 or C-20 would lead to two different compounds 155 and 156 with the same relative configuration but of opposite absolute configuration across the THE ring (Fig. 23). Figure 23 opening
/
opening
,kAi6i9A):20
tetraepi-solamin diacetate 156 OAc
16^
+
OAc
/19
solamin diacetate 155 In the same report (63), the authors have shown that treatment of epomuricenin A 122 with m-CPBA, according to Hoye (74) allowed the paitial synthesis of diepomuricanin-A 121. Again, no details were discussed concerning the expected epimeric mixture at C-19/20 of the products so obtained due to oxidation on either face "a" or "b" (Fig. 24). Figure 24
f35
epomuricenin A 122
-^
I m-CPBA
diepomuricanin A 121 + /?w£'p/-diepomuricanin A
220 Tarn et al. (14) have reported the hemi-synthesis of reticulatacin 154 from dieporeticanin 157, but by treatment with 70 % perchloric acid. Here again, two compounds 154 and 158 were expected with identical relative configuration, but with opposite absolute configuration across the THF ring, because of the opening arising either at C-17 or C-22 (Fig. 25).
dieporeticanin 157 70 % HCIO4, acetone OH
OH
1 8 ^ 2 1
r^rr^^pz-reticulatacin 158 +
OH
o^ I22. •"21
reticulatacin 154 Tam et al. also reported the hemi-synthesis of uieporeticanin 159 (14) and its bisepimer (at C23/24), by treatment with m-CPB A of a natural precursor, dieporeticenin 160 (Fig. 26). Figure 26
-fs?
dieporeticenin 160 m-CPB A
34
tiieporeticanin 159 + toe/?/-tiieporeticanin
5. 2 Hemi-synthesis of corrossolone 161 (9) Corrossolone 161 was isolated from Annoim muricata (48). Gromek et al (9) reported its hemi-synthesis as an diastereomeric mixture at C-15/16/19/20, starting from corepoxylone 5, after treatment with 70 % perchloric acid in acetone. Indeed it has been shown by HPLC that two products, 161 and 162, were obtained, having identical spectroscopic data (^H NMR, ^^C NMR, MS). These two compounds arise from the opening of tlie oxiranes at either C-15 or at C-20 position (Fig. 27).
221 Figure 27 opening
/
V O
OH
opening O
T
/
OH
5l^^o^^ I20. 16 V *j^9
tetraepi-corrossolouQ 162 OH 16^
+
OH j 19
coiTossolone 161 5. 3 Hemi-synthesis of isodeacetyl uvaricin 163 (77) Isodeacetyl uvaricin 163 was isolated from Uvaria narum in 1991 (75), and more recently from Annona bullata and re-named 4-deoxy asimicin (76)! In 1993, Sahpaz et al. {11) reported the hemi-synthesis of 163, starting from tripoxyrollin 164, which after treatment with perchloric acid (9) as already described in section 5.2., led to a complex mixture from which isodeacetyl uvaricin 163 could be characterized as well as other minor unnatural acetogenins (Fig. 28). Figure 28
tripoxyrollin 164
•^37
70 % HCIO4, acetone o ^<^.sO.
35 •;?37
OH T24
34 l 6•r\-,''"<^. ^-y_y23 isodeacetyl uvaiicin 163 + (C-15/16-19/20-23/24)-/7e.xo^/?/-isodeacetyl uvaricin -I- other minor bis-THF compounds
5. 4 Hemi-svnthesis of gigantecin 165 (79) Gigantecin 165 was isolated from Goniothalamus giganteus (78). Gu et al. reported its hemisynthesis (79) starting from gigantetronenin 166, also isolated from G. giganteus (80). The strategy used was based on the same sequence as that used earlier by Roblot (63) and Gromek (9) : namely epoxidation of the isolated double bond of gigantetronenin 166 with m-CPBA leading to two epoxides (due to oxidation on both faces of the unsaturation), followed by treatment with perchloric
222 acid to yield a mixture of the trans compound (38 %), namely gigantecin 165, and the cis compound (37 %), namely C-18/21 cis gigantecin 167 (Fig. 29). Figure 29
threo
gigantetronenin 166 Dm-CPBA 2) 70 % HCIO4, acetone trans threo
threo
1 A
threo
A = trans : gigantecin 165 A = cis : c/5-gigantecin 167
In the same way, gonionenin 168, isolated from Goniothalanius giganteus (79), when treated under the same reaction conditions (m-CPBA, and then perchloric acid), led to a 1:1 mixture of unknown acetogenins: trans -cyclogonionenin 169 and c/5-cyclogonionenin 170 (Fig. 30). Figure 30
U ) ^ 21 I threo trans , gogmonenm 168
1
1) m-CPBA 2) 70 % HCIO4, acetone
A = trans : rrj;75-cyclogognionenin 169 A = cis : d^-cyclogognionenin 170
223
6.
CONCLUSION The total synthesis of acetogenins of Annonaceae remains a challenging goal for synthetic
chemists. Indeed despite their relatively simple structures, these new natural compounds possess a large number of stereogenic centres (5 to 10), for which the stereocontroUed formations require many steps. Therefore, shorter and easier synthetic pathways are still needed for rapid, efficient, and scale up preparations. Furthermore, these compounds have already shown very promising biological properties, and structure-activity relationships have to be studied more elaborately, requiring the access to natural acetogenins and analogues in useful quantities. In the near future, the reported total syntheses of acetogenins of Annonaceae will probably increase considerably in the literature, as a result of creative and talented work of synthetic chemists. Acknowledgements: We wish to thank all our collaborators which appear in the cited articles. By their dedication and talented contribution to the study of acetogenins of Annonaceae, they have allowed us to explore and better understand these relatively new natural products possessing an intriguing large spectrum of biological properties. Special thanks are due to X. Franck for reading the manuscript.
224 REFERENCES 1
Leboeuf M., Cav6 A., Bhaumik P.K., Mukherjee B., Mukherjee R., Phytochemistry, 1982, 2i, 2783-2813.
2
Jolad S.D., Hoffmann J.J., Schram K.H., Cole J.R., Tempesta M.S., Kriek G.R, Bates R.B., J. Org. Chem., 1982,-^7, 3151-3153.
3 4a
Dabrah T. T., Sneden A. T., Phytochemistry, 1984, 23, 2013-2016 Fang X.P., Rieser M.J., Gu Z.M., Zhao G.X., McLaughlin J.L., Phytochem. Anal, 1993, 4, 27-48.
4b
Fang X.P., Rieser M.J., Gu Z.M., Zhao G.X., McLaughlin J.L., Phytochem. Anal, 1993, 4, 49-67.
5
Myint S.H., Cortes D., Laurens A., Hocquemiller R., Leboeuf M., Cav^ A., Cotte J., Qudro A.M., Phytochemistry, 1991, 30, 3335-3338.
6
Duret P., Gromek D., Hocquemiller R., Cortes D., Cav6 A., J. Nat. Prod., 1994, 57, 911916.
7
Cortes D., Myint S.H., Leboeuf M., Cav6 A., Tetrahedron Lett., 1991, 32, 6133-6134.
8
Gu Z. M., Fang X. P., Zeng L., McLaughlin J. L., Tetrahedron Lett., 1994,35, 5367-5368
9
Gromek D., Figadere B., Hocquemiller R., Cave A., Cortes D., Tetrahedron, 1993,49, 5247-5252.
10
Vu Thi T., Chaboche C , Figadere B., Chappe B., Bui Chi H., Cavd A., Tetrahedron Lett., 1994, 35, 883-886.
11
Rieser M.J., Kozlowski J.F., Wood K.V., McLaughlin J.L., Tetrahedron Lett., 1991, 32,
12
Cave A., Cortes D., Figadere B., Hocquemiller R., Laprevote O., Laurens A., Leboeuf M., in:
1137-1140. Recent advances in the acetogenins of Annonaceae, in Phytochemical Potential of Tropical Plants, K.R. Downum, J.T. Romeo, H.E.Stafford, (Ed.); Plenum Press, New York, 1993, pp 167-202. 13
Gromek D., Hocquemiller R., Cave A., Phytochem. Anal, 1994, 5, 133-140
14
Vu-Thi T., Phan-Quan C.-H., Chappe B., Roblot F., Laprevote O., Figadere B., Cave A. , Nat. Prod. Lett., 1994, 4, 255-262.
15
Duret P., Laurens A., Hocquemiller R., Cortes D. Cave A., Heterocycles, 1994, 39, n°2, 741-749.
16
Laprevote O., Girard C , Das B.C., Laurens A., Cave A., Analusis, 1993, 21, 207-210.
17a Laprevote O., Girard C , Das B.C., Cortes D., Cave A., Tetrahedron Lett., 1992, 33, 52315240. 17b Laprevote O., Girard C , Das B.C., Laugel T., Roblot F., Leboeuf M., Cave A., Rapid Commun. Mass Spectrom., 1992, 6, 352-355.
225
17c Lapr^vote O., Das B. C , Tetrahedron, 1994, 50, 8479-8490. 18
Harmange J.C., Figadere B., Cav6. A., Tetrahedron Lett., 1992, 39, 5749-5752.
19
Yamaguchi S., in : Asymmetric Synthesis; Morrison, J. D. (Ed.); Academic Press; 1983; Vol.
20
Potin D., Dumas F., d'Angelo J., J. Am. Chem. Soc, 1990,112, 3483-3486, and
1, pp 125-152. references cited herein. 21
Rieser M.J., Hui Y.H., Rupprecht J.K., Kozlowski J.F., Wood K.V., McLaughlin J.L., Hanson P.R., Zhuang Z., Hoye T.R., J. Am. Chem. Soc, 1992,114, 10203-10213.
22
Gu Z. M., Zeng L., Fang X. P., Colman-Saizarbitoria T., Huo M., McLaughlin J. L., J. Org. Chem., 1994, 59, 5162-5172.
23
Jolad S.D., Hoffmann J.J., Cole J.R., Barry C.E., Bates R.B., Linz G.S., Konig W.A., J. Nat. Prod., 1985, 48, 644-645.
24
Sahai M., Singh S., Singh M., Gupta Y. K., Akashi S., Yuji R., Hirayama K., Asaki H., Araya H., Hara N., Eguchi T., Kakinuma K., Fujimoto Y., Chem. Pharm. Bull., 1994,42, 1163-1174.
25
Rupprecht J.K., Hui Y.-H., McLaughlin J.L., 7. Nat. Prod., 1990, 53, 237-278.
26
Cave, A., Hocquemiller, R., Laprevote, O., 1989, F Patent 1048 N° 88 09 674.
27
Laurens A., Dutartre P., Cortes D., Hocquemiller R., Cave A., Immunomodulating Activity of Annonacin Isolated from A. muricata Seeds; Poster session, 18th lUPAC Symposium, Strasbourg, 30 August 1992.
28
Pr. Jeminet; personnal communication.
29
Londershausen M., Leicht W., Lieb F., Moeschler H., Pestic. ScL, 1991, 33, 427-438.
30
Harmange J.-C, Figadere B., Tetrahedron : Asymmetry, 1993, 4, 1711-1754.
31
Harmange J.-C, Figadere B., Cave A., Tetrahedron Lett., 1992, 33, 5749-5752.
32
Figadere B., Chaboche C , Peyrat J.-F., Cave A., Tetrahedron Lett., 1993, 34, 8093-8096.
33
Gaje J. B., Yu J. G., Khare A., Hu X. E., Ho D. K., Cassady J. M., Tetrahedron Lett., 1993, 34, 5847-5850.
34
Gaje J. B., Yu J. G., Khare A., Hu X. E., Ho D. K., Cassady J. M., Tetrahedron Lett., 1993,34,
5851-5854.
35
Hoppe R., Flasche M., Scharf H. D., Tetrahedron Lett., 1994, 35, 2873-2876.
36
Rama Rao A. V., Reddy K. L. N., Ashok Reddy K., Indian J. Chem., 1993, 32B, 12031208
37
Koert U., Wagner H., Pidun U., Chem. Ber., 1994, 127, 1447-1457.
38
Koert U., Wagner H., Stein M., Tetrahedron Lett., 1994, 35, 7629-7632.
39
Harmange J.-C, Figadere B., Hocquemiller H., Tetrahedron : Asymmetry, 1991, 2, 347350.
226 40
Hoye T. R., Humpal P. E., Jimenez J. I., Mayer M. J., Tan L., Ye Z., Tetrahedron Lett., 1994, J5, 7517-7520.
41
Hoye T. R., Hanson P. R., J. Org. Chem., 1991, 56, 5092-5095.
42
Figad^re B., Harmange J.-C, Hai X. H., Cave A., Tetrahedron Lett., 1992, 33, 5189-5192.
43
Coppola G., Schuster H. F., in: Asymmetric Synthesis: Construction of chiral molecules using
44
Figad^re B., Harmange J.-C, Laurens A., Cav6 A., Tetrahedron Lett., 1991, 32, 7539-
amino acids, Wiley, New York, 1987. 7542. 45
Figad^re B., Chaboche C , Franck X., Peyrat J.-F., Cav6 A., J. Org. Chem., 1994, 59, 7138-7141.
46
Harmange J.-C, PhD thesis, University Paris-Sud, December 1992.
47
Yao Z. J., Wu Y. L., Tetrahedron Lett., 1994, 35, 157-160.
48
Cortes D., Myint S. H., Laurens A., Hocquemiller H., Leboeuf M., Cave A., Can. J. Chem.,
1991,69, 8-11. 49
Bertrand P. Gesson J.-P., Tetrahedron Lett., 1992, 33, 5177-5180.
50
Autissier L., Bertrand P. Gesson J.-P., Renoux B., Tetrahedron Lett., 1994, 35, 3919-3922.
51
Freudenberg K., Dun W., von Hochstetter H., Chem. Ber., 1928, 61, 1735.
52
Sinha S. C , Keinan E., J. Am. Chem. Soc, 1993,115, 4891-4892.
53
Iwai K., Kosugi H., Uda H., Kawai M., Bull. Chem. Soc. Jpn., 1977, 50, 242.
54
Saad J. M., Hui Y. H., Rupprecht J. K., Anderson J. E., Kozolowsky J. F., Zhao G. X., Wood K. v., McLaughlin J. L., Tetrahedron, 1991, 47, 2751.
55
Makabe H., Tanaka A., Oritani T., J. Chem. Soc. Perkin Trans 1,1994, 1975-1981.
56
Makabe H., Tanaka A., Oritani T., Biosci. Biotech. Biochem., 1993, 56, 1028-1029.
57
Trost B. M., Shi Z., J. Am. Chem. Soc, 1994,116, 7459-7460.
58
Hoye T. R., Hanson P. R., Kovelesky A. C , Ocain T. D., Zhuang Z., J. Am. Chem. Soc,
59
Hoye T. R., Hanson P. R., Tetrahedron Lett., 1993, 34, 5043-5046.
60
Pettit G. R., Riesen R., Leet J. E., Polonsky J., Smith C R., Schmidt J. M., Dufresne C ,
1991,113, 9369-9371.
Schaufelberger D., Moretti C , Heterocycles, 1989, 28, 213-217. 61
Koert U., Tetrahedron Lett., 1994, 35, 2517-2520.
62
Pettit G. R., Cragg G. M., Polonsky J., Herald D. L., Goswami A., Smith C R., Moretti C , Schmidt J. M., Weisleden D., Can. J. Chem., 1987, 65, 1433-1435.
63
Roblot F., Laugel T., Leboeuf M., Cave A., Laprevote O., Phytochemistry, 1993, 34, 281285.
64
Scholtz G., Tochtermann W., Tetrahedron Lett., 1991, 32, 5535-5538.
65
Tochtermann W., Scholtz G., Bunte G., Wolff C , Petetrs E. M., Peters H. G., von Schnering, Liebigs Ann. Chem., 1992, 1069-1080.
227
66
Marshall J. A., Welmaker G. S., Synlett, 1992, 537, 538.
67
Marshall J. A., Welmaker G. S., J. Org. Chem., 1994, 59, 4122, 4125.
68
Wang Z. M., Zhang X. L., Sharpless K. B., Sinha S. C , Sinha-Bagchi A., Keinan E.,
69
Kang S. K., Cho H. S., Sim H. S., Kim B. K., J. Carbohydr. Chem., 1992,11, 807-812.
70
Saiah M., Bessodes M., Antonakis K., Tetrahedron Lett., 1993, 34, 1597-1598.
71
Liu Z. Y., Zhang J. J., Chen W., Chinese Chemical Letters, 1993, 4, 663-664.
72
Gravier-Pelletier C , Sani6re M., Charvet I., Le Merrer Y., Depezay J.-C, Tetrahedron Lett.,
Tetrahedron Lett., 1992, 33, 6407-6410.
1994,55, 115-118. 73
Gravier-Pelletier C , Dumas J., Le Merrer Y., Depezay J.-C, J. Carbohydr. Chem., 1992, 11, 969-998.
74
Hoye T. R., Suhadolnik J. C , Tetrahedron, 1986, 42, 2855.
75
Hisham A., Pieters L. A. C , Claeys M., Van den Heuvel H., Esmans E., Dommisse R.,
76
Hui Y. H., Wood K. V., McLaughlin J. L., Natural Toxins, 1992, 4-14.
77
Sahpaz S., Figadere B., Saez J., Hocquemiller R., Cave A., Cortes D., Nat. Prod. Lett.,
Vlietinck A. J., Phy to chemistry, 1991, 30, 1?>13.
1993,2, 301-308. 78
Alkofahi A., Rupprecht J. K., Liu Y. M., Chang C. J., Smith D. L;, McLaughlin J. L.,
79
Gu Z. M., Fang X. P., Zeng L., Song R., Ng J. H., Wood K. V., Smith D. L., McLaughlin
Experentia, 1990, 46, 539-541. J. L., J. Org. Chem., 1994, 59, 3472-3479. 80
Fang X. R., Anderson J. E., Smith D. L., McLaughlin J. L., Wood K. V., J. Nat. Prod., 1992,55,
1655-1663.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 O 1996 Elsevier Science B.V. All rights reserved.
229
The Synthesis of Nonactic Acid. Its Derivatives and Nonactin itself Ian Fleming and Sunil K. Ghosh 1.
INTRODUCTION The actin antibiotics, of which several are represented by the structures 1-9, occur in various
Streptomyces species (1,2). The structures were elucidated as 32-menibered macrocyclic tetraesters. Nonactin itself was isolated in 1955 by Prelog and his co-workers (1) as the first member of the family, and is its lowest homologue and most symmetrical member. All members of the actin family exhibit antimicrobial and antifungal activity, which increase with increasing size of R, and insecticidal activity against mites (3).
0=^
H
\ „^2
H I
^••H
H"/ 0
L /^ W"H
HMV
=0
H„ H
Y XT 1 2 3 4 5 6 7 8 9
nonactin monactin dinactin trinactin tetranactin macrotetralide G macrotetralidc D macrotetralide C macrotetralide B
R' Me Me Me Me Et Et Et Et Et
R3
R2
R3
Me Me Et Et Et Me Me Me Pr*
Me Me Me El Et FV Et Pi^
Et
R^ Me Et Et Et Et Me Pr' Pr* Pr'
It was the first naturally occurring crown ether, and the earliest for which the antibiotic activity could be traced to its ionophoric properties (4). The actins give one-to-one complexes with many alkali and alkaline earth metal ions, with nonactin showing selectivity in the order NH4+ > K+ = Rb+ > Cs+ > Na+ > Ba2+ (5). The first X-ray crystal structure of an ionophore-metal complex was obtained from the nonactin-potassium thiocyanate complex (6). The potassium is bound by coordination to all four tetrahydrofuran oxygens and to the four ester carbonyl oxygens, which creates overall a "tennis ball seam" conformation to the carbon framework as it wraps around the metal. Nonactin is composed of two subunits of the (-i-)-nonactic acid 10 and two subunits of the (-)nonactic acid 12, joined together by lactone linkages in an alternating sequence to give overall the
230 meso configuration (S4 symmetry) (7). The name nonacnn came from its having no optical acrivity, a puzzling feature until the meso structure was deduced. A satisfactory synthesis requires separately each of the enantiomeric nonactic acids, followed by coupling in an alternating sequence, and ring closure. A number of syntheses of nonactic acid and its esters, both racemic and enantiomerically enriched, have been developed over the past twenty years, and four syntheses of the natural product itself. The literature up to 1980 has been thoroughly reviewed (8). OH
.A^oWcOzR H 10 R=H 11 R=Me
(4-)-nonactic acid methyl (+)-nonactate
12 R=H 13 R=Me
H=
(-)-nonactic acid methyl (-)-nonactate
yKA^\^^0,^ H^H_
14 15
2.
R=H R=Me
(+)-8-epinonactic acid methyl (+)-8-epinonactate
16 R=H 17 R=Me
(-)-8-epinonactic acid methyl (-)-8-epinonactale
THE SYNTHESIS OF NONACTIC ACID AND ITS DERIVATIVES The major challenge in the synthesis of nonactic acid is the control of the relative
stereochemistry between the four stereogenic centres, which have 1,2- and
1,3-acyclic
stereochemical relationships (C-2 to C-3 and C-6 to C-8, respectively) and a 1,4-cyclic relationship (C-3 to C-6) on the tetrahydrofuran ring. These challenges have made nonactic acid a favourite target of synthetic chemists intent upon proving the capacity of a method of stereocontrol they have developed. There is much overlap from route to route, as people have often used their new method to set up only one or two of the stereochemical relationships, and have then been content to complete the synthesis using established routes. We distinguish seven common themes among the approaches used for the synthesis of nonactic acid, and have classified them by the methods, numbered 1-7 in Scheme 1, used for the construction of the a.s-2,5-disubstituted tetrahydrofuran derivative 18 (9). These are: 1. Hydrogenation of a 2,5-disubstituted furan. 2. Hydrogenation of the Bartlett type intermediate. 3. Cyclisation of a 1,4-diol derivative. 4. Electrophilic cyclisation of an unsaturated alcohol or enol. 5. Intramolecular conjugate addition of an alkoxide with R^ electron-withdrawing. 6. From a bicyclic intermediate. 7. Ireland-Claisen rearrangement.
The first two approaches, 1 and 2, are based on 5)'«-stereospecific catalytic hydrogenation to control the stereochemistry around the tetrahydrofuran ring. The approach 1 has the disadvantage
231 that it only controls the cis stereochemistry of C-3 and C-6. The approach 2, which shows control from C-6 to the newly developing centre at C-3, as well as being syn stereospecific in creating the centres C-2 and C-3, is popular, and has been applied especially by those research groups who were principally demonstrating a method of 1,3 control for the relationship between C-6 and C-8. Approach 3 is challenging, since it requires a method for controlling the 1,4 relationship between C3 and C-6 in an open chain. Approaches 4 and 5 have not proved easy for setting up C-3 and C-6 with the cis arrangement, in contrast to the approach 6, which is effective for this part of the problem, but has difficulties with the control of C-2 and C-8. The approach 7 has only been used once, with C-2, C-3 and C-6 well controlled by the stereospecificity of the Ireland-Claisen rearrangement, but lacks control at C-8.
^•J3
^'^XK^'
RUCX'^' ftO
RL ^
>k .R^
H
H HO OH
18
= OH
Scheme 1
2.1 Hydrogenation of Substituted Furan Derivatives Although one of the earliest approaches to the synthesis of nonactic acid involved catalytic hydrogenation of an appropriately 2,5-disubstituted furan, the hydrogenation of furan derivatives is limited by the formidable problem of controlling the well separated acyclic stereocentres at C-2 and C-8. The first synthesis of racemic nonactic acid diastereoisomers was carried out by Beck and Henseleit (10) in 1971 (Scheme 2). Methylation of 2-methoxycarbonylmethylfuran 19 gave a mixture of the a-methyl derivative 20 (43%) and the a,a'-dimethyl derivative, which were separated by chromatography. Friedel-Crafts alkylation of 20 with 3-methyl-3-butene-2-one gave the ketofuran 2 1 , which on hydrogenation over rhodium on alumina gave a mixture of the diastereoisomeric d5-2,5-disubstituted tetrahydrofurans 22. Baeyer-Villiger oxidation of the ketone 22 with trifluoroperoxyacetic acid regioselectively gave the ester 23, and hydrolysis followed by esterification gave a mixture of diastereoisomeric nonactic acid methyl esters 24 in 70% yield. The
232 stereochemical control of C-3 and C-6 during hydrogenation was good, but no stereochemical control was observed for C-2 and C-8.
a.
l.NaH.THF C02Me
C02Me 2. Mel
•
19
20 43%
^
r \ 0
^
I
6
BF3.0Et2
20
^
r \
11 ^
^
60%
CF3CO3H, Na2HP04
H2,Rh/Ai203 ^
89%
^
^r-\
J
C02Me
H^ H
21
22 l.KOH 2. HsO-"
9^^
COiMe
73%
H ^ H T
3.CH2N2 70%
23
Scheme 2 Gerlach and Wetter (11) achieved rather better levels of control, also using catalytic hydrogenation for the relative stereochemistry between C-3 and C-6 (Scheme 3). 2-Furylacetone 25 was prepared by a Darzens glycidic ester synthesis from furfural, and was functionalised at the 5position using Eschenmoser's 2-chloro-N-cyclohexyl-propionaldonitrone 26 in the presence of
l.NaOMe,
^N^SS/
MeCHClC02Et
ci
«.;J0
^ 26
OHC
o
3.H2SO4
o
AgBF4
27
25
43% HCl, H2O
l.Cr03,H2S04
9
F~^
H2,Rh
CHO 2. CH2N2 27% (from 25)
NaOMe,MeOH NaBH4
C02Me
O 29
30:31 = 1:4
OH
/—V
OH C02Me ^
CO2MC
98% 32
32:8-epi-32 = 7:10
8-epi-32
Scheme 3 silver tetrafluoroborate (12). The nitrone 27 was hydrolysed to give the aldehyde 28, which was oxidised with chromic acid and esterified to give the ketoester 29 in poor yield. Hydrogenation of
233 the furan ring in 29 over rhodium produced a mixture, diastereoisomeric at C-2, of tetrahydrofurans 30 and 31, which were separated by chromatography. Base-catalysed epimerisation at C-2 favoured (80:20) the natural threo relationship between C-2 and C-3, and the unwanted isomer 30, on equilibration with a catalytic amount of sodium methoxide, could be converted into its diastereoisomer 31. The reduction of the ketone 31 with sodium borohydride took place with poor selectivity, unfavourable to the natural product (7:10), but the methyl nonactate 32 and methyl 8epinonactate could be separated chromatographically. Schmidt and co-workers (13) also synthesised nonactic acid by hydrogenation of a 2,5disubstituted furan derivative (Scheme 4), which they initially carried out in the racemic series. The reaction of 2-lithiofuran with propylene oxide gave the alcohol, which was converted to the acetate 33. The introduction of the three carbon carboxylic acid unit was achieved by Vilsmeier formylation and a Wittig reaction on the aldehyde 34, followed by carbonylation of the alkene 35, catalysed by a complex of |i-dichloro-bis(7i-hexa-l,5-diene)dirhodium(I) with triphenylphosphine. Oxidation of the resulting aldehyde 36 with silver(I) oxide, followed by esterification gave the ester 37. As with the earlier syntheses, they observed poor control for C-2 and C-8 during hydrogenation of 37 over rhodium on alumina, which produced methyl nonactate as a mixture of diastereoisomers 38, 39, 40 and 41 in a ratio of 25:25:25:25. They examined the possibility of equilibrating these diastereo-
Ph3P=CH2 L'
O
2.AC20
O
oocw.
33
Rha)cat.
g,^^
O
57%
34
9^^
f-^
l.Ag20 CHO
/x,^^^4N^^C02Me u ^ u =
^
CHO
9"
C02Me
+
42
39
OH
I—V
X^^C02MQ
^ ' ' ^ V ^ p . ^ v ^ CO2MC
Cr(VI)
H2, Rh/Al203
f-\
• ^ ^ ^ ^
H2,Ni
^
O
I—V
C02Me
H^ H I
H^ H = 43
41
l.TsCl 2. KOAc 39 -I- 40 + 41
-• 3. KOH, MeOH
8-epi-39 + 8-epi-40 + 8-epi-41
4. CH2N2
separate
C02Me 44
Scheme 4 isomers, first by oxidising the 8-hydroxy group to the 8-keto, and regenerating the 8-hydroxy group by catalytic hydrogenation, which gave the mixture of 3 8 , 3 9 , 40 and 41 now in a ratio of
234 19.1:10.8:30.5:39.6. Secondly they found that a Walden inversion sequence could be used to change the relative configuration at C-8. Still in the racemic series, they found that the 8-tosylates gave the C-8 diastereoisomeric acetates on Walden inversion, and hydrolysis gave the diastereoisomers 38,39, 40 and 41 in a ratio of 36.9:32.7:9.6:20.8. The ratio of 8-normal to 8-epi i.e. 38+39 : 40+41 was about 7:3. The racemic methyl nonactate 38 could be separated from the other diastereoisomers by chromatography. Methyl 2-epinonactate 39 could be equilibrated with sodium methoxide to a mixture of methyl nonactate 38 and methyl 2-epinonactate 39 in a ratio of 60:40. Following this exploratory work, the Schmidt group (13) reported the first homochiral synthesis of methyl (-)- and (+)-nonactate using (5)-propylene oxide as the only enantiomerically pure starting material (Scheme 4). This gave the mixture of diastereoisomeric methyl nonactates 38, 39,40 and 41, all with the (5)-configuration at C-8, and chromatographic separation provided 25% of the natural (-) ester 38. The mixture of the other three diastereoisomers (39+40+41) was converted by Walden inversion into a mixture of the same three compounds with C-8 inverted, from which (+)-methyl (25',35,67?,8/?)-nonactate 44 could be isolated in approximately the same amount as its enantiomer 38. White and his co-workers (14) used a Friedel-Crafts acylation of l-(2-furyl)-2-propanol 45, prepared using the same chemistry as Schmidt's (13a), gave the 2,5-disubstituted furan 46 (Scheme 5). Hydrogenation of the furan 46 over rhodium resulted in the c/^-fused tetrahydrofuran alcohol ?"
/"I
Ac20,BF3.0Et2
OAc
OAc
45
2. NaOH, MeOH 52%
2. CrOs, H^ 3. BF3, MeOH 89%
49
^AXV^C
CO2H
l.Ph3P,Et02CN=NC02Et,PhC02H
CO2H
9"
,
^ C02Me
2.NaOMe,McOH 90%
Scheme 5 47, which was oxidised with chromic acid to the diastereoisomeric mixture of ketones 48. The last carbon was introduced by a Wittig reaction on the ketone 48 with methylenetriphenylphosphorane,
235
and the correct oxidation state was obtained by hydroboration, followed by Jones oxidation and esterification, to give the keto esters 50 and 5 1 . The hydroboration took place with high regioselectivity and moderate stereoselectivity (2:1) in favour of the unwanted diastereoisomer 51. The diastereoisomers 50 and 51 were separated, and L-Selectride reduction of the derived acid 52 gave again largely the wrong diastereoisomer, 8-epinonactic acid 54, but the configuration at C-8 could be corrected on the derived ester 55 using a Mitsunobu reaction to give methyl nonactate 56. Still's synthesis (15) (Scheme 6) is unique in having excellent stereocontrol over the larger distances, C-2 to C-8 and C-3 to C-6, but is let down by not solving the problem of relating the C-2 and C-8 pair to the C-3 and C-6 pair. Based on the conformational preference of macrocyclic intermediates (16), the stereogenic centre at C-8 provided stereocontrol for enolate methylation at C2, beautifully solving the problem that most besets the approach using the hydrogenation of a furan 1. Pb(0Ac)4, BF3 2. NaOH, McOH
r^
,
59
Corey-Nicolaou O
""
0
-=.
CO2H
OH
,—(^ ^—1
C °/ =o^ ^i 0. 25-40% •* "O
1. base 2.Md
62
Scheme 6 to establish the relative stereochemistry of C-3 and C-6. The macrocyclic intermediates were constructed using the 2,5-disubstituted furan 59 attached to various spacer groups as in the hydroxy acid 60. Macrolactonisation using the Corey-Nicolaou thiopyridyl ester method gave the lactone 61 having a p-xylyl spacer, which, of those tried, proved to give the best results. Methylation of this lactone gave the diastereoisomer 62 with excellent selectivity (>70:1). Lithium aluminium hydride reduction followed by catalytic hydrogenation provided a 1:1 mixture of the reduced nonactic acid derivative 63 and its diastereoisomer 64. 2.2 Hydrogenation of a Bartlett Type Intermediate The Bartlett approach is based on the hydrogenation of a double bond exocyclic to a tetrahydrofuran ring, simultaneously establishing the desired stereochemical relationship between C2 and C-3, because of the syn stereospecificity of the reaction, and between C-3 and C-6, because hydrogenation takes place stereoselectively on the less hindered face, anti to the side chain on C-6.
236
This approach is very popular because all that is left to control is the 1,3-diol stereochemistry between C-6 and C-8. In 1977, Bartlett and Jemstedt (17) developed a stereocontrolled synthesis of 1,3-diols from homoallylic alcohols (Scheme 7). They found that diethyl 4-penten-2-yl phosphate 65 undergoes an intramolecular cyclisation in the presence of iodine to set up both groups in the ring 66 cis and presumably diequatorial. Arbuzov reaction then gives the cyclic phosphate 67 more than 90% diastereochemically pure. EtO OEt
EiO O
OEt I2. MeCN, 2 5 X
87%
•V-^ r
-EU
EtO
o>"-o 67
66
65
Scheme 7 In their first and racemic synthesis of methyl nonactate 75 (Scheme 8), Bartlett and Jernstedt (18) introduced the 1,3-relationship between C-6 and C-8 in a diastereoselective manner from the diene 68, which gave the analogous cyclic phosphate. Upon treatment with sodium methoxide, the phosphate gave the epoxide 69, which was converted into the syn-l,3-d\o\ 70. Acetylation of the diol followed by ozonolysis gave the aldehyde 71. Aldol condensation of the aldehyde with the silyl
2. NaOMe, THF
1. Ac20,Py 2.03,CH2Cl2
^
^
^
1. MeCH=C(0Me)0SiMe3 TiCl4, CH2CI2, -78°C
OAc OAc
^.
CHO
3. Me2S
90%
OAc OAc CO2MC
2. CrOs, Me2C0, H2SO4
95%
71
l.K2C03,MeOH
H2, Rh/Al203
C02Me 2. (C02H)2
CO2MC
87%
80%
icis:trans = ^5:\5)
l.Et02CN=NC02Et PPh3, PhC02H, THF
^
C02Me
2. NaOMe, MeOH 95%
Scheme 8 enol ether of methyl propanoate in the presence of titanium tetrachloride at -78 °C followed by Jones oxidation of the aldol product gave the p-keto ester 72. The tetrahydrofuran ring was constructed by acetate cleavage and dehydration with oxalic acid giving methyl (E)-8-epi-2,3-dehydrononactate 73 as a single geometrical isomer. The C-2 and C-3 stereocentres of methyl nonactate were then
237 generated stereospecifically by hydrogenation of the double bond using rhodium on alumina, and at the same time the C-3 and C-6 centres were produced selectively (cisitrans = 85:15). Finally, the C-8 centre in the 8-epinonactate 74 was corrected through a Mitsunobu inversion-and-hydrolysis procedure to give methyl nonactate 75. In continuation of this work on acyclic stereocontrol, Bartlett and his co-workers developed a better method for the stereocontrol led synthesis of 1,3-diol derivatives, finding that a rerr-butyl carbonate group was better than the phosphate group in iodocyclisation (19), and applied this technique in an enantiodivergent synthesis (20) of both nonactic acids using the enantiomerically pure carbonate 84, synthesised by the route shown in Scheme 9. Starting in the chiral pool, dimethyl (5')-(-)-malate 76 was converted successively to the diol 77, the epoxide 78 and the dienol 79. Iodocyclisation of the r^rr-butyl carbonate 80 took place with good diastereoselectivity (cis:trans = 6.5:1) to give the cis cyclic carbonate 81, which was reduced with tributylstannane to give the carbonate 82. Ozonolysis gave the aldehyde 83, and the same sequence of steps as in the earlier, racemic synthesis (Scheme 8) then gave the p-keto ester 84, which was the common intermediate for both enantiomers of nonactic acid.
Me02C
OH A^COsMe
l.TsCl,Py 2. PPTS 3. K2C(^
OEt 1. Ll ^ .PPTS 1.
O^
l.Tia4,CH2Cl2,-78X _—.
,
1^
2. MeCH=C(0Me)0SiMe3 3. C1O3, H2SO4 75%
84 Scheme 9
Methanolysis of the carbonate protecting group in 84 (Scheme 10) followed by oxalic acidcatalysed dehydration gave the (+)-enantiomer, (65,8/?)-(E)-2,3-dehydrononactate 85, hydrogenation of which gave the (-) enantiomer 86 of methyl 8-epinonactate as the major product (88%). For the synthesis of methyl (+)-nonactate 88 from the common intermediate 84, the p-keto ester group was used in a nucleophilic attack to invert the configuration at C-6, with the cyclic carbonate group acting as the leaving group. The dehydrononactate 87 thus produced was then converted mostly (88%) to methyl (+)-nonactate 88 by hydrogenation. The methyl (-)-8-
238 epinonactate 86 and methyl (+)-nonactate 88 were used for the synthesis of nonactin itself (Section 3.2). O OH
l.K2C03,MeOH
C02Me
^ C02Me 2. (C02H)2 84 O 76% 1 KH,HMPA.THF 92% C-6 inversion
85
I H2, Rh/Al203
100%
C02Me C02Me lH2,Rh/Al203
H^'HE 86
100%
C02Me
Scheme 10 Johnson and his co-workers developed two routes for the diol 95 corresponding to the carbonate 82 of Bartlett's synthesis. In their first approach (Scheme 11) (21), they coupled the acetal 89, derived from (/?,/?)-2,4-pentanediol and 4-pentenal, with a-trimethylsilylacetone or the trimethylsilyl enol ether of acetone in the presence of titanium tetrachloride. The reaction took place
OSiMes ^/^^^
O
or
Q O ^Q
TiCl4, CH2CI2
-^
o
o
89% or 93% 90
A
90:91=97:3
OTBDMS
91
OTBDMS
iiiiiL
1. TBDMSCl, imidazole
O
9
L-Selectride OH
9
2. HPLC 92 l.Ac20,Py 2. BU4NF 3. PCC 90%
,„„l^ OAc 9 x^\.'^' 94
93 + 8-epi-93 4:l l.base 2. deacetylation
vx''^^^ + 8-epi-94
OH OH
3 chromatography 75%
95
Scheme 11 with high diastereoselectivity (90:91 = 97:3). The rerr-butyldimethylsilyl ether 92 of the major diastereoisomer was reduced with L-Selectride to give a 4:1 mixture of syn 93 and anti (8-epi-93, nonactic acid numbering) isomers, which was quantitatively converted to the corresponding acetates.
239 Desilylation with tetrabutylammonium fluoride gave a mixture of hydroxy acetates, which were oxidised with pyridinium chlorochromate to the ketoacetates 94. p-Elimination under basic conditions and deacetylation gave the mixture of diols 95 and 8-epi-95, which were separated by chromatography. In their second approach (Scheme 12), Johnson and his co-workers used the acetal 96 derived from (35)-butane-l,3-diol instead of from (/?,/?)-2,4-pentanediol (22). As in the earlier work, the Mukaiyama aldol condensation of acetal 96 with acetone trimethylsilyl ether gave a mixture of diastereoisomers 97 and 98 in a ratio of 96:4. In contrast, however, the removal of the chiral auxiliary was achieved without the need for first reducing the ketone group. Oxidation of the aldol OH
^ 97 O
l.PCC
OH
2. Bn2NH2'' CF3CO2" 82%
97:98 = 96:4
n-Bu3B, NaBH4
OH OH
78%
99
98
100 + 2-epi-100 97:3
Scheme 12 product 97 to the aldehyde allowed a selective p-elimination with dibenzylammonium trifluoroacetate. 5j^z-stereoselective reduction of the free aldol product 99 by Narasaka's method (23) using tributylborane and sodium borohydride gave a 97:3 mixture rich in the (2/?,45)-oct-7-ene2,4-diol 100, the diol intermediate in Bartlett's synthesis. Lygo and his co-workers (24) have demonstrated that ether substituents a or p to an epoxide remain unaffected during the reaction of an epoxide with p-ketoester dianions yielding e-hydroxy-pketoesters. These intermediates can be easily cyclised to the cycHc enol ethers needed for the Bartlett O- O"
OMe l.NaH,BnBr 2. MCPBA
102 OBn
l.H2.Pd/C C02Me 104
2.H2,Rh/Al203 95%
54% from 101
OBn OH 103
101 OBn
(C02H)2
CO2MC
OH I—y
OH
^ T ^ H T 105
I—^
-^^-^T^nO^-^ H
1:1
CO2MC
H
106
Scheme 13 approach. Thus the epoxy ether 101 and the p-ketoester dianion 102 gave in one pot the (9-benzylprotected dehydrononactate intermediate 104 as a mixture of diastereoisomers (Scheme 13).
240
Hydrogenolysis using palladium on charcoal, followed by stereoselective hydrogenation of the double bond using Bartlett's conditions gave a 1:1 mixture of racemic methyl nonactate 105 and methyl 8-epinonactate 106, which were separated by chromatography. Lygo (25) has also developed a strategy for the synthesis of nonactic acids and its homologues, using the same chemistry, but starting from 3-butenol 107 instead (Scheme 14). Benzylation, epoxidation and regioselective epoxide opening with the same 3-ketoester dianion 102 gave the intermediate p-ketoester 109. Oxalic acid-catalysed dehydration and cyclisation gave the Bartlett-type intermediate 110. Hydrogenolysis and stereoselective hydrogenation gave the primary alcohol 111, which has all the stereocentres properly established except for C-8, and is suitable for the synthesis both of nonactic acid and its homologues. Oxidation of the alcohol using pyridinesulfur trioxide in dimethylsulfoxide and triethylamine gave the aldehyde 112 in 86% yield. 0" O"
l.NaH.BnBr OH
^
2.MCPBA
^
102
.C02Me
0B„
92%
107
109
1^«
OBn I—. • 110
57% from 108
OBn OH
,ft«
(C02H)2
l.H2,Pd/C 9 "
/—V
Rh/AljOa 93%(d5:fran5 = 8:l)
^^l
SO3, DMSO =
/—\
86%
' 112
Scheme 14 This aldehyde had already been converted (26) to methyl nonactate and methyl 8-epinonactate with high selectivity using titanium tetrachloride catalysed addition of dimethyl zinc and lithium dimethylcuprate, respectively. Lygo also found that dialkylzinc addition under different Lewis acid conditions gave different diastereoisomers with high selectivity (Scheme 15).
0«^vxt^V^C02Me H'^HE
112
• RA.^/3Wc02Me ^ ^ H^Hi 113 Et2Zn,TiCl4 R=Et 4:1 Et2Zn, BF3.0Et2 R=Et 1:10 Me2Zn,TiCl4 R=Me 24:1 Me2CuLi R=Me 1:4.5
A,X\^C02Mc H^Hi 114 85% 80% 85%^^ 78%^^
Scheme 15 Deschenaux and Jacot-Guillarmod synthesised the Bartlett intermediate 124 with a chiral pool source and the Narasaka 1,3-diol synthesis for the C-6 to C-8 relationship (Scheme 16) (27). Methyl (-)-(3/?)-3-hydroxybutanoate 115 (e.e. >99%) was converted to the p-ketoester 116. Reduction of the p-ketoalcohol using the Prasad version of the Narasaka method (23) with diethylmethoxyborane and sodium borohydride giving the diol 117. The acetonide 118 was homologated by way of the
241 alcohol 119 and tosylate 120 to give the nitrile 121 and hence the carboxylic acid 122. The imidazolide of the acid and the magnesium salt of monomethyl malonate gave the p-ketoester 123. Acid catalysed deprotection of the acetonide and concomitant cyclisation of the diol intermediate gave OLi
nS(99%ee)
''''
„6
y^ 9
''"^
y 9
Li^^H4
^^^'-^^^ 118
9
l.TsCl,Py 9 ? ^
96% ^ ^ " V - ^ ^ v , ^ 119
3.NaOH
0
c T
Amberlyst-15 OH C02Me ^^^
j—y
9
9 ^
l.CO(Im)2
^"^.-'-'^^^ 2. COiMe 120 X=OTs 93% — ( 121 X=CN 80% C02)2Mg 122 X=C02H 70% 88% H2, OH .—y
-- --^1 o ^
123
^^^^
y
2.BU4NCN
W 0
in
Rh/Al203
124 (95% ee)
H^ H Y
95%
125
Scheme 16 Bartlett's intermediate 124, and catalytic hydrogenation over rhodium on alumina gave methyl (-)-8epinonactate 125. Kim and Lee have developed syntheses of methyl (+)-nonactate 134 and of methyl (-)-8epinonactate 135 by way of Bartlett's intermediate 133 (=82) (Scheme 17) (28). They used a
MO, iy^2
X\ l.TMSCI,Et3N ^-O TBDMSCl, N-0 . x ^ W 2.TsOH // \ nw imidazole // ' ' ^*S ^ • separation O O2 4. L-Selecuide 126 63% H2,Raney-Ni, o OH Et2B0Me, NaBH4 OH OH boric acid ^ l ^ I ^ o - m n x . . THF.-78°^ , A J k . O T B D M S '• Me^CCOMeh.PPTS 86% 86% 2. BU4NF 129 130 synianti = %:4 „_^
^..-v
^y^ I I
^„
l.TsCl,Et3N,DMAP 2. allylMgCl. Cul
^^^^ I I
3.TsOH 132
am)2C0
133
OH y ^ \ I \ n^\\^^ and x ^ ^ x x f ^ ^ ^ f v ' ^ H ^ HE 135 Scheme 17 dipolar cycloaddition to an acrylic acid group attached to Oppolzer's chiral auxiliary 126 to set up the Bartlett routes
OH \
^
o / /o
73%
131 75% y V I \
Q
^A^ I I
r>c\ XK ^C02Me
first stereogenic centre (29), and Narasaka's method again for the C-6 to C-8 relationship. The syn
242
1,3-diol 130 was synthesised from the isoxazoline 127 after protection of the hydroxy! group with r^rr-butyldimethylsilyl chloride by reductive cleavage with the Raney-nickel-boric acid system and the Narasaka reduction (23) of the p-ketoalcohol intermediate 129. The 1,3-diol 130 was converted to the diol 132 by standard synthetic transformations involving protection of the diol as its acetonide, and deprotection of the 0-silyl group to give the primary alcohol 131. Tosylation, copper-catalysed tosylate displacement with allylmagnesium chloride, and deprotection gave the diol 132, which was converted to its cyclic carbonate 133 (=82) using l,r-carbonyldiimidazole. The rest of the synthesis intersects with that of Bartlett and his co-workers (20). Kim and Lee also developed another enantioselective synthesis of methyl (-)-8-epinonactate (Scheme 18) (30) based on the same nitrile oxide cycloaddition product 127 used in their earlier synthesis. The enantiomerically enriched alcohol 127 (>98%e.e.e.) was converted to its iodide l.PPh3,12. iniidazole O" 0 " 2. < A ^ o M e N-0
N-0
3.CuI 127
H^ 137
I 136 °
59%
I
1. chromatography 2. L-Selectride ^ 3. chromatography 75%
i
2. PCC 75%
1. H2,Raney-Ni, ^"^^^
2.(C02H)2 77%
H^ H= 138 " (major:minor = 91:9)
QT, V I \ /k^O-K/C02Me H H= 139
Scheme 18 using triphenylphosphine, iodine and imidazole, and the iodide treated with the dianion of methyl 2methylacetoacetate to give the p-ketoester 136. Reductive cleavage of the isoxazoline ring followed by oxalic acid-catalysed cyclisation gave the ketone 137 corresponding to Bartlett's intermediate. Rhodium catalysed hydrogenation followed by PCC oxidation provided the ketone 138 where the C-2, C-3 and C-6 centres are correctly established. The last centre at C-8 was regenerated by a stereoselective reduction of the ketone 138 with L-Selectride (14), which provided the methyl (-)-8epinonactate 139. Barrett and Sheth synthesised rerr-butyl (±)-8-0-rerr-butyldimethylsilylnonactate 145 by a stereoselective hydrogenation of 8-0-t-butyldimethylsilyldehydrononactate 144, the 8-epimer and 80-protected analogue of Bartlett's intermediate, and solved the C-6 to C-8 problem in a completely different way by another hydrogenation (Scheme 19) (31). 2,3,5-0-Triacetyl-D-ribonolactone produced the achiral diene 140 on treatment with l,8-diazabicyclo[5.4.0]undec-7-ene (DBU). Hydrogenation of the diene stereoselectively (>97:3) gave the lactone 141, establishing the relative stereochemistry at C-6 and C-8. Diisobutylaluminium hydride reduction of the lactone gave the lactol 142. Wittig reaction, followed by catalytic hydrogenation over rhodium on alumina gave a lactone,
243
and protection of the free hydroxy group gave its silyl ether 143. Claisen condensation and dehydration gave the dehydrononactate intermediate 144, and the usual catalytic hydrogenation over rhodium on alumina then provided the O-silyl derivative 145 of racemic ten-butyl nonactate. AcO
OAc
OAc
DBU
AcO
DIBALH
O 140 l.Ph3P=CHC02Et 2. H2, Rh/Al203 TBDMSO
jP"
-AX.
85%
X ^ Q - ^ OH 142
—'A Pd/CaCOs
94%
3. TBDMSCl 55%
90%
O 141 OLi
2. Ambcrlilc 120 h 75%
143
OTBS
_/^^''
OTBS CO2BU'
H2,Rh/Al203 89%
144
{cis:trans - ^5:\5)
145
Scheme 19 Sutherland and his co-workers developed a concise route for the synthesis of Barrett's intermediate 148, also using hydrogenation of a cyclic intermediate to establish the C-6 and C-8 OH
1. H2,Rh/Al203 O 2. PhCOCl, Py ,x^,>^^^
NaOMe, MeOH
MCPBA
^^
3. Cr03, H-'
93% OBz
/'^-S^Q'^O
H
BzO
148 147
146
Scheme 20 relationship (Scheme 20) (32). Catalytic hydrogenation of methyl hydroquinone, benzoylation and chromic acid oxidation gave the c/5-2,4-disubstituted cyclohexanone 146. Baeyer-Villiger oxidation
^^.^ss^CHO
L-(+)-diisopropyl tariaratc
+ ClMg.
^. 149 (£:Z = 96:4) OH OH
45% conversion OAc OAc
150 RuCl3,NaI04
96% ee K2CO3, MeOH
9^' 9^'
x-^v^^>^0
CO2H
86%
H 154
153
Scheme 21 of this ketone took place regioselectively to give the lactone 147, which on methanolysis gave the racemic lactone 148. Sutherland and co-workers also developed an enantioselective route (Scheme
244 21) (32), based on Sharpless asymmetric epoxidation (33), for the synthesis of Barrett's intermediate 154. This synthesis of Barrett's intermediate also involves the synthesis of the diol 151, a diastereoisomer of the diol 70 in Bartlett's nonactic acid synthesis. 3-Butenylmagnesium bromide and crotonaldehyde gave the diene alcohol 149. The diol 151 was then made in 96% e.e. by kinetic resolution using Sharpless epoxidation, followed by reduction of the epoxide with red-Al. The diacetate 152 obtained from the diol was oxidised with ruthenium trichloride-sodium periodate (34) to give the acid 153, which was converted to Barrett's lactone 154, but now in homochiral form. Batmangherlich and Davidson developed an enantiodivergent route to both enantiomers of tenbutyl nonactate by way of the lactone 156, with C-6 obtained from the a-C of glutamic acid 155 (Scheme 22) (35). The hydroxylactone 156 was protected as its rerr-butyldimethylsilyl ether, which steps
silyl protection
HO2C—< CO2H NH2 155
I. hydrogenation ——1 ^ 2. desilylation 60% from 156
OLi 156
TBDMSO. 157
OBu'
,,^ / \ ^^ ^ , l.Swem HO^^^^COjBu^ ^ H H I 2.Ph3P=CHMe (-)-158 (95% one isomer) 50%
l.NBS.DMS0,H20
OH V"
2. Bu3onH 60%
/—V H^
CO2BU'
OH
,
V"
H H = 160
/ \ ^^ ^ , k^^.^C02Bu' H
H =
159
/"A H^H i 161
160:161 = 1:4
Scheme 22 condensed with the lithium enolate of r^rr-butyl propionate to give the dehydro intermediate 157. Hydrogenation gave the tetrahydrofuran 158 with the correct stereochemistry required for C-2, C-3 and C-6. The chiral centre at C-8 was introduced in the wrong sense by a stereocontrolled bromohydrin formation on a c/^-olefm 159, controlled by the alkoxy group at C-6, and the bromine removed reductively. This gave a 1:4 mixture of r^rr-butyl (-)-nonactate 160 and rerr-butyl (-)-8epinonactate 161. OLi J hydrogcnauon
TsO.
(+)-158 90% one isomer
165 Scheme 23
For the synthesis of the (+)-enantiomer 165, the configuration at C-6 in the common intermediate 156 was inverted (Scheme 23). The p-toluenesulfonate 162 of the lactone 156 was
245 treated with the lithium enolate of tert-butyl propionate to give, by way of the epoxide 163, the (E)alcohol 164, duly inverting the stereochemistry at C-6. Hydrogenation of the dehydro derivative 164 then gave the alcohol (+)-158 with the desired stereochemistry at C-2 and C-3 as well. The centre at C-8 was then set up in the same way as before. Batmangherlich
and Davidson
(35) also resolved racemic methyl nonactate by
chromatographically separating its esters 167 and 168 with (5)-0-acetylmandelic acid (36).
Ph
Ph OAc
C02Me
OAq
^X'^'^vxt^O'W H ^H
167
C02Me
i
168
Honda and his co-workers synthesised methyl (+)-nonactate 179, setting up the C-6 to C-8 relationship by a chelation controlled allylsilane reaction on the aldehyde 169, and the C-3 centre by hydrogenation of the dehydro intermediate 176 carrying two methoxycarbonyl groups (37) (Scheme 24). The thiolactone 175 and dimethyl diazomalonate gave the dehydro intermediate 176 in the presence of dirhodium tetraacetate, by way of a sulfur-ylid rearrangement developed by these
OBn
SiMe3
1
TiCU
OBn OTHP
83%
OBn uan
l.HCl,H20
OBn OTHP
»-
3. Swem 4.NaC102, ^,,^^ KH2PO4, T 75%
2. DHP, PPTS 169
l.B2H6,THF 2.H202,NaOH
I—I
171 R=CH20H 172 R=CHO 173 R=C02H OBn
N2=C(C02Me)2, Rh2(OAc)4
2. Lawesson's reagent 83%
174 X=0 175 X=S
1. H2, Pd/C, 7 atm.
C02Me
» 2. HCl, H2O, MeOH 177
C02Me
l.TBDMSCl, imidazole
TBDMSO
2. KOBu\ Mel 100%
1. BU4NF
OH
CO2MC ' C02Me
/-^ C02Me
C02Me 2. NaCl, DMSO, H2O
H^H 1:1
i
180
Scheme 24 workers (38) and by Takano and his co-workers (39). The dehydro intermediate 176 was hydrogenated using different conditions (10% Pd-C, MeOH-5% HCl) from Bartlett's, and obtained a 4:1 mixture of cis and trans tetrahydrofurans 177. Protection of the hydroxyl group with ten-
246 butyldimethylsilyl chloride followed by methylation of the protected diester gave the ester 178. The stereochemical control for the C-8, C-6 and C-3 centres was good, but no control at C-2 was observed during Krapcho decarboxylation of the malonate derivative 178, which gave a 1:1 mixture of methyl (+)-nonactate 179 and its diastereoisomer at C-2 180.
2.3 Cyclisation of a 1,4-Diol Derivative. This approach requires both a method for setting up the C-3 and C-6 oxygen functions with the correct relative stereochemistry to give the c/5-2,5-disubstituted tetrahydrofuran ring, and their differenriation in order to use the selective displacement of one of them with inversion of configuration. Only two groups have used this approach. Takatori's group prepared both C-3 and C6 diastereoisomers, and made the overall synthesis convergent by separating them, and converting the wrong isomer into the right one with a Mitsunobu sequence. The Fleming group used two silicon-based approaches. In one, the C-3 and C-6 centres were set up with independent absolute control as C-Si bonds, which were later converted to C-0 bonds, and in the other they were set up by moving the chiral information along the chain. Takatori and his co-workers (40) started from the y-dithio-p-hydroxy ester 181 as a homochiral building block derived by yeast reduction of a ketone supplying C-3 ready resolved (Scheme 25). The ester 181 was methylated with stereocontrol at C-2 by the method of Frater (41), and the product converted by way of the C-1-reduced and protected intermediate 182 into the lithium OH l.LDA,THF TBDMSO ^ S ^ J k ^ C O ^ E t 2.McI,THF,HMPA ^S^^XlJ 3.LiAlH4 X^S
181
V^CHO BOMO i«^ bUMU 184 HMPA 73%
^^fpP'BOMO 82%
r
4:TBDMSC1
H" H = 187 "
OTBDMS
^ 182
TBDMSO OTBDMS I I . . o , ^X^ J l.separauon& recycle , OH ,«c 3. TsCl, DMAP, EtsN ^^MO
^
TBDMSO
^ ^ 8 2
49^^
, BOMO
OTBDMS l.Mel.CaCOs 2. PhgP, CBr4
l.ClO3,H2S04, OH^^Me^CO ^H 3.H2,Pd/C 43%
^
^^^
H^ H E 188 "
95%
TBDMSO I
OH
OTBDMS i
^ H^ H= 189
Scheme 25 acetylide 183 by way of the vinyldibromide. This acetylide was added to the known homochiral aldehyde 184 to give the alcohol 185 as a 1:1 mixture of diastereoisomers, but separation, and recycling of the undesired diastereoisomer by a Mitsunobu inversion-hydrolysis sequence (not illustrated, but taking place in 61% conversion yield), overcame the lack of selectivity. Catalytic hydrogenation provided the differentially protected 1,3,6,8-tetraol, which was tosylated to give the C-6 tosylated derivative 186. Deprotection of the C-1 and C-3 hydroxy groups gave the cis-
247
tetrahydrofuran derivative 187, the cyclisation taking place with complete inversion of configuration at C-6. Oxidation of the primary alcohol group gave a carboxylic acid, which was esterified with diazomethane and subjected to hydrogenolysis to give methyl (-)-nonactate 188, and hydrolysis gave (-)-nonactic acid 189. Fleming and his co-workers developed two independent methods for the synthesis of methyl nonactate by ring closure of 1,4-diol derivatives. The stereochemical control needed for the synthesis of the appropriately substituted 1,4-diol derivatives was based on their work on acyclic stereocontrol using organosilicon compounds, and their routes are unique, and in consequence uniquely long, in eschewing cyclic control almost completely. The three aspects of their method of stereocontrol are: the transposition of chiral information from C-1 to C-3 in the electrophilic substitution of allylsilanes (42), the setting up of stereogenic centres with a 1,3 relationship using the hydroboration of allylsilanes (43), and the setting up of stereogenic centres with a 1,2 relationship by alkylation of enolates having a p-silyl group (44). The hydroboration and enolate alkylations leave the phenyldimethylsilyl group in the molecule, and it is converted, with retention of configuration, into a hydroxy group at an appropriate stage (45). Perhaps the most striking feature of these methods of stereocontrol is the sense in which the word "control" really means control: with each method, it is possible to obtain relative stereochemistry in either sense, making the methods equally suitable for the synthesis of any diastereoisomer. In the first route (Scheme 26) (46), the 1,4 diol system was set up by independently introducing silyl groups with absolute stereochemical control, that at C-6 by a stereospecific allylsilane synthesis from a homochiral allylic alcohol derivative, and that at C-3 by conjugate addition of a silylcuprate to an a,p-unsaturated carboxylic acid attached to a chiral auxiliary. Formation of the c/5-2,5-disubstituted tetrahydrofuran was achieved by converting the phenyldimethylsilyl groups into hydroxy groups, and differentiating between them in order to ensure that inversion of configuration took place at the desired centre. The C-2 and C-3 relationship was estabUshed by anti-scltciiwt methylation of a p-silyl enolate, and the C-6 to C-8-relationship was set up by hydroboration-oxidation of a trans allylsilane. The (5)-propargylic alcohol 191 (70% e.e.) was prepared from the ketone 190 using (S)alpineborane following Brown's and Midland's procedures. The alcohol 191 was converted to its carbamate, semihydrogenation of the triple bond of which gave the c/5-alkene 192. Stereospecific silylcupration (47) then gave the (£')-allylsilane 193. Hydroboration with thexylborane followed by alkaline hydrogen peroxide oxidation gave the anti alcohol 194 with high selectivity {antr.syn =95:5). For the synthesis of the (-i-)-enantiomer, this alcohol was subjected to a Mitsunobu inversion to give the syn diastereoisomer, which was protected as its benzyl ether 195. The aldehyde group in 195 was unmasked, and a Wittig-Homer reaction using the phosphonate 196 carrying Koga's chiral auxiliary gave the a,p unsaturated imide 197. Silylcupration on this imide gave an inseparable mixture of diastereoisomeric bis-silyl derivatives 198 with poor selectivity (2:1) in favour of the isomer illustrated. Stereoselective methylation on the p-silyl ester gave the ester 199, conversion of the silyl groups to hydroxy using mercuric acetate and peracetic acid then gave the 1,4-diol derivative, which was hydrolysed to the acid 200. The only problem left to solve was to differentiate
248
the two hydroxy groups, which was achieved by treatment with an excess of benzenesulfonyl chloride. Two things happened: protection of the C-3 hydroxy group as the p-lactone 201 and benzenesulfonylation of the C-6 hydroxy group. The p-lactoneringopened in acidic methanol and ring closure promptly took place, with inversion of configuration at C-6 to give a mixture of 0OH
1. L i - s 2.BuLi,THF O ^
3.AC2O 61%
1. PhNCCEtsN
190
OCONHPh
^. 2.H2,Pd/CaC03 PbO, MnCl2 91% OH
5-alpine borane, THF
SiMe2Ph
9
) ^O
192
ij
h. 70% ee
l.BuUTHF 2.CuI,Ph3P
SiMe2Ph
2. H2O2, NaOH 82%
3. PhMe2SiLi 73%
1.4-02NC6H4C02H,Ph3P, Et02cN=NC02Et 2. NaOH, MeOH
1. thexylborane
OBn SiMe2Ph
l.TsOH,Me2CO,H20 Ph3C0-.,,
3.BnOC(CCl3)=NH,TfOH
rs
•(EiO)20P'^^^ 0 0 86% 196 Ph3C0—., l.McOMgBr 2. LiHMDS
194 Ph3C0—. OBn SiMe2Ph
^3. McI, DMPU 73%
197 OBn SiMe2Ph
PhS02Cl, Py C02H'
PhMe2Si
64%
199 OBn 0S02Ph
OH y — y
l.TsOH,MeOH
C02Me
C02Me O
2.H2,Pd/C 83%
H 202
HE 203
Scheme 26 benzyl methyl (-i-)-nonactate together with other diastereoisomers. Removal of the benzyl protection by hydrogenolysis gave methyl (+)-nonactate 202, which was separated from the other isomers with the major byproduct being its C-2 and C-3 diastereoisomer 203. Two successive reactions independently setting up stereogenic centres has an arithmetical advantage, at some expense in overall yield, with respect to the enantiomeric purity of the major product, as Horeau (48) and Eliel (49) have pointed out. Although the selectivity in the steps leading to 191 and 198 are only 85:15 and 67:33, respectively, the methyl (+)-nonactate 202 and its enantiomer were obtained at the end of the sequence in a ratio of 92:8. This is because the proportion of the major enantiomer 202 is obtained by multiplying 0.85 by 0.67, whereas the proportion of the minor enantiomer is obtained by multiplying 0.15 by 0.33. The enantiomeric purity of the
249 intermediate alcohol 191 could be raised to >97% e.e. by three recrystallisations of its 3,5dinitrobenzoate 204, which would make the whole synthesis capable of delivering methyl nonactate of >99% enantiomeric purity (50).
3,5-(02N)2C6H3COCl T'^v
Et3N,DMAP
^ ^
98%
In the second route (51), Fleming and Ghosh developed an enantiodivergent approach in order to synthesise both enantiomers. Two silyl groups were set up on adjacent centres, destined to become C-3 and C-4, with a known 1,2-relationship between them. The silyl group on C-4 was then made part of an allylsilane 212 so that the stereochemical information could be moved three atoms along the chain by epoxidation, leaving a 1,4 relationship between the remaining silyl group at C-3 and the incoming oxygen atom at C-6 in the alcohol 215. The C-6 to C-8 relationship could then be controlled in either sense by reduction of a p-hydroxyketone using Evans's and Narasaka's methods, and the C-2 to C-3 relationship could be set up reliably by enolate methylation. By a suitable choice of reactions, the common intermediate 215 was converted into both (+)- and (-)-nonactic acid derivatives. The synthesis of the first homochiral intermediate 209 is shown in Scheme 27. The dimethyl meso 3,4-bistolyldimethylsilyladipate 205 was prepared by a samarium(II) iodide induced coupling ^ O ^ ^ SiMe2Tol l.LiOH,MeOH„THF SiMe2ToI Sml2, THF, DMPU ,C02Me '^5^C02Me " i r n r T T T — ^ Me02C' 2.DCC CH2(C02Me)2 SiMe2Tol ToIMe2Si SiMc2Tol 72% 205 206
'U
5'^^^™ ,.Mc3Si(CH,),0H, CO2H ix:c, DMAP •
HO2C ^ r ^^
84% from 205
207
o
2. H2, Pd/C %:4
TolMe2Si CO2H
Me-iSi' O
SiMe2Tol 209
Scheme 27 of the methyl (Z)-2-tolyldimethylsilylacrylate in THF-DMPU in the presence of dimethyl malonate (52). The homochiral mono 2-trimethylsilylethyl ester 209 of the dicarboxylic acid was prepared from the dimethyl ester 205 in four steps. Lithium hydroxide gave the dicarboxylic acid, which was
250 converted into the meso anhydride 206 by treatment with dicyclohexylcarbodiimide. Diastereoselective opening of the me5o-anhydride with Heathcock*s /?-(+)-2-naphthylethanol (99.7% e.e.) (53), the enantiomeric purity of which was raised by Horeau's method (54), gave a 96:4 mixture of diastereoisomeric mono-esters 207 and 208. Esterification of the mixture with 2trimethylsilylethanol gave the mixture of diastereoisomeric diesters, which was hydrogenolysed to give the mono-ester 209 with an e.e. of 92%. The allylsilane 212 and the common intermediate 215 were made from this monoester (Scheme 28). The lithium dianion of the acid-ester 209 was treated with the aldehyde 210 and the mixture of four diastereoisomeric aldols 211 esterified with diazomethane. The four possible diastereoisomers, present in a ratio of 76:9:9:6 were separated and the 2-trimethylsilylethyl ester group removed by treatment with tetrabutylammonium fluoride. The individual diastereoisomeric 1.2LDA,THF,DMPU CHO McsSi 2. O J )
TolMe2S MesSi' O
SiMe2Tol
3. CH2N2 75%
209
211a and 211b: 3. PhS02Cl, Py 4. collidine, heat
SiMe2Tol
211c and 211d: O O 3. Me2NCH(OCH2Bu')2 ^—^ CHCI3, reflux 93% SiMe2Tol .^^^^Ji^^^CQ^H
KH,THF O
O
^\
OSiMe2Tol 214
SiMe2Tol 2. Bu4NfF, THF
P OH SiMe2Tol 4:5,5:6 211a synsyn 76% 211b anu,syn 9% 211c syn,anu 9% 21 Id anti.anti 6%
TolMe2Si SiMc2Tol ^ = Hi
l.KOH,THF,MeOH 2.MCPBA,Na2HP04 O \ CH2CI2 92%
O jC,^^ / / OH O - ^ 213
O
H2,Pt02,MeOH 87% from 213 O 215
Scheme 28 hydroxy acids were each converted to the required trans allylsilane 212, by syn stereospecific decarboxylative elimination by way of their p-lactones for the acids derived from the esters 211a and 211b, and by and stereospecific decarboxylative elimination for the acids derived from the esters 211c and 21 Id, following chemistry developed earlier (55). The methyl ester was hydrolysed to the acid, which was epoxidised using m-chloroperoxybenzoic acid. The epoxide must have been produced with high anti stereoselectivity (antr.syn = 97:3), but it rearranged to the 7-lactone 213 by a stereospecific 1,2-shift of the silyl group from C-4 to C-5, probably with retention of configuration at C-4 and inversion at C-5 (56). The alcohol 213 on treatment with potassium hydride under the conditions of standard Peterson olefmation underwent stereoselective eliminative rearrangement, well precedented in the work of Yamamoto (57), to give the unsaturated acid 214. Deprotection of the 0-
251 silyl ether and hydrogenation of the double bond gave the hydroxy acid 215 in 41% overall yield from the adipate ester 205. The hydroxy acid 215 was the common intermediate for the synthesis of both methyl (+)-nonactate 220 (Scheme 29) and benzyl (-)-nonactate 227 (Scheme 30). The ketal 215 was hydrolysed with pyridinium tosylate and the ketoalcohol reduced stereoselective^ to the and 1,3-diol 216 (antiisyn = 96:4) using Evans's method (58). The C-6 and C-8 hydroxyl groups were differentiated by formation of the seven-membered lactone 217 using Mukaiyama's method (59). The minor enantiomer of the lactone 217 was largely removed because the racemate crystallised, thereby improving the e.e. from 92% to >99%. The 8-hydroxy group was
SiMe2Tol ,C02H l.PPTS,Me2C0 O
O
u y OH 215
SiMe2Tol CO2H
l.TBDMSCl, imidazole 2. LDA, THF, DMPU SiMe2Tol • TBDMSO 3. Mel
O.
92%
TBDMSO
1^
2. separate from racemate 90%
2. Me4NBH(OAc)4 87%
SiMe2Tol
0 218
l.TsCl,DMAP,Py ^. 2. TsOH, MeOH
KBr, AcOOH ^> NaOAc, AcOH 73%
C02Me 220
91%
Scheme 29 protected as its rerr-butyldimethylsilyl ether, and the lithium enolate was methylated to give the lactone 218. Conversion of the tolyldimethylsilyl group into the hydroxyl group with retention of configuration at C-3 was achieved using potassium bromide in peroxyacetic acid, and the hydroxy group in 219 was converted into its tosylate. Methanolysis opened the lactone ring and allowed the free hydroxyl group to displace the tosylate, giving methyl (+)-nonactate 220. The overall yield of (+)-methyl nonactate from the common intermediate 215 was 47%. For the synthesis of benzyl (-)-nonactate (Scheme 30), the hydroxy acid 215 was esterified and deketalised to give the ketoester 221. Stereoselective reduction of the ketone group using Prasad's modification of Narasaka's method (23) gave the syn 1,3-diol (syn:anti = 90:10), which was converted to its acetonide 222. Stereoselective methylation of the open-chain p-silyl ester gave only the ester 223 with the anti relationship between the incoming methyl group on C-2 and the resident silyl group on C-3. Differentiation of the C-6 and C-8 hydroxyl groups was achieved by removing the acetonide, hydrolysing the ester group, and forming the seven-membered lactone 224 using Mukaiyama's procedure (59). As in the earlier sequence, this lactone was enantiomerically enriched (from 92% to >96% e.e.) by removal of the crystalline racemic lactone. The free hydroxyl group in the lactone 224 was protected with r^rr-butyldimethylsilyl chloride, and the lactone opened
252 with sodium benzyloxide to give the benzyl ester in quantitative yield. The C-6 hydroxy group was then converted to its tosylate 225, and the C-3 tolyldimethylsilyl group to hydroxyl, as before. The intramolecular displacement with inversion at C-6 226 then gave directly benzyl (-)-nonactate 227. The overall yield of benzyl (-)-nonactate from the intermediate 215 was 35%. l.Bu2BOMe,NaBH4 SiMe2Tol THF, MeOH C02Me
SiMe2ToI I.CH2N2 CO2H 2. PPTS, Me2C0 86% 215
—
^.
2. (MeO)2CMe2.PPTS
SiMe2Tol 1. PPTS, MeOH 2 cOiMe 2. KOH, THF, McOH
SiMe2Tol C02Me l . L D A , T H F , D M P U ^ 8 ^ ^ 6
r
2. Mel
3. CI-^N"^ Et3N
89% less ??%
4. separate from racemate 83% l.TBDMSCl, imidazole 2.NaOBn,BnOH,THF
"^
^V-/^^'^'^™3.TsaDMAP,Py g \ 85%
SiMe2Tol ^C02Bn
jj TBDMSO
OTs
225
224
KBr, AcOOH AcOH 78%
OH TsO ^ ^ i 226
H ^H
i
227
Scheme 30
2.4 Electrophilic Cydisation of y,8-Unsaturated Alcohols and Enols In their synthesis of racemic methyl nonactate 233 and its 8-epimer 234 (Scheme 31) (26), Baldwin and Mclver controlled the stereochemistry of C-2 and C-3 by conjugate addition of homoallylmagnesium bromide to 2,2-dimethyl-3(2H)-furanone 228 and methylation of the regenerated enolate, which took place with high selectivity (10:1) in favour of the trans dialkylfuranone 229. Conversion of the ketone to the oxime followed by fragmentation with thionyl chloride and protection gave the nitrile, and the now free alcohol group was protected as its 2,6dichlorobenzyl ether 230 {anti:syn = 32:1). Conversion to the corresponding aldehyde with diisobutylaluminium hydride^ followed by exposure to iodine in acetonitrile gave the cyclic iodoaldehyde, which was oxidised to the corresponding acid 231. The iodoetherification took place stereoselectively in favour of the desired stereochemistry at C-6 {cis'.trans = 50:1). Dithiane addition and esterification gave the masked aldehyde 232. After removal of the protecting group, the aldehyde was treated with dimethylzinc in the presence of titanium tetrachloride to give methyl
253
nonactate 233 and methyl 8-epinonactate 234 in a ratio of 24:1. The same reaction using lithium dimethyl cuprate took place selectively (4.5:1) in favour of methyl 8-epinonactate 234 1.
%,y-^MgBT
W
CuBr, MeaS
OCH2C6H3CI2 CN
. SCX:i2, CCI4 •!
^> 2. LDA, THF, Mel
228
l.NH20H,Py
229
56%
3.NaH,THF 4.2,6-Cl2C6H4CH2Br 59%
anti'.syn 10:1
l.DIBALH
C02Me
^2.12, MeCN 3.CrO3,H2S04 54% 1. HgO, BF3.0Et2
C02Me
^.
Me2Zn,TiCl4 Mc2CuLi
24:1 1:4.5
2. Me2Zn, T i C ^ 65% orMe2CiiLi 60%
Scheme 31 Walkup and Park synthesised not only methyl (±)-nonactate 240a but also (±)-homononactate 240b and (±)-bis- 240c and trishomononactate 240d (Scheme 32) (60) starting from hexa-4,5dienal and the appropriate lithium enolate 235 in each case. The relative stereochemistry of C-6 and OLi
Me4N-' (AcO)3BH!•
OHC^x--^^'
MeCN, AcOH, - 4 0 X 236a 236b 236c 236d
235
OH OH
R=Me R=Et R=Pr' R=Bu'
55% 58% 55% 55%
OTBDMS
TBDMSCl,
l.Hg(02CCF3)2
IN
imidazole 237a R=Me 237b R=El 237c R=Pr^ 237d R=Bu'
90% 90:10 80% 96:4 90% 99:1 84% >99:1
^. 2. PdCl2 cat., CuCl, 238a R=Me >98% CO, McOH 238b R=Et >98% 238c R=Pr' >98% 238d R=Bu' 25% + 6-silyloxy-8-ol 75% OH
H2, Rh/Al203 C02Me
y—X CO2MC
CO2MC H^H
cis: trans >98:2 239a R=Me 87% 239b R=Et 70% 239c R=Pr^ 80% 239d R=Bu' 80%
240a-d
1:1
:
241a-d
Scheme 32 C-8 was controlled by Evans' reduction (58) of the p-hydroxyketones 236 giving the anti 1,3-diols 237. The y-silyloxyallenes 238 were then subjected to a one-pot procedure already developed by
254 these workers involving oxymercuration coupled to a palladium-catalysed methoxycarbonylation (61), which gave the tetrahydrofurans 239 with high stereoselectivity (cis.trans >98:2). This short sequence of reactions established efficiendy the required stereochemistry at C-8, C-6 and C-3, but, unfortunately, the final stereogenic centre at C-2 was generated with no control, catalytic hydrogenation gave a 1:1 mixture of the desired products 240 and their C-2 diastereoisomers 241. Iqbal and his co-workers reported a synthesis of 2,5-disubstituted tetrahydrofurans from Y,6unsaturated alcohols (Scheme 33) (62). The stereochemistry of the C-2 and C-3 centres was set up with some selectivity by reduction of the p-ketoester 242. Epoxidation of the terminal double bond
u
I.NaH 2.BuLi
C02Me
NaBH4
C02Me
243 53% OH
CI
V^-N.X^C02Me
242
244
,C02Me
MCPBA ^>
m.^^^^^}^
+ Ho..,^,,X^
78% 243
245
89:11
246
Scheme 33 of the major alcohol 243 with w-chloroperoxybenzoic acid was surprisingly well controlled, with the epoxide undergoing cyclisation under the reaction conditions to give the cis and trans tetrahydrofurans 245 and 246 in a ratio of 89:11. The major product, the alcohol 245, is the racemic methyl ester corresponding to the intermediate 158 in the Davidson and Batmangherlich synthesis of rerr-butyl nonactate (Scheme 22).
o- o-
> T ^ - . XX OMe
TBDMSO.
247 1. base, Mel
OTBDMS
249
O
TBDMSO ^^v
38%
^^^^
251
/—V C02Me
^^ 75psi 86%
l.NPSP,Znl2 '^^™S0 n-^ • / \ ^ /\^C02Mc 2. separation » | jj O j
250
H2, Raney Ni
.TBDMSO,
248
C02Me 2. Lindlar 65%
CO2MC ^g^^
252
^'^\^n'i^-^
CO2H
253 Scheme 34
Ley also used the alkylation of a p-dicarbonyl dienolate 248 to assemble the precursor 250 for an electrophile-induced cyclisation (Scheme 34) (63). The enol of the p-ketoester 250 underwent
255 cyclisation with N-phenylselenophthalimide (NPSP) to give a separable mixture of two diastereoisomers, from which the selenide 251 with the correct C-6 to C-8 stereochemistry was isolated. Raney nickel induced hydrogenolysis of the now superfluous selenide as well as saturation of the C-2 to C-3 double bond, as in Bartlett's synthesis, and gave the 0-silyl protected methyl nonactate 252, which was converted to nonactic acid 253.
2.5 Intramolecular Conjugate Addition ofAlkoxides Gerlach and Wetter established the relative stereochemistry between C-6 and C-8 at the beginning of the synthesis, and made the tetrahydrofuran ring by an intramolecular conjugate addition of the C-6 alkoxide to an a,p-unsaturated ester (Scheme 35) (11). The 1,3-diketone 254, prepared from the dianion of acetylacetone, was reduced with sodium borohydride to give a mixture of the diols 255 and 256 (3:2), which were separated by chromatography. The undesired erythro diastereoisomer was converted to the desired three isomer by tosylation, displacement with acetate ion and hydrolysis, and the combined crops of threo diol 256 were acetylated. Ozonization of the diacetate followed by Wittig reaction of the aldehyde 257 with the carbanion of pmethoxycarbonylethyl diethyl phosphonate gave a mixture of (£") and (Z) isomers 258 {E:Z 85:15). Base catalysed cyclisation of the a,p-unsaturated ester 258 (E:Z = 7:3) gave a mixture in ratios of 100:68:56:71 in which methyl (±)-nonactate 259 was the major product, separated as its rerr-butyl ether and ester. OH OH O
o
O
lin
KNH2
o
255
NaBH4
l.TsCl 2. separate
70% "^ 255:256 3:2 OH OH
3. NaOAc
4. KOH 13%
256 OH OH
OAc OAc
1. AC2O ^> 2. O3, Me2S
256
(ElO)20P^ C02Me ^CHO
257
l.KOH,MeOH.MeCN :N C02Me 258
£;Z7:3
1
66%
2. CH2N2, H-' 97%
9«
r\
C02Me
H^Hi 259
Scheme 35 Sun and Fraser-Reid reported a synthesis of methyl (-)-nonactate starting from D-ribose, C-4 of which (sugar numbering) provided C-6 (nonactin numbering) of the tetrahydrofuran ring (Scheme 36) (64). The ribose-derived aldehyde 260, was converted to the ketone 261 by a Wittig reaction followed by hydrolysis of the enol ether. Raney nickel catalysed hydrogenation of the ketone 261
256 provided the (S)-alcohol 262a with the correct C-8 stereochemistry for methyl (-)-nonactate 265 with high selectivity (9:1), probably stemming from chelation of the nickel to the ring oxygen atoms. In addition, the minor isomer was converted into the major by displacement of its sulfonate with sodium benzoate. The alcohol 262a was hydrolysed and protected as its acetonide to give the aldose 263, which was treated with the phosphorane. Wittig reaction took place followed by intramolecular
OHC
OMe
yOy 6j0
'^'^
A
yOy
H2.Ni /^
o3o
A
260
"I—\ ^ COMe
y—V ^ OMe
2 steps /
\ ^
1. separate 262a 2. HsO""
oTo
3.Me,C(OMe),
A 261
262a R^=OH,R2=H90% 262b R ' = H , R 2 = 0 H 10%
HO
1. Ph3PYC02Me ^^
/\^0E ^
l.benzoylate
^ ^ 3 3. Me2NCH(OR)2 X^o^^^'"^' ^ ^ 4. Ac20,heat H H i 'V' "^ 5. H2, Pd 255 ^^ '^ ^ 6. NaOMe 263 264 78% Scheme 36 conjugate addition of the alkoxide on the unsaturated ester under kinetic control to give a 1:3 mixture ^
2. separate / 3.NaOMe 4. separate and recycle
of the two C-2 diastereoisomers, with the desired isomer 264 the minor component. Under kintetic control, the side-chain at C-3 (nonactin numbering) remains on the upper surface as illustrated, an observation of Moffatt and his co-workers (65). The ratio was improved to 3:2 by equilibration with sodium methoxide by a p-elimination-readdition pathway. After three cycles of equilibration and separation, 90% of the mixture had been converted into the diastereoisomer 264. The acetonide group in the benzoate of 2 6 4 was hydrolysed and the resulting diol subjected to Eastwood deoxygenation (66), which gave the corresponding dihydrofuran. Hydrogenation over palladium then gave methyl (-)-nonactate 265. Sun and Fraser-Reid also synthesised the (+)-enantiomer from the same starting material, which required that the configuration at C-4 be inverted (Scheme 37). The early intermediate 261 prepared from D-ribose was treated with base, which caused epimerisation to give the thermodynamically more stable isomer 266, with an equilibrium ratio of 9:1 as expected from Moffatt's precedents, but surprising at first sight, given that the side chain is endo in the bicyclic system. Nickel-catalysed hydrogenation, selective enough to give the alcohol 267 to the extent of 75%, deprotection of the acetal, and protection of the diol as its acetonide gave the aldose 268. The aldose was treated with the nitrile analogue of the same phosphorane as before to give an epimeric mixture of the nitriles 269. This mixture was epimerised in a few cycles, with separation after each cycle, finally providing the nitrile 270 in 84% yield. The nitrile was used in this sequence because it behaved better in the equilibration steps than the corresponding ester. Eastwood deoxygenation.
257 hydrogenation of the dihydrofuran, and conversion of the nitrile to the methyl ester gave methyl (+)nonactate 271. O
/
V^y Q
OMe
Q
NaOMe, MeOH MeOH NaOMe,
O
OH
OMe
^Jl^ ^ ^ ^^'^\.
90% epimerisation
Q
261
Q
H2. Ni
OMe
JC^^V"
2. separate
J
266
OH
PhsP^CN
267 l.NaOEt,ElOH
0^0
A
2. NaOMe 93%
268
l.benzoylate 2. HsO"^ 269 3. Me2NCH(OR)2 OH ^-AczO^heat JC4^.K^C02Me
OH CN
2. Me2C(OMe)2
OH CN
Ov^O
^
1.H30-'
2. separate 3. recycle and separate 84%
A
5.H2,Pd 6.H2O2 7. NOCl 8. CH2N2 9. NaOMe
0^0
A
270
Scheme 37 2.6 From Bicyclic Intermediates White and his co-workers were the first to use a bicyclic intermediate to control the relative stereochemistry (Scheme 38) (14). They set up the 8-oxa-bicyclo[3.2.1]octene 273 using O
vV Br
fl
q
2.""v^j^O ""2.^^ "'^"^ CF3CO3H, Na2HP04
Zn-Cu 272
Br
° ^
273 .C02Me 220 T
1. NaOMe O 2.NaH,CS2,MeI 71%
274
92% C02Me
MeS2C0 " 1 " 275 OHC
1, (Sia)2BH
CrOs
C02Me
^.
»•
278:279= 1:1
95%
2. H2O2, HO"
^ OHC.,^ H
46% 1. separate 278 2. MeMgl
64%
HE 279
C02Me
C02Me 280
A ^
1:1
281
Scheme 38 Hoffmann's cycloaddition (67) of the oxyallyl cation 272, generated from 2,4-dibromopentan-2-one with LeGoff s zinc-copper couple (68). Hydrogenation followed by Baeyer-Villiger oxidation gave the lactone 274, with C-2, C-3 and C-6 correctly set up. Methanolysis gave a single hydroxyester,
258 which was converted into its xanthate 275. The xanthate on pyrolysis provided the terminal alkene 276,
which was subjected to hydroboration-oxidation to give the primary alcohol 277.
Unfortunately, the configurational identity at C-2 was lost during the hydroboration-oxidation, the alcohol proving to be a 1:1 mixture of C-2 epimers. These were separated after converting the alcohol 277 to the mixture of aldehydes 278 and 279, and treatment of the isomer 278 with methylmagnesium iodide gave methyl nonactate 280 and methyl 8-epinonactate 281 with no selectivity, a problem that was solved later by Baldwin and Mclver (Scheme 31). Warm and Vogel used 7-oxabicyclo[2.2.1]heptan-2-one 284 to control the relative stereochemistry of C-2 and C-3 of methyl nonactate. They also resolved it, and used the (+)enantiomer to synthesise methyl (+)-nonactate (Scheme 39), and the (--)-enantiomer to synthesise methyl (-)-nonactate (69). Zinc iodide-catalysed Diels-Alder reaction between furan and 1cyanovinyl acetate gave the adduct 282, which was saponified to give the racemic ketone. This was hydrogenated using palladium on charcoal, and the enantiomers (+)- and (-)284 were resolved by chromatography of their sulfoximides 285 and 286. Pyrolysis of each diastereoisomer gave the
Aco^cN
"^o^
^ ^ ^OAcC 282
^ C N
2.H2.P(1A:
^-^^
^ ^
4:1 O
42% each
OH
O
' ,
5-
285
(+)-284
l.KHMDS 2. Mel ^> 3. separate from dimethyl product 4. MCPBA, NaHCO^ 59%
,C02Me
C02Me 288:289 1:3 289 1 l.KOH 2. CH2N2 288:289 3 (36%): 4 (27%)
L-Seleciride ^. 82%
C02Me
CO2MC
4
290:291 10:1 l.PhC02H,Ph3P,DEAD
|
2. NaOMe, MeOH 85%
Scheme 39 enantiomerically pure ketones 284 (>99% e.e.). Methylation of the bicyclic ketone (+)-284 followed by Baeyer-Villiger oxidation gave the unstable oxoacetal 287. Addition of one equivalent of the silyl enol ether of acetone to a 1:1 mixture of the acetal 287 and titanium tetrachloride gave a 1:3 mixture of the ketone 288 and its trans isomer 289. However, the undesired isomer 289 could be equilibrated on treatment with potassium hydroxide, by p-elimination and readdition. Acidification
259 and esterification with diazomethane gave the ketones 288 and 289, as a 4:3 mixture. Reduction of the ketone 288 with L-Selectride gave a 10:1 mixture of methyl (+)-8-epinonactate 290 and methyl (+)-nonactate 291. Mitsunobu inversion of the major product and treatment with sodium methoxide gave methyl (+)-nonactate 291. The enantiomeric bicyclic ketone (-)-284 similarly provided methyl (-)-nonactate.
2.7 Ireland-Claisen Rearrangement Ireland and Vevret developed a route for the synthesis of both (+)- and (-)-nonactic acids, with the stereochemistry at C-6 derived from C-4 of D-gulonolactone and D-mannose, respectively (70). For the synthesis of (+)-nonactic acid 301 (Scheme 40), the furanoid glycal 295 was prepared in 10 steps from D-gulonolactone 292 by fairly straightforward functional group manipulations. The
HO
l.HCl,MeOH 2. Me2NCH(OMe)2 3. AC20,130°C
l.Me2C0,H-' 2. DIBALH
OH
•
0
0
3. NaH, BnCl, DMF
6H«
o^^o-^ I
293
HO V-^
l.BuLi 2.EtC0Cl
292
MOMO
\-4 H ^ 294
1. 25°C
MOMO
l.Li,NH3 2.CCl4,Ph3P MOMO OBn 3.Li,NH3
^. 2. CH2N2
C02Me H ^H 299 47% or 54% _
H2,Pt/C C02Me ^> 44% from 295
MOMO
3. LDA, THF, -78°C
H^ 295
11% from 292 r=^
1
OBn 4. 9-BBN 5. NaOH, H2O2 6. P2O5, CH2(OMe)2 OSiMcs"
4. MeaSiCl
H^ 296 l.HCl,MeOH 2. Swem
MOMO
K0H,H20
V—TV
H ^H 298 298:2-epi-298 86:14
^ 3. Mc2CuLi or McMgBr
CO2H
95%
C02Me
300 53%^r40%
Scheme 40 glycal 295 was converted into its propionate ester, which on treatment with lithium diisopropyl amide in THF and trimethylsilyl chloride gave the £-trimethylsilyl enol ether 296. The key step, the Ireland-Claisen rearrangement (71) setting up the stereochemistry at C-2 and C-3, took place at room temperature, and the product mixture was esterified to give a mixture of the C-2 diastereoisomeric
260 esters 297. Catalytic hydrogenation gave the corresponding mixture of esters 298 (86-89:14-11) in favour of the desired isomer. Evidently the Claisen rearrangement had taken place largely with the boat-like transition structure, and suprafacially on the dihydrofuran ring. Deprotection followed by Swem oxidation gave the aldehyde. No stereochemical control was observed in the dimethylcuprate addition to the aldehyde, which gave both diastereoisomers 299 and 300 in approximately equal amounts, in contrast to Baldwin's observation of good control, although in the wrong sense, in this reaction (Scheme 31). They observed somewhat better control when methylmagnesium bromide was used. The enantiomeric glycal 305 was prepared from the D-mannose 302 in 11 similar steps (Scheme 41), and (-)-nonactic acid 310 prepared from it exactly as described for the (+)-enantiomer 301 (Scheme 40).
V
OH l.Me2C0,H-' .OH 2. NaH, BnCI, DMF HO^ 'V-^^^^^*^" HO^'^^^O-^OH 302 Q' -Q
I.Li,NH3 2.CCl4.Ph3PMOMO
^ ^ O ' ^ O B n 3.Li,NH3 ^ H 36% 304 36% from from 302 302
H2, Rh/C
I
jiQ V-a
l.BuLi 2. EtCOCl
3.NaOH.H202 4. KH, MCOCH2CI
MOMO
^ V ^ >> 3.LDA,THF. H^ -78°C 3^5 305 4. Me.SiCL MesSiCl, 25°C 5. CH2N2
V
C02Me
49% from 305
I.AC20,130°C 2. 9-BBN
3.HCl,MeOH O ' ^ r ^ O ' ^ O cB n 4. Me2NCH(OMe)2, CH2CI2 V-n " 303 Me2N f ^ -i^-
MOMO
MOMO M UMU
O^^O
H" H I 307
l.HCi,MeOH 2. Swem
k^^^tN^C02Me H H^ 306
x ' ^ . X h Q ' t v ^ CO2MC H "^ H =
3. Me2CuLi
308 40% OH
307:2-epi-307 89:11
r ^ H^ H= 309 45%
. - ^ X x t ^ Q ' t ^ C02Me
KOH, H2O
H "^ H 308
*- -'^"'^^o'^V' CO2H H ^ H= 310
Scheme 41
3.
SYNTHESES OF NONACTIN The synthesis of nonactin requires that the (+)- and (-)-nonactic acid units be joined together in
an alternating sequence, followed by closing the ends to give the macrocycle. There are two possibihries for ring-formation: (i) cyclodimerisation of a "dimer" and (ii) unimolecular cyclisation of
261 a "tetramer." Both strategies have been used, with several different ways to assemble the dimer and tetramer. In one strategy, the differentially protected nonactic acid enantiomers are coupled to give the protected dimer using standard esterification techniques that preserve the configuration at C-8. In the other strategy, the linear units are coupled by taking one enantiomer of nonactic acid, and using it as a carboxylate nucleophile to displace, with inversion of configuration, the 8-methanesulfonate or tosylate of the 8-epi-diastereoisomer of the other enantiomer, protected at the carbonyl group. 3.1 Synthesis ofNonactin by Unimolecular Cyclisation The first synthesis of nonactin was reported by Gerlach and his co-workers (Scheme 42) (72) in 1975 by the cyclisation of a linear tetraester, but the linear tetraester 317 was a mixture of diastereoisomers because it was made from racemic nonactic acid 311 (prepared in Scheme 35).
CO2H
[ I (+)-311 •^ Bu'OsCMe, MeSOsH 70%
NaH,BnBr ^ OBn
CO2BU'
Py, C i O z S - f J -
70%
COzBu^
H2, Pd/C CO2BU'
C02Bu^
I.CF3CO2H 2. H2, Pd/C
i
4. AgC104, MeCN, 10--*M 5. separate
nonactin (10%) + other diastereoisomers (30%)
Scheme 42 Appropriately protected monomers, the racemic benzyl ether 312 and the racemic rerr-butyl ester 313, were coupled using the mixed anhydride with 2,4,6-trimethylbenzenesulfonyl chloride to give
262 the protected dimer 314 as a mixture, inevitably, of four diastereoisomers, all racemic. Treatment with trifluoroacetic acid removed the rerr-butyl ester group from one portion of the dimer to give the acid 315, and catalytic hydrogenolysis removed the benzyl ether group from the other portion to give the alcohol 316. Activation of the acid 315 with 2,4,6-trimethylbenzenesulfonyl chloride and coupling with the alcohol 316 gave the linear tetramer 317, this time as a mixture of eight racemic diastereoisomers. Deprotection of the rerr-butyl ester with trifluoroacetic acid and of the benzyl ether by hydrogenolysis gave the free linear tetramer, which was cyclised by the Mukaiyama thioester method (73). At this stage the complexity of the mixture became rather less, since there are only four possible diastereoisomers, assuming that there is complete preservation of stereochemical integrity within each nonactic acid unit. Of these four diastereoisomers, three were isolated by chromatography in a ratio of 1:5:2, and the last proved to be nonactin 1. Schmidt and his co-workers (74) were the first to report a synthesis of nonactin from enantiomerically enriched components (Scheme 43). Potassium (-)-nonactate 321, prepared from
4^is^C02Bn H ^H =
H
Asx/o^C02H
H
H
(+)-319 i TsCl,Py 85% ,—V
OTs
H = (-)-320"
OH
An
OTs
C02Bn . A ^ ^ V ^ ^ ^ z "
OH
1
y
O
:
,
1
OH
I
V
C02Bn
O C02Bn
H ^ H = •
H ^H
H" H E
324
I l.H2,Pd/C 2.KHCO3 OH
r—.
O
=
J—.
OTs
-^^oW^o-^-^oV^^^'
/-^
O C02Bn
/ ^ oH^ t H^ =- ^ o
327
OH
y
i
O
r
i
.
O
| 7 4 % from 324 I y 1
O
r
y
v ^C02Bn
H "^ H =
H^H
H "^ H =
H
H
328
I l.H2,Pd/C 2. (PyS)2,Ph3P I 3.AgC104,bei benzene 20% nonactin Scheme 43 the acid 318 (=38) was coupled, with inversion of configuration at C-8, with the 8-epi-tosylate 322, derived from benzyl (+)-8-epi-nonactate 319 (prepared from the methyl ester 40), to produce
263 benzyl (-)-nonactinoyl-(+)-nonactate 324. The diester 325 was prepared similarly using potassium (-)-8-epinonactate 323 and the same 8-epi-tosylate 322 used before. Hydrogenolysis of the "dimer" 324, and conversion to the potassium salt with potassium bicarbonate gave the left-hand component 326. Tosylation of the alcohol 325 gave the 8-tosylate 327, which gave the linear tetraester 328, again with inversion of configuration at C-8, on treatment with the potassium salt 326. Hydrogenolysis, activation and cyclisation following Gerlach produced nonactin in 20% yield together with C-2 and C-8 epinonactins in 12% yield. Fleming and Ghosh synthesised nonactin by cyclisation of the linear tetramer 335 assembled from methyl (+)-nonactate and benzyl (-)-nonactate (Scheme 44) (75). The (9-protected (+)-nonactic
C02Me (+)-329 I l.TBDMSCl, imidazole 2. KOH, THF, MeOH 98% TBDMSO
331 DCC, DMAP 93% TBDMSO C02Bn H ^ H = 100% H2,Pd/C TBDMSO
/—V
TsOH,AcOH 98%
O
Ov^o^^
C02Bn
H
H^A
H E
U H
^
U H
'
334 CI
DMAP, ClOC C ^ C l
95%
CI TBDMSO C02Bn H
H
H
H =
H ^ H =
335 l.H2,Pd/C 2TsOH, AcOH, H2O 3. CI
t nonactin
DMAP, ClOC • ^ - C I
69%
CI
Scheme 44 acid 330 was prepared from the hydroxyester 329 (=220), and coupled with benzyl (-)-nonactate 331 (=227), without inversion at C-8, to give the dimeric ester 332. A portion of this dimer was hydrogenolysed with palladium on charcoal to give the acid 333, while the other portion was
264 deprotected at the hydroxy group using acid to give the alcohol 334. The acid 333 was coupled to the alcohol 334 using the Yamaguchi mixed anhydride method (76) to give the protected linear tetramer 335. The protecting groups were removed to give the free tetramer, which was cyclised in high yield (73%), the best so far achieved, again using Yamaguchi's method. There was no improvement in yield when potassium tetrafluoroborate was present, indicating that coordination to potassium did not help the cyclisation. 3. 2 Synthesis ofNonactin by Cyclodimerisation Schmidt and his co-workers also reported the synthesis of nonactin (Scheme 45) (77) by cyclodimerisation of (-)-nonactinyl-(+)-nonactic acid 336 (=326), which was treated successively OH y i H "^ H
O
=
y y H "^ H
336
/V^V'-carbonyldiimidazole, DBU i
or (PyS)2,Ph3P
"poor yield"
nonactin
Scheme 45 either with carbonyldiimidazole and diazabicycloundecen or with the bispyridyl 2-disulphide and triphenylphosphine to give nonactin in "poor" yield in both cases.
H^ H I (-)-338 ^ MsCl, EI3N, DMAP 91% OMs / — ^ C02Me H
H = 340
C02Me H ^ H = 341 I LiSPr",HMPA 1 CO2H H ^ H l.(PhO)2POCl,Et3N I 2. heat, C6H6, DMAP 16% nonactin + cyclic "dimer" and "oligomers" and polymer
Scheme 46 Bartlett and his co-workers (20) synthesised nonactin by the cyclodimerisation approach (Scheme 46). The potassium salt 339 of (4-)-nonactic acid 337 (=87) and the mesylate 340 of the
265 (-)-8-epiester 338 (=86) gave the dimeric ester 341 with inversion of configuration at C-8, as in Schmidt's synthesis, but working with the enantiomer of each component. The methyl ester was cleaved by lithium n-propyl mercaptide with some difficulty, and with some (25%) epimerisation at the C-2 positions of the nonactic acid units to give the acid 342. Macrolactonisation using Masamune*s procedure (78), gave nonactin in 16% yield, accompanied by the cyclic "dimer" and "oligomers" and polymers. Fleming and Ghosh also synthesised nonactin by cyclodimerisation of the acid derived from the benzyl ester 343 (=334) using Yamaguchi conditions (Scheme 47) (75). Nonactin was isolated in lower yield than by the linear tetramer method, presumably because of the problem of "dimer" and "oligomer" formation.
Q^
o-^^-^/?k^^^^^" H H E 343
l.H2,Pd/C 2.
CI
aoc-0"Ci DMAP a
nonactin (52%) + cyclic "dimer" and "oligomers" and polymer Scheme 47
CONCLUSIONS The syntheses of nonactic acid and its derivatives illustrate many of the most popular methods of stereocontrol used in synthesis. There are examples of absolute control based on (a) resolution, (b) Sharpless asymmetric epoxidation and other methods of kinetic recognition, including an enzymatically controlled reduction, (c) the use of chiral auxiliaries, and (d) starting materials from the chiral pool such as sugars, and malic and glutamic acid. Relative stereochemical control has been achieved by such devices as (a) control on bicyclic frameworks, (b) the use of many different cyclic structures, especially five-membered rings with a predictable stereochemical bias, (c) the similar use of cyclic transition structures for hydride delivery, for enolate alkylations, the Ireland-Claisen rearrangement, and for ring-forming reactions, t>oth pericyclic and ionic, and (d) by the independent synthesis of separated stereogenic centres with absolute control. There are examples of such themes as (a) kinetic and thermodynamic control, (b) of convergent and linear synthesis, (c) of the recycling of unwanted disastereoisomers by Mitsunobu and other inversion processes, and by repeated equilibration and separation, and (d) of the problems of controlling distant stereogenic centres. And there are examples of a very wide range of the common reactions of organic chemistry, including those used in C-C bond-formation, functional group manipulation, and protecting group tactics. The syntheses illustrated here would, on their own, make a surprisingly good basis for an introductory course in organic synthesis.
266 REFERENCES 1
R. Corbaz, L. Ettlinger, E. Gaumann, W. Keller-Schierlein, F. Kradolfer, L. Neipp, V. Prelog and H. Zohner, Helv, Chim. Acta, 1955, 38, 1445.
2
(a) W. Keller-Schierlein, Fortschr. Chem. Org. Naturst., 1968, 26, 161; (b) W. Keller-Schierlein, Fortschr. Chem. Org. Naturst., 1973, 30, 313; (c) J. Dominguez, J. D. Dunitz, H. Gerlach and V. Prelog, Helv. Chim. Acta, 1962, 45, 129; (d) H. Gerlach and V. Prelog, Justus Liebigs Ann. Chem., 1963, 669, 121; (e) J. H. Prestegard and S. I. Chan, / . Am. Chem. Soc, 1970, 92, 4440; (f) M. Dobler, "lonophores and their Structure", Wiley, New York, 1981; (g) B. T. Kilbourn, J. D. Dunitz, L. A. R. Pioda and W. Simon, J. Mol. Biol, 1967, 30, 559.
3
(a) K. H. Wallhausser, G. Ruber, G. Nessenmann, P. Prave and K. Zept, Arzneimittelforschung, 1964, 14, 356; (b) K. Ando, H. Oishi, S. Hirano, T. Okutani, K. Suzuki, H. Okasaki, M. Sawada and T. Sagawa, J. Antibiot., 1971, 24, 347.
4
(a) S. N. Graven, H. A. Lardy and S. Estrada-0, Biochemistry, 1967, 6, 365; (b) S. N. Graven, H. A. Lardy, D. Johnson and A. Rutter, Biochemistry, 1966, 5, 1729.
5
D. H. Haynes and B. C. Pressman, J. Membr. Biol, 1974, 18, 1.
6
M. Dobler, J. D. Dunitz and B. T. Kilbourn, Helv. Chim. Acta, 1969, 52, 2573.
7
K. Mislow, Croat. Chem. Acta, 1985, 58, 353.
8
P. A. Bartlett, Tetrahedron, 1980, 36, 2. '
9
J.-C. Harmange, B. Figadere, Tetrahedron Asymmetry, 1993, 4, 1711.
10 G. Beck and E. Henseleit, Chem. Ber., 1971,104, 21. 11 H. Gerlach and H. Wetter, Helv. Chim. Acta, 1974, 57, 2306. 12 U. M. Kempe, T. K. Das Gupta, K. Blatt, P. Gygax, D. Felix and A. Eschenmoser, Helv. Chim. Acta, 1972,55,2187. 13 (a) J. Gombos, E. Haslinger, H. Zak and U. Schmidt, Monatsh. Chem., 1975, 106, 219; (b) U. Schmidt, J. Gombos, E. Haslinger and H. Zak, Chem. Ber., 1976, 109, 2628; (c) H. Zak and U. Schmidt, Angew. Chem. Int. Ed. Engl, 1975, 14, 432. 14 M. J. Arco, M. H. Trammell and J. D. White, /. Org. Chem., 1976, 41, 2075. 15 W. C. Still, L. J. MacPherson, T. Harada, J. F. Callahan and A. L. Rheingold, Tetrahedron, 1984, 40, 2275. 16 W. C. Still and I. Galynker, Tetrahedron, 1981, 37, 3981. 17 P. A. Bartlett and K. K. Jemstedt, J. Am. Chem. Soc, 1977, 99, 4829. 18 P. A. Bartlett and K. K. Jemstedt, Tetrahedron Lett., 1980, 21, 1607. 19 P. A. Bartlett, J. D. Meadows, E. G. Brown, A. Morimoto and K. K. Jemstedt, J. Org. Chem., 1982, 47, 4013. 20 P. A. Bartlett, J. D. Meadows and E. Ottow, J. Am. Chem. Soc, 1984, 106, 5304. 21 W. S. Johnson, C. Edington, J. D. Elliott and I. R. Silverman, / . Am. Chem. Soc, 1984, 106, 7588. 22 I. R. Silverman, C. Edington, J. D. Elliott and W. S. Johnson, J. Org. Chem., 1987, 52, 180. 23 K. Narasaka and F.-C. Pai, Tetrahedron,
1984, 40, 2233. See also, K.-M. Chen, G. E.
Hardtmann, K. Prasad, O. Repic and M. J. Shapiro, Tetrahedron Lett., 1987, 28, 155.
267 24 B. Lygo, N. O'Connor and P. R. Wilson, Tetrahedron, 1988, 44, 6881. 25 B. Lygo, Tetrahedron, 1988, 44, 6889. 26 S. W. Baldwin and J. M. Mclver, J. Org. Chem., 1987, 52, 320. 27 P.-F. Deschenaux and A. Jacot-Guillarmod, Helv. Chim. Acta, 1990, 73, 1861. 28 B. H. Kim and J. Y. Lee, Tetrahedron Utt., 1992, 33, 2557. 29 B. H. Kim, J. Y. Lee, K. Kim and D. Whang, Tetrahedron Asymmetry, 1991, 2, 27 and 1359. 30 B. H. Kim and J. Y. Lee, Tetrahedron Utt., 1993, 34, 1609. 31 A. G. M. Barrett and H. G. Sheth, J. Chem. Soc, Chem. Commun., 1982, 170. 32 P. C. B. Page, J. F. Carefull, L. H. Powell and L O. Sutherland, / . Chem. Soc, Chem. Commun., 1985, 822. 33 R. A. Johnson and K. B. Sharpless, in Comprehensive Organic Synthesis, ed. B. M. Trost and I. Fleming, Pergamon, Oxford 1991, Vol. 7, ed. S. V. Ley, Ch. 3.2, pp. 389-436. 34 H. J. Karlsen, T. Katsuki, V. S. Martin and K. B. Sharpless, J. Org. Chem., 1981, 46, 3936. 35 S. Batmangherlich and A. H. Davidson, J. Chem. Soc, Chem. Commun., 1985, 1399. 36 J. K. Whitesell and D. Reynolds, J. Org. Chem., 1983, 48, 3548. 37 T. Honda, H. Ishige, J. Araki, S. Akimoto, K. Hirayama and M. Tsubuki, Tetrahedron, 1992, 48, 79. 38 T. Kametani, K. Kawamura and T. Honda, J. Am. Chem. Soc, 1987, 109, 3010. 39 S. Takano, S. Tomita, M. Takahashi and K. Ogasawara, Synthesis, 1987, 1116. 40 K. Takatori, N. Tanaka, K. Tanaka, M. Kajiwara, Heterocycles, 1993, 36, 1489. 41 G. Frater, U. Muller and W. Gunther, Tetrahedron, 1984, 40, 1269. 42 M. J. C. Buckle, I. Fleming and S. Gil, Tetrahedron Lett., 1992, 33, 4479 and references therein. 43 I. Fleming and N. J. Lawrence, /. Chem. Soc, Perkin Trans. 1, 1992, 3309. 44 R. A. N. C. Crump, I. Fleming, J. H. M. Hill, D. Parker, N. L. Reddy and D. Waterson, J. Chem. Soc, Perkin Trans. 1, 1992, 3277. 45 I. Fleming and P. E. J. Sanderson, Tetrahedron Lett., 1987, 28, 4229. 46 M. Ahmar, C. Duyck and I. Fleming, Pure Appl. Chem., 1994, 66, 2049. 47 I. Fleming, D. Higgins, N. J. Lawrence and A. P. Thomas, J. Chem. Soc, Perkin Trans. I, 1992, 3331. 48 J. P. Vigneron, M. Dhaenens and A. Horeau, Tetrahedron,
1973, 29, 1055. See also V.
Rautenstrauch, Bull. Soc. Chim. Fr., 1994, 131, 515. 49 T. Kogure and E. L. Eliel, J. Org. Chem., 1984, 49, 576. See also, M. M. Midland and J. Gabriel, / . Org. Chem., 1985, 50, 1144; C. S. Poss and S. L. Schreiber, Ace Chem. Res., 1994, 27, 9. 50 U. Ghosh, unpublished work. 51 I. Fleming and S. K. Ghosh, /. Chem. Soc, Chem. Commun., 1994, 2285. 52 I. Fleming and S. K. Ghosh, /. Chem. Soc, Chem. Commun., 1992, 1775. 53 P. D. Theisen and C. H. Heathcock, J. Org. Chem., 1988, 53, 2374 and 1993, 58, 142. 54 I. Fleming and S. K. Ghosh, J. Chem. Soc, Chem. Commun., 1994, 99. 55 I. Fleming, S. Gil, A. K. Sarkar and T. Schmidlin, /. Chem. Soc, Perkin Trans. I, 1992, 3351. 56 I. Fleming and S. K. Ghosh, J. Chem. Soc, Chem. Commun., 1992, 1777.
268
57 K. Yamamoto, T. Kimura and Y. Tomo, Tetrahedron Lett., 1984, 25, 2155. 58 D. A. Evans, K. T. Chapman and E. M. Carreira, /. Am. Chem. Soc, 1988, 110, 3560. 59 T. Mukaiyama, M. Usui and K. Saigo, Chem. Lett., 1976, 49. 60 R. D. Walkup and G. Park, J. Am. Chem. Soc, 1990, 112, 1597. 61 R. D. Walkup and G. Park, Tetrahedron Utt., 1987, 28, 1023. 62 J. Iqbal, A. Pandey and B. P. S. Chauhan, Tetrahedron, 1991, 47, 4143. 63 S. V. Ley, Chem. Ind. (London), 1985, 101. 64 K. M. Sun and B. Fraser-Reid, Can. J. Chem., 1980, 58, 2732. 65 H. Ohnii, G. H. Jones, J. G. Moffatt, M. L. Maddox, A. T. Christensen and S. K. Byram, J. Am. Chem. Soc, 1975, 97, 4602. 66 F. W. Eastwood, K. J. Harrington, J. S. Josan and J. L. Pura, Tetrahedron Lett., 1970, 5223. 67 H. M. R. Hoffmann, K. E. Clemens and R. H. Smithers, J. Am. Chem. Soc, 1972, 94, 3940. See also R. Noyori, S. Makino and H. Takaya, J. Am. Chem. Soc, 1971, 93, 1272. 68 E. LeGoff, / . Org. Chem., 1964, 29, 2048. 69 A. Warm and P. Vogel, Helv. Chim. Acta, 1987, 70, 690. 70 R. E. Ireland and J.-P. Vevert, / . Org. Chem., 1980, 45, 4259. R. E. Ireland and J.-P. Vevert, Can. J. Chem., 1981,59,572. 71 R. E. Ireland, R. H. Mueller and A. K. Willard, / . Am. Chem. Soc, 1976, 98, 2868. 72 H. Gerlach, K. Oertle, A. Thalmann and S. Servi, Helv. Chim. Acta, 1975, 58, 2036. 73 T. Endo, S. Ikenaga and T. Mukaiyama, Bull. Chem. Soc Jpn., 1970, 43, 2632. See also E. J. Corey, K. C. Nicolaou and L. S. Melvin, J. Am. Chem. Soc, 1975, 97, 653. 74 J. Gombos , E. Haslinger, H. Zak and U. Schmidt, Tetrahedron Lett., 1975, 3391. U. Schmidt, J. Gombos, E. Haslinger and H. Zak, Chem. Ber., 1976, 109, 2628. 75 I. Fleming and S. K. Ghosh, J. Chem. Soc, Chem. Commun., 1994, 2287 76 J. Inanaga, K. Hirata, H. Saeki, T. Katsuki and M. Yamaguchi, Bull. Chem. Soc Jpn., 1979, 52, 1989. See also: M. Hikota, Y. Sakurai, K. Horita and O. Yonemitsu, Tetrahedron Lett., 1990, 31, 6367. 77 J. Gombos, E. Haslinger, A. Nikiforov, H. Zak and U. Schmidt, Monatsh. Chem., 1975, 106, 1043. 78 T. Kaiho, S. Masamune and T, Toyoda, J. Org. Chem., 1982, 47, 1612.
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 18 © 1996 Elsevier Science B.V. All rights reserved.
269
Total Synthesis of Bioactive Natural Spiroethers, Tautomycin and Oscillatoxin D Akitami Ichihara, Hideaki Oikawa and H. Toshima 1. Introduction There are number of natural spiroethers which have cytotoxic and antitumor activities. Two such spiroethers, tautomycin and oscillatoxin Ds, were selected as target molecules of total synthesis, since it is known that tautomycin, which is one of okadaic acid class compounds, is the specific inhibitor of protein phosphatase and it could play an important role as a tumor promotor, while oscillatoxin Ds have antileukemic activity in the L-1210 cell line, and closely related analogs, aplysiatoxins, exhibit activity as tumor promotors. Therefore the study of the structure-activity relationships of these spiroethers can provide not only useful information on phosphorylation and dephosphorylation mechanisms at intraceller levels, but also about effective structural models for antitumor activity. In the total synthesis of tautomycin highly regio- and stereoselective reductions of the spiroketals have been developed in the synthesis of the spiroketal moiety. The construction of the spiroether units in oscillatoxin Ds has been achieved by a biomimetic pathway involving intramolecular aldol condensation and Michael type addition. The total synthesis provides a certain amount of oscillatoxin Ds which are difficult to obtain from natural sources. 2. Total Synthesis of Tautomycin In 1987, Isono and co-workers reported the isolation of tautomycin 1 from the culture filtrate of a soil fungus Streptomyces spiroverticillatus (1), an amorphous white powder exhibiting potent antifungal activity against Sclerotinia sclerotiolum (2) and inducing a morphological change (bleb formation) of human leukemia cells K562 (2). Since tautomycin enhances phosphorylation mediated by protein kinase C (PKC) in vitro (2), 1 was first assumed to be an activator of PKC, as phorbol dibutyrate. However, 1 does not compete with phorbol ester for binding to cells, and dose not activate PKC significantly in vivo (2). Later, Magae et al. (3) and MacKintosh et al. (4) independently found that enhanced phosphorylation caused by 1 due to the inhibition of protein phosphatase (PP), as found with the well-known tumor promoter, okadaic acid 2 (5, 6). Recently, the reversible phosphorylation of proteins has been recognized to be a major mechanism for the intracellular signal transductions in eukaryotic cells (7). The specific inhibitors of PP become a useful tool for studying such intracellular events. Tautomycin inhibited both type-1 and type-2A PP with IC50 of 22-32 nM (3, 4). Fujiki et al. proposed to classify 1 as belonging to the okadaic acid class of compounds which bind to okadaic acid receptors, PPl and PP2A, and inhibit their activities (8). Interestingly, all of those PP inhibitors (8) such as 2, dinophysistoxin-1, calyculin A, microcystins and nodularin are isolated from marine sponge and algae which are not available in large quantity. On the other hand, 1 is the only compound produced by a soil fungus which can be fermented in large scale. It therefore offers the great advantage that semi-synthesis of tautomycin analogues using a late
270
Tautomycin 1
Okadaic acid 2
Tautomycetin 3
intermediate derived from the degradation of natural 1 is possible once the total synthesis of 1 has been achieved. Isono's group also isolated the structurally related tautomycetin 3 from another soil fungus, Streptomyces griseochromogenes (9, 10). This antibiotic also shows antifungal activity and bleb formation in K562 cells (9). The similar biological activities of 3 to those of 1 strongly suggest that 3 is also a specific inhibitor of PP although this was not tested. The major structural difference in these antibiotics is present in the right hand portion of the molecules: tautomycin possesses a spiroketal moiety while this part of tautomycetin is replaced by a simple dienone. This suggests that the lefthalves of these antibiotics are essential for the inhibition of PP but the right-halves are variable. Furthermore, 1 and 2 show different affinities for PPl and PP2A (3, 4, 5) although the molecular size and partial structure of 1 are similar to those of 2. Thus, a systematic study of the structureactivity relationship of 1 would clarify the structural requirements for the inhibitory activity and enable us to design new specific inhibitors for PPl and PP2A. Our interests have therefore led us to develop an efficient synthesis of 1. In addition, the relative and absolute configurations of 1 (Id) have been determined mostly by NOE experiments and conformational analysis using MM2 calculations for the degradation products and their derivatives. Therefore, confirmation of the structure by total synthesis is necessary to secure its conclusion. In this article, we describe the first total synthesis of tautomycin 1(11, 12). 2-1. Chemical Property of Tautomycin Structurally, tautomycin 1 consists of a polyketide chain including a spiroketal moiety and a
271 unique acyl chain containing a dialkylmaleylanhydride unit. The name "tautomycin" came from the facile interconversion between the anhydride and the diacid (la, lb). Because of the steric congestion of the two alkyl substituents, the hydrolyzed diacid tends to regenerate the anhydride ring (eq. 1) (lb). Under nearly neutral conditions (pH 7.3), this moiety exists as two intereconverting anhydride and diacid forms in about 5:4 ratio (la, lb). Similar equilibration was found in an inhibitor of famesyltransferase chaetomellic acids (13) whose dialkylmaleylanhydride moiety is assumed to mimic a pyrophosphate group of famesylpyrophosphate. In analogy, one could speculate that the anhydride unit of 1 mimics the phosphate of a phosphorylated protein. Additionally, it was found that 1 and its derivatives with the anhydride unit were amenable to serious losses during their purifications on silica gel, possibly due to irreversible adsorption (14). This is one of the difficult problems rose during the handling of 1.
pH7.3
OH
O
H 0 2 C ^ ^
(eq.1) In the structure elucidation of 1, extensive degradations have been carried out by Ubukata et al. (Scheme 1) (la, lb). Alkaline hydrolysis of 1 with cesium carbonate at pH 9 gave the anhydride fragment 4 and anhydrodeacyltautomycin 5 which was further degraded with cesium carbonate at pH 10 to the enone 6 and the spiroketal 7 by retro-aldol cleavage. During the alkaline treatment at pH 9, C3 epimerization of 5 occurred (Id). These results indicated that dehydration of the C22 hydroxy group and epimerization at C3 are major problems and that operations at the final stage of the total synthesis must be carried out under neutral or acidic conditions. O
un
u
0
<
4 R = H 4a R = CH3 (13%. 2 steps)
Scheme 1
5 (64%)
(a) CS2CO3 at pH 9, HsO-MeOH; (b) cone. H2SO4; (c) CS2CO3 at pH 10, H20-MeOH.
Later, we found that a careful alkaline treatment of 1 with 0.3M K2CO3 in methanol at 3°C and methylation with diazomethane gave 5 and 8a in improved yields, thereby avoiding C3 epimerization under harsh conditions (cone. H2SO4) for esterification (Scheme 2) (14). Moreover, mild
272 methanolysis of 1 at 3°C with 4-dimethylaniinopyridine afforded the half-ester 9, and subsequent methylation with diazomethane gave the dimethyl ester 9a which could easily be purified by sihca gel chromatography, unlike 1 (12). Instead of methanol, the use of 2-(trimethylsilyl)ethanol slowly effected regioseiective anhydride opening to give the corresponding half ester 10 which was trapped with diazomethane to give the maleyl diester 10a. Meanwhile, no ring opening took place in the case of the bulky t-butyl alcohol. The half-ester 9 in chloroform slowly converted to 1. Thus, for the protection of the anhydride moiety, we chose the differently protected maleyl diester in which one of the alcohol groups was the acid sensitive group.
RO2C
OCH3
+
RO2C
8 R = H
5 (97%)
^i8a R = CH3 {90%, 2 steps) cord ,
R1O2C R2O2C
/ 9 Ri = CH3 or H, R2 = H or CH3 . / 10 Ri = H, R2 = CH2CH2TMS 9a Ri = R2 : CH3 (70%. 2 steps) ° A 10a Ri = CH3, R2 CH2CH2TMS (40%, 2 steps)
1 (77%)
12 R = TBS 12a R = H (47%) Scheme 2 (a) 0.3M K2CO3, MeOH; (b) CH2N2; (c) DMAP, MeOH, 3°C, 78%; (d) 2-(trimethylsilyl)ethanol, K2CO3; (e) TESOTf, 2,6-lutidine, CH2CI2, -35°C, 64%; (f) 47% HF-CH3CN-H2O (5:86:9), rt; (g) TBSOTf, 2,6-lutidine, CH2CI2, -75 -> 5°C, 70%.
Next, we investigated the protection of the three hydroxy groups in 1 (Scheme 2). Among the various protecting groups attempted, the silyl group was found to be suitable (12). Silylation of 1 with triethylsilyl triflate and t-butyldimethylsilyl triflate gave 11 and 12, respectively. While the former was deprotected with dilute HF to regenerate 1, the latter gave partially deprotected 12a whose C3'-t-butyldimethylsilyl group was resistant to acidic hydrolysis, and other conditions gave only degradation products. These results suggested that the Cs'-stereogenic center is sterically
273 hindered and triethylsilyl group or its equivalent is the protecting group of choice at the final stage. 2-2. Synthetic Plan Investigations on the chemical reactivity of 1 led us to focus on the following issues: (a) timing for the generation of the anhydride unit; (b) C3 epimerization; (c) C22-OH dehydration. For the first issue, we decided to construct the anhydride ring in the final reaction. Since introduction of the anhydride moiety by oxidation at a late stage while keeping other functionalities intact is expected to be difficult, we opted for hydrolysis of a differently protected maleyl diester, such as 13. The second problem can be solved by protection of the C2-carbonyl as an olefin (i.e.14) which is stable to various transformations and which can be converted to the corresponding methyl ketone by a Wacker type oxidation under neutral conditions. Finally, the most difficult third problem can be solved by direct assemblage of the C2J-C22 bond using two large subunits.
Tautomycin 1
DEIPSO s
V
O II
K
EtOzC
13
14 Scheme 3
Incorporating the considerations as shown above, retrosynthetic disconnection of the carbon backbone at the C21-C22 bond divides the target into two subunits, named the Left-wing 13 and the Right-wing 14 (Scheme 3). This rather bald disconnection also eliminates the difficulty of the O24acylation of the main chain and anhydride segment since facile hemiketal formation between C24hydroxy group and C20-carbonyl is expected to become a problem. The stereocontrolled aldol coupling of two key subunits is the key issue (synthetic challenge) of our synthesis since there is no established method for the stereochemically controlled coupling of these highly oxygenated segments. The Left-wing is further divided into dialkylmaleylanhydride segment 15 and the C22-C26 segment 16 (Scheme 4). On the other hand, retro-synthetically, the right wing is sectioned into C19-21 Cs-unit DEIPSO
O
'BUO2C, Et02C 13
15
Scheme 4
274
and C1-C18 spiroketal 17 which is further partitioned into the sulfone 18 and the aldehyde 19 (Scheme 5). The spiroketal 17 is synthetically equivalent to the degradation product 7. 0
^
f
OTES
0
1
"1
H
^^A^ 1
V^'%
^=^
^••H 14 17
{I SOaPh > r 5 OMOM : H
+
OHC
18
H RO^s
19 R = p-nitrobenzoyi
Scheme 5
For the right half of 1, we are aware that oxygenated carbons are present at exactly every five carbons and that the stereogenic centers are located near the oxygenated carbons as shown in partial structure A (Scheme 5). For synthesizing both segments 18 and 19, we planed to develop a new method for spiroketal reduction. At the beginning of our work, the stereochemistry at C15 was not settled; the possibility still remained to apply the method for the preparation of the Cu-Cig subunit. However, considering the eventually established structure of 1 possessing
\'h,\A-syn-\A,\'b-syn
stereochemistry which can be hardly prepared by our spiroketal reduction strategy, we decided to use another method for the synthesis of this segment. 2-3. Regie- and Diastereoselectivity of Spiroketal Reduction 2-3'L
Strategy of Spiroketal
Reduction
The chemistry of spiroketals, especially l,7-dioxaspiro[5,5]undecanes, is well-studied and reviewed in the hterature (15,16). Generally, the ratio of isomers of spiroketals may be controlled by several stabilizing factors such as stereoelectronic effects and 1,3-diaxial interaction of the substituents. Utilizing well-designed spiroketals, one can selectively prepare the most stable isomer by thermodynamic equilibration. Based on this idea, we planned to prepare a spiroketal represented by II which possesses several substituents with established (Cy and C5') and unestablished (€«) configurations. The spiroketal center and its a-position in II can isomerize to the more stable forms by steric effects caused by the established stereocenters on the ring. The prepared most stable isomer
275 is then subjected to reduction which gives another stereogenec center (Scheme 6). If the reduction proceeds regio- and diastereoselectively, this two-step process (thermodynamic equilibrationreduction) can be regarded as a formal remote chiral transfer. Using spiroketals as a template for manipulating functional groups on a ring, several excellent studies have been carried out in the total synthesis of natural products (17). To our knowledge, however, studies on the spiroketal reduction have been limited (18). In order to achieve this type of chiral transfer, we studied the reduction of spiroketals (19).
,''
\ Ri OH
H^
R2
reduction
illA
Mfol
H
IIIB
H
„
IIIC
HID
Scheme 6
reduction Y'
^ 20 21 22 23
Ri Ri Ri Ri
= H, R2 = H = CH3, R2 = H = CH3, R2 =CH20TBDPS = CH3, R2 = CHgOBn
24A 25A 26A 27A
Ri = H, R2 = H 24B R^ = H, Rj = H Ri = CH3, R2 = H 25B Ri = CH3, Rj = H R-, = CH3. R2 =CH20TBDPS 26B Ri = CH3, Rg =CH20TBDPS Ri = CH3, R2 = CH20Bn 27B R^ = CH3, R2 = CH20Bn
Scheme 7
From the examination of molecular models, we anticipated that the Ca methyl group of spiroketals 20, 21, 22 and 23 is large enough to interfere with the coordination of aluminum reagent
276 and Lewis acid at Oe' on these spiroketals. Since the reaction with alane reducing agent usually proceeds with retention of configuration (20), we expected type-A reduction products (24A, 25A, 26A and 27A) to be predominant in the reactions of DIBAH (eq 2). On the other hand, the several proposed mechanisms of silane-Lewis acid (SI-LA) reactions (21) suggested that we could change the stereochemical course by selecting a proper Lewis acid and an appropriate design of the substrates. For studying these reductions, the isomerically pure spiroketals 20, 21, 22 and 23 were synthesized (22,23) utilizing thermodynamic equilibration and subjected to reductions under several conditions (Scheme 7). The experimental results are summarized in Table 1. 2'3'2,
Spiroketal
Reduction
with DIBAH
As we expected, all cases (entries 1-4) in DIBAH reduction yielded type-A products predominantly with retention of configuration at the spiroketal center. In these reactions, selective cleavage of the C-Oe bond suggests that coordination of aluminum reagent predominantly occurred at the less crowded oxygen, Oe. As Yamamoto et al. proposed previously (20), the stereocontrol of DIBAH reduction may originate from the tight ion-pair complex, such as IM-1 (eq. 2), leading to rapid hydride transfer from the aluminum reagent to the oxocarbenium ion with retention. Yields of DIBAH reductions are normally satisfactory except for the sterically hindered ketal 22 (entry 3) which was less reactive and was converted to non-reduced products; the enol ethers 28, 29 were formed as major by-products under the reduction conditions. In the reduction of 23 (entry 4), the formation of a small amount of 27B suggested that chelation controlled reduction occurred to some extent (i.e. eq. 4) although such a reaction path was limited. Table 1. Reduction of a-methylspiroketals Entry
Spiroketal
Reagent
Yield (%)
DIBAH(a) DIBAH(a)
68
24A:24B:others
(94 : 2 : 4)(0
62 <10(d)
25A:25B: others
(100 : 0 : 0)(0
26A:26B:others 27A:27B:others 24A:24B:others
(100:0:0)(g)
1
20
2
21
3 4
22
5
23 20
DIBAH(a) DIBAH(a) Ph2SiH2-TiCl4(b)
6
21
Ph2SiH2-TiCl4(b)
7
22
8
22
Ph2SiH2-TiCl4(b) Et3SiH-SnCl4(^)
9
23
Ph2SiH2-TiCl4(b)
10
23
Et3SiH-SnCl4('^>
70
74 86
Reduction Products
(77 : 8 : 15)(g) (7 : 79 : 13)(0
86 81(e)
25A:25B:others
(100:0:0)(f)
26A:26B: others
(55 : 0 : 45)(g)
89
26A:26B: others
( 1 0 0 : 0 : 0)(g)
quant.
27A:27B:others
(19:81 :0)(g)
27A:27B:others
(0 : 100 : 0)(g)
(a) DIBAH (5 eq), toluene, -12°C ^ RT, (b) Ph2SiH2 (1.2 eq)-TiCl4(1.2 eq), CH2CI2, -70 -> -40°C (1.2 eq), (c) Et3SiH (1.2 eq)-SnCl4 (1.2 eq), CH2CI2, -94 -^ -60X. (d) Major by-products were enol ethers, (e) The products ratio was varied under minute experimental conditions, (f) Determined by GC-analysis on a 25-m OV-1 capillary column, (g) Determined by 270-MHz ^H-NMR analysis of the products mixture.
277
25A - 27A
21 - 2 3
28
2-3-3. Spiroketal
Reduction
(eq. 2)
29
with Silane-Lewis
Acid
The SI-LA reductions also proceeded exclusively with retention and afforded either type-A products (entries 6-8) or type-B products (entries 5, 9 and 10). From the results of the SI-LA reductions of spiroketals 21 and 22, it is assumed that coordination of Lewis acid occurs preferentially at Oe but severe steric hindrance on the rear side of the cleaving C-Og bond prohibits SN2-like attack of the silane reagents (i.e. IM-2A in eq. 3). Thus, the silanes can attack the spiro carbon only after ring opening to form a free oxocarbenium ion (21a) or a "separated ion-pair" intermediate (21b). For stereoelectronic reasons (24), the axial attack of hydride takes place predominantly in IM-2B (eq. 3). It should be noted that significant epimerization occurs at € « during the reaction at a higher temperature than at -20°C (25), which complicates the reaction products. This observation suggests that rapid ring opening and recyclization via IM-4 takes place at higher temperatures (eq. 5). The choice of the combination of silane and Lewis acid also effected the selectivity (entries 7 vs 8 and 9 vs 10). No reaction of 22 was observed with weaker Lewis acid (BF3-OEt2). In the reduction of the least substituted spiroketal 20 (entry 5), the major product was not the expected 24A but rather 24B which could be obtained from IM-3-like intermediate by axial attack (eq. 4). This discrepancy between prediction and experimental data requires further investigation. Switching the protective group from t-butyldiphenylsilyl group on 22 to the benzyl group on 23 resulted in a significant change of regioselectivity in the SI-LA reduction (entry 8 vs 10) (26). This remarkable change can be attributed to the selective chelation of Lewis acid. In the reaction of 23, the Lewis acid formed a chelation complex IM-3 from which product 27B was obtained by axial attack. The high regioselectivity (entry 10) shows that chelation control is apparently superior to steric control due to the CQ methyl group. Corcoran reported similar chelation controlled reactions of acetals (27) although no diastereoselectivity was observed. The foregoing results suggest that either type-A or type-B reduction products can be prepared at will, and both linear chain stereoisomers (epimeric at oxymethine carbon) might be prepared after ring opening of the tetrahydropyran (19). Using spiroketal templates, we achieved a formal double chiral transfer (Cy (and Cg-) to the spirocenter and Ca), respectively, with high regio- and diastereoselectivity by means of equilibration
278 and subsequent reduction. The application of the newly developed methodology shown above is described in the following section.
R'gR'SiH
21 or 22
25A or 26A (eq. 3)
R = H or CH2OTBDPS
23
:^
(eq. 4)
27B MXn
IM-3
H3C
a
IM-2A
Zi=
MXn
li
H3C1
IM-4
^AXOA
IA
(eq. 5)
H3CI
R = H or CH2OTBDPS
2-4. Synthesis of Right-wing 2'4'L
Synthesis of intermediate of racemic Cj-Cjo
segment
In order to examine the reduction of spiroketals, 22 and 23 were prepared (19) as racemic forms starting from the known kinetic iodolactonization product 30 (28) containing a 3:1 mixture of cis- and trans-isomers (Scheme 8). Since these isomers could not be separated at this stage, the following transformations were carried out of the mixture until spiroketal formation. Following the reported procedure (29), treatment of the lactone 30 with potassium benzyloxide afforded the epoxide 31 which was then hydrogenolized with concomitant epoxide opening to give the hydroxylactone 32. After protection of the hydroxy group with benzyl bromide and silver (I) oxide, reduction with lithium aluminum hydride furnished the diol 34. The primary hydroxy group was then selectively converted to the sulfide 35 using diphenyl disulfide and tri-n-butylphosphine. Protection of the remaining hydroxy group as an ethoxyethyl ether and oxidation with m-chloroperbenzoic acid furnished the sulfone 36. Condensation of 2-methyl-5-valerolactone and the sulfonyl carbanion
279 derived from 36 yielded the coupling product 37 as an isomeric mixture in 81% yield. Treatment of 37 with boron trifluoride etherate at 3 to 25°C effected deprotection of the ethoxyethyl group, spiroketalization and thermodynamic equilibration to afford the spiroketal 38 (30) as a single product (containing the C2 epimer of 38 originating from the trans-isomcv of 30). For this equilibration, Lewis acid treatment gave a better result in both yield and diastereoselectivity as compared to other protic thermodynamic equilibration conditions (15,16). Deprotection of the benzyl group in 38 and protection of the resultant alcohol with t-butyldiphenylsilyl chloride and imidazole gave the siloxyspiroketal 39 as crystals suitable for X-ray crystallography. Recrystallization successfully removed the undesired minor C2-epimer. The configuration of 39 was confirmed by X-ray analysis (19). Reductive elimination of the sulfone in 39 did not proceed under standard sodium-amalgam condition. Use of the lithium and monomethyl amine system at -78°C gave a reduction product 39 along with 30% of the ring opened ketodiol. However, use of a large excess of Raney-Ni (W-2) cleanly converted 39 to the desired isomerically pure spiroketal 22 in 73% yield. Desilylation and rebenzylation gave the benzyloxy spiroketal 23.
M
b, c
30
o^o^s^c 32 R = H 33 R = Bn
31 SOaPh
d. e. f. g . RO
34 X = OH, R = H 35 X = SPh, R = EE 36 X = SOgPh, R = EE SOgPh
37
SOzPh
j.k - o »
^O
^0
H
38
22
I, m, n ,0
H
22 R = TBDPS 23 R = Bn
39
reduction ,
23
26A
H
reduction k^O
OH 27B
Scheme 8 (a) KOBn, THF, -20°C, 79%; (b) Hg, Pd/C, EtsO; (c) BnBr, AggO, DMF, 6 1 % {2 steps); (d) LiAIH4, EtaO, 3-^25°C; (e) (PhS)2, n-BugP, Py, 3->25°C; (f) ethyl vinyl ether, PPTS, CHgClgi (g) mCPBA, NaHCOg, CH2CI2, 3->25°C, 45% (4 steps); (h) n-BuLi, 2-methyl-6-valerolactone, Et20-hexane, -80^25°C, 81%; (I) BF3»OEt2, CH2CI2, 3-»25°C, 75%; G) H2, Pd(0H)2/C, EtOH; (k) TBDPSCI, imidazole, DMF, 3-^25°C, 88% (2 steps); (I) Raney Ni (W-2), EtOH, reflux, 73%; (m) TBAF, THF; (n) BnBr, NaH, THF-DMF, 70% (2 steps).
280 As described in the foregoing section, highly diastereoselective reductions of 22 and 23 afforded 26A and 27B, respectively. Although 27B possessed all the required configurations for the synthesis of 1, the opening of the tetrahydropyran ring seemed to require developing a new efficient method. Thus, we chose 26A as a synthetic intermediate in which inversion of the €5 stereochemistry was needed. 2-4-2, Synthesis of chiral Cj-Cjo
segment (17)
Synthesis of the chiral Ci-Cio fragment was started from the aldehyde 40 derived from cisbutendiol (Scheme 9). Brown's crotylboration (30) of 40 using (-)-(E)-crotyldiisopinocampheylc, d, e OHO
OTBDPS
OTBDPS
40
45 X = SOgPh 46 X = SOPh
39
I. J OAc
22
H
H • OTBDPS
OAc
H
H •
47 X = OH 48 X = Br
26A 0, P, q
50 R = CH2OAC 51 R = CHpOH „ , . ^ ., , 19 R = CHO PNBz = p-nitrobenzoyi
49
51 52a (70%)
Scheme 9 (a) (-)-(E)-crotyldiisopinocampheylborane, Et20-THF, -78°C; 3M NaOH, H2O2, reflux; (b) MOMCI, i-Pr2NEt, CH2CI2, 61% (2 steps); (c) 9-BBN, THF; 3M NaOH. H2O2, 4-^25°C; (d) (PhS)2, n-BugP, Py; (e) mCPBA, NaHCOg, CH2CI2, 5->25°C, 94% (3 steps); (f) n-BuLi, 2-methyl-6-valerolactone, Et20-hexane, -65-^25°C, 70%; (g) TMSBr, CH2CI2, -30-»3°C, 89%; (h) Raney Ni (W-2), EtOH, reflux, 88%; (i) EtgSIH, SnCU, CHgClg, -78-^-60''C; AcOH, THF-HjO; (j) AcgO. Py, DMAP, CH2CI2, 78% {2 steps); (k) TBAF, THF; (I) MsCI, Py, DMAP, CHgClg; (m) LiBr, DMF, 70°C, 88% (3 steps); (n) Zn, EtOH-HgO, reflux, 87%; (o) p-nitrobenzoic acid, PPhg, DEAD, CQH^; (p) NaH, MeOH, 5°C, 54% (2 steps); (q) DMSO, (C0CI)2, CH2CI2, -78°C; EtgN, -78-^0°C, 88%; (r) K2CO3, EtOH, quant.; (s) AgsCOg-Celite, CgHe, reflux.
281 borane afforded the 5>'n-adduct 41 in high diastereo- and enantioselectivity. The alcohol 41 was protected as a methoxymethyl ether instead of an ethoxyethyl ether which was not stable under hydroboration conditions. Essentially the same procedure as shown above was used for the conversion of 42 to the sulfone 43. The coupling of lithiated 43 with 2-methyl-5-valerolactone was affected under similar conditions as those of the racemate to furnish the adduct 45 in 61% yield. In order to facilitate desulfurization, the sulfoxide 44 was also coupled with the lactone but this condensation under several conditions always gave low yields. The subsequent three-step conversion of 45 was carried out by treatment with bromotrimethylsilane to afford 39 as a single isomer in 88% yield (31). Deprotection of MOM group and spiroketalization proceeded below -30°C and then equilibration occurred at 0°C to room temperature. After desulfurization with a large excess of Raney Ni (W-2), the resultant spiroketal 22 was reduced by our newly developed procedure. Reduction of 22 with triethylsilane and tin(IV) chloride at -78 to -60°C was followed by acid hydrolysis of triethylsilyl ether to give chiral 26A in 98% yield. Completion of the synthesis of the Ci-Cio segment requires opening of the tetrahydropyran ring and inversion of the C5 configuration (Scheme 9). Acetylation of 26A was followed by desilylation with tetra-n-butylammonium fluoride to give the alcohol 47 which was converted to the bromide 48 by mesylation and substitution with lithium bromide in 88% yield. Reductive ring opening with zinc and acetic acid furnished 49 which was followed by Mitsunobu inversion (32) with p-nitrobenzoic acid and diethylazodicarboxylate and triphenylphosphine) to afford the diester 50. Selective hydrolysis of the ester 50 and Swern oxidation provided the aldehyde 19 in 42% overall yield in 4 steps. For the examination of coupling conditions, the lactones 52a and 52b were also prepared from the acetates 49 and 51, respectively. 2-4-3. Synthesis of Cij-Cjg
segment
We then moved to the synthesis of the Cn-Cig segment (Scheme 10) (29). The alcohol 53 (33) was converted to 55 by a standard C2 homologation procedure in 77% overall yield in 6 steps. Swern oxidation of 55 gave the aldehyde 56 which was submitted to Lewis acid catalyzed crotylstannylation (34). This reaction provided the expected eryr/zro-adducts 57a and 57b in a 3:1 ratio which were separated by medium-pressure silica gel chromatography. Protection of 57a as a MOM ether was followed by hydroboration with 9-borabicyclo[3.3.1]nonane to afford the alcohol 58 which was further converted to the sulfone 18 by the procedure shown above. In order to examine the coupling of the Ci-Cio and Cn-Cig segments and subsequent spiroketal formation, a model study was employed. The carbanion derived from 18 on treatment with n-butyl lithium was condensed with 5-valerolactone to give the adduct 59 in 59% yield. Desulfurization of the P-ketosulfone 59 with sodium amalgam gave 60 in modest yield. Spiroketal formation was then carried out. Treatment with bromotrimethylsilane gave the best result to furnish 61 as a single isomer in quantitative yield. At this stage, we could confirm the C13 and C14 configurations, which were installed by crotylstannation, using NMR data including NOE experiments. In addition, a good correlation of the ^H-NMR data of 61 with that of tautomycin 1 provided us further confirmation for assignment of the C15 configuration which was determined by
282 MM2 calculations of several C14-C15 rotamers in 1 and its C15 epimer (Id).
55 R = CH2OH 56 R = CHO
54
53 OH
I
• other isomers
57a
66:22
57b
OH
57a
, SOaPh
k, I
I.J
TBDPSO
58
'OMOM
18
X OMOM
59 X = SOaPh 60X = H
61
Scheme 10 (a) DMSO, (C0CI)2, CH2CI2, -70°C; EtgN, -70-^25X; (b) (EtO)2P(0)CH2C02Et, NaH. THF, -78-^25X, 98% (2 steps); (c) H2, Pd/C, EtOAc; (d) LiAIH4, Et20, 3-^25°C; (e) TBDPSCI, imidazole, DMF; (f) p-TsOH, MeOH, 79% (4 steps); (g) DMSO, (C0CI)2, CH2CI2. -78*C, EtgN, .78-^25°C; (h) trl-n-butylcrotylstannane, BF3*Et20, CH2CI2, -86->0°C, 92% (2 steps); (i) MOMCI, i-Pr2EtN, CH2CI2, quant.; (j) 9-BBN, THF, 3->25''C; 3M NaOH, H2O2, 3-425°C; (k) (PhS)2. n-BugP, Py; (I) mCPBA, NaHCOg, CH2CI2, 3->25'='C, 82% (3 steps); (m) n-BuLi, 5-valerolactone, Et20-liexane, -78->25°C, 59%; (n) Na(Hg), K2HPO4, MeOH, -20^25''C, 30%; (0) TMSBr, CH2CI2, -70^25°C, quant.
2-4-4. Coupling synthesis
of
of Ci'Cio
segment with Cn-Cig
segment and completion
of the
Right-wing
Coupling of the Ci-Cio segment with the Cn-Cig segment was next examined (Scheme 11). In the condensations of lactones 52a and 52b with lithiated sulfone derived from 18, no or low conversion to the product was observed in both solvent systems, non-polar toluene and ether-nhexane, which gave a satisfactory result in our similar coupling of 23 and 2-methyl-5-valerolactone. On the other hand, the reaction of aldehyde 19 with the sulfone carbanion proceeded smoothly to give adduct 62 which was inmiediately converted by Swem oxidation to 63 in 82% overall yield. The next reductive desulfonylation (36) proved to be difficult. Reduction of 63 with either aluminum amalgam (37) or n-tributyltin hydride (38) only gave uncharacterized reduction products in which the more sensitive nitro group were reduced. Finally, this problem was solved by use of samarium diiodide (39) affording the desired product 64 in 51 % yield. In this product, the nitro group was also reduced into the hydroxyamine (40). After hydrolysis of the benzoate group, cyclization to the spiroketal 65 was effected with bromotrimethylsilane as described above in 72% yield. Spiroketal 65 was then converted to the degradation product 5 in order to confirm the proposed structure of 5. The
283 silyl group of 65 was removed with tetra-n-butylammonium fluoride to give the alcohol 66 and subsequent oxidation with Dess-Martin periodinane (41) afforded aldehyde 17 in 80% yield. Wacker-type oxidation (42) of the alcohol 66 proceeded cleanly to give 67 which was finally converted to 5 by Swem oxidation in 94% overall yield. The synthetic material was identical with 5 derived from natural tautomycin in all respects. Thus, we unambiguously established the C3-C15 absolute configuration in tautomycin. , SOgPh
a, b TBDPSO
18
62 R = H, OH 63 R = 0
d, e, f, g
64 R = p-hydroxyaminobenzoyi
65 R = CH2OTBDPS 66 R = CH2OH 17 R = CHO
-^^
66
OHO,
67
Scheme 11 (a) n-BuLi, 19, Et20-hexane, -78-425°C; (b) DMSO, (C0CI)2, CH2CI2, -78X; EtaN, -78-> 0°C, 82% (2 steps); (c) Smig, THF-MeOH, -78°C, 51%; (d) K2CO3, MeOH, 60°C; (e) TMSBr, CH2CI2, -30->3X, 72% (2 steps); (f) TBAF. THF; (g) Dess-Martin periodinane, Py, CHjClg, 80% (2 steps); (h) O2, PdCl2, CuCI, DMF-H2O; (i) DMSO, (C0CI)2, CH2CI2, -78°C; EtgN, -78^0°C, 94% (2 steps).
0Ti(0'Pr)3
TBSO
17
'
(eq. 6)
68a
+
a I
OMgBr
TMSO
'
68b
For the introduction of the Cig and C19 stereocenters to the aldehyde 17, we first tested Heathcock's asymmetric anti-selective aldol reaction (eq. 6) (43). We anticipated that the stereocenters in 17 would have little effect on diastereoselectivity since the two methylene groups are
284 inserted between C15 and Cig. However, the reaction of 17 with the enolates 68a and 68b proceeded non-selectively to afford 69 as 3 isomeric mixtures in a 2:2:1 ratio. This is probably due to the other oxygen functionalities in 17 causing disorder of metal chelation structure in the transition states. Crotylboration (31,44) provided us a nice solution to this problem (Scheme 12). The reaction of 17 with Brown's (-)-(E)-crotyldiisopinocampheylborane afforded adducts 70a in very high diastereoselection but in modest yield (<50%). Meanwhile Roush's (E)-crotylboronate (44) also gave satisfactory selectivity (9:1) in 97% yield. Although Brown's reagent gave higher diastereoselectivity than Roush's, we chose the latter protocol because the reagent can be stored at -20°C for months and allows facile work-up (44). Introduced stereochemistries were tentatively assigned empirically based on reagent stereofacial selection. Further transformations of 70a needed differentiation between the two vinyl groups. Hence, a hydroxyl group-directed epoxidation of 70a with t-butyl hydroperoxide and vanadate (45) was employed to afford 71 in a 3:1 mixture of diastereomers. After protection of the hydroxy group as the triethylsilyl ether 72, reduction of the epoxide 72 was carried out. Use of lithium aluminum hydride caused concomitant deblocking of the silyl group. On the other hand, reduction with lithium triethylborohydride proceeded smoothly to give the alcohol 73 which was then oxidized with pyridinium dichromate (46) to the methyl ketone 14 in 86% overall yield.
70a
73
71 R = H 72 R = TES
14 Scheme 12 (a) (S,S)-diisopropyl tartarate (E)-crotylborate, MS4A, toluene, -78°C; (b) TBHP, V0(acac)2, CH2CI2; (c) TESCI, EtgN, CH2CI2, 78% (2 steps); (d) LiEtaBH, THF, 90%; (e) PDC, CH2CI2, 95%.
2'4-5, Alternative
synthesis
of Right-wing from
tautomycin
This segment was also obtained from anhydrodeacyltautomycin 5, formed by degradation of tautomycin 1, as shown in Scheme 13 (14). When we treated 1 with triethylsilyl triflate and 2,6lutidine at 3°C, the silyl enol ether 74 was obtained. Considering this finding, we decided to mask the C2 carbonyl as an olefin before the retro-aldol reaction to prevent C3 epimerizafion. After silylation of 5, treatment of the resultant silyl ether 75 with 1.2 eq. of lithium hexa-
285 methyldisilylamide was followed by adding A^-phenyltriflimide to afford the enol triflate 76 in 37% overall yield (47). As we anticipated, regioselective deprotonation occurred at Ci without affecting other functionalities. The palladium-catalyzed reduction of 76 with tributyltin hydride (47) gave the olefin 77 which was desilylated to provide 78 in 99% yield over two steps. Since retro-aldol reactions on 78 under basic conditions (20% cesium carbonate at pH 10 or lithium hexamethyldisilylamide, -20°C) gave products in only moderate yields (<30% of 6, and 30-80% of 17), we applied a thermal reaction (48) (toluene, 170°C) which afforded the retro-aldol products 6 and 17 in good yields.
1
TESOTf, 2,6-lutidine
(eq. 7)
•
CH2CI2, 3°C, 47%
TESO OCH3
76 R = OTf 77 R = H
Scheme 13 (a) TESOTf, 2,6-luticline, CHjClg, -50 -> -25°C, 87%; (b) LiHMDS, TfsNPh, THF, -78 -^ 0°C, 48%; (c) n-BugSnH, Pd(PPh3)4, LiCI, THF, reflux, quant.; (d) 47% HF-CH3CN-H2O (5:86:9), THF, 15°C, 99%; (e) toluene, 170°C (sealed tube)
The degradation products 6 and 17 were identical to the compounds in the reported (lb) and synthetic materials (1 la) in all respects. Thus, we prepared suitably protected synthetic intermediate 17 from 1 in a 26% overall yield (6 steps). An efficient degradative transformation of 1 into the synthetic intermediate 17 could be utilized for the synthesis of various analogues of this antibiotic.
286 2-5. Synthesis of Left-wing 2'5'L Synthesis of Anhydride
Segment n
OH
O
(eq. 8)
V o
u
un
u
u
u
u 0CH3
H3CO H3CO
4b
0 -K
0
0
H3CO
o OCH3
79
o\
o
0 0CH3
H3C0
.^-
0
0
H3C0
0CH3
o ] HaCO^Jl^ 80
0
Scheme 14
Until now, several methods to construct the dialkylmaleate moiety, synthetically equivalent to the dialkylanhydride, have been reported (49,50). In order to find a novel approach, we tested three routes featuring aldol condensation, 1,3-dipolar addition and oxidation of an alkylfuran. At first, we examined the 2ildol route. Based on a biosynthetic study of 1, the anhydride unit could be constructed by condensation of a-ketoglutarate with 2-methylmalonyl CoA (eq. 8) (51). Analogous aldol type reaction of P-ketoglutarate with a-diketo compounds was reported as the Weiss reaction (52). Thus, we applied this methodology to the anhydride synthesis. Our retrosynthetic analysis is shown in Scheme 14. Under buffered methanol at pH 6.8, condensation of dimethyl pketoglutarate with methyl pyruvate proceeded smoothly to afford a single isolable compound 81 in 32% yield along with 24% of the recovered dimethyl p-ketoglutarate (Scheme 15). Unfortunately, the 0 0
H3C0
0
0
JUUL».- - - A H3C0
±1f
citrate buffer (pH 6.8) MeOH, 60°C, 32%
R
"'-V^' NaBKI4. CeCl3 / 81 R = OH MeOH\, 70% ""^ 83 R = H
o o 82 Scheme 15
287 structure of the product 81 was fumarate not maleate 79 (53). Exclusive formation of 81 could be ascribed for considering the cyclized intermediate 82 prior to dehydration. The reduction products 83 and 81 themselves could not be converted to their ring opened forms due to facile recyclization. Thus, we abandoned this route. O
OH
O
u
u
u
CH3O
CH3O
CH3O
CH3O,
4a
84
85
OTHP
•=>
CH3O CH3O
88 87
Scheme 16
Another route involving 1,3-dipolar addition was next attempted (54). The key feature of this route is a cycloaddition of nitrile oxide (55) 88 with citraconate 87 (Scheme 16). Treatment of 89 prepared from 3-bromopropanol with phenyl isocyanate produced the nitrile oxide 88 which was then subjected to cycloaddition with dimethyl citraconate 87 to afford the adducts 86a and 86b as a 1:1 mixture. In this reaction, as expected, 86a, in which the less hindered oxygen is substituted at the quaternary carbon, was the major product (56). Conveniently, however, the undesired adduct 86b could be separated in the following dehydration step. These adducts were next subjected to various reduction conditions (57) in order to obtain intermediate 85 (eq.9). Unfortunately, we could not obtain any of the desired compound even under Curran's conditions (Raney-Ni (W-2), boric acid) which are mild enough to suppress undesired side reactions. All compounds produced in this reduction were retro-aldol products 90-93 probably derived from 85 or its imine form. This
a, b 88
^
H
3
C
0
^
^
HaCO^ O
CH3
86a
1:1
86b
Scheme 17 (a) AgNOg, EtgO; (b) DHP, PPTS, CH2CI2. 70% (2 steps); (c) PhNCO, 87, EtgN, CeHg, reflux, 62%.
288 O
NH2
HgCO
OTHP
H2 86a + 86b ,, , .^ , (1:1 mixture)
H3CO
OTHP
90 •
catalyst
^
91
^
^
,JU^.
H3CO
^ ^
OTHP
92
, ^, (eq. 9)
^
. >\ r ^^ ^ .
H3CO
OTHP
93
extremely facile retro-aldol cleavage from the sterically congested structure of 85. We did not explore this approach further although the formation of 81 via the aldol route suggested that there are some ways to avoid the decomposition. In the third route, we planned to synthesize the anhydride 4b via a furan which can be oxidatively converted to the anhydride (58). This route was tested using commercially available furan 94 (59) which had a striking structural similarity to 4b. Enantioselective reduction of 94 with Bmethyloxazaborolidine (S)-96 (60) and borane dimethyl sulfide complex gave the alcohol 95 in 50% yield. The enantioselectivity was determined by HPLC analysis using Chiralcel OB as 25% ee (61, 62). This result and the successful synthesis via Homer-Emmons condensation (vide infra) led us to stop further investigation of this route. Isobe's group successfully used a similar dialkylfuran route for synthesizing a tautomycin analogue (12).
/==:5f^^^-^0Et
o S^
{S)-96, BH3-(CH3)2S, CH2CI2, -20^-15°C, 50%
94
o \ ^
(eq. 10) ^^ ^
Phph
B CH3
96
With unfruitful results as shown above, we tried Sutherland's approach for constructing the anhydride segment via Homer-Emmons olefination (50) which was finally chosen in our synthetic study (Scheme 18). In this synthesis, differentiation of the three carboxyl group equivalents in the molecule is required. For this purpose, we planned to employ a condensation between triethyl phosphonoacetate and the a-ketoester 101. The synthesis of the anhydride segment 15 began with the alcohol 97 (Scheme 18). Oxidation under Parikh-Doering conditions (63) to the aldehyde followed by a Wittig reaction gave the a,Punsaturated ester 98 in 90 % yield. To set up the required chirality at €3% we selected the Sharpless asymmetric dihydroxylation protocol (64). Treatment of 98 with AD-mix-^ afforded (3'/?,4'5)-diol 99 in 99% yield, the enantiomeric excess being determined as 96% by ^H-NMR analysis of the corresponding bis-a-methoxy-a-trifluoromethylphenylacetic acid ester. The [i-hydroxy group was then selectively protected by oxidation with 2,3-dichloro-5,6-dicyano-l,4-benzoquinone (65a) to afford the benzylidene acetal 100. Subsequent oxidation with Dess-Martin periodinane gave the unstable a-keto ester 101, which was directly subjected to Homer-Emmons olefination (50) to
289 afford the maleate 102a preferentially in 67% yield along with the geometric isomer 102b (28%). When we used p-methoxybenzylidene acetal instead of 3,4-dimethoxybenzylidene acetal (65b) (eq. 11), the condensation proceeded in low yield (42%) and low diastereoselectivity (1.6:1). In this reaction, formation of the by-product 106 indicated decomposition of the acetal to benzaldehyde which then condensed with phosphonate anion. Fortunately, the use of 3,4-dimethoxybenzylidene acetal completely suppressed this undesired decomposition and improved the isomer ratio. Cleavage of acetal in 102a with 2,3-dichloro-5,6-dicyano-l,4-benzoquinone followed by protection with diethylisopropylsilyl chloride (66) and selective deprotection of the primary silyl ether gave 103 in 83 % overall yield. Sequential oxidation with Dess-Martin periodinane and then with sodium chlorite (67) provided the acid 15. Protection of the Cs'-hydroxy group with triethylsilyl group caused deprotection during the final oxidation. a, b
Since t-butyldimethylsilyl group (TBS) resisted acidic
'B^o^c^^jj^;.^^^^^^^
98
97
u
99 (96 % ee)
u
'BuOaC,
BuOzC.
c:
100R = a-OH, H 101 R = =0
102a (67%)
ODEIPS
g. h.
J. k
' B u O z C ^ ^ ^ A s ^ ^ CO2H
T
EtOzC''^
I, m , n
^.
r-
CH3O2C,
CHsOzC
8a
15
103
CO2CH3
Scheme 18 (a) SOg-Py, DMSO, CH2CI2. EtgN; (b) t-Bu02CCH=PPh3, CHgCig, 90 % (2 steps); (c) AD-mix-p, MeS02NH2, t-BuOH-H20, 0°C, 99 %; (d) DDQ, MS3A, CH2CI2, 5°C, 66 %; (e) Dess-Martin periodinane, Py, CH2CI2; (f) Et02CCH(CH3)PO(OEt)2, t-BuOK, THF, -60 -»-20°C (95 %, 2 steps); (g) PPTS, MeOH, 98 %; (h) DEIPSCI, Im, CH2CI2, 89 %; (i) ACOH-H2O-THF (4:1:4), 95 %; (i) Dess-Martin periodinane, Py, CH2CI2, 87 %; (k) NaCI02, 2-methyl-2-butene, t-BuOH, 91 %; (I) CH2N2, 85%; (m) see final section; (n) conc.H2S04, MeOH, 83%.
R 0 ^ 0
IF 104 R = MP 107R=DMP
Et02CCH(CH3)PO(OEt)2,
t- BuOK, THF, -60 -^-20°C
Eto^C. EtOzC
EtOaC'^^
105a R = MP (26%) 105b R = MP (16%) 106 (ca.40%) 108a R =DMP (38%) 108b R =DMP (15%) MP = p-methoxypheny, DMP = 3,4-dimethoxyphenyl
(eq.11)
290 deprotection in triTBS-protected tautomycin 12, diethylisopropylsilyl group was our choice as the C3'-protecting group. After methylation of 15 with diazomethane, anhydride formation (vide infra) and esterification with concentrated sulfuric acid in methanol gave 8a. This triester derived from 15 was identical with the degradation product 8a from natural tautomycin in all respects. 2'5'2,
Synthesis
of
Left-wing
Regioselective opening of the epoxide 109 (68) was conducted by employing titanium tetra-pmethoxybenzylalkoxide to give diols 110a and 110b in a 9.5:1 of isomeric mixture in 69% yield (69). Treatment of this mixture with p-toluenesulfonyl chloride and potassium hydride provided the epoxide 111 which was reacted with thiophenoxide to give the sulfide 112 in 74% yield. Undesired 110b was removed at the stage of oxirane formation. Methylation with sodium hydride and methyl iodide was followed by deprotection of the p-methoxybenzyl group with 2,3-dichloro-5,6-dicyano1,4-benzoquinone to afford the alcohol 16 in 98% yield. Acylation of 16 with the acid chloride derived from 15 or with 15 in the presence of 1,3-dicyclohexylcarbodiimide was unsatisfactory. However, 15 was esterified with the alcohol 16 by the Yamaguchi method (70) to afford 113 in nearly quantitative yield. Finally, oxidation to the sulfoxide followed by Punmierer reaction with trifluoroacetic anhydride and pyridine, and methanolysis furnished 13 in 81% yield. OH
.*^§
>Y"s4^^
OH
+
y
OH
Y^
'
110a
109
OH
OMPM
110b ODEIPS 'BuOzC^^Js,^^ Et02C^
c, d, e
r"
CO2H 15
SPh
OCHg
111
112 R = MPM 16 R = H DEIPSO
g. h Et02C
EtOzC
113
O
^
A 13
Scheme 19 (a) Ti(0MPM)4, toluene, 85°C. 69%; (b) p-TsCI, KH, THF, SOX, 91%; (c) NaOMe, PhSH, MeOH, 50°C, 81%; (d) NaH, Mel, THF-DMF, 3°C ^ r t , 98%; (e) DDQ, CH2CI2-H2O, quant.; (f) 2,4,6-trlchlorobenzoylchloride, EtgN, toluene; 15, DMAP, 60°C,94%; (g) Nal04, dioxane-H20. 97%; (h) TFAA.Py; NaHCOg, MeOH, 84%.
2-6. Completion of Synthesis Next, our attention moved to the crucial coupling of the Left-wing 13 and the Right-wing 14. Aldol reactions dealing with 7C-facial selectivity have been extensively studied in natural product synthesis (71). However, reliable prediction of the stereochemical outcome in this coupling cannot be
291 made because of the complexed nature of both segments. We initially examined the aldol reaction of metal enolates using model compounds 114-117 (Scheme 20). In order to determine the C22 configuration, one of the products 118b was converted to its benzylidene acetal 121. Detailed ^HNMR analysis led to the structure 121, as shown in Scheme 20. Among several different substrates, the obtained synlanti ratios were essentially the same (Table 2). The boron and chlorotitanium enolate (72) gave the desired (225)-adduct prefemtially (entries 5, 6, 8 and 10) but other enolates yielded the undesired (22/?)-adduct (the Felkin adduct, considering the methoxy group as a large group) predominantly. Considering the high diastereoselectivity, we proposed the cyclic transition state X in which the aldehyde adopts a conformation in which stabilizing dipolar interactions exist (eq. 12). Contribution of the chelation of the P-alkoxy group in the aldehyde with the metal enolate was less likely due to the necessity of an energetically unfavored boat-like transition state. In the case of boron and chlorotitanium enolates, Felkin model and P-chelation model in acyclic transition states could not completely excluded. Unfortunately, all attempts to reverse the selectivity using the metal enolate were unsuccessful. OR1
O
OR2
OR1
n-C9Hi9
114Ri=MPM 115 R, = Bz
116Ro = TES 117 R2 = BOM ^
•
OH
O
OR2
OR1
> ^ ^ ^ ^ ^ - ^ " ^ ^ n - C 9 H , 9
+
118aRi=MPM, Rg = TES ^19^ R^ = Bz, R2 = TES 120b Ri =MPM, R2 = BOM
> ^
OH
O
> ^ ^ - ^
OR2
> ^ ^
n-CsH
118b R^ = MPM, Rg = TES 119b R, = Bz ^2 = TES 120b Ri = MPM, R2 = BOM
MP OTES n-CgHic
Scheme 20
121
Table 2. Model aldoI reaction of a,p-diaIkoxyaldehydes and methyl ketones entry aldehyde methyl ketone
conditions
syn/anti ratio
yield (%)
1 2
114 114
116 116
LHMDS, THF, -78°C LHMDS, ClTi(0'Pr)3, THF, -78->-30°C
0:100 0:100
33 45
3 4
114
116
LHMDS, MgBr2«Et20, THF, -78X
0:100
114
116
LHMDS, ZnCl2, THF, -78°C
22:42
47 64
5
114
116
TiCU, Tr2NEt, CH2CI2, -78°C
1:1
80
6
114
116
9-BBN triflate, 'Pr2NEt, CH2CI2, -78°C
33:37
70
7
115
116
LHMDS, THF, -78°C
<1:3.7
<47
8
115
116
TiCU, ^Pr2NEt, CH2CI2, -78°C
19:21
40
9
114
117
LHMDS, ZnCl2, THF, -78°C
<5:18
<23
10
114
117
TiCU, 'Pr2NEt, CH2CI2, -78°C
1:1
36
292 L
H
XH OCH3
-M-..^
CH3O.
^
^
L
I
Y
1
(eq.12)
(22S)-adduct (desired)
(22R)-adduct
OCH3 122a
Ri
(eq. 13)
OTMS
In order to reverse the diastereoselectivity in the aldol reaction, the Lewis acid-catalyzed silyl enol ether addition (73) (Mukaiyama aldol reaction) was examined. Since the Mukaiyama aldol reaction is assumed to be proceeded via an acyclic transition state, a chelation controled aldol reaction of the a-alkoxy aldehyde should be possible (74). In the presence of TiCU, the silyl enol ether derived from 14 was reacted with aldehyde 13, followed by desilylation to afford the desired antiFelkin product 122a as a single adduct (Scheme 21). Based on precedents for chelation-controlled Mukaiyama aldol reaction (74), the exceptional high selectivity in this reaction would be accounted for by chelation of TiCU with the C23-methoxy group of the aldehyde 13 (eq. 13). On the other hand, when the lithium enolate derived from 14 was treated with the aldehyde 13, followed by desilylation, it gave a 1:4 ratio of the two epimers in favour of the undesired (225)-aldol product DEIPSO
EtOzC
O
^
1
^ 13
14 1) aldol reaction 2) HF-CH3CN-H2O
'BuOzC, EtOsC
122a (22R)-adduct (desired) 122b (22S)-adduct
Scheme 21
293 Table 3. Aldol reaction of Left-wing 13 and Right-wing 14 entry
conditions LHMDS, THF, -78X TiCU, ^•Pr2NEt, CH2CI2, -78°C TMSOTf, EtsN, CH2CI2, -78°C then TiCU, CH2CI2, -78^-15°C
yield (%) of 122a
yield (%) of 122b
7.5 15
30 48
54
0
122b. Using the titanium enolate under the conditions described by Evans (72), an increased amount of the (22/?)-adduct 122a was obtained in improved yield (combined yield 63 %) with essentially the same selectivity. Finally, the coupling product 122a was converted to 1 as shown in Scheme 22. Deprotection of the t-butyl group was initially investigated using the model compound 15a (eq.l4). Besides giving satisfactory results with protic acid such as trifuluoroacetic acid and dilute hydrofluoric acid, much milder conditions (trimethylsilyl triflate and 2,6-lutidine) recently developed by Evans et al. (71a) also effected smooth hydrolysis to afford 123. Pd-assisted selective oxidation of the terminal olefin of DEIPSO
I
' B u 0 2 C ^ ^ ^ ^ ' ^ \ ^ CO2CH3 'BUO2C,
TMSOTf(IOeq) 2,6-lutidine (10 eq)
V L X ' ^ ^ ^ CO2CH3
(eq. 14)
CH2CI2, rt (>80%) 15a
123
'BUO2C.
122a
EtOzC
124 (72%)
122b
a, b 125 (82%, 2 steps)
Scheme 22
(a) O2, PdCia, CuCI, DMF-HgO; (b) TESOTf, 2,6-lutidine, CH2CI2.
294 122a afforded the methyl ketone 124 (42). While deprotection of the t-butyl group with trifluoroacetic acid or aqueous HF was unsuccessful, even under Evans's conditions extensive decomposition occurred. Eventually, this problem was solved by use of milder Lewis acid triethylsilyl triflate. Thus, deprotection of the t-butyl group with triethylsilyl triflate and 2,6-lutidine and concomitant ring closure gave 1. In this reaction, the use of a limited amount of 2,6-lutidine (0.4 eq for triethylsilyl triflate) can avoid the silylation of other hydroxy groups. Our synthetic sample was identical in all respects with natural tautomycin. Under essntially the same conditions, 122b was converted to 22/?-epitautomycin 125 in 82% yield. In conclusion, the first total synthesis of tautomycin has been achieved via an efficient aldol coupling of two large subunits. Our synthetic route would provide an efficient way to prepare various analogues of 1 for biological evaluation. 3. Total Syntheses of Oscillatoxin D and 30-Methyloscillatoxin D Oscillatoxin D 126a and 30-methyloscillatoxin D 126b (75) are natural products derived from p-polyketides. They occur with aplysiatoxins 127a-127c (76), potent tumor promoters (77), in the marine blue-green algae belonging to the Oscillatoriaceae: Lyngbya majuscula,
Schizothrix
calcicola, and Oscillatoria nigroviridis. Their structures were mainly determined by spectral studies using EI-MS, IR, ^H-NMR, NOE, and CD spectra (75). They consist of a main spiroether moiety, corresponding to 1oxaspiro[5. 5]undec-4-ene-8-one ring system, (including seven asymmetric centers and possessing a (Z)-olefin, a mera-substituted phenol, and a P-ketocarboxylic acid functions) and a p-hydroxy-ylactone moiety. Oscillatoxin Ds are present as esters of both moieties. On the other hand, aplysiatoxins are present as macro bis-lactones including a spiroacetal. We are interested in the relationship between oscillatoxins and aplysiatoxins from a biosynthetic point of view.
OMe X
126a: Oscillatoxin D (R=H) 126b: 30-Methyloscillatoxin D (R=Me)
Fig. 1
127a: Aplysiatoxin (R=Me, X=Br, Y=OH) 127b: Debromoplysiatoxin (R=Me, X=H, Y=OH) 127c: Oscillatoxin A (R=X=H, Y=OH) 127d: 3-Deoxydebromoaplysiatoxin (R=Me, X=Y=H)
Structures of Oscillatoxin Ds and Aplysiato ins
The polyfunctional structures of aplysiatoxins and oscillatoxin Ds have received much attention as attractive synthetic targets and several synthetic studies have recently been reported (78). In connection with a series of our synthetic studies on aplysiatoxins, the total synthesis of 3-
295 deoxydebromoaplysiatoxin 127d has been completed (79). The synthetic 3-deoxy analogue 127d has been known to exhibit conGiparable activity as a tumor promoter (80). Debromoaplysiatoxin 127b and oscillatoxin A 127c show activity against P-388 leukemia in vivo (81). Oscillatoxin D 126a also displays antileukemic activity against the L-1210 cell line (82), however, its biological activity has been rarely examined because of the limitation of the natural sample for bioassay. Therefore, it is important to supply oscillatoxin Ds synthetically for biological investigations. In this article, we describe the first total syntheses of oscillatoxin D and 30-methyloscillatoxin D. The construction of the spiroether moiety has been achieved by biomimetic intramolecular aldol condensation and intramolecular Michael-type addition as key steps. Furthermore, some analogues of oscillatoxin D, which play an important role on the structure-activity relationship, can be prepared by our synthetic route. 3-1. Retrosynthetic Analysis
1. (i-elimination
^^ 2. Y-lactonization 27^0 OH
[A]
3. aldol-condensation 4. dehydration
Aplysiatoxins Oscillatoxin Ds
5. Michael-type addition
Q. Segment A -I-
Segment B + Segment C
"V
OP
•• ^
1 MeOO
Segment A
I
OMe
kJI QQ^ |
CHO
MeO
Segment B
Segment C
A Possible Transformation of Aplysiatoxins into Oscillatoxin Ds Retrosynthesis Scheme 23
We adopted a possible transformation of aplysiatoxins into oscillatoxin Ds in a retrosynthetic analysis (Scheme 23). Although aplysiatoxins exist as spiroacetals, they are equivalent to a
296 diketoalcohol [A] opening the acetal. The intermediate [A] may be converted into the intermediate [B] by two steps: p-elimination of the C27- carboxyUc acid to form the (Z)-olefin (step 1) and ylactonization between this carboxylic acid and the Cso-hydroxyl group (step 2). Aldol condensation between the C2 active methylene and the C7 ketone (step 3) and subsequent dehydration (step 4) may provide the intermediate [C]. Intramolecular Michael-type addition of the Cn-hydroxyl group to the C7 position (step 5) may complete the transfomation of aplysiatoxins into oscillatoxinDs. In the spiroetherification, it is important to control not only the configuration of the spiro center (C7) but also the conformation of the spiroether ring system, from a synthetic point of view. We expected to solve these problems by utilizing the thermodynamic equilibrium of the spiroether. Therefore, a practical retrosynthetic analysis was made for the intermediate [B], which was divided into three main segments, A (C8-C23), B (C1-C7, C24, C25, C26), and C (C27-C30 or C31). We chose an acetylenic compound as segment A, because its acetylide would act as a good nucleophile for an aldehyde (segment B), and the acetylenic bond would be selectively reduced to a (Z)-olefin at an appropriate step after coupling. The chiral pool method was applied during the synthesis of each segment. 3-2. Syntheses of Three Optically Active Segments 3-2-Jf. Synthesis
of Segment A
The C8-C23 segment of aplysiatoxins was synthesized as an alkyl iodide corresponding to 137 from D-glucose and 3-hydroxyacetophenone in the total synthesis of 3-deoxydebromoaplysiatoxin (79). The C1-C6 carbon unit of D-glucose was incorporated in the Cg-Cn skeleton of aplysiatoxins with sequential four asymmetric centers. The asymmetry of the C15 benzylic position was introduced by a diastereoselective reduction using a chiral reducing reagent. We modified this route and an acetylenic function was newly introduced to segment A. The first task involved the introduction of two methyl groups at the C3 and C5 positions of Dglucose with inversion of those configurations. A known tosylate 128 was readily synthesized in 6 steps from D-glucose (Scheme 24) (83) and possessed the required asynmietry at the C3 position. Treatment of 128 with sodium benzyl alcoholate gave the benzyl ether via an epoxide as an intermediate. The resulting secondary hydroxy 1 group was substituted with inversion by a chlorine atom to provide a chloro benzyl ether 129 in two steps. Acidic hydrolysis of 129 followed by sodium borohydride reduction gave an acycHc triol, which was again protected with 2, 2dimethoxypropane to give the 1, 3-acetonide 130 and the 1, 2-acetonide 131 in 50% and 46% yields, respectively. The desired 131 was further obtained by the acidic isomerization of 130 in almost the same yield and ratio as that obtained by the protection of the triol. Treatment of 131 with sodium methoxide gave an epoxide 132 in quantitative yield. Methylation of 132 with lithium dimethyl cuprate gave an alcohol opened at the C5 position as the main product in 11:1 regioselectivity. The undesired regioisomer was separated off by a silica gel column chromatography. Protection of the secondary hydroxy 1 group as a silyl ether followed by deprotection of the primary benzyl ether by hydrogenolysis gave an alcohol 133. At this stage, the first task was completed.
297
D-glucose
6 steps
a, b, c TsO
>^ I ' 0 '
BnO X t ^ O '
128
°)C
129
ci,e,f
134
OBn
135
OBn
Scheme 24 (a) BnONa / BnOH-THF; (b) MsCl, pyr. / CH2CI2; (c) n-Bu4NCl / PhH, 89%, 3 steps; (d) Amberlite IR-120 (H"^) / dioxane-H20; (e) NaBH4 / MeOH-H20, 88%, 2 steps; (f) 2,2-dimethoxypropane, p-TsOH I acetone , 130: 50%, 131: 46%; (g)p-TsOH / acetone; (h) NaOMe / MeOH, 100%; (i) MeoCuLi / Et20; (j) TBDMSOTf, EtsN / CH2CI2; (k) H2, 10 % Pd-C / MeOH, 81 %, 3 steps; (1) (COCl)2, DMSO, EtsN / CH2CI2, (m) Ph3P=CHC0N(0Me)Me / CH2CI2; (n) H2, 5 % Pd-C / EtOAc; (0) LiC6H4(w-OBn) / THF, 86%, 4 steps; (p) (-)-DIPCl / THE; (q) Mel, NaH / THE; (r) n-BuNF / THE, 74%, 3 steps.
The second task is the elongation of the carbon chain including the incorporation of an aromatic ring and introduction of asymmetry at the benzylic position. Swem oxidation of 133 gave an aldehyde, which was subjected to Wittig reaction with N-methoxy-N-methyl-2(triphenylphosphoranylidene) acetate (84). After catalytic hydrogenation, a saturated amide was obtained, which was treated with a 3-benzyloxyphenyl lithium (generated from the corresponding bromide with n-butyl lithium) to provide a ketone 134. For the diastereoselective reduction of the benzylic ketone 134, an asymmetric reducing reagent, Brown's (-)-diisopinocampheylchloroborane (this reagent provides 5 configuration) (85) was applied to give a sole diastereomer. Using ^HNMR (400MHz) spectroscopy, it was compared with the diastereomixture at the C15 position (1:1) prepared by the other method. The purity could be determined based on the methyl signals of the TBDMS group and the reduced product showed only two singlet methyl signals (86). After methylation of the resulted secondary hydroxyl group, in order to confirm the absolute stereochemistry at the C15 position, the benyl ether was converted to the corresponding phenol. The CD curve ([6]260 +404° in EtOH) of the phenol was quite similar to that of debromoaplysiatoxin (127b), thereby confirming the absolute stereochemistry. Deprotection of the
298 silyl ether gave a secondary alcohol 135 which isomerized to a primary alcohol 136 under the same acidic conditions (Scheme 25) as described in the recycle of 130 to 131. The third task was the introduction of an acetylenic function and the favorable choice of a protective group at the Ci i-hydroxyl group. Treatment of 136 with triphenylphosphine in refluxing carbon tetrachloride gave the chloride 137 in quantitative yield. Then 137 was subjected to the base induced elimination (87) to provide an acetylenic alcohol 138 in 72% yield. The Cn-hydroxyl group, which was involved in the construction of a spiroether, was protected as a silyl (segment Al), a methoxymethyl (segment A2), and a tetrahydropyranyl (segment A3) ethers. In practice, the protective group was exchanged successively for some synthetic operations as described hereinafter. In this way, the synthesis of segment A was completed (88).
O
OH
OMe
0
135
OBn
137
OBn
0
OMe
138
OBn
OMe Segment Al: P=TBDMS Segment A2: P=MOM Segment A3: P=THP
d,e, f OBn
Scheme 25 (a) p-TsOH / acetone, 135: 30%, 136: 70%; (b) PhsP / CCI4, 99%; (c) LDA / THF, 72%; (d) TBDMSOTf, Et3N / CH2CI2, 100%; (e) MOMCl, i-PrjNEt / CH2CI2, 87%, (f) DHP, PPTS / CH2CI2, 97%.
3'2'2.
Synthesis of Segment B The C4 position is the only asynmietric center in segment B. We used methyl (5)-3-hydroxy-
2-methylpropionate as the starting material, which was readily converted into the known alcohol 139 in two steps (Scheme 26) (89). Tosylation of 139 with tosyl chloride and pyridine followed by substitution with sodium iodide gave an alkyl iodide 140. Treatment with iodine, triphenylphosphine, and imidazole in benzene also gave 140 directly in good yield. The lithium enolate of ethyl isobutylate, as the four carbons unit, was alkylated with 140 to provide 141 in quantitative yield. Deprotection of the benzyloxymethyl ether by hydrogenolysis, and Swern oxidation of the resulting alcohol, followed by thioacetalization with 1, 3-propanedithiol gave a 1, 3-dithiane derivative 142. The ester group of 142 was reduced with lithium aluminum hydride to provide an alcohol, which was protected as a r-butyldiphenylsilyl (TBDPS) ether 143. Further
299 elongation by two carbons unit, including the generation of a masked aldehyde, was achieved by treatment of the lithiated 143 with l-bromo-2, 2-dimethoxyethane to afford 144 in quantitative yield. Deprotection of the TBDPS ether followed by oxidation using an activated dimethylsulfoxide method gave an aldehyde, segment B. This segment has one free aldehyde group required for coupling with segment A and two masked carbonyl synthons which can be employed to generate an activated methylene system for intramolecular aldol condensation. Methyl (5)-3-hydroxy2-methylpropionate
a, b
c,d (ore) 140
139
g, h, i
BOMi
COOEt 141
j,k
COOEt
142
OTBDMS
OTBDMS
143
Scheme 26 (a) BOMCl, /-Pr2NEt / CH2CI2, 100%; (b) LiAlH4/ Et20, 93%; (c) TsCl, pyr. / CH2CI2; (d) Nal / acetone, 86%, 2 steps; (e) I2, PhsP, imidazole / benzene, 82%; (f) LDA, ethyl isobutyrate / THF, 100%; (g) H2, Pd-black / EtOH; (h) (C0C1)2, DMSO, EtsN / CH2CI2; (i) 1,3-propanedithiol, BF30Et2 / CH2CI2, 85%, 3 steps; (j) LiAlH4, / Et20; (k) TBDPSCl, imidazole / DMF, 91%, 2 steps; (1) t-BuLi / HMPATHF, then l-bromo-2,2-dimethoxyethane, 99%; (m) n-Bu4NF/THF; (n) SOspyr., DMSO, Et3N/CH2Cl2, 86%, 2 steps.
3-2-5. Synthesis
of Segment C
p-Hydroxy-Y-lactone moieties, segment CI and segment C2, are known compounds whose syntheses have been reported. Therefore, we followed the reported methods and could readily synthesize both segment Cs. Segment CI was synthesized from (/?)-malic acid as a starting material in the following three steps (Scheme 27): (I) diesterification with acetyl chloride in methanol, (II) selective reduction of a-hydroxy ester with boran dimethylsulfide complex and sodium borohydride in THF, and (III) acid catalyzed lactonization of 145 (90). Segment C2 was synthesized from methyl (/?)-lactate as a starting material in the following three steps: (I) acetylation with acetic anhydride and pyridine, (II) intramolecular Claisen condensation of 146 using lithium bis-(trimethylsilyl) amide, and (III) hydrogenation of the enolic olefin 147 in the presence of rhodium catalyst. The required diastereomer (segment C2) was
300 produced in 9:1 diastereoselectivity and was readily separated by a silica gel column chromatography (91).
(/?)-Malic acid
DH
OH
a, b
J:S
145
Segment CI
OH
HoC COOMe
Methyl (/?)-lactate
-
0
^
146
0
.OH
^
147
Segment C2
Scheme 27 (a) AcCl / MeOH; (b) BHgSMes, NaBH4 / THF; (c) TFA / CH2CI2; (d) AC2O / pyridine; (e) LiN(TMS)2 / THF; (f) H2, 5% Rh-C / EtOAc.
3-3. Construction of Spiroethers Ring System 3'3'1,
Synthesis
of Cyclohexenone
Derivatives
by Intramolecular
Aldol
Condensation-Dehydration At first, segment Al (P=TBDMS) was applied for couping with segment B prior to subsequent manipulation. The lithium acetylide of segment Al generated with «-butyl lithium was successfully coupled with segment B to give the diastereomeric alcohol 148a (97% yield, diastereoselectivity: ca. 1:1), which possessed the required C1-C26 carbon skeleton of oscillatoxin Ds (Scheme 28). Oxidation of the secondary hydroxyl group followed by deprotection of the thioacetal with N-chlorosuccinimide gave a diketone 149a (68% yield, 2 steps). When the ^H-NMR of 149a was observed in chloroform-d, there was a slight singlet signal assigned to an aldehyde in addition to its own signals. We considered that the aldehyde signal might be due to the formation of a cyclohexenone derivative 150a. This indicates that intramolecular aldol condensation-dehydration takes place after hydrolysis of the dimethylacetal in a weak acidic medium. Therefore, we examined various acidic conditions in order to hydrolyze the dimethylacetal and further to cyclize to 150a. For example, transacetalization with a catalytic amount of /?-toluenesulfonic acid in acetone gave only a mixture of enols derived from the resulting 1, 3-dicarbonyl system. Fortunately, we could find the optimum condition, which was treating with 50% aq. trifluoroacetic acid / chloroform (1:5) (92), to give 150a in 67% yield. A deprotected product 150d was also obtained in 29% yield. This intramolecular cyclization is regarded as a biomimetic reaction corresponding to the conversion from the intermediate [B] to [C] in Scheme 23. Oxidation of the aldehyde 150a with sodium chlorite to the corresponding carboxylic acid, followed by treatment with diazomethane, provided a stable methyl ester. Partial hydrogenation of the acetylenic linkage in the presence of Lindlar catalyst in ethyl acetate then gave a (Z)-olefin 151a in 80% yield (3 steps from 150a). The TBDMS group was found to be an effective protective group through this sequence of reactions, however, it resisted deprotection since the C n hydroxyl group is sterically hindered by two neighboring methyl
301 groups, and furthermore, the TBDMS group itself is a rather bulky group. The approach of reagents would therefore be restricted, principally, due to steric hindrance of substrate 151a. We next chose an acetal-type protective group, among which methoxymethyl ether is the smallest one and can be deprotected under acidic conditons. The lithiated segment A2 (P=MOM) was coupled with segment B to give 148b in a 93% yield. Oxidation of 148b, followed by deprotection of the 1, 3-dithiane proceeded without difficulty to give 149b in 74% yield. Deprotection of the dimethylacetal and subsequent intramolecular aldol condensation-dehydration was achieved under the same acidic condition as the cyclization of 150a. The desired cyclohexenone 150b and the deprotected one 150d were hence obtained in 56% and 27% yields, respectively. The MOM group compared with the TBDMS group did not influence the yield appreciably until this step. Conversion into a methyl ester, followed by introduction of the (Z)olefin provided 151b in 49% yield (3 steps). The yields in these sequential steps were relatively lower than in the TBDMS series. Deprotection of the MOM ether and subsequent intramolecular Michael-type addition from 151b was possible in one pot.
CHO
Segment A1: P=TBDMS QBH Segment A2: P=MOM Segment A3: P=THP
Segment B
OBn MeO
U °" ^^
OP
OMe
148a: P=TBDMS (97%) 148b: P=MOM (93%) 148c: P=THP (93%)
OBn
149a: P=TBDMS (68%) 149b: P=MOM (74%) 149c: P=THP(72%) OBn
150a: P=TBDMS (67%) + 150d: P=H (29%) 150b: P=MOM(56%) + 150d(27%) 150c: P=THP (8%) + 150d (78%)
OBn 151a: P=TBDMS (80%, 3 steps from 150a) 151b: P=MOM (49%, 3 steps from 150b) 151d: P=H (54%, 3 steps from 150d)
Scheme 28 (a) «-BuLi / THF then Segment B; (b) SOspyr., DMSO, EtsN / CH2CI2; (c) NCS, AgNO^ I 10% aq. CH3CN; (d) 50% aq. TFA / CHCI3 (1:5); (e) NaC102, NaH2P04, 2-methyl-2-butene /t-BuGH - H2O (4:1); (f) CH2N2 / Et20; (g) H2, Lindlar cat. / EtOAc
The deprotected product 150d was also oxidized to the corresponding carboxylic acid without affecting its hydroxy! group. Treatment with diazomethane, followed by partial hydrogenation.
302
gave ISld in moderate yield. From this result, the dianion of the free alcohol 138 was used for coupling with segment B to give a diol. Selective oxidation of the propargyl alcohol with activated manganese dioxide instead of activated DMSO provided a ketone, which could be further converted into the cyclohexenone 150d. Since the overall yield of these four steps was only 13% yield, we examined the other protective group. It was proved that using a protected substrate gave a better result than using a non-protected one. Tetrahydropyranyl group, which might be readily deprotected under weakly acidic conditions, was next introduced to 138. Segment A3 (P=THP) was converted into the diketone 149c in good yield. By treating of 149c in 50% aq. TFA / CHCI3 (1:5), the desired compound 150d was obtained as the main product in 78% yield in one pot. Hydrolysis of the dimethylacetal, intramolecular aldol condensation-dehydration, and deprotection of the THP group proceeded successively. The THP group functioned as the most versatile protective group among the ones used. In this way, the direct precursor of intramolecular Michael-type addition was efficiently synthesized from segment A3 and segment B. 3-3-2, Synthesis
of Spiroethers
by Intramolecular
Michael-type
Addition
The cyclohexenone derivatives (151a and 151b) could be converted to the direct precursor 151d of intramolecular Michael-type (1, 4-conjugate) addition by deprotection. Various inter- and intramolecular 1, 4-conjugate additions by heteronucleophiles have been reported and achieved under not only basic but also acidic or nearly neutral conditions (93). Hence, on our substrates (151a and 151b), deprotection of the TBDMS or the MOM groups and the subsequent cyclization might proceed in one pot under appropriate deprotective conditions. At first, we attempted the deprotection on the TBDMS protected 151a under more than ten conditions. For example, tetra-n-butylanmionium fluoride in THF, the most popular condition which is nearly neutral, did not provide 15Id or a spiroether, but only led to the decomposition of 151a. The acidic conditions used for intramolecular aldol condensation gave a little 15Id with decomposition. No reaction took place under some conditions because of the steric hindrance. Hydrogen fluoride in aq. CH3CN gave 151d in 38% yield accompanied by spiroethers in 8% yield, as shown in Table 4, that was the best result among the attempted conditions on 151a. From this result, we obtained a significant information, namely, that the intramolecular conjugate addition would proceed under acidic conditions. We next explored the deprotection reactions on the MOM protected 151b. Among the several proton and Lewis acids employed in attempts for deprotection, treatment of 151b in cone. HCl / MeOH provided a mixture of four spiroethers (152,153, 154, and 155) which was not accompanied by the deprotected 151d. The stereoisomers were readily separated by preparative TLC on silica gel. We further applied the same acidic conditions on the non-protected 151d, and got a better result from the standpoint of the total yield of four stereoisomers as shown in Table 4. We considered that the four stereoisomers are under thermodynamic equilibrium under these conditions. The spiroether 152 had the same stereogenic centers as oscillatoxin Ds and the three steroisomers (153, 154, and 155) could isomerize to the natural form 152 under their cyclization conditions. The determination of
303 stereochemistries and conformations of the four spiroethers is discussed in the next section. On the other hand, treatment of 151d with potassium r-butoxide in THF gave only two spiroethers 153 and 155 in 43 and 24% yields, respectively. This results may reflect a kinetic controlled cyclization, so that the synthesis of the natural form must be achieved under thermodynamic equilibrium conditions. The mechanism of spiroetherification is also discussed in the next section. TABLE 4. Synthesis of Spiroether by Intramolecular Michael-Type Addition Substrate
Spiroether^ (Yield^
Condition 152
153
154
155
151a
47%HF/aq. ChsCN
8^^ (total yield of four isomers)
151b
conc.HCl/MeOH
10
15
5
14
151d
conc.HCl/MeOH
26
26
6
20
151d
f-BuOK/THF
0
43
0
24
a. The structures of spiroethers are shown below. b. Isolated yield after purification by preparative TLC c. The deprotected product 26d was obtained in a 38% yield.
MeO
MeO
152 (Natural form: 25, 4R, 7R) OBn
153 (7S-isomer)
154 (45-isomer)
155 (2R, 45-isomer)
In this way, the construction of the C1-C26 methyl ester 152, lacking the y-lactone moiety, has been accomplished by an intramolecular Michael-type addition, corresponding to the conversion from the intermediate [C] to oscillatoxin Ds as shown in scheme 23. This reaction may be regarded as a biomimetic reaction. 3-3-3, Stereochemistry
and Conformation
of
Spiroethers
Oscillatoxin Ds have eight or nine asynmietric centers, most of which are concentrated on the spiroether ring system. The relative sterochemistries, determined by ^H-^H NOE experiments, are 25*, 4/?*, 7/?*, 105*, 11/?*, and 125*. Since the cyclohexanone ring has three quaternary carbons (C3, Q , and C7), ^H-^H coupling dose not provide any information of significance about the relative stereochemistry. The absolute stereochemistry at C15 was determined to be 5 from the CD spectrum. In addition to NOE experiments, acid hydrolysis of aplysiatoxins produced the same ylactones which existed in oscillatoxin Ds and the absolute configurations at the corresponding centers in oscillatoxin Ds were most likely the same as those in aplysiatoxins. Therefore, the
304
absolute sterochemistries at all the centers of oscillatoxin Ds are 25", 4/?, 7^, 105, IIR, 125, 155, 29/?, and 30R (75). During the intramolecular Michael-type addition of 15Id, the stereochemistry at the two newly produced stereogenic centers (C2 and C7) must be controlled exactly. We obtained four stereoisomers of spiroethers whose stereochemistries were determined through the detailed analysis of NOE difference spectra as well as comparison to other natural products. The NOE results for the various stereoisomers are shown in Fig. 2. Successive irradiations of the C25 axial methyl proton (5 1.22) and then of the C24 equatorial methyl proton (5 0.88) gave positive NOEs for the H2, H4, H24 protons, and for the Hio, H25 protons, respectively. Furthermore, irradiation of the Hio proton gave positive NOEs for the C22, C23, C24 methyl protons. These results were closely similar to those of natural oscillatoxin Ds, and meant that the H2 and H4 protons had to be located in axial positions, and the ether oxygen had to be attached equatorially to the cyclohexanone ring. Therefore, 152 was assigned as the natural form, possessing 25, 4R, IR configurations.
H (irradiated)
152
153
154
155
NOE
H25
H(enhanced) H2, H4, H24
H24
Hio. H25
Hio
H22. H23, H24
H25 H2
H2, H4, Hgeq, Hg, H24
H4 H11
H2, H25, H26
H24 H11
H4, He, H25. COOMe
H25 H24
Hio. H24 H2, H4, He, H25
He
H2. H24
H4, He, H25 H24
H2.
Fig. 2 The NOE Data of Spiroethers The stereoisomer 153 exhibited NOE enhancements of the Hg olefinic proton in addition to the H2, H4, Hseq., H24 protons when the C25 axial methyl protons (6 1.20) were irradiated. Positive
305 NOEs for the Hg proton based on irradiation of the H2 proton and for the C24 methyl protons (5 0.89) based on irradiation of the H n proton indicated that the double bond had to be attached equatorially in contrast to 152. Hence, it was confirmed that the stereoisomer 153 was the corresponding C7 epimer of the natural form 152. Irradiation of the C24 axial methyl protons (5 1.12) in the stereoisomer 154 gave positive NOEs of H4, Hg, H25 (5 0.87), and ester methyl (on C2) protons. This meant that the H4, H24, and the methoxycarbonyl group had to be attached axially so that the stereochemistry at C4 was inverted to an S configuration. NOE enhancement of the H2 proton by irradiation of the Hn proton, indicated that the double bond existed equatorially while the ether oxygen was axial. Hence, the conformation of the cyclohexanone ring of the C4 epimer 154 flipped in contrast to those of 152 and 153. Irradiation of the C25 equatorial methyl protons (5 0.89) of the stereoisomer 155 gave positive NOEs for the Hio and H24 protons. In addition to the H4, Hg, H25 protons, the H2 proton was enhanced instead of the ester methyl proton, when the C24 axial methyl protons (6 1.23) were irradiated. Therefore, the methoxycarbonyl group of 155 had to be attached equatorially to the cyclohexanone ring, and the other stereochemistries were identical with those of 154. In all steroisomers, the axially oriented methyl group in the g^m-dimethyl group at C^ was observed at a lower magnetic field in the ^H-NMR spectra than the equatorially oriented one. This tendency was the same as that in natural oscillatoxin Ds. In 152 and 153, the C25 and C24 methyl groups occupy axial and equatorial orientations, respectively. In 154 and 155, the orientations of the C25 and C24 methyl groups were reversed as compared to 152 and 153. Except for the natural form, the ether oxygen at C7 was attached axially to the cyclohexanone ring. 3-3-4. Mechanism Addition
of spiroetherification
under acidic and basic
by Intramolecular
Michael-type
conditions
It is known that the 2-cyclohexenone system exists, principally, as two rapidly exchanging envelope (also called sofa) conformations (93, 94). Conjugate addition of a nucleophile can occur to either face of the 2-cyclohexenone. Parallel or anti-parallel (with respect to the axial substituent at C4) attack is possible in principle, however, a nucleophile must approach from an axial direction for satisfying the requirement of the stereoelectronic effect. Anti-parallel attack leads to a favorable chair-like intermediate, whereas parallel attack leads to an unfavorable boat-like intermediate in each case. In an anti-parallel attack, the newly introduced nucleophile forms a rraw^-diaxial arrangement found in a chair conformation. Conversely, parallel attack leads to a 5'>'?i-diaxial arrangement found in a boat conformation. Therefore, anti-parallel attack is favored as this leads to a lower energy intermediate. This description can be applied to our intramolecular conjugate addition in order to explain the reaction mechanism. The precursor 151d of spiroetherification exists as the two envelope conformers [Dl] and [D2], rapidly exchanging with each other (shown in Fig. 3.). The equilibrium between them would lie towards [Dl], because the 1, 3-diaxial interaction due to the C24 and C26 methyl groups contributes negatively in [D2], whereas in [Dl], such an interaction does not exist. In
306 the conformer [Dl], an anti-parallel attack (with respect to the C25 axial methyl group at €5) of the hydroxyl group to the C7 position of the enone leads to a chair-like enol intermediate [E] under acidic conditions. Then protonation of [E] from the axial (top face) direction provides the 75-isomer 153. On the other hand, when conjugate addition proceeds from the conformer [D2], a chair-like enol intermediate [Fl] is formed by anti-parallel attack (with respect to the C24 axial methyl group at C6). Furthermore, [Fl] would flip its conformation to the another chair-like one [F2], because of the 1, 3-diaxial interaction in the former. Protonation of [F2] from the axial (top face) direction provides the natural form, 152. Protonation of [Fl] from the axial (bottom face) direction should provide the 27?-isomer [G], however, in practice, it could not be isolated. Instead of [G], the further isomerized product 155 {2R, 45-isomer), which is more thermodynamically stable, because it lacks 1, 3-diaxial interaction, is produced. The 45-isomer 154 is also produced to a small extent by isomerization of 155. The 1, 3-diaxial interaction between the C24 methyl group and the methoxycarbonyl group would make a relatively less contribution to the thermodynamic stability than the interaction between the two methyl groups.
MeO
152
MeOOC
154
Fig. 3 Equilibrium of Spiroethers under Acidic Conditions
307
Under basic conditions, only two stereoisomers 153 and 155 were produced in ca. 2:1 ratio, respectively. It is considered that the result reflects a kinetic controlled cyclization. An enolate anion corresponding to [E] is produced via a half chair-like transition state [TSl] from [Dl], and then rapid protonation of the enolate from the top face provides 153 as a major product. Through a half chair-like transition state [TSl] from [D2], another enolate anion corresponding to [Fl] is produced. This enolate is rapidly protonated from the bottom face and then epimerization of the C4 stereochemistry leads to 155. In each transition state, it is also considered that the transition state [TSl] (having no 1, 3-diaxial interaction) is more stable than the transition state [TS2] (having a 1, 3-diaxial interaction of two methyl groups). 3-4. The First Total Syntheses of Oscillatoxm D and 30-Methyioscillatoxin D The methyl ester 152 seems to be an important precursor for the total syntheses of oscillatoxin Ds. However the carboxylic acid corresponding to 152 could not be obtained under various conditions. Under the usual alkaline hydrolysis, isomerization occurred to the other stereoisomers and further led to decomposition of the substrate. Some nucleophilic deprotections in aprotic solvents or using Lewis acids were not effective, and mainly resulted in no reaction. Therefore, this route via 152 was suspended, and an alternative biomimetic route via an intermediate [C], as shown in Scheme 23, was carried out according to the initial retrosynthesis. The carboxylic acid obtained from the aldehyde 150d was esterified with segment CI or segment C2 by the Yamaguchi's method (95) to give 156a in 29% yield or 156b in 15% yield. Significant amount of the pyrone derivative 156c was also obtained as a by product in this and in the other esterification. The cyclization seemed to be an intramolecular 1, 6-conjugate addition of the corresponding carboxylate, and followed the favorable "6-endo-Dig." mode of the Baldwin's rule (96). Efforts to increase the yield of esterification were made using a model system. Partial hydrogenation of 156a and 156b proceeded in moderate yields to give the (Z)-olefins 157a and 157b corresponding to [C], which were precursors for the intramolecular Michael-type addition. The conditions (cone. HCl / MeOH) used for the cyclization of the methyl esters 151b and 151d were first applied to 157a, but a fully cyclized product could not be obtained. A product possessing a methoxyl group was detected in ^H-NMR spectrum. It was considered that methanol used as solvent added to the C7 position of 157a, which was more sterically hindered due to the introduction of the y-lactone moiety than that of the methyl ester 151d. Hence intermolecular addition of methanol predominated over the intramolecular reaction on account of the sterically hindered secondary hydroxyl group. Methanol was therefore replaced to a non-alcoholic solvent, dichloromethane, and camphorsulfonic acid was used instead of hydrochloric acid. As a result of these changes, intramolecular cyclization of 157a was accomplished to provide the spiroethers as three stereoisomers: the natural form 158a, 12%, the 75-isomer 159a, 29%, and the 2R, 45-isomer 160a, 11%. The stereochemistries of these isomers were analyzed by NOE experiments similar to those applied for the determining the stereochemistries of the methyl ester series. A small long-range coupling (72,4=1 Hz) between the H2 and H4 axial protons, not explainable as a W-type coupling.
308 characteristically observed in oscillatoxin Ds, was also detected in the iR-NMR spectrum of 158a. The 45-isomer corresponding to 154 was not produced in this case because the contribution of the 1, 3-diaxial interaction between the C24 methyl group and the y-lactone ester would be greater in contrast to that in the methyl ester. The 7S-isomer 159a could be isomerized to the natural form 158a in 24% yield and the IR, 45-isomer 160a in 12% yield, with the recovery of the starting material in 38% yield, under the same acidic condition. It was also possible to isomerize 160a to provide the equilibrium mixture of 158a, 159a, and 160a (97).
R
OBn 156a: R=H (29%, 2 steps) 156b: R=H (15%, 2 steps)
158a: natural form (12%) 159a:7S-isomer(29%) 160a: 2R, 4S-isomer (11%) 158b: natural form (7%) 159b:7S-isomer(29%) 160b: 2R, 4S-isomer (12%)
OBn
158a 158b
e
^-
159a 159b
-£ _ _ ^
158a + 160a 158b +160b
160a 160b
^_ 1 ^ ^
158a + 159a 158b+159b
126a: Oscillatoxin D ( 3 6 % ) j 126b: 30-Methyloscillatoxin D (55%) j
Scheme 29 (a) NaC102, NaH2P04, 2-methyl-2-butene / f-BuOH - H2O (4 : 1); (b) 2, 4, 6trichlorobenzoyl chloride, EtaN / toluene, then Segment CI (R=H) or Segment C2 (R=Me), DMAP / toluene; (c) H2, Lindlar cat. /EtOAc; (d) CSA / CH2CI2; (e) Raney-Ni (W-2) / EtOH
Final deprotection of the benzyl ether (98) of 158a with Raney-Ni (W-2) (99) in ethanol provided oscillatoxin D in 36% yield. The yield was decreased by adsorption of the deprotected phenol on the nickel surface, but the double bond was not reduced under these conditions. Optimization of the deprotection was considered to be possible. The spectral data of the synthetic
309 oscillatoxin D were identical with those of the natural compound. In a similar manner, 157b was converted into 30-methyloscillatoxin D, whose spectral data were also identical with those of the natural compound. The yield of the final deprotection of 158b was improved to 55% yield (100). In conclusion, the first total syntheses of oscillatoxin D and 30-methyloscilIatoxin D, belonging to aplysiatoxins / oscillatoxins family, were accomplished according to a possible biomimetic pathway. The tricarbonyl intermediate, obtained by coupling between Segment A and the Segment B followed by convesion of functional groups, was subjected to intramolecular aldol condensation-dehydration (first cyclization) utilized the p-polyketide character, to afford the cyclohexenone derivative. Thermodynamically controlled intramolecular Michael-type addition (second cyclization) of the enone ester including Segment C provided the natural spiroether ring system of oscillatoxin Ds. The two stereoisomers could be isomerized to the natural form under the same conditions as their cyclization. The two intramolecular cyclizations were the key steps in our total syntheses. ACKNOWLEDGMENTS We are grateful to Dr. Masato Oikawa who actually carried out most of the tautomycin project, and Mr. Tohru Ueno who accomplished the difficult synthesis of the anhydride segment. Their enthusiasm and skillful technique were definitely essential to the compeletion of the tautomycin project. We also thank Kaken Pharmaceutical Co. for the crude sample of tautomycin, and to Dr. Ubukata at the Institute of Physical and Chemical Research and Prof. Isono at Tokai University for the spectra of the degradation products. The oscillatoxin D project was initiated under the direction of Prof. S. Yamamura at Keio University and we are grateful to him for his useful suggestions. We are also grateful to our coworker, Mr. Takashi Goto, who actually carried out the total syntheses of oscillatoxin Ds. These two projects were supported by a Grant-in-Aid from the Ministry of Education, Science, and Culture of Japan. REFERENCES AND NOTES 1.
2. 3. 4. 5. 6. 7.
(a) Cheng, X.-C; Kihara, T.; Kusakabe, H.; Magae, J.; Kobayashi, Y.; Fang, R.-P.; Ni, Z.F.; Shen, Y.-C; Ko, K.; Yamaguchi, I.; Isono, K. J. Antibiot. 1987,40, 907-909. (b) Cheng, X.-C; Ubukata, M.; Isono, K. / Antibiot. 1990, 43, 809-819. (c) Ubukata, M.; Cheng, X.-C; Isono, K. / Chem. Soc, Chem. Commun. 1990, 244-246. (d) Ubukata, M.; Cheng, X.-C; Isobe, M.; Isono, K. J. Chem. Soc, Perkin Trans. 1 1993, 617-624. (a) Magae, J.; Watanabe, C ; Osada, H.; Cheng, X.-C; Isono, K. / Antibiot. 1988, 41, 932937. (b) Magae, J.; Hino, A.; Isono, K.; Nagai, K. J. Antibiot. 1992, 45, 246-251. (c) Kurisaki, T.; Magae, J.; Isono, K.; Nagai, K.; Yamasaki, M. J. Antibiot. 1992, 45, 252-257. (a) Magae, J.; Osada, H.; Fujiki, H.; Saido, T. C ; Suzuki, K.; Nagai, K.; Yamasaki, M.; Isono, K. Proc. Jpn. Acad. Ser. B 1990, 66, 209-212. (b) Hori, M.; Magae, J.; Han, Y.-G.; Karaki, H.; Hartshome, D. J. FEES Lett. 1991, 285, 145-148. MacKintosh, C ; Klumpp, S. FEES Lett. 1990, 277, 137-140. Bialojan, C ; Takai, A. Biochem. J. 1988, 256, 283-290; Haystead, A. J.; Sim, A. T. R.; Carling, R. C ; Honnor, R. C ; Tsukitani, Y.; Cohen, P.; Hardie, D. G. Nature 1989, 337, 78-81. Suganuma, M.; Fujiki, H.; Suguri, H.; Yoshizawa, S.; Hirota, M.; Nakayasu, M.; Ojika, M.; Wakamatsu, K.; Yamada, K.; Sugimura, T. Pro. Nattl. Acad. Sci. USA 1988, 85, 17681771. Cohen, P. Annu. Rev. Biochem. 1989, 58, 453-508.
310
8.
9. 10. 11. 12. 13. 14. 15.
16. 17.
18.
19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
Fujiki, H.; Suganuma, M.; Nishiwaki, S.; Yoshizawa, S.; Yatsunami, J.; Matsushima, R.; Furuya, H.; Okabe, S.; Matsunaga, S.; Sugimura, T. In Relevance of Animal Studies to the Evaluation of Human Cancer Risk; D. Amato, T. J. Slaga, W. Farland and C. Henry, Ed.; Wiley: New York, 1992; pp 337-350. Cheng, X.-C; Kihara, T.; Yinng, X.; Uramoto, M.; Osada, H.; Kusakabe, H.; B.-N., W.; Kobayashi, Y.; Ko, K.; Yamaguchi, L; Shen, Y.-C; Isono, K. 7. Antibiot. 1989,42, 141144. Cheng, X.-C; Ubukata, M.; Isono, K. J. Antibiot. 1990,43, 890-896. (a) Oikawa, M.; Oikawa, H.; Ichihara, A. Tetrahedron Lett. 1993, 34, 4797-4800. (b) Oikawa, H.; Oikawa, M.; Ueno, T.; Ichihara, A. Tetrahedron Lett. 1994,35,4809-4812. (c) Oikawa, M.; Oikawa, H.; Ueno, T.; Ichihara, A. / Org. Chem. 1995. m press. Other synthetic studies: (a) Ichikawa, Y.; Tsuboi, K.; Naganawa, A.; Isobe, M. SYNLETT 1993, 907-908. (b) Naganawa, A.; Ichikawa, Y.; Isobe, M. Tetrahedron 1994, 50, 89698982. (c) Nakamura, S.; Shibasaki, M. Tetrahedron Lett. 1994, 35,4145-4148. Singh, S. B.; Zink, D. L.; Liesch, J. M.; Goetz, M. A.; Jenkins, R. G.; Omstead, M. N.-. ; Silverman, K. C ; Bills, G. F.; Mosley, R. T.; Gibbs, J. B.; Schonberg, G. A.-. ; Lingham, R. B. Tetrahedron 1993, 49, 5917-5926. Oikawa, H.; Oikawa, M.; Ichihara, A.; Ubukata, M.; Isono, K. Biosci. Biotech. Biochem. 1994,55, 1933-1935. (a) Deslongchamps, P., In Stereoelectronic Effects in Organic Chemistry, Pergamon Press: New York, 1983; Vol. 1, pp 4-53. (b) Deslongchamps, P.; Rowan, D. D.; Pothier, N.; Sauve, T.; Saundares, J. K. Can. J. Chem. 1981, 59, 1105-1121. (d) Pothier, N.; Goldstein, S.; Deslongchamps, P. Helv. Chim. Acta 1992, 75, 604-620. (a) Perron, F.; Albizati, K. F. Chem. Rev. 1989, 89, 1617-1661. (b) Boivin, T. L. B. Tetrahedron 1987, 43, 3309-3362. Transformations of spiroketal templates for remote chiral transfer to linear chain subunits in natural products synthesis, see: (a) Totah, N. I.; Schreiber, S. L. 7. Org. Chem. 1991, 56, 6255-6256. (b) Bemet, B.; Bishop, P. M.; Caron, M.; Kawamata, T.; Roy, B. L.; Ruest, L.; Saube, G.; Saucy, P.; Deslongchamps, P. Can. J. Chem. 1985, 63, 2814-2818. (c) Ireland, R. E.; Daub, J. P.; Mandel, G. S.; Mandel, N. S. J. Org. Chem. 1983, 48, 1312-1325. (a) Pettit, G. R.; Albert, A. H.; Brown, P. J. Am. Chem. Soc. 1972, 94, 8095-8099. (b) Pettit, G. R.; Bower, W. J. J. Org. Chem. 1960,25, 84-86. (c) Deslongchamps, P.; Rowan, D. D.; Pothier, N. Can. J. Chem. 1981, 59, 2787-2791. (d) Zhao, Y.-b.; Albizati, K. F. Tetrahedron Lett. 1993, 34, 575-578. (e) Zhao, Y.-b.; Pratt, N. E.; Heeg, M. J.; Albizati, K. F. J. Org. Chem. 1993, 58, 1300-1301. (a) Oikawa, H.; Oikawa, M.; Ichihara, A.; Uramoto, M.; Kobayashi, K. Tetrahedron Lett. 1993, 34, 5303-5306. (b) Oikawa, M.; Oikawa, H.; Ichihara, A. Tetrahedron 1995, 51, 6237-6254. (a) Ishihara, K.; Mori, A.; Yamamoto, H. Tetrahedron 1990, 46, 4595-4612. (b) Ishihara, K.; Hanaki, N.; Yamamoto, H. J. Am. Chem. Soc. 1991,113, 7074-7075. (a) Denmark, S. E.; Almstead, N. G. J. Am. Chem. Soc. 1991,113, 8089-8110. (b) Mori, I.; Ishihara, K.; Flippin, L. A.; Nozaki, K.; Yamamoto, H.; Bartlett, P. A.; Heathcock, C. H. J. Org. Chem. 1990, 55, 6107-6115. The spiroketals 20 and 21 were prepared by the nucleophilic additions of Grignard reagent or dianion prepared by Cohen's procedure to the corresponding lactone, see ref. 19. The preparation of spiroketals 22 and 23 were described in the following section. Kirby, A. J. In The Anomeric Effect and Related Stereoelectronic Effects at Oxygen; SpringerVerlag: New York, 1983; ref. 15a, pp 209-221. When the SI-LA reduction of 21 at -20°C was quenched before completion, twenty six percent of Ca epimer of 21 was detected on GC-analysis of recovered spiroketals. In the similar reaction at -78°C, however, no isomer of 21 was detected. Hydoxy protecting group concerning non- and chelation of metal ions and Lewis acids, see: Chen, X.; Hortelano, E. R.; Eliel, E. L.; Frye, S. V. J. Am. Chem. Soc. 1990,112, 6130-6131 and references cited therein. Corcoran, R. C. Tetrahedron Lett., 1990, 31, 2101. (a) Bartlett, P. A.; Myerson, J. J. Am. Chem. Soc. 1978,100, 3950-3952. (b) Gonzalez, F B.; Bartlett, P. A. Org. Synth. 1985, 64, 175-181. Collum, D. B.; McDonald, J. H.; Still, W. C. J. Am. Chem. Soc. 1980,102, 2118-2120. The similar approarches for preparation of spiroketal with removable equilibration controller at Ca position were reported, see: WilHams, D. R.; Barner, B. A.; Nishitani, K.; Phillips, J. G.
311
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.
59. 60.
7. Am. Chem. Soc. 1982,104, 4708-4710; Iwata, C ; Hattori, K.; Uchida, S.; Imanishi, T. Tetrahedron Lett. 1984, 25, 2995-2998. (a) Brown, H. C ; Bhat, K. S. J. Am. Chem. Soc. 1986,108, 5919-5923. (b)Brown, H. C ; Bhat, K. S.; Randad, R. S. 7. Org. Chem. 1989, 54, 1570-1576. Hanessian, S.; Delorme, D.; Dufresune, Y. Tetrahedron Lett. 1984,25, 2515-2518. (a) Mitsunobu, O. Synthesis 1981, 1-28. (b) Martin, S. F.; Dodge, J. A. Tetrahedron Lett. 1991, 32, 3017-3020. (c) Hughes, D. L. Org. React. 1992,42, 335-656. Mori, K. Tetrahedron 1983, 39, 3107-3109. (a) Yamamoto, Y.; Yatagai, H.; Naruta, Y.; Maruyama, K. J. Am. Chem. Soc. 1980,102, 7107-7109. (b) Yamamoto, Y.; Yatagai, H.; Ishihara, Y.; Maeda, N.; Maruyama, K. Tetrahedron 1984, 40, 2239-2246. Review, see: (c) Yamamoto, Y. Aldrichimica Acta 1987, 20, 45-49. (d) Yamamoto, Y.; Asano, N. Chem. Rev. 1993, 93, 2207-2293. Simpkins, N. S. InSulphones in Organic Synthesis; Pergamon: Oxford, 1993. Corey, E. J.; Chaykovsky, M. J. Am. Chem. Soc. 1965, 87, 1341-1353. Smith, A. B., HI; Hale, K. J. Tetrahedron lett. 1989, 30, 1037-1040. Molander, G. A.; Hahn, G. J. Org. Chem. 1986, 51, 1135-1138. Kende, A. S.; Mendoza, J. S. Tetrahedron Lett. 1991, 32, 1699-1702. Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155-4156. Tsuji, J.; Nagashima, H.; Nemoto, H. Org. Synth. 1984, 62, 9-13. Review, see: Hosokawa, T.; Murahashi, S.-I. Ace. Chem. Res. 1990,1990,49-54. (a) Heathcock, C. H. Aldrichimica Acta 1990, 23, 99-111. (b) Van Draanen, N. A.; Arseniyadis, S.; Crimmins, M. T.; Heathcock, C. H. J. Org. Chem. 1991, 56, 2499-2506. (a) Roush, W. R.; Ando, K.; Powers, D. B.; Palkowitz, A. D.; Halterman, R. L. J. Am. Chem. Soc. 1990,112, 6339-6348. (b) Roush, W. R.; Palkowitz, A. D.; Ando, K. J. Am. Chem. Soc. 1990,112, 6348-6359. (c) Roush, W. R.; Bannister, T. D.; Wendt, M. D. Tetrahedron Lett. 1993, 34, 8387-8390. Mihelich, E. D.; Daniels, K.; Eickhoff, D. J. J. Am. Chem. Soc. 1981,103, 7690-7692. (a) Corey, E. J.; Schmidt, G. Tetrahedron Lett. 1979, 20, 399-402. (b) Herscovici, J.; Antonakis, K. J. Chem. Soc. Chem. Commun. 1980, 561-562. Scott, W. J.; Stille, J. K. J. Am. Chem. Soc. 1986,108, 3033-3040. Westley, J. W.; Evans, J., R. H.; Williams, T.; Stempel, A. J. Org. Chem. 1973, 38, 34313433. (a) Baumann, M. E.; Bosshard, H.; Breitenstein, W.; Gunter, R. Helv. Chim. Acta 1986, 69, 396-403. (b) Newman, M. S.; Stalick, W. M. J. Org. Chem. 1973, 38, 3386. (c) White, J. D.; Dillon, M. P.; Butlin, R. J. J. Am. Chem. Soc. 1992,1992, 9613-961 A. (a) Huff, R. K.; Moppett, C. E.; Sutherland, J. K. J. Chem. Soc. (C) 1968, 2725-2726. (b) Huff, R. K.; Moppett, C. E.; Sutherland, J. K. J. Chem. Soc, Perkin Trans. 1 1972, 25842590. Ubukata, M.; Cheng, X.-C; Isono, K. In Symposium Papers of 33th Symposium on the Chemistry of Natural Products; Osaka (Japan), 1991; pp 643-650. (a) Yang-Lan, S.; Mueller-Johnson, M.; Oehldrich, J.; Wichman, D.; Cook, J. M. J. Org. Chem. 1976, 41, 4053-4058. (b) Bertz, S. H.; Cook, J. M.; Gawish, A.; Weiss, U. Org. Synth. 1986, 64, 27-37. (c) Fu, X.-Y.; Cook, J. M. Aldrichimica Acta 1992, 25, 43-54. Ueno, T.; Oikawa, M.; Oikawa, H.; Ichihara, A., unpublished results. Ueno, T.; Oikawa, M.; Oikawa, H.; Ichihara, A., unpublished results. (a) Carruthers, W. In Cycloaddition Reactions in Organic Synthesis; Pergamon: Oxford, 1990. (b) Curran, D. P. The Cycloadditon Approach to b-Hydroxy Carbonyls; Wiley: New York, 1990. (a) Christ, M.; Huisgen, R. Chem. Ber. 1973,106, 3345-3367. (b) Martin, S. F.; Dupre, B. Tetrahedron Lett. 1983,2^, 1337-1340. Curran, D. P. J. Am. Chem. Soc. 1983,105, 5826-5833. Tanis, S. P.; Herrinton, P. M. J. Org. Chem. 1983, 48, 4572-4580. Related oxidation of furan, see: Sargent, M. V.; Cresp, T. M. In Comprehensive Organic Chemistry, E. Haslam, Ed.; Pergamon: Oxford, 1979; Vol. 4; pp 693-744; Lipshutz, B. H. Chem. Rev. 1986, 86, 795-819. Hughes, M. J.; Thomas, E. J.; Tumbull, M. D.; Jones, R. H.; Warner, R. E. /. Chem. Soc. Chem. Commun. 1985, 755-758. Hughes, M. J.; Thomas, E. J. J. Chem. Soc. Perkin Trans. 1 1993, 1493-1505. (a) Itsuno, S.; Ito, K.; Hirao, A.; Nakahama, S. J. Org. Chem. 1984, 49, 555-557. (b) Corey, E. J.; Bakushi, R. K.; Shibata, S. J. Am. Chem. Soc. 1987,109, 5551-5553. (c)
312
61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.
72. 73. 74. 75. 76. 77.
78.
79. 80. 81. 82. 83. 84.
Corey, E. J.; Bakushi, R. K.; Shibata, S.; Chen, C.-P.; Singh, V. K. J. Am. Chem. Soc. 1987, 709, 7925-7926. (d) Mathre, D. J.; Thompson, A. S.; Douglas, A. W.; Hoogsteen, K.; Carroll, J. D.; Corley, E. G.; Grabowski, E. J. J. / Org. Chem, 1993, 58, 2880-2888. Oikawa, H.; Ichihara, A., unpublished results. Shibasaki group also reported highly enantioselective reduction of the compound related with 95: Nakamura, S.; Shimizu, S.; Nakada, M.; Shibasaki, M. In Symposium Papers of 36th Symposium on the Chemistry of Natural Products', Hiroshima (Japan), 1994; pp 611-618. Parikh, J. R.; Doering, W. v. E. /. Am. Chem, Soc. 1967, 89, 5505-5507. Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.; Hartumg, J.; Joeng, K.-S.; Kwong, H.-L.; Morikawa, K.; Wang, Z.-M.; Xu, D.; Zhang, X.-L. J. Org. Chem. 1992, 57, 2768-2771. (a) Oikawa, Y.; Nishi, T.; Yonemitsu, O. Tetrahedron Lett. 1983, 24, 4037-4040. (b) Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yonemitsu, O. Tetrahedron 1986, 42, 3021-3028. Toshima, K.; Mukaiyama, S.; Kinoshita, M.; Tatsuta, K. Tetrahedron Lett. 1989, 30, 64136416. Kraus, G. A.; Taschner, M. J. J. Org. Chem. 1980,45, 1175-1176. Gorthey, L. A.; Vairamani, M.; Djerassi, C. 7. Org. Chem. 1984,49, 1511-1517. Sharpless, K. B.; Caron, M. J. Org. Chem. 1985, 50, 1557-1560. Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull Chem. Soc. Jpn. 1979, 52, 1989-1993. (a) Evans, D. A.; Ng, H. P.; Rieger, D. L. J. Am. Chem. Soc. 1993, 775, 11446-11459; (b) Evans, D. A.; Calter, M. A. Tetrahedron Lett. 1993, 34, 6871-6874; (c) Roush, W. R.; Bannister, T. D.; Wendt, M. D. Tetrahedron Lett. 1993, 34, 8387-8390; (d) Martin, S. F.; Lee, W.-C. Tetrahedron Lett. 1993, 34, 2711-2714; (e) Paterson, I.; Cumminng, J. G. Tetrahedron Lett. 1992, 33, 2847-2850. Evans, D. A.; Rieger, D. L.; Bilodeau, M. T.; Urpf, F. J. Am. Chem. Soc. 1991, 775, 10471049. Mukaiyama, T. Org. React. 1982,28, 203-331. (a) Reetz, M. T.; Kesseler, K. J. Org. Chem. 1985, 50, 5436-5438. (b) Reetz, M. T.; Jung, A. J. Am. Chem. Soc. 1983,105, 4833-4835. (c) Gennari, C ; Beretta, M. G.; Bemardi, A.; Moro, G.; Scolastico, C ; Todeschini, R. Tetrahedron 1986, 42, 893-909. Entzeroth, M.; Blackman, A. J.; Mynderse, J. S.; Moore, R. E. J. Org. Chem., 1985, 50, 1255-1259. Moore, R. E.; Blackman, A. J.; Cheuk, C ; Mynderse, J. S.; Matsumoto, G. K.; Clardy, J.; Woodard, R. W.; Craig, J. C. J. Org. Chem., 1984,49, 2484-2489. (a) Fujiki, H.; Suganuma, M.; Nakayasu, M.; Hoshino, H.; Moore, R. E.; Sugimura, T. Gann, 1982, 73, 495-496; (b) Suganuma, M.; Fujiki, H.; Tahira, T.; Cheuk, C ; Moore, R. E.; Sugimura, T. Carcinogenesis, 1984, 5, 315-318; (c) Jeffrey, A. M.; Liskamp, R. M. J. Proc. Nat. Acd. Sci. USA, 1986, 83, 241-245. (a) Park, P.; Broka, C. A.; Johnson, B. F.; Kishi, Y. J. Am. Chem. Soc, 1987,109, 6205-6207; (b) Ireland, R. E.; Tharisrivongs, S.; Dussalt, P. H. J. Am. Chem. Soc, 1988, 110, 5768-5779; (c) Walkup, R. D.; Cunningham, R. T. Tetrahedron Lett., 1987, 28, 4019-4022; (d) Walkup, R. D.; Kane, R. R.; Boatman, Jr., P. D.; Cunningham, R. T. Tetrahedron Lett., 1990, 31, 7587-7590; (e) Walkup, R. D.; Boatman, Jr., P. D.; Kane, R. R.; Cunningham, R. T. Tetrahedron Lett., 1991, 32, 3937-3940; (f) Okamura, H.; Kuroda, S.; Ikegami, S.; Tomita, K.; Sugimoto, Y.; Sakaguchi, S.; Ito, Y.; Katsuki, T.; Yamaguchi, M. Tetrahedron, 1993, 49, 10531-10554. (a) Toshima, H.; Yoshida, S.; Suzuki, T.; Nishiyama, S.; Yamamura, S. Tetrahedron Lett., 1989, 30, 6721-6724; (b) Toshima, H.; Suzuki, T.; Nishiyama, S.; Yamamura, S. Tetrahedron Lett., 1989, 30, 6725-6728. Nakamura, H.; Park, P.; Kishi, Y. 58th Annual Meeting of Chem. Soc Jpn., Kyoto, Japan, April 1989, Abstract II, pp 1189. (a) Mynderse, J. S.; Moore, R. E.; Kashiwagi, M.; Norton, T. R. Science, 1977, 796, 538540; (b) Moore, R. E. Pure & Appl Chem., 1982, 54, 1919-1934. Moore, R. E., Personal communication, The antileukemic activity was based on an assay using a limited amount of oscillatoxin D. Kinoshita, M.; Mariyama, S. Bull Chem. Soc Jpn., 1975,48, 2081-2083. (a) Preparation of the Wittig reagent: Evans, D. A.; Kalder, S. W.; Jones, T. K.; Clardy, J.; Stout, T. J. J. Am. Chem. Soc, 1990, 772, 7001-7031; (b) N-Methoxy-N-methylamide
313
(Weinreb's amide) as acylating reagent: Nahm, S; Weinreb, S. M. Tetrahedron Lett., mi, 22, 3815-3818. 85. (a) Brown, H. C ; Singaram, B. 7. Org. Chem., 1984, 49, 945-947; (b) Chandrasekharan, J.; Ramachandran, P. V.; Brown, H. C. J. Org. Chem., 1985, 50, 5446-5448; (c) (-)Diisopinocampheylchloroborane could be readily derived from (-i-)-a-pinene. Now, this reagent is commercially available as (-)-DIP-Chloride™ from Aldrich Chemical Company, Inc. 86. The desired diastereomer (5 configuration): 5 (ppm) -0.021 and 0.017. The undesired diastereomer {R configuration): 5 (ppm) -0.006 and 0.025. 87. (a) Yadav, J . S.; Chander, M. C ; Rao, C. S. Tetrahedron Lett., 1989, 30, 5455-5458; (b) Yadav, J . S.; Deshpande, P. K.; Sharma, G. V. M. Tetrahedron, 1990, 46, 7033-7046. 88. Toshima, H.; Goto, T.; Ichihara, A. Tetrahedron Lett., 1994, 35, 4361-4364. 89. Boekman, Jr., R. K.; Charette, A. B.; Asberom, T.; Johnston, B. H. /. Am. Chem. Soc, 1991,113, 5337-5353. 90. Saito, S.; Hsegawa, T.; Inaba, M.; Nishida, R.; Fujii, T.; Nomizu, S.; Mori wake, T. Chem. Lett, 19S4, 1389-1392. 91. Ortuno, R. M.; Bigorra, J.; Font, J. Tetrahedron, 1988,44, 5139-5144. 92. Elhson, R. A.; Lukenbach, E. R.; Chiu, C.-W. Tetrahedron Lett., 1975, 499-502. 93. Perlmutter, P. Conjugate Addition Reactions in Organic Chemistry, Pergamon Press, Oxford, 1992. 94. Chamberlain, P.; Whitham, G. H. J. Chem. Soc, Perkin Trans. 11,1972 130-135. 95. Inanaga, J.; Hirata, K.; Saeki, H.; Katuki, T.; Yamaguchi, M. Bull. Chem. Soc. Jpn., 1979, 52, 1989-1993. 96. Baldwin, J. E. J. Chem. Soc, Chem. Comm., 1976, 734-736. 97. The namral form and the 75-isomer were obtained slightly with the recovery of the starting 2R, 45-isomer mainly and decompositions decreased their yields. 98. Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yanemitus, O. Tetrahedron, 1986, 42, 3021-3028. 99. Mozingo, R. Org. Synth. Coll. Vol. 3,1955, 181-183. 100. Toshima, H.; Goto, T.; Ichihara, A. Tetrahedron Lett., 1995, 36, 3373-3374.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
315
Aza-Annulation of Enamine Related Substrates with UybUnsaturated Carboxylate Derivatives as a Route to the Selective Synthesis of d-Lactams and Pyridones John R. Stille and Nancy S. Barta
1 . INTRODUCTION Six-membered nitrogen heterocycles are found in a wide variety of naturally occurring alkaloids, and are often incorporated into the design of biologically active pharmaceutical products. Of the many ways to construct these heterocycles, the aza-annulation of enamine substrates with a,p-unsaturated carboxylic acid derivatives has been a versatile and efficient method for the formation of dihydropyridone and pyridone products. The synthetic utility of this approach has led to incorporation of aza-annulation methodology as a key step in the synthesis of a number of interesting heterocychc molecules. A/-AcyJatJon
O
Condensation Conjugate Addition Figure 1.
There are many variants of the aza-annulation approach to heterocycle formation. However, the general process involves the combination of three fundamental components - an amine, an a,punsaturated carboxylic acid derivative, and a carbonyl compound (Figure 1). Although each variant of the aza-annulation reaction displays subtle component dependent differences that influence the interactions in the transition state, a detailed mechanistic analysis will not be presented. However, from a synthetic perspective, a general introduction to the characteristic types of reactions involved will provide a greater understanding of the effects that substrate and reagent properties have on the outcome of the reaction. Combination of the three annulation components results in three different bond-forming processes: (1) condensation of the nitrogen with the carbonyl derivative to generate an enamine, (2) conjugate addition of the carbonyl derivative to the a,p-unsaturated carboxylic acid derivative, which serves as a Michael acceptor, and (3) acylation of the nitrogen to form the amide bond. The two approaches used most often for aza-annulation are shown in Scheme 1. In each case, the initial reaction in the aza-annulation process involves incorporation the nitrogen atom.
316
Whether through initial A^-acylation, typically the use of readily available acrylamide or acrylonitrile, or through enamine formation, the stage is set for the important carbon-carbon bond formation step. Conjugate addition results in formation of the carbon-carbon bond to the acrylate derivative, and the cyclization is completed through either condensation or A^-acylation.
o O O "^ 1
'^'^N^ ^
Conjugate Addition
or
/^-Acylation I
p6
|| I
Condensation Conjugate Addition
^^^N'" 5
^
u R5
Scheme 1.
An alternative approach to aza-annulation again utilizes condensation to form an enamine, which is subsequently //-acylated (Scheme 2). The resultant acrylenamide is a stable, isolable intermediate which can be efficiently converted to the corresponding dihydropyridone through photochemical processes. However, application of this photochemical methodology has been the subject of previous reviews, and will not be discussed here.^ Alternative routes for conversion of the acrylenamide to the dihydropyridone, through the use of Lewis acids, protic acids, and heat, generally has been unsuccessful.
Rk^.H
^^^N'" Condensation
R S - ^
R^
A/-Acylation
R 6 ' ^
. ,
"Conjugate iiinntp Addition"
R^-V >.
Scheme 2.
The electronic and steric effects observed for the various carbonyl substrates and acrylate reagents used in this reaction clearly illustrate the pivotal role of the conjugate addition step in the azaannulation process (Scheme 3). The nature of the carbonyl derivative is highly dependent upon the
317 type of substituent present. Substituents R^ and R^ directly affect the imine-enamine tautomer equilibrium, the regiochemical formation of enamine tautomers, and the reactivity of the enamine. Due to the importance of the carbonyl derived substrate in determining the outcome of the reaction, this review is organized according to carbonyl substrate type.
Iv^ E = CN, CO2R Y = SR, OR, NR2
R5
Scheme 3.
The presence of electron withdrawing groups at the a position (R^) of the acrylate derivative increases the reactivity of the reagent toward conjugate addition, while substituents in the P position (R"^) tend to provide steric constraints that hinder carbon-carbon bond formation (Scheme 3). Of the various acrylate derivatives employed in these reactions, the most frequently used have electron withdrawing functionality such as a carboxylic acid, amide, ester, or nitrile group or a combination of these. Direct pyridone formation can be achieved primarily through the use of either a,p acetylenic esters or acrylate derivatives with P substituents (Y = SR, OR, NR2) that eliminate under the reaction conditions.
H2 Pd or Pt
DDQ or Mn02 0 "'"N-^
R.V B'
r"' "' ROH, H"'
Li, NH3 or NaBH4 or RgSiH, TFA
RO '^ R5
Scheme 4.
318 The dihydropyridone products formed through this aza-annulation process can also be modified to provide other important ring systems (Scheme 4). Pyridone species have been obtained by oxidation of the corresponding dihydropyridone with either DDQ or Mn02. Reduction can be performed to generate the corresponding tetrahydropyridones with either a cis or trans relationship of substituents R^ and R^, depending on the type of reducing agent employed. Acid catalyzed addition of nucleophiles to C-6 has also been reported. The scope and utility of the aza-annulation methodology is described herein, and ;s presented according to the different types of carbonyl derivatives used. Coverage of this subject encompasses the use of enamine substrates with the same oxidation state as aldehyde and ketone functionality, which leads to the formation of 6-lactam products. Substrates of the carboxylate oxidation state, such as ketene related enamine substrates in which R^ = SR, OR, or NR2, result in the formation of products with the oxidation state of glutarimide, and will not be covered in this review. For each class of carbonyl substrate, the effectiveness and selectivity of various a,p-unsaturated carboxylic acid annulation reagents is presented. Although special emphasis is placed on the use of the azaannulation in the synthesis of biologically active natural products, a perspective of the methodology developed for each substrate type is included. Recent use of this methodology for the highly diastereoselective generation of quaternary centers is also covered. 2 . ALKYL IMINE/ENAMINE SUBSTRATES 2.1
Acrylonitrile Reagents
1
NC,
1a: n = 0 lb: n = 1
H20/ACOH (Trace)
Cyclohexylamine (0.05 equiv.), AcOH (0.02 equiv.)
200 °C 2-4 h
^^
<:xn I
8: X = H,H (±)-Cephalotaxine
steps
^ o I J
' [ A
^ o X X
7 -> 8, 50% 7 -» 9, 52% (+16% yield of the isomer formed from 5)
9:X = 0
(±)-8-Oxocephalotaxine
[Ref. 2 and 3]
f"'^"^'
Scheme 5.
\
\
319 Early studies in the area of aza-annulation included an examination of the reaction of aliphatic ketones with acrylonitrile (2, Scheme 5).2 This two-step, one-pot procedure involved initial conjugate addition of the carbonyl substrate to acrylonitrile, followed by hydrolysis of the nitrile and intramolecular condensation to form the dihydropyridone product. Typical reaction conditions were relatively harsh, and required trace amounts of H2O, AcOH, and a primary amine at temperatures that reached 200 °C. Although this procedure provided 4b in 79% yield, initial aza-annulation studies with either la or acetone were less successful, and produced the corresponding dihydropyridones in only 18% and 15% yields, respectively.^ A later report described the cyclization of 3a to give 4a and 5a in 61% yield as a 79:21 ratio of the regioisomers after equilibration of the product mixture with MgS04 in CH2Cl2.^ Alkylation of 4a and rearrangement of 6 to the spirolactam 7 led to the syntheses of (±)-cephalotaxine (8) and (±)-8-oxocephalotaxine (9), and the ester derivatives of 8 and 9 have demonstrated anticancer activity.
O
2
ll
3 Steps (Eq.1)
4-Methylcyclohexylamine (cat.), AcOH (cat.)
60%
10 53%
11a:11b (75:25)
[Ref. 4]
12b
12a:12b (75:25)
Regioselective aza-annulation with acrylonitrile met with somewhat limited success. Azaannulation of 10 with acrylonitrile resulted in the formation of a 75:25 mixture of 11 (eq. 1)."^ The mixture of 11 was taken on to 12, which was used in the synthesis of optically active 2,2'bipyridines and 1,10-phenanthrolines.
A more selective route to 11a was reported, and this
approach started from optically pure (/?)-(+)-pulegone (13, eq. 2).5 In this case, regioselective Michael addition to give 14 was accomplished due to the presence of the isopropylidene moiety of the (/?)-(+)-pulegone. Subsequent rhodium catalyzed deisopropylidenation, nitrile hydration, and cyclization gave l i b , which was taken on to (-)-pumiliotoxin C (15).
1)RuH2(PPh3)4 2) H2O, H* 56% 13 [Ref. 5]
14
^
4 Steps 15 (Eq.2) (-)-Pumiliotoxin C
320 An example of a regioselective aza-annulation reaction with acrylonitrile derivative 17 is illustrated in eq. 3. Regioselective reaction with levulinic acid (16) led to the isolation of 18 in 34% yield.^ Interesting features of this reaction include (1) selective reaction at the terminal methyl group of 16, (2) the compatibility of a carboxylic acid functional group with the reaction conditions, and (3) the elimination of one of the -SMe groups to generate pyridone 18. Treatment of symmetrical ketones cyclohexanone (lb) or acetone with 17 under the same reaction conditions resulted in 48% and 50% yields, respectively, of the analogous pyridone products.^ When a related reagent, bis(methylthio)methylenecyanoacetamide, was used for aza-annulation of cyclohexanone and acetone, yields of 24% and 50%, respectively, were obtained.^ KOH (Powder) DMSO
NC^^CN MeS
O
16
25°C,8h
SMe
'^?
II
(Eq.3)
34%
17
[Ref. 6]
2.2
Acrylamide Reagents Aza-annulation of enamine substrates with acrylamide (20) occurred under milder reaction
conditions. Treatment of 19 with 20 in dioxane at reflux (100 °C) resulted in the efficient formation of 4b with reported yields of 70%'^ and 83%^ (eq. 4). Subsequent study of this reaction led to the quantitative formation of 4b and 5b reported as a 50:50 mixture of the regioisomeric alkenes.^ This methodology was also extended to aza-annulation of the cyclopentanone enamine with 20 (72%).^ The analogous reactions have been performed by the combination of 20 with the benzyl imine (77%)8 or the propyl imine (18%)^^ of cyclohexanone (lb). Limitations arose when the acrylamide was A^-substituted or had a substituent at the p-position (crotylanilide).^ When reagents with these features were used, the reaction did not proceed.
6
1) 1,4-Dioxane, Reflux, 3 h
19
[Ref. 7-10]
(Eq.4)
2) Add H2O, Reflux, 1 h 70-100%
Both symmetrical and unsymmetrical ketones have been utilized in synthetic applications of the aza-annulation reaction of enamine derivatives with acrylamides. The most efficient use of this methodology involved generation of the enamine from the symmetrical ketone 21 (eq. 5).^^ Azaannulation with this enamine led to the formation of 22 as an 85:15 mixture of the regioisomeric tetra- and trisubstituted alkenes (b and a), respectively, which were then oxidized to the corresponding pyridone 23. Formation of intermediate 23 was a key to the synthesis of the
321 nootropic agent (±)-huperzine A (24), which has potential applications for the treatment of Alzheimer's disease. 1) Pyrrolidine, 1) BnCI, KH
2)20 Dioxane,
PhSeCI
Reflux 3) H2O, Reflux
(Eq.5)
2) LDA,
^»
9 Steps
^^
3) Nal04
70% 22a:22b (15:85)
72%
23
10%
24 (±)-Huperzine A
A second application of this methodology involved the use of an unsymmetrical ketone, which led to regioselectivity complications during aza-annulation. The lack of regioselectivity resulted from the generation of 25 as a mixture of regioisomeric olefins (eq. 6). 12 Subsequent azaannulation with acrylamide gave a mixture of regioisomeric products that included 26, which could be isolated from the mixture in 20-25% yield. Heterocycle 26 was used in an approach to (±)lycopodine.
Q
1) 20, Dioxane, Reflux, 12 h
10 Steps (Eq.6)
2) Add H2O, Reflux, 2 h
Ar 25
20-25%
27 (±)-Lycopodine
[Ref. 12]
O
EtOH 50-60 °C, 8 h Ph
29
(Eq. 7)
21%
Equation 7 illustrates the regioselective aza-annulation of an unsymmetrical ketone with a (3substituted acrylate derivative. ^^ Although the p-phenyl substituent served to retard the azaannulation process, reaction of 28 with 29 proceeded due to double activation of the acrylate reagent by both the amide and cyano substituents. Regioselective formation of 30 was the result of steric
322 interactions between the phenyl and geminal dimethyl substituents during aza-annulation, and the influence of such steric interactions was also reflected by the low yield. Another example of a regioselective aza-annulation with a more efficient variant of a |3-aryl substituted acrylate derivative is shown in Scheme 6. Reaction of 2-butanone (b) with 31 in the presence of NH4OAC at 170 °C resulted in regioselective carbon-carbon bond formation to give 335.^"^ In general, this aza-annulation reaction was believed to proceed by initial conversion of 31 to the corresponding amide followed by conjugate addition of the thermodynamic enol tautomer of the ketone to give 32. When the reaction with cyclopentanone (d) was stopped at partial conversion, 32d was isolated, and further heating produced a mixture of 33d and 34d in which R^=H. This methodology was applied to the synthesis of derivatives of 35, which were investigated as pharmaceutical agents useful for treating psychic disturbances and depressive states. ^^ Similar methodology was extended to the thio and imino amide derivatives of 31.^^ However, this approach resulted in less efficient aza-annulation of 2-butanone with requisite 2-cyanoacrylate substrates (2117
a: b: c: d: e: f:
2.5 equiv. NH4OAC, MeNH2 or ArNH2
R^ = Me, R^ = H R^ = Me. R^ = Me R^ = Me, R^ = Ph R \ R 2 =-(CH2)3R \ R 2 =-(CH2)4R \ R 2 =-(CH2)5-
O
O
170 °C 4-6h
59^%
O
1)NH40Ac 25 °C, 3 d 2) Heat, 1 h O H
34 +
[Ret. 16]
Scheme 6.
2.3 a,p-Unsaturated Ester Reagents Heterocycle formation through aza-annulation with acrylate esters was accomplished with both imine and enamine derivatives of aldehydes. Treatment of 36 with methyl acrylate (37) resulted in both mono and bisalkylation at the nucleophilic site, and selective bisalkylation to form 38 could be enhanced through the use of two equivalents of 37 (eq. 8). 1^ Subsequent cyclization of 38 generated the corresponding aza-annulation product 39. Selective monoalkylation was more
323
effective from the analogous enamine 40, which provided 41 in good yield (eq. 9).^^ Cyclization to the 5-lactam followed by reduction gave the useful indole alkaloid synthon 42.
r 37
iPr. |l
2equiv. I
Hydroquinone 110-120 °C 16h 66P/o
36 [Ret. 18]
^J
s1
'^^-N
1)37, MeCN 2) HgO-^
^^^2^^
y ^
Pyridine Heat
^ s ^ COaMe
COaMe
r^r^ Kii COsMe
65%
1) BnNH2 Toluene Reflux, 6 h 2)95°C,12h
(Eq.8)
'^^^N"^
(Eq.9) Bn^^.A^
LiAIH4
^'^^M"'^
89%
96%
T 42
40
^
[Ref. 19]
Ketone-type enamine substrates have also been utilized, as demonstrated through the synthesis of (±)-Na-benzyl-20-desethylaspidospermidine (45, ^-H) from 43 (eq. 10).^^ Treatment with methyl acrylate (37) resulted in conjugate addition to the acrylate and deallylation, and subsequent reduction with NaBH4 gave 44 as a 70:30 mixture of p:a diastereomers. Reduction with LiAlH4 gave 45 (75:25, p:a). However, similar stepwise aza-annulation with the enamine derived from pyrrolidine and a protected 4-piperidinone resulted in only 14% product formation.^^ 1)37, MeOH Reflux, 12 h 2) NaBH4 EtOH, 8 h 53%
(Eq. 10)
^^ p:a
43 [Ref. 20]
(75:25)
45 (p-H) (±)-Na-Benzyl-20-desethylaspidospermidine
Aza-annulation of ketone derived substrates with a,P-unsaturated ester reagents was very effective when reagents that led to direct formation of pyridone products were used. Deprotonation of cyclic imine 46 followed by treatment with 47 resulted in formation of the pyridone derivative 48, which gave (±)-isosophoramine (49) upon deethoxycarboxylation (eq. ll).^^ Similarly, the exocyclic imine 50 was used to generate 52 by treatment with 51, accompanied by elimination of MeOH and removal of the t-Bu group at elevated temperatures (eq. 12).^^
324 1)LDA.THF 2) O O
O
O
EtO^V^OEt EtO
47
OEt
20% H2SO4 Reflux, 14 h
^^
^
26%
53%
48
(±)-lsosophoramine
1)100°C,3h 2) 200 °C, 1 h tBu^
6
OMe
Ph20
OMe
(Eq.12)
71% 50
51
[Ret. 23]
A somewhat unique substrate, 54, was used in an approach to a-amino acid lactone targets, but instead underwent aza-annulation with 55 to give 56 after elimination of an amino group and deprotection (eq. 13).24 Acrylate equivalents, in the form of p-heteroatom substituted esters, such as P-amino esters^^ and P-lactones^^ have also been used to generate 5-lactams, but lactam or pyridone formation was not selective. Pyridone formation was also shown to occur through treatment of 21 with 57 under 200 psi of NH3 at 100 °C to give 58 (eq. 14), a key intermediate in the synthesis of the nootropic agent (±)-huperzine A (24).27 The use of 57 also led to good yields for pyridone formation from enamine derivatives of cyclopentanone and cycloheptanone.^^
AcOH
Ph' Me 54
21% O
^ N ^
MeO,
(Eq. 13)
55
[Ret. 24]
MeO^^O
57 [Ret. 27]
NH3 (200 psi) MeOH, 100 °C
70%
(Eq. 14)
325 2.4 Activated a,p-Unsaturated Carboxylic Acid Reagents The effectiveness of the aza-annulation of cyclohexanone imines with various activated acryhc acid derivatives was the subject of an in-depth study.28 The efficiency of aza-annulation with respect to the properties of the imine substrate, carboxylate derivative, and acrylate substitution was investigated. When different methods for activation of acryhc acid were used in the aza-annulation of 59, acryloyl chloride produced very low mass recovery with significant amounts of the A^-acylated derivative 60 (eq. 15). In the presence of NEt3, 60 was essentially the only product formed. Attempts to convert 60 to either 61 or 62, through the use of Lewis acids, protic acids, and temperatures as high as 200 °C, were unsuccessful. Only photochemical methods have been used to effectively facilitate this conversion (60 -> 61, Rl=CH2Ph, R2=R3=H; hv, 61%).29 0
0
R^
^ " ^
^R3
!
THF
Nx^
59
GO
61
(Eq. 15) 62
[Ret. 28]
R1
R2
R3
X
Yield
Temp (°C)
60(%)
6 1 ( % ]\
62(%)
iBu iBu
H H
H H
Cl-
66 66
72
15
H
H
EtOC02-
25
0 4
81
iBu
32 70 77
13
H2C=CHC02-
33
19 63
N3Imidazole-
69
0
5
72
23
81
66
0
74
26
H2C=CHC02Cl-
69
66
2
>95
34
66
82
EtOC02-
79
66
45
43
ClCl-
32
66
—
EtOC02-
66
66 66
13 69 4
72 15 31 (both) 85 11
iBu
H
H
iBu
H
H
iBu iBu iBu
Me H H
Me Me
iPr iPr iPr
H H H
Me Me
H
H
<3 18 (both) 12
Other activated acrylate derivatives have been used more successfully to minimize formation of 60 with concommitant increase in reaction yield.^^ The corresponding anhydride (X = H2C=CHC02-), mixed anhydride (X = EtOC02-), azide (formed by treatment of acrylic acid with (PhO)2P(0)N3 (DPPA)), and the acyl imidazole all enhanced the aza-annulation process. In general, substitution at the a-position had minimal effect on the outcome of the reaction, but was accompanied by variations in the ratio of 61 to 62. However, when the crotyl derivative was used (R^ = Me) the aza-annulation process was significantly slower, and when compared to the acrylic acid reagents (R^ = H), the amount of 60 was increased relative to the amounts of 61 and 62
326 formed. An increase in the steric contribution of the imine substituent (R^ = iPr) hindered the formation of 60 and resulted in increased formation of 61 and 62 upon treatment with crotonate derivatives (X = C1-, EtOC02-). In related studies with the imines of cyclopentanone, yields were 20% lower on the average, and similar reaction with the imine of butanal gave poor yields.^^ Prior to the above investigation, Hickmott had reported a study on regioselective azaannulation with an unsymmetrical ketone.^^ Treatment of 63 with 64 resulted in selective formation of 66 in preference to 67 (eq. 16). Formation of 66 was the result of conjugate addition of the thermodynamically favored enamine isomer, which led to formation of a quaternary carbon in this process. Treatment of 63 with crotonyl or cinnamyl chloride resulted in iV-acylation of the enamine to give the corresponding analogs of 65 without significant heterocycle formation.
.A U
R^
(Added at Reflux)
63 [Ret. 30]
Benzene
^
R"" = Cyclohexyl
I
||
rS-
+ 38%
65
67
An interesting variant of the annulation of enamine substrates with acrylic acid derivatives was the thermal reaction of enamines with vinyl isocyanates (eq. 17).^^ Treatment of 68 with DPPA produced the intermediate vinyl isocyanate, which underwent cycloaddition with the enamine followed by elimination of the pyrrolidine to give pyridone 70. Pyridone 70 was efficiently converted to the corresponding highly substituted pyridine. (Eq. 17)
OH OMe OMe OMe
a 34% OMe
Acrylic acid species 72, without activation, underwent aza-annulation with 71 to generate 73 almost entirely as the a isomer (eq. 18).32 Subsequent modification of 73 led to the total synthesis of (±)-vallesamidine 74. Treatment of 71 with cinnamic acid, which lacked the NO2 functionality, gave the corresponding aza-annulation product at higher temperature (145 °C) in 65% yield as a 6:1 ratio of diastereomeric products.32 Alternatively, treatment of 71 with cinnamyl chloride/NEts,
327 resulted in formation of the acrylate enamide (88%) analogous to 60. This acrylate enamide could only be converted to the dihydropyridone 73 through photochemical cyclization methodology (42%).
NO2
(Eq. 18)
72
0\_) 71
3 Steps
Dioxane, Reflux
J
80%
42%
74 (±)-Vallesamidine
[Ret. 32]
3 . AROMATIC CONJUGATED IMINE/ENAMINE SUBSTRATES 3.1 a-AryI Enamine Substrates Regioselective aza-annulation reactions have been achieved through the use of aiyl substituted keto enamine substrates. Aryl substitution at the a-position of the enamine provides for regioselective imine-enamine tautomerization at the non aryl P carbon. Early studies in this area compared the utility of pyrrolidine enamine versus benzyl imine derivatives of a-tetralone.^ In the absence of solvent, aza-annulation of 75 with acrylamide led to the formation of 76, but use of imine derivative 77 resulted in significantly higher yield of 76 (eq. 19).^ Similar reactions were observed between an aryl ethyl ketone and acrylamide (72% yield, dioxane, 70 °C) or acrylonitrile when catalyzed by AICI3 at 25 T.33
o •^Y^ 75
TsOH 80-130 °C
H2N 20 TsOH 80-130 °C
20%
54% 76
(Eq. 19) .Bn N'
77
[Ret. 8]
Methyl acrylate (37) was used as a reagent for annulation with the phenyl imine derivative of acetophenone. For example, phenyl imine 78 was converted to 79 by treatment with 37 in the presence of AICI3 (eq. 20).^'^ The enamine formed from acetophenone (80) and NH4OAC was converted to 82 by combination with 81 (eq. 21).^'^ In the case of 81, the aryl substituent provided a steric and electronic hindrance to the aza-annulation reaction, but the reactivity of the reagent was enhanced by the combined effects of the C02Et and CN functionality. Due to the nature of the substituents on the intermediate dihydropyridone, dehydrogenation occurred to form pyridone 82.
328
.O
u
AICI3
(Eq.20)
85%
37 78 [Ret. 34]
O^
NH4OAC EtOH Reflux, 1 h
Eta
(Eq.21) H.^^.CN
44%
80 [Ret. 17]
Other metliods for one-step pyridone formation include aza-annulation with either 57 or 86. The reaction of tetralone 83 and ammonia generated an imine that underwent aza-annulation with 57 to give 84, which was used as a key intermediate in the synthesis of derivatives of antitumor and antiviral alkaloid fagaronine (eq. 22).^^ Reagent 86 converted 85 to the corresponding pyridone 87 by aza-annulation and subsequent elimination of MeSH (eq. 23).^ Reaction of 86 with other aryl alkyl ketones, such as a-tetralone and 2-acetyl thiophene, only produced yields of 28% and 36%, respectively.
MeO^O
„.cX)6
NH3, MeOH 150°C,14h 72%
83
(Eq.22)
» MeO'
57
84
[Ret. 35]
O
KOH (Powder) DMSO 25 °C, 8 h
MeS 85
SMe 86
^
46%
329 Aza-annulation has been particularly valuable in the syntheses of a variety of different benzo[fl]quinolizin-4-ones from 88 (Scheme 7). In order to prepare naturally occurring alkaloids, introduction of a substituent on the 5-lactam product required the use of a p-substituted acrylate derivative. The reaction of 88 with crotonic anhydride in the presence of pyridine revealed a lack of regioselective incorporation of the acrylate derivative.^^ Both 89 and 90 were formed, and attempts to convert 89 to the desired dihydropyridone were unsuccessful. Double activation of the Michael acceptor allowed regioselective addition of 91 to 88, but a retro-Michael process occurred at an intermediate stage (92) which formed pyridone 93 as a result of elimination of diethyl malonate.^^ This elimination process could be avoided through in situ NaBH4 reduction of 92 to give 94.^'^ Alternatively, acrylate derivative 95 could be used, which intramolecularly trapped the enamine as 96 and prevented aromatization to the pyridone.^^ Subsequent cleavage of the epoxyetheno bridge of 96 was accompHshed by treatment with NaBHsCN at pH of 3.
MeO.
MeO^
MeO, MeO'
[Ret. 36]
^ 1 Y^
1 M e O ^ ^ M e O ^ ^ 88
1
EtO^O EtOH 25 °C OEt •r\SIOEt
EtO, Et02C
91
O O
C02Et
92
EtOgC^ ^COsEt
^^^-x OEt EtOgC ^cOgEt Scheme 7.
[Ret. 37]
330 Cyclic imine 88 was also used in the synthesis of several natural products (Scheme 8). Azaannulation of 88 with 97 led to the formation of 98, which was then transformed to the key intermediate 99 by alkylation, reduction, and decarboxylation.^^ Compound 99 was subsequently converted to (±)-tubulosine (lOO),^^ (±)-dihydroprotoemetine (101),^^ and (±)-emetine (102).39
^ M e O ^ ^
1
88
>s^N
1
[Ref. 40]
100
(±)-Tubulosine
(±)-Dihydroprotoemetine
102
(±)-Emetine
Scheme 8.
Aza-annulation of 88 with an unsubstituted acrylate derivative, 37, followed by in situ reduction of the resultant dihydropyridone with NaBH4, provided 103 as a key intermediate in the preparation of ipecac alkaloids (eq. 24)."^^
331
7
MeO. MeO.
37
NaBH4 Benzene/MeOH (1:1)
^
MeO'
(Eq.24)
MeO'
47%
88
MeO.
103
[Ret. 41]
Aza-annulation of 104 with a variety of acrylate reagents has been utilized in the synthesis of indoloquinolizidine alkaloid skeletons (eq. 25). Aza-annulation of 104 was affected with acrylic acid (91%), acrylic acid/DPPA (95%), acryloyl chloride/DMAP (64, 63%), and methyl acrylate (37, 52%) to generate the pentacyclic eburnane skeleton 105.^^2 Carbonyl reduction gave Wenkert's enamine (106), which was carried on in the synthesis of (±)-apovincaniine (107) and the clinically active synthetic analog (±)-Cavinton (108).^2 (Eq.25)
y
5 Steps
HO-
[Ret. 42]
91%
H0-, DPPA
95%
CI-, DMAP
63%
MeO-
52%
^
R=Me: 27% R=Et: 23%
Yield 105: X = 0 106: X = H,H
RO,
d 107: R = Me (±)-Apovincamine 108: R = Et (±)-Cavinton
Imine 109 has been used as an important building block in the synthesis of several natural products. DPPA, in conjunction with acrylic acid derivatives 110, provided efficient annulation when substituents were present on the acrylic acid species, such as the aza-annulation of 109 with cinnamic acid (110b) to give 111b (eq. 26).^^
HO
O DPPA
R^ 109
[Ret. 43]
a b c d e
Me Ph (E)- Me-CH=CHH H
H H H Me Ph
yield 78% 65% 73% 87% 85%
(Eq. 26)
332 Aza-annulation with 97 provided an efficient route to intermediate 112, which was converted to (±)-corynantheal (113), and also constituted the formal total syntheses of (±)-corynantheine and (±)-ajmalicine (Scheme 9).^^ Through a different route, 112 was efficiently converted to (±)dihydrocorynantheol (114).^^
OMe 97
MeO
84%
109
[Ref. 44 and 45] 114 (±)-Dihydrocorynantheol
Scheme 9.
115
°X>
MeCN, 5 h
\=^
4 Steps
^-
84%
OH
109
46% 117 (±)-Deplancheine
MeO^^.0 MeOH/Benzene \ (1:1) 25 °C, 24 h
\^
Q
^^g Acetaldehyde
EtO OEt
NaBH4, MeOH 0 °C, 50 min 87%, (From 109) [Ref. 46]
Scheme 10.
333 The indoloquinolizidine alkaloid (±)-deplancheine (117) was prepared through two complementary aza-annulation procedures (Scheme 10)."^^ When treated with the a-methylene lactone 115,109 was converted to 116 in good yield. However, after a four step sequence, (±)deplancheine was generated as only a 60% component of a three compound mixture. In order to circumvent this problem, 109 was treated with 118 to give the Michael addition product 119, and reductive cyclization completed the annulation process to give 120. Wittig-Homer homologation selectively formed the alkene to give (±)-deplancheine (117). 121
COgMe
(-1:1)
rr'COsMe II tBu02C C02tBu
NaBH4 THF
80%
COaMe
109 tBu02C Rose Bengal 500-W Halogen Lamp 20-25 °C, 2 h
C02tBu
57%
3 Steps 37% C02Me tBu02C
C02tBu
C02Me tBu02C
C02tBu
125 O (±)-Camptothecin [Ret. 47]
Scheme 11.
The aza-annulation methods developed for conversion of 88 to 94 were extended to the synthesis of the antileukemic and antitumor natural product (±)-camptothecin (125, Scheme ll)."*^ Aza-annulation of 109 with 121 in the presence of NaBH4 resulted in heterocycle formation to give 122 without subsequent elimination of the malonate species. A dye sensitized photo-oxidation promoted the rearrangement of the indolo[a]quinolizinone ring to the indolizino[l,2-/?]quinolone ring 123. Compound 123 was converted to 124, which constituted a formal total synthesis of camptothecin (125). 3.2 p-Aryl Enamine Substrates Both acyclic and cyclic aza-annulation substrates with aryl substituents in the P-position can be used effectively in the construction of alkaloid skeletons. Equations 27 and 28 illustrate examples in which aza-annulation can be performed directly from methacrylamide (127) and a carbonyl
334 compound.'^^ In the first case, CsF and Si(0Me)4 were used to promote enolization of 126, which led to formation of 128 (eq. 27).^^ When regiochemical issues arose, as for the unsymmetrical ketone 129, regioselective annulation occurred. Compounds 130 and 132 were formed through conjugate addition at the aryl substituted a-carbon (eq. 28).'*^'^^ Again, reaction with acrylamide reagents required additional enhancement, and in the transformation of 129 to 130, KOtBu was employed in order to activate the carbonyl substrate.^^ An altemative approach, activation of the Michael acceptor through use of 131, also resulted in regioselective aza-annulation to generate 132 (eq. 28).50 u
CsF, Si(0Me)4 80 °C, 5 h
(Eq.27)
••
76% 127
u
131
127 f-BuOK Dioxane 130
57%
^OEt
EtOH Reflux, 5 h 129
(Eq.28)
67%
[Ref. 49]
The use of enamine and imine derivatives of carbonyl substrates was also an effective means of performing aza-annulation. Treatment of either enamine 133 or imine 134 derivatives of (3tetralone with acrylamide resulted in the formation of 135 with (85% yield)^! or without^ solvent (eq. 29). In contrast, aza-annulation of the corresponding methyl enamine of P-tetralone with methyl methacrylate generated a mixture of products.^2
H2N
(Eq.29)
20 TsOH 80-130 °C
135
134
335 Aza-annulation of a number of dimethoxy-substituted p-tetralone derivatives, such as those represented by 136, with acrylamide was used to produce 137. In turn, 137 was an important intermediate in the synthesis of conformationally restricted congeners of dopamine (eq. BO).^^ 1)80°C, 3h 2)130°C,0.5h
H2N
(Eq.30)
> •
81%
20 136
137
OMe
MeO'
OMe
Control of ring fusion through post aza-annulation modification was also employed in the synthesis of benzoquinolinone steriod analogs, which have demonstrated selective and potent inhibition of human type I 5a-reductase enzyme (eq. 31).^^ Aza-annulation was performed with acrylamide (20) and tetralone 138 to regioselectively generate the quaternary carbon of 139, and ionic reduction led to formation of the trans fused product 140. Similarly, aza-annulation of enamine 141 with acrylamide generated enamide 142, and ionic reduction gave 143 as the trans fused ring system (eq. 32).^"^ Subsequent enantioselective syntheses of these molecules are discussed in Section 8.
) NaH, Mel 2) EtaSiH, TFA
,r±? =
(Eq.31) O^ N
139
1) NaH, Mel 2) EtaSiH, TFA O^ N
(Eq. 32)
[Ref. 54]
Formation of the methyl enamine of 144, followed by aza-annulation with methacryloyl chloride provided an 80:20 mixture of the desired tetracyclic system 145 to the N-acylation product
336 146 (Scheme 12).^^ Reductive modification of 146 was selectively performed to access either the trans or cis ring fusion for total synthesis of (±)-festuclavine (147) or (±)-costaclavine (149), respectively.
1) MeNH2 2)
O
CI
\ 66%
1)LiAIH4 2) Mn02 16P/o
[Ret. 55]
Scheme 12.
4 . VINYLOGOUS AMIDE DERIVATIVE SUBSTRATES
R2'
R 2 ^
R3 R3
Section 4.1 Vinylogous Amides p-Enamino Ketones
4.1
Section 4.2 Vinylogous Carbamates P-Enamino Esters
Section 4.3 Vinylogous Ureas p-Enamino Amides
Vinylogous Amide Substrates 4.1.1
Acyclic Dione Derivatives
p-Diketone substrates have been valuable in the aza-annulation reaction with a,P-unsaturated carboxylic acid derivatives, and both acyclic and cyclic P-diketone species have been investigated. The simplest acyclic p-diketone, 150, underwent condensation reaction with BnNH2 to generate the
337 corresponding (3-enamino ketone 151 (Scheme 13).56 Regioselective 5-lactam formation was affected through aza-annulation with this vinylogous amide, and solvent effects played an important role in this reaction. For example, reaction with acryloyl chloride in benzene at reflux generated 152 in 44% yield,56 while the same reaction generated an 94% yield when performed in THF.57,58 x^e resultant vinylogous amide functionality of dihydropyridone 152 was catalytically reduced to give the predominantly cis substituted 6-lactam 153.^"^'58 Epimerization of the diastereomeric mixture, followed by Baeyer-Villiger oxidation established the trans stereochemistry of the oxygen substituent relative to the methyl substituent of intermediate 154, which gave 155 upon base catalyzed hydrolysis.
O
BnNHs, TsOH Benzene Reflux
O
N
AA
150
151
74%
OH
V
154
THF 94% (from 150)
1) DBU (cisrtrans, 24:76) 2) m-CPBA CF3CO2H
O
NaOH H2O
'")b 155
O
O 1 atm of H2 Pd/C , 81% NaaCOg p^ ^"^N
Cis:Trans (90:10)
45%
o
Scheme 13.
[Ref. 56-58]
Aza-annulation with unsymmetrical (3-diketone 156 resulted in regioselective generation of 157, which gave 158 upon reaction with acryloyl chloride in benzene (47%) or THF (96%) (eq. 33) 56,57,59 j h e reaction of cinnamoyl chloride with 157 in benzene gave 158 in 30% yield.^^ O O
BnNH2, TsOH Benzene Bn Reflux
O 156
H
(Eq.33)
CI
" v ^
[Ref. 56, 57, and 58] R H H Ph
Solvent Benzene THF Benzene
156-> 158 47% 96% 30%
338 Enamino ketoester 159 (in equilibrium with the corresponding imidizolidine) efficiently underwent aza-annulation with acryloyl chloride in the presence of pyridine and DMAP to give 160 (eq. 34).60 This dihydropyridone was then converted to 161, the pentacyclic skeleton of the 21epimer of the aspidospermine alkaloids.
(Eq.34) TiCU CICH2CH2CI 80 °C
/I.
63% 161
C02Me
Pyridone products were directly accessible through aza-annulation with p-heteroatom substituted acrylate derivatives. Aza-annulation of 150 with 86 in the presence of K2CO3 led to the formation of 162 upon elimination of MeSH (eq. 35).^ When facilitated by NaOEt, reaction of 150 with 163 resulted in carbon-carbon bond formation through conjugate addition and subsequent elimination of MeOH (eq. 36).^^ Intramolecular lactam formation generated bicyclic species 164.
O
O
-V"
MeS 150
SMe
K2CO3, DMSO 100°C,3h 6P/0
^
(Eq.35)
86
[Ref. 6]
XX
NaOEt
(Eq.36)
69%
150 [Ref. 61]
Pyridone products 166 were also generated through aza-annulation of (3-enamino ketones 165 with 57 (eq. 37).62 p-Enamino ketone 167 could be generated either by the condensation of 150 with ACONH4 or by hydrogenation of isoxazole 168 (eq. 38). Subsequent aza-annulation with 57 gave 169.^2 Application of this methodology to the synthesis of medorinone (170) was completed by the conversion of 169 to 170.
339 DMF 1)25°C,3.5h 2) Reflux, 24 h
MeO,^0 "^N IN - ^ O ^
I
165
57
[Ref. 62]
O
O
(Eq.37)
•
R = Et (43%) 4-F-C6H4 (52%) Thiophene (47%) CgHs (72%)
R^^O
166
ACONH4
AA 150
^
N
^
O
57 DMF 1)25°C, 3.5h 2) Reflux, 24 h
(Eq- 38) u ""N
62%
1) (Me2N)2CHOtBu 2) ACONH4/DMF 59% 170 Medorinone
4.1.2
Cyclic Dione Derivatives
Cyclic p-diketones and their derivatives have also been the subject of aza-annulation studies. Methodology studies with enamine derivative 171 established the effectiveness of acryloyl chloride (64) for the formation of 172 (eq. 39).^^ Aza-annulation with a p-substituted aery late derivative was most efficiently accomplished through the use of diester Michael acceptors, as exemplified by reaction of 173 with 174 in the generation of 175 (eq. 40).^^ Benzene Reflux r ^
[Ref. 56]
64
(Eq. 39)
47%
171
(Eq.40) 71%
340 Aza-annulation of the benzyl enamine derivative of 179 was employed in the synthesis of (±)-5-epipuniiliotoxin (Scheme 14).^^ Condensation of 179 with BnNH2 followed by reaction with acryloyl chloride (64) gave the key bicyclic intermediate 180. Catalytic hydrogenation selectively established the cis ring fusion of 181. Addition of MeMgBr gave 182, and stereoselective dehydroxylation generated 183. Compound 183 was converted to (±)-5-epipumiliotoxin (184) by sequential deprotection, imidate formation, alkylation, and reduction procedures. 1)3atmofH2 Pd/C NaaCOa
1) BnNH2 Benzene Reflux
> •
& c
2)
179
2) (C0CI)2, DMSO, NEt3
THF CI
85%
64
75%
4 Steps t
52%
25%
182
184 (±)-5-Epipumiliotoxin Scheme 14.
[Ref. 64]
A
NC, 2
1
185
1)Mel 2) H2/Pt 3)LiAIH4 4) AcCI/Pyridine 60%
(±)-A/crMethyl-A/[3-Acetylphlegmarine (Mixture of 4 epimers) [Ref. 65]
1)NaH, CS2 2) Mel 3) BuaSnH AIBN
'"^N H.j
OH
341 Acrylonitrile (2) has also been used for heterocycle formation from P-diketone substrates. Conjugate addition of 185 to acrylonitrile produced 186 with previously reported conditions,^ and cyclization under hydrolytic conditions generated the corresponding lactam 187 (Scheme 15).65 Dissolving metal reduction provided the trans ring fused product 188, which was subsequently converted to 189. Reduction of 189 completed the synthesis of the Lycopodium alkaloid (±)-Namethyl-A^p-acetylphlegmarine (190) as a mixture of 4 diastereomers. Direct reaction of 191 with acryhc acid resulted in efficient formation of 192 (eq. 41).^^ Subsequent dehydrogenation at elevated temperatures provided the aromatic species 193, which was a key intermediate in the synthesis of 194, a compound that displays strong ^-blocking activity. Under similar conditions, reaction of 191 with crotonic acid, cinnamic acid, and ethyl acrylate did not generate the corresponding bicychc alkaloid skeletons.
A, 191
10%Pd/C 195°C, 3h Decaline 140 °C 3h 95%
OH
74%
[Ret. 66]
When treated with acrylic acid, bicyclic enamine 195 was converted to the tricyclic vinylogous imide 196, which was then incorporated into the synthesis of the Lycopodium alkaloid annotinine (197) as well as an annotinine degradation product (eq. 42).^'^ u 135°C
23 Steps
63%
<1%
(Eq. 42)
The formation of pyridone products can be accomplished by the use of acetylenic esters. Aza-annulation of 198 with 57 resulted in the formation of 199 (eq. 43).27,68 Compound 199 was converted to 200, which was a key intermediate in the preparation of 4-aza-19-norsteroids. Reaction of 57 with hydrazine enamine 201 gave 202 which allowed access to the aromatized derivative 203 (eq. 44).^^ Similarly, annulation of hydrazine enamine 204 with 205 resulted in formation of pyridone 206, which had a -C02Me substituent (3 to the lactam carbonyl (eq. 45).69
342
O
\
70%
I
H
.dy
57
MeO
(Eq.43) R""N 200
199
198
[Ret. 27 and 68]
1)NaBH4 2) H2SO4 3) Pd/C
MeO^^O
/N.^.H
(Eq.44)
i
50%
30% 57 202
201 [Ret. 69]
.N.^.H
MeO^^O
'^
OMe
28%
(Eq.45)
O'^OMe 204 [Ref. 69]
205
Aza-annulation of aminonaphthoquinone 207 with either 208 (70%) or 209 (75%) resulted in A^-acylation to generate 210 (eq. 46)7^ Subsequent intramolecular condensation under acidic conditions led to the formation of the extended aromatic system of 211.
V^-A -.ju
N
208
°^-
70% O
207 [Ref. 70]
209 75%
o
Xylene, 130 °C 210
(Eq.46) H2SO4 25 °C 0.5 h 88%
O
343
4.1.3 P-Enamino Imine Substrates (Eq.47)
u 37
"
KH THF, 0 °C
H2SO4 THF 60 °C, 4 h
70%
80%
» 214
[Ret. 71]
(3-Enamino imines can also be utilized as masked ketone substrates in the aza-annulation reaction with a,P-unsaturated esters. Annulation of 212 with 37 proceeded with carbon-carbon bond formation by Michael addition followed by cyclization with the least substituted nitrogen to give 213 (eq. 47)7^ Potassium hydride was important in this reaction, as lithium hydride did not promote the aza-annulation reaction. Hydrolysis of the imine functionality gave the corresponding ketone product 214. Reaction of 37 with 215 under similar conditions resulted in lactam formation with the least substituted benzylic nitrogen to give 216 despite the substantially different electronic environment (eq. 48).^^ Through this methodology, p-enamino aldehyde equivalent substrates could be used in aza-annulation reactions, and the aldehyde could be deprotected by hydrolysis to give 217.
(Eq.48)
u H.^.H
^
6
215
MeO' 37 KH THF, 0 °C
H2SO4 THF 60 °C, 4 h
74%
93%
»
f
217
[Ref. 71]
When the reactivity of hydrazine imine 218 was investigated through aza-annulation with 205, pyridone formation also occurred with the least substituted nitrogen to form 219 (eq. 49).'72 Hydrolysis of 219 led to the formation of 220 in good yield.
344 (Eq. 49)
MeO,.0 205
H2SO4 THF 25 °C 12h
H^N'^
^^
218
OMe
O'^OMe MeOH
V
\
80%
80%
220
Ph
[Ret. 72]
4.2 Vinylogous Carbamate Substrates 4.2.1 Acyclic Enamino Esters Condensation of P-keto ester 221 with either BnNH2, Me2NNH2, or a source of NH3 led to the formation of P-enamino esters 221, 222, and 223, respectively. These p-enamino esters could be treated with either acrylic acid,'^^ acryloyl chloride,^^'^'^'^^ or crotonyl chloride,^^'^'^'^^ to generate the corresponding 5-lactam products (eq. 50). Interestingly, p-propiolactone could also be used to affect the same transformation in 55% yield.'^^ When crotonyl chloride was used for azaannulation, the rate of reaction was slowed, but the 1,4,5,6-tetrasubstituted dihydropyridone 226 was formed in good yield.^^'^'^'^^ In these reactions, solvent played an important role in the efficiency of product formation, and reaction in THF provided significantly higher yields of 225 or 226 than did aza-annulation in benzene.
0
0
208
R^NH2
I
R\
R2
^^
^ 221: R^ =H 222: R^ = Bn 223: R^ = MeaN Ref. [73] [56] [56] [57, 59] [56] [57, 59] [59]
X HO CI CI CI CI CI CI
(Eq.50)
0
x^
1. ^ R 2
EtO"^ 0
Solvent PhCI Benzene Benzene THF Benzene THF THF
yield 80% 39% 42% 94% 50% 75% 75%
R^ _RL Product H 224 H 224 H H Bn 225 H Bn H 225 Bn Me 226 Bn Me 226 MeaN H 227
Application of this methodology led to the synthesis of indolizidine and quinolizidine alkaloids. Condensation of 228 with benzyl amine followed by aza-annulation produced dihydropyridone 229 (Scheme 16).^9''7'^ Stereoselective hydrogenation in the presence of Na2C03
345 established the cis stereochemistry (>95:5) of the substituents on 230. This key intermediate was converted to (±)-5-epitashiromine (231, n=3)59 and (±)-lupinine (232, n=4).74
1) BnNHs, TsOH 2) AcryloyI Chloride
'"°>)M0E.
n = 3; 82% n = 4; 80%
228 a: n = 3 b: n = 4
>
[Ret. 59] 231 (±)-5-Epitash(romine
232 (±)-Lupinine
Scheme 16.
An alternative approach to the formation of P-enamino esters was through conjugate addition of a primary amine to an alkyne (eq. 51).^^^^^ This complementary method provided a route from alkyne 205 to p-enamino ester 233, which underwent aza-annulation with acryloyl chloride to give 234. Catalytic hydrogenation efficiently gave the cis substituted tetrahydropyridone species (235).
64 J ^ MeC^O
BnNH2 Benzene Reflux
^ o , ^ , O'^OMe 205
THF " Reflux
Bn,^..H
233
Bn^
0 A^
3 atm of H2 Pd/C MeOH
^
OMe 84% (From 205)
MeO
(Eq.51)
MeO^^ OMe ^ 234
80%
0
^
235
[Ret. 57 and 59]
Conjugate addition to terminal alkyne 57 resulted in a lower yield of the overall enamine formation/aza-annulation sequence, but still provided a route to 236 (Scheme 17).^^ Functionalization of C-5 with a heteroatom was accomplished by modification of the carboxylate substituent. Conjugation with the lactam functionality was removed through catalytic hydrogenation of 236 to give 237. Hydrolysis of the ester efficiently generated the corresponding carboxylic acid 238, which underwent a modified Curtius rearrangement after treatment with DPP A to give 239 in low yield.
346 1) BnNH2, TsOH Benzene, Reflux 2) AcryloyI Chloride THF, Reflux
OMe
ill
S./0
57
'
Q [j
0
3 atm of H2
I^^N^OMe
f O^
Bn
1) NaOH, H2O 2) HCI, H2O
90%
Bn
T
1) DPPA, NEt3, f-BuOH, 2)HCI 3) NaOH, H2O
^ ^ N H 2
^ 1
il
24%
239
237
Bn
236
0
^'
Bn
238
Scheme 17.
[Ref. 58]
Homologation of the lactam carbonyl was also efficiently accomplished through standard methods (Scheme 18).58 Hydrogenation of 240 stereoselectively generated 241, which had a cis relationship of the substituents on the heterocyclic system.57,59 Conversion of 241 to the corresponding thiolactam 242, followed by Eschenmoser contraction/sulfide extrusion gave (3enamino ester 243.^^ Stereoselective reduction of the a,(J-unsaturated ester moiety could be achieved with NaBHsCN (predominantly 244) or H2 with Pd/C (predominantly 245).
sx
3 atm of H2 Pd/C^Na2C03
C02Et
O^^N Bn
^^C02Et
91%
240
Bn
241
Lawesson's Reagent 99%
^
XX
C02Et
S^^N Bn
1) Et02CCH2Br 2) NEt3, PPha
a
COsEt
Et02C^^J^
Bn (99% Yield)
[Ref. 58]
79%
^.'-s^C02Et
' Et02C^^J^
^
Bn
Bn 244 92
245 8
Reagent NaBHaCN, pH 4.0
15
85
H2.10% Pd/C Na2C03 Scheme 18.
242
243
347 The formation of P-enamino esters by conjugate addition of benzyl amine to acetylenic ester 246 was applied in the synthesis of natural products (Scheme 19).^^'^^ In situ treatment of 247 with acrylic anhydride (248) gave 5-lactam 249. Catalytic hydrogenation generated 250 with >98:2 cis stereoselectivity. Conversion to the corresponding methyl ketone followed by epimerization at C5 generated 251 with the stereochemistry desired for further elaborations. Baeyer-Villiger oxidation and protecting group manipulation gave 6-lactam 252. This key intermediate was transformed to the a-D-mannosidase inhibitors (±)-mannonolactam (253)^^ and (±)-deoxymannojirimycin (254),^^ as well as to the antibiotic and anesthetic agent (±)-prosopinine (255).^8''75 U
BnO'
P
246
BnNHs THF Reflux
O
THF Reflux OEt
OEt
247
62% (From 246)
249
EtQ-^O
1 atm of H2 Pd/C Na2C03
80%
Eton
1) CF3CO2H m-CPBA 2) KOH, H2O 3) KOH, BnBr
1) MeMgBr NEt3 2) DBU, 25 °C
3nO^Xj EtO^O
OH OH
OH OH
OH OH
253
(±)-f\/lannonolactam
254 (±)-Deoxymannojirimycin
255 (±)-Prosopinine
Scheme 19.
[Ref. 58 and 75]
Products related in structure to 249 and 250 could be accessed through condensation of benzyl amine with tetronic acid (256) followed by aza-annulation with acrylic anhydride to give 258 (eq. 52).58 Stereoselective generation of the cis ring fusion of 259 was accomplished by catalytic hydrogenation.
348 O
O (Eq.52)
BnNHs TsOH Benzene Reflux
OH ,
1 atm of H2 Pd/C Na2C03 EtOH
THF Reflux
257
83%
71% (From 256)
259
Incorporation of a P-phenyl substituent in the aza-annulation process, by the use of the appropriate acrylate reagent, was even more difficult than the reaction with crotyl derivatives and necessitated the use of doubly activated 260 to access 261 (eq. 53)7^
"'-H^" o
N
O
A^
75%
221
EtO^O
260
o
261
[Ref. 76]
Reaction of 205 with benzotriazole enamine (262), generated through conjugate addition to the corresponding alkyne, led to pyridone 263 (eq. 54)7^ Formation of 263 was accompanied by a minor amount of a side product (264, 8%) in which the aromatic species on the nitrogen participated in the cyclization.
N-N
205 I
N-N
,H N' 262
(Eq. 54)
E
E = C02Me [Ref. 77]
Pyridone formation could also be accomplished by reaction of 208 with 86 in the presence of K2CO3 to yield 265 upon elimination of MeSH (eq. 55).^ Alternatively, reaction of 267 with Penamino ester 266 led to the formation of the pyridone 268 in good yield (eq. SG).^^
349
O
. "'•'Y" K2CO3 DMSO
O OEt
MeS
SMe
II
86
208
(Eq- 55) ^SMe
77%
EtO"^0
265
[Ref. 6]
o Pyridine
MeO" OEt 266
267 MeO.
[Ref. 78]
Y s;
s
II
(Eq.56)
81%
EtO^O
268
0
The use of 1,3-dicarboethoxyallene (269) also provided a route to 5-lactam products through aza-annulation (Scheme 20)7^ Treatment of 221 with 269 resulted in formation of the pyridone derivative 270. Similar chemistry was performed with p-enamino ketones, but studies were far more extensive with the P-enamino ester substrates.^^ Elaboration of 270 led to hydrolysis of the ester to give 271 followed by construction of a lactone ring to give 272. O
1 eq. of NEt3 AcOH^oluene(1:1) 100°C, 5h
EtO'
^^N'^ O
OEt
OEt 221
269
O
>
^ ^ N ^
OEt
70% EtO-^O
270
NaOH, EtOH Reflux, 3 h MeO
01
AcOH
EtO ^ O [Ref. 79]
272
31% (From 270) Scheme 20.
Aza-annulation methodology that involved 269 was applied to the synthesis of (±)camptothecin (125, Scheme 21).^^ Combination of 205 and 273 generated 274, which was transformed to 275 under mild conditions by aza-annulation with 269. Intermediate 275 was then converted to 276, which was carried on to (±)-camptothecin (125).
350 O
EtO
1 eq. of NEta
Eto->.
\
N' H
EtO" 273
EtO EtgO
EtO^'^^^N'^ O
•
MeO O
MeO^A^
O 205
269
MeOH 25 °C
V
OMe
OEt
EtO N
EtO"
OE OEt
^
45% ^
274
O'^OMe 275
OMe
4 Steps
6 Steps
1% 125
(±)-Camptothecin Scheme 2 1 .
[Ret. 80]
An interesting acrylate derivative, 277/278, was also used for pyridone formation (eq. 57).81 Treatment of 221 with 277/278 resulted in formation of the corresponding a-acyl substituted pyridone 279.
OH DMF
^-N'^ O
OEt 221
68% 277
278
(Eq.57) EtO ^ O
279
[Ret. 81]
Malonic acid derivatives have also been effective as 1,3-dielectrophiles for the formation of 4hydroxypyridone products. Although not truly acrylate-type reagents, the tautomeric form of these species was similar in nature to their acrylate counterparts. Reaction of 280 with 281 led to the quantitative formation of 282 (eq. 58).^2 Examples of (J-enamino ketone substrates were also reported for reaction with 281, but their use was limited. ^^
351 O MeO,
MeO,
(Eq. 58)
281
100% 280 [Ret. 82]
The formation of pyridones by aza-annulation with aryl-substituted malonate derivatives was shown to be highly dependent upon the substituent pattern on the aryl ring (Scheme 22).^^ Ring formation was most efficient when the aryl group on the malonate reagent was either unsubstituted (phenyl, 283) or substituted in the 4-position (286). Product formation was significantly decreased when the malonate reagent differed from this substitution pattern. Examples of typical malonate azaannulation reactions are illustrated by the conversion of 221 to 284 by treatment with 283, and the analogous formation of 287 from the reaction of 221 and 286.^^ Subsequent reactions of 284 and 287 were highly dependent on nitrogen substitution. During hydrolysis of 287 with NaOH/H20, the intermediate carboxylic acid species decarboxylated rapidly to give 288. In contrast, hydrolysis of 284 under the same conditions gave the carboxylic acid as an isolable intermediate, and extreme conditions were required to produce 285 through decarboxylation. Both 285 and 288 were found to inhibit the growth of Mycobacterium tuberculosis.^^
1) NaOH/HaO Reflux 2) 220 °C 2,4,6-Trichlorophenol
MeO'
^^^
MeO^O 285
R = Bn OEt 221: R = H 222: R = Bn
S.R = H
220 °C, 30 min
MeO' MeO ^ O [Ref. 83]
OH
80%
Bromobenzene Reflux, 1 h
Scheme 22.
352
^
°^^
OEt
Benzene 20 °C, 2 h
r=\
290 1) KOH, H2O Reflux 2) H C I / ^ ^ 85%
Brs
SOCI2, Benzene C
V.
50%
A very unique reagent (290) for aza-annulation with P-enamino ester 289 was reported (Scheme 23).^'^ Combination of 289 and 290 resulted in the formation of 291, which was further modified through a variety of pathways to produce ring-opening of the cyclopropane ring. The cyclopropyl ring could be opened to place the benzyhc fragment at C-3 (292), remove it entirely through hydrogenation (293), or situate the substituent at C-4 (294). 4.2.2
Cyclic Enamino Esters u
(Eq.59) 296 Ref. [85] [57, 59] [85]
Conditions EtO^^O Conditions Pyridine, Toluene, Reflux THF, Reflux NaH, Et20, 25 °C
CI CI EtO
rield 72% 87% 75%
A variety of aza-annulation chemistry has focused on the conversion of cyclic enamino ester 296 to a variety of substituted indolizidinone products (eq. 59). The simplest aza-annulation process involved treatment of 296 with either acryloyl chloride (64) or ethyl acrylate, which generated the
353 bicyclic product 297.5'7'59,85 when acryloyl chloride was used, reaction occurred in either toluene or THF at reflux, and the reaction in THF produced a slightly higher yield. Deprotonation of 296 followed by treatment with ethyl acrylate, produced comparable results at ambient temperature.^^ Pyridone formation was achieved through a two step process, by sequential conjugate addition and cychzation (eq. 60).^^ Conjugate addition was accomplished through extensive heating of 296 with 57, and cyclization of 298 to 299 was facilitated by the subsequent addition of NaH.
MeO^^O 57 Benzene Reflux, 4 d
.H OEt
(Eq.60)
NaH Benzene Reflux, 1 h OEt
60%
54%
296 MeC^O
[Ret. 86]
The conversion of 296 to 297 was used as a key ring forming step in the synthesis of (±)tashiromine (301, Scheme 24).^9 Stereoselective introduction of the two vicinal stereogenic centers was accomplished through catalytic hydrogenation of 297, which resulted in >95:5 stereoselectivity for generation of 300. Further reduction of 300 gave (±)-5-epitashiromine (231), which was then efficiently converted to (±)-tashiromine (301).
THF Reflux 296
ffr/o
- c6
3 atm of H2 Pd/C Na2C03
Eton
297 E f O ^ O
OEt
95%
91%
CO
HE "^OH 301 (±)-Tashlromine [Ref. 59]
1)LiAIH4 2) H2O, NaOH
1) (CIC0)2, DMSO NEt3 2) Piperidine, pTsOH 3) (C00H)2, H2O 4) LiAIH4 58% Scheme 24.
OH 231 (±)-5-Epitashiromine
354 O
A-N'
(Eq.61) ^R1
O OEt
Conditions
296 [Ret. 85] Conditions Pyridine, Toluene, Reflux Pyridine, Toluene, Reflux NaH, EtaO, 25 °C NaH, Et20, 25 °C
X CI CI EtO EtO
R^ Me Ph Me Ph
rield
Product a b a b
Yield 72% 53%
^ —
— —
63% 74%
O
[Ref. 87]
[Ref. 87]
[Ref. 86]
^ r-*^A^ NH; Etc " O 29%
305
97%
Benzene Reflux 1h
1) Benzene Reflux, 1 h 2) 205 °C, 3 h
HN 306 )
309
54% Benzene Reflux, 48 h
/^N'
296 a: R=Et
O OR
205 MeO > § s 1) Benzene Reflux, 2 h ^^Q 2)NEt3, MeOH 25 °C, 15nnin
O O
EtO OEt
(EtO)20P DP^^-^^OEt 311
MeO^^O 312 [Ref. 88]
74%
72% 312 54%
OEt MeC^'O Scheme 25.
[Ref. 88]
310 [Ref. 86]
355 The reaction of crotonic and cinnamic acid derivatives with 296 was very dependent upon the nature of the carboxylate derivative used in the aza-annulation and the conditions of the reaction (eq. 61). When the corresponding acid chloride was used in the presence of pyridine, only A^-acylation was observed to yield 303.^^ However, when 296 was deprotonated with NaH and then treated with the ethyl ester of either crotonic or cinnamic acid, aza-annulation occurred to give 302. The use of aza-annulation to generate dihydropyridone and pyridone products with substitution p to the lactam carbonyl was performed with a variety of other reagents (Scheme 25). The reaction of 296 with maleic anhydride (304) gave a high yield of the expected dihydropyridone annulation product 305 to generate a -CO2H substituent.^^ In contrast, the use of maleimide (306) under the same reaction conditions gave a good yield of the corresponding Michael addition product, but this species could only be cyclized by heating with NaH to give a low product yield of the amidesubstituted 307.^^ Fumarate derivative 308 was used the prepare the corresponding estersubstituted 309.86 Formation of pyridone products was accomplished in a number of ways (Scheme 25). The reaction of 205 with 296 gave the corresponding conjugate addition, and treatment with NEt3 facilitated cyclization to complete the aza-annulation process and the formation of 310.^6 Disubstituted pyridone 312 was prepared through aza-annulation with either 311 or allene 269 as part of a synthesis of camptothecin precursors.^^
/^N^^ O
EtaN OMe
296b
OMe
92%
313 86%
1) f-BuOK, EtI 2) (CH20)n f H2SO4, H2O
4 Steps 8%
125 (±)-Camptothecin [Ret. 89]
Scheme 26.
Aza-annulation of 296b with 313, the chloro analog of 311, led to formation of bicyclic pyridone 314 (Scheme 26).^^ Modification of this intermediate gave 315, which was then decarboxylated, oxidized, and utilized in the Friedlander quinoline synthesis to give (±)-camptothecin (125).
356 Formation of the corresponding a-substituted 6-lactams was accomplished by reaction of 296 with itaconic anhydride (316) to give dihydropyridone 317,^^ while pyridone formation was accomplished by aza-annulation with diester 47 to generate pyridone 318 (eq. 62).^'^
EtO" 316
"
Benzene Reflux, 1 h
,0H
92%
cu„„
(Eq.62)
OEt
47 OEt Benzene Reflux, 12 d
O
O OEt
60%
EtO'^O
296
318 [Ref. 87]
Application of aza-annulation with cyclic enamino esters was reported in the synthesis of angiotensin converting enzyme inhibitor A58365A (323, Scheme 27).90 Aza-annulation of proline derivative 319, which was obtained in 4 steps from L-pyroglutamic acid, with a-methyleneglutaric anhydride (320) led to the formation of indolizidinone 321 as a mixture of diastereomers. Esterification followed by oxidation with DDQ gave 322, which was converted in 4 steps to the desired target 323. MeO.
Benzene Reflux
MeO. OH
95%
OBn 319
320
BnO ^ O 1)CH2N2 2) DDQ
HO
40%
MeO. 4 Steps
OH
[Ref. 90]
(67:33)
OMe
33% A58365A
BnO^O Scheme 27.
Aza-annulation of 324, the six-membered ring analog of 296, with acryloyl chloride or ethyl acrylate led to the formation of a mixture of isomeric enamide products 325a and 326a (eq. 63).^^ The corresponding crotyl derivatives (b) were also successfully employed in these aza-annulation
357 studies. Regioselective formation was only observed when ethyl 3,3-dimethylacrylate, a derivative disubstituted (R,R = Me,Me) in the P-position, was used, but the yield was poor (34%). The lack of regioselectivity in this aza-annulation process has prevented the otherwise straight forward conversion of 325 and 326 to natural product targets such as (±)-lupinine (232). O (Eq. 63)
OLX„
Conditions
324
EtO^O Conditions
EtO^O
Product
325
326
Pyridine, Toluene, Reflux
CI
H
a
70
30
82%
NaH, Et20,25 °C
EtO
H
a
65
35
49%
Yield
Pyridine, Toluene, Reflux
CI
Me
b
60
40
68%
NaH, Et20,25 °C
EtO
Me
b
52
48
60%
[Ref. 85]
The seven-membered ring analog, 327, showed different properties than those observed for 324 (eq. 64).^^ Complete regioselective formation of 328 was observed for the acryloyl or crotyl derivatives. In these aza-annulation reactions, the use of acid chloride reagents resulted in higher yields than the corresponding ethyl esters.
(Eq. 64)
Conditions
^
327 Conditions Pyridine, Toluene, Reflux
CI
NaH, Et20, 25 °C Pyridine, Toluene, Reflux
EtO CI
Product H H
a a
Me
b
Etc " " O
328
Yield 95% 56% 68%
[Ref. 85]
An interesting variation on the aza-annulation of cyclic enamino ester substrates is illustrated in Scheme 28.^1 The amino acid derivative 329, formed by the reaction of the corresponding amino phenol with 205, underwent conjugate addition with a second equivalent of 205 to give 330. Treatment of 330 with a nucleophile such as EtOH or pyrrolidine resulted in the formation of 332. Formation of 332 was suggested to proceed through intermediate 331.^^ Evidence for this intermediate was acquired by isolation of the nitrogen analog 334, prepared in the same manner by
358 reaction of 333 with 205 (eq. 65).92 Several aromatic and aliphatic 1,2-diamine substrates, including unsymmetrical diamines, were successfully employed in this aza-annulation reaction. Overall, the reactions in Scheme 28 and eq. 65 illustrate an interesting class of conformationally restricted amino acid derivatives.
Me 0 , ^ 0
r^
OMe 329
Dioxane Heat
OMe
O'^OMe 206
HY, CH2CI2 25 °C, 3 h Y = OEt; 89%
Y= N J
MeO
Scheme 28.
[Ret. 91]
MeO,^0
O [Ret. 92]
4.2.3
; 92%
OMe 333
1) Dioxane Reflux (54%) 2) DMSO Reflux, 1 h (64%)
(Eq. 65) OMe
O'^OMe
MeO^O ^
334
205
Tetrasubstituted Enamino Esters
Tetrasubstituted enamino esters have also been employed in the aza-annulation reaction with acryloyl chloride (64) (eq. 66-68).^^ As observed for the examples in which trisubstituted enamino esters were used, the ester functionality directed regioselective enamine formation, and the resulting carbon-carbon bond formation occurred at the more substituted site. In the case of tetrasubstituted enamino ester substrates, carbon-carbon bond formation resulted in the generation of a quaternary carbon with deprotonation to form the enamide functionality which occurred exocyclic to the 6-lactam ring. This process was performed with 335 to generate the fused bicyclic ring system 336, which was reduced by hydrogenation to a mixture of diastereomers 337 (eq. 66).59 p.Keto lactone 338 was exposed to the same reaction conditions to give the corresponding spirocyclic 5-lactam 339 (eq.
359 67).59 Hydrogenation of this annulation product resulted in stereoselective formation of 340. In each of these examples, formation of a stereogenic center occurred, and this methodology was used as the ground work for asymmetric induction in the aza-annulation reaction (see Section 8).
O
A^COsEt
1) BnNH2, TsOH Benzene, Reflux 2) AcryloyI Chloride THF, Reflux > •
89%
U
3 atm of H2 Pd/C, Na2C03 EtOH
r
TcogEt
335
^.
85%
I
336
rco2Et 337
(56:44)
[Ref. 59]
O
O
1) BnNHg, TsOH Benzene, Reflux 2) AcryloyI Chloride THF, Reflux
3 atm of H2 Pd/C, Na2C03 EtOH
^^
Bn^ N
(Eq.67)
83%
84% 339
338
340
[Ref. 59]
Aza-annulation of 341 with acryloyl chloride (64) provided 342, which was reduced to give the indolizidine-type ring skeleton 343 (eq. 68).^^
CI 64
/^NH \As^C02Et C02Et
THF, Reflux
^^ 75%
3 atm of H2 Pd/C, EtOH
-N Et02C C02Et
85%
342
341
^^
(Eq.68) "N Et02C C02Et 343
[Ref. 59]
4.2.4 a-Amido Aza-Annulation Reagents There has been increased use of acrylate reagents with an a-nitrogen substituent, and the corresponding aza-annulation reaction products were a-amido 6-lactams, which represented an interesting class of conformationally restricted peptide analogs. Oxidation of the dihydropyridones that resulted from aza-annulation led to the corresponding pyridones.
360
"°-VY
(Eq. 69)
NEt3, DPPA DMF, 0 °C
344
R = Et; 78% R = H; 88% 345: R = Et 346: R = H
104: R = Et 109: R = H [Ret. 42]
The first methods for aza-annulation with a-amido-derived acrylate reagents, such as 344, involved activation of the carboxylic acid toward acylation through the use of DPPA (eq. 69)."^^ Efficient aza-annulation of 104 and 109 generated the corresponding amino acid derivatives 345 and 346.
NEt3 Dioxane, 80 °C
"•N-P
(Eq.70)
OEt
O
221: R = H 222: R = Bn 266: R = Ph
EtO^O
R = H; 64% R = Bn; 56% R = Ph; 77%
EtO*
348: R = H 349: R = Bn 350: R = Ph
[Ref. 93]
An alternative method for generation of the corresponding pyridone species was performed with 347, prepared from hippuric acid (eq. TO).^^ j ^ e reaction of enamino ester 221 with the novel acrylate derivative 347 gave a-amido pyridone 348 in a single step. Substituted derivatives 222 and 266 reacted in a similar manner to give 349 and 350, respectively.^^ (Eq.71) II
•
H0\V 344
NaH THF '"'^
EtO' 11
'
—^ NaoYr 351
THF
CI
1
EtO'^O
O
H
Vr 352
[Ref. 94]
Aza-annulation has been efficiently performed with the mixed anhydride-type reagent represented by 352, which was generated by deprotonation of acid 344 with NaH followed by treatment with Et02CCl (eq. 71).94 Because attempts to isolate the active annulation species led to reagent decomposition, the reagent mixture was generated in situ, and the structure illustrated for
361 352 was proposed. In order to simplify the presentation of this chemistry, the active species will be referred to as 352.
O
1) BnNHa BF3*OEt2 Benzene Reflux
O
'Y^
2) 352, THF
DDQJoluene Reflux (73%) or Mn02, Xylenes,
Reflux'(90%)
O
H
^"^N^^^'^Y^
OEt 91%
208
0*^061
2) EtOgCCI
O
OEt
H.^,H
68%
H
O
78%
354
30% H2O2 KOH
KOH H2O
1)NaH 3)
O
353
H
O
'I
O^OH 357
355
356
H
OEt Scheme 29.
[Ref. 94]
When the enamine generated from 208 was treated with 352, formation of the corresponding dihydropyridone 353, which was similar to an Ala-Ala dipeptide, occurred in high yield (Scheme 29).^^ Oxidation of 353 with either DDQ or Mn02 gave the corresponding pyridone product 354. The dipeptide analog 354 could be selectively deprotected to generate 355, which was then converted to the tripeptide species 356 through standard peptide coupling techniques. Alternatively, hydrolysis of the protected carboxyl and amino termini could be affected in one step to transform 354 to 357. 1) BnNH2 BF3»OEt2 Benzene, Reflux
BnO^ 246
2) 352, THF
OEt
BnO.
(Eq. 72)
83P/o
[Ref. 94]
Conjugate addition of BnNH2 to an alkyne was also an effective method for generation of the enamine used in this class of aza-annulation reactions. Formation of the Ser-Ala dipeptide analog 358 was accomplished by conjugate addition of BnNHi to 246, followed by annulation with 352 (eq. 72).94 Interestingly, when the same methodology was used to access the Phe-Ala dipeptide
362 analog, the ester substituent controlled regioselective aza-annulation from 359, but kinetic deprotonation resulted in conjugation of the enamine with the phenyl substituent to give 360 (eq. 73).^"^ Enamine formation from 205 followed by aza-annulation generated the Asp-Ala analog 361, which was then oxidized to the corresponding pyridone 362 with DDQ (eq. 74).9'^ In the case of 358 and 360, treatment of the dihydropyridones with DDQ did not result in effective formation of the desired pyridone.
Ph'
1)BnNH2 BF3«OEt2 Benzene.Reflux 2) 352, THF
O OEt
359
(Eq.73)
61%
[Ref. 94]
1) BnNH2 BF3*OEt2 Benzene.Reflux 2) 352, THF
OMe
H
DDQ Toluene Reflux MeO,
P 206
O
71%
OMe
71% 361
^O'^OMe
C^
OMe
[Ref. 94]
/-N'^0
352, THF 77%
c6rY O^OEt
296
^ DDQ 78% I Toluene Reflux
KOH, H2O 25 °C
O^OH [Ref. 94]
82% Scheme 30.
O
H
•v-VV O^OEt
364
363 The dihydropyridone 363 with the features of a Pro-Ala dipeptide was prepared by azaannulation of 296 with 352 (Scheme 30).^^ Conversion to the aromatic ring system 364 was accompHshed by oxidation with DDQ, and hydrolysis of the substituent functionality gave the amino acid 365. 4.3
Vinylogous Urea Substrates Vinylogous urea substrates have also been used in the aza-annulation reaction to form 6-
lactam products. This process was illustrated by the condensation of P-keto amide 366 followed by aza-annulation with acryloyl chloride to give 367 (eq. 75).^'^'^^ Catalytic reduction of the tetrasubstituted double bond led to stereoselective formation of 368. The products formed in this reaction were p-enamino peptide units, and this chemistry can be extended to the preparation of triand tetrapeptide analogs.
1) BnNHs, TsOH Benzene, Reflux 2) Acryloyl Chloride THF, Reflux
^ ^ . - • ^
H2, Pd/C
(Eq. 75)
^^
53%
56%
366 367
H.
Ph
[Ref. 57 and 59]
1) BnNH2, BF3»OEt2 Benzene, Reflux 2) 352, THF
O
H
n
K>
N^'^Y Y ^
O
H
O II
H '
Reflux 76%
7^ Y? 7 EtO^ J^.,J<^ ^N^ ^ 1) {R)-Pheny\ glycine ethyl ester BF3*OEt2 Benzene,Reflux 2) 352, THF [Ref. 94]
DDQ Toluene luiufcfiit;
Ph I
Toluene Reflux 55% 373
(51:49)
Scheme 31.
Bn
364 Peptide analogs have been prepared by enamine formation from 369, followed by azaannulation with 352 to give the corresponding a-amido dihydropyridones (Scheme 31).^^ Condensation of BnNH2 with 369, and treatment with 352 gave 370, the amide derivative of 353. As was observed for the ester substrate 353, oxidation of 370 to the corresponding pyridone was accomplished with DDQ. However, oxidation of the amide derivative was much slower than that observed for 353. When the ethyl ester of phenyl glycine was used to generate the enamine, azaannulation produced 372 in excellent yield as an equal mixture of diastereomers. With this methodology, complex molecules were rapidly constructed in a one-pot procedure from readily available starting materials. The complexity of 372 was translated into the slightly lower yield obtained for oxidation to 373. As discussed in section 4.2.4, this class of products represents conformationally restricted peptide analogs, and 371 and 373 resemble di- and tripeptide units, respectively.
O 1) BnNH2, BF3-OEt2 Benzene,Reflux 2) 352, THF
H
^ ^ K K H
O-^ N 95%
375
OEt 374
H
O
DDQ
H
Ar^r-
78%
V
OEt
OEt
O 86%
Ph EtO.
1) (fO-Phenyl glycine ethyl ester BF3«OEt2 Benzene,Reflux 2) 352, THF
O
H
DDQ Toluene Reflux
m
u
n
IN
OEt [Ref. 94]
Extended peptide analogs have been prepared through aza-annulation of 352 with 374 (Scheme 32).94 Condensation of 374 with BnNH2 followed by treatment with 352 generated 375, which was then oxidized to the tripeptide analog 376. The reaction of 374 with the ethyl ester of phenyl glycine, and then aza-annulation with 352 provided 377 as an equal mixture of diastereomers in excellent yield. Oxidation of 377 to pyridone 378 was accomplished by treatment with DDQ. The use of aza-annulation reagent 347 was also applied to the reaction with P-enamino amide substrates (eq. 76).93 Treatment of 379 with 347 led to formation of the TV-substituted and A^unsubstituted amino acid derivatives 380, and the unsubstituted substrate produced superior yields.
365
P
"•N'" O NHp 379 a: R = H b: R = Bn
jj EtO
O
H
^^^
Dioxane , 85°C (Eq.76)
347
380
._MAr^ HgN^^O
R = H; 70% R = Bn; 45%
[Ref. 93]
An interesting method for the formation of a pyridone, which involved the same processes found for typical aza-annulation, has been reported. In this procedure, a unique acrylamide derivative 382 underwent aza-annulation with the P-keto amide substrate 381 in the presence of ZnCl2 to give 383 (eq. 77).95 O O
O
N ^ \ ^
(Eq.77) ZnClg, EtOH Reflux, 1 h 90%
381
382
[Ref. 95]
The use of the more complex acrylate derivative 385 led to an alternate pathway for product formation (eq. 78). Instead of reaction through what would be envisioned as conjugate addition of 384 to the p-position of 385, the regiochemical combination of 384 and 385 resulted in Cacylation of the acrylate derivative and conjugate addition of the amide nitrogen.^^ Subsequent hydrolytic removal of the imine gave ketone 386. Similarly, reaction of 387 with 385 led to the formation of 388 (eq. 79).96
6 O
^%'^ O
AJ^M-Bn
1) 384
385
o ^
N
Toluene, Reflux 2) HCI/MeOH
^"^ J
^
»
65% N
O
A;^M-^" [Ref. 96]
I
Bn
386
(Eq.78)
366
385
Toluene, Reflux
BnO.
2) HCi/HgOATHF OBn 387
H
50%
O
388
O^^
[Ret. 96]
5 . P-ENAMINO NITRILE SUBSTRATES P-Enamino nitriles also serve as suitable substrates for aza-annulation reactions with various aery late derivatives, and both acyclic and cyclic enamino nitriles have been used in this process for the construction of alkaloid skeletons. As illustrated in equation 80, substituted pyridone 390 was prepared in good yield through aza-annulation of enamino nitrile 389 with acryloyl chloride (64).57,59
64 ^^N'^
A
389
"
THF, Reflux
(Eq. 80)
82%
CN
390
CN
[Ref. 57 and 59]
u
Dioxane Reflux, 3.5 h
•^r
29%
CN
A
^^N^^
389
4 Steps
392
N
393
N
395
CN H.
5 Steps 19%
DMF Reflux, 72 h [Ref. 97]
CN 394 Scheme 33.
367 P-Enamino nitrile 389 also underwent aza-annulation with methyl methacrylate (391) and 57 to give dihydropyridone 392 and pyridone 394, respectively. These intermediates were carried on to the 3-methyl (393) and 7-methyl (395) regioisomers of medorinone, and these species were used in structure activity relationship studies of this cardiotonic agent (Scheme 33).^'^ Another investigation, aimed at the preparation of 4,6-dihalogenated-3-pyridinecarbonitriles for potential anti-fungal activity, showed that enamino nitrile 389 underwent efficient aza-annulation with alkyl or aryl substituted ethyl malonates 396 (eq. 81).^^ The resultant phenyl substituted pyridone 397a was transformed into the pyridine species 398 by treatment with POCI3. Compound 398 was then used to generate several derivatives (399) for use as candidates for biological screening. 396
0
(Eq.81)
U
EtO^^
POCI3
0
EtO^"-^0
CI
X
(R = Ph)
NaX
^^
81%
CN
CN
CN
CN
396
397
389
399
a: R = Ph; 68% b: R = nBu; 36% c: R = Bn; 33%
[Ref. 98]
- v'V'' a: X = OMe(85%) b: X = SEt(84%)
Cyclic (3-enamino nitriles have been successfully employed in heterocycle synthesis through aza-annulation with both methyl acrylate and acrylonitrile. Monocyclic P-enamino nitrile 400 underwent aza-annulation with methyl acrylate in the presence of NaH to give 401, which was reduced with diborane to give 402 and 403 in a ratio of 5:1 (Scheme 34).^^ Quinolizidine 402 was carried on directly to (±)-epilamprolobine (404), or alternatively, epimerized with NaH to 403 and transformed into (±)-lamprolobine (405).
Cp
U
H
"^^ 400
CN
M e O ^ 37 NaH Et20, 25°C ^ - j ^ 77%
^ 401
402 B2H6
CN
-"^
+
CD
403
[Ref. 99]
H • CN 38%
8%
Scheme 34.
"cN
3 Steps 46%
NaH Relflux"^ 70% 3 Steps 42%
(±)-Epilampro!obine
CO. 405 (±)-Lamprolobine
368 Isoquinoline derivative 406 was smoothly converted to the emetine-related species 408 through the aza-annulation reaction with acrylonitrile and sodium ethoxide followed by hydrolysis of the intermediate imine 407 (eq. 82).!^
1
NC, 2
NaOEt HOEt 71%
MeO'
(Eq.82)
AcOH H2O Reflux J
96%
MeO*
MeO'
MeO [Ret. 100]
6 . OTHER ENAMINE SUBSTRATES Phosphonate groups also provided regiochemical control in the aza-annulation reaction. Condensation of 409 with BnNH2, followed by reaction with acryloyl chloride gave 410 in good yield (eq. 83).^^'^^ Catalytic hydrogenation of 410 was less efficient than the corresponding ester and amide derivatives and resulted in the formation of a mixture of diastereomeric products 411.
O
O
x'-'^^^f^OEt OEt
1)BnNH2, TsOH Benzene, Reflux 2) Acryloyl Chloride THF, Reflux 72%
o „ ][ *^""N " ^ EtO.p 410
409
(Eq.83)
3 atm of H2 Pd/C, Na2C03 EtOH
^
67%
EtO' ^O 411
Cis.Trans (78:22)
[Ret. 57 and 59]
Aza-annulation with sulfone derivatives appeared to be somewhat dependent on the sulfone substituent. Stepwise enamine formation and aza-annulation from 412 led to the formation of 414 in better yield than conversion of the analogous methyl derivative 413 to 415 (eq. 84).^'7'5^
O
O
1) BnNH2, TsOH Benzene, Reflux 2) Acryloyl Chloride THF, Reflux
(Eq.84)
O 412: R = Ph 413: R = Me [Ref. 57 and 59]
R = Ph; 69% R = Me; 45%
414: R = Ph 415: R = Me
369 7 . PYRIDINE SUBSTRATES Several examples of aza-annulation have been reported with substrates in which the nitrogen involved in formation of the 6-lactam product was a constituent of a pyridine ring. In general, (1) the nucleophilic carbon of the substrate was activated by carbonyl functionality, (2) these substrates require reaction with doubly activated Michael acceptors, and (3) generation of the unsaturated pyridone ring was an important driving force for aza-annulation. Pyridine derivatives will undergo aza-annulation at elevated temperatures. For example, reaction of ketone 416 with the diester 47 at 150-190 °C led to formation of quinolizone 417 (eq. 85). ^^1 The ketone substituent R played an important role on the outcome of the reaction, in which substrates with an alkyl substituent gave much greater yields than those with a phenyl substituent. j
O
Li
150-190 °C 3h
O
* ^"Y
Neat
OEt
OEt
^^
a: R = Me; 43% b: R = Et; 48% c:R = Ph; 18%
R^^O
417
416 [Ref. 101]
Substrate 418, which was less hindered without the methyl substituent in the 6-position of the ring, gave slightly higher yield than the analogous reaction of 416 (Scheme 34). ^^^ Deethoxycarboxylation of this class of products was demonstrated by the conversion of 419 to 420. A mixed ester and ketone-activated reagent (421) was also effective for aza-annulation in the construction of 422 at ambient temperature. O
O
O
120-130 °C
O OEt
EtO
Cone. HCI Reflux 30min 73%
419
418
39%
X^o O
[Ref. 102]
^
X^o
0
O
O
EtO^>^^
25 °C
421 ^OEt
HOEt
422
O
Scheme 34.
Aza-annulation was more efficient with pyridine substrates activated by the presence of ester functionality. As an example, the reaction of reagent 421 with ester stabilized pyridine substrate
370 423 resulted in a significantly higher yield than with the analogous ketone substrate 418 (eq. 86).102 O u
O u
Etc 421
OEt
25 °C NaOEt HOEt
O
O (Eq.86)
86%
[Ref. 102] O
O (Eq.87)
47 "^oPt 180°C
EtO^^O
R = H; 52%, 67% R = Me; 80%
4 Steps
EtO^^O 425: R = H 426: R = Me
423: R = H 424: R = Me
12% (From 425)
OH 232 (±)-Lupinine
[Ref. 103 and 104]
Treatment of 423 with 421, again at an elevated temperature (180 °C), led to the formation of 424 (eq. S1)A^^^^^^ In the case of the ester-stabilized pyridine substrates, substitution with an electron donating substituent such as a methyl group significantly enhanced the yield, as observed for the conversion of 424 to 426. Compound 425 was converted to (±)-lupinine by reduction and deprotection. 1) DMF, NaH 1 h, 60 °C 2) AcOH, Reflux 72h
(Eq.88)
MeO'
427: Y = OH 428: Y = N02
429: Y = OH 430: Y = N02
Doubly activated acrylonitrile reagents have also been utilized in the aza-annulation of pyridine substrates. Aza-annulation of 423 with 427 or 428 was performed at a mild temperature
371 (60 °C) by generation of the corresponding enolate with NaH (eq. 88).^^^ Through variation of the aromatic substituent Y, derivatives 429, 430, and others were prepared. Alternatively, 423 could be treated with 431 at 120 °C to give 432, which was converted to a number of heterocycle substituted derivatives related to 433 (eq. 89). 1^5 j ^ all of these cases, derivatives of 429, 430, and 433 exhibited selective inhibitory activity against IgE-antibody formation. As a result, these compounds have potential for treatment of diseases such as allergic rhinitis, atopic dermatitis, allergic bronchial asthma, and hypersensitiveness.^^^
||
SMe
MeCN Reflux 10 h
433
A nitro-substituted acrylate derivative 434 has also been used as an aza-annulation reagent. Treatment of 423 with 434 resulted in the formation of 435, which was reduced to the a-NH2 derivative 436 (eq. 90).^^^ When the pyridine substrate was substituted in the 6-position, reaction proceeded through an alternate pathway, and aza-annulation did not occur. ^^^ O NO2
EtO'
(Eq. 90) NH2
Zn, HCi
OEt
434
42%
45%
A more efficient approach to a-NHR substituted carbonyl derivatives was through the use of 347 (eq. 91). Aza-annulation of 423 with 347 produced 437 in good yield.107
(Eq.91) AcOH 79% EtO ^O 423 [Ret. 107]
EtO 347
437
372 Similar reactivity was observed for the reaction of 347 with nitrile activated substrates. Treatment of 438 with 347 generated the a-NHR substituted carbonyl derivative 439 (eq. 92). ^^'^
a 438
(Eq.92) AcOH 68%
0 ^ '
N
EtO
347
[Ref. 107]
(Eq.93) AcOH, 4 h
»
79% (From 438) 438
440
N
[Ref. 108]
An interesting reagent for aza-annulation was the acrylate-type reagent 440. When 438 was treated with 440, formation of 441 resulted (eq. 93).^^^ Subsequent reaction in AcOH resulted in cyclization to complete the aza-annulation process. An example of aza-annulation that involved the use of this reagent with a substrate activated by NO2 substitution, 443, was also reported (eq. 94) 108 Formation of 444 resulted, and cyclization gave 445 in low yield. (Eq. 94)
CI AcOH, 4 h
CI O 443
NO2
»
22% (From 443)
440
NO2
[Ref. 108]
EtO" 47 446: Y = OEt 447: Y = NH2 [Ref. 109]
0 M '^OEt OEt
NaOEt, HOEt 25°C, <15min Y = OEt; 72% Y = NH2; 81%
OH 0
C
r ^ i T
kAJ
Y'^O 448: Y = OEt 449: Y = NH2
(Eq. 95) OEt
373 Pyridone derivatives themselves can be used for aza-annulation. The pyridone tautomers 446 and 447 underwent facile aza-annulation with 47 at ambient temperature (eq. 95).^^^ In the presence of NaOH, in which the anion was stabilized through conjugation with ester (446) or amide (447) functionality, reaction occurred immediately at ambient temperature to give 448 and 449. In addition, extended pyridone derivative 450 reacted with 47 to give 451 for the purpose of pharmaceutical drug investigation (Scheme 35).^^^ Treatment of 450 with 47 resulted in initial reaction at the amino substituent to give 451, and addition of more 47 provided 452 in 44% yield. Combination of the steps, through reaction with an excess of 47 in Dowtherm at reflux, gave a 40% yield of 452 from 450. Modification of 452, by deprotection of the amino group, gave 453. Conversion to 454 was accomplished by treatment with NaN02/H2S04, and hydrolysis of 454 gave 455. In pharmacological studies, compound 453 displayed anthypoxic activity when administered to mice, while 454 and 455 exhibited a passive cutaneous anaphylaxis inhibition in rats. 110
OEt Benzene Reflux 451
450 68%
Ion Exchange (Dowtherm) Reflux 44%
452 NH3 EtOH
NaOH H2O OH
Eton
OEt
[Ret. 110]
NaN02 H2SO4 -*-
OEt
86%
Quant. 455
Quant.
454
453
Scheme 35.
8 . ASYMMETRIC INDUCTION IN THE AZA-ANNULATION REACTION Due to the importance of enantioselective synthesis in the construction of optically pure target molecules, methodology for asymmetric synthesis has received increased attention. An important
374 contribution to these studies has been the control of absolute stereochemistry through asymmetric aza-annulation reactions. Substrates in this section contain a stereogenic center, which was utihzed to selectively generate additional asymmetry. Several important features of the aza-annulation reaction are illustrated in this section. In some examples, asymmetric carbon-carbon bond formation resulted from the aza-annulation process, while other aza-annulation reactions led to a 5-lactam skeleton with an asymmetric substituent attached to the ring, as seen for the synthesis of (-)-pumiliotoxin (eq. 2).^ In this case, asymmetry was not generated through aza-annulation, but rather, peripheral asymmetry directed diastereoselective reduction of the aza-annulation product. Development in the area of asymmetric aza-annulation reactions paralleled achievements in the analogous area of asymmetric Michael addition reactions with chiral imines.^^l Induction of asymmetry has been primarily controlled through substitution at the nitrogen of the imine or enamine that becomes incorporated into the heterocycle. Restricted rotation of this asymmetric substituent led to preferred conformational isomers, which provided stereofacial bias for carbon-carbon bond formation. 8.1
Alkyl and Aryl Substituted Enamine Substrates Asymmetric induction in the aza-annulation reaction through the use of a chiral auxiliary was
pioneered by d'Angelo and coworkers in their work with 456. Even though annulation was not accomphshed with the use of methyl crotonate, crotonyl cyanide produced an equimolar mixture of 458 and 459 in moderate yield (Scheme 36).^^^ Both 458 and 459 reflect regioselective Michael addition at the most substituted a-carbon. The vicinal methyl substituents that resulted from azaannulation were oriented in a cis relationship.
Ph. /
456 20 °C 24 h t Ph
458 OMe
(4050% Combined Yield)
SPh 460 (95:5)
/
H'^N'^
OMe
OMe
ACOH/H2O SPh 461 [Ret. 112]
Q NH4OH
'^^N'^^y^
80% (From 456) 462 Scheme 36.
463
375 The acrylic acid derivative 460 was successfully used for asymmetric carbon-carbon bond formation to form 461, and in situ hydrolysis provided 462 with good selectivity. 1^2 j^jg g^^p. wise aza-annulation process was completed by treatment with NH4OH, which led to the formation of 463. In the examples shown in Scheme 36, both reactions resulted in deracemization of the initial racemic ketone substrate through formation of the carbon-carbon bond.
,x?
N I {S)-465 ^** Ph Toluene 110 °C
CI CI
NaHC03 (aq.) CHCI3 25 X •
79%
100%
464 30 equlv. NaBHaCN HCO2H, 25 °C 1) CF3CO2H Reflux 2) NaH, Mel DMF, Reflux f
72% 20% [Ref. 113]
13%
Scheme 37.
Chiral imine substrates, in which the enamine tautomer was stabilized through conjugation with an aryl group, have been studied (Scheme 37).^ ^3 xhe impetus for this work stems from the structural similarity of the resultant heterocyclic analogs to steroidal enzyme substrates, and 140 and 143 have demonstrated selective inhibition of human Type I steroid 5-a-reductase (eq. 31 and 32). Condensation of 464 with (5)-465 gave 466, which was then treated with acryloyl chloride under Schotten-Bauman conditions to give optically pure 467 in high yield (Scheme 37). Although azaannulation did not result in the formation of a stereogenic center, subsequent reduction of 467 led to the stereoselective formation of the two trans fused ring systems 468 and 469 (trans/cis, 5:1). Chromatographic separation of the the two trans isomers gave 468 and 469 in 20% and 13% yield, respectively. Removal of the chiral auxiliary from 468 followed by methylation gave 143.
376
N I ^**
(S)-465 Ph
XI
Toluene 110 °C
[Ref. 113]
JT j r
X,
O-^^CI
64
THF, >20 °C^
^'vlJkJ^
Scheme 38.
The angular methyl derivative of 143 was prepared stereoselectively through a related approach (Scheme 38).^^^ In this reaction sequence, condensation and aza-annulation of the methylated substrate 138 led to diastereoseiective formation of 471 through asymmetric carboncarbon bond formation (25:1). Selective reduction of 471 generated 472 along with the cis isomer (6:1), and 472 was methylated to give 140.^ ^^ 8.2
Vinylogous Amide Substrates Enders and coworkers reported studies in which the RAMP and SAMP chiral auxiliaries were
employed in the aza-annulation process (Scheme 39).^ ^"^ Condensation of 179 with RAMP provided a route to the optically active enamino hydrazone 473, which was then metalated with nBuLi to generate the corresponding anion. Aza-annulation of 473 with 474 produced intermediate 475, which could be cyclized slowly (2 d) at 60 °C to give 476.
Alternatively, heterocycle
formation could be facilitated by an increase in reaction temperature (toluene, heat). Removal of the chiral auxiliary gave 477 in 50-52% overall yield from 179 in >99:1 enantiomeric purity. Substituents on the aromatic ring did not have a measureable effect on the yield of the aza-annulation reaction.
377
ri
•A„
1)n-BuLi,TMEDA THF, -78 °C 2) Ar OMe -78 °C
^^^''^OMe
OMe r^-^o MeO ^ O 474 3) NH4CI, -30 °C
RAMP
*N'
Ar
OMe
^ - j s ^ 473
179
475 or Toluene Reflux
60 °C 2cl R
Ar
179-> 476
a
H
Ph
50%
b
Me
Ph
50%
c
Me
4-MeO-C6H4
52%
COgMe
R
OMe Zn, AcOH Reflux 6-12 h
C
V^^' O
O OMe
>99.i Enantioselectivity Scheme 39.
[Ret. 114]
8.3
476
Vinylogous Carbamate Substrates Stereoselective generation of quaternary carbon centers has been studied for a variety of
tetrasubstituted p-enamino ester substrates. Asymmetric enamines have been generated by condensation of either optically active 465 or an amino acid derivative with the corresponding p-keto ester. Formation of 479 from 478, followed by treatment with acryloyl chloride (64) gave 480 in good yield with highly diastereoselective formation of the 5-lactam (eq. 96).^ ^^ The high degree of stereoselectivity observed in product formation resulted from (1) the geometry of the enamine which was fixed by the constraints of the ring, and (2) the relative steric demands of the Me and Ph of the chiral auxiliary.
0
0
A 1 478
[Ref. 115]
H NH2 (R)-465 EtaO'BFs Benzene Reflux
Ph ^;?
H^N'^ 0
64 THF, Reflux
479 a b
0 1 2
O ^ " " ' ' 480 478-^480 76% 85%
Ratio 97:3 >97:3
378 Acyclic enamines were formed selectively as essentially a single geometric isomer through the stabilization that results from intramolecular hydrogen bonding. Aza-annulation of acyclic substrates such as 481a and 481b also resulted in the formation of 5-lactam products with high diastereomer ratios (eq. 97).^ ^^ Similar reactivity and stereoselectivity was observed with the analogous lactone derivative 484 in the formation of 486 (eq. 98). 11^ u
Ph ^^ H^NHs {R)-465 ^-^\^OEt 481
(Eq.97) H^N'^ O
Et20»BF3 Benzene Reflux
64 OEt
Y 482
Afii -:.4a3 92% 58%
a Me b OBz
[Ret. 115]
OEt
THF, Reflux
Ratio
97:3 92:8
u
Ph ,^
V
H'^NH2
Et20»BF3 Benzene Reflux
(Eq.98)
CI
(R)-465
64 L ^ 485
THF, Reflux 80% (From 484)
[Ref. 115]
1)
O
R^R2
487 V H NH2 Et20»BF3 Benzene, Reflux
O
(Eq. 99)
OEt 2)
O
478b THF, Reflux [Ref. 115]
a a b
Ri R2 C02Et Ph C02Et Ph iPr C02Me
Ratio (488:489) 79:21 98:2 43:57
Temp 66 °C -33 °C 66 °C
Yield 63% 77% 43P/o
Amino acid derivatives have also been explored as potential chiral auxiliaries in the asymmetric aza-annulation reaction. As reported for the Michael addition to acrylate derivatives, the reaction outcome has also shown sensitivity to the special balance of complementary steric demands of the methyl and phenyl substituents. The degree of diastereoselectivity in carbon-carbon bond
379 formation was dependent on the presence of a phenyl substituent. When the methyl group of phenethylamino auxiliary (465) was replaced with an ester group, as in the phenyl glycine derivative 487, stereoselectivity in the formation of 488a and 489a dropped considerably (79:21), and decreased reaction temperatures were required to achieve selective product formation (eq. 99).^^^ Further alterations in the source of asymmetry, through the use of the valine derivative b, led to minimal induction of asymmetry in the generation of 488b and 489b. Substituted acrylate derivatives have also been employed in the asymmetric aza-annulation reaction (Scheme 40). Aza-annulation of 479b with crotonyl chloride (490) demonstrated several important features of this reaction.l ^^ First, concomitant formation of two stereogenic centers gave 491 with high internal asymmetric induction, while high relative asymmetric induction resulted from the amine substituent. However, the presence of a methyl substituent at the p-position of the acrylate derivative slowed the reaction significantly, and resulted in a poor yield.
O
O
(^V^OEt k ^ 478b
H NH2 {R)-A65 Et20*BF3 Benzene Reflux
V
H^N'^0 479b 352, THF, -33 °C
492 THF, -33 °C 2)NaH
(73:27)
1) (52:48) 2) (83:17)
Substitution at the a-position of the acrylate derivative did not appear to significantly change the aza-annulation process, and formation of the quaternary center was highly stereoselective (Scheme 40).^^^ In contrast to the observations for the formation of 491, stereochemical control at the 3-position of the 5-lactam was only moderate. Annulation of 479b with methacryloyl chloride
380 (492) resulted in the formation of 493 with >98:2 stereoselective generation of the quaternary center, while a 52:48 kinetic ratio of isomers at C-3 resulted. Equilibration of C-3 by treatment with NaH provided an increased diastereomer ratio of 83:17. Slightly higher diastereoselectivity was obtained in the formation of 494. Modification of the products that resulted from the aza-annulation of tetrasubstituted enamine substrates with acrylate derivatives was very limited. The aza-annulation of benzyl ester 496 with the mixed anhydride, a mixture (497) preformed from Et02CCl and sodium acrylate, provided a route to 498 in >98:2 diastereoselectivity (eq. 100), which allowed access to the carboxylic acid derivative 499 through catalytic hydrogenation. ^ ^^ Further elaboration of either the ester or the acid derivative was unsuccessful, possibly due to the steric congestion around the reactive functionality. Extended hydrogenation did not reduce the enamine functionality, as observed in related substrates, and 498 was relatively stable to acidic hydrolysis conditions.
In addition, DCC (A^,A^'-
dicyclohexylcarbodiimide) coupling of acid 499 with either benzyl amine or glycine ethyl ester was unsuccessful. (Eq. 100)
^) % / (fl).465 H NH2 O
3 atm of H2 Pd/C EtOH
Et20»BF3 Benzene, Reflux
O
OBn
^
98%
EtOaCCI +
OH
497
496
THF, Reflux 70%
[Ref. 116]
(>98:2)
8.4
Vinylogous Urea Substrates Studies of aza-annulation reaction with tetrasubstituted P-keto amide substrates have also been performed. Investigations centered around those substrates that were analogous to the ester species described in Section 8.3. In general, the amide substrates were found to react 20-25% slower than their ester counterparts, and as a result, greater diastereoselectivity was observed.^^^ 1)
6" O
O
R^ R2
V
0
R\R2
H ^ N ^
H-^N
R\R2
H NH2 Et20«BF3 Benzene, Reflux
NHBn
NHBn 2)497 THF, Reflux
^ ^
500
[Ref. 116]
a b c d
0
R2 R1 Ph Me C02Et Ph iPr C02Me Bn C02Et
"*"
501 Ratio (501:502) >98:2 2:>98 >98:2 >95:5
0
(Eq.
^
NHBn ^
0
Yield 99% 96% 90% 46%
502
381 Condensation and aza-annulation of 500 provided 501 as a single observable isomer in excellent yield (eq. 101).^i^ The (/?)- or (5)-stereoisomer of the quaternary carbon could be obtained depending on the chiral auxiliary used, and for each auxiliary, high selectivity was obtained. In the case of the phenylalanine derivative d, a lower yield was obtained, possibly due to the decreased steric protection of the enamide functionality which then allowed electrophilic reactions such as hydrolysis to occur. The analogous five-membered ring substrate 503 also gave highly stereoselective reaction upon condensation with (/?)-465 and subsequent aza-annulation to give 505 (>98:2), but the yield for this reaction was low due to the sensitivity of 505 to hydrolysis (eq. 102).
^
(f?)-465
o NHBn
503
H-^^N'^ O
Et20»BF3 Benzene Reflux
\
I
(Eq.102)
487 THF, Reflux NHBn
50% (From 503)
505
^ - ^
504
(>98:2)
[Ret. 116]
Similarly, aza-annulation with an acyclic substrate resulted in a high degree of stereocontrol. These results suggested that intramolecular hydrogen bonding of the intermediate enamine controlled the enamine geometry and served to restrict rotation of the chiral auxihary (eq. 103).^ ^^ In this case, 507 was sensitive to hydrolysis, and isolation was performed after hydrolysis to 508.
1) H'^^NHs Ph ^^
(FJ)-465
u
Et20*BF3 Benzene, Reflux
u NHBn
2)497 THF, Reflux
506
H2O pTsOH
O
(Eq.103)
H-^N' NHBn
NHBn 82% (From 506)
508
(>98:2)
[Ref. 116]
Reaction of 500 with (/?)-465 followed by aza-annulation with a substituted acrylate derivative 352, gave 509 (eq. 104).^ ^^ Although the quaternary carbon was formed almost exclusively as the R isomer, an equimolar mixture of the substituent at C-3 of the 5-lactam resulted.
382
1) H ^ N H s {R)-A65 EtaO'BFs Benzene, Reflux
6" O
O
(50:50) (Eq. 104)
.Ph. . c Q / H
2) 352 THF. Reflux
tfr
NHBn
67% (From 500)
500 [Ret. 116]
O NHBn 509
8.5 p-Imino Sulfoxide Substrates The use of P-imino sulfoxide substrates has followed a different strategy than other asymmetric aza-annulation reactions. Instead of generating a stereogenic center during the azaannulation process, a chiral sulfoxide was used to modify the 5-lactam product in an asymmetric fashion after formation of the heterocyclic ring. The use of p-imino sulfoxide substrates led to a number of appUcations in natural product synthesis. For example, 510 was deprotonated to generate the corresponding a-sulfinyl ketimine anion, and addition of methyl acrylate resulted in the formation of 511 (eq. 105).^^"^'^^^ The next step involved stereoselective reduction of the enamide functionality with NaCNBHa to give 512 as the only diastereomer. Final reductive removal of the sulfoxide functionality and reduction of the carbonyl gave (/?)-(-)-indolizidine (513).
1)LDA 2) -30 °C-25 °C 2h, O MeO " \ rr II 37
510
NaCNBHa AcOH CF3CO2H
5 y ^ N - ^
1)25°C.2h 2)50°C.4h^
'if / ^ ^ ^
i7^
1) Raney-Ni EtOH 2) LiAIH4 83%
(Eq. 105)
CO H
513
[Ref. 117 and 118]
Further extension of this methodology demonstrated that cyclic acrylate derivatives could be used to construct tricyclic ring systems with the formation of stereogenic centers during azaannulation (Scheme 41).ll7,ll8 When treated with 514, the cis to trans ring fusion obtained for product formation was 70:2 for 515:516. Compound 515 was reduced to 517. Although azaannulation with 518 gave slightly lower selectivity in the formation of 519 and 520, formation of the cis ring fusion was still favored, and a good selectivity was obtained.
383
1) A?-BuLiorLDA 2) -78 °C- 25 °C
1) /7-BuLiorLDA 2) -78 °C- 25 °C
jy^-° xr^f'o 515
70%
71%
H
516
jy'-° jy^-f--o
2>/o
519
60%
520
1) NaCNBHa AcOH 2) Raney-Ni EtOH, Reflux
H
517
[Ret. 117 and 118]
Scheme 41.
The use of 521 led to the synthesis of a number of natural products through the azaannulation of this P-imino sulfoxide (Scheme 42).^ ^'^'^ ^^ Application of methyl aery late in this azaannulation process led to the formation of 522. The chiral auxiliary was then used to provide moderate stereochemical control in the reduction of the enamide alkene to 523 and 524 in a 1.9:1.0 ratio. Compound 523 was then reduced to remove both the sulfoxide and the lactam carbonyl to give (-)-l,2,3,4,6,7,12,12b-octahydroindolo[2,3-a]quinolizine (525).^^'^'^^8
384 1)LDA,THF -78 °C, 1 h 2) - 25 °C, 4 h 37
J JT/c
90%
NaCNBHs AcOH 25 °C, 3 h
1) Separation 2) Raney-Ni THF/EtOH 65 °C 3) LiAIH4, 25 °C 525 H-1,2,3,4,6,7,12,12b-0ctahydroindolo[2,3-a]quinolizine
75%
^'^ •• '-^^ r
\ 523
[Ret. 117 and 118]
ii 524
Scheme 42.
The reaction of 521 with 518 has been used to generate 526, a very versatile pentacychc intermediate in the synthesis of natural products (Scheme 43). ^ 17,118 Removal of the sulfoxide gave 528, which could be further reduced to give (-)-alloyohimban 529. Alternatively, 528 could be treated with LDA to cause epimerization of the stereogenic center a to the lactam carbonyl, and subsequent reduction gave (-)-yohimban 530. Initial reduction of 526 in the presence of the chiral auxiliary, was found to give 531 in slight preference to 532, and 531 could be reduced under standard protocol for these molecules to give (+)-3-ep/-alloyohimban (533).11'7'11^
385 1)LDA,THF 2) 25 °C, 1 h 60 °C, 14 h 518
35% (Recovered 521)
NaCNBHa X 80% A c O H / (From 526)
88% \ Raney-Ni (From 526) \ EtOH
Raney-Ni EtOH
^-
89% (From 532)
1)LDA 2) AcOH 3) LiAIH4
533 (+)-3-ep/-Alloyohimban [Ref. 117 and 118]
529 (-)-Alloyohimban
43%
530 (-)-Yohimban
Scheme 43.
Substituted acrylate derivatives have also been employed in the asymmetric synthesis of a natural product. In a model study, the deprotonation of 510 and aza-annulation with 534 led to a 60:40 mixture of 535 and 536 (eq. 106). ^ 1^
386 1)n-BuLi 2) O H MeO^"^
O
Y
O
H ^OtBu
OtBu
H ,N„.OtBu
To
534
34%
(Eq. 106) 535
510
536
(60:40)
[Ref. 119]
This methodology was applied to the substituted analog 537, which also gave a 60:40 ratio of diastereomeric products in 55% yield (Scheme 44). 1^^ In this case, isomers 538 and 539 could be separated and then carried through the same sequence of parallel steps to give (-)-slafraniine (540) and (-)-6-epislaframine (541). 119
1)/>BuLi 2) O H MeC
O
OtBu
H
O
H
M
^OtBu
Y o
537
538
(60:40)
539
SiRs = Si(Me)2tBu 4 Steps
28% (From 538)
4 Steps
13P/o (From 539)
.»NH2 [Ref. 119]
^ AcO 540 (~)-Slaframine Scheme 44.
AcO 541 (-)-6-Epislaframine
OtBu
387
9. (1)
(2) (3) (4) (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) (33) (34) (35) (36) (37) (38)
REFERENCES (a) Ninomiya, L; Naito, T. Heterocycles 1981,75,1433. (b) Ninomiya, L; Miyata, O. In Studies in Natural Products Chemistry, Atta-ur-Rahman, Ed.; Elsevier: New York, 1989; Vol. 3, p. 399. (c) Ninomiya, L; Naito, T. In The Alkaloids; Brossi, A., Ed.; Academic Press: New York, 1983; Vol. 22, p. 189. Vill, J. J.; Steadman, T. R.; Godfrey, J. J. / Org. Chem. 1964, 29, 2780. Kuehne, M. E.; Bommann, W. G.; Parsons, W. H.; Spitzer, T. D.; Blount, J. P.; Zubieta, J. J. Org. Chem. 1988, 55, 3439. Chelucci, G.; Cossu, S.; Scano, G.; Soccolini, F. Heterocycles 1990, 57, 1397. Murahashi, S.-L; Sasao, S.; Saito, E.; Naota, T. J. Org. Chem. 1992, 57, 2521. Tominaga, Y.; Kawabe, M.; Hosomi, A. / Heterocycl Chem. 1987, 24, 1325. Stork, G. Pure and Appl Chem. 1968,77,383. Ninomiya, L; Naito, T.; Higuchi, S.; Mori, T. J. Chem. Soc, Chem. Commun. 1971, 457. El-Barbar>% A. A.; Carlsson, S.; Lawesson, S.-O. Tetrahedron 1982, 38, 405. Hickmott, P. W.; Rae, B.; Pienaar, D. H. S. Afr. J. Chem. 1988, 41, 85. (a) Xia, Y.; Kozikowski, A. P. J. Am. Chem. Soc. 1989, 777, 4116. (b) Kozikowski, A. P.; Xia, Y.; Reddy, E. R.; Tuckmantel, W.; Hanin, L; Tang, X. C. J. Org. Chem. 1991, 56, 4636. Stork, G.; Kretchmer, R. A.; Schlessinger, R. H. J. Am. Chem. Soc. 1968, 90, 1647. Paronikyan, E. G.; Sirakanyan, S. N.; Lindeman, S. V.; Aleksanyan, M. S.; Karapetyan, A. A.; Noravyan, A. S.; Struchkov, Y. T. Chem. Heterocycl. Compd. (USSR) (Engl. Transl.) 1990,25,953 (Khim. Geterotsikl Soedin. 1989, 1137). (a) Sammour, A.; Alkady, M. Ind J. Chem. 1974, 72, 51. (b) El-Kady, M.; El-Hashash, M. A.; Sayed, M. A.; El-Sherif, M. Ind J. Chem., Sect. B 1981, 20, 491. Briet, P.; Berthelon, J.-J.; Depin, J.-C. European Patent 0 000 306, 1979. Chem. Abstr. 1979,97:20479v. Elgemeie, G. E. H.; Elghandour, A. H. H. Bull. Chem. Soc. Jpn. 1990, 63, 1230. Kambe, S.; Saito, K.; Sakurai, A.; Hayashi, T. Synthesis 1977, 841. Takahashi, K.; Miyake, A.; Hata, G. Bull. Chem. Soc. Jpn. 1972, 45, 2212. Norman, M. H.; Heathcock, C. H. J. Org. Chem. 1988, 53, 3370. Gardette, D.; Gramain, J.-C; Sinibaldi, M.-E. Heterocycles 1990, 57, 1439. Borne, R. P.; Fifer, E. K.; Waters, I. W. J. Med Chem. 1984, 27, 1271. Wenkert, E.; Chauncy, B.; Dave, K. G.; Jeffcoat, A. R.; Schell, F. M.; Schenk, H. P. J. Am. Chem. Soc. 1973, 95, 8427. Ito, K.; Yokokura, S.; Miyajima, S. J. Heterocycl. Chem. 1989, 26,111>. Kmetic, M.; Stanovnik, B.; Tisler, M.; Kappe, T. Heterocycles 1993, 35, 1331. (a) Meyers, A. I.; Reine, A. H.; Sircar, J. C ; Rao, K. B.; Singh, S.; Weidmann, H.; Fitzpatrick, M. J. Heterocycl. Chem. 1968,5,151. (b) Horii, Z.-i.; Iwata, C ; Ninomiya, I.; Imamura, N.; Ito, M.; Tamura, Y. Chem. Pharm. Bull. 1964, 72, 1405. Shabana, R.; Rasmussen, J. B.; Olesen, S. O.; Lawesson, S.-O. Tetrahedron 1980, 36, 3047. Kozikowski, A. P.; Reddy, E. R.; Miller, C. P. J. Chem. Soc, Perkin Trans. 1 1990, 195. Paulvannan, K.; Stille, J. R. J. Org. Chem. 1992, 57, 5319. Ninomiya, I.; Naito, T.; Higuchi, S. J. Chem. Soc, Chem. Commun. 1970, 1662. Hickmott, P. W.; Sheppard, G. J. Chem. Soc (C) 1971, 1358. (a) Rigby, J. H.; Balasubramanian, N. /. Org. Chem. 1984, 49, 4569. (b) Rigby, J. H.; Qabar, M. Synth. Commun. 1990, 20, 2699. (a) Dickman, D. A.; Heathcock, C. H. J. Am. Chem. Soc 1 9 8 9 , 7 7 7 , 1528. (b) Heathcock, C. H.; Norman, M. H.; Dickman, D. A. J. Org. Chem. 1990, 55, 798. (a) Aranda, V. G.; Barluenga, J.; Gotor, V. Tetrahedron Lett. 1973,2819. (b) Barluenga, J.; Muniz, L.; Palacios, F.; Gotor, V. J. Heterocycl. Chem. 1983, 20, 65. Aranda, V. G.; Barluenga, J.; Gotor, V. Tetrahedron Lett. 1974, 977. Janin, Y. L.; Bisagni, E.; Carrez, D. J. Heterocycl. Chem. 1993, 30, 1129. Kametani, T.; Terasawa, H.; Ihara, M. J. Chem. Soc, Perkin Trans. I 1976, 2547. Kametani, T.; Terasawa, H.; Ihara, M.; Fukumoto, K. Heterocycles 1977,6,37. Ihara, M.; Noguchi, K.; Ohsawa, T.; Fukumoto, K.; Kametani, T. Heterocycles 1982, 79, 1829.
388 (39) Kametani, T.; Suzuki, Y.; Terasawa, H.; Ihara, M. / Chem. Soc, Perkin Trans. I 1979, 1211. (40) Kametani, T.; Suzuki, Y.; Ihara, M. Can. J. Chem. 1979, 57, 1679. (41) Bhattacharjya, A.; Bhattacharya, P. K.; Pakrashi, S. C. Heterocydes 1983, 20, 2397. (42) (a) Danieli, B.; Lesma, G.; Palmisano, G. /. Chem. Soc, Chem. Commun. 1980, 109. (b) Danieli, B.; Lesma, G.; Palmisano, G. Gazz. Chim. Ital. 1981, 111, 257. (43) Danieli, B.; Lesma, G.; Palmisano, G.; Tollari, S. Synthesis 1984, 353. (44) (a) Kametani, T.; Kanaya, N.; Ihara, M. Heterocydes 1981, 76, 925. (b) Kametani, T.; Kanaya, N.; Hino, H.; Huang, S.-P.; Ihara, M. J. Chem. Soc, Perkin Trans. I 1981, 3168. (45) Kametani, T.; Kanaya, N.; Hino, H.; Huang, S.-P.; Ihara, M. Heterocydes 1980, 14, 1771. (46) Calabi, L.; Danieli, B.; Lesma, G.; Palmisano, G. Tetrahedron Lett. 1982,25,2139. (47) Kametani, T.; Ohsawa, T.; Ihara, M. J. Chem. Soc, Perkin Trans. I 1981, 1563. (48) Corriu, R. J. P.; Perz, R. Tetrahedron Lett. 1985,26,1311. (49) Singh, B. Synthesis 1985, 305. (50) Singh, B. U.S. Patent 4 347 363,1982. Chem. Abstr. 1982, 97:216018n. (51) Tanabe Seiyaku KK Japanese Patent J4-8 023 779, 1984. (52) Horii, Z.>I.; Iwata, C ; Tamura, Y.; Nelson, N. A.; Rasmusson, G. H. J. Org. Chem. 1964, 29, 2768. (53) (a) Cannon, J. G.; Hatheway, G. J.; Long, J. P.; Sharabi, F. M. J. Med. Chem. 1976,19, 987. (b) Cannon, J. G.; Suarez-Gutierrez, C ; Lee, T.; Long, J. P.; Costall, B.; Fortune, D. H.; Naylor, R. J. J. Med. Chem. 1919,22, 341. (c) Cannon, J. G.; Hamer, R. L.; Ilhan, M.; Bhatnagar, R. K.; Long, J. P. J. Med. Chem. 1984, 27, 190. (d) Cannon, J. G.; Chang, Y.; Amoo, V. E.; Walker, K. A. Synthesis 1986, 494. (54) Jones, C. D.; Audia, J. E.; Lawhom, D. E.; McQuaid, L. A.; Neubauer, B. L.; Pike, A. J.; Pennington, P. A.; Stamm, N. B.; Toomey, R. E.; Hirsch, K. S. J. Med. Chem. 1993, 36, 421. (55) Ninomiya, I.; Kiguchi, T. J. Chem. Soc, Chem. Commun. 1976, 624. (56) Hickmott, P. W.; Sheppard, G. / Chem. Soc (C) 1971, 2112. (57) Paulvannan, K.; Stille, J. R. Tetrahedron Lett. 1993,34,8197. (58) Cook, G. R.; Beholz, L. G.; Stille, J. R. J. Org. Chem. 1994, 59, 3575. (59) Paulvannan, K.; Stille, J. R. J. Org. Chem. 1994, 59, 1613. (60) Huizenga, R. H.; van Wiltenburg, J.; Pandit, U. K. Tetrahedron Lett. 1989,50,7105. (61) Augustin, M.; Frank, J.; Kohler, M. J. Prakt. Chem. 1984, 326, 594. (62) Singh, B.; Lesher, G. Y. J. Heterocyd. Chem. 1990, 27, 2085. (63) Lielbriedis, I. E.; Kampare, R. B.; Dubur, G. Y. Latv. PSR Zinat. Akad. Vestis., Kim. Ser. 1990,2, 212. (64) Paulvannan, K.; Stille, J. R. Tetrahedron Lett. 1993, 42, 6673. (65) (a) Leniewski, A.; Szychowski, J.; MacLean, D. B Can. J. Chem. 1981, 59, 2479. (b) Leniewski, A.; MacLean, D. B; Saunders, J. K. Can. J. Chem. 1981, 59, 2695. (66) Shono, T.; Matsumura, Y.; Kashimura, S. J. Org. Chem. 1981, 46, 3719. (67) (a) Wiesner, K.; Jirkovsky, I.; Fishman, M.; WiUiams, C. A. J. Tetrahedron Lett. 1967, 1523. (b) Wiesner, K.; Jirkovsky, I. Tetrahedron Lett. 1967, 2077. (c) Wiesner, K.; Poon, L.; Jirkovsky, I.; Fishman, M. Can. J. Chem. 1969, 47, 433. (68) (a) Sluyter, M. A. T.; Pandit, U. K.; Speckamp, W. N.; Huisman, H. O. Tetrahedron Lett. 1966, 87. (b) Dubas-Sluyter, M. A. T.; Speckamp, W. N.; Huisman, H. O. Rec Trav. Chim. Pays-Bas. 1972, 91, 157. (69) Wolf, U.; Sucrow, W.; Vetter, H.-J. Z. Naturforsch. 1979, 34b, 102. (70) Marcos, A.; Pedregal, C ; Avendano, C. Tetrahedron 1994, 50, 12941. (71) Barluenga, J.; Jardon, J.; Gotor, V. Synthesis 1988, 146. (72) Barluenga, J.; Iglesias, M. J.; Gotor, V. Synthesis 1987, 662. (73) SchroU, G.; Klemmensen, P.; Lawesson, S.-O. Ark. Kemi. 1967,26, 317. (74) Paulvannan, K.; Schwarz, J. B.; Stille, J. R. Tetrahedron Lett. 1993,54,215. (75) Cook, G. R.; Beholz, L. G.; Stille, J. R. Tetrahedron Lett. 1994, 35, 1669. (76) Knoevenagel, E.; Fries, A. Chem. Ber. 1989, 31, 761. (77) Sanna, P.; Nuvole, A.; Sequi, P. A.; Paglietti, G. Heterocydes 1993, 36, 259. (78) Capuano, L.; Boschat, P.; Miiller, I.; Zander, R.; Schranmi, V.; Hadicke, E. Chem. Ber. 1983,776, 2058.
389 (79) Danishefsky, S.; Etheredge, S. J.; Volkmann, R.; Eggler, J.; Quick, J. J. Am. Chem. Soc. 1971, 93, 5575. (80) Volkmann, R.; Danishefsky, S.; Eggler, J.; Solomon, D. M. J. Am. Chem. Soc. 1971, 95, 5576. (81) Heber, D. Arch. Pharm. 1987,520,445. (82) Ziegler, E.; Hradetzky, P.; Belegratis, K. Monatsh. Chemie 1965, 96, 1347. (83) Dannhardt, G.; Meindl, W.; Schober, B. D.; Kappe, T. Eur. J. Med. Chem. 1991, 26, 599. (84) Ried, W.; Batz, F. Liebigs Ann. Chem. 1972, 762, 1. (85) Brunerie, P.; Celerier, J.-P.; Huche, M.; Lhommet, G. Synthesis 1985, 735. (86) Nagasaka, T.; Inoue, H.; Hamaguchi, F. Heterocycles 1983, 20, 1099. (87) Nagasaka, T.; Inoue, H.; Ichimura, M.; Hamaguchi, F. Synthesis 1982, 848. (88) Danishefsky, S.; Etheredge, S. J. J. Org. Chem. 1974, 39, 3430. (89) Shen, W.; Cobum, C. A.; Bommann, W. G.; Danishefsky, S. J. J. Org. Chem. 1993, 58, 611. (90) Fang, F. G.; Danishefsky, S. J. Tetrahedron Lett. 1989, 30, 3621. (91) Kawahara, N.; Nakajima, T.; Itoh, T.; Ogura, H. Synthesis 1985, 644. (92) Kawahara, N.; Nakajima, T.; Itoh, T.; Ogura, H. Heterocycles 1983, 20, 1721. (93) Chiba, T.; Takahashi, T. Chem. Pharm. Bull. 1985, 33, 2731. (94) (a) Beholz, L. G.; Ph.D. Thesis, Michigan State University, 1994. (b) Barta, N. S.; Stille, J. R. Unpublished results. (95) Seidel, M. C. J. Org. Chem. 1972, 37, 600. (96) Sato, M.; Yoneda, N.; Kaneko, C. Chem. Pharm. Bull. 1986, 34, 621. (97) Singh, B.; Lesher, G. Y.; Brundage, R. P. Synthesis 1991, 894. (98) Kappe, C. O.; Kappe, T. Monatsh. Chemie 1989,120, 1095. (99) Yamada, Y.; Hatano, K.; Matsui, M. Agr. Biol. Chem. 1970, 34, 1536. (100) Openshaw, H. T.; Whittaker, N. J. Chem. Soc. 1961, 4939. (101) Thyagarajan, B. S.; Gopalakrishnan, P. V. Tetrahedron 1965, 21, 3305. (102) Thyagarajan, B. S.; Gopalakrishnan, P. V. Tetrahedron 1965,27,945. (103) Boekelheide, V.; Lodge, J. P., Jr. J. Am. Chem. Soc. 1951, 73, 3681. (104) Bohlmann, V. F.; Ottawa, N.; Keller, R. Liebigs Ann. Chem. 1954, 587, 162. (105) Kurashina, Y.; Miyata, H.; Momose, D.-I. European Pat 309 260, 1989. Chem. Abstr. 1989,777:153656. (106) Thyagarajan, B. S.; Gopalakrishnan, P. V. Tetrahedron 1964, 20, 1051. (107) Kolar, P.; Tisler, M. J. Heterocycl. Chem. 1993,50, 1253. (108) Forti, L.; Gelmi, M. L.; Pocar, D.; Varallo,M. Heterocycles 1986, 24, 1401. (109) Adams, R.; Reifschneider, W. J. Am. Chem. Soc. 1959, 81, 2537. (110) Tonetti, I.; Primofiore, G. 11 Farmaco 1980, 35, 1052. (111) (a) Sevin, A.; Masure, D.; Giessner-Prettre, C.; Pfau, M. Helv. Chim. Acta 1990, 73, 552. (b) d'Angelo, J.; Desmaele, D.; Dumas, F.; Guingant, A. Tetrahedron: Asymm. 1992, 5, 459. (c) Pfau, M.; Tomas, A.; Lim, S.; Revial, G. J. Org. Chem. 1995, 60, 1143. (112) d'Angelo, J.; Guingant, A.; Riche, C; Chiaroni, A. Tetrahedron Lett. 1988, 29, 2667. (113) Audia, J. E.; Lawhom, D. E.; Deeter, J. B. Tetrahedron Lett. 1993, 34, 7001. (114) Enders, D.; Demir, A. S.; Puff, H.; Franken, S. Tetrahedron Lett. 1987, 28, 3795. (115) Barta, N. S.; Brode, A.; Stille, J. R. /. Am. Chem. Soc. 1994, 116, 6201. (116) Benovsky, P.; Stille, J. R. Unpublished results. (117) Hua, D. H.; Bharathi, S. N.; Takusagawa, F.; Tsujimoto, A.; Panangadan, J. A. K.; Hung, M.-H.; Bravo, A. A.; Erpelding, A. M. J. Org. Chem. 1989, 54, 5659. (118) Hua, D. H.; Bharathi, S. N.; Panangadan, J. A. K.; Tsujimoto, A. J. Org. Chem. 1991, 56, 6998. (119) Hua, D. H.; Park, J.-G.; Katsuhira, T.; Bharathi, S. N. J. Org. Chem. 1993, 58, 2144.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
391
Selective Reactions and Total Synthesis of Inositol Phosphates Yutaka Watanabe
1.
INTRODUCTION Biologically and structurally interesting natural products have
stimulated chemists to accomplish their total synthesis. For this purpose various useful synthetic and analytical methodologies have been devised. Such developments have resulted in the realization of efficient total syntheses of these molecules and their analogues. These achievements have greatly contributed in elucidating
their
physiological roles, understanding interactions of substrates with their recognized proteins
(enzymes) at the molecular
level, and
creating useful medicinal substances. Inositol derivatives present a good
representative
example
of
these
years. The structure of inyo-inositol configurational
isomers
of
developments
in
recent
(1), which is one of the nine
inositols
(1, 2 , 3 , 4 , 5,6-cyclohexane-
hexaols) and which is most widely distributed in its derivatives in nature (in animals, plants, and most bacteria), is illustrated below where three types of structures are given (Scheme 1-1) . HO OH
«oj"
l_lc J LOH OH
myo-lnositol
Scheme
1-1
Neurotransmitters, photons and hormones stimulate receptors on the surface of the target cells causing mobilization of the calcium ion
in
intracellular
stores,
thereby
triggering
physiological
responses to occur. The physiological process had been known but it was not clear how the outer information
(first messengers), which
cannot permeate the cell membrane, was transmitted to the calcium stores.
In
1983,
experimentally
this
that
question
was
D-myo-inos i t ol
resolved
by
showing
1 , 4 , 5 - t r i sphosphate
392 [ Ins (1,4 , 5 ) P3 / 2, transmittance
by
second
stimulating
messenger] cellular
mediates
calcium
the
signal
stores.^
Thus,
Ins(1,4,5) P3 is released to cytosol in cells by receptor-regulated hydrolysis of phosphatidyl-inyo-inositol 4,5-bisphosphate [PI(4,5)P2/ 3} in plasma membrane and stimulates a calcium (Ca^^) store by way of the receptor resulting in liberation of calcium ion which causes a
variety
of
biological
responses
(metabolism,
secretion,
contraction, neural activity, and cell proliferation).^ The other hydrolysis product, 1,2-di-O-acyl glycerol (DAG like 2-arachidonoyl1-stearoyl-sn-glycerol
shown in scheme 1-2) is also recognized to
act as a second messenger to stimulate protein kinase C
(PKC).^
These messengers of organic compounds were identified about 3 0 years after the discovery of cyclic AMP in 1957. ^ Disclosure of this signal
transduction
system has helped
to clarify
the processes
involved in various biological pathways. The present understanding (Scheme
of the metabolic cascade of PI(4,5)P2 is summarized below 1-2) .5 O II
*co'CO-i HO-
DAG
PLC
OH
•H03PO •H03PO. 5 OH O" PI(4,5)P2
•HO3PO •HO3PO 5 OH
OPOgH^
lns(1,4,5)P3 PI(4)P -
PI
1 Ins
I
>^^^/(l,4;
• (3.4)
I
insPg
lnsP2 '(1,4) >'^^(1.3)
' PI(4,5)P2
-^
lnsP4
(1.4,5) (1,4,5)
. * ^
> (1,3,4)
^ ^ ^
InsPc
(1.4,6) (3.4.6) InsPeP* (P*=pyrophosphate)
Scheme
1-2
Although inositol itself and many of its derivatives have been discovered over the years,^ the chemistry and biochemistry of inositol had been little investigated. However, in 1961, PI(4,5)P2
393 isolated from beef brain was already structurally characterized by Ballou'^ who described in the literature"^^ that phosphoinositides might be involved
"in the active transport of certain types of
molecules".^ In their work, the structure of Ins(1, 4, 5)P3 , which was obtained by the chemical hydrolysis of PI(4,5)P2, was confirmed. Ballou's
group
also
reported
phosphatidylinositols inositides
from
structure
elucidation
of
glycosyl
(GPIs), a series of mannosylated phospho(Scheme
mycobacteria
1-3) .^
recently, structurally similar glycosyl phosphatidylinositol
More (GPI)
anchors which hold membrane enzymes in the cell membrane through a covalent bond have been found although their physiological role is not clear yet
(Scheme
1-3) .^^
A GPI is also hydrolyzed by the
insulin action to the inositol phosphoglycan which seems to be a second messenger.
S3^
HO HO
HOHO ^ ^ —
•
HO
SliiJlTo
OH
HOHO-
-HS3^ HO
(R = fatty acid residue) Protein—C-N" H
^ P - 0 ^ OH HO ^ • HO HO HOHO HO ' O ^ ^ - ^
HO O^
-
0 Phosphatidylinositol pentamannoside
HO
O-^^HO-
HO.
HO / V n ' ^ e ^ ' " ' " ^
' V^/Cu° ^=^
^^-T?;A HOA.--^
HO
OH
o'^J^^^f-^o HO
Scheme
GPI anchor-protein
1-3
Furthermore, recent researches have shown that there is another type
of metabolic
tyrosine
pathway
kinase-linked
of
inositol
receptors
phospholipids
embedded
in the cell
where the membrane
394 participate. Thus, binding of growth factors and hormones such as insulin to the receptors causes activation of PLCyi or phosphatidylinositol 3-kinase (PI 3-kinase) which respectively hydrolyzes or phosphorylates PI(4,5)P2 resulting in the formation of Ins(1,4,5}P3 and DAG or PI{3,4,5)P3. In the trigger reaction of the old PI cycle described above, PLCpi is activated by the G-protein~ linked receptors in the plasma membrane resulting in the hydrolysis of PI(4,5)P2 (Scheme 1-4). Hormones G protein-linked receptors
acetylcholine histamine vasopressin
PLCft
Hormones (insulin) Growth factors
DAG
lns(1,4,5)P3
PI(4,5)P2
Tyrosine kinase-linked receptors
co-H •HOgPoi^'^O
^mA
•HO3PO •HO3PO
P-0-
6-
OH
PK3,4.5)P3
Scheme
1-4
In organic chemistry, structurally characteristic features of inositol
phosphates
physiological chemists
to
and
related
compounds
characteristics as mentioned the
importance
of
as
well
above
inositol
have
as
their
awakened
chemistry
biochemistry. At present, the biological roles of many
and
inositol
derivatives are unclear. To disclose their functions, their chemical synthesis and analogues are quite useful. From the viewpoint of organic synthesis, they are structurally unique and challenging to synthetic chemists. These facts have directed researchers to prepare various inositol compounds. At earlier stages (around 1984) of the researches in the race to chemically synthesize Ins(1,4,5)P3 and related compounds, there were some problems to be solved: How to perform
multiple
including
phosphorylation
vicinally
of
situating
several
hydroxyl
polyhydroxyls;
groups
how
to
straightforwardly and conveniently protect inositol hydroxyls; how to
conveniently
derivative.
gain
access
to
an
optically
active
inositol
395 Problems
with
In the preparation of myo-
phosphorylation:
inositol phosphates and related compounds, the most crucial problem to be solved is the multiple phosphorylation of polyol derivatives. Especially vicinal diols 4 are very difficult to be transformed to the diphosphates 7, mainly because the monophosphorylation product 5 is prone to cyclization to the 5-membered cyclic phosphate 6 rather than undergoing facts
(Scheme
the second phosphorylation
stimulated
efforts
to
develop
a
new
1-5) . These
phosphorylation
methodology, and in 1987 two types of new phosphorylation methods employing P(III) and P(IV) reagents were successfully introduced for the synthesis of Ins(1,4,5)P3 and Ins(1,3,4,5)P4.
OH
K
OH
Schesne
As
well
as
the
exhaustive
1-5
phosphorylation
of
polyols,
a
regioselective partial phosphorylation of inositols is quite useful especially for introducing a phosphate function at the 1 position of a
1,2-diol
leading
to
the
phosphatidylinositol,
as
discussed
later. Such a methodology, however, was not known until the report on the phosphite-phosphonium approach in 1993.^^ Problems specified
with
For
protection:
free hydroxyl
groups, a
the
phosphorylation
short
access
to
a
of
the
properly
protected inositol is required. Although the protection technology has
developed
derivatives, hydroxyls
is
enormously,
in
the
case
of
the straightforward protection still
quite
difficult
due
inositol
and
its
of some of the six to
their
similar
reactivities. Some useful protecting methods have been
recently
reported in relation to the synthesis of inositol polyphosphates.
Problems
with
optically
active
inositol
derivatives:
Most
inositol phosphates are optically active. Therefore, a generally applicable procedure for getting a chiral derivative is required. This subject has been reinvestigated since starting the synthetic race for obtaining Ins (1, 4, 5) P3 . ^^ In most of the cases where myoinositol is chosen as a starting material, conventional
optical
396 resolution procedures which comprise the derivatization of a racemic compound
to
the
diastereomeric
mixture
and
their
subsequent
separation have appeared using a variety of chiral auxiliaries. Among
them,
camphanic
esters
are
most
frequently
used
as
diastereomeric derivatives resulting in the successful achievement of the optical resolution of various jnyo-inositols.^^ It should be noted
that although optical resolution is often mentioned
to be
cumbersome, there are available now a number of successful reports. A variety of chiral starting materials such as D-and L-quebrachitol, D-glucurono-6,3-lactone, D-glucose, galactinol, (-)-quinic acid, and D-pinitol has been also utilized. Although this methodology allows the avoidance of optical resolution, it does not always provide a concise synthesis. Another choice in obtaining a chiral compound is the asymmetric synthesis which includes enzymatic reaction. The last subject only is discussed in detail in the text. Several review articles and books concerning the synthesis of inositol phosphates and phospholipids are available.^^ This text does
not
cover
all
reports
on the
synthesis
of
inositols
but
principally deals with synthetic strategies and total synthesis of inositol
derivatives which
involve
selective
reactions
such as
stereoselective and regioselective ones. Abbreviations All Bn BOM Bz Cbz CSA DABCO DAST DCC DDQ DEAD DIBAL DMAP DMF DMPM DMSO EE GPI HMPTA LDA Lev mCPBA MEM Ment
Allyl Benzyl Benzyl oxyme thy 1 Benzoyl Carboxybenzyloxy D-10-Camphorsulfonic acid l,4-Diazabicyclo[2.2.2]-octane Diethylaminosulfur trifluoride Dicyclohexylcarbodiimide 2,3-Dichloro-5,6-dicyano-l,4benzoquinone Diethyl azodicarboxylate Diisobutylaluminium hydride 4-(Dimethylamino)pyridine iV, i\J-Diinethylformainide 3,4-Dimethoxyphenylmethyl Dimethylsulfoxide 1-(Ethoxy)ethyl Glycosyl phosphatidylinositol Hexamethylphosphoric triamide Lithium diisopropylamide Levulinyl m-Chloroperbenzoic acid Methoxyethoxymethyl Menthyl
MOM MPM NIS NMO PCC Ph PI PTC Py SEM SM TBDMS TBDPS TBPP TFA TfOH THF THP TIPS TMS Tr TsOH XEPA
Methoxymethyl p-Methoxyphenylmethyl iV-Iodosuccinimide i\7-Methylmorpholine-i\r-oxide Pyridinium chlorochromate Phenyl Phosphatidylinositol Phase transfer catalysis Pyridine 2-(Trimethylsilyl)ethoxymethyl Starting material t-Butyldimethylsilyl t-Butyldiphenylsilyl Tetrabenzyl pyrophosphate Trifluoroacetic acid Trifluoromethanesulfonic acid Tetrahydrofuran Tetrahydropyranyl 1,1,3,3-Tetraisopropyldisiloxanyl Trimethylsilyl Tri tyl (Triphenylmethyl) p-Toluenesulfonic acid o-Xylylene iV, iV-diethylphosphoramidite
397 2 . PHOSPHORYLATION 2.1 Exhaustive Phosphorylation To achieve
total
synthesis
of
inositol phosphates
derivatives, multiple phosphorylation most
crucial
step. Use
of
and
of polyol derivatives
dianilinophosphoric
chloride
only method of choice for this purpose. However, deprotection
of
several
the is
furthermore
dianilinophosphoric
esters
the same molecules is quite difficult, while phosphorylation 2,3,6-tribenzyl-jnyo-inositol
is the
8 was
its reactivity
not satisfactory for perphosphorylation of inositols and spontaneous
related
in
of D-
9 with the chloride 8 giving 11 in 60%
yield was accomplished after exploring proper reaction conditions in the first total synthesis of Ins(1,4,5)P3 the
phosphorylation
positional reported
isomer
product
12 was not
of 9 under
similar
(Scheme 2-1).^^^ formed
at all
conditions.
1987, it was
that the reaction of tetrabenzyl pyrophosphate
with alkoxides generated in polyphosphorylated
shown in Scheme
products
in good
smoothly gave
yields.
Examples
the are
2-1.1^^
BnO.
BnO^^A^OBn
PO
PO^^'OBn OP OP 11 12 (/^Phosphoryl group introduced)
MoBn or
HO'* Sr"OBn OH 9 (R^=H, R2=Bn) 10 (R^=Bn, R 2 = H ) {PhNH)2P(0)-CI 8, Py [(BnO)2P(0)]20 13, n-BuLi (BnO)2PN/-Pr2 22, tetrazole then mCPBA
ItsX
60%
0%
70%
40%
88%^^^
^P-NEt2, tetrazole then mCPBA °
OP
OBn
OR'^
(fjf"
(TBPP, 13)
by the action of a strong base such
situ
as NaH,^^*^ KH,!^ or butyllithium^^^ on inositols desired
In
However,
from 1 0 , a
87%
97%
23
Scheme
2-1
It is interesting to note that when the monolithium salt of the vicinal
diol
diphosphate
14
was
treated
15 was produced
with
equimolar
in 2 6% yield
together with 46% of the starting material monophosphate was observed at all equiv
of
butyllithium
and
2.5
amounts
(based on
of
TBPP,
the
pyrophosphate)
(SM) and no corresponding
(Scheme 2-2).^^ equiv
of
TBPP
However, when
were
used,
15
2.4 was
obtained in 81%. A similar phosphorylation procedure for the simple 1, 2-trans-cyclohexanediol
16
gave
an
identical
result.
However,
398 product distribution was found to depend on electrophiles as shown in Scheme 2-2. when benzoic anhydride was used as an electrophile, the monobenzoate 18 (El=Bz) became the predominant product yield),
contrary
experiments
to the result
of the pyrophosphate
(92%
13. These
indicate that a salt exchange reaction between
the
lithium salt 19 and the monophosphate 18 [El= (BnO) 2? (0) ] formed first is much faster than the substitution reaction of 19 with TBPP while benzoic anhydride reacts faster with 19 than its salt exchange with the monobenzoate 18.
rvBuLi
BnO.
0 'C, 2 h
-78 'C, 10mln
^^^
{1 equiv each)
16
a:^'
rvBuLi
El OH
(BnO)2P{0)0
1.1 eq 2.5 eq
1.0 eq 2.4 eq
^"^^
Bn04
t(BnO)2P(0)]20
17
'OH
O-EI
18
(BnO)2P(0)-0(0)P(OBn)2 (BnO)2P(0)-F BnOC(0)-CI Ts-CI PhC(0)-OC(0)Ph
94: 6 79:21 45 : 55 34:66 8:92
a.
,0-EI 0-EI
17
Scheme The
second
phosphitylation
phosphorylation using
and
intermediate
subsequent 24
perphosphorylation reactivity the
of
synthesis
may of
provide
of a
with
and y i e l d
protected
which
21 such as
( 2 2 ) ^'^ a n d
oxidation
polyols
the reagent, of
methodology
phosphoramidites
diisopropylphosphoramidite (XEPA)IS
2-2
its
cyclic
the
more
general to
N,N-
analogue
resultant
respect
involves
dibenzyl
method
1,4,5-
for
applicability,
(Scheme 2-3) . In Scheme
myo-inositol
23
phosphite
and
2-1,
2,4,5-
399 trisphosphate from the corresponding p h o s p h o r y l a t i n g methods i s summarized.
ROH
(R'0)2PNR"2 21 tetrazole
•
[O]
ROP(OR')2 g^
Scheme
triols
•
using
various
ROP(0)(OR')2 ^5
2-3
2.2 Reaioselective Phosphorylation The methods described in the preceding section lead to smooth phosphorylation
of all hydroxyl groups
in inositol derivatives.
Contrary to this exhaustive phosphorylation, the regioselective one is also quite useful, especially for the selective formation of the phosphoric
diester
functions
at
the
C-1
position
in
1,2-diol
derivatives of myo-inositol in the synthesis of phosphoinositides. These 1,2-diol derivatives are easily accessible because 1,2-cisdihydroxyl
groups
of jnyo-inositol
ketalization in comparison with trans
are
easily protected
by
its
diols. The diol derivatives
thus obtained comprise one equatorial and the other axial hydroxyl groups. Since the former is generally more reactive than the latter, several electrophiles were selectively introduced at the 1 position of the diols as shown in Scheme 2-5. However, phosphorylation at C1 in a 1,2-diol was extraordinarily difficult when known procedures were
used
because
migration
of
the
phosphate
function
and
cyclization occured with ease.^^ Meek et al. reported a regioselective phosphitylation using a P(III) reagent, dimethyl phosphorochloridite Ins(l,4,5)P3
in the synthesis of
(Scheme 2-4).20 Thus, the dibenzoate 26 which can be
readily derived from myo-inositol in three steps, was treated with 3.3
molar
equivalent
of
the
ethyldiisopropylamine at -40
chloridite
in
the
presence
of
°C to afford 1,4,5-trisphosphite
in
high yield, the C-2 hydroxyl group remaining free. The resulting phosphite was acetylated and oxidized to give the fully protected form 27 of Ins(1,4,5)P3, which was finally deprotected. Treatment of the
1,4-dibenzoate
26
in refluxing aqueous pyridine afforded a
mixture of dibenzoates, from which the desired 2,4-dibenzoate 28 was isolated and transformed to Ins(1,3,4,5)P4, as above. In general, preparation of 1-phosphate
derivatives has been
accomplished according to a tedious sequence starting from 1,2-diols 29 as follows: temporary protection of the OH-1 in 29 with the allyl, MOM, silyls
[Ph2{t-Bu)Si, Me2(t-Bu)Si, EtsSi], and benzoyl
400 OAc BzO^^A^OH OBz
HO'
(MeOgPCI
I.AcCI, DMAP
'-Pr2NEt DMF, -40 'C
2. H2O2
BzO^^A^OP(0)(OMe)2
Py.HgO 100'C
ca. 88%,
LHBr.AcOH 2. LiOH
lns(1,4,5)P3
OBz HO.
OBz 0P(0)(0Me)2 27
{MeO)2{0)PO 94%
OH
(MeOgPCI
I.HBr.AcOH -^ 2. LiOH
/-PrgNEt DMF, -40 "C
lns(1,3,4,5)P4
Scheme 2-4 groups, protection of OH-2 with Bn, THP, ArCO, EE (used in 3-deoxy3-fluoro derivative), deprotection of OH-1, and phosphorylation of OH-1
in
32
(Scheme
regioselective
2-5) .17c, 21
alkylation,
;^s
acylation,
shown and
in
the
silylation
scheme, at
the
1
position in the 1,2-diol derivatives 29 have been documented with ease
by
conventional
procedures
but
attempts
at
selective
phosphorylation at C-1 of these molecules have failed. OH
OH
RO.
Jv^OH
RO'
T^'''OR
RO.
+R1
+R2
^X^OR^
^As^OH 1 RO'^ Y^""OR
N
RO^' ^i""OR OR 31
RO'^ ^ I "'OR OR 30
OR 29
0R2
0R2
yK^OR^
OR 32
R^=AII, MOM, Ph2{f-Bu)Si, Me2(f-Bu)Si, EtgSi, PhCO R2=Bn, THP, ArCO, EE (for 3-deoxy-3-fluoro-mya-inositol)
Scheme
2-5
Jastorff et al. reported a novel approach to the preparation of inositol 1-phosphate via the selective ring opening of phosphate
derivatives
1,2-cyclic
34 with alcohols (Scheme 2-6).22 The best
selectivity was observed using t~BuOH at low temperature (0 'C) even though the chemical yield was low (28%) . A phosphorus derivative which has the reactivity between highly reactive P(III) and less reactive 0=P(V) is expected to be suitable for
the
selective
phosphorus generated in
P(IV) situ
of pyridinium
phosphorylation.
derivative,
Thus, a
phosphonium
salt
tetracoordinate 38
which
was
from the corresponding phosphite 37 by the action
bromide perbromide
reacted with various
alcohols
401 O II O—R^- OBn
YS-°
BnOP(NEt2)2 1 eirazcNe then mCPBA
BnO'
ROH 0 'C, 22 h
y "OBn OBn 34 0
OH BnO^ BnO^'
BnO,
»0-P-OBn OR 'OBn
+
7 25 >100
:
BnO^
OBn 35 ROH=MeOH hPrOH f-BuOH
49% yield 39 28
1 1 1
Scheme 2-6
resulting
in the
Arbusov-type
formation
of the phosphoric
decomposition
39
of
triesters
(Scheme
40 via
2-7).
This
phosphorylation methodology can be ideally applied to 1,2-dihydroxy inositol derivatives and 1-phosphates are thus selectively obtained. The
method
opens
a
convenient
way
to
synthesize
phosphatidylinositols. Phosphorylation of 1,2:4,5-dicyclohexylideneinyo-inositol 49a using glyceryl phosphites was also achieved at C-3 regioselectively by the phosphite-phosphonium
approach to afford
43b and 43c. The former product 43b was then subjected
to the
phosphorylation using XEPA giving the fully protected phosphatidylinyo-inositol 4-phosphate 44 give PI(4)P
quantitatively which was deprotected to
(Scheme 2-8). Similar
selective phosphorylation
was
observed in the reaction of 49b with glyceryl phenyl phosphate in the presence
of iV-mesitylenesulfonyltriazole
(Scheme
2-9) .^3
Phosphorylation of 49b with diphenyl phosphorochloridate was also recorded.^^ At present, methods for exhaustive and partial phosphorylation of
inositol
derivatives
have
been
established.
These
methods
facilitate greatly the syntheses of inositol phosphates, inositol phospholipids and their analogues.
402
PyHBra
37 (Ri=Bn, Me)
, +^0R2 (R^O)2P: Br
^
(R^0)2P0R2
CH2CI2
R30H
rR^OUP^^'^
EtaN
•
'•Br
R^O-P, 40
39
38
„ OR^
O17H35CO2
j
O17H35OO2 Product ?^ ".OR BnO,^A^O-P:
0R2
T OBn oV>c OBn 41
BnO
OH
(BnO)3P
(MeO)2PO—'
nupn—> (MeO)2PO
a
b
c
94% (-42 'C, 1.5 h)
85% (0 'C, 0.5 h)
93% (-42'C, 1.5 h)
94% (0 X , 0.5 h)*
^ OR^
BnO,
OP
42
Q JoR^ HO
I
95% (0 'C, 0.5 h)
61% (-18 'C, 1.25 h)* SM: 20% 4-P: 6% 1,4-P2: 6%
O
P=(3C^P(0) Scheme
85% (-20*C, 1.5 h)* SM: 11% 4-P: 4% 1,4-P2: 6%
*Py/CH2Cl2(1:10)
2-7
O )-P-OMe
XEPA tetrazole 43b
^-0' ^0-P
thenmCPBA
T
''o
U02CC17H35
0-y-\
'-02CC,7H35
PI(4)P
& ' Scheme
2-8
C15H31CO2-1 Ci5H3.C02-[^W_^^
/
ONa
C15H31CO2—I C,5H3iC02^
c^i^o-p-o-J
Me HO
H0''V''0 ' 45
49b I
(PhO)2P(0)-CI EtgN, DMAP r.t, 4 h
60% (4-P: ca. 5%, 1,4-P2: <1 %)
0,1^0P(0)(0Ph)2
HO''y''0
07^
46
Scheme
30% (4-P: trace, 1 .A-Pg: 8%) 2-9
403 3. Protection of Inositols 3.1 Ketalg Cyclohexylidene and isopropylidene monoketals 47 and diketals 48, 49, and 50 are well-known protected myo-inositol
derivatives
(Scheme 3-1). Compounds 47 ketalized at the 1,2-cis-positions have been utilized conveniently for the synthesis of various inositol phosphates
since 47
can be regioselectively
functionalized
and
prepared in good yield by the conventional ketalization procedure and subsequent partial deprotection of the less stable trans-type ketals from the diketal mixture formed first in the reaction.^^ Three diketals^^ have been also often employed for the synthesis of target
inositol
derivatives,
because
they
have
the
following
advantages: (1) A trans vicinal hydroxyl moiety as well as the 1,2cis-diol are conveniently protected;
(2) the more reactive trans-
ketal function may be chemoselectively removed after modification of the two remaining free hydroxyls. R O-VR
R 0-\-R
R 0~\-R
R 0,^1^0 HO** Y^^" OH OH
R 0'* X
R O—\-R
HOvAyO
"OH
HC^A^O
o**Sr " O H
OH
Ho^^Sr'"O
R-yO
O—AR R' 49 50 a: [R-R=(CH2)5], b: R=Me
R
47
48
Scheme
3-1
3.1.1 The Reaction of 1.2:4,5-Diketals The
reaction
of
1,2:4,5-diketals
49 with
a
variety
of
electrophiles has been reported, and it was found that the use of a limited quantity of the reagents and control of reaction temperature may result in a fairly good regioselective monosubstitution at the 3 position. Benzylation of dicyclohexylidene-inyo-inositol
49a with
benzyl bromide in the presence of NaH under toluene under refluxing conditions was reported to produce selectively the 3-benzyl ether 51 (R=Bn)
in
70 (Ref.l3a)
and
60%27
yields
(Scheme
3-2) .
Benzylation^ld, 28 ^nd allylation^^ using BaO and Ba{0H)2 in DMF afforded successfully the 3-alkylation products 51 (R=Bn and All) in 70
and
68%
yields
respectively.
These
authors
claimed
that
combination^^ of BnBr and NaH resulted in poorer selectivity {48%
404 for 3-ether and 16% for 6). Some reports on alkylation of 49 showed also poor selectivities.24,30 camphor ketal 52 was transformed to the
3-benzyl
ether
53 in good
yield
(BnBr,
NaOH,
HMPTA,
Benzene).^^ In general, the 3-hydroxyl group is more reactive toward electrophiles than OH-6. Furthermore, as mentioned later, the OH-3 in the 1,2-monoketal 47 has also the highest reactivity among the four hydroxyls. These facts may be explained in terms of steric hindrance. Moreover in order to get good regioselectivity in the alkylation, a low concentration of the alkoxide should be kept in the solution by using a weak base such as barium hydroxide or by employing a less polar solvent, since control of reactivity of an alkoxide which has
inherently
a strong nucleophilicity
is very
difficult. Chung and Ryu^^ have mentioned that solvent dependent change of the regioselectivity in benzylation might be due to the difference in the reactivity of the alkoxide in each solvent.
49a
|3
6j
•'OH
BnBr, NaH, PhMe, refl: 70% BnBr, BaO, Ba(0H)2, DMF, r.t, 4 d: 70% (diBn: 3%) AIIBr, BaO, Ba(0H)2, DMF, 20 'C, 72 h: 68%
BnBr, NaOH HMPTA, Benzene 45'C
n^Z
68%
Scheme
Regioselective acylation of 49a was accomplished by treatment with iV-acyl imidazole
in the presence of CsF to give 3-0-acyl
products 54 in good yields along with a small amount of 3,6-diacyl derivatives
(Scheme 3-3) . When benzoyl chloride was used, the 6-
acyl product was formed to some extent in addition to two products. Interestingly employment of benzoic anhydride showed the opposite tendency giving the 6-benzoate as the main product.^2 Chung and Ryu reported a comprehensive study on alkylation, acylation,
trifluoromethylsulfonylation,^^
silylation, and
405
phosphorylation of diisopropylidene i n o s i t o l s 49b and 50b with good t o moderate s e l e c t i v i t i e s . ^ 4 Yield, %
HOyL^O
a
O^^Y"OH O
^CO-N^^
RCO,^0
CsF.DMF
49a
-
0'^T"OH /\"0
p^^^.,^ 3-acvl ^^
6-acvi 3,6-diacvl
A>BnOPhCO-lm 79 CHgCO-lm 70
-
18 27
(PhC0)20 11 PhCO-CI 53
41 13
10 17
\_J 54 Scheme 3-3
3.1.2 The Reaction
of 1.2:5,6-Diketals
Highly regioselective (regiospecific) alkylation of 1,2:5,6dicyclohexylideneinositol 50a was achieved by reaction of its stannylene derivative with CsF as shown in Scheme 3-4 (Refs.17a,34) whereas use of tetrabutylammonium bromide in place of CsF with application of heat gave no selectivity.^5 Exceptionally, methoxybenzylation (n-Bu2SnO then MPM-Cl, CsF, and KI) of 50a gave a poor regioselectivity (3:1).^6 in the case of the isopropylidene homologue 50b, regioselectivity of an identical benzylation in the presence of CsF decreased.^4 Q S F increases the nucleophilicity of the alkoxides in 3,4-0-stannylene derivative 57, and consequently, allows the reaction to proceed at a lower temperature (room temperature) than that without CsF. The CsF-promoted stannylene methodology therefore provides the methodology for highly
Snn-Bu2
1. n-Bu2SnO, 2. BnBr, n-Bu4NBr, PhMe, 70 *C BnBr, n-Bu4NHS04, CH2CI2, 5% NaOH, refl. •D.-M. Gou, W.-R. Shieh, P.-J. Lu, and C.-S. Chen, Bioorg. Med. Chem., 2 (1994) 7-13.
Scheme
3-4
406 efficient and regioselective nucleophilic substitution. It should be noted
that this procedure was first used
for an intramolecular
nucleophilic substitution of the mesylate derivative as a key step in
the
total
synthesis
group^'7 and somewhat
of
later
octosyl
acid
A
by
the monoalkylation
Danishefsky's
and acylation of
dimethyl tartrate via its stannylene derivative in the presence of CsF
was
reported.-^^ This
efficient
procedure
was
used
for
benzylation of the 5,6-diol 186 giving the 6-0-substitution product 187 in 80% yield which was then converted to Ins (1, 3 , 4, 5) P4 (Scheme 4.7) 34a,39 synthesis (Scheme
The method was also conveniently
of
inositol
6-3) .^^
benzylation although
it
A
50a
of
utilized
1, 4, S-trisphosphate^'^^
phase
transfer
showed
poor
sometimes
catalysis
provides
more
procedure (Scheme
selectivity
for the
and PI{3,4,5)P3 for
3-4)2 6a
selective
product
distribution^-^ in the alkylation of the 4,5-diols rather than the Ba(OH)2-'^2 or NaH-mediated^^ procedure. Acylation of 1,2:5,6-diketals 50a showed remarkable selectivity for the 3-OH group. ^4 rpj^^ benzoylation product converted
into
6-deoxyinositol
1-phosphate
58 from 50a was
according
to the
sequential procedures shown in Scheme 3-5.'^'^
Qo
Ov^A^OR
l2,Ph3P.lm^
o"Sr "OH r. (^
Qo
O^^A^OR
^2% o;'Y^! \ 1 '
O
/ \
50a: R=H > g^^, p 58: R = B z ^
2.KOH
61%
3. (PhO)2P(0)CI 4.Si02
68%
Qo
0,^A^OP(0)(OPh)2
«V ^ *
°^
59
H2 PtOo
60
^^/^ 6-Deoxy-lns(1)P
Scheme
3-5
B e n z y l a t i o n of 1 , 2 : 3 , 4 - d i c y c l o h e x y l i d e n e i n o s i t o l prpc2 6a and t i n - m e d i a t e d ^ 5 (without CsF) c o n d i t i o n s selectivity.
48a under showed no
3 . 1 . 3 The r e a c t i o n of 1,2-Ketals The r e l a t i v e r e a c t i v i t y of the four free hydroxyl groups i n k e t a l s 47 of n i y o - i n o s i t o l can be e s t i m a t e d based on t h e r e p o r t s several
substitution
chlorosilanes.
reactions
with
acyl
chlorides
The r e a c t i o n of 1 , 2 - c y c l o h e x y l i d e n e - i n y o - i n o s i t o l
1,2of and 47a
407 with mandelic acid chloride afforded selectively diastereomeric 3mandelates
61 and its opposite diastereomer
{>20 relative to 6-
regioisomer 62) in 55% yield from which one diastereomer 61 with >96% d.e. was obtained by precipitation from ether-hexane
(Scheme
3-6).^^ Other mono-acylation products were not detected
in the
reaction while peracylated derivatives were formed. An identical result was obtained in the reaction of the same substrate 47a with benzoyl chloride.^^ Treatment of 47a with 1.1 equiv of TBDPS-Cl and imidazole
afforded
combined yields
the
3-silyl
ether
61
predominantly in
75%
(61/62=20) while 2 equiv of the reagent gave the
3,6-disilyl ether. ^'^ From these results the order of
reactivity for
electrophilic substitution can be estimated at least as 0H-3>0H6>OH-4,OH-5. On the other hand, acetylation of optically active la in
the
presence
of
MS
4A without
a
tertiary
amine
in N, N-
dimethylacetamide showed different regioselectivity, giving the 4acetylated product 63 and the 5-acetylation product 64 in a ratio of about
1:1
(Scheme
3-6). The mixture
of
the
two
mono-acylated
products was separated by recrystallization or by chromatography after phosphorylation of the acetates. When the tertiary amine was employed in this reaction, many spots were observed on TLC. The acetates 63 and 64 were converted to D-Ins(1,4,5)P3 Ins(l,4,6)P3
(Ref.48) respectively
by
(Ref.46) and D-
phosphorylation
and
deprotection.
9c
R-CI
HS° OH 74% (ca1 :1)
Py
OH
47a AC2O, MS 4A AcNMeg, r.t.
D-lns(1,4,5)P3
*AcO
61
H&°
OH
62
TBDMSO"^^COCI, -40 °C 55% (61/62>20) PhCOCI, r.t. 47% (for 61) TBDPSCI, -10'C 75% (61/62=20)
9. HOV . . Z - 7 — - ^ O H OAc 63
9c
HO HO
OH 64 D-lns(1,4,6)P3
Scheme
3-6
1,2-Camphor k e t a l 65, which can be conveniently prepared by the r e a c t i o n of m y o - i n o s i t o l w i t h D-camphor dimethyl ketal r e g i o s e l e c t i v e l y (Scheme 4-4),^^''^'^ showed i d e n t i c a l r e a c t i v i t y
408 with that of 47. Thus, treatment of 65 with 1.1 equiv of TBDPS-Cl in the presence of imidazole in pyridine at -10 °C gave the 3-silyl ether 66 in 88% yield together with <5% of the G-isomer^"^ while the same reaction in acetonitrile at room temperature provided 66 in low yield
(30%)
(Scheme
3-7).^^'"^^ Employment
of 2.1
equiv of
the
silylating reagent under the conditions using pyridine gave 84% yield of the 3,6-disilyl ether 67, which was also obtained by the reaction of the monosilyl ether 66 with the reagent under similar conditions. Acylation using pivaloyl chloride in pyridine was also studied. On increasing the molar ratio of the chloride the major product changed from the 1-monoester 69 to the 3,5-diester 70 and then to the 3,5,6-triester 71 in moderate yields. The tripivalate 71 was converted to Ins (1,4,5) P3 in five steps.-^^'^^ The reactivity of the hydroxyl groups different
in the tetrol
65
toward
pivaloylation
from that to silylation. Interestingly,
the
was
3-0-TBDPS
ether 66 was acylated regioselectively with pivaloyl chloride to give the 6-0-substituted product 68 in 78% contaminated with ca. 15% of an unidentified isomer. This result is different from that in the case of acylation of 69 where the 3,5-dipivalate 70 was obtained predominantly,
indicating
that
the
remote
silyl
substituent
regulated the direction of substitution site.
Imidazole, Py HO..Z-r^^CZoH ^^
Ph2(f-Bu)SiCI
H0^Z-r^^^^!>-^^'^''^")P^2
1.1 mol: 88% (for 66)
OH
78%
^^
3
R=^BuCO
Py, r.t. f-BuCOCI
1.1 mol: 44% (for 69: R=R'=H) 2.2 mol: 40% (for 70: R=H, R'=f-BuCO) 5.0 mol: 48% (for 71: R= R'=:f-BuCO) RO7 5 OH
Scheme
3-7
^
-02CNBU
730/^
409 O O OCr^Pr n-PrCC^X^OH
0 . pp^^^ (i.ieqf
HO T "OH OH
0
? ^'^^^^ />PrCO,^A^O(p^Pr
Py--"^
HO T " 0 H OH
72
? ^'^^^'^ />PrC0,^4s^0H
H O T "OC^^Pr OH
73 50%
Scheme
0
? ^'^^^' A>PrCO,^A^OH
H O T "OH Ofn-Pr
b
74 25%
75 5%
3-8
In relation to the substitution reaction of the 1,2-ketal 47, the reaction of 1,2-dibutyroyl inositol 72 with 1.1 eguiv of butyric anhydride
in pyridine was
found to give predominantly
the 3-0-
acylation product 73 in 50% yield together with smaller amounts of 74 and 75.^*-' One product 74 was used to prepare racemic 2,3,6-triO-butyroyl-myo-inositol 1,4,5-trisphosphate, which was proved to be a membrane-permeant Ins{l,4,5)P3 agonist (Scheme 3-8). 3.2 Protection with Digilox^ne
OH HO^OH
i 0 ^
0-Q HoJsJo
HO T "OH TsOH HO^Sr"OH OH ^Z" OH 1 47a APr2Si(CI)-0-Si(CI)/-Pr2 I ' Py.13h,r:t. i94%
- ^ J o " T "OH OH 76
' ^'-oTT^ _SC^.V"oR.
Al%
^oi^O^yXv^O o T X
^--^ ^/^
-^S^^vA^O ^"oT J ^fO^ T "OLev
/
^=>
'"^^^^'^ '"^^5)P
^
,^ ^ ^,« »ns(1,3,4)P3
^ "^^
PI(3A5)P3 PK3,4)P2 GPi
"^^
i: BzCI, Py; ii: BzCI, DMAP; iii: CH3CO(CH2)2C02H. DCC, DMAP
Scheme
3-9
The reaction of 1,2-cyclohexylidene-inyo-inositol dichloro-1,1,3,3-tetraisopropyldisiloxane a completely
47a with 1,3-
(TIPS-CI2) takes place in
regioselective manner to afford
the
3,4-disiloxane
derivative 76 in quantitative yield. ^-^ Compound 76 has been shown to be a very useful synthetic intermediate by the synthesis of various inositol phosphates and phospholipids
(Scheme 3-9). The usefulness
410 of 76 is based on suitable regulation of the reactivity of the two free
hydroxyl
groups
by
the
steric
bulkiness
of
the
diisopropylsilylene group at the side of C-4 as well as the easy availability of 76 as illustrated in the scheme. Furthermore, the two protecting
groups
in 76
can
be
distinguished
chemically,
consequently selective removal of the desired one is quite easy. In the
case
of
similarly
protected
1,2:3,4-diketals
48, a highly
regioselective reaction will not be expected and the function over the cis
trans-ketal
always has to be first removed at an earlier
stage of the synthesis. The completely regioselective acylation at the 6-position of 76 was achieved by benzoylation and levulinoylation. The monobenzoate 77 thus obtained was phosphorylated quantitatively by the method using PCI3 and the four-step deprotection procedure to give 5-phosphate
inositol
[Ins{5)P]
myo-
(Scheme 3-10). On the other hand,
phosphorylation of 77 using butyllithium and TBPP afforded the 6phosphate 81 resulting from the initial migration of the benzoyl group and subsequent phosphorylation of the less hindered OH-6. The phosphate
81
was
levulinic
ester 79
synthesis
of
converted
inositol
phosphatidylinositol which
are
to
is a pivotal
racemic
inositol
synthetic
phospholipids
and
GPIs,
to
play
an
important
The
for the
especially
3 , 4 , B-trisphosphate^^ g^^d
postulated
Ins(4)P.
intermediate
for
3 , 4-bisphosphate
role
in
a
new
intracellular signaling system connected with the tyrosine kinaselinked
receptors.
The
efficiency
of
their
synthesis
depends
largely on the usefulness of the disiloxanyl protecting group.
Olns(4)P -^ .0^....... 83x90x85%
TIPS5 II J J ^ o ' T "'OP ^o^r"'nP ^^^
81
-* -* 4^/» 47%
77 77
•• TIPS II J J ^^^ o T "^" Puant •o'*N^"OBz
P=(BnO)2P(0)
• 89x85x61%
lns(5)P
^^
80
1:1. PCl3, 2. BnOH, 3. f-BOgH; ii: 1. (HOCH2)2 , TsOH, 2. EtgNHF, 3. Hg, Pd-C, 4. NH3, MeOH; iii: n-BuLi, TBPP
Scheme
3-10
Dibenzoylation of 76 can be also realized by using a combination of benzoyl chloride and DMAP in 88% yield. The resultant benzoate 78 was
converted
to
the
optically
active
menthoxyacetate
83 by
successive decyclohexylidenation, selective acylation at C-1, and optical resolution
(Scheme 3-11). A diastereomeric mixture of 83
411 and the other isomer can be separated effectively by a chiral column chromatography while the racemic diol 82 was difficult to separate by
the same
column.
Derivatization
of
a
racemate
to the
diastereomers bearing a proper chiral auxiliary is very useful for optical resolution by a chiral chromatography. Benzoylation of 83 followed by removal of the menthoxyacetyl and TIPS groups gave the 1,3,4-triol which was then phosphorylated by the method using XEPA.
V, 76%
78 97%
OBz OBz R^=R2=H R^=MntAc,
R2=H
R^=MntAc,
R2=B2
R^=H,
R2=:BZ
i: (HOCH2)2, TsOH; ii: MntAcCI; iii: BzCI, DMAP, EtN/'-Prg; iv: NHgNHg; v: EtgNHF; vi: XEPA, tetrazole then mCPBA
Scheme 3-11 Azidodeoxy-myo-inositol
87 was efficiently protected with 2.5
and 1.2 equiv of TIPS-CI2 under conditions shown in Scheme 3-12 respectively
giving the corresponding bis-disiloxane 88 and mono-
disiloxane 90 in high yields respectively. These were utilized to prepare the tritium-labeled jnyo-inositol analogue 89 and 3-azido-3deoxy-myo-inositol 2 , 4 , 5-trisphosphate 91.^-^
TlPS-Cig (2.5 eq) /
OH
/
TIPS
73"C,40h 80%
0'**' T ^ \ 1 TIPS-0
I.AC2O, DMSO 95%
1
TIPS
2. NaBT4
88
^OH "OH
HO OH 87
\ \
\
TIPS-CI2 (1.2 eq) r.t, 10h 95%
OH
^
TIPS
OP03Na2
1.NaH, [(BnO)2P(0)]20 61% 2. a) TMS-Br, b) H2O ONaOH 68%
Scheme 3 - 1 2
OH _ . NaaOgPO
412 Bruzik and Tsai applied the protection strategy using the TIPS group to the synthesis of various optically active precursors for inositol
phosphates
and
inositol phospholipids
from
1,2-camphor
ketal 65 of niyo-inositol. ^^ The parent myo-inositol itself can be protected regioselectively by
the
TIPS
derivative
group
to
the
1, 6:3,4-bis (disiloxanylidene)
92 in 66% yield together with trace amounts of other
inositol derivatives inositol
give
(Scheme 3-13).^^ A similar type of protected
is difficult
to obtain by any known method. The novel
intermediate
92 was benzoylated followed by removal of the TIPS
group
phosphorylation
and
tetrakisphosphate 94, which Ins ( 1 , 3 , 4 , 6 ) P4 .
Thus,
a
to
form
was
finally
fairly
the
fully
protected
deprotected
concise
to
synthesis
give
of
the
tetrakisphosphate has been completed. OH lns(1,3,4,6)P4
93: R=Bz 97%
p= ( Q C ! ^^^^^
i: TlPS-Og, Py; ii: B2CI, Py, refl.; iii: aq. HF; iv. XEPA, tetrazde then mCPBA; v: Hg, Pd-C; vi: MeONa
Scheme
3-13
3.3 Reaction of mvo-Inositol Orthoformate The
orthoformate
95 of myo-inositol,
which
was
originally
reported by Lee and Kish,^^ has been utilized frequently
for the
synthesis of various inositol derivatives. It was first employed for the synthesis Ins (1, 3 , 4 , 5) P4 by two groups . ^^ ' ^'^^'55 rpj-^^ Merck group found that action of one molar equivalent each of NaH and several electrophiles on the orthoester 95 afforded the 4-0-monosubstituted products 96 in good yields, together with a trace of 4,6-dibenzyl ether
99
in the case
of benzylation
(Scheme
3-14) .
The
high
regioselectivity and monosubstitution may be rationalized in terms of
the
formation
of
a
thermodynamically
preferable
chelated
intermediate 97 and subsequent substitution at the fixed anion site. Introduction of the second benzyl group to 96 was not so selective while benzylation of 95 under catalysis
conditions
selectivities
gave
acid-catalyzed
the 2,4-dibenzyl
(Scheme 3-14). ^'^^
or phase ether
transfer
98 with good
413
,-/^o
, - ^ 0 1.NaH(1eq)
HO, HOI
OH
95 TBDMS-CI 2,6-lutidine
J^P HO.
2.RX
BnBr: 75% p-MeOPhCHsCI: 6 7 % AIIBr: 8 0 % [(BnO)2P(0)]20: 7 2 % BnOGHgCI: 6 7 %
HOI OR
96
97
o/^o •J3
58%
HO, BnO I OBn 99
B n O li
98 TBDMSO,
HOI OR 100: R=H - \ BnBr. r-BuOK 101: R=Bn-*^
I OH
Cl3CC{NH)0Bn, TfOH
81%
<10%
BnBr, n-Bu4NHS04 5% NaOH, CH2Cl2.refl.
64%
16%
Scheme
3-14
The 4-allylation product 96(R=A11) was successfully used for the synthesis of DL-inyo-inositol
1,3,4,5-tetrakisphosphate
(Scheme 3-
3^5)^15,55 Thus, the remaining hydroxyl groups in 96 was benzylated followed by rhodium-catalyzed migration of the double bond in the allyl group and hydrolysis to afford tetrol 102. Phosphorylation of 102 and deprotection using a hydrogenolysis procedure gave very concisely the target tetrakisphosphate. This is the shortest way to racemic Ins(1,3,4,5)P4 reported to date. OBn
OBn
(BnO)2P{0)0,
0P(0)(0Bn)2 lns(1,3,4,5)P4
96 (R=AII)
(BnO)2P{0)0"
'OBn 0P(0){0Bn)2 103
i: NaH, BnBr (86%); ii: RhCI(PPh3)3, DABCO; iii: aq. HCI, MeOH, refi. (56%); iv: NaH, TBPP (70%); v: Hg, Pd-C (quant.)
Scheme
3-15
Optically active D- and L-Ins(1,3,4,5)P4 were also prepared from the orthoester 95 by Vasella et al. although the process required many additional steps compared with the synthesis of the racemic material described above. The orthoester 95 was
regioselectively
silylated to give 58% of 100 which was then mono-benzylated by the chelation-controlled reaction mentioned above to afford the ether 101 in 99% yield
(Scheme
3-14) . Chiral
synthetic
intermediates
414 from 100 or 101 by enzymatic
were prepared starting
or chemical
(Scheme 4-8) .
procedures
The 2,4-dibenzoate 104 of the orthoester 95 showed interesting (Scheme
reactivities
3-16).^^
Thus,
silver
oxide-assisted
allylation of 104 with equimolar amount of allyl bromide afforded the 4-allyl ether 105 (40%) together with 20% of the
4,6-diallyl
ether 106(R=A11) where both products did not bear the benzoyl group at C-4 in the starting material. Alkylation of 104 with excess of alkyl
halides
in the presence
of Ag20
formed
the
4,6-dialkyl
ethers 106 in high yields (80% for R=benzyl) in which the 4-benzoyl group
was
absent
intermediate complex
again.
The
107
diol
was
ruled
out
in the reaction by some control experiments
intermediate
108
was
proposed
in
the
as
an
and a
alkylation
and
debenzoylation. In the solvolysis of 104 in methanol and pyridine at 40 "C, the benzoate function at C-2 was kept intact. In this case, participation of the axial hydroxyl was also suggested.
D<^P
Vo
J
RO" 106^^
excess RX (BnBr, AIIBr, EtI) 80o/o(R=Bn)
•,"^0
-•T^o
10
J
0 - ^ 0
AIIBr AgjO
2 BzO' OH ^^^^^ 104
40% (dially deriv: 20%)
MeOH.Py 40'C
^"°0H 105
O.v^Ph solid
HCM HOI ( OH
o - ^
107
BzO
Scheme
? 108
3-16
Selective C-0 bond cleavage in the orthoester function in 109 was achieved by the action of a Lewis acidic reagent (Scheme 3-17), Thus, treatment of 109 with DIBAL produced the 1,3-acetal high yield together with a trace of the 1,5-acetal
111 in
(ca. 20:1). The
high selectivity was rationalized by the coordination of bulky DIBAL to the less hindered 0-5 resulting in the cleavage of the linkage between the orthoester carbon and oxygen at the 5 position in 110. On the other hand, the action of trimethylaluminium
on 109 gave
selectively the 1,5-acetal 113 accompanied with the ring opened 5isopropyl ether 114 which was presumably formed from the acetal 113.
The trimethylaluminium-assisted
different
regioselective
ring opening
from the case of DIBAL may be explained by the first
415 coordination of the less bulky former reagent to 0-2 in 109 followed by activation of the oxygen at C-3 in 112 resulting in the formation of
the
113.
1,5-acetal
Both bicyclic
acetals
111 and 113
are
synthetically interesting as is illustrated by the reaction of the further benzylated 1,3-acetal 115 with TiCl4 giving rise to the 1,2acetal product 118, which was assumed to be formed via 116 and 117, as shown in Scheme 3-17. ^ ^ AIH/-BU2 AIH/-BU2 CH2CI2, r.t.
O ^
oJ
Bnojil^
BnO,
BnO
93-100%
BnO I OBn 110
BnO
MegAI '
109 MegAI CH2CI2, r.t.
o'"^!^0 ^O '
BnO.
OBn
111 H^V
OBn
OH
3 BnO
Me
B-i^'
Bno' OBn 113 84%
OBn
112
BnO OH ^ - r - O B n W--^X^OGHMeo
CI4 ^^^
^O /
OBn TiCL BnO I BnO' OBn 115
-78 'C, 30 min
OBn
BnoJW BnO
BnO.
OBn
116
4 71%
HO.
OBn OBn OBn 118
Scheme
3-17
3.4 Reaioselective Direct Functionalization of Inositols Inositols such as myo-inositol and chiro-inositol themselves are difficult
to
be
transformed
straightforwardly
into
suitably
protected forms mainly because the reactivities of the six hydroxyl groups are not so different. Their low reactivity and low solubility in organic
solvents are also the reasons
for preventing
useful
derivatization. Monoketals and diketals had been the only successful examples of the direct functionalization of the parent inositols. In
416 addition to these protected forms, some practically usable synthetic intermediates have been recently reported. The introduction of
the
TIPS group to inyo-inositol itself^^ is described in section 3.2. The
tin-mediated
strategy
was
applied
to
and
myo-
chiro-
inositol. Thus, treatment of myo-inositol with 5 equiv of dibutyltin oxide
and
bromide
allyl
bromide
furnished
in
two major
the
presence
tetraallylated
of
tetrabutylammonium
products
119 and 120.
Parallel results were obtained when crotyl bromide was used. When 3 equiv
of
the
tin oxide were used,
major
product
(Scheme
3-18).^'^
1,3,4-triallyl
Five
membered
ether
cyclic
derivatives like 123 may be formed predominantly at the site
and
then
the
more
reactive
121 was a
alkoxide
stannylene 1,2-cis-diol
reacts
with
an
electrophile, as is shown in Schezoe 3-18. ^^ OH
OH AIIO^ A ^ O A I I
OH AIIO^ A ^ O A I I
Alia' T 'oAii OH
Alio
119
aO(3eq) UNBr ^ AIIBr(7.7eq) CH3CN 17h,refl.
T "OH OAII 120
25%
26%
OH A I I O ^ A ^ ,OAII i I I ^ur/"^"CiH Alio T ^ ^ OH 121
^^Snn-Bu2
OH
BnO*'*
n-BrigSnO n-Bu4NBr
OH
BnBr, PhMe 7h,refl.
BnO^Js^OBn
^^"
l''^'''OH
B n o ' T "OH OBn
OBn 122
124
Ln-BugSnO />BU4NI
OH
HO' HO.
OH
»
Scheme An
identical
inositol
(125)
procedure and
HO'Ss^PC OPO3H2
HO 126
32%
125
HgOgPOH2O3PO '
HO'Ss^^COBn
2. BnCI CH3CN, refl.
HO
OH
BnO BnO
HO 127 3-18
was used
tribenzyl
ether
in the benzylation 126
was
isolated
of as
chiroa
major
product in 3 2 % yield along with some other products. The ether was converted to L-chiro-inositol an analogue hydroxyl
of
Ins(1,4,5)P3
(Scheme
3-18).59
2,3,5-trisphosphate
with an additional
axial
126
127 which is and
inverted
417 Borylated myo-inositol 128 was transmetallated with dibutyltinbis(acetylacetonate)
(1 equiv) to afford the butyltin derivative,
which was then treated with (-)-menthyl chloroformate
(1.3 equiv.)
and iV-methyl imidazole to give a 2:1 (70% yield) mixture of 129a and b (Scheme 3-19).^^ Recrystallization of the mixture gave optically pure 1-0-carbonate 129b in 30% yield and the next recrystallization afforded the other diastereomer 129a in an optically pure form in 2 0% yield. This procedure provides a method for the convenient and straightforward simultaneous
regioselective
optical
protection
resolution.
A
of
similar
myo-inositol
and
transmetallation
methodology was applied to alkylation [benzylation (63% yield) and allylation
(53%)]
and
methoxybenzoylation
acylation
[benzoylation
(50%)] at C-1
(43%)
of myo-inositol
regioselectivities where tributylstannyl
in
and
p-
moderate
enolate of pentane-2,4-
dione was used instead of bis-acetylacetonate.^^ The diastereomers 129b and 129a were conveniently transformed into D- and L-Ins(1,4,5)P3 as shown in Scheme 3-19.^^ 0BEt2 EtzBO. JL ^0BEt2 ^ ^ ^^^ 1. n-Bu2Sn(acac)2 EtsBO
T 0BEt2 0BEt2
Ment-0(0)CQ,
2
128
D-lns(1,4,5)P3
\
OCfi OC(0)CI
"^
129a 20%
R'^O
OBn
90%
OR'^
-132: R^=H, R^-R^=isopropylidene ^ > 1 3 3 : R'=R^=H £ 5% ^ 1 3 4 : R^=R2=(BnO)2P(0)
. /- 130: R=H ' ' ^ 131: R=Bn 91%
i: 2,2-dimethoxypropane, TsOH; ii: BnBr, AggO; iii: MeONa; iv: PTS; v: (BnOlgPNf-Prg, tetrazole then mCPBA; vi: Hg, Pd-C
Scheme
3-19
Benzoylation of myo-inositol with 3.5 equiv of benzoyl chloride in pyridine at 90 ' C for 2 h gave
1,3,4,5-tetra-O-benzoyl-myo-
(135) as the main product
(37% yield), the quantity of
inositol
which was more than that statistically expected
(Scheme 3-20).^^
When the reaction was conducted by using 2.5 equiv of the chloride, 1, 3 , 5-tri-O-benzoate
137 was conveniently
isolated by a column
chromatography even though in low (15%) yield. ^-^ The benzoates thus
418 formed
which
interesting groups
can
be
separated
from
each
other^^
intermediates
because
replacement
phosphate
functions
provides
with
of
are
these
various
quite
benzoyl inositol
polyphosphates. For example, 135 was concisely converted to racemic Ins(1,3,4,5)P4 (Scheme 3-20) 62 OH OBz
BZO4
myo-inositol
OH
HO' OBz 137
lns(1,3,4,5)P4 OBn
Pd-C
OR
quant
XEPA. tetrazde / ' 1 0 2 : ' ^ = ^ ^ ^ Q thenmCPBA \.^4Q.
R^Cl
y{0)
90%
Scheme From
the
Ins(1,4,5}P3
symmetrical
3-20
tribenzoate 137,
optically
active
and Ins(1,3,4,5)P4 were efficiently prepared in the
shortest way, as shown in Scheme 4-2. 3.5 R^qjoselegtive Protection of 1,2-Diolg The 1,2-diol derivatives of iTiyo-inositol comprise the equatorial hydroxyl group at the 1-position and the axial one at C-2, so that selective modification of the less hindered former group is easier than
that
benzylation
of
the
employ
latter.
Old
conditions
procedures using
for
a hydroxide
allylation base
and
such
as
powdered NaOH or KOH and a hydrocarbon solvent such as benzene or toluene at refluxing temperature. More regioselective alkylation is achieved by using a stannylene intermediate^lb especially in the presence of CsF. The method was applied to get 1-0-MEM derivative 142 in high yield and selectivity.'^^ The MOM group was introduced to OH-1 of 141 in a completely selective manner without CsF even at higher temperature.^^ Some examples are presented in Scheme 3-21.
419 OBn AilO^^Js^OMEM
OH AIIO^^Js^OBn
Alio*' Y^^''OBn OAII 142
BnO^' X^^''OBn OBn 143
1. n-BugSnO 2. MEM-CI, CsF 3. BnBr, NaH 1. n-BugSnO
2. AIIBr, nSu^Nl 95 'C. PhMe
62%
85%
OBn
85%
1. n-BugSnO 2. MOM^NEtg-CI, 60 'C
OH
OH AIIO^^Js^OAII
BnO^J^^OMOM
B n O * ' T "OAII OAII 144
BzO^* T ^ " ' O B n OBz 145
Scheme Various
OH
1. n-BugSnO 2. BnBr, PhMe, n-Bu4NBr, 2 h, refi.
regioselective
3-21
acylation,
silylation, ^'^^' 21ci, 66 ^^d
carbamoylation^'^ of 1,2-diol derivatives of inyo-inositol have been reported where the substituents were introduced at C-1 without
the
aid
of
the
tin
intermediate^^^ at
room
directly or
lower
temperatures because these electrophiles have enough reactivities toward alcohols
(Scheme 3-22) . Me,
r.u.n
OH
I
BnO^ J L ^OSiEtg (BnO)2P(0)0'
T OBn OP(0)(OBn)2 147
95% 'OBn
85%
EtaSiCI Py
NaOCN CF3CO2H
OH
BnO^^A^OC(0)NH2
Scheme 3-22
420 The triethylsilyl ether 147 thus formed regiospecifically from the d i d 150 [R1=(BnO)2?(0), R2=Bn], which was optically resolved by a chiral column chromatography, was transformed to Ins(2,4,5)P3 and Ins(l,4,5)P3 (Scheme 3-23}.^^ At this stage, temporary protection of OH-1 with the silyl group is not necessary, i.e. 150 can be directly phosphorylated by the phosphite-phosphonium approach as described in the section on phosphorylation (Scheme 2-6).^^ The diol 150 was used furthermore as a versatile synthetic intermediate for the synthesis of myo-inositol 1,2-cyclic-4,5-trisphosphate 152 (Scheme 3-23},{8} 2-acyl analogues of Ins(1,4,5)P3, and inositol phospholipid.
o BnO,
O
Cl2P(0)0-
(BnO)2P(0)IO (BnO)2P(0)IO
OBn
Py quant.
y Phospholipids
(BnO)2P(0)IO' T " O ^ " (BnO)2P(0)IO 151
^^'^ -HO3PO''' T^^'"OH ^"^"^"HOgPO 152
0P(0)(0Bn)2 Bn04.^i^^0SiEt3
I.PCI3 2. BnOH
^•^•^"^^^^^^ (BnO)2P(0).O^^V""OBn 84% (BnO)2P(0)IO 153 BnO, (BnO)2P(0)IO (BnO)2P(0)IO 154
0P(0)(0Bn)2
I.H2, Pd-C
OBn
2. NH4OH quant.
H2
lns(2,4,5)P3
^ci-C ^
lns(1,4,5)P3
i: 1. B2CI, DMAP, Py (97%); 2. aq. AcOH, TsOH (90%); 3. PCI3, BnOH, then /-BUO2H (85%) Scheme
3-23
As described above, the l-(or 3-) hydroxyl group of myo-inositol has higher reactivity over other equatorial hydroxyls. This tendency was observed in the following examples.^^ Phosphorylation of 155 with diphenyl phosphorochloridate produced the 1,3-diphosphate 156 predominantly in moderate yield together with 1,5-diphosphate 157 (Scheme 3-24).15,55 Benzylation of the triol 155 under PTC conditions led to the 1-benzyl ether 158 in 71% yield (overall yield from the orthoester 95 in three steps). The resultant tetrabenzyl ether 158 was again regioselectively acylated at C-3 with camphanic acid chloride for resolution (Scheme 3-24) A'^
421 OBn
OBn
HO.,A^OH Bno' T
"OBn
OH 155
(PhO)2P(o)ci
OBn
POyX^op
myX^• OF
DMAP. E\,H " BnO^ T " O B n "" BnO^^ T "OBn CH2CI2, 25 'C OH OP 58% 156 82 : 18 157 F=(PhO)2P(0)
71% BnBr,n-Bu4NI (from orthoester 95) i aq. NaOH, CH2CI2 OBn
OBn
BnO^^As^OH
(.)-camphanic acid chloride
BnO^ T " OBn OH
B n O ^ ^ A ^ O ^ ^
Py. 0 'C BnO 47 + 44% (for diastereomers)
158
Scheme
X " OBn OH
3-24
On the other hand, the sterically more hindered axial hydroxyl at the 2-position in 1 was ingeniously benzoylated without affecting OH-1
(Scheme 3-25). Thus, 1,4-di-O-benzoyl-inyo-inositol
(26) was
converted to the orthoester 160 by the regioselective reaction at the 2,3-cis-diol site in preference to the 5,6-trans
one. Subsequent
hydrolysis of the orthoester function of 160 in 80% aq. acetic acid afforded the 2-benzoylated product 161, which was transformed into racemic Ins(1,4,5)P3 as shown in Scheme 3-25.^^
PhC(0Me)3
OMe Ph-7^0 0^^^,x\^OBz
TsOH, DMF Il0-C,12h
BzO'
OH
HO^yA^OBz D,n***k^'', BZO Y OH OH
OBz aq.AcOH
110-C
HO^
J^^OBz 1. (EtO)2PCI 2. H2O2
B2O''
51%
26 OBz 1.TMS-Br
OBz
(EtO)2P(0)a
lns(1,4,5)P3 2. KOH, 60 'C
BzO"
0P(0)(0Et)2 OP(0)(OEt)2
162
Scheme
4.
3-2 5
ACCESS TO CHIRAL INOSITOLS An enantioselective reaction is a useful tool to obtain a chiral
compound.
However,
in a practical
sense, employment
of
such a
reaction in a total synthesis is limited mainly because it is not always
easy
to
get
the
product
with
high
optical
purity. A
422 b i o c a t a l y s i s system can a l s o be u t i l i z e d f o r t h i s p u r p o s e and s e v e r a l c h i r a l i n o s i t o l d e r i v a t i v e s with h i g h l y o p t i c a l p u r i t y have been a c c o r d i n g l y p r e p a r e d . In t h i s s e c t i o n , chemical asymmetric p r o c e s s e s and b i o c a t a l y s i s s y s t e m s a r e d i s c u s s e d . 4 • 1 K i n e t i c R e s o l u t i o n i n Chemical 4.1.1 Enantioselective OH BzO 1 ^OBz "^^-yX/
2 /—s MsCI, Me-N 0
HO
DMAP.THF.OC
,V-
I OH OBz
Processes
Tartarovlation OH BzO^^^A^OBz O II , R*CO OBz
/Y*"0H
'
HO^V'"OIROBz
163a
137 163a /163b
R^COgH
\y^"-|^OMe A Q ^ O H 0
OH Bz04,^A^0Bz T T ^
96:4
Yield. % chem. opt.
64
92
163b 163a /163b
R'COzH O •^y^ OMe ([^^^tvOH
l>164a
0
Yie\6, % chem. opt.
4 : 96
56
92
2 : 98
40
96
27:73
40
46
L-164b
g
<:xX
OMe OH O D-164b O
98:2
62
96
O L-164C
P
OEt OH O
(XT ""^^ 'o-Sr^H
97 : 3
42
94
0 B z O v ^ OMe
BzO^*'Sr^*^
D-164C
BzO^
L-164d
OH I ^OBz
OH
X •X' ^ . 0 'v^^s^OH
rX--°oV165
'OBz OBz 166 (83:17)
(88:12)
Scheme
4-1
When symmetrical 1,3,5-tri-O-benzoyl-inyo-inositol derived
by
benzoylation
of inyo-inositol
in
one
137, which is step,
reacts
enantioselectively at the enantiotopic C-4 or -6 position with an optically active compound, the chiral inositol derivative can be obtained essentially without losing half of one of the enantiomers due to jneso-compound. On the basis of this consideration, various chiral auxiliaries were examined and only tartaric acid monoesters 164
were
found
to
be
extraordinarily
efficient
(Scheme
4-
D.'^O Thus, treatment of the benzoate 137 with isopropylidene or
423 cyclohexylidene D-tartaric acid D-164a or D-164b in the presence of methanesulfonyl chloride and N-methylmorpholine at 0 °C furnished enantioselectively 163a accompanied with a small amount of 163b in a 96:4 or 98:2 ratio keeping the lower reactive hydroxyl group at Cacids L-164b and L - 1 6 4 C
2 intact. The reaction using L-tartaric
yielded opposite enantiomers 163b predominantly. When an acyclic dibenzoyltartrate L-164d was used, the enantioselectivity decreased dramatically. When the tartaroylation using D-164b was applied to a diastereoselective reaction of the racemic mixture, one of the pair of
165 or 166 was converted
enantiomers
in a
fairly
selective
manner to the corresponding 6- or 3-0-acylated derivative in ratios of 88:12 or 83:17 respectively. R'COgH
EtgSiCI Imidazole, DMF
163a
- ^
BzCI
137
MsCI
myo-lnositoi
Py
98% OSiEta
BzQ^JL
rco Y
HO, J L .OH
EtMgBr
0SiEt3
Et20 refl.,3h
OBz 167 74%
^Z"
po^ X ,op
XEPA
N»**\^''^ HO J OSiEtg OH 168
Tetrazole thenmCPBA 90%
MeMgBr, EtgO, refl., 2 h OSiEtg
BzO^ J L ^OH HO'
OSiEtg
0SiEt3
OBz
T OH
Hg, Pd'C
D-lns(1,3,4,5)P4
BzQ, J L ^OP
I.H2, Pd-C
2. NaOMe Tetrazole ,. ^ ^ ^ .,, 0SiEt3 thenmCPBA po'' I OSiEts quant 97% OP
170
I 'OSiEtg OP 169
quant OSiEtg
XEPA
PO'
D-Jns(1,4,5)P3
O
171
«-c(o)=o°i^°^« '-0C>'°' Scheme
4-2
The enantioselective tartaroylation of tribenzoate 137 has led to regioselective protection as well as optical resolution. The tetraacyl derivative 163a with high optical purity thus obtained was shown
to
be
a
suitable
material
Ins(1,3,4,5)P4 and D-Ins(1,4,5)P3
for
the
preparation
of
D-
(Scheme 4-2).^^ Thus, 163a was
silylated with triethylsilyl chloride to afford the disilyl ether 167 in 98% yield which became optically pure by recrystallization. Removal
of
the
four acyl groups
from 167
can
be
accomplished
successfully by the action of the ethyl Grignard reagent to give the tetrol
168
in high yield.'^^ Several
conventional
deprotecting
424 procedures for esters using nucleophiles such as ammonia, hydrazine, sodium methoxide, and DIBAL afforded various silyl group-migrated products. When the methyl Grignard reagent, which is less reactive than the ethyl one, was used the benzoyl group at C-3
in 167
remained intact resulting in the formation of 170. The tetrol 168 and the triol 170 were respectively
subjected
to a sequence of
reactions involving phosphorylation using XEPA and hydrogenolysis on Pd-C in aqueous MeOH for deprotection with accompanying removal of the silyl groups, and in the latter case additional methanolysis, giving rise to the target inositol phosphates. Both sequences are the shortest preparative
routes developed
to date
to optically
active D-Ins(1,3,4,5)P4 and D-Ins(1,4,5)P3. 4.1.2 Chiral Spiroketal Synthesis
I
jL0N>Me 173
B2O. OBz
CSA, CHCI3 refl.
70% 174 l.aq. NaOH, 96% 2. BnBr, NaH, n-Bu4NI, 74% 3. 95% TFA, 63%
177 CSA, CHCI3, refl.
BnO, OBn BzO OBz
0 2 X ^ ^ 8 " ' " " " ' ° " 91%^ 2. BnBr, NaH, ' n-Bu4NI. 87% 3. 95% TFA ,18%
176 Me
BnO OBn
D-175
H g S Z i h ^ g f " L-175
ph'^^oqf ] Me
177
^^-^
Scheme
178
4-3
Using a strategy similar to the asymmetric acylation described above (4.1.1), the symmetrical 2,5-dibenzoate 172 was transformed to the
optically
pure
enantioselectivity
by
diol the
174
in
reaction
70% of
yield one
with
pair
of
complete the
two
enantiotopic vicinal diols with the novel chiral pyranyl pyran 173 with a C2 (Scheme
axis
in the presence
of
camphorsulf onic
4-3). ^^2 This unique spiroketal, which
acid
is named
(CSA)
dispoke
425 174,
controls
regioselective
the
direction
pathways
of
the
resulting
in
enantioselective the
formation
of
and the
thermodynamically more stable adduct. This dispoke adduct 174 is anomerically
stabilized
tetrahydropyranyl
because
the oxygen
substituents
of
the
ring are disposed axially. Furthermore 174 has
equatorial methyl groups whereas the other unfavorable adduct would have
less
dispoke
stable
axial
adduct was
side
chain
converted
substituents. The
"matched"
to D-1, 2 , 5 , 6-tetra-O-benzyl-myo-
inositol D-175 as shown in Scheme 4-3. When the dipyran 177 was used instead of 173, the dispoke 176 was the
"matched" adduct,
which was converted to the opposite enantiomer L-175. The dipyran 178 was also shown to produce L-175 via an adduct similar to 176. 4.1.3 Camphor Ketal Formation Ketalization of jnyo-inositol substance,
was
carried
(1), which is also a symmetrical
out with
D-camphor
dimethyl
ketal
17 9
producing several products, from which after partial hydrolysis of the trans-ketal a diastereomeric mixture of four 1,2-cis-ketals 65a, b,
c, and d were obtained
initially
formed
(Scheme 4-4) .^^ When a mixture of the
ketalization
products
was
exposed
to
acidic
conditions in a solvent system of CHCI3, MeOH, and H2O, the major 65a
monoketal
precipitation
was
obtained
after
preferentially,
decomposition
of
resulting
the
from
trans-ketals
the and
equilibration. The latter procedure provides a practical method for the preparation of a chiral 1,2-protected jnyo-inositol which can be transformed into various optically active inositol derivatives.'^'^' ^^
OH
nor^-^ d h T ^ 1
/d^
1.H2SO4.DMSO
Meo'''''
-^z ^ n
2.TsOH.CHCl3-MeOH-H,0
179
^'^^
65a
OH 47%
65b
OH 65a
I.H2SO4, DMSO 2. partial hydrolysis
OH
^ ^ O ^ ^ ;_ ^ ,OH
&°» • ^ s : ^ - >s^r;^f^ OH
13%
65c
Scheme
OH
17%
4-4
65d
23%
426 4.2 Biocatalvsis Routes In organic synthesis, a variety of enzyme-catalyzed reactions have been examined recently with expectations of highly regio- and enantioselective reactions and functional group transformation under mild
conditions.'7^ In the field
of
inositol
chemistry,
such
selective reactions using isolated enzymes and microbes have been demonstrated providing useful chiral synthetic intermediates. 4.2.1 Enzvme-aided Enantioselective Hydrolysis Treatment
of
racemic
cyclohexylidene-myo-inositol
3 , 4-di-0-ac e t y 1 - 1 , 2 : 5 , 6 - d i - 0180 with bovine cholesterol esterase
(CE) yielded a mixture of the fully deacylated diol (-)-50a (51% chemical yield with 85% ee) and the monoacetate (+)-181 (38% yield with
86%
ee)
with
high
optical
purities
(Scheme
Enantioselective hydrolysis of racemic 4-butyrate or porcine pancreatic lipase
4-5) .^ ^
(±)-182 with CE
(PPL) was also demonstrated yielding
the diol (-)-50a and starting material (+)-182 with highly optical purities.^^^
Cholesterol esterase
O/'Y^OAc
DMF 23*C,168h
H0'V\P ( ) >—^ (-)-50a
180
o\^"oAc 51% (85% ee)
( \ N—f
38% (86% (+)-i81
o^'Sr "OH >98% ee (after recrystallization) (+)-50a
OyA^OH
182
(+)-182 CE PPL
Scheme
86%ee 95%ee
4-5
HO,. A ..0
{.)-50a 93%ee 88%ee
427
The monoacetate ( + ) - 1 8 1 was h y d r o l y z e d followed by r e c r y s t a l l i z a t i o n t o p r o v i d e t h e d i o l (+)-50a w i t h o p t i c a l p u r i t y g r e a t e r t h a n 98%. The o p t i c a l p u r i t y of an i n o s i t o l d e r i v a t i v e contaminated with the other enantiomer often i n c r e a s e s by r e c r y s t a l l i z a t i o n . The d i o l (+)-50a was r e g i o s e l e c t i v e l y b e n z y l a t e d v i a a s t a n n y l e n e approach, followed by a c e t y l a t i o n t o p r e v e n t t h e m i g r a t i o n of t h e c i s - k e t a l during h y d r o l y s i s of t h e t r a n s - k e t a l i n t h e next s t e p . The f u l l y p r o t e c t e d d e r i v a t i v e 183 t h u s formed was c o n v e r t e d t o I n s ( 1 , 4 , 5 ) P 3 a c c o r d i n g t o t h e p r o c e d u r e s shown i n Scheme 4 - 6 . 3 4 a , 7 4
n-BugSnO Q O ^ Y ' ^ O H
W
BnBr.CsF
(+)-50a
OAc
AC2O, EtgN
O: T \_J
'OBn
"^"'
Q T OBn
t55 I.AcCl, MeOH.CHgCIs 72% 1 2 . KOH/MeOH
I.Pd-C.Hg
lns(1,4,5)P3
(BnO)2P-N/-Pr2
2. AcOH
o'\x^^ tetrazole R 0 T OBn thenmCPBA OR 97%
98%
, ^,^ . HO \ OBn OH
184
185 R=P(0)(0Bn)2
Scheme 4-6 The antipode diol
(-)-50a was used for the synthesis of D-
Ins(1,3,4,5)P4 and D-Ins(1,3,4)P3 which were obtained in multigram quantities according to conventional and convenient sequences shown in Scheme
4-7. 34a,39
-^
novel
phosphatidyl inos itol
3,4,5-
(-)-50a
(Scheme
trisphosphate was also synthesized starting from 6-3) .40 Porcine
liver
esterase
(PLE)
effected
enantioselective
transformation of the symmetric dibutyrate 196 of the orthoester 95 to the monoester 197 in 83% yield with >95% ee (Scheme 4-8). ^'^^ A diastereomeric chemically
pair
by
of chiral
resolution
intermediates
of
the
were
also
corresponding
obtained
iV- ( i^) - 1-
phenylethylcarbamates 198. These chiral orthoester derivatives were converted which
to D- and L-2,4-di-O-benzyl-jnyo-inositol
in turn,
were
transformed
D- and L-102
to D- and L-Ins ( 1 , 3 , 4 , 5 ) P4
428 respectively. Racemic Ins(1,3,4,5)P4 as well as the optically active compound described here were synthesized by using the orthoester strategy.
However,
sequence (Scheme
compared
the chiral with
synthesis
the synthesis
required
a much
of the racemic
longer
material
3-15) . OBn ^^yX^O^
(-)-50a
OBn AIIQ^r^OAII
Viii
D-lns(1,3,4,5)P4 HO^^ I " OBn OH
^^'
AIIO*^ T '' OBn OAII
D-102
189
75% I
'% |\ 97% I Vii OH _ .
AiiQ^o
... ^ Aiio^y 0
i,v ^ AiioyyoH
••^x/* Alio I OH OH 186
80%
91%
ok-/% AIIO T OBn OAII 188
^.
OH
^AIIO^X^OAII
95%
.»\/s AIIO T OBn OAII D-144
95' ...^ O H ^ , , ^ „ ^ OH _ . „ ^ ^ OBn ^ ^ A I I O y T OH y, ^ A I I O y T O A I I ^ ^ R O y y O R
O-^ AIIO^O
Alio T OBn OBn
Alio
190
» OBn OBn
AIIO T OBn OBn
191
^ ^
RO T OBn OBn
192
Viii r ' ' 9 3 " '^=AII " V l 9 4 : R=H 79% '^ *<195: R=(BnO)2P(0)
i: AIIBr, NaH, DMF; ii: CH3COCI, CHgOH-CHgClg; iii: n-BugSnO, BnBr, CsF; IV: BnBr, NaH, DMF; V: CH3COCI, CH3OH-CH2CI2; Vi: n-BugSnO. AIIBr, CsF; Vii: BnBr, NaH, DMF; Viii: Pd-C, TsOH, CH3OH-H2O; iX: (BnO)2P-N/-Pr2, tetrazole then mCPBA; X: Hg, Pd-C, 60% EtOH.
Scheme
•^o
J
"•'^''""ioCn-Pr 196
p^
4-7
LO
J
83%095o.ee) "'^^"''l, 197
Scheme
LP '"°
4-8
4.2.2 Enzym^-aidec^ Enaritips^leptive Acyl^tion Enzymatic
esterification
in an organic
solvent
which
is a
reverse process of the hydrolysis described above, were examined and 1,2:5,6-,
50a, 1,2:3,4-di-O-cyclohexylidene-inyo-inositol
1,2-cyclohexylidene-inyo-inositol
48a and
47a, which are synthetically useful
intermediates but difficult to obtain in optically active forms, have been separated to each enantiomer respectively. Acetylation of racemic
50a with acetic
anhydride
catalyzed with
a lipase
from
429 sp. (Lipase P) afforded the optically pure 3-0-acylated
Pseudomonas
product 199 in 25% chemical yield regioselectively together with the L-50a-rich Candida
starting material
(Scheme 4-9). "^^ When a lipase from
(Lipase AY) was applied to the same substrate
cyclindracea
(±)-50a with phenyl and p-nitrophenyl anhydride,
the
4-0-acetylated
acetate as well as acetic
product
(-)-181 with
100% ee,
contrary to the case of Lipase P, was enantio- and regioselectively obtained accompanied with an almost optically pure diol L-50a. The racemic
1,2:3,4-dicyclohexylidene
derivative
48a
can be also
efficiently resolved by enantio- and regioselective acylations using Lipase AY. Experiments concerning solvent effects on this reaction showed that relatively less polar solvents such as ethyl ether and benzene acetate,
gave better
results
than polar
solvents
such as ethyl
acetone, THF, and dioxane. These water-soluble
ethers did not allow the reaction to proceed.
(±)-50a Lipase Lipase Lipase Lipase
P AY AY AY
25%(100%ee) AC2O AC2O AcOPh AcO/>N02Ph
AC2O (C2H5CO)20 (C5HiiCO)20
L-50a
(-H81
199
48%(100%ee) 38%(100%ee) 40%(100%ee)
OH OAc 200 50% (96% ee) 54% (84% ee) 50% (92% ee)
73% (36% ee) 51%(98%ee) 53% (92% ee) 55% (84% ee)
HO' OH (+)-48a 48%(100%ee) 37%(100%ee) 43%(100%ee)
Lipase CES AcgO, Dioxane OH (+)-201 49% (98% ee)
Scheme
4-9
OH (+)-47a
49%(100%ee)
cyclic
430 Contrary
to the above
results, dioxane
effective on a Lipase CES (from Pseudomonas
as
a
solvent
was
sp.)-catalyzed selective
acetylation of 1,2-cyclohexylidene-inyo-inositol 47a; furnishing the 3-acetate
(+)-201 and L-1,2-protected starting material (+)-47a in
theoretically
quantitative
selectivities
and
chemical
yields
(Scheme 4-9).^^ Lipase P gave a comparable result to that in the case of Lipase CES. It is interesting to note that the Lipase CES recovered by filtration after the reaction lost most of the activity but the reusable enzyme can be obtained by treating with water and drying, indicating the necessity of hydration of the enzyme for maintaining the activity. Both enantiomers (+)- and (-)-47a were employed to synthesize D-Ins(l,4,5)P3
(Scheme 3-6),^^ D- and L-Ins(1,4,6)P3
(Ref.48), D-
and L-Ins(l,4,5,6)P4,'76 and PI(3,4,5)P3 (Scheme 6-2).^7 A lipoprotein lipase from Pseudomonas
sp. effected not only
regioselective acetylation of the orthoester 95 without
observing
the acylation of the axial hydroxyl groups but also enantioselective butyroylation (>95% ee) of symmetrical 4,6-dibenzoate 203 which was derived from the acetate 202 by benzoylation and acidic methanolysis (Scheme
4-10) .^8 Regioselective
chemical
acylation
of
the
equatorial hydroxyl group in 95 was also reported using benzoyl (64% yield) and p-nitrobenzoyl chlorides (51%).^9
o
AcO"^
HO.
HOI 95
I OH
Lipoprotein lipase (from Pseudomonas sp.)
H O Il OH c 202 I.BzCi.Py 2. HCi, MeOH
OH HO^ ^ I ^
BzO
i OH 204
^02Cn-Pr
OBz
AvPrCOo^"^ Lipoprotein lipase (fronn Pseudomonas sp.)
(>95% ee)
Scheme
4-10
4 . 2 . 3 Microbial Oxidation of Arenes Mutant o x i d a t i o n {Pseudomonas putida) of a r e n e s 205 produces l , 2 - c i s - d i h y d r o x y c y c l o h e x a - 3 , 5 - d i e n e s 206. The chemical s y n t h e s i s of such n o v e l compounds i s v e r y d i f f i c u l t . Furthermore, when
431 monosubstituted benzenes such as toluene, fluoro-, chloro-, bromo-, and
iodobenzene
are used,
the c ojTiresponding' almost
cis-diols 206 can be obtained
optically
pure
(Scheme 4-11). These advantages
have
promoted their use as synthetic starting materials in a large number of total syntheses of various complex molecules. The diol derived
from benzene has been used
inositol
phosphates
by Ley's
206{R=H)
for the synthesis of some m y o (Scheme
group
5-3 and 5 - 4 ) . In this
section, the transformation of chiral diols is treated.
6
Pseudomonas putida
R=H, Me, F, CI, Br, I
205
Scheme
X 207
206
4-11
I.OSO4, MNO 85% 2. LiAIH4 85%
86% 86%
HO' > r ^ ^ ^ 0 OH
HO' 0 H O ' ^^^ ^ ^ ^ ^^^ O OH
208
209
mCPBA
,6: Br
O ' ^
1.MeOH,Al203 90% 2. HCI. H2O LMeOH.AIgOa 89%
MeO^S^O
2..LiAIH4
2. HCI, H2O
HO^' Y ' ^ O H OH (-)-210 (+)-210
OH 212
211
OH
t^o-
HO,, A^^OH
'••rr' &^^
OH
HO^^^Y^^OH OH (-)-213
(+)-213
Scheme Homochiral
OH MeO^ X ^OH
I.OSO4, NMO 63%
4-12
bromo-isopropylidene
207
ketal
derived
from
206(R=Br) has two double bonds, among which the olefin at C-3 is the more
electron
rich.
Therefore,
oxygenation
reactions
occur
preferably at C-3 and C-4 rather than at C-5 and C-6, as supported by calculation.SO Thus, osmylation and epoxidation of 207 proceeded in complete regio- and stereoselective manners to give the diol 208 and They
the epoxide were
then
211 in high converted
yields to
respectively
( + )-
and
(Scheme
(-)-pinitol
4-12) . 210
via
stereoselective oxygenation according to the procedures shown in the
432 scheme. By similar approaches, chloro- and bromo-cis-diols 206 (R=C1 and
Br)
were
transformed
into
both
enantiomeric
conduritol
E
epoxides (-)- and ( + )-213 respectively. ^^ Three stereoisomeric
inositols were prepared from the highly
functionalized derivative 215 which can be derived from 214 in one step by the reaction with KMn04 (Scheme 4-13).^2 Treatment of 215 with AI2O3
in an aqueous medium afforded
ketoalcohol
216 in 85% yield, which was converted efficiently to
ailo-inositol
stereoselectively
the
(217) essentially as a single product. The epoxy diol
215 was converted under basic conditions to D-chiro-inositol
(125)
with more than 95% selectivity by attack of the hydroxide ion on one side of the epoxide carbon atom while its treatment under acidic conditions furnished neo-inositol
(219) as a minor product along
with 125 (3:7) resulting from the attack of H2O from the other side. CI
KMn04
AI2O3 H2O, 80'C
HOvJC^OH
H2
HO.,Jss^OH
H O ' S ^ OH "^^^
H O * ^ ^ ^ OH
OH
OH
216
217
OH HO^ > ^ ^ Q ^ r / \ HOs K> ^ ^0 OH
HO^ " T "OH OH 219
{neolchiroZ-1)
OH HO^ X ^OH resin(H*
un^^k^ HO V OH OH D-125
218 Scheme
4-13
4.2.4 Employment of Metabolic Enzymes In
a
living
system,
the
transformation
of
a
substance
is
catalyzed by its recognized enzyme. Therefore, all natural compounds and
their
analogues
can,
in principle, be prepared
by
a
sole
enzymatic reaction or by consecutive enzymatic reactions along the metabolic pathway in vitro
or in
vivo.
One of the metabolic enzymes for Ins(1,4,5)P3, Ins(1,4,5)P3 3kinase
isolated
from
rat
brain
cortex
was
Ins(1,3,4,5)P4 from Ins(1,4,5)P3.83 j^yo-Inositol directly without protection to fluorescent
used
to
prepare
(1) was converted
1-phosphatidylinositol
433 analogues by reaction with the corresponding cytidine diphosphate diacylglycerols 220 in the presence of PI synthase (rat liver microsomes used as the source of its activity.) (Scheme 4-14). ^"^ These analogues were used for studying the metabolism and intracellular transport of these lipids in living cells.^^ NH2
OH
0
N
•0-P-O-
RCO2 R'COa
PI synthase
NO2
n=5,11
Scheme Provided
that
(rat liver microsomes)
L-o-P=0 )—( 6" HO OH
OH
a reasonable
4-14 amount
of
the enzyme(s)
can be
obtained, the enzymatic preparation of the desired molecules would be practically substantial
realized.
In order
quantities, an over
to obtain
expression
a pure
of
enzyme
in
this protein
is
required. For this purpose, the purified protein is necessary but its purification is not always easy, at present. 4.3 Ferrier Reaction In the biogenesis of myo-inositol, D-jnyo-inositol 3-phosphate is known to be derived from glucose 6-phosphate by the action of myoinositol 3-phosphate
synthase. This transformation bears a close
resemblance to the Ferrier reaction.°" This biomimetic sequence is suitable
for the preparation
of some optically
derivatives. An exomethylenetetrahydropyrane
active
inositol
compound 222 derived
from D-glucose are transformed to the corresponding oxocarbocycles 223 by a Hg2+-assisted to
aldol-type
reaction
in an
(Scheme 4-15). The resultant hydroxyketone 223 was
aqueous solvent dehydrated
intramolecular
afford
the
enone
224
which was
in turn
reduced
stereoselectively with NaBH4 and CeCl3 yielding 225. The conduritol derivative
225
has
been
shown
to
be
a
versatile
synthetic
intermediate. Thus, the benzylation product 226 of 225 has a C2 axis, therefore its osmylation, which is conducted by the approach of OSO4
from both sides of the double bond, produced
a single
product. ^'^ The tin-mediated regioselective introduction of the MOM
434 group to the diol with the inyo-configuration followed by benzylation and hydrolysis furnished the triol 9 which is the phosphorylation substrate laminitol
leading
to Ins(1,4,5)P3 . In a similar
(229) and mytilitol
manner, {-)-
(230) were synthesized. ^'^^ The same
type of compounds as 226 with a C2 axis is derived from D-6,3glucronolactone, and osmylation and epoxidation on the double bond were similarly demonstrated as shown in Scheme 5-5.^^
R2O\...--VA R'O
acetone-H20 ^'^l^y^ ^e«-
n OM M e.
222
R 2 O \ ^ - Y ^ O H —^ 77% "''
"R^O ^
l.OsO„NMO
BnO Q ^
R^O 224
Bno
MOM-CI.EtaN, r.t.
2 2 7 : R=H 2 2 8 : R=MOM 1. BnBr, NaH 2. aq. HCI
R^oX.^
223
(Ri=Bn. R 2 = M 0 M )
BnO
^
' 100% '"^"'
^"0 226
^Z"
,
,
1
NaBH4 CeCl3
OJPK^
R^O 89% (2 steps)
225
96% 90%
BnO
HO
HO
HO
OH^ ^"^isn 9
Me ^ ^ i H 229
Me ^ ^ 230
Scheme 4 - 1 5 Glucosaminyl-chiro-inositol phosphate 237 and its myo-homologue 241
have
been
proposed
as partial
structures
in unidentified
substances acting as second messengers of insulin action. These substances are postulated to be released resulting from hydrolysis of glycosyl phosphatidylinositols bound to the plasma membrane after the binding of insulin to target cells. ^^ Both compounds 237 and 241 were synthesized from the Ferrier product (Scheme 4-16).^^ The tribenzyl
ether
225 was subjected
to a hydroxyl
group-assisted
stereoselective epoxidation with mCPBA and its regioselective ring opening with
allyl alcohol
in the presence of borontrifluoride
etherate gave a chiro-inositol 232 in good yield, which was then benzylated at the equatorial site via the stannylene intermediate. The remaining axial hydroxyl of the product 233 was glycosylated by the Schmidt's method using glycosyl trichloracetimidate and TMS-OTf giving the a-glycosyl-chiro-inositol 234 and subsequently, according to the sequences shown in the scheme, 237 was obtained. The myo-
435
i n o s i t o l homologue 241 was prepared by using s i m i l a r p r o c e d u r e s . For t h e t r a n s f o r m a t i o n of t h e c h i r o - f o r m 238 t o t h e myo-one 239, t h e Mitsunobu r e a c t i o n , i n v e r s i o n of t r i f l a t e with b e n z o a t e , and some o t h e r a t t e m p t s were not s a t i s f a c t o r y but t h e o x i d a t i o n - r e d u c t i o n sequence proved s u c c e s s f u l . HO
mCPBA
HO
BnO,\,...^-^
BnOi^----^
BnO
93%
.„ ^ u
HO
BP
BnO
225
OH i
BnoX-'-vA 75%
BnO ^ ^ j j
231
232 n-Bu2SnO, A>-Bu4NBr
.OAc
BnO
^O
^ ^
^S^-^°'^^^>^^'^
BnO^^^i^
55/0 TMS-OTf
BnO r BnOj^,,
^p^^
OH
. r 234: R=AII > 235: R=H " W 236: R=(BnO)2P(0) ^•NH3
2. Hg. Pd-C rOH HO-^jA^^ ^O H V I uo ^ 0 H O - ^ ^ "^OPOaH237
OAII j. [ir(COD)(Ph2MeP)2]PF6, Hg then Ig, THF. HgO • ^38 ii: 1. (BnO)2PNAPr2, tetrazole, 2. RUCI3. Nal04. 70% 1.PCC 2. (S)-Alpine hydride 580/, -90'C ^^^^ 1. (BnO)2PNAPr2 ^ J^^iJPv ^ Setraiole BnO-^Y^-^QH 2. RUCI3, ^^\Q^^^ ^ " O ^ ' R O I MPMO ^.^.--^ ^"^OAII BnO'T"'^-'-^ 83% 239 BnO-W^OP(0)(OBn), 240 \ . ^ O H 1.CF3C02H,84%\^ H O - ^ T - ^ O 2. glycosylation. 65% HO-^^--^^ r^u 3NH3 "^ HoNlJdO. ? i i 4. deallylation ^ 0-^>^--r\OH 5. Hg, Pd-C 28% (overall) "HOaPO A-.-^-^^-V-OH 241
Scheme An
acetoxy-substituted
4-16
exomethylene
derivative
242 is a
promising substrate for the Ferrier reaction since the resultant product has all the oxygen functionalities of the inositol skeleton. Thus, vinyl acetate 242, which is derived from the corresponding protected glucopyranoses by oxidation and 0-acetylation, was treated with mercuric trifluoroacetate in aqueous acetone at 0 °C to form the oxymercuration intermediates 243 which cyclized by addition of the chloride
ion to give a mixture of diastereomers
(Scheme 4-
17).^^ The Ferrier reaction recorded fairly good stereoselectivities and
compound
245,
having
the
desired
configuration,
was
predominantly formed. Bender and R. J. Budhu found furthermore that
436 the cyclization of the silylated organomercurial intermediate 242c was promoted by various Lewis acids, among which extremely superior selectivity
SnCl4
showed
(Table 1) . In the case of the silyl
ether 242, two other products 247c and 248c were formed along with 245c and 246c, and interestingly the NMR analysis showed that both predominantly adopt the conformation in which the three silyloxy groups and the hydroxyl group are axial. The major product 245 was then reduced in a completely stereoselective manner to generate an equatorial hydroxyl group resulting in the formation of the myoinositol derivative 249. OAc
RO
Hgs^OAc XHg
Hg(02CCF3)2 MeCOMe/H20(4:1) OMe 0 "C, 10 min
242
NaCI
RO RO. o\^..<^ RO 243
25*C,20h
Z/E. 95:5 - 97:3
Q
EtgSiO X O ' lAcO?
0, ^°0H 245
^^ 246
EtoSiO EtaSiO
Yield, % of 245 245:246:247:248 242a 242b 242c
RO-V^^OAc RO OH
BaSiO I
V
247c: X=OH, Y=H 248c: X=H, Y=OH
NaBH(0Ac)3 AcOH, CH3CN HO
0 OAc EtaSiO-^^^
/
81:19: 0: 0 85:15: 0: 0 64:48:25:3
57 59 50
a: R=Me, b: R=Bn, c: R=Et3Si
249
Scheme
4-17
Table 1. Lewis acid -promoted Ferrier reaction of 242c® Lewis acid EtaAICI SnCU SnCU^ TiCU a-Br-9-BBN ZnCl2 BF3»OEt2
245c:246c:247c:248c
Combined yield, %
33:<1:62:5 87:ca.1:10:2 67:ca.1:12:11 24:4:66:6 87:<2:10:3 62:<2:37:1 no reaction
f Lewis acid=2 equiv., CH2CI2, -78 °C. ^ (E)-242c used. ^ Yield not determined.
83 83 c 79 52 C 0
^ '
J
437 Estevez and Prestwich converted the diol 249(R=MPM), obtained according to the same sequence, into a photoaffinity probe 251 with an azido function for Ins (1, 3 , 4 , 5 ) P4 (Scheme 4-18) .^^ rpj-^^ aminopropyl ester obtained after the hydrogenolysis procedure was also immobilized to a resin for affinity chromatography. BOMO
^
9
HO
250
^ ^
? 9^
251
i: BOM-CI, />Bu4NBr, Proton sponge, 69%; ii: NaOH, MeOH, 95%;iii: (BnO)[CbzNH(CH2)30]PNAPr2, tetrazole then mCPBA, 85%; iv: DDQ, 78%;v: (BnO)2PN/-Pr2, tetrazole then mCPBA, 73%; vi: Hg, Pd-C, quant.;vii: 4-aziclosaiicylic acid A/-hydroxysuccinimido ester
Scheme
4-18
For the synthesis of glycosyl inositols, the disaccharides 252 were chosen and transformed to the target molecules based on the Ferrier reaction as shown in Scheme 4-19.^-^ The procedures before and after the Ferrier reaction were identical with those described above and the allyl alcohol 254 was thus obtained. The selectivity of osmylation on 254 was improved from 2;1
(256/255) to 5.6:1 by
using its acetate 254(R=Ac) as the substrate. BnO ^OBn UvS.^0
BnoXZ^rA
l.Nal
v-oTs
BnO o - T < ^
n-Bu,Ni
9,
2. DBU
BnOA-^-T-^
252
GAL
i
o
OR
B n O \ - - r A BnO^^ OH 256
60%
GAL "1
'7;:^-T^^v--q O'
BnO-^^^^"^ 253
^ " ° OMe
GAL .w
1.Hg(02CCF3)2 r.t., 12h,92% 2. MsCI. Py.81% 3. NaBH4, CeCls •78 'C, 87%
BnO ^ ^ ^
GAL OR
9*^
BnoV--A.OH BnO 255
Scheme
" ^ ~ ^•^920/0 (for R=Ac)
OR
B n O X . ^ aln °^^ 254
4-19
Contrary to the results described above
(Scheme 4-19), highly
diastereoselective osmylation of cyclohexene derivatives 258 with the more bulky substituent, TBDMS, on one side of the double bond, which were derived
from oxanorbornenone 257, was
reported. This
afforded the suitably protected chiral myo-inositol derivatives 259 (Scheme
4-20).^^
438
^:^^Okc
Os04,Et3NO aq. acetone r.t.,24h
60% (diastereoselectivity=91:9) 88% (diastereoselectivity=92:8) R=TBDMS ^ Ph
O T^*OTBDMS OTBDMS
Scheme 4 - 2 0
5. NUCLEOPHILIC SUBSTITUTION Nucleophilic
substitution
with
inversion
or retention of
configuration of an inositol ring carbon is used to obtain inositol derivatives stereoselectively with the desired configuration. The reaction has been often employed when myo-inositol derivatives are derived from other starting materials such as natural products and arenes. 5•1 Transformation of Alcohols to Halides ^ ? MeO yj^a^T- OH y>r—/-OH f Ql_j HO
DAST \ 20 C, 45min 57%
260
MeO p . ^ - . - , J > 7 ^ OH
"HO r MMeO eO _ •OH F . . / ' - ' ^ ' " " - ^ OH OH
I OH HO
•I
261
262
BBr, P OH
OH 263 OH
C OH
OH 265
OH 264
9 O2CC15H31 0 - P - 0 > ^ x ^ O2CC15H31 OH
OMe H? MeO /-—<^0H
L^
I OH HO
^OH
260
^'^^
HO
^>
H* 84-97%
PhaP, CCI4
^
90 'C, 6 h 34%
Scheme 5 - 1
"^
C l j l - r - ^ O H OH 266
439 L-Quebrachitol the rubber
260, which can be obtained from an exudate of
tree, is a useful chiral
starting material
for the
preparation of inyo-inositols as well as chiro-inositols. Optically active 3-deoxy-3-fluoro-D-myo-inositol
263 can be very efficiently
obtained in a two-step procedure which involves the reaction of 260 with DAST giving two isomers 261 and 262, and subsequent treatment with BBr3 (Scheme 5-1). ^^ The fluoride 263 which is an inhibitor of the growth of transformed fibroblasts^^ is converted to analogues of
Ins(l,4,5)P3
(Ref.97)
phosphatidylinositol,^^ 264
and
and
265. The chloro homologue 266 was also prepared from L-quebrachitol by inversion of the configuration with the chloride ion (Scheme 51).99 5.2 Use of Trifluoromethanesulfonic Esters o R
R
'T^o "
R
R
"T^o R'=H, camphanate
"
R
"-^o "
267
49
R 268
Nu: BzOLi, AcSK, n-Bu4NF. NaNg, nBu4NN03, (BnO)2P(0)OLi
269
270
271
LiN3, r.t., 9.75 h: 48% LiCI, Imidazole, r.t., 12 h: 36% (olefin: 45%) KSAc, r.t., 14 h: 51% (olefin: 8.6%) BzOCs, 18-crown-6, 100 'C, 10 h: 56% (olefin: 41%) BzOK, 18-crown-6, 80 'C, 4 h: 18% (olefin: 85%)
Scheme
5-2
Recently trifluoromethanesulfonic esters of inositols have been employed in the SN2-type substitution reaction, among which several reports are described here. In order to prepare D-chiro-inositol 125 and
its derivatives,
the
former
of which was
identified
as a
constituent of a putative insulin mediator for rat liver, readily available jnyo-inositol triflates 267 were converted to chiro-forms 268 by the S-^2 reaction with various nucleophiles
(Scheme 5-2).^^
On the other hand, the triflates 269 derived from quebrachitol 260
440 with chiro-configuration were transformed to myo-type derivatives 270 by reaction with some nucleophiles in order to prepare optically active 3-modified substrates
or
myo-inositol analogues which are evaluated as (Scheme
inhibitors
reactions were
accompanied
produces
271.
like
The
5-2).^^^
with varying
reaction
In these cases, the
amounts
using
cesium
of
elimination
benzoate
as
a
nucleophile in the presence of 18-crown-6 furnished the substitution product 270 predominantly whereas the potassium salt afforded the cyclohexene derivative 271 as the major product. This sequence was used for the synthesis of D-myo-inositol 1-phosphate.^^^ Preparation of D- and L-Ins(1,4,5)P3 using D-chiro- (derived from D-pinitol) and L-chiro-inositol 125 (from L-quebrachitol) as starting materials was reported where triflates were also used for transformation of the chiro-configuration to the myo one.^^2 5.3 Use of Epoxides Epoxides
are
good
substrates
for
functionalization
and
epimerization by the ring opening with nucleophiles. Ley's reports are described here in some detail.-^^-^ Their inositol chemistry is based
on
microbial
oxidation
dihydroxy-3,5-cyclohexadiene pivotal
synthetic
of
benzene
leading
to
1,2-cis-
206(R=H) which may be used as the
intermediate
for
the preparation
of
various
inositols and related substances. The cyclic carbonate 272 derived from the diol 206 was epoxidized stereoselectively using mCPBA to give the a-epoxides 273a and its p-isomer in a 4.6:1 ratio (Scheme 5«3)^104 rpj^g regioselective ring opening of the major isomer 273a was conducted with a chiral nucleophile, (+)-1-phenylethanol in the presence of HBF4 affording the 1:1 diastereomeric mixture of 273 in 67% yield. Separation and conversion led to D- and L-Ins(1,4,5)P3, respectively
as
stereoselectively
shown with
in
Scheme
mCPBA
5-3. Diol 275
with
assistance
of
was
oxidized
the
hydroxyl
group(s) followed by acetonization to afford the ^-epoxide 276 in 87% yield together with 5% of the a isomer. The ring opening of 276 with an alcohol proceeded with moderate regioselectivity to furnish 277 with a new protecting group. On the other hand, the epoxide 276 was
transformed
to
a
putative
insulin
mimic
287.^^5
Thus,
regioselective reductive ring opening with LiAlH4 yielded 279 in 81% yield together with 12% of the regioisomer. Oxidation of 279 with tetrapropylammonium perruthenate
(TPAP) followed by treatment with
TBDMS-Cl gave the silyl enol ether 280 which was then reigo- and
441 O.,.
a : -^ a:>" ^ a:>
205
(2 steps)
272
206
273a
^-^-UoJ^)
Ri^OH I
OBn
I
67%
OR
OR
r
, 0 * ^ , x ^ v ^ 0 ^ Ph vii,87% (a-isomer: 5%)
"^^ 'OH 275
<
viii, 89%
273a: R=H 274: R=Bn 100%
273b
OBn
OBn
'"^ V r \ £ ^
0
0.y.^OH
^
1
OBn
^•^Y'^^o .
I
(2 steps)
279
OBn won
»v-
HO' Y
88%
D-lns(1,4,5)P3
T^
O
OTBDMS
X
OBn
Ph^^O.y'^-iO,
P-V^oX
85% OBn
.X.,
Ph^^O^Y^^^O. RO^'^V^O
OR
BnO
BnO'*' Y
" '^'
OBn 287
XII
281
78%
OH
278
--^ ™ TX°x HO*'
,,,
NH?
'4^
0P(0)(0Bn)2
/
OBn
I
OTBDMS
OH
H O - ^ ^ O H I ^ ^ ^ OH OPO3H2
c.
0
280
P*''
:6c°x
1
277
^iv Ph-^o^^A^o
OH
H O ' Y ^ '
\^
n
^ -
xiii 81% (isomer: 12%)
Ph-^Oi
r
0P(0)(0Bn)2 (BnO)2(0)POX, 100% xi, 67%
58% (isomer; 37%)
276
i
P.JL. h^Oi^y^-X^O
X
OMPM
c
80%
282: R=TBDMS 283: R=H
(f 284: R=MPM, 59% (p: 2A°A > 2 8 5 : R=H, 76% W 286: R=(BnO)2P(0), 72%
i: Pseudomonas putida, ii: (MeO)2CO, MeONa; iil: mCPBA; iv: HBF4; v: BnBr, AggO; vi: EtsN, MeOH; vii: mCPBA: vlii: 2,2-dimethoxypropane, CSA; ix: NaH, 95 *C, 44 h; x: Hg, Pd-C; xi: LDA, TBPP; xii: Hg, Pd-C then aq. TFA; xiii: UAIH4, r.t.; xiv: tetrapropylammonium perruthenate, A/-mehtylmorpholine A/-oxide ttien TBDMS-CI. EtgN, 65 'C, 2 days; xv: BHg'THF, r.t., then f-BuOgH; xvi: MPM-CI, NaH, 80 'C, 8 h; xvii: tetrabutylammonium fluoride; xviii: TMS-OTf, -20 *C; xix: DDQ; xx: (BnO)2PNA-Pr2, tetrazole then mCPBA; Hg, Pd-C, then aq. AcOH, 50 'C
Scheme 5-3 stereoselectively hydrated by the usual hydroboration and oxidation procedure yielding 281 in 76%. The inyo-configurational skeleton 281 thus prepared
bears
a variety
of protecting
groups
and
it
is
therefore a promising synthetic intermediate for the preparation of various inositol derivatives. The free hydroxyl group at C-6 in 281
442 can be glycosylated, but in the project it was treated with MPM-Cl and NaH at 80 "C resulting first in the migration of the silyl group and subsequently p-methoxybenzylation at C-1. Since the hydroxyl group at C-1 usually reacts with an electrophile in preferance to other hydroxyls, a sluggish reaction at C-6 promotes the migration of the silyl group to OH-6 and the resulting free hydroxyl of C-1 reacts
favorably with
the chloride. These consecutive
resemble the migration
of
the benzoyl
group
and
reactions subsequent
phosphorylation in the TIPS-protected derivative 77 (Scheme 3-10) . The fully protected 282 was transformed to the target molecule 287 according to the procedures shown in the scheme (Scheme 5-3). The epoxy carbonate 2 7 3a can be transformed to the key intermediate 289 by a sequence similar to that mentioned above which includes a regioselective ring opening with benzyl alcohol leading to 288 and stereoselective epoxidation convert
(Scheme 5-4) . In order to
the epoxide 289 to a variety of Ins(1,4,5)P3
analogues
modified at C-6, it was treated with some nucleophiles to afford the regioselectively ring opened products 290-293. They were converted to
the final
products
by a
sequence
of reactions
involving
debenzylation, phosphorylation, and deprotection (Scheme 5-4).^^^ OH
OBn
BnO-
XXh
OBn
BnO*Y^\^
0,y
MeONa, MeOH
BnO«»^^^^V^ O MeO
289
288 LiAIH4 EtgO ^efl., 2 h
I
BnO^Y'-^^O
V-o^ OH
291
OH 290
DAST THF refl., 4 d
Me2Cu(CN)Li2 ^^THF, -30 'C. 22 h
OBn
OBn
.
refl., 72 h
74% (regioisomer: 21%)
OBn
BnO-Y'-^-^O
.
BnO^Y''^^ O , .
.V-o>^ „Mo>< OH
OH
292
293
76%
59%
(regioisomer: 12%)
(regioisomer: 15%)
Scheme
73% (regi
5-4
The functionalities of the six carbon atoms in D-glucurono-6,3lactone 294, which is commercially available, are properly protected except for the two hydroxyls at C-2 and -5. Epimerization at the C-5 carbon atom and cyclization between the terminal C-1 and C-6 carbons will lead the carbocycle with the myo-configuration.^8 Thus, 294 was
443
^303: R^=Tr, R^=H ^ 304: R^=Tr. R2=Bn or All ^305: R^=H, R2=Bn or All
OR
OR
307, 33 (myo) R=AII: R=Bn:
30% 25%
OR
308 (chiro)
309 {scyllo)
27% 25%
9% 11%
i: acetone, TsOH [64%]; ii: TsCI, Py [88%]; iii: DIBAL [88%]; iv: K2CO3, MeOH [87%], v: MeOH, HCI [00%]; vi: NaH, BnCI [73%]; vii: aq. H2SO4 [91%]; vlii: NaBH4 [75%]; Ix: TrCI. Py [80%]; x: NaH, AIIBr [93%] or BnCI, xi: aq. HCI [50% for 305 (R=Bn) from 303, 91% for 305 {R=AII)]; xli: (C0CI)2, DMSO, EtgN; xlii: TICI4, Zn(Cu)
Schezne 5-5 converted
to the substrate
297
for the epimerization
in Scheme 5-5. The tosylate exists as the cis
and trans
as
shown
forms, 296
and 297, which are readily interconvertible under basic conditions, therefore both isomers are converted with epimerization via the trans
isomer 297
to the epoxide 298, which
was
subjected
to
methanolysis to yield the methyl acetal 299 highly efficiently as a single product. Further transformation was carried out as shown in the scheme and the dialdehyde 306 was subjected
to a reductive
cyclization with low valent titanium complexes to produce the inositol chiro-
myo-
derivatives 307(R=All) and 33(R=Bn) accompanied with the 308 and scyllo-isomers 309. Similar titanium reagents are
known to transform simple dicarbonyl compounds into
1,2-cis-diols
stereoselectively. -^^^ A
stereoselective
intramolecular
pinacol
coupling
of
the
dialdehyde 312, which was derived from D-mannitol in 9 steps, was
444 recently accomplished by using samarium diiodide as the coupling reagent compound
giving
the
cis-diol
(Scheme
92:8)
313
5-6). l^*^
(cis vs
trans-isomer,
scyllo-
Similarly, D-1, 4 , 5 , 6 - tetra-0-
benzyl-myo-inositol was obtained by the reaction of the protected dialdehyde derived from L-iditol in 56% yield together with the two trans-isomers, scyllo-
and chiro-inositols (4% each).^^^
Ph2(f-Bu)Si04
SmU 0Si(f-Bu)Ph2 8 6 % (2 steps)
Scheme The
mixture
of
5-6
33, 308, and 309
can be
converted
to
the
cyclohexene derivative 310 with a C2 symmetry axis by reaction with triphenylphosphine and triiodoimidazole and by respective osmylation and
epoxidation
obtained
of
310, myo-
33 and chiro-inositols
OSO4 PhgP
33 + 308 + 309
were
BnQ
Y^
mCPBA
33 BnO,
'v >
BnO*'* X'^"' OBn
OBn
308 (R=Bn)
HoO
OBn 311
310
Scheme
5-7
The modification at C-2 of myo-inositols starting
308
(Scheme 5-7) .
can be
demonstrated
from the 1,2-diol derivative 33 by a novel sequence of
reactions involving migration of a phosphonio function from C-1 to C-2 and subsequent nucleophilic substitution (Scheme 5-8).^^^ Thus t e t r a b e n zy 1 - myo-inositol
33
was
treated
with
iodine,
triphenylphosphine, and imidazole to give the epimerized iodide 314 with the scyllo-configuration together with a small amount of an elimination nucleophilic
product
310. The iodide 314 was again subjected
substitution
with
phenylmethanethiolate
to
after
protection of the remaining hydroxyl group as the TBDMS derivative and removal of the silyl group gave 315 which was then transformed to the 2-thio-myo-inositol derivative 316. The formation of the 2-
445 iodide 314 can be explained by the migration of the initially formed 1-phosphonium
intermediate
320 and then substitution of 322 with
the iodide ion. Considering that the Mitsunobu reaction takes a reaction pathway similar to that described above, 33 was allowed to react with p-nitrobenzoic acid resulting in the formation of the expected scyllo-inositol 318 in excellent yield.
BnO.OH BnO BnO. BnO
PPh3.l2 Imidazole PhMe, refl.
33 PPh3, DEAD p-NOgPhCOgH PhMe, refl. BnO ^nO-T"^-^^ BnO. BnO 318
61%
BnO BnO BnOBnO 315
I.NaOH 2. NaH, BnBr BnO
310
9%
SBn
AcO?^^ AcO AcO. AcO 316
B n O ^ / ^ - T ^ - - ^ ^ OBn BnO 319
NHo
OAc IJiS^r
HOJ "
g2Z:;:^0H HO.
91%
HO
317
„ ^ OPPhg BnO J "^ B nOJ y-pphg Dnu B n O ^ - ^ o ' ^ ^ ^ B n O ^ - ^ O H — BnO 321
314
BnO 322
Scheme
6.
BnO
Na. liq. NH3 85% 1 then AC2O, Py
BnOT"^—^-7^ OBn
BnOjOH
^ Bn^SSZ;::^
1.TBDMS-OTf,90% 2. BnSNa. DMF, 120 'C, 2.5 h, 84% 3. /7-BU4NF, 99%
OgCPh-p-NOg
BnO' BnO.! ^ 2 ^ z : ^ o p p h 3 - — BnO 320
BnO
BnO
^-"2Z;::^(H BnO BnO 314 70%
5-8
SYNTHESIS OF PHOSPHATIDYLINOSITOLS As described in the Introduction, various inositol phospholipids
exist
in nature and are likely
to play
important
physiological
roles although their biological properties in many cases are still unclear. The synthesis of phosphoinositides has been reported over the years.^^^ In this section, a newly discovered PI(3,4,5)P3 and glycosyl phosphatidylinositols are treated. 6.1 Synthesis of Phosohatidvlinositol 3,4,S-Trisphosohate Falck et al. have reported the syntheses of various inositol derivatives
starting
from
(-)-quinic
acid
3 2 3 . ^^^
These
transformations involve interesting reactions but generally require
446 many steps. The enone 324 was converted
into the i77yo--inositol
derivative 329 with a variety of protecting groups and was finally converted
to
D-I ns { 1 , 3 , 4 , 5 ) P4 . L-Ins ( 1 , 5 , 6 ) P3 . ^ ^ ^
PI(3,4,5)P3 (Scheme 6-1).^^^
HO
902Me 1
CO2H
HO^'^^OH
68%
, \
Q
u=
OH 323
^ sPh
^ SPh
i, ii iii, iv
^
80% HO*' T
'ODMPM
' SEMO'
=
ODMPM
^^"
(S
^
N—/324
325
326 83% j vii, viii
OTBDMS
.»*k^'^ 66% »»»»\^% 90% o**\x^''' SEMO* \ ODMPM SEMO 5 ODMPM SEMO = ODMPM OBn OBn OBn xiii, 90% 329 xiii, 90% 323 327 XV, 81% / V xiv, 80% xiv, xvi, 80% / \ v XV, 89% OP{0)(OBn)2 L-lns(1,5,6)P3 (BnO)2P(0)0^ is.^OBn xvi 00%
D-lns(1,3,4,5)P4
•
'
»^/o
(BnO)2P(0)0*''^Y"'''OH
....^Jji^^xvi ^"^"-*-
OBn
PI(3,4,5)P3
330 i: DIBAL, -78 'C; ii: (PhS)2, n-BugP; iii: DMPM-Br, NaH; iv: MeOH, HCI; v: rvBugSnO then SEM-CI, CsF. -15 to 0 'C; vi: BnBr, KH; vii: mCPBA, (EtO)3p, viii: BnBr, KH; ix: OSO4, NMO; x: Pb(0Ac)4; xi: TBDMS-OTf, EtgN; xii: BH3. MeaS then mCPBA; xiii: HCI, MeOH; xiv: (BnOlgPNAPrg, tetrazole then mCPBA; xv: DDQ; xvi: Hg, Pd-C, 50 psi; xvli: 1,2-di-O-palmitoyl-sn-glycerol 3-(benzyi MA^diisopropylphosphoramidite), tetrazole then mCPBA
Scheme A
versatile
syntheses
intermediate
of PI(3,4,5}P3
79
6-1 was
involved
(Scheme 6-2).^^
in
two
concise
Since optically active
1,2-cyclohexylidene-jnyo-inositol 47a can be obtained enzymatically (Scheme 4-9) , the target molecule with the natural
configuration
can be obtained.'^'^ Chen's
synthesis
of
PI
used
the
enzymatically
resolved
dieyelohexylidene ketal (-)-50a for the optically active substrate and Pd black for the final deprotection
(Scheme 6-3).40 p^ black
seems to be better as a catalyst for hydrogenolysis than Pd-C which was
used
to deprotect
the
same molecule
in Scheme
6-1
since
adsorption of PI phosphate on the active carbon is possible. Indeed, hydrogenolysis
on Pd-C of a similarly protected
derivative
343
resulted in no isolation of any inositol derivative at all while the same reaction of 333 and 335 in Scheme 6-2 caused no problem. ^^^
and
447
oQ
OH
PO^A^OH
85% ji 72%
OLev
OP
83%
po***Sr '''OLev
POI*'
OP 331
79
I^^'OLev OP 332
OH
vi, 100%;vii,73% viii, 96%;xi, 100% GICH2CO2
O
'o''
T
PO^^X^O-ff-OBn CICH2CO2
P^Js^O-P-OMe TIPS T
ODSG
-
36%
OOSG
O
P'O^^As^ 0-P-OMe aosG
.XI
y'^''OLev
21%
OP 333
PI(3,4,5)P3
'OLev
'o'
OSiEta
OLev
y^ c OP
334
335 P=(BnO)2P(0)
•cc::
02CC17H35
P(0)
DSG=^
1^
02CC17H35
i: n-Bu4NF»3H20. PhC02H; ii: (BnO)2pN/-Pr2, tetrazole then mCPBA; iii: TFA, commercial CH2CI2, iv: (BnO)2PO-DSG, PyHBrg, lutidine; v: 1. H2, Pd-C, 2. NH2NH2; vi: EtgSi-CI; vii: TsOH, (CH20H)2; viii: (MeO)2PO-DSG. PyHBrs, EtgN; ix: [CICH2C{0)]20, Proton sponge; x: 47% aq. HF; x: XEPA, tetrazole then mCPBA; xi: 1. H2, Pd-C, 2. NH2NHC(S)S-HEt(/-Pr)2N*. 3. PhSH, EtgN, 4. NHgNHg
Scheme
Alio,
OH I OH
I OBn OAll 336
OR^ 1 .OR^
rr
Alio,
tf..
Alio
6-2
Alio'*
R O y T
f?
(BnO)2P(0)0
OBn
V- °1
(BnO)2P(0)0 ) I
OBn
<
339: R=H 340: R=(BnO)2P(0) gQo/,
^vJk>OH
( B n O ) 2 P ( 0 ) q ^ T ^ O - P - OBn PI(3,4,5)P3
RO
OR
jjj / - 1 4 2 : R^=MEM, R2=Bn . > 337: R^=H, R2=Bn '^ * ^ 3 3 8 : R^=Ac, R2=Bn QJO/^
OBn
I OBn oVc
88%
jT " O B n OAll
OAc
O2OC15H31
^h 02CCi5H3i
342
PO'
4^''0Bn OP 341 P=(BnO)2P(0)
i: 1. n-BugSnO, 2. MEM-CI, CsF; ii: BnBr, NaH; Iii: ZnBrg; iv: AcgO, DMAP; v: Pd-C, TsOH, aq. MeOH; vi: (BnO)2PN/-Pr2, tetrazole, then mCPBA; vii: NaOH; viii: 1,2-di-O-palmitoyl-sn-glycerol 3-(benzyl A/,A/-diisopropylphosphoramidite), tetrazole then mCPBA; ix: Hg, Pd black (50 psi)
(BnO)2P(0)Q,)
OH
Jf
0-P-OBn
(BnO)2P(0)0'
•" Y'"OBn h O2CC17H35
(BnO)2P(0)0
O2CC17H35
343
Scheme
6-3
448 6.2 Synthesis of Glvcosvl Phosphatidvlinositols Total synthesis of the glycosyl phosphatidylinositol anchor of Trypanosoma brucei has been achieved by Murakata and Ogawa (Scheme 6.4)^114 rp]-^e hexasaccharide core 353 additionally involving the myo-inositol residue at the one terminal, in which all glycosidic linkages are a was assembled by stepwise stereoselective aglycosylation using glycosyl sulfide, fluorides, and chloride as glycosyl
donors.
Introduction
of
two phosphoryl
functions
was
accomplished efficiently by the H-phosphonate method which involved the
initial
formation
of
the
H-phosphonic
and
pivalic
mixed
anhydride 356 and subsequent oxidation with I2 as shown in Scheme 6-5.
BnO HO' MPMO-
OBn
BnO-X^^X"^
v
y
346 ^^^ 348
1
I
r.n^
AcO-^-'^^n
. ^
BnO
BnO
BnO-V^' HO-
347
N3
I BnO-7 3nO-, -, oBnO-i>—r~^ N3 n 349
^ n ^ BnOi n BnO-^^'' ' ^ ^ ^ BnO-
BnO r* T .
- 350 BnO _ _
AcO
BnO'^^'' BnOnO-^*-"^^
no-^-^-^n BnO-\ ?o
BnOBnO'' BnO-
y/
351 CI
BnO-^^-"-^^ 3^9 F F 352
BnO
BnO. OBn ^3 i -- — r - X ^ I - ^ OBn J3 A
V/CoBn
Bntf^
353
MPM
O II
-P-0-| H-P-0-| OH [-OCOC13H27 ^OCOCi3H27
u II
. K-fj»-OCH2CH2NHGbz OH 355
354 GPI-anchor Scheme
6-4
449
f-BuCOCI • ^
354
0 0 II •' ^ ROH ^.B^Q»o-P-0-j • l^ hOCOCi3H27 '- OCOC13H27
356
O '' ^ RO-P-0-| X rOCOCi3H27 ^ OCOC13H27 >,aq.Py Q 357: X=H 358: X=OH
Scheme
6-5
Fraser-Reid et al. reported the synthesis of the oligosaccharide backbones of two GPI anchors of Trypanosoma
brucei^-^^
and rat brain,
which involved highly stereocontrolled glycosylations.^^^ 6.3 Synthesis of Phosphatidvlinositol Mannosides Reports on the preparation of 2-a-mannopyranosyl-1-phosphatidylD-myo-inositol were published in 1977 (Ref.117) and 1 9 8 9 . ^ ^ After these reports, recently the 2,6-dimannoside
{PIM2/ 363) has been
synthesized. Van Boom's group-^-^^ introduced the mannosyl moieties to the 2 position and subsequently to the 6 position in myo-inositol derivative
359 to yield 361. Dimannoside 361 was
phosphorylated
after removal of the allyl group by the H-phosphonate method and finally deprotected giving the final product PIM2 (Scheme 6-6) . lO-N^lO 360 SEt
OH BnO^ J^
, ^OAII
°TV'
BnO'
7 ^ "OMPM BnO ocQ
IV
BnO-^
10
BnO-^^
H
Q
BnO \ Qg^ r
j^\ OAlir U
Txy^. >^'^ ^"^
= 'O—^^ BnO 361 \
O H-P-O-r . OH h0C0Ci5H3i
A H 0 ^ 2 \ ?Q HOr ^ > L \ HO' HQ
-"is^ai
362
POCOC15H31 O .••OCOC,5H3i n HO-P-O-J O OH OH
HO' HO HO 363
Scheme
O
6-6
In t h e following method, in c o n t r a s t t o t h e above one, all p r o c e s s e s s t a r t i n g from m y o - i n o s i t o l were performed i n completely
450 regiospecific manners by using only two protecting groups, cyclohexylidene and TIPS, for the inositol skeleton (Scheme 67)^120 Thus, 76 was glycosylated at C-6 with mannopyranosyl phosphite 364 followed by deprotection of the ketal to yield the aglycoside 366 which was separated into two diastereomers. The desired stereoisomeric triol 366 was phosphorylated only at C-1 by the phosphite-phosphonium method and the glycosylation of the resultant phosphate 368 took place regio- and stereoselectively at C-2 affording 369 together with no P-glycoside. The inert nature of the hydroxyl group at C-5 might be attributed to steric hindrance owing to the mannosyl moiety at C-6 as well as the TIPS group. Deprotection of the carbonate of 369 was efficiently carried out by using
the
ethyl
Grignard
reagent
and,
after
acylation
and
deprotection of 370, gave PIM2 in good yield.
BnO-vBnO 1.n-Bu4NF-3H20 PhCOOH 89% 2. PhSH, EtgN 89%
' Ci7H35C02~|
OBn 2. C17H35COCI O^OBn 73%
^^^f 0"
Scheme
7.
OBn
PiMo
quant.
6-7
CONCLUSION A variety of methodologies and strategies for the synthesis of
inositol derivatives has been reported in the recent decade. We can now prepare any of the desired inositols. Although development of
451 new methodologies is still interesting and important in the field of inositol chemistry, our efforts should be directed to create useful probes having a variety of functions which interact with receptors and metabolic enzymes in order to understand various cell signaling processes
including the metabolisms of inositols at the molecular
level.
8.
ACKNOWLEDGMENT The author is deeply grateful to Tomoko Nakamura for assistance
in the preparation of the manuscript.
9. 1 2 3 4 5
6 7
8 9 10 11 12 13
14
REFERENCES H. Streb, R. F. Irvine, M. J. Berridge, and I. Schulz, Nature, 306 (1983) 67-68. M. J. Berridge and R. F. Irvine, ibid., 312 (1984) 315-321. M. J. Berridge, Nature, 361 (1993) 315-325. Y. Nishizuka, Nature, 334 (1988) 661-665. T. W. Rail, E. W. Sutherland, and J. Berthet, J. Biol. Chem. , 224 (1957) 463-475. T. W. Rail and E. W. Sutherland, ibid., 232 (1958) 1065-1076. S. B. Shears, in: J. W. Putney, Jr. (Ed.), Advances in Second Messenger and Phosphoprotein research. Vol. 26. Chapter 4: Metabolism of Inositol Phosphates, Raven Press, New York, 1992, pp. 63-92. Newly discovered pathways were added by suggestion of Prof. Masato Hirata of Kyushu University. D. J. Cosgrove, Inositol Phosphates, Their Chemistry, Biochemistry and Physiology, Elsevier, Amsterdam, 1980. (a) C. Grado and C. E. Ballou, J. Biol. Chem., 236 (1961) 5460. (b) R. V. Tomlinson and C. E. Ballou, J. Biol. Chem., 236 (1961) 1902-1906. (c) H. Brockerhoff and C. E. Ballou, J. Biol. Chem., 236 (1961) 1907-1911. see also: L. E. Hokin and M. R. Hokin, Biochim. Biophys. Acta, 84 (1964) 563-575. Y. Chuan and C. E. Ballou, Biochem., 4 (1965) 1395-1404. M. A. J. Ferguson, S. W. Romans, R. A. Dwek, and T. W. Rademacher, Science, 239 (1988) 753-759. Y. Watanabe, E. Inada, M. Jinno, and S. Ozaki, Tetrahedron Lett., 34 (1993) 497-500. For pioneering work, see: V. I. Shvets, B. A. Klyashchitskii, A. E. Stepanov, and R. P. Evstigneeva, Tetrahedron, 29 (1973) 331-340. (a) D. C. Billington, R. Baker, J. J. Kulagowski, and I. M. Mawer, J. Chem. Soc., Chem. Commun. (1987) 314-316. (b) J. Gigg, R. Gigg, S. Payne, and R. Conant, J. Chem. Soc. Perkin Trans. I, (1987) 1757-1762. (c) J. P. Vacca, S. J. deSolms, and J. R. Huff, J. Am. Chem. Soc, 109 (1987) 3478-3479. V. I. Shvets, A. E. Stepanov, V. N. Krylova, P. V. Gulak, myoinositol and Phosphoinositides, Science Publishing House, Moscow, 1987. D. C. Billington, Chem. Soc. Rev., 18 (1989) 83122. B. V. L. Potter, Natural Product Reports, 7 (1990) 1-24.
452
15 16
17
18 19
20 21
22 23 24 25
26
A. B. Reitz (Ed), ACS Symposium Series 463, Inositol Phosphates and Derivatives, Synthesis, Biochemistry, and Therapeutic Potential, American Chemical Society, Washington, D.C., 1991. D. C. Billington, The Inositol Phosphates, Chemical Synthesis and Biological Significance, VCH, Weinheim, 1993. B. V. L. Potter and S. R. Nahorski, in: A. P. Kozikowski (Ed), Drug Design for Neuroscience. Chapter 14: Synthetic Inositol Polyphosphates and Analogues as Molecular Probes for Neuronal Second Messenger Receptors, Raven Press, NY, 1993, pp 3 83-416. S. Ozaki, Y. Watanabe, M. Hirata, T. Ogasawara, T. Kanematsu, and M. Yoshida, in: A. P. Kozikowski (Ed), Drug Design for Neuroscience. Chapter 15: Inositol 1,4,5-Trisphosphate Affinity Chromatography: Fishing out novel Ins(1,4,5)P3-recognizable Proteins, Raven Press, NY, 1993, 417-434. D. C. Billington and R. Baker, J. Chem. Soc. , Chem. Commun. , (1987) 1011-1013. (a) S. Ozaki, Y. Watanabe, T. Ogasawara, Y. Kondo, N. Shiotani, H. Nishii, and T. Matsuki, Tetrahedron Lett., 27 (1986) 31573160. (b) Y. Watanabe, H. Nakahira, M. Bunya, and S. Ozaki, Tetrahedron Lett., 28 (1987) 4179-4180. (c) S. Ozaki, Y. Kondo, N. Shiotani, T. Ogasawara, and Y. Watanabe, J. Chem. Soc. Perkin 1, (1992) 729-737. (a) K.-L. Yu and B. Fraser-Reid, Tetrahedron Lett., 29 (1988) 979-982. (b) G. Baudin, B. I. Glanzer, K. S. Swaminathan, and A. Vasella, Helv. Chim. Acta, 71 (1988) 1367-1378. (c) C. E. Dreef, C. J. J. Elie, P. Hoogerhout, G, A. van der Marel, and J. H. van Boom, Tetrahedron Lett., 29 (1988) 6513-6516. Y. Watanabe, Y. Komoda, K, Ebisuya, and S. Ozaki, Tetrahedron Lett., 31 (1990) 255-256. A. V. Luk'yanov, A. I. Lyutik, V. I. Shvets, and N. A. Preobrazhenskii, Zh. Obsh. Khim., 36, (1966) 1029-1031. V. N. Krylova, A. I. Lyutik, N. P. Gornaeva, and V. I. Shvets, Zh. Obsh. Khim., 51, (1981) 210-214. G. Lin, C. F. Bennett, andM.D. Tsai, Biochem., 29 (1990) 2747-2757. J. L. Meek, F. Davidson, and F. W. Hobbs, Jr., J. Am. Chem. S o c , 110 (1988) 2317-2318. (a) P. A. Gent, R. Gigg, and C. D. Warren, Tetrahedron Lett., (1970) 2575-2578. (b) P. A. Gent, R. Gigg, and C. D. Warren J. Chem. Soc. (C) , (1969) 2367-2371. (c) D. J. R. Massy and P. Wyss, Helv. Chim. Acta, 73 (1990) 1037-1057. (d) C. E. Dreef, R. J. Tuinman, A. W. M. Lefeber, C. J. J. Elie, G. A. van der Marel, and J. H. van Boom, Tetrahedron, 47 (1991) 4709-4722. (e) T. P. Zubkova, Z. Ya. Khrapkova, I. K. Sarycheva, and N. A. Preobrazhenskii, Zh. Org. Khim., 4 (1968) 2226-2228. C. Schultz, T. Metschies, B. Gerlach, C. Stadler, B. Jastorff, Synlett, (1990) 163-165. M. Jones, K. K. Rana, J. G. Ward, and R. C. Young, Tetrahedron Lett., 30 (1989) 5353-5356. S.-K. Chung and Y. Ryu, Carbohydr.Res., 258 (1994) 145-167. For cyclohexylidene ketal synthesis: S. J. Angyal, G. C. Irving, D. Rutherford, and M. E. Tate, J. Chem. S o c , (1965) 6662-6664. C. Jiang and D. C. Baker, J. Carbohydr. Chem., 5 (1986) 615-620. And Ref. 21c. For isopropylidene ketal synthesis: Ref. 21b. For the synthesis of dicyclohexylidene ketals: (a) P. J. Garegg, T. Iversen, R. Johansson, and B. Lindberg, Carbohydr. Res., 130 (1984) 322-326. And Ref. 21c. For the synthesis of diisopropylidene ketals: (b) J. Gigg, R. Gigg, S. Payne, and R. Conant, Carbohydr. Res., 142 (1985) 132-134. (c) J. Gigg, R. Gigg, S. Payne, and R. Conant, J. Chem. Soc. Perkin I, (1987),
453
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
2411-2414. (d) R. F. de la Pradilla, C. Jaramillo, J. JimenezBarber o, M. Martin-Lomas, S. Penades, and A. Zapata, Carbohydr. Res., 207 (1990) 249-257. AndRef. 24. J. P. Vacca, S. J. deSolms, J. R. Huff, D. C. Billington, R. Baker, J. J. Kulagowski, and I. M. Mawer, Tetrahedron, 45 (1989) 5679-5702. C. J. J. Elie, R. Verduyn, C. E. Dreef, D. M. Brounts, G. A. van der Marel, and J. H. van Boom, Tetrahedron, 46 (1990) 82438254. C. E. Dreef, C. J. J. Elie, G. A. van der Marel, and J. H. van Boom, Tetrahedron Lett., 32 (1991) 955-958. M. S. Shashidhar, J. F. W. Keana, J. J. Volwerk, and 0. H. Griffith, Chem. Phys. Lipids, 53 (1990) 103-113. J. J. Kulagowski, Tetrahedron Lett., 30 (1989) 3869-3872. K. M. Pietrusiewicz, G. M. Salamo' nczyk, K. S. Bruzik, and W. Wieczorek, Tetrahedron, 48 (1992) 5523-5542. S. Ozaki, Y. Kondo, H. Nakahira, S. Yamaoka, and Y. Watanabe, Tetrahedron Lett., 28 (1987) 4691-4694. F. Tagliaferri, S.-N. Wang, W. K. Berlin, R. A. Outten, and T. Y. Shen,Tetrahedron Lett., 31 (1990) 1105-1108. (a) D.-M. Gou, Y.-C. Liu, and C-S, Chen, Carbohydr. Res., 234 (1992) 51-64. (b) D.-M. Gou, W.-R. Shieh, P.-J. Lu, and C.-S. Chen, Bioorg. Med. Chem., 2 (1994) 7-13. P. J. Garegg, B. Lindberg, I. Kvarnstrom, and S. C. T. Svensson, Carbohydr. Res., 173 (1988) 205-216. C. Murakata and T. Ogawa, Tetrahedron Lett., 31 (1990) 24392442. S. Danishefsky and R. Hungate, J. Am. Chem. Soc. , 108 (1986) 2486-2487. S. Danishefsky, R. Hungate, and G. Schulte, J. Am. Chem. S o c , 110 (1988) 7434-7440. N. Nagashima and M. Ohno, Chem. Lett., (1987) 141-144. D.-M. Gou and C.-S. Chen, Tetrahedron Lett., 33 (1992) 721-724. D.-M. Gou and C.-S. Chen, J. Chem. S o c , Chem. Commun., (1994) 2125-2126. C. E. Dreef, W. Schiebler, G.A. van der Marel, and J. H. van Boom, Tetrahedron Lett., 32 (1991) 6021-6024. P. Westerduin, H. A. M. Willems, and C. A. A. van Boeckel, Tetrahedron Lett., 31 (1990) 6915-6918. C. E. Dreef, R. J. Tuinman, C. J, J. Elie, G.A. van der Marel, and J. H. van Boom,Reel. Trav. Chim. Pays-Bas, 107 (1988) 395397. R. Baker, J. J. Kulagowski, D. C. Billington, P. D. Leeson, I. C. Lennon, and N. Liverton, J. Chem. S o c , Chem. Commun., (1989) 1383-1385. See also: S. J. Angyal, M. E. Tate, and S. D. Gero, J. Chem. S o c , (1961) 4116-4122. K. S. Bruzik, J. Myers, and M.-D. Tsai, Tetrahedron Lett., 33 (1992) 1009-1012. L. Ling and S. Ozaki, Tetrahedron Lett., 34 (1993) 2501-2504. L. Ling and S. Ozaki, Carbohydr. Res., 256 (1994) 49-58. K. S. Bruzik and M.-D. Tsai, J. Am. Chem. S o c , 114 (1992) 6361-6374. Y. Watanabe, T. Ogasawara, S. Ozaki, and M. Hirata, Carbohydrate Research, 258 (1994) 87-92. G. M. Salamo'nczyk and K. M. Pietrusiewicz, Tetrahedron Lett., 32 (1991) 6167-6170. C. Schultz, G. Gebauer, T. Metschies, L. Rensing, and B. Jastorff, Biochem. Biophys. Res. Commun., 166 (1990) 1319-1327. Y. Watanabe, M. Mitani, T. Morita, and S. Ozaki, J. Chem. S o c , Chem. Commun., (1989) 482-483.
454 52 53 54 55 56 57 58 59 60 61 62 63 64
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
Y. Watanabe, H. Hirofuji, and S. Ozaki, Tetrahedron Lett., 3 5 (1994) 123-124. A. P. Kozikowski, A. H. Fauq, G. Powis, P. Kurian, and F. T. Crews, J. Chem. Soc. , Chem. Cornmun. , (1992) 362-364. H. W. Lee and Y. Kishi, J. Org. Chem., 50 (1985) 4402-4404. D. C. Billington, R. Baker, J. J. Kulagowski, I. M. Mawer, J. P. Vacca, S. J. deSolms, and J. R. Huff, J. Chem. Soc. Perkin 1, (1989) 1423-1429. T. Banerjee and S. M. Srikantian, Tetrahedron Lett., 35 (1994) 8053-8056. T. Desai, J. Gigg, R. Gigg, S. Payne, S. Penades, and H. J. Rogers, Carbohydr. Res., 216 (1991) 197-209. Gigg eq 1, 2; T. Desai, A. Fernandez-Mayoralas, J. Gigg, R. Gigg, and S. Payne, Carbohydr. Res., 205 (1990) 105-123. C. Liu, S. R. Nahorski, and B. V. L. Potter, J. Chem. S o c , Chem. Commun., (1991) 1014-1016. A. Aguilo, M. Martin-Lomas, and S. Penades, Tetrahedron Lett., 33 (1992) 401-404. A. Zapata, R. F. de la Pradilla, M. Martin-Lomas, and S. Penades, J. Org. Chem., 56 (1991) 444-447. Y. Watanabe, T. Shinohara, T. Fujimoto, and S. Ozaki, Chem. Pharm. Bull., 38 (1990) 562-563. Y. Watanabe, T. Fujimoto, T. Shinohara, and S. Ozaki, J. Chem. S o c , Chem. Commun., (1991) 428-429. S. Ozaki and Y. Watanabe, in: A. B. Reitz (Ed), ACS Symposium Series 463, Inositol Phosphates and Derivatives, Synthesis, Biochemistry, and Therapeutic Potential. Chapter 4: Synthesis of Inositol Polyphosphates and their Derivatives, American Chemical Society, Washington, D.C., 1991, pp. 43-65. Y. Watanabe and N.Sasaki, unpublished results. Y. Watanabe, T. Ogasawara, H. Nakahira, T. Matsuki, and S. Ozaki, Tetrahedron Lett., 29 (1988) 5259-5262. M. Brufani, M. C. Cesta, L. Donnarumma, L. Filocamo, G. Gostoli, S. Lappa, Carbohydr. Res., 228 (1992) 371-376. C.-Y. Yuan, H.-X. Zhai, and G.-H. Weng, Chinese J. Chem., 12 (1994) 174-178. S.-K. Chung, Y.-T. Chang, and K.-H. Sohn, Korean J. Med. Chem., 4 (1994) 57-65. Y. Watanabe, A. Oka, Y. Shimizu, and S. Ozaki, Tetrahedron Lett., 31 (1990) 2613-2616. Y. Watanabe, T. Fujimoto, and S. Ozaki, J. Chem. S o c , Chem. Commun., (1992) 681-683. P. J. Edwards, D. A. Entwistle, C. Genicot, K. S. Kim, and S. V. Ley, Tetrahedron Lett., 35 (1994) 7443-7446. C.-H. Wong and G. M. Whitesides, Tetrahedron Organic Chemistry Series Vol. 12, Enzymes in Synthetic Organic Chemistry, Pergamon, Oxford, 1994. Y.-C. Liu and C.-S. Chen, Tetrahedron Lett., 30 (1989) 16171620. L. Ling, X. Li, Y. Watanabe, T. Akiyama, and S. Ozaki, Tetrahedron Lett., 33 (1992) 1911-1914. L. Ling, Y. Watanabe, T. Akiyama, and S. Ozaki, Bioorg. Med. Chem., 1 (1993) 155-159. S. Ozaki, L. Ling, T. Ogasawara, Y. Watanabe, M. Hirata, Carbohydr. Res., 259 (1994) 307-310. Y. Watanabe, M. Tomioka, S. Ozaki, manuscript to be prepared. P. Andersch and M. P. Schneider, Tetrahedron Asymmetry, 4 (1993) 2135-2138. S. Ozaki, Y. Koga, L. Li, Y. Watanabe, Y. Kimura, M. Hirata, Bull. Chem. Soc. Jpn., 67 (1994) 1058-1063.
455 80
T. Hudlicky, J. D. Price, F. Rulin, and T. Tsunoda, J. Am. Chem. S o c , 112 (1990) 9439-9440. 81 H. A. J. earless, Tetrahedron Lett., 33 (1992) 6379-6382. 82 M. Mandel and T. Hudlicky, J. Chem. Soc. Perkin I, (1993) 741743. 83 S. Cerdan, C. A. Hansen, R. Johanson, T. Inubushi, and J. R. Williamson, J. Biol. Chem., 261 (1986) 14676-14680. 84 A. E. Ting and R. E. Pagano, Chem. Phys. Lipids, 60 (1991) 8391. 85 R. E. Pagan and R. G. Sleight, Science, 229 (1985) 1051-1057. 86 For a review article, see: R. J. Ferrier and S. Middleton, Chem. Rev., 93 (1993) 2779-2831. 87 (a) K. Sato, S. Sakuma, S. Muramatsu, and M. Bokura, Chem. Lett., (1991) 1473-1474. (b) K. Sato, M. Bokura, and M. Taniguchi, Bull. Chem. Soc. Jpn., 67 (1994) 1633-1640. 88 Y. Watanabe, M. Mitani, and S. Ozaki, Chem. Lett., (1987) 123126. 89 For review articles, see: M. P. Czech, J. K. Klarlund, K. A. Yagaloff, A. P. Bradford, and R. E. Lewis, J. Biol. Chem., 263 (1988) 11017-11020. A. R. Saltiel and P. Cuatrecasas, Am. J. Physiol., 255 (1988) Cl-Cll. 90 C. Jaramillo and M. Martin-Lomas, Tetrahedron Lett., 32 (1991) 2501-2504.C. Jaramillo, J.-L. Chiara, and M. Martin-Lomas, J. Org. Chem., 59 (1994) 3135-3141. 91 S. L. Bender and R. J. Budhu, J. Am. Chem. S o c , 113 (1991) 9883-9885. 92 V. A. Estevez and G. D. Prestwich, J. Am. Chem. S o c , 113 (1991) 9885-9887. 93 H. B. Mereyala and S. Guntha, J. Chem. Soc. Perkin 1, (1993) 841-844. 94 0. Arjona, A. de Dios, R. F. de la Pradilla, and J. Plumet, Tetrahedron Lett., 32 (1991) 7309-7312. 0. Arjona, A. Candilejo, A. de Dios, R. F. de la Pradilla, and J. Plumet, J. Org. Chem., 57 (1992) 6097-6099. 95 A. P. Kozikowski, A. H. Fauq, and J. M. Rusnak, Tetrahedron Lett., 30 (1989) 3365-3368. 96 A. P. Kozikowski, A. H. Fauq, G. Powis, D. C. Melder, J. Am. Chem. S o c , 112 (1990) 4528-4531. 97 A. P. Kozikowski, A. H. Fauq, J. Am. Chem. S o c , 112 (1990) 7403-7404. 98 A. P. Kozikowski, W. Tiickmantel, and G. Powis, Angew. Chem. Int. Ed. Engl., 31 (1992) 1379-1381. 99 A. H. Fauq, A. P. Kozikowski, A. Gallegos, and G. Powis, Med. Chem. Res., 3 (1993) 17-23. 100 S. C. Johnson, J. Dahl, T.-L. Shih, D. J. A. Schedler, L. Anderson, T. L. Benjamin, and D. C. Baker, J. Med. Chem., 3 6 (1993) 3628-3635. 101 T. Akiyama, N. Takechi, S. Ozaki, Bull. Chem. Soc. Jpn., 65 (1992) 366-372. 102 W. Tegge and C. E. Ballou, Proc Natl. Acad. Sci. USA, 86 (1989) 94-98. 103 S. V. Ley and L. L. Yeung, Spec. Publ. Royal Soc. Chem., Ill (Molecular Recognition: Chemical and Biochemical Problems II), (1992) 183-191. S. V. Ley, Pure Appl. Chem., 62 (1990) 20312034. 104 S. V. Ley, M. Parra, A. J. Redgrave, and F. Sternfeld, Tetrahedron, 46 (1990) 4995-5026. 105 S.V. Ley and L. L. Yeung, Synlett, (1992) 997-998. 106 E. J. Corey, R. L. Danheiser, and S. Chandrasekaran, J. Org. Chem., 41 (1976) 260-265. J. E. McMurry and J. G. Rico,
456
107 108 109 110 111
112 113 114 115 116 117 118 119 120
Tetrahedron Lett., 30 (1989) 1169-1172. J. E. McMurry and N. 0. Siemers, Tetrahedron Lett., 34 (1993) 7891-7894. J. L. Chiara and M. Martin-Lomas, Tetrahedron Lett., 35 (1994) 2969-2972. J. P. Guidot, T. L. Gall, and C. Mioskowski, Tetrahedron Lett., 35 (1994) 6671-6672. Other report: J. L. Chiara, W. Cabri, and S. Hanessian, Tetrahedron Lett., 32 (1991) 1125-1128. J. P. Guidot and T. L. Gall, Tetrahedron Lett., 34 (1993) 46474650. A. E. Stepanov and V. I. Shvets, Chem. Phys. Lipids, 25 (1979) 247-263. A. E. Stepanov and V. I. Shvets, ibid., 41 (1980) 151. J. R. Falck and A. Abdali, in: A. B. Reitz (Ed), ACS Symposium Series 463, Inositol Phosphates and Derivatives, Synthesis, Biochemistry, and Therapeutic Potential. Chapter 11: Enantiospecific Synthesis of Inositol Polyphosphates, Phosphatidylinositides, and Analogues from (-)-Quinic Acid, American Chemical Society, Washington, D.C., 1991, pp 145-154. J. R. Falck and A. Abdali, Bioorg. Med. Chem. Lett., 3 (1993) 717-720. Y. Watanabe and T. Ogasawara, unpublished results. C. Murakata and T. Ogawa, Tetrahedron Lett., 32 (1991) 671-674. C. Murakata a n d T . Ogawa, Carbohydr. Res., 235 (1992) 95-114. D. R. Mootoo, P. Konradsson, and B. Fraser-Reid, J. Am. Chem. S o c , 111 (1989) 8540-8542. U. E. Udodong, R. Madsen, C. Roberts, and B. Fraser-Reid, J. Am. Chem. S o c , 115 (1993) 7886-7887. A. E. Stepanov, V. I. Shvets, and R. P. Evstigneeva, Zh. Obs. Khim., 47 (1977) 1653-1656. C. J. J. Elie, C. E. Dreef, R. Verduyn, G. A. van der Marel, and J. H. van Boom, Tetrahedron, 45 (1989) 3477-3486. C. J. J. Elie, C. E. Dreef, G. A. van der Marel, and J. H. van Boom, J. Carbohydr. Chem., 11 (1992) 715-739. The results was presented at the 69th National Meeting of The Chemical Society of Japan, March, 1995. Y. Watanabe, T. Yamamoto, T. Okazaki, and S. Ozaki, manuscript to be prepared.
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
457
Synthesis of Phytosphingolipids Tadao Kamikawa
1.
INTRODUCTION Lipids occur throughout the living world in microorganisms, plants and animals. Recently, it has become well known that lipids not only contribute to the structure of cells and provide an energy store, they also participate in the transmission of chemical signals in living organisms. The majority of lipids are based on a glycerol backbone. Another important group of acyl lipids, however, have sphingosine-based structures. Sphingolipids have a rather specific distribution in cells where they are concentrated in the outer leaflet of the plasma membrane. Glycosphingolipids are recognized as mediators of cell-cell recognition and communication, cellgrowth regulators, cell immune response, and cell oncogenic transformation. The analysis and identification of sphingolipids have recently been substantially improved by the use of gas-liquid chromatography, fast-atom bombardment mass spectrometry, nuclear magnetic resonance spectroscopy and selective enzymatic cleavage. Despite these efforts, however, many important members of this class of biomolecules remain relatively inaccessible. Isolation of pure compounds is still difficult because of the diversity and heterogeneity of lipids. Effective synthetic routes to these compounds are, therefore, extremely important to investigate their chemical and biological properties. The chemistry of sphingolipids is rather well documented,! so the recent developments in the synthesis of phytosphingolipids will be presented in this review. 2.
BASIC STRUCTURE OF SPHINGOLIPIDS Sphingosine is a long-chain amino alcohol. Several long-chain amino alcohols which occur in nature are shown in Figure 1.
NH2 sphinganine, dihydrosphingosine, D-e/yf/)ro-2-amino-1,3-dihydroxyoctadecane
458
4-sphingenine, sphingosine, D-eAyf/7ro-2-amino-1,3-clJhydroxy-frans-4-octaclecene OH
OH
NH2
4-hydroxysphlnganine, phytosphingosine, D-f/bo-2-amino-1,3,4-trlhyclroxyoctaclecane OH
OH
NH2
4-hyclroxy-8-sphJngenine, dehydrophytosphingosine, D-rtoa-2-amino-1,3,4-trihydroxy-c/s-8-octadecene
Fig. 1. Long-Chain Amino Alcohols
Both the amino and the alcohol moieties of sphingosine can be substituted to produce the various sphingolipids {Fig. 2).
y Cerebroside
Ceramide < Sugar
^ Natural Sphingolipid
Sugar—Sugar
Sphingosine (Long-Chain Base)
Fig. 2. Sphingolipids
Ceramides are N-acylated sphingosines. This acyl linkage is resistant to alkaline hydrolysis and therefore can be easily distinguished from the 0-esters found in glycerol-based acyl lipids. Attachment of hexosides to ceramides yields cerebrosides. The naming "cerebroside" was originally used for the galactosyl ceramide of the brain but is now widely used for monoglycosyl ceramides. Further attachment of hexosides to cerebrosides yields natural glycosyl sphingolipids. These are conveniently written in abbreviated form.
459 e.g., Galal-44Galpl-^lCer (Gal: galactopyranosyl; Cer: Ceramide). Some glycosphingolipids contain one or more molecules of sialic acid in the sugar residues of ceramide oligosaccharide. These lipids are called gangliosides. Esterification of the alcohol moiety of the sphingosine base with phosphocholine yields sphingomyelin. 3.
BIOLOGICAL ACnVITIES OF SPHINGOLIPIDS Monogalactosyl ceramides are the largest single component of the myelin sheath of nerves. They are also found in the lung, kidney, liver, spleen, serum and almost all tissues although in trace amounts. Essentially all the glycosphingolipids are immunologically active, either in heptenic reactivity in vitro or in antibody-producing potency. Glycosphingolipids play key roles in many biological processes. They have been shown to be cell-surface receptors for viral and bacterial toxins, ^ regulators of cell proliferation by interacting with transmembrane signal transducers and to be mediators of cell-cell recognition events.^ Gangliosides (e.g. I) at the surface of leukocytes, which carry the Sialyl Lewis^ or Sialyl Lewis^ epitopes, are likely to play an important part in inflammatory responses.^ L^OH
CO.H
o,
.OH
OH
OH ^Ci3H2
HO
un^OH
I
NHAc
OH
<^
HN
C17H35 O
H60H 1 Sialyl Lewis'^ Ganglioside
Recently, even simpler ceramides and cerebrosides have been found to show interesting biological activities. Marine sponges are rich sources of bioactive cerebrosides. Hirsch's group^ isolated halicerebroside A 2, which exhibited mild antitumor activity against P388 leukemia cells, from the Haliclona sp. sponge.
460 Kawai and Ikeda^ found t h a t the fruiting body formation of basidomycete, Schizophyllm commune, could be stimulated by some cerebrosides in its mycelia. They identified one of the active substances as 3 . Structurally related but highly unsaturated cytotoxic glycosphingolipid 4 has beer solated from the sea star Ophidiaster opidiximusJ
3 4
R = (CH2)i3Me; R^ = Me; R^ =H R = (CH2)i9Me; R^ = H; R^ = Me; 10,11 -dehydro
Natori's group^ isolated agelasphins 5 from the sponge Agelas maurittamis. These substances are the first example of a-galactosylceramides showing antitumor activity from natural sources.
^
Agelasphin-7a Agelasphin-9a Agelasphin-9b Agelasphin-11
R= R* R= R=
-(CH2)nM6 -(CH2)i2Me -(CH2)iiCHMe2 -(CH2)iiCH(Me)Et
Kobayashi's group^ isolated novel phytosphingosine-derived azetidine alkaloids, penaresidines 6, with potent actomyosin ATPase activating activity from the sponge Penares sp.
HO H
Penaresidin A R^ = OH; R^ = H Penaresidin B R^ = H; R^ = OH
New antifungal agents, sphingofungins 7 , were isolated and chai-acterized.io These compounds have been shown to be specific inhibitors
461 of serine palmitoyl transferase (potent inhibitors of sphingosine biosynthesis). In connection with this, sphingofungins have the opposite stereochemistry at C(2) and C(3) to that of D-(+)-erythro-sphingosine.
Sphingofungin Sphingofungin Sphingofungin Sphingofungin
A B C D
R U C ( N H ) N H 2 ; R^ = H R^ = H; R^ = H R^ = H; R^ = Ac R^ = Ac; R^ = H
4.
SYNTHESIS OF PHYTOSPHINGOLIPIDS 4.1. Strategic Consideration The stereochemical features of the ph)^osphingosines include three consecutive chiral centers. Great efforts have been devoted to exploit how to control the stereochemistry. In the synthetic sequences, the alcohols are manipulated in a protected form which can be unmasked at later stages in the synthesis. The amino group has almost always been introduced via the azide group. In many cases, obtaining optically pure material is of foremost importance, particularly when the samples are used in a biological test. Because the early work on the synthesis of ph3rtosphingosines was carried out with racemates, in this review I will focus on the work since the 1980s. 4.2 Chiral Pool Synthesis The most effective approach to phytosphingosines has been to use natural products as starting materials. The advantage of these methods is that the desired chirality is contained within the starting material and that the raw materials are usually available not only in bulk quantities but also at low cost. On the contrary, these methods often require lengthy protections and functional transformations. (a) From D-Glucosamine In the first straightforward synthesis of phytosphingosine, Gigg's group 11 used D-glucosamine as the starting material. Scheme 1 outlines the synthesis of N-benzoyl Cis-phytosphingosine 15. The mesyloxy oxazoline 8, which is readily prepared from 2-benzamido2-deoxy-D-glucose, was hydrolyzed in dry methanolic hydrogen chloride to the furanoside 9. Treatment of 9 with sodium methoxide yielded the oxazoline 10,
462 which was then oxidized with periodic acid to the aldehyde 11. Wittig olefination of 1 1 with n-tridecylidene triphenylphosphorane followed by catalytic hydrogenation yielded the saturated oxazoline 12. Cleavage of the oxazoline ring of 12 with acid gave the ammonium salt 13. Base treatment of 13 gave the corresponding N-acyl derivative 14 without inversion of the configuration via a cyclic intermediate. The glycosidic linkage was hydrolyzed by acid, and the free sugar derivative was reduced with sodium borohydride to give N-benzoyl phytosphingosine 15.
X.
HO
HO-i HCI/MeOH
Wo
HO
n OMe O-1
HOH
MeONa 52% from 8
NHBz
T
Ph
OHC o
9^
HIO4
LCigHzyPPhaBr PhLi
^^^^^7
—I
n
^'
9 1 % from 10
^'
T
Ph
Ph
11 C13H27 — 1 0 ^ ' ^ ®
K2CO3
12 C13H27
1.HCI 2. NaBH4
80% from 12 PhCOO "^NHgCl' 13
OMe HCI
2.H2/10%Pd-C
T
10
HO
CH3(CH2)i3
NHBz
U
Scheme 1
More recently, Murakami's groupie reported the synthesis of the phytosphingosine 22 using a modem version of Gigg's method (Scheme 2). Thus 4,6-O-ethylidene-N-benzoyl-D-glucosamine IS was reduced with NaBH4 and selective protection of the primary hydroxyl group followed by mesylation of the secondary alcohol gave the dimesylate 17. Treatment of 17 with pyridine gave the phenyloxazoline IS. Deacetalization and base treatment afforded the 4,5-epoxide 19 as a single product.
463 Chain-elongation was achieved, via the tosylate of 19, with dodecylmagnesium bromide and CuBr to give the coupling product 20. Treatment of 20 with Nal and chlorotrimethylsilane gave the iodohydrin which without isolation was reduced with BuaSnH to give 21. Finally functional interconversion yielded 22. 1.NaBH4 2. /-Bu2Ph2SiCI/py; MsCI/EtgN
Me-'^O OH NHBz
86%
Me
0"^0
NHBz OTBDPS
py/EtgN
OMs OMs
16
17
OTBDPS
1. TiCIVPhSH 2. KaCOg/MeOH
OTBDPS
73%
18
I.TsCI/DMAP/EtgN 2. n-Ci2H25MgBr CuBr
—
84%
^
19
OTBDPS
C12H;
Nal/MegSiCI
C12H;
OTBDPS
20 n-BuaSnH AIBN 88%
a,
OTBDPS
C12H25
1.HCI 2.NaOH/H20 3. NaOH/aqEtOH 75%
OH NH2 JkJL/OH 6H 22
21 Scheme 2
(b) From D-Galactose In the second synthesis of ph5^osphingosine, outlined in Scheme 3, Giggs groupie made possible the preparation of 1,3,4-tri-Obenzyl-L-galactitol 23 from D-galactose in good yield. They proved 23 to be an efficient starting material.
464 1,3,4-TYi-G-benzyl-L-galactitol was converted into the isopropylidene mesylate 24. Treatment of 2 4 with lithium azide in DMF gave the azide 25 with inversion of the configuration. Reduction of the azide 25 with lithium aluminum hydride gave the amine 26 To protect the amiino group, Nethoxycarbonylphthalimide was used with triethylamine as solvent to give the phthaloyl derivative 27 Mild acid hydrolysis of the phthalimide 2 7 gave the corresponding glycol which was cleaved with sodium metaperiodate. The aldehyde formed was subjected to a Wittig olefination with n-tridecylidene triphenylphosphorane. Hydrogenation of the double bond with concomitant hydrogenolysis of the benzyl groups gave the triol 28. Deprotection of 28 with hydrazine hydrate followed by benzoylation and hydrolysis of Obenzoyl groups gave 15. OH
OBn
BnO.
LMegCO p-TsOH
OMs OBn
LiNo
BnO.
••
63%
2. MsCI/py 83%
N3
OBn
N-ethoxycarbonyl phthalimide
NH2 OBn
LiAIH4
BnO^
BnO^ OBn
OBn
25
26
PhthN
OBn
BnO^ OBn 27 I.H2NNH2H2O 2. BzCI/py 3. KOH 72%
1. HCI/MeOH 2. Nal04
^.
+ 3. rhC-,3H27PPh3Br PhLi 4. Hg/Pd-C 27% BzHN
PhthN
i
y
EEtsN
^ - f - 72% from 25 OH ''(CH2)i3CH3
OH
28
OH ^(CH2)i3CH3 OH 15
Scheme 3
Essentially similar synthesis of the ceramide part 3 4 of wheat flour glycosyl ceramides from 3,4,6-tri-Obenzyl-D-galactose 29 has since been published by Ogawa's groupie (Scheme 4).
465 BnO/OBn
OH
OBn 1. MsCI/py
BnO^ OH '^^
2. fvCigHgyPPhgBr A>BuLi
29
30
j^
OBn
BnO>.
C14H29
C14H21
95%
OBn
72%
N3
NaNa
OBn
1. HS(CH2)3SH EtaN 2. C23H47COOH
CL,
32
CH3
NH
/BU3N
r
72%
H2/10%Pd-C
OBn
BnO.
^.
2. TsNHNHs AcONa
OBn
65%
OMs OBn BnO.
C 12^21
94%
C14H29
C14H29
Scheme 4
An alternative versatile starting material is 2,4-Obenzylidene-D-threose 35, which is readily obtained from D-galactose. Retrosynthetic analysis indicates that the 2,4-disubstituted D-threose derivative and carbanions would offer a shorter route to D-ribophytosphingosines as well as D-ribodehydrophytosphingosines (Scheme 5). NH2 OH
N3
OH
OH
O
RO, OR
Scheme 5
Schmidt's group ^^ reported the first synthesis of the lactosyl ceramides 44 (Scheme 6). Reaction of excess n-tetradecylmagnesium bromide with 35 in THF at 60°C afforded a 1:1 mixture of D-arabino- and L-xyZo-octadecanetetrol derivatives, 36a and 36b, respectively. Reactions at lower temperature or in different solvent did not result in a dramatic change in the ratio of 36a and 36b. It is interesting, however, that the addition of salts, such as the copper bromide-dimethyl sulfide complex or titanium tetrachloride, as a catsdyst led
466
to preferential or exclusive formation of 36b. This preferential formation of 36b is probably due to si-face attack of the nucleophile on the metal complex with three oxygen functions as in A.
s/face A
After separation of the diastereomers, the required C(2)-0-activation of 36a was simply done by treatment of methanesulfonyl chloride in pyridine to give 3 7 . This favorable chemoselectivity is attributed to the oxygen atoms of the 1,3-dioxane ring causing increased nucleophilicity of C(2)-OH through accumulation of lone-pair orbitals and/or through higher acidity because of hydrogen bonding. Treatment of 37 with sodium azide afforded the azide 3 8 which was deprotected by acid treatment to give 39, For the synthesis of glycosphingolipid, the azide 3 9 was first partially protected with 2,2-dimethoxypropane and p-toluenesulfonic acid to give the 3,4-O-isopropylidene derivative 40. Glycosylation of 4 0 with the lactosyl donor 41 using Schmidt's protocol^^ gave the 1-O(p-Iactosyl)phytosphingosine derivative 42. Reduction of the azide 42 followed by acylation of the resulting amine with palmitoyl chloride gave 43 which was deprotected to furnish the desired glycophytosphingolipid 44. OH O^ J^ Ph^^U^oCZjLy^ 35
+
BrMg(CH2)i3CH3
OH
O^f
Ph-..^O^X^l^(QH2)^3CH3 H 36a 35%
THF/60°C
OH
+
0^»i*
^^^--^^•^--^T^(CH2)i3CH3 HO 36b 36%
MsCI/py OMs lOH
NaN3/DMF
Ph^0CZf.(CH2)i3CH3 H ^ ^"" "
63% ""'°
O
37
^
OH
HCI
P^-^0CZ^^0H2).3CH3 H ' - ' — -
65%
38
467
39
41 OAc (
OAc
^^^^
ISJ3
Aco-r;:^^or^-q o,A_(CH,,,3CH3
^^"^^
ArON
69%
^ ^
OAc
^'^^
OAc
..
o
1. LiAIH4/NiCl2
^'^^
2. CH3(CH2)i4COCI
X
<(
55%
1. MeONa/MeOH
^^7^iZ^.9Z\Z:^O.J^ 1 ^ 0 AcO^OAc
n
ecu
AcO-^*^^T^
(CH2)i3CH3
^ r ~ < O N / 0
2.CF3CO2H 50%
43
HO"
OH OH
HO^oH
OH ,0H
0^^(CH2)i4CH3
T
NH
OH ''(CH2)i3CH3
"^
OH
44
Scheme 6
New glycosphingolipids, agelasphins 5, have been isolated from the marine sponge Agelas mauritianus.^ Although a number of p-galactosylceramides have been known, agelasphins are the first cerebrosides having an a-galactosyl linkage. It is also interesting that all agelasphins showed antitumor activity. The synthesis of agelasphin-9b 45 by Akimoto's group, i'^ outlined in Scheme 7, starts with the derivative 46 obtained from 29 . A series of reactions, (a) Wittig olefination, which accompanied formic ester hydrolysis, (b) mesylation, and (c) azidation gave the azide 48. The primary hydroxyl group of 4 8 was protected, and the remaining secondary hydroxyl groups were benzoylated. Selective deprotection gave the alcohol 49. Reduction of 4 9 afforded the corresponding amine, which was acylated with p-nitrophenyl {2R)acetoxytetracosanoate 50 to give the protected ceramide 51.
468 To obtain the a-glycoside, 51 was reacted with the galactosyl fluoride with a non-participating protecting group at C(2) 52 under Mukaiyama's conditions 1^ to give 53 as a single product albeit at a low yield. Deprotection of 53 afforded the target molecule 45. OHCO
OBn
Br Ph3P(CH2)8CH=CHCH(CH3)2
^vAjA CHO
^>
BnO,
n-BuLi
OBn
68%
46 OH
l.MsCI/py 2. Hg/Pd
OBn
BnO,
3. NaN3 OBn
68% 47
N3
I.TrCI/py 2. BzCI/py/DMAP
OH
3. TsOH/MeOH
67%
N3
OBz
I.Hg/Pd-black OAc
OBz
2.
O:
*"]]
49
40%
OAc
SnCl2/AgCI04
•Y"^(CH2)2iCH3 NH
MS4A
OBz
^.
pOBn BnoLo (foenVvF
OBz 51
OBn
52
36%
OAc BnoL-o (OBn\
"O.. Y''^^^'^2)2iCH3 NH OBz
1. Hg/Pd-black
2. MeONa BnO
C22H45-n
50
OBz 53
^^
469 OH (QH N|
NH OH
Scheme 7
(c) From iV-Acetyl-D-Mannosamine In 1992 Van Middlesworth group reported the isolation 19 and structure elucidation 10 of sphingofungins 7, a new family of antifungal metabolites produced by Aspergillus Jiimigatus ATCC 20857. Their stereochemistry at the C(14) still remained unknown. Mori's group^O has developed the method for synthesizing {14R]- and (14S)-isomers of sphingofungin D. Retrosjnithetic analysis suggests that (i?)-l,2-epoxyoctane 54, 1-heptyne 55, and N-acetyl-D-mannosamine 56 can serve as the starting materials (Scheme 8). OH OH ^COgH OH NHAc
.vNHAc
H U0—- \\ M
H54
55
NHAc NMAC
HO-^\MR
HO-^'-'^^^^OH 56
Scheme 8
Scheme 9 outlines the synthesis of the (2S, 3 R 4K,5S, 14S)-isomer 57. Cleavage of 54 with the acetylide derived from 55 yielded 58, which was subjected to the acetylene-zipper reaction to give the (9R)-terminal acetylene
470
59. Mitsunobu inversion followed by protection converted {9R)'59 to the corresponding (9S)-TBS ether 60. The TBS ether 60 was then converted into the (2)-iodide 61 uia the alkenylstannane. On the other hand the polar building block 65 was prepared from 56. Oxidation of 56 with bromine water gave 62 . Protection of the uic-diol of 62 as benzylideneacetal was followed by further protection of the remaining hydroxyl group to afford 63. Hydrogenolysis of the benzylidene group of 63 by transfer hydrogenation with cyclohexene and the Pearlman catalyst gave 64, which was oxidized with sodium periodate to give the aldehyde 65. Coupling of 65 with 61 was carried out with chromium (II) chloride and nickel (II) chloride to yield 66 as a diastereomeric mixture. Removal of the TBS protecting group and acetalization of the 3,5-diol gave 5 7 and 67. The coupling of 6 5 with the (9i?)-iodide followed by similar functional transformations yielded (2S, 3R 4JR, 5S, 14i?)-57. Li/f-BuOK H2N(CH2)3NH2
n-BuLi/BF30Et2 THF
55
54
88%
''O
80% 1.Et02CN=NC02Et Ph3P/PhC02H 2. KgCOs/MeOH 3. TBSCI/imidazole DMF
59 80% 1. n-BuaSnH AIBN 2. I2/Et20 87%
HO-^
NHAc
Br2/H20 41%
56
unlloH lYoNHAc
1.PhCH(0Me)2 HBF40Et2/DMF
o=° —" 62
2. TBSOTf/CH2Cl2 2,6-lutidine 64%
471 P h - < JOTBS
QNHAC
\QF=O 63
Pd(OH)2 cyclohexene »^ EtOH 85%
H0-| lOTBS HOH QNHAC
Nal04/H20 CH2CI2
64
95%
OTBS
OHCl QNHAC
Llr° 65
61
+ 65
OH OTBS
CrCl2/NiCI;
,vNHAc
1. aqHF/MeCN 2. Me2C(OMe)2
^
TsOH/DMF
67 31%
Scheme 9
W) From D-Mannltol Another efficient and simple approach, adopted by the Mulzer group,2i was based on the ready availability of (i?)-2,3-0-isopropylidene glyceraldehyde 68 from inexpensive D-mannitol in multigram quantities (Scheme 10). After the usual Wittig olefination and selective protection, 68 was converted into the mono-benzoate 7 0 , which was subjected to a Mitsunobu reaction with triphenyl-phosphine, diethyl azodicarboxylate and phthalimide to give the phthalimide olefin 71. Cis-Hydroxylation with OSO4 converted 71 into a 2:1 mixture of diols 72a and 72b. Successive removal of the protecting groups of 72a yielded D-ribo-C is-phytosphingosine 22.
472
Ph3P=CHCi4H29-n
^ '
-6
80%
O^'V^ V-6
1.60%HOAc 2N H2SO4 BzO C14H29
OH
2. BzCI/py 82%
70
69
68 PhgP/DEAD
, - Y ^ C14H29
PhthNH
^R R2
OSO4
BzO' PhthN
»>
BzO
A/-methylmorpholineA/-oxide
77% ^^
C14H29
PhthN ^R R""
90%
72/) R^ R^ « H; R \ R"* = OH OH
OH KOH 79%
C14H29
BZO^^YV^''^^
85%
NH2 OH
NH2 OH 22
73
Scheme 10
(e) From D-Xylose Two novel sphingosine-derived azetidine alkaloids, penaresidins 6, were isolated as potent actomyosin ATPase activators from the Okinawa marine sponge Penares sp.^ The first synthesis of straight-chain model compound 74 from D-xylose was developed by us.22 Retros3aithetic analysis indicates that D-xylose is the most suitable candidate for assembling the requisite stereochemistry (Scheme 11). PHNWOMs
H HO^ -^^^'
P^O
BnO^^^A^
Y^^"H
OBn
OH
BnO-
P^O
foBnVOH
OBn
OH 75
Scheme 11
C12H2! OBn
74
BnO>.^^As^CHO
OP^
BnO.
473
Scheme 12 outlines the synthesis of the straight-chain model compounds 82. The Grignard reaction of 76, derived from 3,5-di-O-benzyl-Dxylose 7 5 , with dodecylmagnesium bromide in the presence of dilithium tetrachloro-cuprate afforded a 1:1 mixture of syn and anti adducts 7 7 in good yield. Because the diastereoselectivity was very bad, the adducts 7 7 were reoxidized and reduced with zinc borohydride to give, after acetylation, exclusively the desired acetate 78 [antvsyn = 95:5). It is well known that the reduction of a-keto alcohols with Zn(BH4)2 leads to anti-diols with good selectivity.23 Selective deprotection, mesylation and azidation gave the azide 79 . Reduction of 79 gave the requisite Cie-phytosphingosine 8 0 in quantitative yield. Construction of the azetidine ring began with mesylation giving the N,0dimesylate, which upon treatment with sodium hydride afforded 81. Unfortunately many attempts towards reductive cleavage of the mesyloxy group were unsuccessful, so the N-protection with a removable group, Omesylation and cyclization sequences were tried (Scheme 13). Both routes gave the pure azetidines 83 and 85, respectively. However, deprotection of both N and O protecting groups followed by acetylation always gave isomers 84a and 84b. in a 1.5:1 ratio. Comparison of the iH NMR spectra of natural penaresidin acetates (crude) and that of the synthetic triacetates 84a and 84b showed good resemblance in chemical shifts, coupling constants and the ratio of the isomers. This isomerization seems to occur, when both the NH group and the C(3)-OH group are acetylated, through intramolecular nucleophilic attack giving an azabicyclobutyl ion-pair or an azetidinyl ion-pair, which is then attacked by an acetoxy ion from the both sides of the cation with nearly equal probability of aiffording the isomers (Scheme 14). BnO-i
o
PMBO
fcBnVOH
BnO.
OHO
53% OH
^>
2. Zn(BH4)2 3. AC20/py/DMAP
PMBO
OAc
BnO.
C12H25
OBn 78
BnO.
C12H21
OBn 77
76
1. (C0CI)2/DMS0 EtaN
OH
PMBO
».
94%
OBn
75
81%
Ci2H25MgBr Li2CuCl4
1.DDQ 2. MsCI/py 3. LiNa/DMF 69%
BnO,
474
LiAIH4
1. MsCI/py/DMAP 2. NaH/DMF ^ 62%
NH2 OH ^
BnO.
100%
OBn
BnO-
H ^
ivia 1
WC12H25
-^^*^Y^'"H OBn
80
5r
Ms 1 1. Na/liqNHg ^ 2. ACgO/py/DMAP
^V/ \A^^2^2S AcO--^^'^v/^"H OAc
68%
82
Scheme 12 1.TsCI/py/DMAP 2. MsCI/py/DMAP 3. NaH/DMF 49%
Ts 1
H ^ v \>Ci2H25 BnO^x^-^X/^'/^
1.Na/liqNH3 2. Ac20/py/DMAP 62%
Ac 1
H ^ y v^Ci2H25 ACO-^N*'Ay/^/^
OBn
OAc
83
04a
Boc
Ac 1
on
OU
+
BnO'^o*'%/^'/u
£
T
1.(Boc)20/NaOH 2. MsCI/py/DMAP 3. NaH/DMF 72%
H ^ v v>Ci2H25 Ac0^x^"^\/^''l_j
>H25
OAc
1.TFA 2. H2/Pd-C 3. Ac20/py/DMAP 50%
OBn 85 or
04^
1.TFA 2. AC20/py/DMAP 3. H2/Pd-C/H* 4. Ac20/py/DMAP 23%
Scheme 13
" Ac I
"
^
fteC^Oi Ac
...<|>„. _
R' <'I'>-R^ H" O A C
AcO H
Ac I
R^
O^R'
Acd H Scheme 14
Ac
=^ R^ O^^' AcO H
475
(f) From L-Ascorbic Acid New ceramide digalactosides were isolated by Hayashi's group from the marine sponge Halichondria japonica.'^^ To the major glycosphingolipid was assigned the structure 86, except for the stereochemistry, using FAB/MS, IR and iH NMR spectroscopy. The first synthesis of the ceramide 8 7 and therefore the structure determination were described by us.25
"T
^cT/OH OH
OH
0=
NH OH OH 86
The synthetic strategy (Scheme 15) for assembling the phytosphingosine 88 was the stereoselective ring opening of the epoxide 90 with 2-alkyl-2-lithio1,3-dithiane A, followed by functional transformation to give a precursor B. The requisite consecutive stereochemistry of the target intermediate C could be obtained by means of Dondoni's protocol^G using 2-trimethylsilylthiazole 89 (see p. 2?). (2R)-Hydroxydocosanoic acid 91 could also be derived from the epoxide 90. The synthesis began with the treatment of 3 , 4 - a n h y d r o - l , 2 - 0 isopropylidene-D-erythritol 9 0 with 2-alkyl-2-lithio-l,3-dithiane 92 to give 93 (Scheme 16). Reductive desulfurization of 93 and transacetalization of the resulting 94 by the following reaction sequence (1. acidic hydrolysis; 2. protection of the primary hydroxyl group; 3. ketalization; and 4. basic hydrolysis) afforded the primary alcohol 95. The Swern oxidation of 9 5 yielded the aldehyde 96. To create a new chiral center at C(2) and to introduce a hydroxymethylene group simultaneously, Dondoni's method^S was used. Treatment of the aldehyde 96 with 2-(trimethylsilyl)thiazole 89 afforded the highly diastereoselective adduct 97, but it had an undesired configuration at the new chiral center. The high diastereoselectivity is attributed to the preferred transition state D. Attempts to invert the configuration at C(l) were fruitless, owing to severe steric hindrance. So a two-step oxidation-reduction sequence was used to obtain the C(l)-C(2) syn product. Reduction of the ketone 98 with NaBH4 in the presence of CeCls afforded the best result (99:97 = 85:15). The
476 preferred formation of 99 may be rationalized based on the transition state E by assuming that complexation of the cerium ion occurs between the O atom of the isopropylidene group and the N atom of the thiazole ring and that the hydride ion attacks from the less hindered side. A series of protection, methylation, reduction, hydrolysis and reduction steps provided the alcohol 100, which was subjected to transformation into the alcohol 101. Final conversion of the alcohol 101 into phytosphingosine 102 was carried out in the usual manner. The EDC-mediated condensation of (2K)-benzoyloxydocosanoic acid^^ and the amine 102 followed by debenzylation gave t h e ceramide monobenzoate 103. OH
NHR O H '^{CH2)iiCHMe2
HOOC''^C2oH4i 91
87
R = (2R)-C2oH4iCH(OH)CO
88
R=H
OH OP^ P^O.
(CH2)iiCHMe2 OP^
HOOy^C2oH4i OH
C
OP"*
H
>-SiMe3
89
+
OHC.
"Y'^(CH2)iiCHMe2 p3 OP^
V-0
(CHgjgCHMea
90
B
OH
Scheme 15
A
- ^
OH
L-ascorbic acid
477
Raney Ni (CH2)9CHMe2
HQ (CH2)9CHMe2
I.PTSA/MeOH 2. PivCI/py 11—^
HO-^ (CH2)iiCHMe2 )—( ^ ^
3. Me2C(OMe)2/PPTS 4. LiOH
(CH2)iiCHMe2 W
o
^
o
1.lL/>-SiMe3 " 75 2.TBAF
97%
(CH2)iiCHMe2
O
O
O
99%
NaBH4 CeClgyHgO
O
4. CuCl2/CuO
X
^ ^
1.PTSA/MeOH 2. BnBr/NaH
1. PMBCI/NaOH 2. Mel 3. NaBH4
^S OH || / ) — ( (CH2)iiCHMe2 o
3. DDQ
^-^^^^^ 9d
61%
OH OBn o n 1 ^ ^""-^ ^ ^ " ^ ( C H 2 ) i i C H M e 2 "^g^
81%
^MsCI ^' '"'^^ 3. LiAIH4 56%
100
101 QBz
NH2 OBn BnO A J^ ^^•-^^Y"^(CH2)iiCHMe2 OBn
^
gj
90
/ ^
(C0CI)2/DMS0 Et3N
^
96%
P'^'^^ H2)iiC H 0 H 2^C- -- ^^ ( ( C (CH2)iiCHMe2 0
o o
MeOH
X
^ ^
(CH2)nCHMe2
86%
95
^S p [I / ^ - ^
95
| ^ W ^N W
^
(C0CI)2/DMS0 EtgN 1 ^
V / \
64%
\0
89%
1. (2R)-benzoyloxydocosanoic acid ^^^ ^ 2.H2/Pd-C 84%
Ov^^A^ 7 (CH2)9Me NH OH HO X A ^^^^^^Y'^(^^2)iiCHMe2 OH 103
Scheme 16
478
SiMea
(g) From (S)-Malic Acid Guanti's group^s has developed a new "electrophilic amination" method (Scheme 17) for the P-hydroxyester 107, derived from dimethyl(S)-malate 105, with di-tert-butylazodicarboxylate 108. By performing the reaction at -50 °C a moderate selectivity {erythro:threo = 67:33) was observed. The two isomers 109e and 109t were easily separated and 109e was converted into the N , 0 isopropylidene acetal 110. Reduction of 1 1 0 with calcium borohydride and successive protection and deprotection gave the primary alcohol 112, The Swem oxidation of 1 1 2 followed by treatment with lithium tetradecyne in the presence of HMPA preferentially yielded the anti adduct {113a:113s = 85:15). The alkyne 113a was converted by the u s u a l reaction sequence into tetraacetyl D-ribo-Cis-phytosphingosine 104. OH
A^CO, Me
BH3/NaBH4{cat)
Me02C
OH HO^^^^As^COgMe
83%
105
OH TBDRSO^^^A^COaMe 107
''•LDAn-HF
109B
71%
^.
71%
106
OH
OH T B D P S O ^ ^ ^ ^ C O J Me
TBDPSOs^^^x^s^COaMe
Boc'' "NHBoc
Boc'^^'NHBoc
2. BocN=NBoc 108 62%
MeOC(CH3)=CH2 PTSA
f-BuPhgSiCI imidazole
109t
109e
NaBHVCaClg
TBDPSO, COaMe 110
0^N'^'"= TBDPSO.
N.
90%
Boc
OH 111
479
1. MEMCI/EtsN 2. n-Bu4NF
V
86%
i
xBoc
T "°^ ^OMEM
rr2
X.
^I
, ^ ,Boc ^I
A..
I
1. (C0CI)2/DMS0 EtgN
?
2. n-Ci2H25CECLi THF-HMPA 71%
LAcOH/HCI 2. ^. H2/Pt02 n2/riU2
T T OH
7
^Boc
^°^ K^^^^
113a : anti 113s : syn
_ . .,L,. ^'^'^ 1 1^^"^
3.AC2O
^^ ^^
rrja
A 104
Scheme 17
(h) From (S)-Serine Diastereoisomeric D-ribo, D-Iyxo-, D-arabino-, and D-xyZo-Cie-phyto sphingosine tetraacetates were S5aithesized from the oxazoline 114, derived from (S)-serine, by Komori's group29 (Scheme 18). Treatment of the alkenylalane 115, prepared from n-tridecyne and DIBAH, with the oxazohnecarbaldehyde 114 gave a mixture of 116a and 116b. After chromatographic separation, the allylic alcohol 116a was then oxidized with vanadyl acetylacetonate and tert-butylhydroperoxide to give a mixture of the diastereomeric epoxy alcohols 117a and 117b in a 3:2 ratio. The epoxy alcohol 117a was reduced with DIBAH to give an inseparable mixture of 1 1 8 a and 118b. Debenzylation and acetylation gave D-ribo-Ciephytosphingosine tetraacetate 119a and its isomerl 1 9 b . Other diastereomers of phytosphingosine were synthesized by a similar method.
r^ HO. L-serine
^
'^V > =MN ^^^
115
480
(CH2)ioCH3 OH 116b
17%
V0(acac)2 TBHP Ph.
Ph.
OH 117b
33%
DIBAH
BnHN
BnHN H O \ . A s ^ - - v . ^ (CH2)ioCH3
OH ^(CH2)iiCH3 OH
OH
118a
118b
OH
1. Pd-C/cyclohexene/HCI 2. Ac20/py
AcHN AcO.
AcHN
OAc (CH2)iiCH3
+
AcO.,^A.,^-s,^(CH2)ioCH3 OAc OAc 119b 24% from 117a
Scheme 18
4.3 Chiral Induction Obtaining optically active compounds by chiral induction represents a more refined solution to organic chemists. The kinetic resolution developed by Katsuki-Sharpless^O for sdlylic alcohols is superior in enantiotopic face differentiation and in versatility. Another interesting chiral induction method has been developed by Dondoni^i using 2-(trimethylsilyl)thiazole as a masking formyl group. These methods are
481 more efficient because stereoselectivity.
the desired isomer is obtained
with
high
(a) Katsuki-Sharpless Epoxidation The first asymmetric synthesis of a phytosphingosine was accomplished by Komori's group^^ (Scheme 19). Racemic 120 was kinetically resolved by asymmetric epoxidation using (+)-diisopropyl tartrate as a chiral auxiliary to give the (4R)-allylic alcohol 121 and the (4S)-epoxy alcohol 122. Protection of 121 followed by ozonolysis gave the aldehyde 123. The aldehyde 123 was then converted into the epoxy alcohol 126 by means of 1. Homer-Emmons reaction, 2. DIBAH reduction and 3. Katsuki-Sharpless epoxidation. After conversion of 126 into the benzyl urethane 227, it was treated with sodium hydride to give the 2-oxazolidinone 128 via intramolecular base-catalyzed epoxide opening. Subsequent hydrolysis of 128 followed by cleavage of the benzyl and the MOM ether group gave the Cie-phytosphingosine 129. They also reported the synthesis of acanthacerebroside A^s 134^ which was isolated from the starfish Acanthaster planci.^^ DCC-mediated condensation of 129 and (R)-2-acetoxytetradocosanoic acid 130 afforded the ceramide monoacetate 131. The glycosidation of 131 v/ith 2,3,4,6-tetra-Oacetyl-p-D-glucopyranosyl bromide 132 in the presence of silver triflate and molecular sieves gave a mixture of the p-monoglycoside 133 and a diglycoside in 37 and 15% yield, respectively. Hydrolysis of 133 yielded acanthacerebroside A 134. Ti(0Pr^4/(+)-DIPT TBHP n-Ci2H25
^-012^25
+
n-Ci2H25
120
1.MOMCI 2. O3; Me2S 121
OMOM n-Ci2H25 ^
"CHO
OMOM
(EtO)2P(0)CHC02Et ggo/^ ^^Q^ ^21
^-^i2H25 124
123
DIBAH 82%
OMOM
^.
^OH ^-C 12^25 125
C02Et
Ti(0PrV(-)-DIPT TBHP ^ 63%
OMOM n-Ci2H25 126
482 OMOM NaH 72%
97% r27
OMOM, n-Ci2H25
MOMO
-'/AN ^
[I 0
Bn
n-C,2H25
A
Bn.
P
N--\
f ^ OH
^
t25 LNaOH 2. Pd-C/cyclohexene HCI
OH n-Oi2H25
3. HCI
NH3 CI
y
^^
OH
59%
t29
Scheme 19 OAc HoAy.C22H46 CI HgN
OAc 130
OH
HO.
(CH2)iiCH3 OH
NH
73%
AcO-
131
] ^ (CH2)2lCH3 ^,^ Q,^
AcO-j
/-O
^AcV AcO^^ OAc
AgOTf 37%
o X
A
^^-'"^Y'^^^"2)iiCH3 OH 133
OH "^f
1 ^
^(CH2)iiCH3 OH
OAc
t32
(CH2)2iCH3
VO-^x^Y^(CH2)iiCH3
HO^'—f OH
OH
H0>.
DCC/HOBT
129 -OBr DAcS AcO , OAc
Y'^^^^2)2lCH3
OH 134 Scheme 20
K2CO3
483
(b) Dondoni Carbon Chain Extension Method Chain elongation of aldehydes via 2-(trimethylsilyl)thiazole (2-TST) 135 involves two key steps: A, construction of a chiral hydroxyalkyl chain at C(2) of the thiazole ring (functionalization); B, the liberation of the aldehyde by cleavage of the thiazole ring (unmasking). The mechanism of the functionalization step is shown in Scheme 21.30 R
R
r-N'^OSiMes
RCHO
RCHO
,. - N
OSiMes
^S^SiMea 135
136
137
i
/ R
r\
\
R N-^0SiMe3
- RCHO
^g-^^^^OSiMea
R
R
p
138
140
139
Scheme 21
The initial reaction is 1,2-addition of the carbon-silicon bond of 1 3 5 to an aldehyde giving the thiazolium 2-ylide 136. The 2-ylide 136 may then react with a second aldehyde molecule to give a 2:1 adduct 137 which in turn react with a third aldehyde molecule to give a 3:1 adduct 139. Silyl migration of 1 3 7 followed by removal of an aldehyde molecule would yield the 2substituted thiazole 140. Dondoni's group^S described the synthetic utility of 2-TST for various Nprotected a-amino aldehydes. Application for the synthesis of phytosphingosine is outlined in Scheme 22.
y.„-Boc OHO 141
BU4NF ^S^SiMea 135
85% OSiMea
484 \ /
^Boc
\ /
,Boc
OSiMes
OSiMea
142a
1^2b 1.Mel 2. NaBH4 3.HgCl2
I
65% 1. BnBr/NaH
-hn:
Boc
BU4NI
0^>v,.CHO ^^^ y OH
Boc
Ov
'
1. BnBr 2. Mel
143
^1 ^ ^ N
2.2-TST 64%
^
S-N
QBn CHO
Boc
OH
+
N. i V // OBn S -
0
144a
n-Ci3H27PPh3 Br
>J .Boc 7-N OBn
Raney Ni 'C12H21
n-BuLi
70%
66%
-N
Boc
OH C14H2
I.CF3CO2H/H2O 2. Ac20/py
AcHN _
OAc _ C14H29
57%
Scheme 22
The reaction between equimolar a m o u n t s of 2-TST and N-tertbutoxycarbonyl-L-serinol acetonide 141 occurred smoothly at room temperature to give, after desilylation of the resulting adducts with tetrabutylammonium fluoride, a separable mixture of amino alcohols 142a and 142b in high diastereoselectivity [142a:142b = 92:8). This high anti diastereoselectivity may be attributed to the Felkin-Anh open-chain model for as3mimetric induction,^^ and the preferred transition state is presented in A.
485
A
The unmasking protocol consists of three sequential operations: Nmethylation, reduction, and hydrolysis. Thus 142a was converted into the aldehyde 143, which after protection was subjected to a further one-carbon homologation. The addition of 2-TST in dichloromethane at room temperature was rather unselective (ds = 60%) but became quite diastereoselective by using tetrahydrofuran at 0 °C giving adducts 144a and 144b in an 85:15 ratio. The anti configuration of the major isomer 144a is again consistent with the non-chelate Felkin-Anh model for diastereoselection. Protection and unmasking provided the aldehyde 145. The synthesis of D-ribo-Cis-phytosphingosine 104 was achieved by a series of Wittig olefination, reduction, hydrolysis and acetylation steps. (c) Catalytic Asymmetric Aldol Reaction Kobayashi's group^^ developed a new enantioselective synthesis of Cis phytosphingosine using catalytic asymmetric aldol reactions as a key step (Scheme 23). The key catalytic aldol reaction of acrolein with the ketene silyl acetal 148 derived from phenyl a-benzyloxyacetate was carried out by using tin(II) triflate, chiral diamine 149, and tin(II) oxide. The desired aldol product 150 was obtained in high diastereo- and enantio-selectivities [aywanti = >98:<2, 96% ee (syn)]. After reduction and protection of 150, the resulting 151 was oxidized with m-chloroperbenzoic acid to give preferentially the aepoxide 152 (a:p = 74:26). Regioselective ring opening of 152 by the Grignard reagent and a catalj^ic amount of copper iodide gave the alcohol 153. The alcohol 153 was then converted into the azide 154 by the following reaction sequence: 1. protection, 2. debenzylation, 3. mesylation and 4. azidation. After the removal of the MOM and the acetonide groups of 154, the azide group was reduced to afford the phytosphingosine 104. BnO,
OSiMe3 148
OPh
Sn(OTf)2/SnO ry^^^_r\ ^ N " ^ H \==/
9^
9
l ^^" 150
I.DIBAL 2.Me2C(OMe)2 TsOH 78%
486
—1^
96%
OBn
0 ^ 0
OiaHayMgBr
m-CPBA
Cul
OBn
151
CiaH; 13"27 OH
97%
152
I.MOMCI/APrjNEt 2. H2/Pd-C 3. MsCI/py 4. NaNa
0
^13^27
76%
MOMO
V
Ng
0
1.ACOH/H2O 2. PhaP/HgO-py
OAc "^OAc
3. ACgO/EtgN DMAP
154
OBn
153
OAc NHAc 104
48%
Scheme 23
4.4 Synthesis of Gangllosides Although glycosphingolipids have been isolated from a wide variety of sources,3S there have been no reports of the occurrence of glycosphingolipids in fish roe. Li's group39 found that the major ganglioside from the roe of striped mullet [Mugil cephalus) had the structure of the oligosaccharide moiety identical to t h a t of the G M 2 gangUoside from the h u m a n brain: GalNAcpl^4(NeuAca2^3)Gaipi-»4Glcpl-^l'Cer 155. This ganghoside, however, differed from brain GM2 in its ceramide portion. The most striking differences are the presence of large amounts of phytosphingosines (over 80% of the total long-chain bases). In spite of important biological activity, only a few sjnithetic studies of these particular gangliosides have been reported.^O HO/OH
A ^ : ^ oO )H
riA
^OH (CH2)iiCH3 OH
<.„,,
HO.
NH
OH /(CH2)i3CH3
OH HO^NHAc
155
Siba's group^i reported the synthesis of ganglioside GM5 156, isolated from eggs of the sea urchin Anthocidaris crassispina as a main component.^^
487 The synthesis starts with commercially available phytosphingosine (Scheme 24). T h u s 2 2 was AT-acylated with n-docosanoic acid Nhydroxysuccinimide ester. After protection of the primary alcohol, the other secondary hydroxyl groups were benzoylated and the silyl ether was deprotected to give the protected ceramide 157. Preparation of the oligosaccharide moiety was carried out with the 2methylthio derivative of N-acetylneuraminic ester 158. The donor 158 was coupled with the acceptor 1 5 9 in the presence of dimethyl(methylthio)sulfonium triflate (DMTST) and molecular sieves to give the disaccharide 160 (a:P = 3:1). After isolation of the a-isomer, deprotection and then Nglycolylated with AT-hydroxysuccinimide ester followed by methylation and acetylation gave the benzyl glycoside 162. The benzyl glycoside 1 6 2 was hydrogenated and then converted into trichloroacetimidate 163. The activated donor 163 was coupled with 1 5 7 to give the protected ganglioside 164. Final deprotection of 1 6 4 afforded the ganglioside 156. 1. CH3(CH2)2oC02NSu EtgN 2. f-BuPh2SiCI
NH2 OH C14H21
22%
22 ACO
PAC
-SMe
AcO'" AcHN
DMTST/MS-4A
+ ^5^_
AcO
OAc
AcO'" AcHN
159 C02Me
AcO
61%
OAc
158
AcO
C14H21
3. BzCI/py 4. n-Bu4NF
OH
AcO AcO
1. NaOMe 2. NaOH
-»>
3. H2NNH2
OBn
63%
OAc
160
1. ACOCH2CO2NSU 2. CH2N2 3. Ac20/py 79%
AcO AcO" ACOCH2COHN
OAc
C02Me
1.H2/Pd-black 2. CI3CCN/DBU
^~
AcO 162
AcO !,^^Si^^^.^:L^OBn AcO OAc
93%
488 AcO AcO«ACOCH2COHN
OAc
COgMe r57/BF3-OEt2
P: AcO
AcO AcO
163
28% AcOA O^NH CCI3
AcO AcO ACOCH2COHN
OAc
C02Me
AcO
0-A AcOAcO
164
0;^C2iH43 N NH AcO
CO2H H0CH2C0HN--^^T'*^ O HO HO HO 156
OBz C14H2
1. NaOMe 2. NaOH
•
85%
OBz
0
OH ^14^2-
Scheme 24
5.
CONCLUSIONS In summary, synthetic methods for controlling the consecutive three chiral centers of phytosphingosines have been described. Some modern synthetic approachs to sphingoglycolipids are also described. Considerable efforts have been devoted to the search for new biologically active sphingolpids, especially from marine organisms. However, because of t±ie difficulty of isolating the pure compounds, the structure elucidation relies largely upon mass spectrometry. Thus the syntheses of sphingolipids gain in importance as well the structure determination and understanding of their functions. Finally I will call attention to novel composite microstructures from lipids and minerals. When a-hydroxy fatty acid galactocerebroside and a mixture of anioic sulfated galactocerebrosides are suspended in ethylene glycol multilayer tubules are produced. When the tubule suspension is then treated with an acidic ferric chloride solution, the tubules became covered with shards of the mineral lepidocrocite (y-FeOOHj.'^^ Although the direct application of this organic-inorganic composite is still uncertain, but it seems to have a possibility of designing new materials.
489 ABBREVIATIONS Ac
acetyl
acac AIBN
acetylacetonate
Bn
2,2'-a2obisisobutyronitrile aqueous benzyl
Boc
t-butoxycarbonyl
n-Bu
n-butyl
i'Bu
iso-butyl
t'Bu
tert-butyl
Bz
benzoyl
DBU
l,8-diazabicyclo[5.4.0]undec-7-ene
DCC
iV,7V'-dicyclohexylcarbodiimide
DDQ DEAD
2,3-dichloro-5,6-dicyano-1,4-benzoquinone diethyl azodicarboxylate
DIBAH, DIBAL
d i i s o b u t y l a l u m i n u m hydride
DIPT
diisopropyl t a r t r a t e
aq
DMAP
4-(dimethylamino)pyridine
DMF
dimethylformamide
DMSO DMTST
dimethylsulfoxide
EDC
N-[3-(dimethylamino)propyll AT-ethyl carbodiimide
Et HMPA
ethyl
dimethyl(methylthio)sulfonium triflate
hexamethylphosphoric triamide
HOBT
1 -hydroxybenzotriazole
LDA liq Me MEM
lithium diisopropylamide liquid methyl p-methoxyethoxymethyl
MOM
methoxymethyl
MS
molecular sieves
Ms
methanesulfonyl (mesyl)
Ph
phenyl
PhthNH
phthalimide
Piv PMB
trimethylacetyl (pivaloyl) 4-methoxybenzyl
PPTS
pjn-idinium p-toluenesulfonate
Pri PTSA, p-TsOH
iso-propyl p-toluenesulfonic acid
490 py SuNOH TBAF TBDPS TBHP TBS TFA Tf TMS Tr Ts 2-TST
pyridine N-hydroxysuccinimide tetrabutylammonium fluoride t-butyldlphenylsiiyl t-butyl hydroperoxide t-butyldimethylsilyl txifluoroacetic acid trifluoromethanesulfonyl trimethylsilyl trityl 4-toluenesulfonyl (tosyl) 2-(triinethylsilyl)thiazole
REFERENCES For a review see: R.R. Schroidt, in: W. Bartmann, K.B. Sharpless (Eds), Stereochemistry of Organic and Bioorganic Transformations, VCH, Weinheim, 1987, pp. 169-189; E.J. Reist and P.H. Christie, J . Org. Chem., 35 (1970) 4127- 4130; H. Newmann, J. Am. Chem. S o c , 96 (1973) 4098-4099; P. Tkaczuk and E.R. Thornton, J. Org. Chem., 46 (1981) 4393-4398; B. Bemet and A. Vasella, Tetrahedron Lett., 24 (1983) 5491-5494; R.S. Garigipati and S.M. Weinreb, J. Am. Chem. S o c , 105 (1983) 4499-4501; K. Mori and Y. Funaki, Tetrahedron Lett., 25 (1984) 5291-5294; K. Koike, Y. Nakahara and T. Ogawa, Glycoconjugate J., 1 (1984) 107-109; M. Obayashi and M. Schlosser, Chem. Lett., (1985) 1715-1718; W.R. Roush and M.A. Adam, J. Org. Chem., 50 (1985) 3752-3757; R. Julina, T. Herzig, B. bemet and A. Vasella, Helv. Chim. Acta, 69 (1986) 368-373; R.R. Schmidt and P. Zimmermann, Tetrahedron Lett., 27 (1986) 481-484; R.H. Boutin and H. Rapoport, J. Org. Chem., 51 (1986) 5320-5327; K. Koike, M. Sugimoto, Y. Nakahara and T. Ogawa, Carbohydr. Res., 162 (1987) 237-246; M.A. Findeis and G.M. Whitesides, J. Org. Chem., 52 (1987) 2838-2848; Y. Ito, M. Sawamura and T. Hayashi, Tetrahedron Lett., 29 (1988) 239240; P. Herold, Helv. Chim. Acta, 71 (1988) 354-362; P. Zimmermann and R.R. Schmidt, Liebigs Ann. Chem., (1988) 663-667; Th. Bar and R.R. Schmidt, Liebgs Ann. Chem., (1988) 669-674; K. Mori and T. Kinsho, Liebigs Ann. Chem., (1988) 807-814; S. Nimkar, D. Menaldino, A.F. Merrill and D. Liotta, Tetrahedron Lett., 29 (1988) 3037-3040; P. Gamer, J.M. Park and E. Malecki, J. Org. Chem.,53 (1988) 4395-4398; K. Ohashi, S. Kosai, M. Arizuka, Y. Yamagiwa and T. Kamikawa,
491 Tetrahedron, 45 (1989) 2557-2570; K.C. Nicolaou, T. Caulfield, H. Kataoka and T. Kumazawa, J. Am. Chem. Soc, 110 (1989) 7910-7912; H. Shibuya, K. Kawashima, M. Ikeda and I. Kitagawa, Tetrahedron Lett., 30 (1989), 7205-7208; A. Dondoni, G. Fantln, M. Fogagnolo and P. Pedrini, J. Org. Chem., 55 (1990) 1439-1446; S. Takano, Y. Iwabuchi and K. Ogasawara, J. Chem. Soc. Chem. Commun., (1991), 820-821; Y. Hirata, Y. Yamagiwa and T. Kamikawa, J. Chem. Soc., Perkin Trans 1, (1991), 2279-2280; J.S. Yadav, D. Vidyanand and D. Rajagopal, Tetrahedron Lett., 34 (1993) 1191-1194; M.A. Petersen and R. Polt, J. Org. Chem., 58 (1993) 4309-4314; B. GuUbert and S.L. Flitsch, J. Chem. Soc., Perkin Trans. 1, (1994) 1181-1191; A. SoUadie-Cavallo and J.L. Koessler, J. Org. Chem., 59 (1994) 3240-3242. 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
K.-A. Karlsson, Annu. Rev. Biochem., 58 (1989) 309. For a review see: S.-L Hakomori, J. BioL Chem., 265 (1990) 1871318716. For review see: T.A. Springer, Nature, 346 (1990) 425-434. S. Hirsch and Y. Kashman, Tetrahedron, 45 (1989) 3897-3909. G. Kawai and Y. Ikeda, Biochem. Biophys Acta, 754 (1983) 243-248. W. Jin, K.L. Rinehart, and E.A. Jares-Erijman, J. Org. Chem., 59 (1994) 144-147. T. Natori, Y. Koezuka and T. Higa, Tetrahedron Lett., 34 (1993) 55915592. J. Kobayashi, J.-f. Cheng, M. Ishibashi, M. Walchli, S. Yamamura and Y. Ohizumi, J. Chem. Soc. Perkin Yrans. 1, (1991) 1135-1137. F. VanMiddlesworth, C. Dufresne, F.E. Wincott, R.T. Mosley and K.E. Wilson, Tetrahedron Lett., 33 (1992) 297-300. J. Gigg, R. Gigg and C D . Warren, J. Chem. Soc. (C), (1966) 1872-1876. T. Murakami, H. Minamikawa and M. Hato, Tetrahedron Lett., 35 (1994) 745-748. J. Gigg and R. Gigg, J. Chem. Soc. (C), (1966) 1876-1879. K. Koike, Y. Nakahara and T. Ogawa, Agric. Biol. Chem., 54 (1990) 663667. R.R. Schmidt and T. Maier, Carbohydr. Res., 174 (1988) 169-179. R.R. Schmidt, Angew. Chem. Int. Ed. Engl., 25 (1986) 212-235. K. Akimoto, T. Natori and M. Morita, Tetrahedron Lett., 34 (1993) 55935596. T. Mukaiyama, Y. Murai and S. Shoda, Chem. Lett., (1981) 431-432. F. VanMiddlesworth, R.A. Giacobbe, M. Lopez, G. Garrity, J.A. Bland, K. Bartizal, R.A. Fromtling, J. Polishook, M. Zweerink, A.M. Edison, W.
492
20 21 22
23 24 25 26 27 28 29 30
31 32 33 34 35 36 37
Rozidilsky, K.E. Wilson and R.L. Monaghan, J. Antibiot., 45 (1992) 861867. K. Mori and K. Otaka, Tetrahedron Lett., 35 (1994) 9207-9210. J. Mulzer and C. Brand, Tetrahedron, 42 (1986) 5961-5968. Y. Yamagiwa, H. Nakashima, T. Hirakl and T. Kamikawa, Sjmiposium Papers of 36th Symposium on the Chemistry of Natural Products, (1994) 752-759. J. Seyden-Penne, Reductions by the Alumino- and Borohydrides in Organic Synthesis, VCH Publishers, Weinheim, 1991. A. Hayashi, Y. Nishimura and T. Matsubara, Bochem. Biophys. Acta, 1083 (1991) 179-186. H. Nakashima, N. Hirata, T. Iwamura, Y. Yamagiwa and T. Kamikawa, J. Chem. Soc. Perkin Trans. 1, (1994) 2849-2858. A. Medici, G. Fantin, M. Fogagnolo, and A. Dondoni, Tetrahedron Lett., 24 (1983) 2901-2904. K. Yamagata, Y. Yamagiwa and T. Kamikawa, J. Chem. Soc. Perkin Trans. 1, (1990) 3355-3356. G. Guanti, L. Banfi and E. Narisano, Tetrahedron Lett., 30 (1989) 55075510. S. Sugiyama, M. Honda and T. Komori, Liebigs Ann. Chem., (1990) 1069-1078. T. Katsuki and K.B. Sharpless, J. Am. Chem. Soc, 102 (1980) 59745976; V.S. Martin, S.S. Woodard, T. Katsuki, Y. Yamada, M. Ikeda and K.B. Sharpless, J. Am. Chem. Soc, 103 (1981) 6237-6240; J.G. Hill, E. Rossiter and K.B. Sharpless, J. Org. Chem., 48 (1983) 3607-3608; R.M. Hanson and K.B. Sharpless, J. Org. Chem., 51 (1986) 1922-1925. A. Dondoni, in: R. Scheffold (Ed.), Modem Synthetic Methods 1992, VCH, Basel, 1992, pp. 377-437. S. Sugiyama, M. Honda and T Komori, Liebigs Ann. Chem., (1988) 619625. S. Sugiyama, M. Honda and T. Komori, Liebigs Ann. Chem., (1990) 1063-1068. Y. Kawano, R. Higuchi, R. Isobe and T. Komori, Liebigs Ann. Chem., (1988) 19-24. A. Dondoni, G. Fantin, M. Fogagnolo and P. Pedrini, J. Org. Chem., 55 (1990) 1439-1446. M. Cherest, H. Felkin and N. Prudent, Tetrahedron Lett., 18 (1969) 2199-; N.T. Anh, Top. Curr. Chem., 88 (1980) 144. S. Kobayashi, T. Hayashi and T. Kawasugi, Tetrahedron Lett., 35 (1994) 9573-9576.
493 38 39 40
41 42 43
S. Hakomori, in: D.H. Hanahan (Ed.), Handbook of Lipid Research, Vol. 3, Plenum, New York, 1983, pp. 1-165. Y.-T. Li, T.A.W. Koemer and S.-C. Li, J. Biol. Chem., 256 (1984) 89808985. K. Kameyama, H. Ishida, M. Kiso and A. Hasegawa, Carbohydr. Res., 209 (1991) C1-C4; K.C. Nicolaou, C.W. Hummel and Y. Iwabuchi, J. Am. Chem. S o c , 114 (1992) 3126-3128; A. Toepfer, W. Kinzy and R. Schmidt, Liebigs Ann. Chem., (1994) 446-464. T. Yamamoto, T. Teshima, U. Saitoh, M. Hoshi and T. Shiba, Tetrahedron Lett., 35 (1994) 2701-2704. H. Kubo, A. Irie, F. Inagaki and M. Hoshi, J. Biochem., 108 (1990) 185192. D.D. Archibald and S. Mann, Nature, 364 (1993) 430-433.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 18 © 1996 Elsevier Science B.V. All rights reserved.
495
New Developments in Brassinosteroid Research G. Adam, A. Porzel, J. Schmidt, B. Schneider and B. Voigt
1.
INTRODUCTION The brassinosteroids are a new class of native plant-growth regulators whose unique
structure as well as interesting biological properties have stimulated a broad range of research activities. ^""^ The first member, named brassinolide, was isolated in 1979 by Grove et al. from rape pollen (Brassica napus) and structurally elucidated by X-ray analysis as (22/?,23/?,24S)2a,3oc,22,23-tetrahydroxy-24-methyl-B-homo-6a-oxa-5a-cholestan-6-one (1). In the bean second internode bioassay brassinolide (1), even in ng amounts, induced growth stimulation including cell division as well as cell elongation.^ In the meantime altogether 36 characterized native brassinosteroids are known (Scheme 1)^'^ of which the very recently discovered five compounds are discussed in section 2. Most of them exhibit, like brassinolide (1), a 7-numbered ring B lactone structure or, as castasterone (3), a common steroidal skeleton with a 6-oxo function. Other structural variations lie in the substitution pattern of rings A/B and/or different alkylation at C-24, the latter reflecting corresponding phytosterols as biogenetic precursors. All hitherto found brassinosteroids possess the 22/?,23/?-diol structural feature. There are hints that about 40 more members exist which are, however, not yet completely characteiized.^ With the isolation of 23-0-(3-D-glucopyranosyl-25-methyldolicho-sterone
(36) and 23-0-p-D-
glucopyranosyl-2-epi-25-methyldolichosterone (37) from immature seeds of Phaseolus vulgaris, the first brassinosteroid glucosides have been detected^, a conjugate type characteristic also of the classical phytohormones.^ Another type of brassinosteroid derivative, teasterone 3-myristate (9), was found recently by Asakawa et al. in anthers of Liliiim longiflorumP Further new conjugated brassinosteroids, detected after feeding experiments as metabolites in cell cultures, are presented in section 4. The native brassinosteroids occur in a broad spectrum of monocots and dicots^'^ as well as in gymnosperms, suggesting an ubiquitous distribution of these plant growth regulators in the plant kingdom. Among them are the important cultivated plants like rape, tea, rice, wheat, rye, sugar beet, Chinese cabbage, beans, sun flower, raddish, pine species and datepalm. Brassinosteroids have also been found in the green alga Hydrodictyon reticularum^^ and in the fern Eqidsentm arvense^K With the exception of roots, they were detected in every part of the plants e. g. in pollen, seeds, leaves, shoots and sheaths. In comparison with vegetative organs
496
Castasterone (3)
Brassinoiide(l)
2-Epicastasterone (4)
3-Epicastasterone (5)
Teasterone ( 8 ) ; R=H Teasterone-myristate ( 9) :R=CO{CH2)i2CH3
Typhasterol (7)
2,3-Diepicastasterone ( 6 )
6-Deoxocastasterone (10)
OH =
OH ^
3-Epi-6-deoxocastasterone ( 11)
24-Epibrassinolide (12)
24-Epicastasterone (13) OH
OH =
3,24-Diepicastasterone ( 14)
28-Norbrassinolide (15)
OH
Dolicholide(17)
Brassinone(16)
OH
Dolichosterone(18)
Scheme 1 Structures of brassinosteroids from plants
6-Deoxodolichosterone ( 19)
497
24-Ethylbrassinone (20)
6-Deoxohomodolichosterone ( 23)
2,3-Diepi-25-methyldolichosterone ( 26)
Homodolicholide ( 21)
Homodolichosterone ( 22)
25-Methyldolichosterone ( 24)
2-Epi-25-methyldolichosterone ( 25)
3.Epi-2-deoxy25-methyldolichosterone ( 28)
2-Deoxy-25-methyldoiichosterone ( 27)
3-Dehydroteasterone (31)
6-Deoxo-25-methyldolichosterone ( 29) OH
Secasterone (32)
6-Deoxo-24-epicastasterone ( 33)
6-Deoxo-28-norcastasterone ( 34)
HO,
.2^=^^H 2-Deoxybrassinolide ( 35)
Scheme 1 (continued)
23-0-P-D-Glucopyranosyl25-methyldoiichosterone ( 36)
HO''
^ ^ ^ : ^ ^ H OH 23-0-(5-D-Glucopyranosyl2-epi-25-methyldolichosterone ( 37)
498 which have only ng-amounts/kg, pollen contains remarkably higher concentrations e.g. in the case of Brassica napits and Pimis thimbergii about 100 |ig/kg of brassinolide (1) was detected.^^ Of special interest is the finding that insect galls of DistyUwn racemosum contain 300 times more brassinosteroid than the leaves of this plant.^^ In regard to the found antiecdysone activity of some brassinosteroids^"^'^^ such an accumulation possibly reflects a protective function in this special case. Biological active brassinosteroids behave as "botanical juvenile hormones" and enhance the growth of young plant tissue and stimulate, in submicromolar concentrations (lO'^-lO'^^M), the metabolic, differentiation and growth processes. ^^-^^ Strong growth promoting effects on higher fungi after exogenous application of brassinosteroids have also been reported.^^ The influence on the root development of mono and dicots seems to be of a more complex nature21'22 where in the case of Picea abies an increase of percent rooting and rooting quality has been observed.^^ Of practical interest are also reports on increased resistence against stress factors like drought, temperature, pathogenic infection, herbicidal injury and salinity after exogenous application of brassinosteroids.^'^"* Their high bioactivity at very low concentrations, non-toxicity as well as their native character are suitable prerequisites for a further development of brassinosteroids as ecologically soft stress-modulators and bioregulators in agriculture and horticulture. The multiple physiological properties of brassinosteroids involve, among others, membrane effects, activation of photosynthesis and specific enzymes as well as interaction with other phytohoiTnones and external signals e.g. Hght, temperature and gravitropism.2'^'^'7'25 with the evidence of uptake and transport^^'^^, localisation of endogenous brassinosteroids^^ and formation of specific proteins after exogenous treatment^^'^^ brassinosteroids fulfil the criteria to constitute a new family of plant hormones.^^ First results on the molecular mode of action including the moleculai* cloning and characterization of a brassinosteroid regulated gene have been published veiy recently by Zurek et al. ^^'^^ In the following recent advances in new brassinosteroid structures, synthesis and metabolism with special consideration of results of our laboratoiy are discussed.
2.
NEW NATIVE BRASSINOSTEROIDS The isolation and structure elucidation of native brassinosteroids is summarized
comprehensively by several reviews and books.^-^ This section describes recent studies on the occurrence of new brassinosteroids found since 1991. 2.1.
Isolation and purification of brassinosteroids Brassinosteroids are extracted preferably by methanol or methanol/ethylacetate, partitioned
between water and chloroform as well as 80% methanol and n-hexane. The purification is carried out by successive chromatographic methods such as silica gel column chromatography, Sephadex LH-20 chromatography, ion exchange chromatography and preparative HPLC. Concerning the
499 Plant material
Extraction with MeOH Concentrate
Partition between H2O and CHCI3
H2O
CHCK Concentrate
Partition between «-hexane and 80% MeOH
«-Hexane
80% MeOH Concentrate
Silica gel chromatography MeOH-CHCl^ gradient system bioassay
Bioactive fraction(s) Bioactive fraction(s) DEA-chromatography
Preparative HP LC bioa ssay
Methaneboronation/trimethylsilylatic >n of the bioactive fraction(s)
Scheme 2 Isolation and purification of brassinosteroids
Sephadex LH-^10 chromatography bioassay
GC/MS analysis
500 extraction and chromatography, several modifications in solvent and gradient systems or choice of the HPLC phases are useful. The application of solid-phase extraction was also described.^^ The total amounts of brassinosteroids in plant material are usually very low (ng - to ^ig level / kg plant material). Therefore, it is necessary to use a sensitive bioassay system for monitoring brassinosteroid containing fractions during chromatography. A typical isolation and purification procedure of brassinosteroids is shown in Scheme 2. The bioassays for detection of brassinosteroids include the bean second-internode test^, the rice-lamina-inclination test^^, the wheat leaf-unrolling test^^ and a radioimmunoassay^'^ which are especially practicable. Among them the rice-lamina inclination bioassay is the one mostly used. Originally developed for auxins, this assay was later found to be highly sensitive to brassinosteroids: Leaf segments consisting of the second leaf lamina and the second lamina joint and sheath from etiolated rice seedlings were excised and floated on distilled water containing an equivalent of the fraction to be tested. The angle of inclination of the lamina joint indicates the presence of brassinosteroid activity, depending on the concentration (Fig. 1). The bioassay profile by preparative HPLC analysis is shown in Fig. 2. 2.2.
Analysis of brassinosteroids Combined gas chromatography-mass spectrometry (GC/MS and GC/SIM) are the most
commonly used microanalytical techniques for analyzing brassinosteroids. For this, the brassinosteroids have to be converted into volatile derivatives which is carried out preferably by methaneboronation of the vicinal hydroxyl groups using methaneboronic acid and pyridine.^^ Brassinosteroids possessing two vicinal hydroxyl groups yield bismethaneboronates leading to characteristic key ions in their EI mass spectra.^^ In case of the 2-deoxybrassinosteroids the methaneboronation is followed by a trimethylsilylation to derivatize the hydroxyl group at C-3. Such combined methaneboronation/ silylation also allows a differentiation between 3-epimeric brassinosteroids, such as typhasterol (7) and teasterone (8), by both GC and MS.^^ GC-CIMS^^ and GC-CIMS/MS"*^ were successfully applied to improve both the sensitivity and the abundance of the molecular ions. Recently, the LC/MS technique was used as an additional method for the structure elucidation of teasterone 3-myristate (9).^ Brassinosteroids were also determined by HPLC, e. g. as bis-9-phenanthrene-boronates using fluorimeuic detection.'^^ 2.3.
Structure of new brassinosteroids From Secale cereale, besides the common brassinosteroids castasterone (3), typhasterol
(7), teasterone (8), 6-deoxocastasterone (10), brassinone (16) and 24-ethylbrassinone (20), the hitherto unknown brassinosteroid 32 could be isolated.'^^ The mass spectrum of this compound obtained by GC/MS after methaneboronation shows a molecular ion at m/z 470 and is very similar to that of the 3,6-diketo compound 3-dehydroteasterone (31) isolated from Distylium racemosum^ and Triticum aestivum.^^ However, the GC retention data are quite different. Compound 32 displays significant key ions at m/z 454, 399, 316, 286, 245, 155 and 138/137 (Table 1, Scheme 3). The base peak at m/z 155 is the typical ion for a saturated brassinosteroid side chain
501
0.1 ppm 24-epi-BR
0.01 ppm 24-epi-BR
Control (H2O)
Fig.l Rice-lamina inclination bioassay (24-epi-BR == 24-epibrassinolide, 13)
Rice lamina angle (grade)
0.01 ppm BR 0.001 ppm BR 0.0001 ppm BR
Control
20
30 Retention time (min)
40
Fig. 2 Distribution of biological activity determined by the rice-lamina inclination bioassay after reversed-phase HPLC of the purified extract of germinated seeds ofRaphanus sativus (BR = brassinolide, 1)
502 with two hydroxyls at C-22 and C-23. The appearance of an [M-0]+-ion at m/z 454 suggests the presence of an epoxy function. In agreement to 31 both the molecular ion and other key ions show a mass shift of 42 amu compared with castasterone (3), Therefore, this compound should also possess only one oxygen function at ring A. This is supported by the ions at m/z 137 and/or 138 also appearing in 3,6-diketo-5a-steroids.^^ For the final identification of 32 suitable reference compounds having a 2,3-epoxy function were synthesized (see, Section 3). The results of the GC/MS investigation of the corresponding 24-epimeric 2a,3a- (38, 39), 2p,3P-epoxy (32, 40) as well as of the 3,6-diketo compounds (31, 41) are given in Table l.'^^ All six reference compounds can be separated by capillary GC. Both the GC retention time (RRj = 0.935 with respect to castasterone) and the mass spectrum of the isolated compound from Secale cereale were found to be in good agreement with the synthesized epoxy compound 32. Therefore, this brassinosteroid could be regarded as (22/?,23/?,24.S)-22,23-dihydroxy-2,3-epoxy-24-methyl-5a-cholestan-6-one for which we proposed the name secasterone (32), representing the first brassinosteroid with a 2,3-epoxy function. Compound
38 32 39 40 31 41
Substitution at Ring A
Configuration atC-24
2a,3a-epoxy 2p,3P-epoxy 2a,3a-epoxy 2p,3P-epoxy
24S 24S 24R 24R 24S 24R
3-0X0 3-0X0
3^e<—< 287 ^ 286 (-H) 138 . 137 (-H)^'
; 260^ 0 j 259 (-H) 245 ^'-H Scheme 3 Mass spectral fragmentation of methaneboronates of the epimeric 2,3-epoxy brassinosteroids 32, 38-40 and the 3,6-diketo brassinosteroids 31 and 41
503
Table 1 GC/MS data of the synthesized 2,3-epoxy- and 3,6-diketo brassinosteroids (as methaneboronates) in order of increasing relative retention time RRt^ Compound
38
32
39
40
3lt>
41
RRt
0.920
0.941
0.950
0.972
0.991
1.020
Key ions (m/z)
Rel. int. (%)
470 M+
80
87
67
100
454[M-0]+ 439 [M-0-CH3J+
29 31
16 11
28 31
17 6
426[M-0-CO]+
12
5
11
4
49
42
399 (C23/C24)
2
9
10
12
8
8
316 (C20/C22)
17
26
23
31
22
14
12
11
13
8
11
12
9
10
16
8 8
7
245 (C14/C15)
29
24
33
17
23
19
155 (C20/C22)
100
100
100
93
100
100
29
4
22
13
7
11
287 (C17/C20) 286 260 (C15/C16) 259
138 (RingB) 137
7
^ Relative retention time with respect to castasterone (R^ = 10.98 min); ^ from ref. 45. The seeds of the Leguminosae Ornithopus sativus Brot. contain the two brassinosteroids castasterone (3) and 24-epicastasterone (13). This was not only the first detection of 13 in a higher plant but also the first described co-occurrence of these two 24-epimeric compounds.'^'^ An investigation of young shoots of this plant led to the detection of two new 6-deoxobrassinosteroids (33 and 34) besides the known 6-deoxocastasterone (10)."^^ The bismethaneboronate of compound 33, eluted later than 10 in the GC, displays the same M+-ion at m/z 498 in the EI mass spectrum as 6-deoxocastasterone (10). Also the other fragment ions are in good agreement. The final proof of the structure being that of 6-deoxo-24-epicastasterone was furnished by direct GCMS comparison with an authentic sample synthesized from 24-epicastasterone (see, Section 3). The presence of 6-deoxocastasterone epimers was also indicated in grains of Triticum aestivum, one of them probably being identical with compound 33.^^ The bismethaneboronate of the second new 6-deoxobrassinosteroid 34 from Ornithopus sativus showed, besides a molecular ion at m/z 484, significant key ions at m/z 469, 343, 313,
504
m/z 498
0
111111111111111 [ 11111
I » I I 1 1 I I •! 1 1 1 1 I I I n
1 1 1 1 I 11 1 1 1 r r 1 1 1 1 1 1 1 I M 1 1 1 1 I 1 1 I I 1 1 1 I > 1 1 I I I 1 1 1 1 I i r 1 1 I I I I I I I I n
7.417
TIC 8.334 0
11' 11111111111111111111 r 111111 n 11 n n u 111111 > 1111111111111111111 M11111111111 n 11 T< n f 1111111111111 R T
7.000
7.200
7.400
7.600
7.800
8.000
8.200
8.400
8.600
Fig. 3 GC/MS ion profile of bismethaneboronates of 6-deoxo-24-epicastasterone (33) and 6-deoxo-28-norcastasterone (34) isolated from shoots of Orniihopus sativus Brot.
8.800
9.000
505 100
1273
%
Me-B
50
79
273 ^"-H
484 67
97 205
108 121
50
100
m
213 8288
[145
200
150
469
1313 319 ^^^ \ / 343 367
228
414 / 427
tAfc/lHlMli/"\i^t M v'/«-r . . M l ^ ' . ' k - V /' , , I •.'II 250
300
400
350
450
500
Fig. 4 EI mass spectrum of the bismethaneboronate of 6-deoxo-28-norcastasterone (34)
100
156
%
156
376
Me
/
404 (-H)
••.
o'\
490 ^
195
TMSO""
50 i
• 332 211 Ar-'-H
545 85 75 95 121 H07I139
490 96
177
ilii# m
50
100
150
200
211 I
332
287 316 I
375^04
531
476 470 I 440 I,
60 V'T^'^'^
250
300
350
400
450
500
_L
550 m/z
Fig. 5 EI mass spectrum of the methaneboronate/trimethylsilyether of 2-deoxybrassinolide (35)
m/z
506 156
100 %
TMSO'"'"
50 J
195 95 85il 121 73 I I1071 139
50
II
100
545 476 A90
177
m
150
332
211
375 404
269 287 316 I
l;ll.ltlLji.i^.r,ili.^i,.^ .^. yil.
200
250
y-t
300
470 440 ^, ,
531 560
4^
"H f
350
400
450
500
550 m/z
Fig. 6 EI mass spectrum of the methaneboronate/trimethylsilyether of 2-deoxy-24-epibrassinolide (42)
545
TMSO'
213 227
/ 150
200
269 297 250
300
391 350
400
470 450
531 500
550 m/z
Fig. 7 EI mass spectum of the methaneboronate/trimethylsilyether of 2-deoxy-3,24-diepibrassinolide (43)
507
288, 273, 205 and 141 (Figs. 3 and 4). Prominent ions at m/z 288, 273 (base peak) and 205 are characteristic of 6-deoxobrassinosteroids with a 2,3-diol moiety in ring A.^^ j ^ g jon at m/z 141 indicates that 5 exhibits hydroxyls at C-22 and C-23 but no methyl at C-24 in the side chain. Both the molecular ion and other key ions in 34 showed a mass shift of 14 amu compared with 10 and 33. Therefore, compound 34 is proposed to be 6-deoxo-28-norcastasterone. We found another new brassinosteroid 35 in the seeds of Apium graveolens. The mass spectrum of the methaneboronate/trimethylsilyl derivative of this compound shows a molecular ion at m/z 560. Both the M+-ion and the key ions at m/z 545, 531 and 470 (loss of methyl, ethyl and trimethylsilanol, respectively) appeared with a mass shift of 16 amu compared with typhasterol (7) and teasterone (8), indicating an additional oxygen function. The key ions at m/z 404 and 156 characterizing the side chain are complementary ions arising by cleavage of the bond C-20/C-22 (Fig. 5). The ion at m/z 332, also appearing in the EI mass spectrum of the bismethaneboronate of brassinolide (1), represents a key ion for B-homo-6a-oxa lactonetype brassinosteroids with hydroxyls at C-22 and C-23 as well as a methyl at C-24.39 Further important key ions appear at m/z 375 / 376 (cleavage C-17/C-20), 211 (ring B-cleavage), 195 and 177. The GC retention data and the EI mass spectrum of the methaneboronate/silyl derivative of compound 35 were compared with those of synthesized 2-deoxy-24-epibrassinolide (42) and 2-deoxy-3,24-diepibrassinolide (43) (see, Section 3). Compound 35 is eluted earlier than compounds 42 and 43. The difference in the GC retention data of 35 and the 24/?-conrigurated 42 (relative retention times RR^) is typical for other 245/24/?-epimeric brassinosteroids.^^'^3''^5,48 -r^g gj xn2i^^ spectrum of the methaneboronate/trimethylsilyl derivative of 35 is consistent with that of 42, but quite different from 43 (Figs. 5-7). Therefore, the new brassinosteroid can be regarded as 2-deoxybrassinolide (35). A similar methodology was used for the identification of the new member homoteasterone (30) from seeds of Raphanus sativus.^^ The distribution of brassinosteroids in plants investigated since 1991 is summarized in Table 2. Among them six new members along with teasterone myristate (9) from the anthers of Lilium longifoliurrfi, and a new type of brassinosteroid conjugates could be identified. The occurrence of brassinosteroids in species of several plant families hitherto not investigated has been verified. Our investigations also showed that 24epicastasterone (13), firstly found only in the green alga Hydrodictyon reticulaturn}^, is widely distributed in higher plants. It represents the only brassinosteroid in Phoenix dactylifera.^^
3.
SYNTHESIS OF NEW BRASSINOSTEROIDS The original structure of brassinosteroids and the requirement for reference compounds and
sufficient amounts of brassinosteroids for biological studies has tremendously stimulated the synthesis of such phytohormones and their analogues up till now.2.4»60-62 Much efforts have especially been focussed to developing convenient and effective methods for constructing the brassinosteroid side chain with (22/?,237?)-diol function, which is essential for a high bioactivity.^^ Starting from suitable phytosterol precursors with a A^^ double bond the alkyl
508 Table 2 Distribution of brassinosteroids detected since 1991
Plant family
Brassinosteroids
Apium graveolens (seed)
Umbelliferae
2-Deoxybrassinolide (35)
52
Beta vulgaris (seed)
Chenopodiaceae
Castasterone (3) 24-Epicastasterone (13)
53
Cassia torn (seed)
Leguminosae
Brassinolide (1), Castasterone (3), Typhasterol (7), Teasterone (8), Brassinone (16)
54
Daucus carota ssp. sativus (seed)
Umbelliferae
Brassinolide (1), Castasterone (3), 24-Epicastasterone (13)
55
Distylium racemosum (leaves)
Hamamelidaceae 3-Dehydroteasterone (31)
44
Lilium elegans (pollen)
Liliaceae
Brassinolide (1), Castasterone (3), Typhasterol (7), Teasterone (8)
56
Lilium longiflorum (anthers, pollen)
Liliaceae
Brassinolide (1), Castasterone (3), Typhasterol (7), Teasterone-myristate (9) 3-Dehydroteasterone (31)
9,44, 57
Omithopus sativus (seed, shoots)
Leguminosae
Castasterone (3), 6-Deoxocastasterone (10), 24-Epicastasterone (13), 6-Deoxo-24-epicastasterone (33), 6-Deoxo-28-norcastasterone (34)
47,48
Phoenix dactylifera (pollen)
Palmae
24-Epicastasterone (13)
Raphanus sativus (seed)
Cruciferae
Brassinolide (1), Castasterone (3), Teasterone (8), Homoteasterone (30)
Rheum rhabarhanim (panicles)
Polygonaceae
Brassinolide (1), Castasterone (3), 24-Epicastasterone (13)
59
Secale cereale (seed)
Gramineae
Homobrassinolide (2), Castasterone (3), Typhasterol (7), Teasterone (8), 6-Deoxocastasterone (10), Brassinone (16), Secasterone (32)
43
Triticum aestivum (grain)
Gramineae
Castasterone (3), Typhasterol (7), Teasterone (8), 6-Deoxocastasterone (10), 3-Dehydroteasterone (31)
45
Plant species
Reference
(plant part)
51 50,58
509 substituent at C-24 has a significant influence on the stereochemistry of the hydroxylation to 22,23 diols with osmium tetroxide. Thus, the (245)-alkyl substituent directs the hydroxylation prefentially to the undesired (22»S,23iS')-configuration and also starting with (24/?)-alkylated A22precursor more of the corresponding (225,23iS^-diol is formed.^2 Since stereoisomers with the unnatural (22iS',235)-diol function are inactive or less potent growth stimulators, an improved method for obtaining the natural (227?,23/?)-stereoisomers was required. In 1990 an enantioselective method for the osmium-catalyzed asymmetric dihydroxylation of olefins using potassium ferricyanide (K3Fe(CN)6) as the co-oxidant was reported.^ Applying this method with the chiral ligand dihydroquinidine p-chlorobenzoate (DHQD) for the dihydroxylation of the (22£',247?)-24-methyl substituted steroidal side chain a 8:1 ratio of (22R,23R) and (22iS',23»S')-stereoisomers was formed.^^ The reaction was carried out at room temperature in tert. butanol-water 1:1 (v/v) using 1.0-2.0 mmol DHQD, 6.0 mmol K3Fe(CN)6, 6.0 mmol K2CO3, 0.1 mmol Os04 for 1.0 mmol olefin for 4 - 6 days with stirring. Based on this procedure an improved sequence for converting cheap ergosterol (44) to 24epibrassinolide (12), which is the most important brassinosteroid for biological large scale experiments^^, was published (Scheme 4).^'^ Thus, ergosterol mesylate (45) was transformed to isoergosterol (46) which was oxidized with chromium trioxide in pyridine to the corresponding enone 47 . Reduction with lithium in liquid ammonia afforded a cyclopropyl ketone 48 which was isomerized to the A2-6-ketone 49 by heating with pyridinium hydrochloride and lithium bromide in dimethylacetamide at 160° C. Upon hydroxylation of 49, employing the method of asymmetric dihydroxylation with the chiral ligand DHQD, the yield of the desired (22/?,23/?)-stereoisomer 24-epicastasterone (13) was raised to 80 %, besides obtaining 8 % of the (225,235)-epimeric trisepic as taster one (50). The reaction rate was increased substantially in the presence of methane sulfonamide as an additive. Direct conversion of 13 to 24-epibrassinolide (12) was accomplished by Baeyer-Villiger oxidation with trifluoroperoxyacetic acid (CF3CO3H). The overall yield of 24epibrassinolide (12) starting from ergosterol (44) was 26 % in seven steps. To study the biosynthesis and metabolic pathways of brassinosteroid, labeled precursors are required. For such purpose we have developed an effective procedure for labeling 24epibrassinolide (12) with deuterium or tritium (Scheme 5).^^ Thus, reaction of the tetraacetoxy derivative 51 of 24-epicastasterone (13) with 2H2O in dimethylformamide in the presence of triethylamine afforded smoothly the corresponding tris-deuterated compound 52 (R=2H) as the main product. The position of the introduced deuterium is confirmed in the NMR spectrum which lacks the signals for the 5a- and 7a,7p-protons at (6) 2.57 and 2.33, respectively. Baeyer-Villiger oxidation of 52 (R=2H) with CF3CO3H in dichl(6ro)iiethane afforded the 6-oxo-6a-oxa-lactone 53 ( R = 2 H ) besides traces of the isomeric 5a-oxa-6-oxo-lactone, which were separated by Si02chromatography. Treatment of the deuterated 53 with K2CO3 in methanol/water followed by acidification with HCl in tetrahydrofuran yielded the [5,7,7-2H3]24-epibrassinolide (12, R=2H, Scheme 5). This procedure can be used also as a simple method for the introduction of tritium using 3H2O as a labeling reagent. Thus, starting from 51 without isolation of the intermediates the
510
44 R = H 45 R = Ms
OSO4, K2CO3, K3Fe(CN)6 CH3SO2NH2, DHQD t-BuOH/H20
OH
=
24-Epicastasterone (13) (22R, 23R)
Trisepicastasterone (50) (22S, 23S)
H2O2 / (TFA)20 CHCI3
OH
=
H3C0.
24-Epibrassinolide (12)
Scheme 4 Synthesis of 24-epibrassinolide (12) from ergosterol (44)
511 desired [5,7 J-3H3]24-epibrassinolide (12, MBq/mmol was obtained.
R=3H,
Scheme 5) with a specific radioactivity of 222
OAc =
OAc =
AcO. AcO'
AcOi, AcO' 53 R = 2H R = 3H
12 R = 2H R = 3H
o -
HOii,
Oifi./
XoX
HO«''
54 R = H R = 2H R = 3H
13 R = 2H R = 3H
Scheme 5 Synthesis of labeled 24-epibrassinolide (12) In an improved procedure the diisopropylidene derivative 54 (R=H) was used for the tritiation to afford the corresponding 5,7,7-tris-labeIed intermediate 54 (R=3H) which was oxidized under simultaneous deprotection directly to the desired tritiated 24-epibrassinolide (12, R = 3 H ) with a specific radioactivity of 232 MBq/mmol. Therefore, starting from 13 this modification represents a smooth and simple pathway for labeling the biologically important phytohormone 12 in only three steps. On the other hand acid treatment of the diisopropylidene derivative 54 with R = ^H or ^H afforded the corresponding labeled 24-epicastasterone 13 (R = 2H or 3H, Scheme 5).
512 3.1
Synthesis of secasterone and further epimeric 2,3-epoxy brassinosteroids The structural determination of endogenous brassinosteroids, present only in minute
amounts in plant material, requires the availability of corresponding reference standards. Thus, for the final identification of the new brassinosteroid secasterone (32) isolated from Secale cereale (see. Section 2), the four epimeric brassinosteroids with 2,3-epoxy function derived from castasterone (3) and 24-epicastasterone (13), respectively, were synthesized (Scheme 6).^^ For synthesis of both (24/?)-configurated 2,3-epoxides 39 and 40 the 3a,5-cyclo-A22.5. ketone 48 was used as key intermediate. The enantioselective modification of the osmiumcatalyzed dihydroxylation of (22E)-olefm 48 using K3Fe(CN)6 as the co-oxidant and DHQD as the chiral ligand gave 73 % of the desired diol 55 with (22/?,23/?)-configuration. However, direct isomerization of the unprotected diol 55 with pyridinium hydrochloride and lithium bromide in dimethylacetamide led to a ring A saturated 3-chloro derivative. The same reaction starting from the isopropylidenedioxy derivative 56 smoothly afforded the A2-6-keto acetonide 57, which was deprotected with 2 N HCl to give 22,23-diol 58. Epoxidation of 58 with m-chloroperbenzoic acid (MCPBA) afforded, via attack from the less hindered a-side, stereoselectively (22/?,23/?,24/?)22,23-dihydroxy-2a,3a-epoxy-24-methyl-5a-cholestan-6-one (39). For synthesis of the (24^)-configurated 2a,3a-epoxy compound 38 the known*^^ diacetyl ketone 59 was used. Hydrolysis to the (22/?,237?)-diol 60 followed by epoxidation with MCPBA gave 38. To prepare the (24/?)-2P,3p-epoxide 40 the A2-6-keto acetonide 57 was transformed with N-bromosuccinimide (NBS) in dimethoxyethane (DME) to the bromohydrin 62. Acid deprotection to 63 followed by hydrogen bromide elimination with sodium methoxide led to the desired compound 40. Using a procedure similar to the one described for the preparation of the 2p,3p-epoxy compound 40, the known (245)-configurated A^-G-keto acetonide 61^^ was transformed via the bromohydrin 64, deprotection to 65 and HBr elimination, to the (245')-2p,3p-epoxide 32, which was found to be identical with the native secasterone from Secale cereale (see, Section 2). The spectral data of the new compounds are in agreement with the given structures. The observed low field shifts (A 5 + 0.09) of the 19-methyl singlet in comparison to 39 confirm the pconfiguration of the 2,3-epoxy function in compound 40. The same shift was found also for both (24^)-epimers 38 and 32, respectively. 3.2.
Synthesis of 3-dehydroteasterone, 3-dehydro-24-epiteasterone and 6-deoxo-24-epicastasterone 3-Dehydroteasterone (31), the first naturally occurring 3,6-diketo brassinosteroid from
Distylium racemosum and Triticum aestivum, respectively, was synthesized from typhasterol (7)'*5 or teasterone ( 8 ) ^ by oxidation of their con-esponding isopropylidenedioxy derivatives 66 and 67, respectively, with pyridinium chlorochromate and subsequent deprotection (Scheme 7). For the synthesis of the 24-epimer 41, expected also as a native brassinosteroid, the 3,5cycloketone 48 was directly solvolyzed with aqueous H2SO4 to give the 3p-hydroxy-6-ketone
513
48
OR 1.HCI, MeOH
'^H 57 58 59 60 61
(24R), (24R). (24S), (24S). (24S),
62 63 64 65
(24R), (24R), (24S), (24S).
2. MCPBA
R = MesCC R= H R = Ac R=H R = Me2CC[
R = MezCC R= H R = MesCC R=H
OH
°''L
'H 38 (24S) 39 (24R)
32 (24S), secasterone 40 (24R)
Scheme 6 Synthesis of secasterone (32) and further epimeric 2,3-epoxy brassinosteroids
514
68J2 Subsequent Jones oxidation led to the 3-dehydro derivative 69, which afforded, upon asymmetric dihydroxylation, the 3-dehydro-24-epiteasterone (41).69
31
66 3a-OH 67 3P-0H
69
68
41
Scheme 7 Synthesis of 3-dehydroteasterone (31) and 3-dehydro-24-epiteasterone (41) Two new brassinosteroids could be detected from the shoots of Omithopus sativus (see, Section 2). One of them, 6-deoxo-24-epicastasterone (33), was synthesized by us from 24epicastasterone (13) via the corresponding thioacetal 70 and subsequent reductive elimination of the thioketal group by reaction with tri-n-butyltin hydride (Bu3SnH) in the presence of 2,2'azabis-2'-methylpropionitrile (AIBN) (Scheme %).^^ OH ^
13
Scheme 8 Synthesis of 6-deoxo-24-epicastasterone (33)
70
33
515 3.3.
Synthesis of 24-epiteasterone, 24-epityphasterol, 2-deoxy-3,24-diepibrassinolide, and 2-deoxy-24-epibrassinolide To investigate their possible occuirence in plants, we have developed convenient methods
for the synthesis of 24-epiteasterone (71) and 24-epityphasterol (75) as well as their corresponding B-homo lactones 43 and 42, respectively (Scheme 9). For the synthesis of compound 71 the (24/?)-3P-hydroxy-6-ketone
68 was used. Asymmetric catalytic
dihydroxylation of the A^^ double bond of 68 gave the (22/?,23/?)-diol 71 as the main product, besides U'aces of its (225',23iS')-epimer. Baeyer-Villiger oxidation of 71 with CF3CO3H led to a 1 : 0.6 mixture of 2-deoxy-3,24-diepi brassinolide (43) and its 5a-oxa-6-oxo isomer 72, which were separated by preparative HPLC. The corresponding 3a-hydroxy lactone 42 was synthesized from 68 using the Mitsunobu procedure (diethyl azodicarboxylate/triphenylphosphine/formic acid) for inversion of the hydroxy function in position 3J'^ The resulting 3a-formyloxy ester 73, upon hydrolysis afforded the 3aalcohol 74. Asymmetric dihydroxylation of 74 yielded 24-epityphasterol (75) as the main product. Baeyer-Villiger oxidation of 75 led to 2-deoxy-24-epibrassinolide (42) and its isomeric lactone 76 in a 1 : 0.6 ratio.^"^ The spectral data of all new compounds are in agreement with the given structures. Especially the ^H NMR spectra confirm the lactone/isolactone stmctures of 43 and 72 as well as of 42 and 76, respectively. Both isomeric B-homolactone series are easy to differentiate by their characteristic H-5 and H-7 chemical shifts (Fig. 8), whereas in the case of 6-oxo-6a-oxa-lactones the H-5 signal appears as a double doublet at 5 2.86 (43) or 3.18 (42) and that of H-7 (2 H) at 6 4.06 (43) or 4.10 (42), while in the isomeric 5a-oxa-6-oxo-lactones H-5 resonates at 5 4.26 (72) or 4.62 (76) and H-7 at 6 2.48 (72) or 2.49 (76). Also the observed opposite circular dichroism allows a clear differentiation between both the isomeric lactone series (Fig. 9; 6a-oxa-lactones 43 and 42: A£ -h0.201 and +0.265, respectively, at 215 nm; 5a-oxa-lactones 72 and 76: As -0.143 and -0.123, respectively, at 212 nm). 3.4.
Synthesis of A^-T-oxygenated and A^''7-unsaturated brassinosteroids Investigations of the plant extract of celei^ (Apium graveolens) suggest the occun'ence also
of A^-7-oxygenated brassinosteroids.^^ Therefore, we have developed a strategy for the synthesis of such compounds (Scheme 10).''^ Starting from stigmasterol (77) via isostigmasterol (78) the catalytic dihydroxylation of the A^^ double bond led to the (22^,23/?)-diol 79 as the main product, which was isomerized'^^ to the new (22/?,23/?)-22,23-dihydroxystigmasterol (80). Subsequent acetylation of the three hydroxyl groups and allylic oxidation with chromic acid in dichloromethane led to the enone derivative 81, which was hydrolyzed to the enone triol, 82. Reduction of 82 to the 7p-hydroxylated compound 84 was achieved with sodium borohydride in the presence of cerium trichloride in tetrahydrofuran/methanol. On the other hand, the reduction of 82 with L-Selectride in tetrahydrofuran at -78° C^'^ gave the corresponding 7a-hydroxylated allylic alcohol, 83.
516
O
O
68
71 CF3CO3H CHCI3
N—COsEt
II
N—C02Et, PhaP, HCO2H, benzene
OH
=
HO O
75
73 R = HCO 74 R = H
CF3CO3H CHCI3
OH
=
HO''
76
Scheme 9 Synthesis of 24-epiteasterone (71), 24-epityphasterol (75), 2-deoxy-3,24-diepibrassinolide (43), and 2-deoxy-24-epibrassinolide (42)
517
M|]tMjii»Mii»»|iiii|Uiijiiii)iiii|iiii|nnjini|Mii|ini[nii|MM|iiiijMn|Mi»jiiu|iiiijniiiiii>}iiiiiiiii|»Mi|n
4.6
4.2,
3.8
3.4
3.0
2.6
ppm
zll_tizt
M|r»>jNn|
|i>ii|iiii|iinjHii|iiiijiniiiiii|iiii|iiM}iiii|ii
4.6
4.2
3.8
3.4
3.0
2.6
ppm
I L-j^^->j** t»[iiiijiiii|ini}!ii>inii}iiiipMi|ini|iiiijiiii|tiii{Mii]iiMjini|wiijiiii|iiu[uri|iiii{iiii|iiiijii[i|iwijini|n
4.6
4.2
3.8
3.4
3.0
2.6
ppm
HO^
^^ £ O
76 i»inimHi|i»iijnii|iiii|iiMiiiMpir>|iiiijiiMpiM|iiii|iiM{iin|iiii[ini|iiii|iMi|Mii|iin|iiii|Mii|iiii|Uii|ii
4.6
4.2
3.8
3.4
3.0
2.6
ppm
Fig. 8 Low-field region of ^H NMR spectra of the isomeric lactones 76, 72, 42 and 43 (500 MHz, solvent: CDCI3)
518 3.000E+01
I
I
» '
I
'
I
'
•
'
'
I
'
'
'
'
I
I
111 I 1 11 I 1 1 1 I I I I I 11 ' '
'J
OH
CD [mdegj
-3.000E+01 3.000E+01
CD [mdeg]
I I I I I I I I I I I I
300.0
Fig. 9 Circular dichroism spectra of the isomeric lactones 43 and 72 as well as 42 and 76 (in trifluoroethanol)
519
HC^
^NNHS02^^^^_V-CH: 85
86
SchemelO Synthesis of A5-7-oxygenated and A5,7-unsaturated brassinosteroid analogues 82, 83, 84 and 86
520 For the synthesis of the corresponding A^.^-unsaturated brassinosteroid analogue 86 the A^-7-keto derivative 82 was reacted with toluene-4-sulfonohydrazine in dry tetrahydrofuran under anaerobic conditions at 75° C to give the corresponding tosylhydrazone, 85. Reductive elimination of compound 85 with lithium hydride in toluene at 100° C^^ yielded (22/?,23;?,245)-22,23dihydroxy-28-homoergosterol (86), the structure of which was confirmed by spectral data7^
4.
METABOLISM OF BRASSINOSTEROIDS
4.1.
Current status of brassinosteroid biosynthesis 79
In 1991, Yokota et al., published a hypothetical pathway of brassinosteroid biosynthesis.
The authors suggested phytosterols, e. g. campesterol, as biogenetic precursors of brassinosteroids. Teasterone (8) and typhasterol (7) may be regarded as the first compounds of the biosynthetic sequence bearing some of the major characteristics of brassinosteroids: trans-fustd A/B-ring system, vicinal hydroxyl groups in the side chain at C-22 and C-23, and oxygenation at C-6. Following this hypothesis, which is confirmed also by the occurrence of 6deoxobrassinosteroids, e.g. compounds 10, 33, and 34 in Ornithopus sativus (see. Section 2), bis-hydroxylation of the side chain should occur prior to 2a-hydroxylation. The final step of this hypothetical pathway is the Baeyer-Villiger type oxidation of castasterone (3) to yield brassinolide (1). This sequence was based on common occurrence and mechanistic considerations rather than on experimental results. The bioactivity, measured by means of the rice lamina inclination bioassay, increased with each biosynthetic step, and was used as a supporting argument for the suggested pathway. In the mean time several of the proposed biosynthetic steps have been supported by experimental results. Feeding experiments using [26,28-^H]labeled precursors and GC-MS analysis have established the biosynthetic sequence teasterone (8) —> typhasterol (7) —> 80 castasterone (3) —> brassinolide (1) in Catharanthus roseus. Teasterone (8) was demonstrated to serve as a biosynthetic precursor of typhasterol (7) in crown-gall and non-transformed cells of Catharanthus roseus in which both compounds are endogenous. This conversion probably 81
proceeds via 3-dehydroteasterone (31) as an intermediate naturally occurring brassinosteroid in Triticum aestivum
which has recently been identified as a and Distylium racemosum.
This result
is in analogy with ecdysteroids where epimerization has been demonstrated to occur through a 3dehydro type compound.
After feeding of [26,28-2H]3-dehydroteasterone (31), labeled
typhasterol (7) was detected as a major product and labeled teasterone (8) as a minor one. This reversibility of inversion suggested an 3-epimerase system, similar as shown for ecdysteroids.^^ When [26,28-2H]typhasterol (7) was administered to cultured cells of Catharanthus roseus, GCMS revealed castasterone (3) in extracts obtained after 24 h and 48 h, respectively. Because of the co-occurrence of teasterone (8), typhasterol (7), castasterone (3), and brassinolide (1) in several plant species, the proposed biosynthetic sequence might be ubiquitous in higher plants. The transformation of castasterone (3) to brassinolide (1) via a Baeyer-Villiger type reaction previously
521 has been demonstrated for crown gall cells of Catharanthus roseus using ^H labeled castasterone 84
2
(3). This was later confirmed by H-labeling experiments in cell suspension cultures of the same 85 species. The complete pathway (Scheme 11) between teasterone (8) and brassinolide (1) was 80
established by the same authors after feeding of deuterated precursors.
Phytosterol
Teasterone (8)
Typhasterol (7)
Castasterone (3)
3,6-Dehydroteasterone (31)
Brassinolide (1)
Scheme 11 Biosynthesis of brassinosteroids in Catharanthus roseus However, there are a lot of open questions in the biosynthesis of brassinosteroids. An interesting aspect which remains to be studied is the mechanism of the Baeyer-Villiger reaction in a biological system. Furthermore, the early steps of the brassinosteroid biosynthesis between phytosterols and teasterone (8) are an open field because there exist no labeling studies. It would be quite interesting to clarify if there are several parallel pathways between the phytosterols and brassinosteroids or if there is only one route, for instance the pathway outlined in Scheme 11. The co-occurrence of phytosterols and brassinosteroids with corresponding side chains implicate the former variant. However, it was also argued that major sterols do not account for the transformation to brassinosteroids in a proportional ratio, indicating rather selective transformation of 24-methyl and 24-methylene sterols. Investigations on the enzymatic level are not yet known. 4.2.
Brassinosteroid conjugates In contrast to the biosynthesis, aspects of interconversion and metabolism of brassino-
steroids have been poorly investigated until now. As assumed for classical phytohormones, also in the case of brassinosteroids, different types of conjugates may be involved in the biosynthesis,
522 transport, compartmentation and storage processes. However, only a few brassinosteroid conjugates are hitherto described. Whereas both the glycosides 23-0-p-D-glucopyranosyl-25methyldolicosterone (36) and 23-0-p-D-glucopyranosyl-2-epi-25-methyldolicosterone (37) occur native in Phaseolus viilgaris^^ 23-0-P-D-glucopyranosyl-brassinolide (87) was identified as a 79 87
metabolite of brassinolide in Vigna radiata. '
OH OH
23-0-p-D-Glucopyranosyl-brassJnollde(87)
Remarkably, all these glycosides bear the sugar moiety at the 23-OH. Until now no 3-0glucoside of any brassinosteroid has been found, although this is a commonly glucosylated position in other steroidal compounds.
88
Based on reports on fractions with significant activity in
89
87 90
the rice lamina inclination test and corresponding fractions in metabolic studies * a wide spread occurrence of hitherto unknown glucosides and hydrophilic non-glucosidic brassinosteroid conjugates is likely. Before starting our investigations, 23-0-p-D-glucopyranosyl-brassinolide (87) was the only known metabolite of plant origin derived from an exogenously applied 91
brassinosteroid. Acyl type conjugates were very frequently described for various phytosterols.
The first fatty acid conjugate, teasterone-3-myristate (9), was isolated recently from the anthers of Lilium longiflorwn. Acyl glucosyl conjugates, another common type of phytosterol conjugates in plants,
88
are expected to exist for brassinosteroids too. The lack of knowledge on the interconversion and the fate of brassinosteroids in plant
systems prompted us to study the metabolism of selected compounds of this type in detail. At the beginning of the metabolic studies on brassinosteroids we had to choose the compounds to be investigated as well as suitable plant systems. 4.3.
Prerequisites for metabolic studies 24-Epicastasterone (13) and 24-epibrassinolide (12) are naturally occurring brassino-
steroids with significant bioactivity in the rice lamina inclination test as well as in other bioassays. Both compounds are readily available
and a procedure for radioactive labeling with tritium was
established^^ (see, Section 3). [5,7,7-3H]24-Epicastasterone (13 in Scheme 5: R = 3 H ) and [5,7,7^H]24-epibrassinolide (12 in Scheme 5: R=^H) were used in our experiments. Following the biosynthetic sequence shown in Scheme 11 and experimental results, '
the 6-keto- and oxa-
lactone series are closely related biosynthetically. However, in Oryza saliva, castasterone (3), as an example for the 6-kelo compounds, may not serve as a precursor for brassinolide (1)^^ but may either be physiologically active per 5e or act as a precursor for unknown active brassinosteroids.
523 Cell suspension cultures possess several advantages over whole plants as objects in metabolic studies.
92
Thus, hitherto known experiments on the biosynthesis of brassinosteroids
were mainly performed using cell cultures of various plant species. In our experiments cell suspension cultures of Lycopersicon esculentum and Ornithopus sativus were used. Both cell lines are fast growing, well characterized and easy to handle. Preceeding experiments indicated that the cell growth of both species were not significantly influenced by concentrations up to 10'^ M of exogenously applied compounds 12 and 13. 4.4.
Metabolism of 24-epibrassinolide in cell cultures oi Lycopersicon esculentum The use of radiolabeled precursors allowed the measurement of the distribution of both the
parent compound and the metabolites in cell suspension cultures. In tomato cell cultures [5,7,7^H]24-epibrassinolide (12) was rapidly taken up by the suspended cells. As shown by TLC of ceU extracts obtained at day 4 after administration of the labeled compound, 24-epibrassinolide (12) has been converted to several hydrophilic metabolites. The structures of the major metabolites, clearly separated by TLC (see, insertion c and d in Fig. 10) were determined by MS and NMR analysis as 25-P-D-glucopyranosyloxy-24-epibrassinolide (88) and 26-P-D-glucopyranosyloxy93 94
24-epibrassinolide (89). '
_L
The FAB-MS spectra of both compounds (m/z 659 [M-t-H]"*", m/z
681 [M-i-Na]"^) exhibited nearly identical fragmentation patterns and relative intensities with negligible differences between the corresponding peaks (Fig. 10 a, b). The fragmentation patterns (e. g. m/z 409) indicated an additional fifth hydroxyl group located in the terminal part of the side chain beyond C-23. The positions of the new hydroxyl groups and of the glucosyl moieties at these newly functionalized carbon atoms in both compounds were unambiguously established by detailed NMR investigations (see, Section 5). Thus, the metabolic conversion of 24epibrassinolide (12) to 25-p-D-glucopyranosyloxy-24-epibrassinolide (88) and 26-p-Dglucopyranosyloxy-24-epibrassinolide (89), respectively, is a two-step process including hydroxylation to 25-hydroxy-24-epibrassinolide (90) and 26-hydroxy-24-epibrassinolide (91), respectively, and subsequent glucosylation (Scheme 12). These compounds represent the first brassinosteroids of plant origin with a hydroxyl group at the C-25 or C-26 position. Furthermore, the glucosides 88 and 89 are the first brassinosteroid glucosides are not to have the glucose moiety at C-23. The compounds 90 and 91 were not found in a non-glucosidated state in the cultured cells but were obtained by acid or enzymatic hydrolysis of the glucosides 88 and 89, respectively. The results of the rice lamina inclination test (RLIT) indicated an extraordinary high activity of 25-hydroxy-24-epibrassinolide (90). This compound is about ten times more active than 24epibrassinolide (12), indicating that the hydroxylation at C-25 is an activating step in the brassinosteroid metabolism. Therefore, 25-hydroxy-24-epibrassinolide (90) is, next to brassinolide (1), one of the most active brassinosteroids known until now. In comparison with 25hydroxy-24-epibrassinolide (90), the 26-hydroxylated metabolite (91) was clearly less active. As in other groups of steroidal hoimones, for instance vitamine D metabolites, hydroxylation at C-25 seems to be essential for high activity.
524
100
409
% 90
•Qc
60 50
479
497
lit.lll^l[^ll , t l l , t . t l , . , . . . ^ ! kf.hlfk ^Af
500
100
.11,11 I l ^ l i l j t l l
700
m/2
l4lH|>lltl^l.l^.tl^L ^i.ijil.^1 lvL,i^.t.jL*^|L,Jltiii^4l', "J 500 700 600
m/z
300
600
409
%
90
60
479
40
497
20
461
349 379
10
300
400
yMiw
Fig. 10 FAB mass spectra of (a) 25-P-D-glucopyranosyloxy-24-epibrassinolide (88) and (b) 26-P-D-gIuco-pyranosyIoxy-24-epibrassinolide (89). Insertions (c) and (d): Radio-TLC profiles of extracts of Lycopersicon esculentum cultured cells
525 12
/
25-Hydroxy-24-epibrassinollde (90)
OH
\
26-Hydroxy-24-eplbrasslnollde (91)
OH
rr
25-p-D-Glucopyranosyloxy-24-eplbrasslnolJde (88)
=
26-p-D-Glucopyranosyloxy-24-epJbrassinolide (89)
Scheme 12 Metabolism of 24-epibrassinolide in cell suspension cultures of Lycopersicon esculentum In comparison with the aglycones, the glucosides 88 and 89 exhibited less but also significant activity in the RLIT which may be due to hydrolysis within the test system. These findings suggest that 25-hydroxy-24-epibrassinolide (90) and its 25-0-glucoside (88) are not detoxification products of exogenously applied 24-epibrassinolide (12) but could be regarded as final members of the biosynthetic chain of brassinosteroids. The hydroxylation at C-25 and C-26 of 24-epibrassinolide (12) found in tomato cell cultures provided the opportunity to study these reactions in more detail. Hydroxylations in general are expected to be catalyzed by cytochrome P-450 dependent monooxygenases which are commonly characterized by their sensitivity to carbon monoxide and specific inhibitors. Thus, treatment of tomato cell cultures with various monooxygenase inhibitors simultanously with administration of 24-epibrassinolide (12) was supposed to influence the pathway of hydroxylation. From this approach infomiations on the specificity and the character of the enzymes involved were expected. The ratio of 89 : 88 within extracts of inhibitor non-treated cells was about 1 : 1. Tetcyclasis changed this ratio in favour of
26-p-D-glucopyranosyloxy-24-
epibrassinolide (89). This finding is quite opposit to the effect of cytochrome c which inhibited the formation of 89. Consequently, the concentration of 25-p-D-glucopyranosyloxy-24-
526
OH
=
1.0
Rf
Fig. 11 Radio TLC profiles of brassinosteroid glycosylation in cell cultures of Lycopersicum esculentum: (a) regiospecific glucosylation of 25-hydroxy-24-epibrassinolide (90), (b) non-specific glycosylation of 26-hydroxy-24-epibrassinolide (91)
527
epibrassinolide (88) was significantly increased. The different yields of 88 and 89, respectively, after ti*eatment with these inhibitors suggested that the hydroxylation of 24-epibrassinolide (12) at C-25 and C-26, respectively, is catalyzed by two regiospecific enzymes of different types.
This
was confirmed by CO poisoning, which is a principal criterium for a cytochrome P-450 96 involvement in an oxidation reaction. Following exposure to carbon monoxide (CO : O2 9:1), hydroxylation at C-26 was drastically reduced. The ratio of 89 : 88 was found to be 8 : 92 in this experiment. The CO poisoning effect was partially reversible by light. In contrast to the C-26 hydroxylation sensitive to carbon monoxide, the C-25 hydroxylation, non-typically for a cytochrome P-450 enzyme, was completely resistant. To examine the regiospecificity of the glucosyltransferases involved in the metabolism of 24-epibrassinolide (12) in cell cultures of Lycopersicon esculentum, tritium labeled compounds 90 and 91 obtained by enzymatic hydrolysis of biosynthetically prepared 88 and 89, respectively, were applied. After 4 days of incubation, radio TLC and reversed phase HPLC indicated that only 25-p-D-glucopyranosyloxy-24-epibrassinolide (88) was formed from 90 (Fig. 11 a). FAB-MS of 25-P-D-glucopyranosyloxy-24-epibrassinolide (88), obtained from re-application experiments confirmed the position of the glucose moiety at the terminal part of the side chain. These results strongly suggested that 25-hydroxy-24-epibrassinolide (90) did not undergo conversion except to 25-p-D-glucopyranosyloxy-24-epibrassinolide (88) nor was it subjected to any remarkable catabolism. In contrast, H labeled 26-hydroxy-24-epibrassinolide (91) under the same conditions afforded at least four glucosyl conjugates (Fig. 11 b). Among them the major compound was identical with 26-P-D-glucopyranosyloxy-24-epibrassinolide (89). Thus, it was concluded that 91 may be glucosylated at various hydroxyl groups. Comparing the behaviour of 90 and 91 in reapplication experiments it has to be stated that the glucosylation of the 25-hydroxy compound (90) proceeds in a highly regiospecific manner at this position while the glucosylation of the 26hydroxy compound (91) was less regiospecific. This implied that the aglycone (90) and its 25-0glucoside (88) could be involved in the mode of action of 24-epibrassinolide (12) in plants. 4.5.
Metabolism of 24-epicastasterone in cell cultures of Lycopersicon esculentum 24-Epicastasterone (13) is assumed to be the immediate biogenetic precursor of 24-
epibrassinolide (12), analogously with the related couple castasterone (3) and brassinolide (1) in the (245")-series.
In further experiments H labeled 24-epicastasterone (13) was fed to cell
suspension cultures of Lycopersicon esculentum and the extract obtained four days after administration was shared into two halfs. One part was subjected to enzymatic hydrolysis (aglycone fraction) while the other one was further used in the glucosidic state. Four brassinosteroid glucosides were identified as major metabolites within the nonhydrolyzed fraction. As described for the formation of 88 and 89 after administration of 12, hydroxylation and glucosylation also occurred at C-25 and C-26 of the side chain of 13 yielding 25-P-D-glucopyranosyloxy-24-epicastasterone (92) and 26-P-D-glucopyranosyloxy-24-epicastasterone (93) (Scheme 13). The structures of both glucosides were established by FAB-MS (92: m/z 665 [M]"^) and (93: m/z 665 [M]"*") and NMR. The fragmentation patterns of both compounds
528 OH
OH
=
=
OH
26-p-D-Glucopyranosyloxy24-epicastasterone (93)
25-p-D-Glucopyranosyloxy24-epicastasterone (92)
t
OH = OH
\ - \
OH
^Ky HO""^- ' ' ' ' ^ N /
13 I—
I-
25-Hydroxy-24-epicastasterone
26-Hydroxy-24-epicastasterone
-J
(95)
(94) OH
=r
3-Dehyclro-24-eplcastasterone (96)
i
14
25-Hydroxy-3,24-diepicastasterone (97) OH
=
HO—V
OH 2-O-p-D-Glucopyranosyl3,24-diepicastasterone (98)
HO-
HOHO.
.S^:^' 3-0-p-D-Glucopyranosyl3,24-diepicastasterone (99)
Scheme 13 Metabolism of 24-epicastasterone in cell suspension cultures of Lycopersicon esculentum
529 (m/z 393, bond fission between C-23 and C-24) confirmed the position of the glucosyloxy moieties at the terminal part of the side chain beyond C-23. Obviously, the metabolic pathway which was established for the 24-epibrassinolide (12) is also operating for 24-epicastasterone (13), namely the regiospecific hydroxylation of 13 and glucosylation of the newly formed hydroxyl groups in the intermediates, 25-hydroxy-24-epicastasterone (94) and 26-hydroxy-24epicastasterone (95). Furthermore, 2-0-p-D-glucopyranosyl-3,24-diepicastasterone (98) and 3-03-D-glucopyranosyl-3,24-diepicastasterone (99) were detected and the structures were elucidated by NMR analysis of the non-separated mixture of both compounds (see. Section 5). These compounds represent the first brassinosteroid glucosides bearing the sugar moiety at the ring A hydroxyl group. The 3p-conriguration in 99 suggested epimerization at C-3 prior to glucosylation. This was confirmed by the isolation of related aglycones from the hydrolyzed parts of the extract. These metabolites belong to a metabolic sequence starting with the oxidation of the 3a-0H. The first compound of this sequence is 3-dehydro-24-epicastasterone (96), a new 3,6-diketobrassinosteroid. It has to be regarded as an intermediate in the epimerization to 3,24-diepicastasterone (14), 2 which is known as a naturally occurring compound. Compound 14 is a branching point in this metabolic sequence. It is either glucosylated at 3P-0H or at 2a-0H yielding 99 and 98, respectively, or it can be hydroxylated at C-25 to give 25-hydroxy-3,24-diepicastasterone (97). For MS fragmentation, see Scheme 16. Alternatively, compound 97 may be derived also from the intermediate pentahydroxylated 25-hydroxy-24-epicastasterone (94). The latter compound 94 and 26-hydroxy-24-epicastasterone (95) were not detectable in a non-glucosylated state, probably due to very small endogenous pool sizes. Analogously, the intermediate pentahydroxylated metabolites 90 and 91 were also not detectable in the cell culture medium. Scheme 13 shows the metabolic pathways of 24-epicastasterone (13) in cell suspension cultures of Lycopersicon esculentum so far currently known.
97
All these compounds, both glucosides and aglycones, were exclusively
isolated from the suspended cells of Lycopersicon esculentiim. The medium of this cell culture did not contain significant amounts of any brassinosteroid metabolite, 4.6.
Metabolism of 24-epicastasterone and 24-epibrassinolide in cell cultures of Ornithopus sativus A quite different distribution of metabolites was found in cell suspension cultures of
Ornithopus sativus. Surprisingly, from 1 h after the beginning of the experiments over the remaining incubation time of several days, the distribution of the radioactivity between the cells and the medium did not change significantly. About 40% of the radioactivity after feeding of 24epibrassinolide (12), and 25% after feeding of 24-epicastasterone (13), were found in the medium and the remainder was present within the cells. There was also a clear compartmentation of the different types of metabolites between the cells and the culture medium. While in the medium the non-conjugated metabolites were almost solely found, the cells did contain both hydrophilic and lipophilic conjugates. From the medium brassinosteroid-derived pregnan-like compounds were 98 isolated and their structures were elucidated by MS and NMR analysis. Following apphcation of
530 24-epibrassinolide (12), 2a,3P-dihydroxy-B-homo-6a-oxa-5a-pregnane-6,20-dione (102, m/z 364 [M]"^), and after exogenous application of 24-epicastasterone (13), 2a,3P-dihydroxy-5apregnane-6,20-dione (108, m/z 348 [M]"*") and 2a,3P.6P-trihydroxy-5a-pregnane-20-one (109, m/z 350 [M]"^) were found. '
These compounds are the first side chain degradation products of
brassinosteroid origin described in plant material. Compounds 102 and 109 seem to be the final products of catabolism of 12 and 13, respectively, in Ornithopus sativus cell cultures and two completely elucidated metabolic sequences revealing these compounds were established (Scheme 14 and 15).^^^ 12 OH
3,24-Dlepibrasslnolide (100)
=
25-Hydroxy-3,24-diepibrassinollde (103)
OH
(20R)-Hydroxy-3,24-diepibrassinoflde(101)
3,24-Diepibrassinollde-
3p-laurate (104)
o
3p-myrlstate (105) 2a,3p-Dlhydroxy-B-homo6a-oxa-5a-pregnane-6,20-dione (102)
3p-palmitate (106)
Scheme 14 Metabolism of 24-epibrassinolide in cell suspension cultures of Ornithopus sativus Starting from 24-epibrassinolide (12) and 24-epicastasterone (13), in both cases the metabolism involves epimerization of the 3a-hydroxyl group to the equatorial 3p-0H, leading to compounds 3,24-diepibrassinoIide (100) and 3,24-diepicastasterone (14), respectively. The mass
531 spectra resemble those of the parent compounds 12 and 13, respectively. However, the H,^H coupling constants in the H NMR spectra indicated axial position of H-3 (see, Section 5). In the next step, hydroxylation at C-20 takes place. The mass spectra of the resulting (20/?)-hydroxy3,24-diepibrassinolide (101, m/z 497 [M]"^) and (20/?)-hydroxy-3,24-diepicastasterone (107, m/z 481 [M] ), respectively, are characterized by molecular ions of low intensities (about 1%) and very strong fragment ions of m/z 365 [M]"^ (79) for 101, and m/z 349 [M]"^ (100) for 107, indicative for bond fission between C-20 and C-22 (fragment b in Scheme 16, R2 = OH, R3 = R4 = H). Obviously, this bond between C-20 and C-22 is destabilized by C-20 hydroxylation and
13
14 3,24-Dlepicastasterone-
3(3-)aurate (110)
0
(20R)-Hydroxy-3,24-dlepicastasterone (107) ap-palmitate (112)
2a,3n-Dihydroxy5a-pregnane-6,20-dJone (108)
2a,3p,6{^-Trihydroxy5a-pregnane-20-one (109)
Scheme 15 Metabolism of 24-epicastasterone in cell suspension cultures of Omithopus sativus it is hence accessible to enzymatic attack. This assumption was confirmed by the very small concentrations of metabolites 101 and 107, respectively, suggesting rapid side chain cleavage between C-22 and C-20. The 20,22,23-trihydroxy structural feature, even more then the general structure of brassinosteroids, resembles the ecdysteroids which frequently bear a 20-hydroxyl
532 group.101,102 Within this metabolic sequence of 24-epibrassinolide (12) (Scheme 14), the pregnane-like compound 102 is the final product. The seven-membered lactone ring structure obviously prevents further conversions, which in the metabolic sequence of 24-epicastasterone (13) (Scheme 15) via reduction of the 6-keto group led to 2a,3p,6P-trir ydroxy-5a-pregnane-20one (109),^^ The structure elucidation of another minor metabolite of 24-epibrassinolide (12) from the cell culture medium of Omithopus
sativus revealed the presence of 25-hydroxy-3,24-
diepibrassinolide (103, m/z 497 [M+H]*). In contrast to 25- and 26-hydroxy-24-epibrassinolides (90 and 91) which were found in the glucosidic state in Lycopersicon esculentuniy this compound occurred as an aglycone. In the mass spectra, the fragments m/z 379 of 103 (fragment a in Scheme 16, R p -COO, R2 = R4 = H, R3 = OH) and m/z 365 of 101 (fragment b in Scheme 16, Rj = COO, R2 = OH, R3 = R4 = H),
C4•,'i'i...!
OH HQ,,,
R^-j-O^
J3
^ R 1 ^
R
! . - • ( ]1
R2
1R1
^co ^co ^co ^co ^co -co-co-co-co-co-
R3
R4
H
OH
H
25-Hydroxy-3,24-diepicastasterone
(97)
OH
H
H
(20R)-Hydroxy-3,24-clieplcastasterone
(107)
H
H
lauryl
3,24-Dieplcastasterone-3p-laurate
(110)
H
H
myristyl
3,24-Dieplcastasterone-3n-myristate
(111)
H
H
palmltyl
3,24-Diepicastasterone-3n -palmltate
(112)
-0- •
OH
H
H
(20R)-Hydroxy-3,24-dlepibrassinollde
(101)
-0-
H
OH
H
25-Hydroxy-3,24-diepibrassinolJde
(103)
H
H
lauryl
3,24-Dleplbrasslnollde-3p -laurate
(104)
-0-0-0-
•
H
H
myristyl
3,24-DleplbrassmolJde-3p -myristate
(105)
H
H
palmltyl
3,24-Dleplbrassmolide-3(3 -palmltate
(106)
Scheme 16 EI-MS fragmentation of brassinosteroid metabolites of the 3,24-diepi series respectively, represent diagnostic ions of pentahydroxylated brassinosteroids derived from 12 (for 25-hydroxylation in the first case and for 20-hydroxylation in the second one). As found for 3,24-diepicastasterone (14) in L esculentum
cell cultures, 3,24-
diepibrassinolide (100) is a branching point in the metabolism of 24-epibrassinolide (12) in
533 Ornithopus sativus. Besic'e 25-hydroxylation as a minor metabolic reaction in 0. sativus, the fatty acid esters were mainly formed from 3,24-diepibrassinolide (100). However, these lipohilic 99
metabolites were not present in the medium but only inside the cells.
After purification of the
lipophilic fraction of the cell extract by TLC and separation by HPLC (RP-8), the structures of these metabolites were elucidated by spectroscopic methods. Three fatty acyl esters were derived from both 100 and 14, all in nearly the same quantity. 3,24-Diepibrassinolide-3P-laurate (104, m/z 662), -3p-myristate (105, m/z 690), -3P-palmitate (106, m/z 718) were found as metabolites of 24-epibrassinolide (12), and 3,24-diepicastasterone-3P-laurate (110, m/z 646), -3P-myristate (111, m/z 675), -3P-palmitate (112, m/z 702) were metabolites of 24-epicastasterone (13). The position of the fatty acid residue at ring A can be deduced from the fragment ions which appear after fission a, b, or c (Scheme 16). The base peaks of the acyl-conjugated 3,24diepibrassinolides (104 - 106) appear at m/z 361 (a-RCOOH) and of the acyl-conjugated 3,24diepicastasterones (110 - 111) at m/z 346 (a+H-RCOOH), respectively. The NMR spectra of the fatty acyl conjugates are very similar to each other. In comparison to the spectrum of the nonconjugated compounds 100 and 14, respectively, H-3p exhibits a downfield shift of about 1.2 ppm due to an ester bond at this position. The H, H coupling constants establish that epimerization has occurred at C-3. In addition to the signals of the genines, the H NMR spectra exhibit signals of the fatty acid methylene protons (5 1.25) and the terminal methyl groups (5 0.88). Our results represent the first report of fatty acid conjugates as metabolites of exogenously applied brassinosteroids. The function of these fatty acyl ester derivatives of brassinosteroids still remains unknown. However, they may be compartmentalised within membrane structures as generally described for phytosterol acyl esters. 103 Several minor hydrophilic compounds were detected in cultured cells of O. sativus which may be glucosides, but the major part of radioactivity was associated with the acyl ester fraction. Summarizing die results of our studies on die metabolism, it can be stated that there are two principle pathways of brassinosteroid conversion in plants: First, hydroxylation in the terminal part of the side chain followed by glucosylation of the newly formed hydroxyl group. This pathway, at least hydroxylation at C-25 of 24-epibrassinolide (12), significantly increases the bioactivity of the 25-hydroxylated compound 88 compared with the parent substance and therefore can be regarded as an activation reaction. Second, catabolic side chain removal was found. This pathway starts with epimerization at C-3, followed by hydroxylation at C-20 and bond fission between the vicinal hydroxyl groups at C-20 and C-22. Conjugation at C-2 and C-3, respectively, with glucose or at C-3 with fatty acids seems to require equatorial position of the corresponding hydroxyl group which is a result of the preceeding epimerization. Surprisingly, the expected biosynthetic Baeyer-Villiger oxidation of 24-epicastasterone (13) to yield 24-epibrassinolide (12) was not observed either in cell cultures of Lycopersicon esculentum or in Ornithopus sativus in our experiments. Altogether, 26 metabolites of exogenously applied brassinosteroids, (except compound 14) described for the first time, have been found in our studies until now.
534 5.
NMR SPECTROSCOPY OF BRASSINOSTEROE) METABOLITES NMR spectroscopy is a powerful tool for structural elucidation of brassinolid metabolites.
Based on complete and unambiguous assignments of proton and carbon NMR signals of the main brassinosteroids^^"*'^^^ the structures of metabolites can be determined on the basis of changes in chemical shifts and coupling constants as well as by correlations found in two-dimensional NMR experiments. Modem NMR spectrometers with cryomagnets and special designed probeheads have made it possible to record proton detected one- and two-dimensional NMR spectra of very small amounts of natural compounds. Nowadays, direct ^^C measurements can be done with amounts of brassinosteroid metabolites down to about 1 -2 fimol, whereas proton detected spectra can be recorded with amounts down to about 0.1 - 0.2 jimol of metabolites. So-called inverse-detected ^H-^^C chemical shift correlation spectra allow the assignments of carbon signals even in cases in which no direct ^^C NMR spectra can be recorded because of poor signal-to-noise ratios. Since in these inverse heteronuclear shift correlation experiments magnetization instead of ^-^C magnetization is detected, the sensitivity is significantly better than in conventional 2D experiments using ^^C detection. Important hints regarding structural changes during metabolic processes can often be achieved by inspection of the relevant parts of the proton NMR spectrum from the metabolite in comparison with the parent brassinosteroid. Metabolic hydroxylation of the methine carbons in the brassinosteroid side chain results in a change of the coupling patterns and in a low-field shift of the adjacent side chain methyl group proton signals (Fig. 12). For example, the ^H NMR spectrum of 24-epibrassinolide (12) shows four side chain methyl group doublets at 5 0.97 (Me-21), 0.91 and 0.86 (Me-26 and Me-27) and 0.83 (Me-28), whereas the proton signals of Me-26 and Me-27 in 25-P-D-glucopyranosyloxy-24-epibrassinoHde (88) appear as singlets at 5 1.36 and 1.30. Hydroxylation of a side chain methyl group gives rise to disappearance of the corresponding methyl doublet in the high-field region and to the appearance of two new proton multiplets in the low-field region. The absence of several methyl group ^H NMR signals indicates degradation of the side chain. Thus, the ^H NMR spectrum of 2a,3P-dihydroxy-5a-pregnane-6,20-dione (108) exhibits only two methyl singlets in the high-field region (5 0.81 and 0.62; Me-19 and Me18, respectively). The methyl singlet at 5 2.13 of 108 is assigned to Me-21 in a 20-ketopregnane side chain moiety. An unchanged side chain is proved by the occurrence of four methyl doublets with the same chemical shifts as found for the feeded brassinosteroid. Even if the methyl region of the proton NMR spectra is superimposed by signals of impurities or by incompletely separated minor metabolites, which is not unusual if the amounts of metabolites are very small, it is possible to recognize an unchanged side chain by its fingerprint region in the ^H-^H chemical shift correlated 2D NMR (COSY) spectrum (Fig. 13). The H-^H COSY spectra correlate proton chemical shifts through homonuclear coupHngs. Starting from a separated, readily assigned signal the appearance of cross peaks allow identification and assignment of the complete spin system. For instance, H-5a shows correlation peaks with H-4a and H-4p, which correlate with H-3p (or H-3a), the latter showing further cor-
535
ppm
Fig. 12 ^H NMR high-field region of 24-epicastasterone (13)^ and 25-P-D-glucopyranosyloxy-24epicastasterone (92)^ (500 MHz, solvent: 0.16 ml CD3 OD, ^ 2.0 mg, ^ 1.9 mg) * May be reversed relation with H-2p. Finally, H-2P correlates with H-la and H-lp. Providing that H-5a is assigned by its chemical shift and coupling pattern (doublet of doublet), simply mapping the ^H-^H couplings by a COSY spectrum is sufficient to assign all signals of ring A protons. Unfortunately, assignment of all the side chain protons starting from a methyl signal suffer from the frequent overlapping of H-20 and H-24. Chemical shifts and coupling constants (or linewidth) of signals in the low-field region of the proton NMR spectrum are significant for changes at rings A and B or in the side chain of the brassinosteroid. Thus, epimerisation at C-3 (3a-0H —> 3p-0H) results in a high-field shift of H-5a due to the absence of the deshielding 1,3 diaxial interaction with 3a-0H. On the other hand a dramatic enlargement of the linewidth of the H-3 multiplet (Fig. 14) is observed, since H-3p is an equatorial oriented proton, whereas the axial H-3a shows a different coupling pattern because of the large axial-axial vicinal coupling constants. Esterification of 3-OH results in a downfield shift of H-3 of about 1 ppm. Finally, glycosylation leads to new spin systems in the low-field proton region, which can be recognized in a ^H-^H COSY 2D NMR spectrum.
536 OH
JQA^I/ 3" J &
J
(ppmll 1.0
I^ ' i.1.2i
D
*-H
i.6i
i.8i
e
2.0 2.2 2.4i 2.6 2.8' 3.0-
i
3.2i
^
3.41 3.6
r[iiiiinii|iiii|iiiiniii|riii|im|iiir]iiii|ini|mi)irn
3.4
3.0
2.6
2.2 Fl
ri|mi[iiii|iin|iiii|iiii|iiii|iiii|iin|iiii[iiii|iiii[ifiniiii
i.B
1.4
1.0
0.6
(ppm)
Fig. 13 %-^H 2D COSY spectrum of 3,24-diepicastasterone (14) (500 MHz, solvent: 0.16 ml CDCI3, 0.05 mg) Marked: fingerprint region of an unchanged 24-epicastasterone (or 24-epibrassinolide) side chain
537
CO.
d
|Hi>|iin|ini[iiii)iiii|iiii|inijiiin"ii|iiii|iiii[iiiniiiijini|iiiijiiir|iiiijii 4.0
3.B
3.6
3.4
3.2
3.0
2.8
2.6
2.4
ppm
Fig. 14 Low-field region of ^H NMR spectra of 24-epicastasterone (13)^ and (20R)-hydroxy3,24-diepicastasterone (107)^ (500 MHz, solvent: 0.16 ml CDCI3, ^ 2.0 mg, ^ 0.2 mg) After assignment of proton signals in the low-field and the methyl region, this assignments can be transferred to carbon signals by a ^H-^^C shift correlation via one bond (Fig. 15), which should be carried out as the proton detected experiment (HMQC: heteronuclear multiple quantum correlation)^^ for sensitivity reasons. Such two-dimensional spectra show correlation peaks at the ^-^C chemical shift in one dimension and at the ^H chemical shifts (of those protons which are bound directly to the carbon) in the other dimension. Mostly, complete sU'uctural elucidation requires information from ^H-^^C shift correlations via two or three bonds (so-called "long-range" correlations). Again, for sensitivity reasons the proton detected version (HMBC: heteronuclear multiple bond correlation)^^^ is the experiment of choice. Each angular methyl group gives four correlations via ^JQ J^ and ^J^^ ^ with carbon signals, whereas each side chain methyl group gives three correlations (Table 3). While methyl groups always cause strong correlation peaks in the HMBC spectrum, correlations between methine or methylene protons and carbons via two and three bonds may be weak or even undetectable, depending on the ^H-^^C coupling constant over two or three bonds and on the proton signal multiplet splitting. Because Me-18 (proton singlet) and Me-21 (proton doublet) have a mutual HMBC correlation to C-17, these signals are easy to assign. In the same way Me-26 and Me-27 can be as-
538
-
"1
(ppm)I
f
20-
_
0.0
II
30-]
-
11
i»«
•
40-
_ -
•II
• 10
60-
70-
i*
III
c-
50-
•
1
* ' "•*
0
-
# • ' ' 1 ~T 1 1 r i • ' ' '
4.0
3.5
1' ' ' ' 1' • 3.0 2.5 F2
T—r"i—1—r—J J J p—1 J 1 J 1 1 p J
2.0
1.5
1.0
(ppm)
Fig. 15 ^H-^^C one-bond shift correlated 2D NMR spectrum (HMQC) of 3-dehydro-24epicastasterone (96) (500 MHz, solvent: 0.16 ml CDCI3, 0.4 mg)
539 Table 3 Expected correlations between ^H NMR methyl group signals and ^^C NMR signals in ^H-'^^C shift correlation 2D NMR spectra via two and three bonds for brassinosteroids
c
1 5 9 10 12 13 14 17 20 22 23 24 25 26 27
Me-18
Me-19
Me-21
Me-26
Me-27
X X
X X X
Me-28
X X X X X X X X
X X X X X X
X
signed due to their mutual correlation with C-25 and C-24. C-22 and C-23 can be assigned by the correlation with Me-21 and Me-28, respectively. Considering the correlations found in the HMQC spectrum, the proton NMR signals of H-22 and H-23 can be readily assigned, too. In a similar manner, the assignment of the methyl group proton signals and a lot of carbon signals is possible in the case of the metabolic side chain hydroxylation. For example (Fig. 16), two ^H methyl group singlets of (20/?)-hydroxy-3,24-diepicastasterone (107) (5 1.30, side chain methyl group because of three HMBC correlations; 5 0.88, angular methyl group because of four HMBC correlations) show a mutual HMBC correlation to a carbon signal at 5 56.7, which therefore has to be assigned to C-17. Consequently, the side chain methyl group (5 1.30) is assigned to Me-21 and the angular methyl group (5 0.88) to Me-18. Me-21 exhibits two further HMBC correlations to carbon signals at 5 78.0 and 74.7, respectively. The former carbon signal gives no correlation in the HMQC experiment and it has to be therefore a quartemary carbon. The carbon signal at 5 77.4 is known to belong to a methine carbon from the HMQC spectrum. Considering the ^^C chemical shifts, both carbons must be hydroxylated. Thus, hydroxylation at C-20 has taken place in the course of the metabolic conversion. The HMBC correlation between the anomeric proton of the sugar unit and the three bond distant steroid carbon in glycosylated metabolites is very important for elucidation of the glycosylation site. Thus, the ^H NMR signal of the anomeric proton of glucose in 25-|3-D-glucopyranosyloxy-24-epibrassinolide (88) exhibits, apart from coirelations with glucose carbons, one correlation to a genin carbon (Fig. 17) which can be assigned to C-25 by its HMBC correlations with the two ^H methyl group singlets at 5 1.36 and 1.30 (Me-26 and Me-27). Unfortunately, such HMBC correlations involving anomeric proton signals may be weak or superimposed by
540
25-d C-25
30-d 35H
AoA
C-12
- > OOo C-lOv
C-24
45-§ 50-j 55
0®0
-4-
C-17 - • dQO
60-:] 65 70H
C-23
- > oOOo
C-22
75H
C-20 I I j I I t I I I I I I I I I i I I I I I I I I I I I I I I I I I I I I I I I 1 I "I I 1 I I I I I 1 I I r I I r 1 1 I I I I T
1.30
1.20
1.10
1.00 F2
0.90
0.80
0.70
(ppm)
Fig. 16 Part of the ^H-^^C long-range shift correlated 2D NMR spectrum (HMBC) of (20R)-hydroxy-3,24-diepicastasterone(107) (500 MHz, solvent: 0.16 ml CD3OD, 0.2 mg); for the shown spectrum CD3OD was used as solvent instead of CDCI3 because the cross-peaks are more separated
541
_^
Me.28
F2 (ppm)
Me-26/ Me-27
1.5i 26
2 . OH
OH ^^^
'
CH3 '^CH3
2.5H
3.0"
3.5H
4.0H
glc-Hl*
4.511 r {I u 1 j I M 1111111n11111111111111 n 1}! 111 j n 111
86 85
84 83 82 8 1 80 79 78 Fl
77
(ppm)
Fig. 17 Part of the ^H-^^C long-range shift correlated 2D NMR spectrum (HMBC) of 25-P-D-glucopyranosyloxy-24-epibrassinolide (88) (500 MHz, solvent: 0.5 ml CD3OD, 1.8 mg) other correlations as in the case of 26-p-D-glucopyranosyloxy-24-epibrassinolide (89). The determination of the glycosylation position is also possible based on ^H acetylation shifts after complete acetylation of the metabolites. Protons at the acetoxylated carbons resonate about 1 ppm downfield in comparison with the non-derivatized metabohte, whereas those at the glycosylation sites are only slightly shifted. For example, the non-separated mixture of the two metabolites 2-0-p-D-glucopyranosyl-3,24-diepicastasterone (98) and 3-0-p-D-glucopyranosyl3,24-diepicastasterone (99) show after acetylation in the ^H-^H COSY spectrum two sets of H-2p/H-3a signals at 5 4.92/3.61 (major component) and 3.76/4.74 (minor component), respec-
542 lively. Consequently, 99 represents the major and 98 the minor glucosylated metabolite. Nuclear Overhauser effects (NOE) contain information about the spatial proximity of protons (or hetero nuclei). Detailed NOE investigations can allow estimation of configuration and/or conformation. NOE measurements can be carried out one-dimensional as NOE difference spectrum or as a two-dimensional NOESY or ROESY experiments. In case of brassinosteroid metabolites it can be advantageous to perform the NOE difference experiment because of its better overall sensitivity, especially if only some NOE interactions will yield the desired information. In such a manner NOE enhancements found for (20R)-hydroxy-3,24-diepicastasterone (107) (Fig. 18) by irradiation of methyl group and side chain proton signals suggest (20R) configuration. This is in agreement with the Me-21 ^H downfield shift of 0.36 ppm compared with 24epicastasterone (13). A similar downfield shift has been found also for corresponding 20-hydroxylated cholesterols^^^, pregnanes^^^ and withanolides^^^.
X I
' 1
mt »W»i.>»iiiMHi», iiitli^^Wm^mnyU*! n n » » » n W , » > w y A L,gW^M
X I
7
\
»J\^
i
X I 0'^-^imAml>*frm^mt^<„^,0> * « t i H M M « i » * « i i ^
OH ^21 OHf OH
U^
CH3 I '
'**»t,Jft*^',%,
27
CH3 26
l_^A^aJ^4>V^
xJ^^
^»ffyU
LJJIL
M|iiiiiiiiijMii|iiii|Miit.iii|iiii|inijiMi|nii|iiii{iin)iiM|[MijiMi|nii|iiii|iiii[iiii|inijiiiitnii|iiinMiijniu
3-6
3.2
2.8
2.4
2.0
1.6
1.2
ppm
Fig. 18 ^H NMR NOE difference spectra of (20R)-hydroxy-3,24-diepicastasterone (107) (500 MHz, solvent: 0.16 ml CDCI3,0.2 mg)
543 ^H and ^^C chemical shifts of brassinosteroid metabolites are shown in Tables 4 and 5, respectively.
Table 4 ^H chemical shifts of parent brassinosteroids 12 and 13 and metabolites 14, 88, 89, 92, 93,96, 97 and 99 -107; chemical shifts are obtained from the ^H NMR (500 MHz), 2D COSY or 2D HMQC spectra; * may be reversed; ^ solvent: CDCI3; ^ solvent: CD3OD; ^ solvent: 95 vol. % CDCI3 + 5 vol. % CD3OD; n. d.: not detected because of poor signal-to-noise ratio and/or overlapping with other signals
H
12a
13a
14a
88^
89^
92b
93^
96^
1 1.86/1.55 1.74/1.55 2.06/1.24 1.82/1.60 1.82/1.60 1.68/1.57 1.68/1.57 2.54/1.44 2 3.60 3.65 3.77 3.60 3.64 3.60 3.65 4.25 3.98 3.39 3.92 3 4.05 3.92 3.94 3.96 4 2.10/1.93 1.92/1.72 1.96/1.60 2.07/1.80 2.05/1.80 1.76/1.66 1.78/1.67 2.72/2.52 3.12 2.69 2.32 3.21 5 3.20 2.72 2.64 2.73 7 ca.4.10 2.30/2.00 n.d. 4.18/4.08 4.19/4.08 2.20/2.11 2.21/2.11 2.39/2.00 n. d. 1.72 1.76 1.69 1.69 8 1.80 1.79 1.85 1.34 1.40 n.d. 1.42 9 1.30 n.d. 1.37 1.43 1.81/1.44 1.68/1.34 1.68/1.41 1.68/1.42 n.d. 11 1.80/1.40 1.65/1.34 n.d. 12 1.99/1.22 2.02/1.28 n.d. 2.04/1.27 n.d. 2.09/1.32 2.07/1.31 2.06/1.30 n.d. 14 1.18 1.31 1.25 n.d. 1.37 1.37 1.33 1.73/1.24 1.59/1.14 1.59/1.15 1.60/1.15 n.d. 15 1.68/1.22 1.58/1.11 n.d. n.d. 1.99/1.35 2.00/1.37 2.02/1.34 16 1.99/1.25 1.98/1.30 n.d. 2.00/1.39 1.60 1.57 n.d. 1.59 1.56 n.d. 1.57 17 1.56 0.74 0.68 0.77 0.78 0.70 0.71 0.68 0.73 18 1.04 0.76 0.92 0.81 0.90 0.90 0.76 0.76 19 1.54 1.59 1.48 1.46 1.58 1.60 1.46 1.47 20 0.98 0.97 0.98 0.98 1.06 0.97 0.98 1.05 21 3.66 3.69 3.69 3.65 3.67 3.70 3.69 3.66 22 3.41 3.54 3.54 3.35 3.41 3.41 3.35 3.37 23 1.50 1.99 1.93 1.50 1.98 1.93 1.48 1.51 24 1.89 2.20 2.19 1.90 1.90 1.90 25 3.71/3.41 0.92 0.92 0.92 3.72/3.40 1.35 0.92 1.36 26* 0.87 0.87 1.29 0.87 0.87 1.30 27* 0.87 0.87 0.85 0.85 0.89 0.85 0.85 0.85 0.90 0.85 28 1' 2' 3' 4' 5' 6'
4.23 4.22 4.56 4.56 3.17 3.17 3.12 3.12 3.36 3.332 n.d. 3.33 3.24 n.d. 3.28 n.d. 3.24 3.24 n.d. 3.25 3.81/3.60 3.86/3.65 3.80/3.61 3.86/3.64
544 Table 4 Continued H 1 2 3 4 5 7 8 9 11 12 14 15 16 17 18 19 20 21 22 23 24 25 26* 27* 28 1' T y 4' 5' 6* -CH2-CH3
97b
99b
100^
101^
102^
103^
104^
105^
n.d. n.d. 2.22/1.20 2.24/1.21 n.d. 1.99/1.22 2.05/1.23 iTd 3.72 3.54 3.72 3.51 3.52 3.46 3.47 3.63 4.55 4.56 3.39 3.34 3.40 3.40 3.48 3.60 n.d. n.d. n.d. n.d. 2.08/1.93 2.09/1.96 1.83/1.50 2.01/1.52 2.97 2.91 2.97 2.91 2.90 2.46 2.91 2.45 4.12/4.03 4.14/4.01 4.12/4.03 4.12/4.01 4.13/4.02 4.12/4.02 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n,d. n.d. n.d. n.d. n.d. n.d. 1.42 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 1.34 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n. d. n.d. n.d. n.d. n.d. n.d. n.d. 1.54 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.72 0.87 0.70 0.71 0.66 0.71 0.71 0.73 0.79 0.78 0.97 0.96 0.97 0.97 0.97 0.97 1.49 n.d. 1.49 1.43 n.d. n.d. 1.02 0.98 0.96 1.33 2.13 0.97 1.00 0.97 3.66 3.66 3.65 3.43 3.68 3.63 3.68 3.48 3.38 3.33 3.75 3.41 3.41 3.46 1.72 1.47 n.d. 1.49 1.77 n.d. n. d. 1.95 n.d. 2.01 n.d. n.d. 0.91 1.22 0.92 0.95 1.30 0.92 0.92 1.20 0.86 0.87 0.82 1.28 0.87 0.87 0.83 0.84 0.83 0.76 0.82 0.85 0.85 4.36 n.d. n.d. n.d. n.d. 3.84/3.65 1.26 0.88
1.25 0.88
545 Table 4 Continued
H 1 2 3 4 5 6 7 8 9 11 12 14 15 16 17 18 19 20 21 22 23 24 25 26* 27* 28 -CH2-CH3
106^
107a
108^
109^
2.05/1.23 2.08/1.26 1.94/0.98 3.60 3.61 3.63 3.39 3.40 3.45 1.96/1.61 1.98/1.60 1.81/1.70 2.34 2.32 1.28 3.88 4.13/4.02 2.32/n. d. n.d. 1.82/1.19 n.d. n.d. n.d. n.d. 1.34 n.d. n.d. n.d. 1.66/n. d. n.d. n.d. n.d. n.d. n.d. n.d. 2.08/1.31 1.31 n.d. n.d. n.d. 1.84/n. d.* n.d. n.d. n.d. 1.59/n. d.* n.d. n.d. n.d. 1.90 n.d. n.d. n.d. 0.64 0.84 0.62 0.71 0.80 0.81 1.07 0.97 2.13 2.12 n.d. 1.34 0.97 3.35 3.68 3.77 3.41 1.60 n.d. 2.02 n.d. 0.95 0.92 0.83 0.87 0.77 0.85 n.d. 3.68 4.55 n.d. 2.97
1.25 0.88
110^
IIP
112^
n.d. 3.80 4.58 n.d. 2.38
n.d. 3.79 4.57 n.d. 2.38
n.d. 3.80 4.58 n.d. 2.38
n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.68 0.81 n.d. 0.98 3.69 3.41 n.d. n.d. 0.92 0.87 0.85
n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.67 0.81 n.d. 0.98 3.69 3.41 n.d. n.d. 0.92 0.87 0.85
n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.68 0.81 n.d. 0.98 3.70 3.41 n.d. n.d. 0.92 0.87 0.85
1.26 0.88
1.25 0.88
1.25 0.88
546 Table 5 ^^C chemical shifts of parent brassinosteroids 12 and 13 and metabolites 88, 89, 92, 93, 96, 99 and 107 ^^C chemical shifts of 12, 13 and 88 are obtained directly from the ^^C{^H} NMR (126 MHz) spectra; ^^C chemical shifts of 89, 92, 93, 96^ 99 and 107 are obtained from the ^H (500 MHz) detected HMQC and HMBC spectra; * may be reversed; ^ solvent: CDCI3; ^ solvent: CD3OD; n. d.: not detected because of poor signal-to-noise ratio and/or overlapping with solvent signals
c
12^
13^
88^
89^
92^
93^
96^
99b
107^
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26* 27* 28
41.2 67.9 68.0 31.1 40.9 176.8 70.5 39.1 58.0 38.2 22.2 39.6 42.2 51.2 24.7 27.6 52.5 11.5 15.3 40.1 12.3 72.4 76.0 41.4 26.9 22.1 17.2 10.8
39.9 68.0 68.1 26.2 50.7 212.9 46.6 37.7 53.6 42.5 21.1 39.3 42.7 56.4 23.8 27.6 52.5 11.7 13.4 40.1 12.3 72.4 75.9 41.4 26.9 22.0 17.1 10.7
42.3 69.0 69.2 32.8 42.2 179.3 71.8 40.5 59.4 39.2 23.3 41.0 43.7 52.5 25.8 29.0 54.7 12.1 15.8 43.4 13.8 73.1 79.0 45.6 84.1 20.9 27.0 13.9
42.2 69.0 69.2 n.d. 42.2 n.d. 71.8 n.d. 59.4 39.1 n.d. 41.0 43.6 52.4 n.d. n.d. 54.1 12.2 15.8 41.1 13.0 73.7 77.8 36.3 33.4 75.2 12.1 11.6
40.9 69.1 69.4 27.8 52.0 214.5 47.4 39.0 55.0 43.6 22.3 40.8 44.0 57.8 24.9 29.0 54.7 12.2 13.7 43.2 13.7 73.2 79.0 45.5 84.1 20.9 26.9 13.8
40.9 69.1 69.4 27.7 52.0 n.d. 47.4 39.1 55.0 43.6 22.4 40.8 44.0 57.8 24.8 28.6 54.1 12.3 13.8 41.0 13.0 73.8 77.8 36.2 33.4 75.1 12.0 11.5
47.8 72.0 211.4 35.0 58.6 208.2 46.3 37.8 53.4 42.1 21.8 39.2 42.8 56.3 23.8 27.6 52.6 11.8 13.8 40.2 12.4 72.6 76.4 41.4 27.0 22.0 17.2 10.8
45.8 n.d. n.d. 28.9 57.0 n.d. n.d. n.d. 54.8 39.8 n.d. 40.7 43.9 57.8 n.d. n.d. 54.1 12.2 15.6 41.6 13.0 73.4 77.3 42.7 27.9 22.5 17.4 11.1
44.3 72.1 75.8 27.8 56.7 209.8 46.4 n.d. 53.9 42.5 21.5 39.8 43.5 56.4 22.8* 23.6* 54.9 13.8 14.3 77.8 23.2 75.4 71.4 42.2 26.7 21.4 16.0
98.1 75.1 78.1 71.6 77.9 62.8
105.0 n.d. n.d. n.d. n.d. 62.9
98.1 75.1 n.d. n.d. n.d. 62.7
104.8 75.1 67.3 71.7 77.9 62.8
r T 3' 4' 5' 6'
102.8 74.9 n.d. n.d. n.d. 62.5
9.9
547 REFERENCES 1
H. G. Cutler, T. Yokota and G. Adam (Eds), Brassinosteroids - Chemistry, Bioactivity, Applications, ACS Symp. Ser. No. 474, American Chemical Society, Washington DC, 1991. 2 V. Marquardt and G. Adam, in: W. Ebing (Ed.-in-Chief), Chemistry of Plant Protection. Vol. 7, Springer, Berlin, 1991, pp 103-139. 3 A. Sakurai and S. Fujioka, Plant Growth ReguL, 13 (1993) 147. 4 V. A. Khripach, F. A. Lavich and V. N. Zhabinskii, Brassinosteroids (in Russian), Science and Technics, Minsk, 1993. 5 M. D. Grove, G. F. Spencer, W. K. Rohwedder, W. K. Mandava, J. F. Worley, Jr., J. D. Warthen, G. L. Steffens, J. L. Flippen-Anderson and Jr. J. C. Cook, Nature 281 (1979) 216. 6 T. Yokota, S.-K. Kim and N. Takahashi, in 13th Int. Conf. on Plant Growth Substances. Abstr. No. 168, Calgary, 1988. 7 S.-K.Kim, in ref. 1, pp. 26-35. 8 W. Schliemann, Naturwissenschaften 78 (1991) 392. 9 S. Asakawa, H. Abe, Y. Kyokawa, S. Nakamura and M. Natsume, Biosci. Biotech. Biochem. 58 (1) (1994) 219. 10 T. Yokota, S.-K. Kim, Y. Fukui, N. Takahashi, Y. Takeuchi and T. Takamatsu, Phytochemistry 26 (1987) 503. 11 S. Takasuto, H. Abe and K. Gamoh, Agric. Biol. Chem. 54 (1990) 1057. 12 G. Adam and V. Marquardt, Phytochemistry 25 (1986) 1787. 13 N. Ikekawa, S. Takatsuto, T. Kitsuwa, H. Saito, T. Morishita and H. Abe, J. Chromatogr. 290 (1984) 289. 14 K. Richter and J. Koolman, in ref. 1, pp. 265-278. 15 K. Richter and G. Adam, Naturwissenschaften 78 (1991) 138. 16 N. B. Mandava, J. M. Sasse and J. H. Yopp, Physiol. Plant. 53 (1981) 453. 17 W. J. Meudt, Plant Physiol. 83 (1987) 195. 18 J. M. Sasse, Physiol. Plant. 80 (1990) 401. 19 T. Iwasaki and H. Shibaoka, Plant Cell Physiol. 32 (1991) 1007. 20 J. Gartz, G. Adam and H.-M. Vorbrodt, Naturwissenschaften 77 (1990) 388. 21 J. G. Roddick, A. L. Rijuenberg and N. Ikekawa, Physiol. Plant. 87 (1993) 453. 22 J. G. Roddick, Phytochemistry 37 (1995) 1277. 23 H. Ronsch, G. Adam, J. Matschke and G. Schachler, Tree Physiology 12 (1993) 71. 24 K. Hammada, FFTC Book Ser. 1, 34 (1986) 188. 25 G. Adam and U. Petzold, Naturwissenschaften 89 (1994) 210. 26 P. Allevi, M. Anastasia, R. Cerana and P. Ciuffreda, Phytochemistry 27 (1988) 1309. 27 T. Yokota, K. Higuchi, Y. Kosaka and N. Takahashi, in: Progress in Plant Growth Regulation, (C. M. Karssen, L. C. Van Loon, D. Vreugdenhil, Eds.). Dordrecht: Kluwer 1992, pp. 298-305. 28 J. M. Sasse, T. Yokota, P. E. Taylor, P. G. Griffiths, Q. N. Porter and D. W. Cameron, ibid. pp. 319-325. 29 O. N. Kulajewa, E. A. Burkhanowa, A. B. Fedina, N. V. Danilowa, G. Adam, H.-M. Vorbrodt and V. A. Khripach, Dokl. Akad. Nauk SSSR 305 (1989) 1277. 30 S. D. Clouse and D. Zurek, in ref 1, p. 122-140. 31 J. M. Sasse, in ref. 1, pp. 158-166 32 D. M. Zurek and S. D. Clouse, Plant Physiol. 104 (1994) 161. 33 D. M. Zurek, D. L. Rayle, T. C. McMorris and S. D. Clouse, Plant Physiol. 104 (1994) 505. 34 K. Gamoh, I. Yamaguchi and S. Takatsuto, Anal. Sciences 10 (1994) 913. 35 K. Wada, S. Marumo, N. Ikekawa, M. Morisaki and K. Mori, Plant Cell Physiol. 22 (1981)323. 36 K. Wada, H. Kondo and S. Marumo, Agric. Biol. Chem. 49 (1985) 2249. 37 T. Yokota, S. Watanabe, Y. Ogino, I. Yamaguchi and N. Takahashi, J. Plant Growth Regul. 9 (1990) 151. 38 S. Takatsuto, B. Ying, M. Morisaki and N. Ikekawa, J. Chromatogr. 239 (1982) 233. 39 N. Ikekawa and S. Takatsuto, Mass Spectroscopy (Japan) 32 (1984) 55. 40 H. Abe, T. Morishita, M. Uchiyama, S. Takatsuto and N. Ikekawa, Agric. Biol. Chem. 48 (1984) 2171.
548 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 76 77 78 79 80 81 82
R. D. Plattner, S. L. Taylor and M. D. Grove, J. Nat. Prod. 49 (1986) 540. K. Gamoh, K. Omote, N. Okamoto and S. Takatsuto, J. Chromatogr. 469 (1989) 424. J. Schmidt, B. Spengler, T. Yokota, M. Nakayama, S. Takatsuto, B. Voigt and G. Adam, Phytochemistry 38 (1995) 1095. H. Abe, C. Honjo, Y. Kyokawa, S. Asakawa, M. Natsume and M. Narushima, Biosci. Biotech. Biochem. 58 (1994) 986. T. Yokota, M. Nakayama, T. Wakisaka, J. Schmidt and G. Adam, Biosci. Biotech. Biochem. 58 (1994) 1183. F. J. Hammerschmidt and G. Spiteller, Tetrahedron 29 (1973) 3995. J. Schmidt, B. Spengler, T. Yokota and G. Adam, Phytochemistry 32 (1993) 1614. B. Spengler, J. Schmidt, B. Voigt and G. Adam, Phytochemistry (1995) in press. T. Yokota, M. Morita and N. Takahashi, Agric. Biol. Chem. 47 (1983) 2149. J. Schmidt, T. Yokota, B. Spengler and G. Adam, Phytochemistry 34 (1993) 391. A. K. Zaki, J. Schmidt, F. M. Hammouda and G. Adam, Planta Medica 59 (1993) A 613. J. Schmidt, B. Voigt and G. Adam, Phytochemisu-y (1995) in press. J. Schmidt, C. Kuhnt and G. Adam, Phytochemistry 36 (1994) 175. K.-H. Park, J.-D. Park, K.-H. Hyun, M. Nakayama and T. Yokota, Biosci. Biotech. Biochem. 58 (1994) 1343. J. Schmidt and G. Adam, in preparation. H. Suzuki, S. Fujioka, T. Yokota, N. Murofushi and A. Sakurai, Biosci. Biotech. Biochem. 58 (1994) 2075. H. Abe: in ref. 1, pp. 200-207. J. Schmidt, T. Yokota, G. Adam and N. Takahashi, Phytochemistry 30 (1991) 364. J. Schmidt, U. Himmelreich and G. Adam, Phytochemistry (1995) in press. K. Mikami and S. Sakuda, J. Chem. Soc, Chem. Commun. 1993, 710. T. G. Back, P. G. Blazecka and M. V. Krishna, Can. J. Chem. 71 (1993) 156 . B. G. Hazra, P. L. Joshi, B. B. Bahule, N. P. Argade, V. S. Pore and M. D. Chordia, Tetrahedron 50 (1994) 2523 . T. Yokota and K. Mori, Mol. Struct. Biol. Act. Steroids 317 (1992) (M. Bohl and W. L. Duax, Eds., CRC Boca Raton, Florida). H. L. Kwong, C. Sorato, Y. Ogino, H. Chen and K. B. Sharpless, Tetrahedron Letters 31 (1990) 2999. L.-F. Huang, W.-S. Zhou, L.-Q. Sun and X.-F. Pan, J. Chem. Soc, Perkin Trans. 1 1993 1683. H. G. Cutler, in: Natural and Engineered Pest Management Agents, (P. A. Hedin, J. J. Menn, R. M. Hollingworth, Eds.), ACS Symp. Ser. No. 551 (1994) 85. T. C. McMorris and P. A. Patil, J. Org. Chem. 58 (1993) 2338. A. Kolbe, V. Marquardt and G. Adam, J. Labeled Comp. Radiopharm. 31 (1992) 801. B. Voigt, S. Takatsuto, T. Yokota and G. Adam, J. Chem. Soc, Perkin Trans. 1 1995 in press. M. Aburatani, T. Takeuchi and K. Mori, Agric Biol. Chem. 51 (1987) 1909. S. Takatsuto, N. Yazawa, M. Ishiguro, M. Morisaki and N. Ikekara, J. Chem. Soc, Perkin Trans. 1, 1984 139. M. J. Thompson, N. Mandawa, J. L. Flippen-Anderson, J. F. Worley, S. R. Dutky, W. E. Robbins and W. Lusby, J. Org. Chem. 44 (1979) 5002. O. Mitsunobu, Synthesis 1 (1981) 1. B. Voigt, J. Schmidt and G. Adam, in preparation. B. Hellrung, B. Voigt, J. Schmidt and G. Adam, in preparation. K. Wada and S. Marumo, Agric Biol. Chem. 45 (1981) 2579. A. Amann, G. Ourisson and B. Luu, Synthesis 11 (1987) 1002. W. G. Dauben and D. S. Fullerton, J. Org. Chem. 36 (1971) 3277. T. Yokota, Y. Ogino, H. Suzuki, N. Takahashi, H. Saimoto, S. Fujioka and A. Sakurai, in ref. 1, pp. 86-96. H. Suzuki, S. Fujioka, S. Takatsuto, T. Yokota, N. Murofushi and A. Sakurai, J. Plant Growth Regul. 13(1994)21. H. Suzuki, T. Inoue, S. Fujioka, S. Takatsuto, T. Yanagisawa, T. Yokota, N. Murofushi, A. Sakurai, Biosci. Biotech. Biochem. 58 (1994) 1186. N. P. Milner and H. H. Rees, Biochem. J. 231 (1985) 369.
549 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
G. F. Weirich, M. J. Thompson, J. A. Svoboda, Arch. Insect Biochem. Physiol. 12 (1989) 201. T. Yokota, Y. Ogino, N. Takahashi, H. Saimoto, S. Fujioka and A. Sakurai, Agric. Biol. Chem. 54 (1990) 1107. H. Suzuki, S. Fujioka, S. Takatsuto, T. Yokota, N. Murofushi and A. Sakurai, J. Plant Growth. Regul. 12 (1993) 101. T. Yokota, S.-K. Kim, Y. Kosaka, Y. Ogino and N. Takahashi, in: K. Schreiber, H. R. Schutte, G. Sembdner, (Eds.), Conjugated Plant Hormones, Structure, Metabolism and Function, Proceedings of the Int. Symp. Conj. Plant Horm., 3-7 Nov., 1987, Gera, p. 288-296. H. Suzuki, S.-K. Kim, N. Takahashi and T. Yokota, Phytochemistry 33 (1993) 1361. Z. A. Wojciechowski, in: G. W. Patterson and W. D. Nes (Eds.), Physiology and Biochemistiy of Sterols, Am. Oil Chem. Soc, Champain, Illinois, 1991, pp. 361-395. J. Schmidt and G. Adam, (unpublished results). C. D. Schlagnhaufer and R. N. Arteca, J. Plant Physiol 138 (1992) 191. L. Dyas and L. J. Goad, Phytochemistry 34 (1993) 17. M. H. Zenk Phytochemistry 30 (1991) '3861. B. Schneider, A. Kolbe, A. Porzel and G. Adam, Phytochemistry 36 (1994) 319. T. Hai, B. Schneider and G. Adam, Phytochemistry (1995) in press. N. Ikekawa, Bioorg. Med. Chem. Lett. 3 (1993) 1789. 0 . Kappler, C. Hetru, F. Durst, J. Hoffmann, in: J. Koolman (Ed.), Ecdysone - From Chemistry to Mode of Action, Thieme, Stuttgart, 1989, pp. 161-166. T. Hai, B. Schneider, A. Porzel, G. Adam, Phytochemistiy (1995) in press. A. Kolbe, B. Schneider, A. Porzel, B. Voigt, G. Krauss and G. Adam, Phytochemistry 36 (1994) 6-71. A. Kolbe, B. Schneider, A. Porzel, J. Schmidt and G. Adam, Phytochemistry (1995) (in press). A. Kolbe, B. Schneider, A. Porzel and G. Adam, Phytochemistry (1995) in press. H. H. Rees, in: J. Koolman (Ed.), Ecdysone - From Chemistry to Mode of Action, Thieme, Stuttgart, 1989, pp. 28-38. R. Lafont, D. H. S. Horn, ibid. pp. 39-64. L. J. Goad, in: B. V. Charlwood, D. V. Banthorpe (Eds.), Terpenoids, Vol. 7 of "Methods in Plant Biochemistry", Academic Press, 1991, pp. 369-434. A. Porzel, V. Marquardt, G. Adam, G. Massiot and D. Zeigan, Magn. Reson. Chem. 30 (1992) 651. T. Ando, M. Aburatani, N. Koseki, S. Asakawa, T. Mouri and H. Abe, Magn. Reson. Chem. 31 (1993) 94. A. Bax and S. Subramanian, J. Magn. Reson. 67 (1986) 565. A. Bax and F. M. Summers, J. Am. Chem. Soc. 108 (1986) 2093. A. Mijares, D. I. Cargill, J. A. Glasel and S. Liebermann, J. Org. Chem. 32 (1967) 810. C. H. Robinson and P. Hofer, Chem. Ind. (London) (1966) 377. G. Adam and M. Hesse, Tetrahedron 28 (1972) 3527.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
551
Structure Elucidation and Synthesis of the Lignans from the Seeds of Hernandia ovigera L. Masao Arimoto, H. Yamaguchi and S. Nishibe
1 Introduction The lignans are groups of natural products whose carbon skeltons are constructed by the linking of C^Cj-units (1), which are formed biogenetically through the shikimate pathway. The term "lignan", reflecting the woody tissue from which many examples derive, was introduced by Haworth (ref. 1), and it is applied to structures that are composed of two C6C3-units, linked / ? - / ? ' (8-8'). The frequent occurrence of this linkage can be ascribed to ^ - /?' coupling of radicals (2), derived by oxidation of, for example, p-hydroxycinnamyl precursor.
etc.
Lignans, which are main area of our concern in this article, are distributed widely in the plant kingdom. Many biologically active lignans are isolated from medicinal plants. A number of reviews about total syntheses and biological activities have been reported (ref. 2). The plants of Hernandiaceae
comprise about sixty-five species in four generic
kinds. The plants of Hernandia ovigera L. are widespread in the tropics and the subtropics. In Japan, these plants grow on the beaches of Okinawa and Ogasawara islands. In the Okinawa region, the plants have been used as therapeutic agents. A number of groups have been involved in the isolation and structure elucidation of aporphine-type and isoquinoline-type alkaloids from these plant materials as well
552 as in the determination of their biological activities. On the other hand, studies of the non-alkaloidal constituents have been scarcely reported. In 1942, Hata has isolated hernandion (3) and isohernandion (4) from the seeds of Hernandia ovigera L. collected in Taiwan (ref. 3a). In 1972, Furukawa et al. have isolated desoxypicropodophyllin (4) from the barks of the roots of Hernandia ovigera L. collected in the Bonin Islands (ref. 3b). In 1973, Nishino and Mitsui have isolated epiaschantin (5) and epimagnolin (6), which are dioxabicyclooctane-type lignans, from the leaves of Hernandia ovigera and carried out their structure elucidation (ref. 4).
0CH3
<::oci^° "
= H Q
6Y>
HgCO^
o'l ^Ro z
OCH3 3:23'trans-H 4 : 2,3-cw-H
5:Ri,R2=OCH20 6 : Ri=R2=OCH3
Ri
In recent years, new improved methods and techniques used for the isolation of compounds from plant materials, structure elucidation and determination of their biological activity have been developed, and studies in these fields are rapidly progressing. This article comprises a review on the structure elucidation and synthesis of new lignans from the seeds of these plants, research on which has been carried out by the author and co-workers. 2 Extraction and Isolation of the Lignans from the Seeds of Hernandia
ovigera
2.1 Extraction and Isolation of the Lignans (ref. 5) The seeds of Hernandia ovigera (8 kg), which were collected in Okinawa, were crushed into small species. Dried seeds (2 kg) were extracted with MeOH (31 x 3) under reflux. The combined extracts were evaporated under reduced pressure to give a residue (1.09 kg). This residue was treated with petroleum ether to remove soluble materials, which afforded the insoluble materials (518 g). This solid residue was extracted with benzene under reflux. After removal of the solvent under reduced pressure, an oily substance (242 g) was obtained. This oily material was recrystallized from EtOH to afford crude crystals of desoxypodophyllotoxin
553 (3) as needles. The crude crystals were further recrystallized from EtOH to give pure desoxypodophyllotoxin (3) (100 g), which is identical with hernandion isolated by Hata. Desoxypodophyllotoxin (3) and hernandin ( 1 5 ) , having crystalline forms different from desoxypodophyllotoxin contained in the mother liquor, were separated by decantation. The resulting crude crystals were repeatedly purified by recrystallization from EtOH. The mother liquor was concentrated under reduced pressure. The resulting residue, from which the remaining oily materials were removed with hexane extraction, was subjected to silica gel column chromatography (CHCl3:EtOAc = 10:1) to afford fractions 1-^6 as shown in Chart 1. Fraction 2 afforded desoxypodophyllotoxin (3) (60 g) and trace amounts of hernandin (1 5) (0.5 g) by recrystallization from EtOH. The residue, which was obtained by
Chart 1 Separation of the Components from the Seeds of Hemandia ovigera Seeds I extd. with MeOH I coned. MeOH extract extd. with petroleum ether 1 sol.
insol.
coned. extd. with benzene benzene extract recryst. from EtOH 3 + 15
mother liquor coned. extd. with hexane 1 sol.
insol.
silica gel column chromatography
Fr. 1
Fr. 2
Fr. 3
Fr. 4
recryst. from EtOH
Fr. 5
14
1 3 + 15
mother liquor I coned. I silica gel column chromatography
1
Fr. 6
n—I 3
•T
r
5
T 10
16
13
17
4
11
12
554 5
H 4
H3CO
OCH,
OCH^
7: Ri=R2=R3=H 8: Ri=OH,R2=OCH3.R3=H 9: Ri=R3=H. R2=OCH3 13: Ri=OH, R2=R3=OCH3
11: Ri=R3=H. R2= a-OH (2,3-tran5) 12: Ri=R3=H, R2= /? -OH {2,3-trans) 15: Ri=OCH3, R2=R3=H (2,3-&-ans)
5
H3CO" ^
H
"OCH3 OCH3
10: R=OH 14: R=H
Figure 1
concentration of the mother liquor, was repeatedly purified by silica gel column chromatography (CHC^thexane = 1:1) to give bursehernin (7) (9 g) and (-)-yatein (9) (150 mg). Desoxypicropodophyllin (4) (3.6 g) and dehydrodesoxypodophyllotoxin (14) (100 mg) were obtained from fractions 3 and 4 respectively. Fraction 6 was further subjected to silica gel column chromatography (hexane:EtOAc = 5:1) which led to the separation of fractions 6 - 1 ^ 5 . Podorhizol (8) (4 g) was obtained by recrystallization of fraction 6-3. Concentration of the mother liquor of recrystallization under reduced pressure afforded a residue which was purified by preparative thin layer chromatography (hexane:EtOAc = 1:1) to afford hernolactone (16) (40 mg). Fraction 6-4 was repeatedly subjected to preparative thin layer chromatography (CHCl3:EtOAc = 2:1 and hexane:EtOAc = 1:1) to afford 5'-methoxypodorhizol (1 3) (215 mg), epipodophyllotoxin (1 2) (40 mg), and podophyllotoxin (11) (20 mg). Fraction 6-5 was purified repeatedly by recrystallization from MeOH and EtOAc to afford dehydropodophyllotoxin (10) (80 mg).
555 3 Chemical Propaties of the Lignans 3.1 Desoxypodophyllotoxin (3) (ref. 5) Hata first isolated desoxypodophyllotoxin, which showed mp 167'^168°C and [a ]j) -116'' (c=2 in CHCI3), from the seeds of Hernandia ovigera, and he named it "hernandion". However this compound was identical with anthricin which was isolated from Anthricus
sylvestris
HOFFM. (ref. 6). Subsequently it was found to be
identical with a product resulting from dehydroxylation of podophyllotoxin (1 1) by Hartwell and co-workers (ref. 7). Since this compound has been given several names, it is reasonable to call it "desoxypodophyllotoxin" on account of the structural relationship with podophyllotoxin according to Hartwell's proposal (ref. 8). The IR spectrum included bands at 1765 ( C = 0 ) , 1580, 1501 (arom. ring) and 935 (OCHjO) c m ' \ The mass spectrum displayed the molecular ion at m/z 398 in agreement with the formula C22H22O7. The ^H-NMR spectrum (CDCI3) showed two singlets at d 3.75 and 3.80 due to the three methoxy groups. Three singlets at d 6.34, 6.52 and 6.67 were assigned to the four aromatic protons. Two singlets at d 5.92 and 5.95 were assigned to the methylenedioxy group. A broad singlet at d 4.60 and two multiplets centered at d 2,1 A and 3.07 were attributed to the C-4 benzyl methine proton and four protons at the C - 1 , C-2 and C-3 positions. A set of AjB doublet of doublets at d 3.92 and 4.47 was assigned to the a - and /? -methylene protons of the lactone junction. 3.2 Desoxypicropodophyllin (4) (ref. 5) Compound (4), mp 171 ~ 17313 and [ c J^ -35° (c=2 in CHCI3), displayed the molecular ion in its mass spectrum at m/z 398 in agreement with the formula C22H22O7. The IR spectrum included bands at 1775 ( C = 0 ) , 1590, 1510, 1482 (arom. ring) and 930 (OCH2O) cm"'. The ' H - N M R spectrum (CDCI3) showed two singlets at d 3.78 and 3.82 due to the three methoxy groups and three singlets at d 6.36, 6.50 and 6.60 due to the four aromatic protons. A set of A2B doublet of doublets resonating at d 3.98 and 4.44 was assigned to the lactone methylene protons, while a set of A2B doublet of doublets resonating at d 2.50 and 2.88 was due to the C-1 methylene protons. A 1-H doublet at d 4.38 was assigned to the C-4 benzylic methine proton. A 1-H doublet of doublets at d 3.34 was ascribed to the C-3 methine proton adjacent to the carbonyl group. A 1-H multiplet at d 3.02 was assigned to the C-2 methine proton. Two singlets at d 5.90 and 5.93 were attributed to two protons of the methylenedioxy group. On the basis of the above spectroscopic studies, structure (4) was assigned to desoxypicropodophyllin which is the epimer of desoxypodophyllotoxin (3) (ref. 9). Although it is known that desoxypodophyllotoxin (3) isomerizes to desoxypicro-
556 podophyllin (4) by treatment in an alkaline solution, it can be presumed that isomerization does not occur during the extraction and separation process. Therefore desoxypicropodophyllin (4) is essentially present in the seeds of this plants. 3.3 Bursehernin (7) (ref. 5) Compound (7), which showed [ a ]p -28.7° (c=2 in CHCI3), displayed the molecular ion in its mass spectrum at m/z 370 in agreement with the formula C21H22O6. The base peak at m/z 151 represented the 3,4-dimethoxybenzyl cation, (CH30)2C6H3CH2'. The IR spectrum included bands at 1765 ( C = 0 ) , 1609, 1585, 1500 (arom. ring) and 935 (OCHjO) cm ^ The ^H-NMR spectrum (CDCl^) showed six protons, corresponding to the aromatic protons, at d 6.42-^6.80. Two singlets at ^ 3 . 8 1 and 3.84 were assigned to the two methoxy groups. A 2-H singlet at ^ 5.95 was due to the methylenedioxy group. A set of AjB doublet of doublets resonating at ^ 3.82 and 4.12 was assigned to the methylene protons of the lactone junction. On the basis of the above spectroscopic studies, structure (7) was assigned to the bisbenzyl-7 -butyrolactone-type lignan, which is actually identical to
(')-trans-
2-(3,4-dimethoxybenzyl)-3-(3,4-methylenedioxybenzyl)- 7 -butyrolactone isolated from Bursera schlechtendalii
by Cole and co-workers (ref. 10). Therefore this
compound was named bursehernin after consultation with Dr. Cole. 3.4 Podorhizol (8) (ref. 5) Compound (8), mp 125'--126'C and [ a l^ -47.4° (c=l in CHCI3), displayed the molecular ion in its mass spectrum at m/z 416 in agreement with the formula ^22^24^8- T^^ base peak at m/z 135 represented the methylenedioxybenzyl cation, (OCH20)C6H3CH2'. The IR spectrum included bands at 3590, 3500 (OH), 1760 ( C = 0 ) , 1582, 1495, 1480 (arom. ring) and 930 (OCH2O) c m ' \ The ' H - N M R spectrum (CDCI3) showed a 9-H singlet at d 3.80 due to the three methoxy groups and two 1-H doublets at d 6.31 and 6.60, a 2-H singlet at d 6.48 and a 1-H singlet at d 6.22 due to the five aromatic protons. Two 1-H doublets at d 6.92 and 6.94 were assigned to the methylene protons of the methylenedioxy group. A 1-H doublet at d 5.25 was assigned to the methine proton adjacent to the hydroxy group. A set of AjB doublet of doublets resonating at d 3.98 and 4.40 was attributed to the a - and /? -methylene protons of the lactone junction, while a set of A2B doublet of doublets resonating at d 2.25 and 2.47 was due to the benzylic methylene protons. Additionally two multiplets centered at S 2.65 and 2.80 were ascribed to the C-2 and C-3 methine protons respectively. On the basis of the above spectoscopic studies, structure (8) was assigned to podorhizol which was isolated from the roots of Podophyllum
emodi WALL, and
557 Podophyllum
peltatum L. by Kuhn and co-workers (ref. 11).
3.5 Structural Determination of 1,2,3,4-Dehydrodesoxypodophyllotoxin
(14)
(ref. 12) Compound (14) was easily purified by recrystallization of the crude product obtained by silica gel column chromatography of fraction 4. It showed mp 2 7 6 ~ 27810 and the molecular formula CjiHjgO^. It afforded no optical rotation and the characteristic absorption of its UV spectrum suggested that this compound was a naphthalene-type lignan. The ^H-NMR spectrum (CDCI3) showed a 6-H singlet at d 3.85 and a 3-H singlet at d 3.97 due to the three methoxy groups. A 2-H singlet at d 5.38 was assigned to the methylene protons of the lactone junction. A 2-H singlet at S 6.09 was assigned to the methylene protons of the methylenedioxy group, while four singlets at S 6.55, 7.12, 7.21 and 7.71 were due to the five aromatic protons. Structure of this compound (14) was confirmed as 1,2,3,4-dehydrodesoxypodophyllotoxin, by synthesis from desoxypodophyllotoxin by reaction with 2,3-dichloro-5,6-dicyanoquinone (DDQ) as described in section 8.2 later, and comparison of all spectroscopic data. Although compound (14) is a known compound, this is the first report of its natural occurrence in plants. 3.6 Structural Determination of (-)-Yatein (9) (ref. 13) This compound (9), which was obtained as a pale yellow glassy solid on purification of fraction 2 by silica gel column chromatography, showed nearly identical spectroscopic data with that of bursehernin (7). The mass spectrum of compound (9), which showed [ a ]^ -29.5° (c=0.48 in CHCI3), displayed the molecular ion at m/z 400.1520 in agreement with the formula C22H24O7. The base peak at m/z 181 represented the 3,4,5-trimethoxybenzyl cation, (CH30)3C6H2CH2*. The IR spectrum included bands at 1770 ( C = 0 ) , 1595, 1505 and 1490 (arom. ring) c m ' \ The ' H NMR spectrum (CDCI3) showed five protons, corresponding to the aromatic protons, at S 6 . 3 6 ~ 6 . 7 2 . Two 1-H singlets at d 5.93 and 5.94 were assigned to the methylene protons of the methylenedioxy group. A set of A2B doublet of doublets resonating at S 3.90 and 4.18 was assigned to the a - and /? -methylene protons of the lactone junction ( / = 9 . 5 , 7.5 Hz). A 9-H singlet at d 3.83 was assigned to the three methoxy groups. Two multiplets centered at d 2.88 and 2.90 were attributed to six protons at the C-2, C-3 and two benzylic positions. Concerning the absolute configuration, a circular dichroism (CD) spectrum showed negative Cotton effect at 239 and 275 nm, suggesting that the absolute configurations of the C-2 and C-3 positions are 2R and 3R (ref. 14). The above spectroscopic studies led to the conclusion that this compound is
558 (-)-yatein (ref. 15). (-)-Yatein is widely distributed in many plants such as Burseraceae (ref. 10), Umbelliferae (vtf. 16), Cupressaceae (rtf. 17), Piperaceae (ref. 18), and Hernandiaceae (ref. 19). It was synthesized by Koga et al (ref. 20). 3.7 Structural Determination of (-)-Dimethylmatairesinol (17) (ref. 21) Compound (17) was isolated from the mother liquor, from which (-)-yatein (9) was obtained by using preparative thin layer chromatography, as colorless needles, mp 123 — 1 2 5 1 : , with an optical rotation of [ a l^ -24.8° (c=1.41 in CHCI3). The mass spectrum of compound (1 7) displayed the molecular ion at m/z 386.1728 by means of high resolution mass spectrometry, in agreement with the formula C22H26O6. Mass fragment peaks were seen at m/z 208, 177, 151 and 107. Fragment peaks at m/z 208 and 151 represented the fragment ion (18) and the benzyl cation (1 9 ) , which are similar to that of bursehernin (7), respectively.
C*H
H3CO H3CO OCH3 OCH3
18 : m/z 208
OCH3
fragment ion of 17
17 : (-)-dimeihylmatairesinol
Figure 2
The UV spectrum showed absorption maxima at 230 and 281 nm. The IR spectrum included bands at 1770 ( C = 0 ) , 1605, 1600 and 1520 (arom. ring) cm"\ The ^H-NMR spectrum (CDCI3) showed six protons, corresponding to the aromatic protons, at d 6 . 4 9 - 6 . 7 9 . Four 3-H singlets at d 3.82, 3.84, 3.85 and 3.86 were ascribed to the four methoxy groups. A set of A2B doublet of doublets at d 3.89 and 4.12 was assigned to the a - and /? -methylene protons of the lactone junction ( / = 9 . 2 , 6.9 Hz). Two multiplets centered at d 2.57 and 2.59 were assigned to six protons due to the four benzylic protons and the two methine protons of the C-2 and C-3 positions. These physical data coincided with that of (-)-dimethylmatairesinol (17) isolated from Cinnawonium
camphola SIEB. (ref. 22). Although direct comparison with
the authentic sample was unavailable, additional confirmation was obtained by means of a circular dichroism spectrum. The CD spectrum showed negative Cotton
559
Chart 2
Circular Dichroism Curves of Dibenzylbutyrolactone in EtOH
bursehernin (7)
podorhizol (8)
5'-methoxypodorhizol (1 3)
(-)-yatein (9)
(-)-dimethylmatairesinol (1 7)
effects at 233 and 276 nm as in the cases of compounds (7), (8), (9), (1 7) and analogous compounds (ref. 14) which is shown in Chart 2, suggesting that the absolute configurations at the C-2 and C-3 positions are 2R and 3R respectively. On the basis of the above spectroscopic studies, structure (17) was assigned to the compound [(-)-dimethylmatairesinol]. 3.8 Structural Determination of 5'-Methoxypodorhizol (13) (ref. 21) Fraction 6-4 was purified by means of preparative thin layer chromatography, followed by recrystallization from benzene-hexane. Compound (13) was obtained as colorless needles, mp 1 2 7 ~ 129°C, with an optical rotation of [ a l^ -49.1° (c= 1.0 in CHCI3). In the UV spectrum, absorption maxima were seen at 235 and 281 nm. The mass spectrum displayed the molecular ion at m/z 446.1583 by means of high resolution mass spectrometry, in agreement with the formula C23H26O9. Fragment peaks were seen at m/z 249, 197, 181, 165, 138 and 125. The IR spectrum included bands at 3600, 3450 (OH), 1760 ( C = 0 ) , 1635, 1595, 1510 (arom. ring) and 935 (OCUfi)
c m ' \ The ' H - N M R spectrum (CDCI3) showed a 12-H singlet at
d 3.84 due to the four methoxy groups and two 1-H doublets at d 5.97 and 6.03
560 and a 2-H singlet at d 6.50 due to the four aromatic protons. Two doublets at d 5.92 and 5.95 were assigned to the methylene protons of the methylenedioxy group, A 1-H broad singlet at d 5.29 was assigned to the methine proton adjacent to the hydroxy group. A set of AjB doublet of doublets resonating at d 3.98 ( / = 9 . 0 , 5.7 Hz) and 4.38 ( / = 9 . 0 , 7.8 Hz) was assigned to the a- and ^ - m e t h y l e n e protons of the lactone junction, while a set of AjB doublet of doublets resonating at d 2.26 ( / = 1 3 . 8 , 7.8 Hz) and 2.46 (7=13.8, 7.8 Hz) was due to the benzylic methylene protons. A 1-H doublet of doublets at d 2.66 (7=6.3, 3.0 Hz) was assigned to the methine proton adjacent to the carbonyl group. Additionally a broad singlet at d 2.56 was ascribed to a hydroxy group which disappeared on the addition of DjO. The mass fragment peaks at m/z 249 and 165 represented the fragment ion (2 0) and the benzyl cation (2 2) respectively, revealing that one methoxy group existed in the same benzene nucleus together with the methylenedioxy group. The fact that
OCH3 13 : 5'-methoxypodorhizol
H3CO 22 : m/z 165 Fragment Ion of 13
Figure 3
two aromatic protons which have small 7 values appear at 8 5.91 (d, 7=1.5 Hz) and 6.03 (d, 7=1.5 Hz) in the ' H - N M R spectrum shows that a methoxy group is situated at the C-5' position. The CD spectrum shows negative Cotton effects at 238 and 280 nm, as in the cases of 7, 8 and 9 (Chart 2), revealing that the absolute configurations at the C-2 and C-3 positions are 2 5 and 3R respectively. Regarding the absolute configuration at the C-6 position, it was known that the C-6 proton appeared at d 5.28 (7=2.0 Hz) in podorhizol (8) (C-6 S) and at d 4.83 (7=7.5 Hz) in epipodorhizol (C-6 R) (ref. 23). In our compound, the C-6 proton appeared at S 5.28 (7=3.0 Hz), indicating that the absolute configuration at the C-6 position
561 is S. From the above results, compound (13) was determined to be 5'-methoxypodorhizol isolated from Hcrnandia cordigera (ref. 24). 3.9 Proof of the Existence of Podophyllotoxin (1 1) and Epipodophyllotoxin (1 2) Fraction 6-4 was repeatedly purified by preparative thin layer chromatography (CHCljiacetone = 10:1) to afford a small amount of compound (11 ) and epipodophyllotoxin (1 2 ) . The ^H-NMR spectrum (CDCI3) of compound (1 1 ) showed a 6-H singlet at S 3.76 and a 3-H singlet at (^3.81 due to the three methoxy groups. A 2-H singlet at S 6.37 and two 1-H singlets at S 6.51 and 7.12 were assigned to the four aromatic protons. Two 1-H doublets at S 5.97 and 5.99 were assigned to the methylene protons of the methylenedioxy group. A 1-H doublet at d 4.77 was assigned to the methine proton at the C-1 position adjacent to the hydroxy group. A 1-H triplet at S 4.09 was due to the (^ -methylene proton of the lactone junction. Two multiplets centered at d 2.80 and d 4.58 were attributed to the methine protons at the C-2 and C-3 positions, the methine proton at the C-4 position and the a -proton of the lactone methylene group respectively. The ^H-NMR spectrum (CDCI3) of compound (1 2) showed a 6-H singlet at d 3.74 and a 3-H singlet at d 3.80 due to the three methoxy groups. Four aromatic protons resonated as a 2-H singlet at d 6.28 and two 1-H singlets at d 6.55 and 6.88. A set of AjB doublet of doublets resonating at d 4. 36 (7=8.4, 5.3 Hz) and 4.39 (7=12.9, 8.4 Hz) was assigned to the a - and /? -methylene protons of the lactone junction. A 1-H doublet of doublets at S 3.29 and 1-H doublet at S 4.61 were assigned to the C-3 and C-4 methine protons respectively. A 1-H doublet at ^ 4 . 8 0 was attributed to the C-1 methine proton adjacent to the hydroxy group. Two 1-H doublets at S 5.97 and 6.00 were ascribed to the methylene protons of the methylenedioxy group, while a multiplet centered at S 2.90 was due to the C-2 methine proton. On the basis of the above spectroscopic studies, the compounds were identified as podophyllotoxin (1 1) and epipodophyllotoxin (1 2) were confirmed by comparison of their spectra data with authentic samples. 4 Studies on 1,2,3,4-Dehydropodophyllotoxin (10) (ref. 13) 4.1 Structure of 1,2,3,4-Dehydropodophyllotoxin
(10)
Fraction 6-5 was repeatedly purified by recrystallization from methanol and ethyl acetate to afford compound (10) as needle crystals, mp 275~280°C ( d e c ) . Compound (1 0) showed no optical rotation and was almost insoluble in most organic solvents. The mass spectrum displayed the molecular ion at m/z 410.1013 by high resolution mass spectrometry, in agreement with the formula C22Hi80g. The
562 mass fragment peaks were seen at m/z 395, 367, 350, 337, 281 and 139. The UV spectrum showed absorption maxima at 222, 262, 322 and 355 nm. The IR spectrum included bands at 3500 (OH) and 1770 (C=0) cm ^ The ^H-NMR spectrum (CDCI3) showed a 6-H singlet at d 3.83 and a 3-H singlet at S 3.95 due to the three methoxy groups. Two 1-H singlets at d 7.10 and 7.48 and a 2-H singlet at S 6.52 were assigned to the four aromatic protons. Two 2-H singlets at S 5.37 and 6.10 were attributed to the methylene protons of the lactone junction and the methylenedioxy group respectively. A 1-H singlet at d 12.00 which disappeared on addition of DjO was assigned to the hydroxy group. On the basis of the above spectroscopic studies, structure (10) was assigned to a phenylnaphthalene-type lignan. 4.2 Structure of 1,2,3,4-Dehydropodophyllotoxin Methyl Ether (23) (ref. 13) The following reaction was carried out in order to determine the position of the hydroxy group. A suspension of compound (10) in ether-acetone (1:2) was methylated with diazomethane in ether. After usual work-up, the resulting residue was purified by silica gel column chromatography (CHCl3:acetone = 9:1). The obtained product was recrystallized from ethanol to give compound (2 3) as colorless needles, mp 2 6 8 ~ 2 7 0 * C The mass spectrum of the methyl ether ( 2 3 ) , which
Table 1
H-NMR Chemical Shifts of 23 and Related Compound 24 and 25
H3CO 5 . ^
OCH3
OCH3
24 Ri=0CH3 R2=R3=H
25 Ri=R3=H R2=OCH3
Ci-H
8.15
7.64
Cg-H
—
6.98
C5-H
6.85
—
Cx-OCHs"^
4.20
3.47
a) Meihoxy group on ring A and ring B
23
7.57 7.07
4.09
563 showed no optical rotation, displayed the molecular ion at m/z 424.1161 by high resolution mass spectrometry, in agreement with the formula CjjHjoOg. Mass fragment peaks were seen at m/z 409, 381, 350, 224 and 171. The IR spectrum included a band at 1770 (C=0) cm'\
The ^H-NMR spectrum (CDCI3) showed a 6-H
singlet at d 3.84 and two 3-H singlets at d 3.96 and 4.09 due to the four methoxy groups. Two 1-H singlets at d 7.07 and 7.57 and a 2-H singlet at d 6.52 were assigned to the four aromatic protons. Two 2-H singlets at S 5.52 and 6.09 were attributed to the methylene protons of the lactone junction and the methylenedioxy group respectively. The methyl ether (2 3) was compared with analogous known lignans. The ^HNMR spectra of compound (2 3) differed from that of dehydro- /? -peltatin methyl ether (2 4) which has the methoxy group at the C-8 position (ref. 25) and from that of dehydrohernandin (25) which has the methoxy group at the C-5 position (ref. 11) as shown in Table 1. Therefore the methoxy group of compound (23) was presumed to be located at the C-1 position of ring B. Although compound (2 3) is a well-known compound (ref. 26), no precise spectral data have been given except in one report (ref. 27). We therefore decided to synthesize compound (2 3) in order to make a direct comparison with an authentic sample. Although Gensler and coworkers derived compound (2 3) from podophyllotoxone in one step by means of selenium oxide treatment (ref. 26), their method gave an unsatisfactory result in our reexamination and we prepared compound (2 3) by another route as shown in Chart 3. 4.3 Synthesis of Dehydropodophyllotoxin (1 0) from Podophyllotoxin (11) (ref. 13) Podophyllotoxin (11 ) was used as the starting material, and it was silylated with ferr-butyldimethylsilyl chloride (TBDMSCl) in the presence of imidazole in dimethylformamide at 75*C (ref. 28). The resulting product was purified by silica gel column chromatography (CHC^iEtOAc = 10:1) to give podophyllotoxin-TBDMS form (2 6) (2,3-trans)
and picropodophyllin-TBDMS form (2 7) (2,3-Cis) which
isomerized at the C-3 position in a ratio of 5.2:1. terr-Butyldimethylsilylpodophyllotoxin (2 6) was obtained as a colorless amorphous powder, which showed an optical rotation of [ a ]^ -86.2
(c=0.62 in
CHCI3). The mass spectrum displayed the molecular ion at m/z 528.2174 by high resolution mass spectrometry, in agreement with the formula C28H3608Si. Mass fragment peaks were seen at m/z 4 7 1 , 397, 3 5 1 , 313, 282, 259, 229, 185, 181 and 168. The IR spectrum included bands at 1780 (C=0) and 935 (OCUfi)
cm"\ The
^H-NMR spectrum of compound (26) showed two 3-H singlets at (^ 0.11 and 0.28 and a 9-H singlet at d 0.95 due to the two methyl and tert-butyl
groups adjacent to
564 TBDMSO
H
" iH 6 HgCO"^ 0CH3
OCH3
OCH3
26 : 2,3-trans, 3,4-0/5
^®
27 : 2,3-cys, 3,4-frans
i TBDMSCl-imidazole / DMF, 75X: ii DDQ / benzene, reflux iii BU4NF / THF, room temp.
H3CO'
^OCHo OCH3 10
Charts
Si-atom. A 6-H singlet at d 3.75 and a 3-H singlet at S 3.81 were assigned to the three methoxy groups. Two 1-H doublets at S 4.57 and 4.80 were assigned to the C-4 methine proton adjacent to the aryl group and the C-1 methine proton adjacent to the 0-silyl group respectively. A 2-H singlet at S 6.38 and two 1-H singlets at d 6.48 and 6.94 were due to the four aromatic protons, while two 1-H doublets at ^ 5 . 9 5 and 5.98 and two 1-H signals at ^ 4 . 0 0 and 4.50 were attributed to the methylenedioxy group and the methylene protons of the lactone junction respectively. A 2-H multiplet centered at d 2.85 was assigned to the C-2 and C-3 methine protons. On the other hand, rerr-butyldimethylsilylpicropodophyllin (2 7) was obtained as needle crystals by recrystallization from MeOH. It showed mp 169^171*C and an optical rotation of [ a ]^ +56.6° (c=0.47 in CHCI3). The mass spectrum displayed the molecular ion at m/z 528.2179 by high resolution mass spectrometry, in agreement with the formula C28H3608Si. Mass fragment peaks were seen at m/z 4 7 1 , 412, 397, 3 5 1 , 313, 282, 259, 229, 185, 181 and 168. The IR spectrum included bands at 1780 (C=0) and 940 ( O C H p ) cm"\ The ' H - N M R spectrum of compound (27) showed two 3-H singlets at d 0.15 and 0.20 and a 9-H singlet at d 1.01 due to the two methyl and tert-buty\
groups adjacent to Si-atom. A 6-H singlet at d
3.84 and a 3-H singlet at d 3.87 were assigned to the three methoxy groups. Two 1-H doublets at S 4.47 and 4.51 were assigned to the C-4 methine proton adjacent to the aryl group and the C-1 methine proton adjacent to the O-silyl group respectively. A 2-H singlet at d 6.38 and two 1-H singlets at d 6.48 and 6.94 were due
565 to the four aromatic protons, while two 1-H doublets at S 5.90 and 5.92 and two 1-H signals at S 3.93 and 4.37 were attributed to the methylenedioxy group and the methylene protons of the lactone junction respectively. A 1-H multiplet centered at d 2.60 was assigned to the C-2 methine proton, while a 1-H doublet of doublets resonating at ^ 3 . 2 1 was assigned to the C-3 methine proton adjacent to the carbonyl group. Subsequent aromatization of compound (2 6) using DDQ in dry benzene afforded compound (2 8 ) . The crude product was purified by silica gel column chromatography (hexanerEtOAc = 7:3) and by recrystallization from ethanol. Pure tert-butyldimethylsilyldehydropodophyllotoxin (2 8) was obtained as colorless needles, mp 245'^248°C, which showed no optical rotation. The mass spectrum displayed the molecular ion at m/z 524.1878 in agreement with the formula C28H320gSi. Mass fragment peaks were seen at m/z 509, 423, 409 and 350. The UV spectrum showed absorption maxima at 206, 264, 318 and 352 nm, and the IR spectrum included bands at 1770 ( C = 0 ) and 940 ( O C H p ) c m \ The ' H - N M R spectrum showed a 6-H singlet at <^ 0.28 and a 9-H singlet at S 1.01 due to the two methyl and rerr-butyl groups adjacent to Si-atom. A 6-H singlet at S 3.84 and a 3-H singlet at d 3.96 were assigned to the three methoxy groups. Two 2-H singlets at S 5.33 and 6.08 were attributed to the methylene protons of the lactone junction and the methylenedioxy group respectively, whereas a 2-H singlet at d 6.53 and two 1-H singlets at d 7.08 and 7.46 were due to the four aromatic protons. Desilylation of compound (2 8) gave compound (10) by using tetrabutylammonium fluoride in THF. The crude product was subjected to high performance liquid chromatography (HPLC) (column, microporasil 3 i.d. x 250 mm; eluent, CHCI3; flow rate, 0.5 ml/min.). Pure dehydropodophyllotoxin was found to be identical with compound (10) on direct comparison in all its spectral data. 5 Studies on Hernolactone 5-1 Structure Elucidation of Hernolactone (1 6) (ref. 25) Fraction 6-3 was purified by preparative thin layer chromatography of the mother liquid, from which podorhizol (8) was separated by recrystallization from EtOH, to afford compound (1 6 ) , named hernolactone, in a small quantity as a pale yellow solid. It showed an optical rotation of [ a ]^ -30° (c=0.36 in CHCl^). The mass spectrum displayed the molecular ion at m/z 432.1765 by high resolution mass spectrometry, in agreement with the formula C23H280g. Mass fragment peaks were seen at m/z 2 5 1 , 238, 208, 194, 181, 167 and 151. In the UV spectrum, absorption maxima were seen at 225 and 272 nm. In the IR spectrum, the hydroxy group and the carbonyl group absorptions were observed at 3585 and 1775 c m ' \
566 respectively. The ^H-NMR spectrum (CDCI3) showed two 2-H singlets at d 6.21 and 6.35 due to the four aromatic protons. A 1-H singlet at d 5.46 was assigned to a hydroxy proton which disappeared on addition of DjO. A 6-H singlet at ^ 3 . 8 1 and a 9-H singlet at ^ 3.82 were assigned to the five methoxy groups. Other signals were due to the presence of a lactone methylene (d 3.92 and 4.20), two benzylic methylene, and two protons of the lactone junction ((^ 2.94 and 2 . 4 6 ^ 2.68). In fact the signals of the four protons including the two C-2 and C-3 methine protons of the lactone junction and the C-5 benzylic methylene protons appeared together at d 2 . 4 6 - ^ 2 . 6 8 as a multiplet. The other C-6 benzylic methylene protons appeared at d 2.94 as a mutiplet, while the C-4 a - and ^ -methylene protons of the lactone junction appeared at d 3.92 and 4.20 as two doublet of doublets, showing that this compound (16) is of the trans-2,3-bisbenzylbutyrolactone-type, like hinokinin (2 9) (ref. 29), dimethylmatairesinol ( 1 7 ) (ref. 22) and (-)-yatein (9) (ref. 15) as shown in Figure 4. In the case of (-)-yatein (9), the trans structure was clearly confirmed by comparison with the cis-type compound (ref. 23). From the above results, four possible structures, la, lb, Ila and lib as shown in Figure 5, can be presumed for the structure of hernolactone. In the
C-NMR spectrum of hernolactone, one methoxy group signal was seen
5 H 4
29 : Ri, R2=R3, R4=OCH20; R5=H 17: Ri=R2=R3=R4=OCH3; R5=H 9 : Ri, R2=OCH20; R3=R4=R5=OCH3
Figure 4
H3C0^ 3
H3CO OCH3
Ila : Ri=H, R2=CH3 lib : Ri=CH3, R2=H
la: Ri=H, R2=CH3 lb : Ri=CH3, R2=H
Figure 5
567 to be shifted to lower magnetic field (60.9). The chemical shifts of the methoxy groups were examined in comparison with those of simple compounds such as 1,2,3-trimethoxybenzene, 2,3-dimethoxyphenol and 2,6-dimethoxyphenol (ref. 30). In these compounds, the signal of a methoxy group situated between two methoxy groups or between the methoxy and the hydroxy groups is slightly shifted to the lower magnetic field (approx. 60 ppm) as compared with ordinary methoxy groups (approx. 55 ppm ) as shown in Figure 6. If hernolactone has the structure lb or lib in which a methoxy group of ring A is situated between the methoxy and the hydroxy groups, two such signals due to the methoxy groups at the C-4' and C-4"
137.5
135.4 60.4 55.4 \ 0 C H 3
59.9
134.1
55.3 X ? "
152.1 T |ri49.0 103.8 ^ ^ 1 0 7 . 9 123.6
152.7 104.7 122.9
T ||146.4 ^ i X ^ 104.2 118.0
Figure 6
H3CO
OCH^3
O
Chart 4
OCH3
Fragment Ion of Hernolactone (16)
568 Table 2
^^C-NMR Data of Hemolactone
Carbon
Chemical shift
1 2 3 4 5 6 1• 1" T 2" 3' 3"
178.7 46.5 41.3 71.3 38.8 35.2 129.0 133.5 105.2 106.3 147.2 153.3
Carbon 4' 4" 5' 5" 6' 6" 3'-OCH3 3"-OCH3 5'-OCH3 5"-OCH3 4"-OCH3
Chemical shift 133.6 137.0 147.2 153.3 105.2 106.3 56.3 56.1 56.3 56.1 60.9
positions should be seen. Accordingly the structures lb and lib were rejected. The assignment of every carbon was deduced by means of distortionless enhancement by polarization transfer (DEPT) and heteronuclear correlated 2D spectroscopy (HETCOR) (ref. 31) (Table 2 ). In the mass spectrum, fragment peaks were seen at m/z 167, 181, 194, 238, 251 and 265, and each fragment ion was assigned as shown in Chart 4 (ref. 22,29,32). The peak at m/z 238 shows that the position of the carbonyl group of the lactone ring is as in la, and hence the alternative structure Ila is rejected. As regards the absolute configurations of the lactone junction, negative Cotton effects appeared at 237 and 277 nm in a circular dichroism spectrum, as in the case of thujaplicatin methyl ether (3 0) and analogous compound (ref. 14) and (-)-yatein (9), suggesting that the absolute configurations at the C-2 and C-3 positions are IR and ZR. In addition, the spectral data of the methyl ether of compound (1 6) coincided well with those of (-)-cubebininolide (ref. 33). In conclusion, compound (16) (hemolactone) is confirmed to be (2/?,3/?)-3-(4'-hydroxy-3',5'-dimethoxybenzyl)-2- ( 3 " , 4",5"-trimethoxybenzyl)- y -butyrolactone.
30 : Ri=0CH3, R2=R4=OH, R3=H 9 : Ri, R2=OCH20, R3=H, R4=OCH3 16 : Ri=R3=R4=OCH3, R2=0H H3CO
Figure 7
569
thujaplicatin methyl ether (30) hernolactone (1 6) (-)-yatein (9)
Chart 5
Circular Dichroism Curves in EtOH
5.2 Synthesis of (±)-HernoIactone (ref. 34) Hernolactone (1 6) is a new bisbenzyl- y -butyrolactone-type lignan isolated from Hernandia ovigera. Although the syntheses of b i s b e n z y l - / -butyrolactones have been reported by many groups (ref. 2), its synthesis was carried out according to the known procedure (ref. 35) shown in Chart 6. Commercially available syringaldehyde (31) was reacted with benzyl chloride in the presence of anhydrous potassium carbonate in dimethylformamide to afford the benzyl ether (32) as pale yellow needles, mp 63.1 ^ 6 3 . S^C. Benzyl syringaldehyde (32) was converted to the corresponding phenyldithioacetal (33) by reacting with thiophenol in the presence of boron trifluoride-etherate followed by condensation with butenolide by Michael addition in /j-butyllithium to give the crude lactone ( 3 4 ) . It was purified by means of HPLC (column, Cosmosil 5Ci8, i.d. x 250 mm ; eluent, CH3CN ; flow rate, 1.5 ml/min.) to afford the product (34) as a pale yellow solid. The mass spectrum displayed the fragment ion corresponding to the loss of a thiophenyl group at m/z 449.1418 in agreement with the formula C26H25O5S (M^-SPh). The IR spectrum included a band at 1770 (C=0) cm'\
The
^H-NMR spectrum (CDCI3) showed a 15-H multiplet centered at S 7.34 and a 2-H singlet at d 6.87 due to the aromatic protons. Two sets of AjB doublet of doublets resonating at ^ 4 . 3 4 (/=9.6, 8.1 Hz) and 4.46 ( / = 9 . 6 , 7.5 Hz), and (^2.68 (/= 18.0, 9.6 Hz) and 2.92 ( / = 1 8 . 0 , 7.8 Hz) were assigned to the four a - and /? -
570
HoCO
HaCO^^^^^CHO
J7
HO
PhS SPh HgCO^^^^Xj^ iii
BzO' OCH3 32
0CH3 31
PhS
ii
OCH3 33 PhS
SPh
SPh
H3CO,
-•
16
BzO
OCH3
i C6H5CH2CI, K2CO3; ii C6H5SH, BFs-etherate; iii butenolide, n-BuLi; iv 3,4,5-(CH30)3C6H2CH2Br. HMDS-HMPA; v Raney-Ni
Chart 6
methylene protons of the lactone junction adjacent to the oxygen atom and the carbonyl group. A 6-H singlet at d 3.75 and a 2-H singlet at d 5.05 were ascribed to the two methoxy groups and the benzylic methylene protons respectively. A 1-H quintet at S 3.23 was attributed to the C-3 methine proton situated in the lactone junction. Compound (34) was allowed to react with lithium hexamethyldisilylamide prepared from hexamethyldisilazane and n-butyllithium in THF at -SO'C (ref. 35c). The generated anion was reacted with 3,4,5-trimethoxybenzyl bromide (ref. 36) derived from 3,4,5-trimethoxybenzyl alcohol and phosphorous tribromide in ether to give the alkylated product. After usual work-up, the resulting residue was purified by silica gel column chromatography (hexaneiEtOAc = 2:1) to afford pure product (35) as a pale yellow solid. The absorption band centered at 1765 cm"^ in the IR spectrum was assigned to the C = 0 stretching vibration. The ^H-NMR spectrum of the compound (35) showed a 15-H multiplet centered at d 7.29 and two 2-H singlets at d 6.23 and 6.88 due to the aromatic protons. Two 6-H singlets at d 3.68 and 3.74 and a 3-H singlet at d 3.82 were assigned to the five methoxy groups. A 2-H doublet of doublets resonating at d 2.79 (7=13.8, 5.7 Hz) and 3.18 ( / = 1 3 . 8 , 4.2 Hz) was assigned to the C-6 benzylic methylene protons. Two 1-H doublets of triplets resonating at S 2.98 ( / = 8 . 1 , 3.9 Hz) and 3.38 (/= 6.0, 4.2 Hz) were attributed to the C-3 and C-2 methine protons of the lactone junction respectively. A set of A2B doublet of doublets resonating at d 3.62 (J= 9 . 9 , 8.7 Hz) and
571 4.41 ( / = 9 . 9 , 3.6 Hz) was ascribed to the C-4 methylene protons in the lactone junction, while a 2-H singlet at 8 5.03 was due to the benzylic methylene protons of the protecting benzyl ether. Subsequently by removal of the protecting groups of compound (35) with Raney-Ni, compound (16) was obtained as needles, mp 1 1 6 ^ 1 1 8 ' ' C . The spectral data were identical with those of natural compound (1 6). 6 Studies on Hernandin 6.1 Structure Elucidation of Hernandin (15) (ref. 12) Compound (15) was obtained as floating feather-like crystals which were deposited in the upper layer of the recrystallizing mother liquid of the desoxypodophyllotoxin fraction, while desoxypodophyllotoxin remaining in the solution crystallized as heavy prisms at the bottom of the flask. The upper layer containing the light feather-like crystals was decanted off and recrystallized from ethanol. It showed mp 210'^213°C and an optical rotation of [ o- j ^ -70° (c=0.7 in CHCI3). The mass spectrum displayed the molecular ion at m/z 428 in agreement with the formula C23H240g. In the IR spectrum, the absorption band centered at 1770 cm"^ was assigned to the C = 0 stretching vibration. In the *H-NMR spectrum (CDCI3), signals due to the lactone methylene protons at d 3.89 and 4 . 4 5 , the methylenedioxy group at d 5.92 and 5.94, the three aromatic protons at S 6.38 and 6.44 and the four methoxy groups at S 3.63, 3.76 and 3.80 were apparent. From the above facts, compound (1 5) was presumed to be tetrahydronaphthalene-type lignan with a trimethoxyphenyl group, a methoxy group and a methylenedioxy group. A known lignan with the same molecular formula and functional groups was reported as /?-peltatin A methyl ether (36) (ref. 37), mp 162-^163°C, [ o ID -70° (in CHCI3). By comparison of the ^H-NMR spectra of both compounds, as shown in Table 3, it was apparent that the greater part of the signals of compound (1 5) and ( 3 6 ) coincided almost exactly, with the exception that an aromatic
0CH3,
„
572 Table 3 Comparison of ^H-NMR Spectrum of Compound (1 5) with that of Compound (36) ( ^ , ppm. CDCI3)
0 arom. H (3H)
C4-H
H2
CH3O X 4
Cl,2,3-H
12H
0 Cg
36
—
C2S6'
C5
6.36
6.28
2H
IH
(s)
(s)
6.44
6.38
IH
2H
(s)
(s)
1 5
0 4.59
3.30-2.40 4H
5.93 IH (d)
4.48 IH (dd)
4.87
3.10-2.58
5.92 IH (d)
3.89 IH (dd)
5.94 IH (d)
4.45 IH (dd)
(d)
(m)
C3'.5- C5
3.97 4.04 3.81 3.76 IH (dd) 3H 3H 6H
(m)
4H
C4
5.91 IH (d)
(d)
—
Cg
(s)
(s)
—
(s)
3.80 3.76
3.63
3H
6H
3H
(s)
(s)
(s)
—
proton of compound (15) in the tetralin ring appeared at a lower magnetic field than in compound ( 3 6 ) ( 1 5 ; ^ 6 . 4 4 , 3 5 ; ^ 6.28) and a methoxy group of compound ( 1 5 ) in the tetralin ring appeared at a higher magnetic field than in compound (3 6) ( 1 5 ; (^3.63, 3 6 ; ^ 4 . 0 4 ) . In the mass spectra of compounds (1 5) and ( 3 6 ) , as shown in Chart 8, not only the molecular ion peaks at m/z 428 but also other prominent ion peaks at m/z 260, 215, 2 0 3 , 181 and 168 were common to both compounds. The mass spectra of analogous lignans have been reported by Duffield and Pelter (ref. 38). Therefore each fragment ion was assigned as shown in Chart 8b. The above results show that compound (1 5) has a trimethoxyphenyl group and a tetralin ring, and that a methoxy and a methylenedioxy group on the benzene nucleus. By aromatization using DDQ, compound (15) was led to the naphthalene-type compound ( 3 7 ) , mp 3 1 0 ^ 3 1 3 ° C , which had no asymmetric carbon. In the IR spectrum of compound ( 3 7 ) , the absorption band centered at 1770 cm'^ was assigned to the C = 0 stretching vibration. The ^H-NMR spectrum showed two 1-H singlets at d 6.98 and 7.64 and a 2-H singlet at S 6.40 due to the four aromatic protons at the C-8, C-1 and C-2' ,6' positions. Two 2-H singlets at d 5.30 and
573 428
100%-,
a)
181
10 -J
203
215 260 168
lil..klliJlli 100
illl ill III 200
i I III. II lit
400
m/z 428
100%-,
10 H
-k^-WS<° lllllllllllll 100 Chart 8
IILIIIIIIMLHII 200
Ji I Ki^Mi J h i 3f)0
400
m/z
M a s s S p e c t r u m of C o m p o u n d (1 5 ) (a) and ( 3 7 ) (b)
6 . 0 6 w e r e a s s i g n e d to the m e t h y l e n e p r o t o n s of the l a c t o n e j u n c t i o n a n d t h e m e t h y l e n e d i o x y g r o u p r e s p e c t i v e l y . T w o 3-H s i n g l e t s at d 3 . 4 7 and 3 . 9 2 and a 6-H s i n g l e t at S 3 . 8 0 w e r e d u e to the four m e t h o x y g r o u p s . T h i s w a s c o m p a r e d with 1 , 2 , 3 , 4 - d e h y d r o - /? - p e l t a t i n m e t h y l ether ( 3 8 ) , mp 2 7 3 . 5 ' - ' 2 7 5 . 5 ° C , w h i c h h a d b e e n d e r i v e d from c o m p o u n d ( 3 6 ) by the s a m e r e a c t i o n (ref. 2 5 ) . T h e ^ H - N M R s p e c t r u m of c o m p o u n d ( 3 8 ) s h o w e d t w o 1-H s i n g l e t s at S 7 . 8 5 and 8 . 1 5 and a 2-H
574 singlet at S 6.20 due to the four aromatic protons at the C-5, C-1 and C-2', 6' positions. Two 2-H singlets at ^ 5 . 3 6 and 6.03 were assigned to the methylene protons of the lactone junction and the methylenedioxy group respectively. Two 3-H singlets at d 3.95 and 4.20 and a 6-H singlet at d 3.83 were due to the four methoxy groups. The fact that compounds (37) and (38) are different compounds offers evidence that the original compounds (1 5) and (3 6) are not stereoisomers.
OCH,
H3CO OCH3
15 : Ri=H, R2=OCH3 (3- /? -H) 36 : Ri=0CH3. R2=H (3- /? -H) 39 : Ri=H. R2=OCH3 (3- a -H)
OCH,
H3CO OCH^
37 : Ri=H, R2=OCH3 38 : Ri=0CH3, R2=H 14:Ri=R2=H
r Ar
type B Figure 9
In the ^H-NMR spectrum of compound ( 3 7 ) , the methylenedioxy group and one of the methoxy groups appeared at d 6.06 and 3.47 respectively, suggesting that the methylenedioxy group was at the C-6 and C-7 positions on the basis of the report by Gilchrist and co-workers (ref. 39), while the methoxy group ( d 3.47) was at the C-5 position (ref. 40) since the signal was shifted to much higher magnetic field than that of the C-8 methoxy group of compound (3 8) (S 4.20) by the anisotropic effect of the C-4 phenyl group. As regards the position of the lactone carbonyl group of 2,3-naphthalide lignans, it was reported that in the ^H-NMR spectrum, the lactone methylene protons and the C-1 proton should appear at S 5.32-^5.52 and d 7 . 6 ~ 7 . 7 respectively in type A, but at d 5 . 0 8 - 5 . 2 3 and d 8.25 in type B (ref. 41), In the ^HNMR spectrum of compound ( 3 7 ) , they appeared at d 5.30 and d 7.64, respectively, showing that compound (37) belongs to type A. On the basis of the above spectroscopic studies, compound (1 5) was presumed to be 2-hydroxymethyl-5-methoxy-6,7-methylenedioxy-4-(3' ,4' , 5 ' -trimethoxyphenyl)-l,2,3,4-tetrahydronaphthoic acid lactone. It was reported by Klyne and Sakakibara (ref. 42) that optical rotatory disper-
575
sion (ORD) and circular dichroism can be applied to determine the absolute configuration of the C-4 aryl group in 4-aryltetralin-type lignans. They clarified that all 4- a -aryltetralin-type lignans gave a positive first Cotton effect at 280 — 290 nm in their ORD and CD spectrum, whereas 4-/?-aryltetralin-type lignans gave a negative Cotton effect. As shown in Chart 9, the authors measured the ORD and CD spectra
CD curves
Chart 9
ORD and CD Spectra of 4-Aryltetralin-type lignans compound (1 5)
desoxypodophyllotoxin (3 )
j5-peltatin A methyl ether (3 6)
retroresinolide
of compound ( 1 5 ) , desoxypodophyllotoxin (3) and /?-peltatin A methyl ether (36) and compared them with the reported data for retroresinolide (C-4- /? -aryl configuration) (ref. 42b). Although the effect was not marked in the ORD spectrum, a clear positive effect was observed in the CD spectrum suggesting that compound (1 5) had C-4- a -aryl configuration. It is well known that the 2,3-trans
configuration of
these lignans is converted to the cis type by base treatment (ref. 5, 9, 37c). On heating with sodium methoxide, compound (15) was readily converted to an isomer named picrohernandin ( 3 9 ) , mp 9 0 ~ 9 2 ° C , [a]^+100°
. The fact that the optical
rotation changed to dextrorotatory suggests that the configuration at the C-2 and C-3 positions of compound (15) is trans. In the IR spectrum, the absorption band centered at 1770 cm"^ was assigned to the C = 0 stretching vibration. The ^H-NMR spectrum showed a 1-H singlet at d 6.40 and a 2-H singlet at d 6.34 due to the three aromatic protons and two 3-H singlets at d 3.90 and 3.80 and a 6-H singlet at d 3.76 due to the four methoxy groups. A 2-H singlet at d 5.92 was due to the
576 methylene protons of the methylenedioxy group. Two 2-H multiplets centered at ^ 2.62 and 4.25 were assigned to the methylene protons of the C-1 benzylic position and the lactone junction respectively. Two 1-H multiplets centered at d 3.10 and 3.46 were attributed to the C-2 and C-3 methine protons in the lactone ring respectively. On the other hand, the crystal structures of 5' -demethoxy- /? -peltatin A methyl ether and 2'-bromopodophyllotoxin were clarified by X-ray analyses (ref. 43) and it was established that the configuration of these lignans were 2- a -H, 3- /? -H and 4- a -aryl. In view of the above experimental results and the conclusions obtained from X-ray analyses of analogous lignans, the structure of hernandin is presumed to be 5-methoxydesoxypodophyllotoxin ( 1 5 ) . To confirm this, the absolute configuration of compound (15) was determined by X-ray analysis. Colorless needle crystals of compound (15) were grown by slow evaporation of an ethanol solution at room temperature. Oscillation and Weissenberg photographs showed the space group to be P2i2i2j. The density was measured by floatation in a mixture of water and saturated aqueous potassium iodide. A crystal, 0.4 x 0.6 x 0.2 mm in size, was mounted on Rigaku automated four-circle diffractometer. Graphite-monochromated CuX^ radiation was used. The unicell dimensions were determined by least-squares calculation with 2 6 values of 25 high-angle reflections. The atomic numbering of compound (15) is shown in Figure 10, and the crystallographic data are summarized in Table 4.
Table 4 Chemical formula Molecular weight Crystal system Space group Cell constant
a/A b/A c/A Volume/A "^ Z 3
D m/gcm D x/gcm' fx{Cu-K^)/cm^ F(OOO)
Crystal Data C23H24O8
428 Orthorhombic P2i2i2i 8.519(2) 13.125(3) 18.497(4) 2068.2 (9) 4 1.371(1) 1.376 8.84 904
577
Intensity data were collected on the diffractometer by using the w —16 scanning mode and a scan rate of 4 ' /min. Stationary background counts of 5 s were taken at both limits of each scan. Four reference reflections were monitored periodically and showed no significant intensity deterioration. Corrections were made for Lorentz and polarization factors, but not for absorption effects. A total of 2024 unique reflections, of which 24 had no net intensities, were measured to the limit 2/9=130° . The structure was solved by the direct method using the MULTAN program (ref. 44). An E-map, calculated by using 310 reflections ( I E l ^ 1 . 3 9 ) with the phase set of the highest combined figure of merit (2.788), revealed the locations of all
C(15)
C(13)
C(2)
Figure 10 Atomic Numbering of compound (15)
the nonhydrogen atoms. The structure was refined by the block-diagonal leastsquares method with anisotropic temperature factors. All the hydrogen atoms, found on a difference Fourier map, were included with anisotropic thermal factors. The quantity minimized was 2 w ( I F^ I - I F^ I ) . In the last refinements, the following weighting scheme was used: a; =0.47 for F^=0.0, a; = 1.0 for 0 < F ^ ^ 12.0, and a; = 1.0/[1.0+0.370(F„-12.0)] for F „ > 1 2 . 0 . The final R value was 0.044. The final positional parameters with their estimated standard deviations are listed in Table 5. The atomic scattering factors for all atoms were taken from the International Tables for X-ray Crystallography (ref. 45). All numerical calculations were carried out at the Crystallographic Reseach Center, Institute for Protein Reseach, and the Computing Center of Osaka University using the UNICS programs (ref. 46). As a result, the presumed structure of compound (1 5) based on the chemical studies was confirmed by the X-ray crystallographic method. An ORTEP drawing of the molecule of compound (15) is shown in Figure 11.
578 Table 5 The Final Atomic Coordinates (xl04 for C, O Atoms, xI03 for H Atoms) with Estimated Standard Deviations in Parentheses SSSSBSSSBSaSBBSS
Atom
X
y
z
Atom
X
0(1) C(2) 0(3) C(4) C(5) C(6) C(7) C(8) C(9) 0(10) C(ll) C(12) C(13) C(14) C(15) C(16) 0(17) C(18) 0(19) C(20) C(21) C(22) C(23) C(24) C(25) 0(26) C(27) 0(28)
10046(3) 9496(5) 7818(3) 7425(4) 5981(3) 5861(3) 4253(3) 4251(4) 2852(4) 3361(3) 5058(5) 5607(4) 7122(4) 7200(3) 8676(4) 8744(4) 4636(3) 4474(5) 1487(3) 3853(3) 4969(3) 4531(4) 3015(4) 1940(4) 2347(3) 5523(3) 7032(5) 2629(3)
4168(2) 3175(3) 3200(2) 4083(2) 4408(2) 5371(2) 5666(2) 6828(2) 7358(2) 8216(2) 8335(3) 7238(2) 7064(2) 5980(2) 5621(2) 4682(3) 3858(2) 2936(3) 7136(2) 5099(2) 4941(2) 4422(2) 4056(2) 4187(2) 4710(2) 4199(2) 4580(4) 3526(2)
2690(1) 2474(3) 2478(1) 2836(2) 3055(1) 3397(1) 3687(1) 3756(1) 4087(2) 4415(1) 4311(2) 4185(2) 3777(2) 3474(4) 3251(2) 2951(2) 2923(1) 3328(2) 4071(1) 4388(1) 4927(1) 5554(2) 5644(1) 5091(2) 4466(2) 6116(1) 6117(2) 6257(1)
C(29) 0(30) C(31) H(2a) H(2b) H(7) H(8) H(lla) H(llb) H(12) H(13a) H(13b) H(15) H(18a) H(18b) H(18c) H(21) H(25) H(27a) H(27b) H(28c) H(29a) H(29b) H(29c) H(31a) H(31b) H(31c)
2243(6) 488(3) -429(5) 995(6) 994(6) 339(4) 435(5) 524(4) 549(5) 568(5) 803(5) 723(5) 960(4) 358(6) 446(6) 518(7) 603(4) 154(4) 754(5) 686(7) 757(5) 209(5) 133(5) 297(6) -136(5) 16(6) -68(7)
Figure 11
y 4172(3) 3772(2) 3506(4) 259(3) 308(4) 548(3) 707(3) 881(3) 872(3) 683(3) 723(3) 757(3) 602(3) 256(3) 306(3) 247(4) 526(3) 484(3) 431(4) 558(4) 438(4) 378(3) 461(4) 468(4) 306(3) 297(4) 417(4)
ORTEP Drawing of Compound (1 5)
z 6847(2) 5215(1) 4623(2) 283(2) 193(3) 333(2) 320(2) 387(2) 473(2) 472(2) 413(2) 333(2) 330(2) 321(2) 387(2) 328(3) 486(2) 412(2) 660(2) 606(3) 572(2) 734(2) 673(2) 700(3) 476(2) 426(2) 439(3)
579 6.2 Synthesis of Hernandin (15) (ref. 34) Hernandin (1 5) is a new compound which belongs to the category of phenyltetralin-type lignans. Many studies have been reported on the syntheses of this type lignans (ref. 2, 47), and some of them were carried out in connection with the syntheses of steganacin (ref. 48) and deoxyschizandrin (ref. 49). A route via an itaconic acid derivative, obtained by Stobbe condensation of a benzophenone derivative with diethyl succinate, followed by cyclization (ref. 41a, 50) was first undertaken through the scheme outlined in Chart 10.
.--• 1 5 H3CO" ^
^OCHa OCH3
41 : Ri, R2=OCH20, R^^OCE^ 42 : Ri=OCH3,R2, R3=OCH20
Chart 10
However, various attempts to prepare compound (41) by condensation of compound (40) with 3,4,5-trimethoxybenzoic acid gave poor results, although a method employing Nafion-H (ref. 51) and polyphosphoric acid ester (ref. 52) afforded compound (4 1) in low yield (below 30 %) accompanied with a by-product ( 4 2 ) . Compound (42) was identified by comparison with an authentic sample, which was synthesized by the condensation of 2-methoxy-3,4-methylenedioxybenzaldehyde and 3,4,5-trimethoxy-l-bromobenzene followed by oxidation of the resulting benzhydrol. Moreover, the condensation of compound (4 1 ) with diethyl succinate to give compound (4 3) in the next step was unsuccessful. Another method for the syntheses of phenyltetralin-type lignans, a route involving the condensation of two phenylpropanoid-type compounds followed by an intramolecular Diels-Alder reaction (ref. 2a), was found to be applicable to the synthesis of hernandin ( 1 5 ) . 3-Methoxy-4,5-methylenedioxycinnamyl alcohol (46) (ref. 16, 53) and 3,4,5trimethoxyphenylpropiolic acid (49) (ref. 54) were prepared through the scheme shown in Chart 11. 3-Methoxy-4,5-methylenedioxybenzaldehyde (44) (ref. 55) obtained from vanillin via 5-iodovanillin and 5-hydroxyvanillin was converted to the corresponding ethyl cinnamate (4 5) by means of the Horner-Emmons reaction
580 COOEt
CHoOH
COOEt
CHO
COOEt
B r ^ l
J-
'Br
COOH
ill
OCH3
OCH3
OCH3
47
48
49
i) (EtO)2P(0)CH2COOEt; ii)LAH; Hi) Br2; iv) KOH.
Chart 11
(ref. 56) using triethyl phosphonoacetate in the presence of anhydrous K2CO3 in 90% yield. Compound (4 5) was subsequently reduced by lithium aluminium hydride (LAH) at -15*C to afford compound (4 6) in 80% yield. On the other hand, compound (49) was obtained from 3,4,5-trimethoxybenzaldehyde via ethyl 3,4,5trimethoxycinnamate (4 7) and the corresponding dibromo compound (4 8) followed by debromination. The process of bromination of compound ( 4 7 ) and subsequent debromination were achieved by Klemm's method (ref. 54). Condensation of compound (46) and (49) was achieved as shown in Chart 12 by means of iV,N-dicyclohexylcarbodiimide (DCC) in the presence of p-toluenesulfonic acid (p-TSA) in pyridine solution (ref. 57), affording compound ( 5 0 ) , mp HO^C, in 80% yield. In the IR spectrum, the absorption bands centered at 2200 and 1700 cm"^ were assigned to the triple bond and the C = 0 stretching vibration. The ^H-NMR spectrum showed two 1-H singlets at 8 6.57 and 6.64 and a 2-H singlet at d 6.84 due to the four aromatic protons. A 6-H singlet at d 3.86 and two 3-H singlets at 8 3.89 and 3.91 were assigned to the four methoxy groups. A 2-H doublet at d 4.86 (/= 6.9 Hz) was assigned to the allylic methylene protons. A 1-H doublet of triplets resonating at ^ 6.18 (/=16, 6.9 Hz) and a 1-H doublet resonating at ^ 6.61 (/=16 Hz) were attributed to the olefinic protons. A 2-H singlet at ^ 5.98 was due to the methylene protons of the methylenedioxy group. Subsequent intramolecular DielsAlder reaction of compound (5 0) was carried out in dimethylformamide at elevated
581
DCC
HaCO
///
H3CO
51 or/and 52 or/and 53
OCH3
50
H3CO
^
"OCH3 OCH3
51 : Ri, R2=OCH20, R3=OCH3 52 : Ri=0CH3. R2. R3=OCH20
53
Chart 12
temperature (ref. 58). In this reaction, the formation of three kinds of products ( 5 1 ) , ( 5 2 ) and ( 5 3 ) is presumed. As a result, the reaction product was obtained as an amorphous powder. However, in the ^H-NMR spectrum of this crude product, the methylenedioxy group signals appeared at two different positions, i.e., at 8 5.96 (/=14.7 Hz) and at 8 5.69 (J=25.5 Hz), suggesting that this product was a mixture of two compounds. The ratio of each component was found to be 1.8:1 from the integration intensity of the methylenedioxy group signals. The mixture was subjected to silica gel column chromatography (hexane:EtOAc:CHCl3 = 2:1:1) to afford two compounds, compound (51 ) as the major product and compound (5 2 ) . Compounds (51 ) and (5 2) showed the same molecular formula of C23H22O8 and the same molecular ion at m/z 426.1313 in the mass spectrum. In the ^H-NMR spectrum, most of the signals of the two compounds were analogous. The signals of the four methoxy groups appeared as three singlet peaks, showing the existence of two equivalent methoxy groups. No vinyl proton was observed. These facts rule out the existence of compound ( 5 3 ) . A methoxy group in compound (51) was observed at a higher magnetic field ( d 3.35) than that in compound (5 2) {8 3.95) due to the anisotropy of the trimethoxyphenyl group. Although it has been reported that the catalytic hydrogenation of analogous 1,2-dihydronaphthalene lactones usually gave the phenyltetralin lactones with all-cis configurations (ref. 9, 40), we examined the
582 direct hydrogenation of compound (5 1) expecting the formation of compound(1 5 ) , because no example was known of compounds which have any functional group at the C-5 position. Reduction of compound (51) was carried out with 10% palladium carbon (Pd-C) in acetic acid for 4 h under atomospheric pressure. The reaction mixture was chromatographed on a silica gel column and an amorphous product (5 6) was obtained as the sole product in low yield with recovery of the starting material and the expected compound (15) was not formed. Compound (5 6) was presumed to be isopicrohernandin, having 2,3-ciS and 3,4-cis configurations, by comparison of the ^H-NMR and IR spectra with those of compound (15) and picrohernandin ( 3 9 ) . The mass spectrum of compound (56) displayed the molecular ion at m/z 428 in agreement with the formula C23H24O8. In the IR spectrum, the absorption band was seen at 1765 cm'^ due to the C = 0 stretching vibration. The ^HNMR spectrum showed a 2-H singlet at S 6.35 and a 1-H singlet at d 6.49 due to the three aromatic protons and three singlets at d 3,73, 3.80 and 3.90 corresponding to the four methoxy groups. A 2-H multiplet centered at d 2.93 and two 1-H multiplets centered at d 2.70 and 3.10 were assigned to four protons at the C-1, C-2 and C-3 positions. Two 1-H triplets at S 3.28 and 4.37 were attributed to the methylene protons of the lactone junction, while two 1-H doublets at d 5.95 and 5.97 were due to the methylene protons of the methylenedioxy group. A 1-H doublet at d 5.10 was assigned to the methine proton at the C-4 position. The unsuccessful attempt at direct hydrogenation of compound (51 ) meant that a new strategy was required. In the hydrogenation of olefins over heterogeneous catalysts, the stereochemistry of reduction has been found to be influenced not only by the bulk of neighboring functional groups but also by the attractive interactions between some of these groups and the catalyst surface. Thompson reported that the stereochemistry was controlled by the neighboring functional group in the catalytic
3
R
O ^
^
R
J
O ^
S
R R
R=CH20H
95%
5%
R=C00CH3
15%
83%
R=C0NH2
10%
90%
Chart 13
J
n^ O
583 hydrogenation of tetrahydrofluorenes, and in the case of the hydroxymethyl group, absorption of the compound on the catalyst surface occurred from the same side as the hydroxy group, as shown in Chart 13 (ref. 59). Considering this information, another method was worked out as described in Chart 14. The lactone ring of compound (51) was cleaved by potassium hydroxide and careful neutralization gave the unsaturated hydroxy acid (5 4 ) . Crude compound (5 4) was hydrogenated on Pd-C in EtOH and the reaction mixture was subsequently acidified with concentrated hydrochloric acid. After usual worked-up, the reaction mixture showed the existence of two products on thin layer chromatography.
15
OCH3
Ai^ -
i) 7% KOH; ii) neuu-alized with 2% HCl; iii) 10% Pd-C, H2 4.5 atom, 45°C; iv) cone. HCl; v) DCC
Chart 14
Each compound was isolated by preparative thin layer chromatography. One coincided with compound (5 6) obtained by the direct hydrogenation of compound (51) judging from the ^H-NMR spectra. Another compound (5 7) showed the molecular ion at m/z 446.1575 in agreement with the formula C23H26O9 in its mass spectrum. In the IR spectrum, the absorption band centered at 3480 c m ' showed the existence of a hydroxy group, while a carboxyl group was indicated by the band at 1730 c m \ The fact that the remaining compound (5 7) unlactonized in concentrated hydrochloric acid suggested that this hydroxy acid had a 2,3-trans
configuration. Its ' H -
NMR spectrum (CDCI3) showed a 2-H singlet at d 6.25 and a 1-H singlet at d 6.42
584 due to the three aromatic protons. A 6-H singlet at d 3.74 and two 3-H singlets at d 3.57 and 3.79 were assigned to the four methoxy groups. A 2-H triplet at S 3.69 was attributed to the methylene protons adjacent to the hydroxy group, while three multiplets centered at S 2.40, 2.75 and 3.00 were assigned to the four protons at the C-3, C-2 and C-1 positions. A 1-H doublet at ^ 4 . 7 2 and two 1-H doublets at d 5.88 and 5.89 were due to the C-4 methine proton and the methylenedioxy group respectively. It seems reasonable that the 2,3-Cis hydroxy acid (5 5) would be more easily lactonized than the corresponding trans isomer (5 7) on account of the proximity of the hydroxy and the carboxyl groups. Lactonization of compound (57) was achieved by means of DCC in chloroform solution (ref. 60) affording a crystalline compound, mp 215-^218*C All spectral data were identical with those of natural hernandin. In conclusion, ( ± ) - h e r n a n d i n (1 5) with 2y3-trans and 3,4-cis configuration was synthesized. 6.3 Cyclization Reaction of 5'-Methoxypodorhizol (13) (ref. 21) Bisbenzyl- / -butyrolactone can be converted to a phenyltetralin-type lignan. Therefore it was expected that hernandin (1 5) would be derived from compound (13) by cyclization. According to the Stevenson's method (ref. 61), compound (1 3) was treated with excess trifluoroacetic acid in dichloromethane for 12 h at room temperature. As a result, two compounds (5 8) and (5 9 ) , which had melting point of 1 8 0 ^ 1 8 2 * 0 and 204~206*C respectively, were isolated by means of
H3CO. 13
CF3COOH CH2CI2
H3CO
OCH,
OCH,
Chart 15
preparative thin layer chromatography using a mixture of benzene, chloroform and ethyl acetate (2:1:1). The mass spectra of both compounds displayed the molecular ion at m/z 428 in agreement with the formula C23H24O8, and in the IR spectrum, the absorption band at 1780 cm'^ was assigned to the C = 0 stretching vibration. The CD spectra showed a positive Cotton effect at 272 and 284 nm in compound ( 5 8 ) ,
585
Chart 16 CD Curves of 4-AryltetraIin Lignan in EtOH hernandin ( 1 5 ) ;
compound ( 5 8 ) ;
compound (59)
Table 6 Comparison of Chemical Shift ((^ ) of Compound (58) and (59)
0
arom. H(3H)
C4-H
C1.2.3-H
H2 CH3O X 4
H2C
12H
0 0
" C g " ^ C2',6'
58
59
4.30
6.40
6.47
IH
2H
(s)
(s)
(d)
6.36
6.51
4.24
IH
2H
(s)
(s)
C3\4',5'
C7
5.90 IH (d)
3.92 IH (dd)
3.30
3.81
3H
9H
(m)
5.91 IH (d)
4.45 IH (dd)
(s)
(s)
3.00-2.35
5.66 IH (d)
3.95 IH (dd)
3.83 3.80 3.91 3H
6H
3H
5.77 IH (d)
4.49 IH (dd)
(s)
(s)
(s)
3.30-2.30 4H
4H (d)
C5
(m)
586 and at 270 and 286 nm in compound (5 9 ) , in contrast to that of natural hernandin (1 5) as shown in Chart 16. In the ^H-NMR spectrum of compound (58) as shown in Table 6, a methylenedioxy group appeared at d 5.90 and 5 . 9 1 , similar to hernandin (15) (d 5.92, 5.94). However, the methoxy group in the tetralin ring which was observed at d 3.63 in compound (1 5), resonated at a slightly higher magnetic field ( d 3.30), and a proton assigned to the C-4 position appeared at a higher magnetic field and had a larger / v a l u e ( ^ 4 . 3 0 , / = 9 . 4 Hz) than that of compound (1 5) ( ^ 4 . 8 7 , / = 4 . 8 Hz). In the ^H-NMR spectrum of compound (5 9 ) , the C-4 proton appeared at a higher magnetic field than that of compound (1 5 ) , as in compound ( 5 8 ) with a large / value ((^ 4.24, / = 1 0 . 3 Hz). However, the environment of the methylenedioxy and the methoxy groups in the tetralin ring is different from that in compound (1 5) or compound (58). Thus, the signal of the methylenedioxy group ( d 5.66, 5.77) appeared at a higher magnetic field than those of compound ( 1 5 ) and compound ( 5 8 ) , and by contrast, the signal of the methoxy group ( S 3.91) was at a lower magnetic field than those of compound (15) and compound ( 5 8 ) . The above results suggest that the C-4 configuration in both compound (5 8) and (59) is 4 /? aryl. In conclusion, compound (5 8) is isohernandin, 2-hydroxymethyl-5-methoxy6,7-methylenedioxy-4-(3' ,4' , 5 ' -trimethoxyphenyl)-l,2,3,4-tetrahydronaphthoic acid lactone, and compound (59) is 2-hydroxymethyl-5,6-methylenedioxy-7-methoxy-4-(3',4',5'-trimethoxyphenyl)-l,2,3,4-tetrahydronaphthoic acid lactone, a positional isomer of compound (15) in relation to the methylenedioxy and the methoxy groups. The measurement of specific rotation was unable due to the small quantities. 7
Total Synthesis of 4-Aryltetralin-Type Lignan (ref. 62) In section 6,2, the synthesis of (±)-hernandin which has 2,3-trans,
3,4-Cis
configuration, was described. This method was adopted for the syntheses of analogous 4-phenyltetralin-type lignans and we succeeded in synthesizing ( ± ) - d e soxypodophyllotoxin (3) and ( ± ) - /? -peltatin A methyl ether ( 3 6 ) . The syntheses were carried out through the scheme in Chart 17. The starting materials, tra/]s-3,4-methylenedioxycinnamyl alcohol (61a) and tra/]S-2-methoxy-3,4-methylenedioxycinnamyl alcohol (61 b) were prepared from the corresponding benzaldehyde via substituted ethyl cinnamate by means of the Horner-Emmons reaction and lithium aluminium hydride reduction. Condensation of compound (61 a) or (61 b) with compound (47) gave compound ( 6 2 a ) or (62b), respectively, followed by ring closure to afford compound (63a) or (6 3 b ) . Intramolecular Diels-Alder reaction of compound (6 2) led to the formation of the aromatized compounds as by-product in both cases. Moreover, in the reaction of
587
61a : R=H 61b : R=OCH3
Ar 64
COOH
U-
3 : R=H 36 : R=0CH3
i) DCC; ii) DMF, A; iii) 7% KOH; iv 10% Pd-C, H2; v) p-TsOH, benzene
Chart 17
compound (62a) to compound (63a), an isomer, 5,6-methylenedioxy-l,2-dihydronaphthalene lactone was also detected among the products. The existence of these by-products was determined by analysis of the ^H-NMR spectra. The product ratio was as shown in Table 7. Compound (63) was successively hydrolyzed to give the corresponding unsaturated hydroxy acid ( 6 4 ) . The next step was slightly modified from that used in the synthesis of hernandin (1 5), in which the experiment was done in two steps after the hydrogenation of the 1,2-dihydrohydroxy acid to distinguish the objective hydroxy acid {2,3-trans
type) from another possible lactone
(2,3-ci5 type). To improve the yield of the desired 2,3-trans
lactone, the hydro-
genated products were directly treated with p-toluenesulfonic acid without isolating
588 the hydroxy acid and the resultant two sorts of lactones were isolated by preparative thin layer chromatography. In the attempt of the synthesis of compound (3), compound (3) and isodesoxypicropodophyllin (65a) were obtained in 35 and 26% yields, respectively, while in the synthesis of compound (36), compound (36) and
Table 7
Product Ratio of Diels-Alder Reaction
Ar 62a : R=H
5.9
62b : R=0CH3
9
: :
O
1
2.4
1
iso- /? -peltatin B methyl ether (6 5 b ) were obtained in 48and 27% yields, respectively. Recently Achiwa et al. (ref. 63) reported the synthesis of (-)-desoxypodophyllotoxin (3) from the corresponding l , 2 - d i h y d r o - ( - ) - 6 4 a , which was derived from (-i-)-podorhizon utilizing our method followed by cyclization. These results showed that this method was advantageous for the synthesis of 4-phenyltetralintype lignans which have the same configuration with the natural products. 8 Some Reactions of Lignans and Related Compounds 8.1 Cleavage Reaction of Methylenedioxy Group with Lead Tetraacetate (ref. 25) For achieving the aromatization reaction of tetralin-type lignans, the methods using palladium black (ref. 64) for picropodophyllin and lead tetraacetate for 3,4dehydrodesoxypodophyllotoxin (ref. 65) have been known. However when desoxypodophyllotoxin (3) was reacted with lead tetraacetate in acetic acid in order to obtain the aromatized product, the desired naphthalene-type lignan could be not obtained. Instead of the objective compound, compound ( 6 6 ) , which showed mp 2 3 8 ^ 2 4 0 ° C and an optical rotation of [ a ]D -116° (c=0.5 in EtOH), was obtained in 54% yield. The elemental analysis agreed with the formula C21H22O7. In the IR spectrum (KBr), the absorption bands at 3420 and 1770 c m ' were assigned to the OH and C = 0 stretching vibrations respectively. The ' H - N M R spectrum (DMSO-dJ showed a 2-H singlet at d 6.30 and two 1-H singlets at d 6.34 and 6.59 due to the four aromatic protons. A 3-H singlet at d 3.61 and a 6-H singlet at d 3.63 were
589 assigned to the three methoxy groups. A 1-H triplet at d 3.93 was attributed to one of the methylene protons of the lactone junction, and a 2-H multipiet centered at d 4.40 was assigned to the C-4 methine proton and another proton of the lactone methylene protons. Two 2-H multiplets centered at S 2.65 and 2.88 were due to four protons at the C-1, C-2 and C-3 positions. Two 1-H singlets at S 8.74 and 8.80 were ascribed to the hydroxy protons which disappeared on addition of DjO. On the basis of the above spectroscopic data, it was clear that compound (6 6) was not aromatized compound (1 4), and it was confirmed to be the product of cleavage of the methylenedioxy group. Although the cleavage reaction of the methylenedioxy group could be readily carried out by orthophosphoric acid or boron trichloride (ref. 66), the above experimental fact was very interesting. In addition, the cleavage reaction was successfully applied to desoxypicropodophyllin (4), which is the isomer of compound (3), and /5 -peltatin A methyl ether (36) under the same conditions to afford the corresponding products (6 7) and (6 8) in 4 5 % and 8% yields, respectively. While our experiments were under progress, Taguchi and co-workers reported the same reaction in gomisin (69) isolated from cbinensis
Shizandra
B^J^L. in benzene to afford the corresponding diphenolic compound (7 0 ) .
OCHo
H3CO
H3CO OCH
OCH3
3 : R=H (2,3-trans) 4 : R=H (2,3-cy5) 36 : R=OCH3 {23-trans)
"xx>
66 : R=H (23-trans) 67 : R=H (2,3-cis) 68 : R=OCH3 (23-trans)
71 72 73 74
OCH3
69 : Ri, R2=OCH20 70 : Ri=R2=OH
: Ri=C00CH3, R2=CH2 : Ri=C00CH3, R2=CHOCOCH3 : Ri=CH2CH2CH3, R2=CH2 : Ri=CH2CH2CH3, R2=CHOCOCH3
Figure 12
They presumed that the mechanism proceeded via the acetoxy form as an intermediate, since they isolated the acetoxy compound (7 2) in the case of the reaction with lead tetraacetate and methyl piperonylate (71 ) (ref. 67). The reaction with
590 3,4-methyienedioxypropylbenzene (73) in acetic acid afforded 3,4-dihydroxypropylbenzene in 14% yield, and in benzene, in 40% yield without the need of silica gel column chromatography. These facts support the foregoing experiment results of Taguchi et aL. It seems that the acetoxy compound (7 4) is very unstable and it is easily converted to the dihyroxy form by hydrolysis. 8.2 Reaction with 2,3-Dichloro-5,6-dicyanoquinone (DDQ) (ref. 25) The aromatization reaction involving conversion of tetralin-type lignans to naphthaline-type lignans was unsuccessful, as described previously. Therefore it carried out using chloranil or DDQ for desoxypodophyllotoxin (3), its isomer (4)
75
6^ 76 : Ri=R2=OCH3, R3, R4=OCH20 77 : Ri, R2=OCH20, R3=R4=OCH3
Figure 13
and /? -peltatin A methyl ether (3 6 ) . Although the starting materials were recovered in the case of the reaction with chloranil, however when DDQ was used with compound (3) in benzene under reflux, the reaction proceeded smoothly to afford the corresponding aromatized products (14) in 84% yield, and in reactions with desoxypicropodophyllin (4) and /?-peltatin A methyl ether ( 3 6 ) , in 64% and 5 1 % yields respectively. On the other hand, Ghosal et al. reacted suchilactone (7 5 ) , an arylidenebutyrolactone lignan, with 2 moles DDQ in benzene under reflux to afford retrochinensin (7 6 ) , via a redox reaction (ref. 68). We carried out the same reaction for bursehernin (7) with 3.3 moles DDQ to give the corresponding aromatized compound (77) in 50% yield via ring closure. Compound (7 7) was spectroscopically identical to chinensin (ref. 69) reported by Dr. Ghosal. The above studies demonstrated one step conversion from dibenzyl-7 -butyrolactone lignans to naphthalene-type lignans via ring closure reaction.
591 9 Synthesis of Related 4-Aryltetralin-Type Lignan from Desoxypodophyllotoxin 9.1 Syntheses of /? -Peltatin A and B Methyl Ether from Desoxypodophyllotoxin (ref. 70) We were interested in the syntheses of related 4-aryltetralin-type lignans by utilizing desoxypodophyllotoxin (3), which is available in large quantities from the plant seeds. The synthetic route to /? -peltatin A methyl ether (36) and B methyl ether (83) (ref. 37) from desoxypodophyllotoxin (3) is shown in Chart 18.
8
1 H HO.
H3CO
^Y
OCH3
OCH3
OCH^
3:3-/?-H 4 : 3- a -H
HO-'X:!:^^^^
78
OCHo 79
_
H3C0'^ ^
OCH3
OCH3 OCH3
OCH3
81 : 3 - / ? - H
66
82 :3-a-H
Chart 18
36: 3-/3-H 83:3-a-H
Koford et al. obtained a 2'-bromo-compound by direct bromination of podophyllotoxin (1 1) (ref. 71). This suggests that the introduction of bromine at the C-5 or C-8 positions of ring A is difficult by direct bromination of desoxypodophyllotoxin (3). To confirm this, we carried out the direct bromination of desoxypodophyllotoxin (3) with N-bromosuccinimide in dimethylformamide and obtained the monobromoderivative (78) in 83% yield. The mass spectrum of this compound ( 7 8 ) ,
592 which showed mp 175~177"C and [ a ]^ -106' (c=1.0, in CHCI3), displayed the molecular ion at m/z 478 in agreement with the formula C22H2iBr07. The ^H-NMR spectrum showed three 1-H singlets at ^ 6.13, 6.36 and 6.60 due to the three aromatic protons, while the three methoxy groups afforded three 3-H singlet peaks at d 3.63, 3.84 and 3.90. This showed that bromine had been introduced in ring C. The cleavage of the methylenedioxy group of compound (7 8) with boron trichloride gave the corresponding dihydroxy compound (7 9), which afforded the molecular formula CjiHj^BrO^, mp 2 2 4 - 2 2 6 t and [ a ]^ -129' (c=0.5 in EtOH). The IR spectrum included bands centered at 3360 (OH) and 1760 (C=0) cm"\ In the ^HNMR spectrum of compound (79), the three aromatic protons were observed at d 6.23 and 6.60 as two singlets, while the three methoxy groups of ring C appeared independently as three 3-H singlets at d 3.60, 3.75 and 3.80. In the usual 4-aryltetralin lignans, such as desoxypodophyllotoxin (3), hernandin (15), ^-peltatin A methyl ether (36) and /?-peltatin B methyl ether (83), which have no functional groups except for the three methoxy groups on ring C, the signals of the C-3' and C-5' methoxy groups resonate as a 6-H singlet. These results lead to the presumption that bromine in compound (78) was at the C-2' position. In order to introduce bromine into ring A, compound (6 6) was used as the starting material in the expectation that an increment of the electron density at the C-5 or C-8 positions would be produced by the phenolic hydroxy groups. Bromination of compound (6 6) by iV-bromosuccinimide in DMF afforded a monobromo compound (8 0), which showed mp 2 2 0 - 2 2 2 t and [ a J^ -131' (c=0.5, in MeOH), in 80% yield. In the ^H-NMR spectrum of compound (80), the C-3' and C-5' methoxy signals appeared at S 3.68 as a 6-H singlet whereas the two aromatic protons at the C-2' and C-6' positions appeared as a 2-H singlet, suggesting that the bromine had been introduced at the C-5 or C-8 positions of ring A. However, it was difficult to determine whether the bromine was at C-5 or at C-8 in its ^H-NMR spectrum. X-Ray analysis could not be carried out because an attempt to get a suitable single crystal of compound (80) was unsuccessful. Temporarily setting aside the problem of the position of the bromine, we turned our attention to the methylenation of compound (8 0) to prepare the corresponding derivative (81) by Miller's method (ref. 72) which consists of reaction with cesium fluoride (CsF) or potassium fluoride (KF) and methylene halide in DMF. In the preliminary experiment, the reaction of compound (66) with CsF and methylene chloride in DMF afforded desoxypicropodophyllin (4) in 53% yield together with a trace of desoxypodophyllotoxin (3). This suggested that the trans (2- a -H, 3-^-H) configuration of the original compound (6 6) was converted to the cis (2a -H, 3- a -H) configuration. For the purpose of finding conditions under which the original 2,3-trans configuration could be retained, various methods were
593 examined for the methylenation of compound (66). It was found that compound (6 6) was preferentially obtained in 4 3 % yield by using methylene iodide in place of methylene chloride. On the basis of these preliminary experiments, the methylenation of compound (80) was examined under various conditions and two compounds ( 8 1 ) and ( 8 2 ) , which had the same empirical formula C22H2iBr07, were isolated in each experiment. The ratio of the yields of these compounds however depended markedly on the sort of methylene halide used. The reaction with CsF and methylene iodide or methylene bromide gave predominantly compound (81) in 6 8 ^ 7 0 % yield and compound (81) in 10-^12% yield. In contrast, the use of methylene chloride gave compound (81 ) in 16% yield and compound (82) in 58% yield. The use of KF with methylene iodide or methylene bromide afforded a mixture consisting of the same products. However the total yields of methylenated products were inferior to those obtained with CsF for compounds (81 ) and (8 2).
Table 8 Yield of Methylenation Products from Compound (66) and (80) Starting material
Metal halide (eq mol)
Methylene halide
Reaction conditions Temp.(°C) Time (h)
66 66 66
CsF (5) CsF (5) CsF (5)
CH2CI2 CH2Br2 CH2I2
120 120 120
2 2 2
80 80 80 80 80 80 80 80
CsF (3) CsF (5) CsF (5) CsF (5) CsF (5) CsF (10) KF(5) KF(5)
CH2Br2 CH2Br2 CH2I2 CH2Br2 CH2CI2 CH2Br2 CH2I2 CH2Br2
60 60 120 120 120 60 120 120
3 3 2 2 2 3 2 2
Product yields
(%) 3 Trace 39 43 8 1
51 68 70 64 16 62 52 50
4
53 15 12
82
9 10 12 18 58 12 4 5
The results are summarized in Table 8. Compound (81) showed mp 2 0 7 ^ 2 0 9 ° and an optical rotation of [ c ]p -148" (c= 0.5 in CHCI3), while compound (82) showed mp 176^^ 178° and an optical rotation of [ a l^ -12.5° (c=l in CHCI3). The specific rotations of these two compounds indicated that compound (81) was 8-bromodesoxypodophyllotoxin and compound (8 2) was 8-bromodesoxypicropodophyllin. In order to clarify the position of the bromine and to confirm the absolute configurations at the C-2 and C-3 positions, X-ray diffraction analysis was applied to a single crystal of compound (81) obtained from the ethanol solution by slow evaporation at room temperature.
594 A crystal with dimension of 0.3 x 0.4 x 0.4 mm was used for the X-ray study. The crystal was orthorhombic, space group P 2^2^2^ with cell dimensions of a =20.720 (2), 6=10.322 (2) and c=9.686 (2) A. The observed density [D„=1.521 (2) g • cm"^] showed that there were four molecules in a unit cell [volume=2072 (2) A \ D A = 1 . 5 3 0 g • cm'^]. The intensity data were collected on a Rigaku automated four-circle diffractometer by using graphite-monochromated Cu-K^ radiation ( X = 1.5418 A). By means of the w-2 6 scanning technique, 2034 independent reflections at less than sin ^/A =0.589A"^ were colleted and used for the structure determination. Corrections were made for Lorentz and polarization effects. Although the linear absorption coefficient of this crystal is relatively large [/i (Cu--K'J= 33.56 cm'^], absorption correction was not made because of the small size of the crystal used. The structure was solved by a combination of the heavy atom and direct
Table 9 Atomic Coordinates (xlO ) with Their Estimated Standard Deviations in Parentheses •!' I..I 1
s
•
I." Mil'
Atom
X
Br 0(1) 0(2) 0(3) 0(4) Od') 0(2') 0(3') C(l) C(2) C(3) C(4) C(5) C(6) C(7) C(8) C(9) C(10) C(ll) C(12) C(13) C(l') C(2') C(3') C(4') C(5') C(6') C(7') C(8') C(9')
7414(0) 7750(2) 7118(2) 4162(2) 3679(2) 3316(2) 4045(2) 5233(2) 5918(3) 5245(3) 4862(3) 5078(3) 6097(3) 6743(3) 7119(3) 6870(3) 6197(3) 5818(3) 7747(3) 4812(4) 4163(3) 4820(3) 4188(3) 3929(3) 4288(3) 4920(3) 5181(2) 2895(4) 3608(4) 5909(3)
.1
:aaa
y 5125(1) 2791(5) 1511(5) 6282(5) 4432(6) 3815(6) 5168(5) 5929(5) 5563(6) 5836(6) 4567(6) 3700(6) 2604(6) 2446(6) 3200(6) 4178(6) 4436(5) 3625(6) 1685(9) 6714(7) 5025(8) 4133(6) 3776(6) 4139(6) 4864(6) 5211(6) 4848(6) 3276(11) 6271(9) 6169(8)
z 4189(1) 6183(6) 7599(6) 4595(5) 5232(6) 9736(5) 11530(4) 10894(5) 4635(7) 5115(6) 5135(6) 6321(6) 7042(6) 6959(6) 6122(6) 5388(7) 5456(6) 6279(6) 7037(9) 4255(8) 5028(6) 7727(6) 8069(7) 9313(7) 10251(6) 9907(5) 8645(6) 8683(10) 11515(8) 10717(8)
595 methods. The positional parameters were refined by the block-diagonal leastsquares method with anisotropic temperature factors for nonhydrogen atoms and isotropic ones for hydrogen atoms. The final R value was 0.074. The atomic numbering of compound (81 ) is shown in Figure 14. The final atomic coordinates are listed in Table 9.
Br
|C(8) C(l)
Figure 14 Atomic Numbering of compound (81)
Figure 15
ORTEP Drawing of compound (81 )
All numerical calculations were carried out on an ACOS-900 computer at the Computation Center of Osaka University by using the UNICS program (ref. 46). A stereoscopic view of the molecular conformation is shown in Figure 15. The structure of compound (81) was confirmed to be 8-bromodesoxypodophyllotoxin.
596 The structure of compound (81 ) as 8-bromodesoxypicropodophyllin had not been proved at this stage, but was clarified by the fact that compound (81) was converted to compound (8 2) on treatment with basic reagents, as described later. Subsequently, several methods were examined to convert compound (81) into ^ peltatin A methyl ether (3 6) and B methyl ether (8 3) by direct substitution of the bromine with a hydroxy group and methoxy group. Attempts to convert compound (81) to /? -peltatin A methyl ether (36) or B methyl ether (83) with potassium methoxide in the presence of dicyclohexyl-I8-crown-6-ether (ref. 73) failed and afforded only compound (8 2) in quantitative yield. This result shows that compound (8 2) is a stereoisomer of compound ( 8 1 ) . Introduction of a hydroxy group by means of the Grignard reaction (ref. 74) was also unsuccessful, affording merely the demethylated product since no Grignard reagent is formed even on treatment with active magnesium (ref. 75). The reaction of compound (81) with cuprous iodide and sodium methoxide (ref. 76) furnished two substances, which were isolated by preparative thin layer chromatography. One of them, mp 189-^ 191*0, [ a ]^ +9. l" , was compound (83) in which the 2,3-configuration had been converted to the cis type, although the introduction of a methoxy group had been achieved. The other one was desoxypicropodophyllin (4) produced by debromination of the starting material (81) together with conversion of the 2,3-configuration. Similar reaction of compound (8 2) gave rise to compound (8 3 ) . The structure of compound (8 2) was confirmed as 8-bromodesoxypicropodophyllin. The above results suggest that mild reaction conditions are required to obtain /? -peltatin A or A methyl ether (3 6) from compound (81) without the conversion of the original 2,3-trans
configuration.
Finally, an attempt was made by using Buck and Kobrich's method (ref. 77), which yields phenols by the reaction of lithiated aryl compounds with nitrobenzene. Compound (81 ) was treated with n-butyllithium at -100°C in THF in the presence of tetramethylethylenediamine followed by the addition of nitrobenzene. The crude products, which were difficult to isolate as phenolic compounds, were immediately methylated with diazomethane in ether solution. The products were subjected to preparative thin layer chromatography, affording two substances. One, obtained from the upper layer, was recrystallized from ethanol as colorless needles. It showed mp 1 6 2 . 5 ^ 1 6 3 . 5 1 ) and an optical rotation of [a ]^ -110' (c= 0.14 in CHCI3). The other one, isolated from the lower layer, was recrystallized from ethanol as colorless needles. It showed mp 185^187°C and an optical rotation of [a ID 4-1 r
(c=0.3 in CHCI3). These products were identified as compounds (36)
and (8 3) by direct comparison with all spectral data of the corresponding authentic samples. Thus, syntheses of /?-peltatin A and B methyl ethers were achieved from desoxypodophyllotoxin.
597 9.2 Synthesis of Epipodophyllotoxin (1 2) and Podophyllotoxin (11) (ref. 78) The preparation of 1-halo compounds of podophyllotoxin (11) and analogous lignans has been reported by several workers (ref. 66, 79). They obtained predominantly 1- /? -bromo- or -chloro compounds from 1- a -hydroxy lignans such as podophyllotoxin (ref. 79a,b) and demethylenepodophyllotoxin dimethyl ether (ref. 66). These 1- /? -halo compounds are easily hydrolyzed to afford 1- /? -hydroxy compounds. On the other hand, no report has appeared on the preparation of 1-halo compounds from desoxypodophyllotoxin (3). In the present work, we aimed at developing syntheses of podophyllotoxin (11) and epipodophyllotoxin (11) from desoxypodophyllotoxin (3) via the 1-bromo derivative, which could be obtained by a radical bromination. Treatment of desoxypodophyllotoxin (3) with equiv. mol of N-bromosuccinimide in the presence of benzoyl peroxide in carbon tetrachloride under reflux, followed by silica gel column chromatography, gave two compounds which contained no bromine. One of them showed mp 160-^ 162''C, [a ]^ -69° and the molecular formula C22H220g, while the other one gave mp 273'^275"C, [ c Ip ± 0 ° and the molecular formula CjzHjgO^. These were identified as epipodophyllotoxin (1 2) and dehydrodesoxypodophyllotoxin (1 4) respectively by direct comparison of their ' H - N M R and IR spectra and mixed melting point determination with authentic samples. The use of excess 7V-bromosuccinimide did not improve the yield of epipodophyllotoxin (1 2), but afforded 2' - and 8-bromodehydrodesoxypodophyllotoxin. We presumed that epipodophyllotoxin (1 2) was produced by the hydrolysis of the 1-bromo derivative of desoxypodophyllotoxin (3) during silica gel column chromatography. The same result was reported in the reaction of the benzyl butyrolactone compound with N-bromosuccinimide followed by hydrolysis on silica gel to give the hydroxy compound (ref. 47a). In order to isolate the bromo
Ri.?^
H3CO" ^
OCH3 OCH3
11 :Ri=H, R2=0H 12:R2=H, Ri=OH 84 : Ri=Br. R2=H 85 : Ri=H, R2=Br 86 : Ri,R2=0
Figure 16
H3CO
598 compound, Silic AR type 60A (Mallinckrodt Works) was used in the chromatography of the reaction product to give a colorless solid containing bromine. The mass spectrum of this solid showed the molecular ion at m/z 476 and 478 in agreement with the formula C22H2iBr07. The ^H-NMR spectrum indicated that this substance was a 7:3 mixture of 1- /? -bromodesoxypodophyllotoxin ( 8 4 ) and 1- a -bromodesoxypodophyllotoxin ( 8 5 ) based on the following findings. Peaks due to the aromatic protons were each observed at d 6.90, 7.18 (Cg-H), 6.47, 6.49 (C5-H), 6.27, 6.40 (Cj. 6-H) as singlets while two 1-H doublets resonating at d 5.28 (/= 10.26 Hz) and 5.62 (/=3.42 Hz) were attributed to the C-1 methine proton adjacent to the bromine. The ratio of the 1- /? -bromo compound (8 4) and the I- a -bromo compound (8 5) was determined from the integration intensity of the C-1 methine proton in the ^H-NMR spectrum. The low yield of compound (8 5) can be ascribed to the steric hindrance of the C4- a -phenyl group on Cj- a -hydrogen.
Chart 19
Recrystallization of the above mixture from hexane-ether-benzene afforded a substance which showed mp 151 ~ 154°C and an optical rotation of [ a ]jy +11.4' (c=0.4 in CHCI3). The mass spectrum displayed the molecular ion at m/z 476 and 478 in agreement with the formula CzjHjiBrO,. In the IR spectrum, the absorption band at 1780 cm'^ was assigned to the C = 0 stretching vibration. The ^H-NMR spectrum showed two 1-H singlets at d 6.47 and 6.90 and a 2-H singlet at S 6.27
599 due to the four aromatic protons at the C-5, C-8 and C-2' ,6' positions. A 6-H singlet at 6 3.81 and a 3-H singlet at 3.75 were assigned to the three methoxy groups. A 1-H singlet at d 5.62 was attributed to the C-1 methine proton. The substance was identical with an authentic compound (8 4) prepared from podophyllotoxin ( 1 1 ) (ref. 79a). However, an attempt to isolate compound (8 4) as crystals was unsuccessful. Treatment of the mixture with aqueous acetone or with silica gel in hexane-ethyl acetate (1:1) gave epipodophyllotoxin (12) as the sole product. When the mixture was refluxed in methanol, only epipodophyllotoxin methyl ether (87) was obtained as shown in Chart 19. These results suggest that substitution of secondary bromine at the C-1 position with a hydroxy or methoxy group proceeds by an 5 ^ 1 reaction mechanism, and the nucleophile attacks from the non-hindered /?-side (ref. 66) as shown in Figure 17.
H
.»
V--"
Figure 17
Since it was difficult to obtain podophyllotoxin (1 1) from the 1-bromo compound, we examined another route to podophyllotoxin (1 1) from epipodophyllotoxin (12) Via podophyllotoxone (86) (ref. 80). Although compound (86) was obtained from podophyllotoxin (1 1) in fairly good yield by oxidation with manganese dioxide (ref. 80), this reagent gave unsatisfactory or negative results with isopodophyllotoxin (ref. 81) and epiisopodophyllotoxin (ref. 82). We also had a negative result with epipodophyllotoxin ( 1 2 ) , recovering only the starting material. Another reagent, pyridinium chlorochromate (ref. 83), was found to be effective, affording compound (86) in 76.2% yield from epipodophyllotoxin (1 2).
600 As a reagent for the reduction of these keto-lignans with retention of their lactone functions, zinc borohydride was recommended by Gensler and co-workers (ref. 80). For example, podophyllotoxin (1 1) and DL-epiisopodophyllotoxin were obtained in good yield from the corresponding keto-lignans (ref. 80, 81). However, good results were unobtainable with picropodophyllone and L-isopodophyllotoxone (ref. 80, 82). We tried to reduce compound (8 6) with a variety of reagents and examined the ratio of podophyllotoxin (11) and epipodophyllotoxin ( 1 2 ) in the resulting products. The results are summarized in Table 10.
Table 10 Stereoselective Reduction of Compound (86) with Various Reducing Agents conditions ,
.
Ratio of products '
a)
reducing agent'
Zn(BH4)2 NH3 • BH3 t -BuNH2 • BH3 [(CH3)2CH]2NH • BH3 Borane-2,6-lutidine BH3 • T H F NaBH4 NaAlH2(OCH2CH20CH3)2 LiAlH(t -BuO)3
Time (h) 3.5 1 3.5 2.5 45 17 17.5 20 1.5
Solvent Et20-benzene Et20 Et20 Et20 Et20 THF Et20 Et20 Et20
Tem.(r) r.t. r.t. r.t. r.t. r.t. - 7 5 - 10 r.t. - 7 5 - 10 r.t.
11:12 7.5:1 7:1 6:1 2.5:1 2.6:1 1:2 4:1 32:1 20:1
yield (%) 96 95 91 10 3 16 19 59 79
a) Molar ratio of the reducing agent to (86) was 10:1. b) Determined from the integration intensity in the ^H-NMR spectrum.
Sodium bis-(2-methoxyethoxy)aluminium hydride (ref. 84) and lithium tri-tertbutoxyaluminium hydride (ref. 85) showed excellent stereoselectivity for podophyllotoxin compared with other reagents. However, these reagents have the disadvantage that low temperature (-75°C) is required to obtain satisfactory yield. Although the stereoselectivity was inferior to that of the above-mentioned aluminium hydride complexes, borane-tert-butylamine complex (ref. 86) and borane ammonia complex (ref. 87) gave the products in more than 90% yield. The reactions proceeded at room temperature, and borane-(err-butylamine complex did not require the use of a nitrogen stream. In conclusion, two kinds of 1-hydroxyphenyltetralin-type lignans, epipodophyllotoxin (1 2) and podophyllotoxin (11), were obtained from desoxypodophyllotoxin (3), which is available in large quantities from the seeds of ovigera L..
Hernandia
601 10 Another Cyclization Reaction of 2,3-Dibenzylbutyrolactone Type Lignan to 4-Aryltetralin-Type Lignan 10.1 Reaction of Arctigenin Monoacetate (88) and Trachelogenin Diacetate (89) with Lead Tetraacetate (ref. 88) Arctigenin monoacetate (8 8) was treated with lead tetraacetate in acetic acid. Purification of the oxidation product by chromatography on a silica gel column afforded 5-acetoxyarctigenin monoacetate (9 0) without cyclodehydrogenation. The ^H-NMR spectrum (CDCI3) showed a signal at d 5.57 ( I H , d, /=8 Hz) due to the C-5 proton, indicating the insertion of an acetoxyl group at one of the benzylic positions. It was shown that acetoxylation in compound (8 8) could be selective because the resistance to oxidation with lead tetraacetate is assumed to depend on the reduced reactivity of the benzylic position due to the presence of an acetoxyl group at the p-position of the aromatic ring. The configuration of the inserted acetoxyl group was established as 5S by a comparison of its ^^C-NMR spectrum (CDCI3) with those of the analogous naturally occurring 5-a77o-hydroxymatairesinol, 5-hydroxymatairesinol and parabenzlactone. The configuration of compound (9 0) was determined in the derived alcohol from its ^^C-shift of 73.9 ppm and in the 5R alcohol parabenzlactone from its shift of 75.4 ppm. Similar shifts were reported in a ^^C-NMR analysis of 7-ephedrine and d-pseudoephedrine (ref. 89). Treatment of trachelogenin diacetate (8 9) with lead tetraacetate in acetic acid also gave 5-acetoxytrachelogenin diacetate ( 9 1 ) . The configuration of the compound (91) is assumed to be 5 5, considering the attack of an acetoxyl radical at the C-5 position from the sterically less hindered side against a hydrogen atom attached to the C-3 position on the butyrolactone ring, as in the case of compound (8 8). This is supported by the coupling constant (/=8 Hz) of the proton at the C-5 position in the ^H-NMR spectrum (CDCI3) of compound ( 9 1 ) .
Pb(0Ac)4 AcOH OCH3
OCH3 OCH3
OCH3
90 : R=H 91 : R=OAc
88 : R=H 89 : R=OAc
Chart 20
602 10.2 Pyrolysis Reaction of 5-Acwtoxydibenzylbutyrolactone-Type Lignan 5-Acetoxydimethylmatairesinol (9 2) and 5-acetoxymethyltrachelogenin (9 3) gave stereoselectively 4-aryltetralin-type lignans (9 4) and (9 5) by the pyrolysis at 260'C, respectively.
JCCP°
H3C0.
ACOM^ H
1
ll
y^OCHa
0CH3 0CH3
OCH3 92 : R=H 93 : R=OH HaCO^ HgCO^
94 : R=H 95 : R=OH
ACO 5
H3CO,
AcO^ H
OCH^
"OCH3
Chart 21
Interestingly, 5-acetoxymethyltrachelogenin monoacetate (9 6) gave naphthalene-type lignan (9 7) by the pyrolysis.
REFERENCES 1 R. D. Haworth, J. Chem. Soc, 194 2, 448; Nature (London), 14 7, 225 (1941). 2 a) R. S. Ward, Chem. Soc. Rev., 11, 75 (1982). b) D. A. Whiting, Nat Prod. Reports, 2, 191 (1985); idem, ibid., 4, 499 (1987). c) D. C. Ayres and J. D. Loike, "Lignans", Cambridge University Press (1990). 3 a) T. Hata, Nippon Kagakukaishi, 6 3, 1540 (1942). b) H. Furukawa, F. Ueda, M. Ito, K. Ito, H. Ishii and J. Haginiwa, Yakugaku Zasshi, 9 2, 150 (1972). 4 C. Nishino and T. Mitsui, Tetrahedron Lett., 19 7 3, 335. 5 H. Yamaguchi, M. Arimoto, K. Yamamoto and A. Numata, Yakugaku Zasshi, 9 9, 674 (1979). 6 K. Noguchi and M. Kanan, Yakugaku Zasshi, 6 0, 629 (1940).
603 7 J. L. Hartwell, A. W. Schrecher and J. M. Johnson, J. Am. Chem. Soc, 7 5, 2138 (1953). 8 J. L. Hartwell and A. W. Schrecher, J. Am. Chem. Soc, 7 6, 4034 (1954). 9 A. W. Schrecher and J. L. Hartwell, J. Am. Chem. Soc, 7 5, 5916 (1953). 10 P. B. Mcdoniel and J. R. Cole, J. Pharm. Sci., 6 1, 1992 (1972). 11 M. Kuhn and A. von Warburg, Helv. Chim. Acta, 6 7, 1546 (1967). 12 H. Yamaguchi, M, Arimoto, M, Tanoguchi, T, Ishida and M, Inoue, Chem. Pharm. Bull., 3 0, 3212 (1982). 13 M. Tanoguchi, M. Arimoto, H. Saika and H. Yamaguchi, Chem. Pharm. Bull, 3 5, 4162(1987). 14 S. Nishibe, S. Hisada and I. Inagaki, Yakugaku Zasshi, 9 4, 522 (1974). 15 J. Harmatha, M. Budesinsky and A. Trka, Collect. Czech. Chem. Commun., 4 7, 644 (1982). 16 M. Kozawa, N. Morita and K. Hata, Yakugaku Zasshi, 9 8, 1486 (1978). 17 H. Erdtman and J. Harmatha, Phytochemistry, 1 8, 1495 (1979). 18 S. K. Koul, S. C. Taneja, K. L. Dhar and C. K. Atal, Phytochemistry, 2 2, 999 (1983); L. P. Badheka, B. R. Prabhu and N. B. Mulchandani, ibid, 2 5, 487 (1986). 19 P. Richomme, J. Bruneton, P. Cabalion and M. M. Debray, J. Nat. Prod., 4 7, 879 (1984); M. -C. Chalandre, C. Pareyre and J. Bruneton, Ann. Pharm. Fr., 4 2, 317 (1984); P. Richomme, J. Bruneton and A. Cave, Heterocycles, 2 3, 309 (1985). 20 K. Tomioka, H. Mizuguchi and K. Koga, Chem. Pharm. Bull., 3 0, 4304 (1982). 21 M. Tanoguchi, E. Hosono, M. Kitaoka, M, Arimoto and H. Yamaguchi, Chem. Pharm. Bull, 3 9, 1873 (1991). 22 D. Takaoka, N. Takamatsu, Y. Saheki, K. Kono, C. Nakaoka and H. Hiroi, Nippon Kagaku Kaishi, 1975, 2192. 23 W. M. Kamil and P. M. Dewick, Phytochemistry, 2 5, 2093 (1986). 24 P. Richomme, J. Bruneton, and A. Cave, Heterocycles, 2 3, 309 (1985). 25 H. Yamaguchi, M. Arimoto, M. Tanoguchi and A. Numata, Yakugaku Zasshi, 101, 485 (1981). 26 W. J. Gensler, R Johson and A. D. B. Sloan, J. Am. Chem. Soc, 8 2, 6074 (1960); H. Kofod and C. J ^ rgensen. Acta, Chem. Scand., 8, 1296 (1954); L. Fa, S. Tianmin and F. Fengyong, Yao Hsueh Hsueh Pao, 1 4, 241 (1979); C. Xingrue, H. Zhibi and Z. Guangfang, ibid, 1 5, 158 (1980); H. P. Plaumann, J. G. Smith and R. Rodrigo, J. Chem. Soc, Chem. Commun., 1980, 354. 27 B. J. Arnold, S. M. Mellows and P. G. Sammes, J. Chem. Soc, Perkin Trans. 1,1 973, 1266. 28 A. S. Kende, M. L. King and D. P. Curran, J. Org. Chem., 4 6, 2826 (1981). 29 J. E. T. Corrie, G. H. Green, E. Ritchie and W. C. Taylor, Aust. J. Chem., 2 3, 133 (1970). 30 E. Wenkert, H. E. Gottlieb, O. R. Gottlieb, M. O. S. Pereira and M. D. Formiga, Phytochemistry, 1 5, 1547 (1976). 31 A. Bax, J. Magn. Reson., 5 3, 517(1983). 32 D. Takaoka, M. Imooka and M. Hiroi, Bull. Chem. Soc Jpn., 5 0, 2821 (1977); S. Nisibe, K. Okabe and S. Hisada, Chem. Pharm. Bull., 2 9, 2078 (1981); L. P. Badheka, B. R. Prabhu and N. B. Mulchandani, Phytochemistry, 2 5, 487 (1986).
604 33 S. K. Koul, S. C. Taneja, K. L. Dhar and C. K. Atal, Phytochemistry, 2 3, 2099 (1984); B. R. Prabhu and N. B. Mulchandani, ibid, 2 4, 329 (1985). 34 M. Tanoguchi, T. Kashima, H. saika, T. Inoue, M. Arimoto and H. Yamaguchi, Chem. Pharm. Bull., 3 7, 68 (1989). 35 a) D. N. Kirk, L. M. McLaughlin, A. M. Lawson, K. D. R. Setchell and S. K. Patel, J. Chem. Soc, Perkin Trans. 1,1985, 35; b) A. Pelter, R. S. Ward, P. Satyanarayana and P. Collins, ibid, 1983, 643; c) Y. Landais and J. -P. Robin, Tetrahedron Lett., 2 7, 1785 (1986). 36 M. Kato, Y. Hayashi, T. Miwa and T. Sakan, Nippon Kagaku Zasshi, 8 5, 225 (1964). 37 a) J. L. Hartwell and W. E. Detty, J. Am. Chem. Soc, 7 2, 246 (1950); b) J. L. Hartwell, A. W. Schrecher and G. Y. Greenberg, J. Am. Chem. Soc., 7 4, 6285 (1952); c) E. Bianchi, K. Sheth and J. R. Cole, Tetrahedron Lett., 1969, 2759. 38 A. M. Duffield, J. Heterocycl. Chem., 4, 16 (1967); A. Pelter, J. Chem. Soc, (C) 1968, 74. 39 T. Gilchrist, R. Hodges and L. Porte, J. Chem. Soc, 1962, 1780. 40 L. H. Klemm, K. W. Gopinath, D. Hsu Lee, F. W. Kelly, E. Trod and T. M. McGuuire, Tetrahedron, 2 2, 1797 (1966). 41 a) Z. Horii, M. Tsujiuchi and T. Momose, Tetrahedron Lett., 1969, 1079; Z. Horii, K.Ohkawa S. Kim and T. Momose, Chem. Pharm. Bull., 1 6, 2404 (1968); J. Chem. Soc, Chem. Commun., 196 8, 653; b) M. Okigawa, T. Maeda and N. Kawano, Tetrahedron, 2 6, 4301 (1970). 42 a) W. Klyne, R. Stevenson and R. J. Swan, J. Chem. Soc (C), 1966, 893; R. J. Swan, W. Klyne and H. MacLean, Can. J. Chem., 4 5, 319 (1967); b) J. Sakakibara, H. Ina and M. Yasue, Yakugaku Zasshi, 9 4, 1377 (1974). 43 a) R. B. Bates and J. B. Wood, J. Org. Chem., 3 7, 562 (1972); b) T. J. Petcher, H. P. Weber, M. Kuhn and A. Von Warzburg, J. Chem. Soc, Perkin Trans. 2,1973, 288. 44 G. Germain, P. Main and M. M. Woolfson, Acta Crystallogr., Sect. A, 2 7, 368 (1971). 45 "International Tables for X-ray Crystallography", Vol. IV, Kynoch Press, Birmingham, 1974. 46 The Universal Crystallographic Computing System (1979). Library of Programs, Osaka University, Computing Center. 47 a) S. Takano, S. Otaki and K. Ogasawara, Heterocycles, 2 3, 2811 (1985); b) A. R. Beard, M. G. B. Drew, J. Mann and L. T. F. Wong, Tetrahedron, 4 3, 4207 (1987). 48 a) T. Ishiguro, H. Mizuguchi, K. Tomioka and K. Koga, Chem. Pharm. Bull., 3 3, 609 (1985); b) K. Tomioka and K. Koga, Tetrahedron Lett., 1979, 3315. 49 Y. Landais, A. Lebrun, V. Lenain and J. -P. Robin, Tetrahedron Lett., 2 8, 5161 (1987). 50 W. J. Gensler, C. M. Samour, S. Y. Wang and F. Johnson, J. Am. Chem. Soc, 8 2, 1714 (1960). 51 G. A. Olah, R. Malhotra, S. C. Narang and J. A. Olah, Synthesis, 1978, 672. 52 K. Shimizu and S, Fukushima, Yakugaku zasshi, 8 7, 1079 (1967). 53 T. Kurihara, M. Kikuchi, S. Suzuki and S. Hisamitsu, Yakugaku Zasshi, 9 8, 1586 (1978).
605 54 L. H. Klemm, K. W. Gopinath, G. C. Karaboyas, G. L. Capp and D. H. Lee, Tetrahedron, 2 0, 871 (1964). 55 B. A. McKittrick and R. Stevenson, J. Chem. Soc, Perkin Trans. 1,1984, 709. 56 J. Villieras and M. Rambaud, Synthesis, 1983, 300. 57 K. Holmberg and B. Hansen, Acta Chem. Scand., Ser. B, 3 3, 410 (1979). 58 B. S. Joshi, N. Viswanathan, V. Balakrishnan, D. H. Gawad and K. R. Ravindranath, Tetrahedron, 3 5, 1665 (1979). 59 a) H. M. Thompson, J. Org. Chem., 3 6, 2577 (1971); b) H. M. Thompson and R. E. Naipawer, ibid., 3 7, 1307 (1972). 60 V. N. Aiyar and F. C. Chang, J. Org. Chem., 4 0, 2384 (1975). 61 P. A. Ganeshpure and R. Stevenson, J. Chem. Soc, perkin Trans. 1,1 981, 1681. 62 T. Kashima, M. Tanoguchi, M. Arimoto, and H. Yamaguchi, Chem. Pharm. Bull., 3 9, 192 (1991). 63 T. Morimoto, M. Chiba and K. Achiwa, Tetrahedron Lett., 3 1, 261 (1990). 64 E. Spath, F. Wessely and L. Kornferd, Chem. Ber., 6 5, 1536(1932). 65 R. D. Haworth and T. Richardson, J. Chem. Soc, 1936, 348. 66 E. Schreier, Helv. Chim. Acta, 4 6, 75 (1963); idem, ibid., 4 7, 1529 (1964). 67 Y. Ikeya, H. Taguchi, I. Yoshioka and H. Kobayashi, Chem. Pharm. Bull., 2 6, 3257 (1978); Y. Ikeya, H. Taguchi and I. Yoshioka, ibid., 2 7, 2536 (1979). 68 S. Ghosal and S. Banerjee, J. Chem. Soc, Chem. Commun., 1979, 165. 69 S. Ghosal, R. P. S. Chauhan and R. S. Srivastava, Phytochemistry, 1 3, 2281 (1974). 70 H. Yamaguchi, S. Nakajima, M. Arimoto, M. Tanoguchi, T. Ishida and M. Inoue, Chem. Pharm. Bull., 3 2, 1754 (1984). 71 H. Koford and C. J ^ rgensen. Acta Chem. Scand., 9, 1327 (1955). 72 a) J. H. Clark, H. L. Holland and J. M. Miller, Tetrahedron Lett., 1976, 3361; b) H. Ishii, E. Kawanabe, N. Fukasaku, T. Yokoshima and M. Akasu, Chem. Pharm. Bull, 3 0, 2761 (1982). 73 D. J. Sam and H. E. Simmons, J. Am. Chem. Soc, 9 6, 2252 (1974). 74 N. J. Lewis, S. Y. Gabhe and M. R. DeLaMater, J. Org. Chem., 4 2, 1479 (1977). 75 R. D. Rieke and S. E. Bales, J. Am. Chem. Soc, 9 6, 1775 (1974). 76 R. G. R. Bacon and S. C. Rennison, J. Chem. Soc, (C), 1969, 312; R. G. R. Bacon and J. R. Wright, ibid., 1969, 1978. 77 P. Buck and G. Kobrich, Tetrahedron Lett., 1967, 1563; P. Wiriyachitra and M. P. Cava, J. Org. Chem., 4 2, 2274 (1977). 78 H. Yamaguchi, M. Arimoto, S. Nakajima, M. Tanoguchi and Y. Fukada, Chem. Pharm. Bull., 3 4, 2056 (1986). 79 a) J. L. Hartwell and A. W. Schrecker, J. Am. Chem. Soc, 7 3, 2909 (1951); b) D. C. Ayres and C. K. Lim, J. Chem. Soc, Perkin Trans. 1,1 972, 1350. 80 W. J. Gensler, F. Johnson and A. D. B. Sloan, J. Am. Chem. Soc, 8 2, 6074 (1960). 81 W. J. Gensler and F. Johnson, J. Am. Chem. Soc, 8 5, 3670 (1963).
606 82 V. N. Aiyar and F. C. Chang, J. Org. Chem., 4 2, 246 (1977). 83 E. J. Corey and J. W. Suggs, Tetrahedron Lett, 1975, 2647. 84 V. BaZant, M. Capca, M. Cerny, V. Chvalovsky, K. Kochloefi, M. Kraus and J. Malek, Tetrahedron Lett., 1968, 3303; T. Nakata, T. Tanaka and T. Oishi, ibid, 2 4, 2653 (1981). 85 A. Kende, M. L. King and D. P. Curran, J. Org. Chem., 4 6, 2826 (1981). 86 F. C. Chang, Synthetic Commun., 11, 875 (1981). 87 B. L. Allwood, H. S. Zavareh, J. F. Stoddart and D. J. Williams, J. Chem. Soc., Chem. Commun., 1984, 1461. 88 S. Nishibe, M. Chiba, A. Sakushima, S. Hisada, S. Yamanouchi, M. Takido, U. Sankawa and A. Sakakibara, Chem. Pharm. Bull, 2 8, 850 (1980). 89 K. Yamasaki and K. Fujita, Chem. Pharm. Bull., 2 7, 43 (1979).
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 O 1996 Elsevier Science B.V. All rights reserved.
607
Studies on the Absolute Configuration of Some Liverwort Sesquiterpenoids Motoo Tori
1.
INTRODUCTION Bryophytes are divided into three classes, Musci, Hepaticae (liverwort) and Anthocerotae, which are known to include 14000,6000 and 300 species, resp)ectively. Various kinds of biologically active compounds have been discovered from bryophytes. For instance there are compounds with hot or bitter taste, ichthyotoxicity, insect antifeedant, antiinflammatory, antimicrobial, antifungus and antitumor activities (1). In Hverworts, there exist oil bodies, while in Musci, no oil bodies are present. This can help us to identify the liverworts. Interestingly, the absolute configuration of some terpenoids isolated from liverworts are the same as those found from higher plants while others have the opposite. For example caryophyllane, drimane, germacrane and guaiane type terpenoids have the same absolute configurations as those of higher plants, while aromadendrane, aristlane, gorgonane, verticillane type terpenoids have the opposite (Fig. 1). A complex example is found in Frullania tamarisci in which (-)-fmUanolide belonging to the eudesmane type was isolated, while morphologically similar species Frullania dilatata produced the (+)-enantiomer (Fig. 2) (2). Thus the absolute configuration of terpenoids found in liverworts are very complicated and need careful analysis. Since the biosynthetic pathways are not clear, we can not discuss these. If the functional groups are not adequate for elucidation of the absolute configuration or the quantity of the isolated substances is minute, it is quite difficult to know the absolute configuration. Thus the synthesis is a necessary tool for such purposes and we have started synthetic work on terpenoids in this area. The main strategy used was optical resolution of the secondary alcohols, and we used the CD spectra for determination of the absolute configurations of the corresponding ketones. The chiral auxiliary used was commercially available l(5)-(-)-camphanic chloride. We now review total synthesis of optically active compounds having hydrindane skeleton as well as a cyclohexanone derivative and their absolute configurations.
608
caryophyllane
drimane
germacrane
guaiane
Skeletons of compounds having the same absolute configuration as those of higher plants
aromadendrane
aristolane
gorgonane
veiticillane
Skeletons of compounds having the opposite absolute configuration to those of higher plants Fig. 1. Skeletons of terpenoids isolated from liverworts.
(-)-fmllanolide Frullania tamarisci [a]D-113°(CHa3)
(+)-frullanolide Frullania dilatata [a]D+114°(CHCl3)
Fig. 2. The complex example of frullanolide
609
2.
Chiloscyphone (-)-Chiloschyphone (1) is a sesquiterpene ketone isolated from Chiloscyphus
polyanthos and its absolute configuration has not been determined (3).
The
structure of this compound was first reported in 1972 (4), but later a total synthesis of this compound revealed that this original structure was wrong (5). Quite recently the correct structure was proposed by Connolly and his co-workers to be that depicted as in 1 (3). The synthesis of this correct structure has not been achieved yet. We plan to introduce a\y-dimethyl groups by 1,4-addition, followed by an intramolecular aldol cyclization to yield a hydrindane skeleton (Scheme 1). The optical resolution must be possible at the stage of compound 4. 3,4-Dimethylcyclohexenone was treated with vinyl magnesium bromide in the presence of cuprous bromide to afford a ketone 2. After protection of the carbonyl group with ethylene acetal, hydroboration, oxidation, and methylation with diazomethane yielded a methyl ester 6. The methyl ester 6 was treated with LDA/allyl bromide to give 7, followed by ozonolysis to afford an aldehyde 8.
I
COoMe
C02Me
(-)-chiloscyphone (1)
Scheme 1. Synthetic plan of chiloscyphone.
>
610
Fig. 3. The NOEs detected for enones 3 and 9.
Acid treatment of 8 resulted in deacetalization and intramolecular aldol cyclization to afford three products, in which the enone 3 was the major. The stereochemistries of the enones 3 and 9, were determined by NOE experiments. On irradiation of methyl groups at C-4 (5 1.03) and C-5 (5 1.08) of the enone 3, a NOE at H-6 (8 3.03) was observed in both cases (Fig.3). While in the case of the enone 9, when H-6 (5 3.06) was irradiated, a NOE at H-4 (6 1.91) was observed. Therefore the cis arrangement of the 1,2-dimethyl groups introduced by 1,4-addition was established at this stage as well as the configuration of the methoxycarbonyl group. Since the alcohol 10 has hydrogen bonding with the carbonyl group at C-1 and acid treatment under similar conditions leads to the enone 3, the stereochemistry was determined as shown in the formula. Hydrogenation of the enone 3 afforded a cis dihydro derivative 11, which was further reduced to give two isomeric alcohols, 4 and 12 (Scheme 2) (6).
611
b, c, d, e, b
COzMe
/—\
CHO COzMe
+ COzMe
1
\
? +
I " COjMe
COjMe
10
C02Me
11
COaMe
Scheme 2. a) CH2=CHMgBr,CuBr«SMe2AHF; b) HOCH2CH2OH, TsOH; c) BH3'THF; H202/NaOH; d) Jones; e) CH2N2; f) LDA, CH2=CHCH2Br; g) O3; Zn-AcOH; h) TsOH/Me2CO-H20; i) H2, Pd-C;j)NaBH4/MeO
612
CamO HO
"9H
H COaMe (+)-4
13 COjMe
f^-
HO
CamO
COoMe
COzMe (-)-4
14
Cam
HO
CamO
H
HO COjMe 15 COoMe 12
COzMe (+)-12
"?H
CamO „ : H
= COzMe 16
COzMe (-)-12
0 (+)-4 or (+)-12
(-)-4 or COaMe (-)-ll
(-)-12 J
=
COaMe
(+)-ll
Scheme 3. a) l(5)-(-)-camphanic chloride, DMAP/CHjClj-Py; b)KOH/MeOH;c) Jones.
613
Both alcohols were treated with l(5)-(-)-camphanic chloride in the presence of DMAP in pyridine to afford a mixture of diatereoisomers, which were separated by HPLC. In the case of alcohol 4, diastereoisomers 13 and 14 were obtained and these were then treated with KOH/MeOH to yield optically active alcohols, (+)-4 and (-)-4. Jones reagent was then used to oxidise these and afforded the ketones, (-)-ll and (+)-ll, respectively. The CD spectrum of (-)-ll showed a negative Cotton effect, and that of (+)-ll a positive. Using the Octant rule, the absolute configurations were assigned as shown in Scheme 3 (7). These results were confirmed by the X-ray analysis of ester 14 (8). The alcohol 12 afforded the ketones (-)-ll and (+)-ll through 15, 16 and 12. The absolute configurations of 15,16 and 12 were determined as shown in Scheme 3.
"9H COiMe (+)-4
(+)-chiloscypholone (18)
(-)-chiloscyphone (1)
isochiloscyphone (19)
Scheme 4. a) POCl3/Py/100°C; b) LiAlH4; c) Swem; d) CH2=C(Me)MgBr; e) Jones
614
One of the alcohols, 12, afforded the disubstituted olefin as a sole product on dehydration, while the other, 4, gave only the trisubstituted olefin 17. The optically active chiloscyphone, (-)-l, [a]^ -15.r (Ht. -24.4') (3), was synthesized in 4 steps from (+)-17, which was obtained from (+)-4 (Scheme 4). The absolute configuration of the alcohol (+)-4 is as shown, therefore the absolute configuration of the natural product was determined as depicted in (-)-l. The absolute value of the specific rotation was a little bit smaller in the synthetic one which is presumably due to the minute quantity of the samples. However we have further confirmed the absolute configuration by the independent synthesis of the antipode (+)-l, [o]^ +24.2% starting from (-)-4 (7, 8). (+)-Chiloscypholone (18) was also isolated as a natural product (3), and it yielded (-)-chiloscyphone (1) and isochiloscyphone (19) on dehydration. Thus the absolute configuration of (+)-18 was also elucidated (Scheme 4).
3.
Tamariscol The liverwort sesquiterpene, (-)-tamariscol (20), isolated from Frullania
tamarisci is a very unstable compound under acidic conditions and has been found in the European species but not in the Japanese species. This is a very strange phenomenon and at first we thought that these two species are different. However a detailed inspection of the Japanese species revealed that those collected at higher altitude or in the northem area do contain tamariscol, but those collected in other parts of Japan do not (9). Connolly and his co-workers investigated the structure of this compound using 2D-NMR techniques including a 2DINADEQUATE and established the structure formulated as 20. However the absolute configuration remained undetermined (10). Tamariscol belongs to a pacifigorgiane skeleton, in which pacifigorgiol (21), isolated from the soft coral Pacifigorgia admsii, is included (Fig. 4) (11). The absolute configuration of this compound is also not known.
615
(-)-tamariscol (20) Frullania tamarisci
pacifigorgiol (21) Pacifigorgia adamsii
Fig. 4. Pacifigorgiane-type sesquiterpenes.
We are interested in compound (-)-20, which has a pleasant fragrance. Later the quality of the fragrance was recognized by companies world wide, although the original paper just mentioned a "pungent taste" (10). Thus we are interested in not only the structure but also the absolute configuration of tamariscol. It is important to know the absolute configuration for developing this compound as a perfume. We started the synthesis to confirm the relative configurations at C-6 and C-7 first. This was previously assigned by consideration of the ^^C NMR chemical shifts and by comparing with known compounds (10). From the literature, compound 23 can be obtained from 22 (12). The relative structure of the enone 23 was verified by independent synthesis of pacifigorgiol (21) (12). It was thought that reductive acylation of the enone 23, through the epoxide 25, may lead to the natural product, but this assumption was not very good. However compound 20 was prepared by alkylation of 27, but the yield was very poor (Scheme 5).
616
c=:> I H COOMe
^
CH2OH
25
H I
>
3
27 tamariscol (20) Scheme 5. Synthetic plan of tamariscol.
At first, the enone 23 was catalytically reduced to afford almost exclusively a cis fused ketone which was expected from the preceding examples. Birch reduction gave cis and trans products in a ratio of ca. 5:4 (13). In general six-membered ring ketones produce trans products preferentially under Birch reduction conditions. In this case the trans product was not the preferred product and the ratio did not change when several different conditions were tried. Thus Birch reduction followed by carboxylation with carbon dioxide of the enolate and methylation afforded the ester, whose ratio of the cis and trans products was ca.
617
3:2. The stereochemistries were estabhshed at the stage of the ketones 38 - 40. The enolizable P-ketoester was cw-fused compound 28, and the non-enolizable one (keto-form) was trans-fustd compound 24. Reduction of the carbonyl group, mesylation or benzoylation, and then base treatment yielded the corresponding a,P-unsaturated ester in each case. Further reduction afforded the allyl alcohols 29 and 30 (Scheme 6).
c, d, e, f a,b
'O^^OMe
23 g, h, i, f
30
Scheme 6. a) NHsCliq.), ^BuOH; C02/rHF/-78^C; b) CH2N2; c) NaBH4AleOH; d) MsCl/Py; e) DBU/PhH; f) LiAlH4; g) L-Selectride; h) BzCl/Py; i) LDA.
The cis and trans allyl alcohols, 29 and 30, were epoxidized and acetylated to afford the acetates 31, 32, 35, and 25. The stereochemistries were revealed at a later stage. The epoxide 31 was treated with Me^CuLi to give a diol 33 in good yield (Scheme 7a). The other epoxides, 32 and 35 were treated similarly to give diols 34 and 36, respectively, however the epoxide 25 did not react at all. The reason for this is not yet clear, but one reason may be that the epoxides tend to open in the direction to yield diaxial alcohols. In order to obtain an a-hydroxy aldehyde, each diol, 33, 34, and 36, was treated with NCS/Me^S to yield an aldehyde with a methylthiomethyloxy group (for example 37). However these
618
aldehydes decomposed under Wittig reaction conditions (Scheme 7b). A route along this pathway was therefore abandoned and instead we turned our attention to the synthesis of ketone 27.
a- c
29
31
31
33
38
32 CH2OH
34
39
Scheme 7a. a) LiAlH4; b) mCPBA; c) Ac20/Py; d) Me2CuLi; e) PCC
The diols, 33, 34, and 36, were oxidized with PCC to afford in good yield the ketones 38 - 40 respectively. The stereochemistries were determined by a detailed inspection of the NMR spectra of these compounds. The hydrogen at C-1, a to the carbonyl group in 38 and 39, will show the same coupling pattern if these are the trans fused compounds. However, if these are cis fused compounds they have different coupling patterns as they adopt different conformations due to
619
O'^l H CH2OAC
35
35
Hcri H CH20H
36
40
O'^i H
\ ^ J CHO 37
Scheme 7b. a) LiAlH4; b) mCPBA; c) AcsO/Py; d) Me2CuLi; e) PCC; f) NCS/Me2S the difference in the configuration of the methyl group at C-3. In fact, compound 38 displays H-1 at 5 2.70 (br t, J=6.4 Hz) and compound 39 at 5 2.79 (q, J=8.5 Hz). The coupling patterns of these protons are quite different from each other and thus show that these are the cis fused compounds. Therefore compound 38 adopts a non-steroidal conformation having an equatorial methyl group at C-3, while in the case of 39, the methyl group is equatorial and adopts a steroidal conformation. Thus ketone 40 is a trans fused compound and the stereochemistry of the methyl group at C-3 is axial, because a NOE between the hydrogen at C-1 and the methyl group at C-3 was observed. So the methyl group at C-3 of 40 is a-axial (Fig. 5).
620
M
"IT 'fi
v " ^ ^ ^
V'
(brt,y=6.4Hz)
2.79 (q, 7=8.5 Hz)
38
39
5HI2.70
8H,
H
'P^ ^•"^NOE 5H,2.13
(ddd, 7=12.7,10.3,7.3 Hz) 40
Fig. 5. Conformations and coupling patterns of ketones 38 - 40.
As ketone 27 was not obtained directly through 25 it was thought that isomerization of the other ketones may give this compound. The MM2 calculation predicts that the trans isomer 27 is the most stable among the four possible isomers. Thus 38 was treated with KfiOjfMeOH under isomerization conditions. Unexpectedly compound 27 was not the most abundant formed, but compound 38 was the major one and compounds 39 and 40 were formed only in minute amounts as expected. The reaction mixture was refluxed overnight, but the ratio did not change, suggesting that the equilibrium was complete and the energy difference between 38 and 27 must be very small. In order to obtain the desired ketone, the enone 23 was converted into 41, followed by 1,4-reduction to afford 26 as shown in Scheme 8. Hydroboration-oxidation gave the alcohol 42, presumably by attack from the less hindered p face. In fact, oxidation of 42 with PDC afforded the ketone 38 . Isomerization of the ketone 38 under K^COg/MeOH conditions yielded 27 . The ratio did not change even if it was started from 39 or 40 (Scheme 8).
621
23
c
41 H
,.-s^p i H HO 4:2
26 H
d • ' • ^ ^ "
.g>
Q,
^
nH
0
38
H
O
H
39
40
Scheme 8. a) Ph3P=CH2; b) Li/NH3(liq.); c) BH3.THF; H202/NaOH; d) PDC; e) K2C03/MeOH/reflux.
The stereochemistry of alkylation was next investigated.
2-
Methylcyclohexanone (43) was treated with MeLi to afford predominantly the trans methylated product 45 . However, when the aluminium compounds MAD or MAT were present, they coordinate to the carbonyl group and cause the configuration to be reversed (14). Very complex cases using other sophisticated alkylating reagents are not reported. When 2-methyl-l-propenyl magnesium bromide was reacted with the ketone 27, the isomer of natural product 46 was the
622 sole product. However the lithium salt of this bromide afforded natural product 20, but the yield was very poor. Several trials did not change the yield. In these cases alkylation with groups other than a methyl group did not result in a reversed stereochemistry, and the methodology obtaining the reversed stereochemistry has not been developed yet (Scheme 9). Therefore the relative stereochemistry of tamariscol was confirmed to be completely correct.
MeLi
44 none MAD
8 93
Me2C=CHX
X=MgBr X=Li
tamariscol (20)
46
0 1
100 30
Scheme 9. Stereochemistry of alkylation
623
As the total synthesis was completed we turned our attention to the absolute configuration next. The yield of alkylation was so bad that we decided to degrade the natural product. Collection and extraction of Frullania tamarisci afforded (-)-tamariscol (20), which was then epoxidized and treated with LiAlH/Etp. The reaction did not proceed in ether but, in benzene solution the allyl alcohol 48 was produced. Oxidation with NalO^ gave the ketone (+)-27, which was completely identical with the racemic synthetic product. The CD spectrum of (+)-27 showed the negative Cotton effect at 294 nm and the absolute configuration was suggested by the Octant rule as depicted in the formula. Independently (-)-27 was prepared from (-)-carvone (49) using a 12 step reaction (15, 16). Thus (-)-carvone (49) was reduced with NaTeH to give the trans-dihyAxo derivative 50, which was reduced and treated with dihydropyrane to afford compound 51. Ozonolysis and WittigHomer reaction provided the methyl ester 52, which was converted into the diol 53 in three steps. Swem oxidation and aldol condensation afforded the enones 54 and 55 after separation. The ketone (-)-39 was obtained by catalytic reduction of 55. As already mentioned, (-)-39 was isomerized into the optically active ketone (-)-27, whose CD spectrum showed the positive Cotton effect at 294 nm, which is opposite to that obtained for the natural one. Therefore the absolute configuration of the natural compound, (-)-20, should be assigned as depicted in the formula (Scheme 10). Each ketone, 27, 38, 39 or 40 has a pleasant fragrance. The fragrance of (-)-tamariscol (20) is woody, earthy floral, a little bit different from these ketones mentioned. A minute quantity of such a ketone as an impurity may alter the quaUty of fragrance. An easy method for qualitative evaluation of the quality of fragrance is being waited.
624
(-)-20
g.h
1, m \^»**'
HH
O
(-)-38
OH
OTHP
OH
52
53
>+ L J ^> %^»'+* o (-)-27
(-)-39
40
CD[ (+)-27]: Ae-1.33(CHCl3)
Scheme 10. a) mCPBA; b) LiAlH4/PhH; c) NaI04; d) NaTeH/EtOH; e) NaBH4; f) DHP/PPTS; g) O3; PPhj; h) (MeO)2P(0)CH2C02Me/ NaH; i) Hj/Pd-C; j) LiAlH4; k) PPTS/MeOH; 1) Swem; m) PhCOjH/ EtjN; n) Hj/PtOj; 0) KzCOj/MeOH
625
4.
Conocephalenol Conocephalenol (56) has been isolated from the European liverwort
Conocephalum conicum and has a similar structure to tamariscol (20) (17). The first compound of this class was brasilenol (57), which was isolated from the alga Laurencia obtusa oi ihtrnxdihrdxiohAplysia brasiliana (Fig. 6) (18). The planar structure of conocephalenol (56) was revealed by Connolly and his group by using a 2D-INADEQUATE spectrum and is shown in Fig. 6 (17). However the ^H NMR spectrum of this compound is very congested in an upfield region and has no characteristic signals. Hence it was difficult to determine the stereochemistry by only spectroscopic methods. It is very interesting from the evolution point of view that similar terpenoids with the same skeleton have been isolated from both liverworts and algae. Since the absolute configuration of no terpenoids in this class has been elucidated, we have started to synthesize conocephalenol (56), whose relative and absolute configuration has not been determined.
OH
brasilenol (57) Laurencia obtusa Aplysia brasiliana
(-)-conocephalenol (56) Conocephalum conicum
Fig. 6. Brasilane-type sesquiterpenes
626
COOMe
H
63
0
O 23
"^ 'O'^OMe
COOMe
59 +
<
58
COOMe
V-
60
O
65
\
OH
66 conocephalenol (56)
Scheme 11. Synthetic plan of conocephalenol
627
Since conocephalenol (56) has a very similar structure to tamariscol, the synthetic strategy used was similar to those used in the tamariscol synthesis. If the trisubstituted double bond of the a,p-unsaturated ester, 61 or 62, could be isomerized into the ester, 63, with the tetrasubstituted double bond, compound 56 might easily be prepared. If this isomerization does not proceed, this could be accomplished through the epoxide 64. The optically active compound might be synthesized from the ketone 65 or its alcohol (Scheme 11). In order to verify the relative configuration, the synthesis of the racemic compound was first undertaken. The enone 23 was methylated twice to afford 58 followed by reductive acylation to give 59 and 60. From the previous experience, the enolizable compound, 59, must be assigned to a cis fused one, while the ketonic compound, 60, to a trans fused one. Reduction and elimination afforded the esters 61 and 62, respectively. Attempted isomerization of these compounds into 63 under various conditions failed. The epoxide, 64, was prepared by epoxidation of the ally lie alcohol obtained by reduction of the ester, 61. The diol obtained by LiAIH^ reduction was acetylated to afford a monoacetate, 67, which was dehydrated to give the trisubstituted- and tetra-substituted olefins. These were separated, the desired acetate was hydrolyzed and finally oxidized to the aldehyde 68. The aldehyde was methylated, oxidized and again methylated with MeLi to yield conocephalenol (56), whose spectral data was completely identical with those of the natural one, confirming that the relative configuration is 1R*,9S* (Scheme 12) (19). The optically active compound may be obtained by resolution of the camphanoyl derivative at a certain stage. The initial target is the ketone 71. The enone 58 was catalytically reduced and treated with NaBH^ to afford a cis fused alcohol 69, which was dehydrated to 70 and further hydroboration-oxidation gave an alcohol 71. This alcohol was treated with l(5)-(-)-camphanic chloride and the resulting diastereoisomers were separated with the HPLC to give (+)-72 and (-)-73. Compound (-)-73 was hydrolyzed followed by Jones oxidation to yield a ketone (-)-75 (Scheme 13). The ketone (-)-75 is a d^-fused hydrindanone, which adopts only one conformation due to the presence of the gem-dimethyl groups (Fig. 7).
528
„jdy^:^-;X|5.;ri5 23
H COOMe
58
60
59
i,e,f,c " " I L /
60
1H
COOMe 61
62
1.J
i,e
H COOMe 61
k,i,l
64
m, 1, m
T' OH conocephalenol (56)
Scheme 12. a) LDA/MeI(twice); b) Li/NHjOiq.); COj; c) CH2N2; d) NaBH4/MeOH; e) AcaO/Py; f) tBuOK; g) L-Selectride; h) BzCl/ Py; i) LiAlH4; j) mCPBA; k) SOCI2; 1) PDC; m) MeLi
629 The CD spectrum of the ketone (-)-75, has a negative Cotton effect and thus suggests the absolute configuration shown in Fig. 7.
H
58
"d^ 70
69
i H OCam.
71
(-)-73
(+)-72
(-)-74
Scheme 13. a) Hj/Pd-C; b) NaBH4/MeOH; c) POClj/Py; d) BHj'THF; HzOj/NaOH; e) l(S)-(-)-cam.-Cl/CH2Cl2-Py/DMAP; f)KOH/MeOH; g) Jones.
630
+ • • • —
+
Fig. 7. Conformations and the absolute configuration of ketone (-)-75.
Alkylation of the cis hydrindanone, (-)-75, was first investigated. Grignard reaction of 2-bromopropene afforded two products, 76 and 77, in a ratio of 3:1 in 45% yield. Compound 77 could be concluded to be cis concerning the isopropenyl group and hydrogen at C-6 as a NOE between the isopropenyl and one of the dimethyl groups was observed. The lithium salt of 2-bromopropene afforded 77 as the sole product, but the yield was only 20%. For dehydration in the next step, it is desirable that the hydrogen at C-6 and the hydroxyl group at C-5 are trans to each other as in 77. This is supported by the fact that 76 gave 78, and 77 afforded 79 in three steps in both cases (Scheme 14). The yield of the desired 77 was not very good anyway. This is attributable to the cis fused ketone 75. Thus in the case of a trans derivative, the isopropenyl group may be introduced to avoid a 1,3-diaxial repulsion of the axial methyl group at C-3 to afford a trans product concerning the hydrogen at C-6 and the hydroxyl group at C-5. Therefore the
631
ketone 75 was next isomerized. The trans derivative is more stable than the cis by 1.4 kcal/mol as predicted by the MM2 calculations. In fact the isomerization under K^COg/MeOH/reflux resulted in (-)-75 : (-)-80 = 3:2. This is again not compatible with the result of the calculation. The reaction of the trans ketone (-)-80 with the hthium salt of 2-bromopropene at -78T afforded (-)^l in 91% yield. The product is an axial alcohol, which was predicted as mentioned above. The evidence is that the NOE between the methyl group at C-3 and isopropenyl
y^ ^
THF 75
76
77
R=MgBr
45%
3
1
R=Li
20%
0
100
a, b, c
76
77
78
79
Scheme 14. a) O3; b) MejS; c) SOCIj/Py
632
group was not observed. After ozonolysis, dehydration with thionyl chloride gave the ketone (-)-79 with the tetrasubstituted double bond as the sole product, confirming the implication mentioned above. Methylation with MeLi afforded (+)-56, whose specific rotation was [a]j^ +5.85' (Scheme 15).
H
/
(-)-80
(->81
c, d
OH (-)-79
(+)-conocephalenol (56)
Scheme 15. a) KjCOs/MeOH; b) CH2=C(Me)Li; c) O3; Me2S; d) SOCl2/Py; e) MeLi
The specific rotation of natural conocephalenol was not given in the original report, therefore we tried to isolate natural product ourselves. However, the liverwort Conocephalum conicum collected in Japan does not contain conocephalenol at all, while the liverwort collected in Europe does contain conocephalenol. We have isolated pure conocephalenol (56) from the German species and the specific rotation is [a]^ -All", confirming that the absolute configuration of the natural one should be formulated as depicted in Fig. 6.
633
5.
Brasilane sesquiterpene isolated from Laurencia implicata Wright, Konig, and Sticher, at ETH, reported that they isolated a brasilane
type sesquiterpene from Laurencia implicata and the structure was proposed as depicted in 82 (20). When we saw this structure we had a feeling that this must be something else, because we know at least two compound, tamariscol (20) and conocephalenol (56), which are very similar to this molecule. Thus we started to synthesize this molecule in order to know the absolute configuration. First of all we wanted the racemic compound for confirmation of the relative configuration. Since this molecule has an a-oriented methyl group in the five-membered ring we have to change the starting compound to another compound other than 58. Therefore we plan to prepare 82 by intramolecular aldol cyclization as shown in scheme 16a. The 1,4-addition of the vinyl group into 5,5-dimethyl-2-cyclohexen-l-one, acetalization, Wacker oxidation, Homer-Emons olefination, and hydrogenation afforded the ester 85. Reduction of the ester 85 followed by deacetahzation and Swem oxidation gave the keto-aldehyde 86 which was subjected to aldol conditions to yield the isomeric enones 87 and 88. The separation was not accomplished at this stage, but was carried out after hydrogenation of the mixture. Since we had a sample of compound 75 the configuration of the other ketone, 89, was immediately assigned as a-oriented. The ketone 89 was alkylated to afford the alcohol 90 in 70% yield. The similar two step reaction gave the enone 91, which was methylated to yield the desired compound 82. that
The ^H NMR spectmm was similar to
634
a,b,c,d,e
y^^
f,g,h CHO
85
86
H f
88 H ?
l,m
87
HOn ft 90 H ?
91
82 Scheme 16a. a) CH2=CHMgBr/CuBr/Me2S; b) HOCH2CH20HA'sOH; c) PdCyCuClz/DMF-HjO; d) (MeO)2P(0)CH2C02Me/NaH; e) H^dC; f) LiAlH4; g) TsOH/THF-HzO; h) Swem oxid; i) TsOH/PhH/reflux; j) H2/Pt02/hexane; k) CH2=C(Me)MgBr; 1) O3; Me2S; m) SOCI2; n) MeLi; o) 2,4-dinitrophenylhydrazine/camphorsufonic acid
635
Scheme 16b. a) H2/Pt02/hexane; b) NaBH4; c) Jones oxid.; d) CH2= C(Me)MgBr, e) O3; MejS; f) SOCI2; g) MeLi; h) l(S)-(-)-camphanyl chloride/DMAP/Et3N; HPLC.
636
reported in the literature, but the ^^C NMR spectrum was completely different (Scheme 16a) (21). Thus we suspect that the real structure must be the one having the P-oriented methyl group (namely 83). It was much easier for us to prepare compound 83, because we have already made compound 78. Simply methylation of the enone 78 afforded the desired compound 83. The ^H NMR spectrum was somehow similar, but the ^^C NMR spectrum was again completely different from that reported (Scheme 14 and 16b) (21). The correctness of the structures of our synthetic products were confirmed by X-ray crystallography of the hydrazone, 92, which was prepared from the enone 91 with 2,4-dinitrophenylhydrazine in the presence of camphorsulfonic acid. The other evidence came from X-ray analysis of the camphanoyl ester 96, which was derived from ketone 80. However this case is rather complicated as when the enone 88 was hydrogenated, not only the cis product 75 but also the trans product 80 was obtained in minute amount. This mixture was reduced by NaBH^ to afford three products, 93, 71, and 94. The two major products are of course the isomeric cis compounds, while the minor one is the trans alcohol 94. The trans one was converted into camphanoyl esters and these diastereoisomers were separated by HPLC. The ester 96 was crystallized and its structure was analyzed by X-ray. Thus the structures of both isomers were confirmed (Scheme 16b). At this stage we have analyzed the original spectmm which appeared in the literature (20). According to this there is a big coupling constant which is attributed to the one between H-1 and H-9, which are cis to each other. However, we doubt this assumption. The big coupling must be a trans relationship between H-1 and H-6, whose dihedral angle is 180° to each other. This may explain other small couplings with H-1 as shown in Fig. 8. Therefore we suspect that the real structure for this sesquiterpene must be the trans derivative with the [J-oriented secondary methyl group as in 84. In order to prepare this molecule we can go back again to the synthesis of conocephalenol. Compound 62 is a good candidate for our synthesis (Scheme 17). Thus the ester 62 was methylated with MeLi to give exactly the same compound as the natural product. Both the ^H and ^^C NMR spectra were identical with those reported (20).
637
c, d,e
1H CO^Me
62
H ^Q^
84
g,h 58
j,k
1, c, d, e, f
98
(-)-84 Scheme 17. a) Li/NHj; COj; b) CHjNj; c) L-Selectride; d) AcjO; e) tBuOK; f) MeLi; g) Li/NHj; h) NaBH4; i) l(S)-(-)-camphanyl chloride/DMAP/Et3N; HPLC; j) KOH; k) Jones; 1) LDA/NCCOjMe
638
^=HJ*^
UKA'
r^*proposed cis derivative
/ J-. \i\
large ca .180^ — three protons
large ca.180* — only one proton
Fig.B. Qiem-3D expression of cis and trans fused compounds.
The optically active compound can be prepared in the same way as mentioned above. We started with the ketone 58, which was treated under Birch reduction conditions followed by NaBH^ to afford a P-alcohol 97. The camphanoyl derivative of this alcohol was separated by HPLC to give 98 and 99. The ketone (+)-100 derived from 98 showed a negative Cotton effect in the CD spectrum. The absolute configuration of this ketone must be formulated as depicted for (+)-100 from the Octant rule. The ketone (+)-100 was converted to (-)-84 in five steps as used in the former synthesis. The specific rotation of (-)-84 is [a]^ -11.5' (c=4.2, CHCI3) (lit. [a]^ -3.0° (c=0.17, CHCI3) (20)). Therefore the absolute configuration of the natural product is formulated as (-)-84.
639
6.
Tridensone Tridensone was isolated recently by Wu and Chen from the liverwort B^zzanza
tridens (22) and its structure was determined to be 101. However, the absolute configuration has not been established. The structure of 101 is very rare and unique from the biogenetic standpoint. It is considered as a seco-eudesmane-type skeleton derived from the C-C bond fission of C-6 and C-7 (Fig. 9). The same group has already reported the isolation of tridensenal (102) from the same liverwort (23). This is considered to be derived from an eremophilane-type sesquiterpene by C-C bond fission of C-5 and C-6 (Fig. 9). These two sesquiterpenoids are thus closely related from the biogenetic point of view. Establishment of the absolute configuration of either terpenoid presumably suggests all the absolute configurations of these family of sesquiterpenoids. We have synthesized the compound proposed (22), 101, and found that the original structure should be revised to 109 (24).
13 1
4
'vA r^
7
11 1 ^
^4
y // 10^
0
12 1
= 15
tridensone (101)
T
seco-eudesmane
CHO
tridensenal (102)
seco-eremophilane
Fig. 9 Tridensone (101) and tridensenal (102) and their presumed skeletons
640
Optically active compounds having a quaternary chiral center next to a carbonyl group can be synthesized by 1,4-addition of the enamines of the chiral amine auxiliary (25). This methodology using (5)-(-)-phenetylamine was reported by d* Angelo et al. (26) and we used this strategy to construct the chiral centre at C-5 of tridensone (101). We started from the known chiral ketone 103 (25).
OH k
A
o
b,c
C00CH3
(+)-103
104
0SiMe2'Bu
OSiMCi'Bu
e,f (+)-101a + (-)-lOlb
107
Scheme 18. a) LiAlH4, EtjO; b) TBDMSCl, NEtj, DMAP; c) PDC/CH2CI2; d) LDA/Mel; e) PhsPCHj+Br-, BuLi; f) Bu4N*F; g) Swem oxid.; h) CH3CH(Br)CH3/Mg, Et20; i) Jones Oxid. then AgNOj-SiOj
641
The chiral ketone (+)-103 was synthesized by a literature route (25). The ketone (+)-103 was reduced by LiAlH^ to afford a mixture of diols 104. The primary alcohol of this diol 104 was selectively protected by a TBDMS group followed by oxidation with PDC to give the ketone (+)-105. Methylation of (+)-105 at this stage afforded a mixture of the diastereoisomers (ca, 3:2) 106 after several trials using different conditions. This ratio could not be altered by quenching the enolate or under equilibration conditions. Thus the mixture of 106 was treated with triphenyhnethylphosphonium bromide/BuLi followed by deprotection afforded a mixture of alcohols 107. Swem oxidation of 107 to an aldehyde, alkylation by isopropyl magnesium bromide and Jones oxidation gave a mixture of the diastereoisomers of the desired ketone and its isomer, which was separated by silver nitrate-impregnated silica gel column chromatography to afford pure (+)-101a and (-)-lOlb (Scheme 18). 13
13
I
kt><
14
•
15
15
(+)-101a
(-)-lOlb
^•"'^s^
^.'^^S.
^X*>s^
^ 1 ^ f\
NOE
Fig. 10 NOE's detected for (+)-101a and (-)-lOlb by NOESY experiment.
642
OMe OMe
metachromin D (108)
TABLE 1 NMR data for tridensone and synthetic compounds
c 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
101
101a
101b
108'
33.4
33.4
34.0
33.98
37.3
37.3
37.1
37.20
21.8
21.8
21.8
21.89
41.4
41.4
38.7
38.72
39.5
39.6
39.1
39.05
158.1
158.2
158.8
159.42
30.7
30.7
34.6
35.35
35.4
35.5
35.5
35.40
215.3
215.5
215.4
41.0
41.0
41.0
18.4
18.4
18.4
18.3
18.3
18.4
26.3
26.3
25.0
24.98
104.9
104.9
103.8
103.71
19.2
19.3
19.5
19.61
* Only the corresponding data are cited and the numbering is applied for the tridensone system (22).
643
The relative stereochemistry of these ketones were unambiguously established by NOESY experiments (Fig. 10) as shown in the formula. Namely, in the case of (+)-101a, NOEs are observed between the higher part methyl group signals (H-13 and H-15) and both of the e;co-methylene signals (H-14). On the other hand, only one NOE between the lower field e;co-methylene signal (H-14) and one of the higher part of the methyl signals (H-13 or H-15) was observed, in the case of (-)-lOlb. Furthermore the other NOE between the higher field ejco-methylene signal and one of the methylene signals assigned to H-7 (by COSY experiment) was observed. These results clearly show that compound (+)-101a has two methyl groups cis to each other and adopts the conformation shown in Fig. 10. The other compound (-)-lOlb has trans-dimcihyl groups and is recognized as having the conformation shown in Fig. 10. The ^H NMR spectrum of natural tridensone (101) was identical with that of synthetic (+)-101a, not with (-)-lOlb, which has the desired trans-dimtihyl groups. Then we compared the ^^C NMR data of the natural product with those of the synthetic products and the compound having the similar partial structure, metachromin D (108), which was isolated from the marine sources (27). The results are shown in Table 1. On comparison the chemical shifts of the synthetic (-i-)-lOla were identical with those of the natural compound 101 reported by Wu and Chen (22). The data of (-)-lOlb can be compared with a cyclohexane part of those of metachromin D (108) (27). The ^^C NMR data also indicate that the assignment of the relative configuration of the natural tridensone (101) was incorrect and it should be revised to the isomer 101a having cis -dimethyl groups. Since the specific rotation of the synthetic (+)-101a was [a]j^ +12.7' (c 0.85 in CHCI3) {lit., [a]^ A5.V (c 0.93, CHCI3)) (22), the absolute configuration of the natural compound should be the antipode of (+)-101a as shown in the formula 109. 7.
Conclusion What can be predicted now that the absolute configuration is solved ? It is
likely that 1 is derived from the eremophilane skeleton by ring contraction (3). That is, the biosynthetic route of this liverwort might be the same as that of the
644
(-)-chiloscyphone (1) H^
'H
(-)-tamariscol (20)
H
OH
(-)-conocephalenol (56)
Fig. 11. Plausible biosynthetic pathways of three sesquiterpenoids
645
Fig. 12. Possiblebiogeneticpathway of tridensone (101)
higher plants, namely the normal type compounds are produced. However, in the case of 20, if this compound comes from caryophyllene as shown in Fig. 11, the pacifigorgiane skeleton may be from the ent series (10). Although the biosynthesis of the brasilane skeleton is not yet clear, a p-substituted methyl group wiU presumably be produced on cyclization of humulene (Fig. 11) (27). Similar compounds have been isolated from the algae and their absolute configurations have not been determined. When these are apparent we can compare the results and more interesting discussions can be carried out. Tridensone may be derived from the ent series of eudesmane-type compound (Fig. 12). It is very important to compare the absolute configuration of these kinds of terpenoids found in liverworts with one another. Acknowledgments We thank Professor Y. Asakawa (this University) for valuable discussion and encouragement. We also thank Professors J. D. Connolly (Glasgow University, Scotland) and C. -L. Wu (Tamkang University, Taiwan) for sending spectra of the natural products. This work has been carried out by Dr. M. Sono, Mr. T. Hasebe,
646
Mr. K. Nakashima, Mrs. Y. Sono (nee Nakaki) and other undergraduate students, to whom many thanks are due. We thank Mr. S. Takaoka (this University), Dr. K. Ogawa and Mr. S. Yoshimura (The University of Tokyo) for X-ray analysis and MM2 calculation, respectively. This work was partly supported by The Research Foundation for Pharmaceutical Sciences.
REFERENCES 1 Y. Asakawa, in "Progress in the Chemistry of Organic Natural Products" (Ed. by W. Herz, H. Grisebach, G. W. Kirby), Springer-Verlag, Wien, 42, 1982,pp.l-285. 2 H. Knoche, G. Ourisson, G. W. Perold, J. Foussereau, and J. Maleville, Science, 166 (1969) 239-240. 3 J. D. Connolly, L. J. Harrison, and D. S. Rycroft, /. Chem, Soc. Chem. Commun., (1982) 1236-1238. 4 A. Matsuo, Tetrahedron, 28 (1972) 1203-1209. 5 J. -L. Gras, J. Org. Chem., 46 (1981) 3738-3741. 6 M. Tori, T. Hasebe, and Y. Asakawa, Chem. Lett., (1988) 2059-2060. 7 M. Tori, T. Hasebe, and Y. Asakawa, Bull. Chem. Soc. Jpn., 63 (1990) 1706-1712. In this ref. the sign of the CD spectrum was somewhat confusing. The absolute configuration is correct as expressed in the ref 8. 8 M. Tori, T. Hasebe, Y. Asakawa, K. Ogawa, and S. Yoshimura, Bull. Chem. Soc. Jpn., 64 (1991) 2303-2305. 9 Y. Asakawa, M. Sono, M. Wakamatsu, K. Kondo, S. Hattori, and M. Mizutani, Phytochemistry, 30 (1991) 2295-2300. 10 J. D. Connolly, L. J. Harrison, and D. Rycroft, Tetrahedron Lett. 25 (1984) 1401-1402. 11 R. R. Izac, S. E. Poet, W. Fenical, D. van Engen, and J. Clardy, Tetrahedron Lett., 23 (1982) 3743-3746. 12 M. Martin and J. Clardy, Pure & Appl. Chem., 54, (1982) 1915-1918. 13 M. Tori, M. Sono, and Y. Asakawa, Chem. Pharm. Bull., 37 (1989) 534-535. 14 K. Maruoka, T. Itoh, M. Sakurai, K. Nonoshita, and H. Yamamoto, J. Amer. Chem. Soc, 110 (1988) 3588-3597. 15 M. Tori, M. Sono, and Y. Asakawa/. Chem. Soc. Perkin Trans 1, (1990) 2849-2850. 16 M. Tori, M. Sono, Y. Nishigaki, K. Nakashima, and Y. Asakawa, J. Chem. Soc, Perkin Trans 1, (1991) 435-445.
647
17 18 19 20 21 22 23 24 25 26 27
28
J. D. Connolly, in "Studies in Natural Products Chemistry" (Ed. by Attaur-Rahman), Elsevier, Amsterdam, 2,1988, pp. 261-275. M. O. Stallard, W. Fenical, and J. S. Kittredge, Tetrahedron, 34 (1978) 2077-2081. M. Tori, M. Sono, K. Nakashima, Y. Nakaki, and Y. Asakawa, /. Chem. Soc, Perkin Trans 1, (1991) 447-450. A. D. Wright, G. M. Konig, and O. Sticher, /. Nat. Products, 54 (1991) 1025-1033; idem., Phytochemical Analysis, 3 (1992) 73-79. M. Tori, K. Nakashima, M. Seike, Y. Asakawa, A. D. Wright, G. M. Konig, and O. Sticher, Tetrahedron Lett., 35 (1994) 3105-3106. C. -L. Wu and C. -L. Chen, Phytochemistry, 31, (1992) 4213-4217. C. -L. Wu, S. -J. Chan, M. Tori, H. Furuta, A. Sumida and Y. Asakawa, /. Chi. Chem. Soc. (Taipei), 37 (1990) 387-391. M. Tori, K. Kosaka, and Y, Asakawa, J. Chem. Soc. Perkin Trans. 1, (1994)2039-2041. M. Pfau, G. Revial, A. Guigant and J. d'Angelo, /. Amer. Chem. Soc, 107 (1985) 273-274. J. d'Angelo, D. Desmaele, F. Dumas and A. Guingant, Tetrahedron: Asymmetry, 3 (1992) 459-505. M. Ishibashi, Y. Ohizumi, J. -F. Chen, H. Nakamura, Y. Hirata, T. Sasaki and J. Kobayashi, /. Org. Chem., 53 (1988) 2855-2858; J. Kobayashi, K. Naitoh, T. Sasaki and S. Shigemori, /. Org. Chem., SI (1992) 5773-5776. M. Tori, Rev. Latinoamer. Quim., 22/3 (1991) 73-83.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
649
Bioactive Gymnemic Acids and Congeners from Gymnema sylvestre Shashi B. Mahato
1.
INTRODUCTION R. Br. is a more or less pubescent, woody climber
Gymnema sylvestre
found in the Deccan Peninsula, extending to parts of northern and western India. It is also cultivated occasionally as a medicinal plant. It belongs to the family Asclepiadaceae. The unique property of the leaves of the plant to inhibit temporarily the ability to taste sweet substances has been known in India since antiquity. The plant is popularly known as 'Mesasringi' although it is also called
'Gurmar' to signify its sugar destroying property. The
leaves have been used
in traditional Indian medicine as a stomachic, diuretic 2 3 and for the treatment of cough, throat trouble, eye pain and diabetes.' The plant has been attracting the attention of researchers since the last century and Falconer
reported in 1847 the antisweet activity of the plant. Isolation
of the antisweet principle as an amorphous monobasic acid was reported by Hooper *
who named it gymnemic acid and suggested it to be a glycoside. A
modified method of isolation of gymnemic acid in white crystalline form was reported by Shore acid.
In
o
Pfaffmann i s o l a t e d gymnemic a c i d i n c r y s t a l l i n e 9 mp 199°C ( d e c . ) and Pfaffmann a l s o r e p o r t e d t h e d e t e c t i o n of g l u c o s e ,
form,
1959,
in 1892 who suggested it to be a derivative of anthranilic
arabinose,
Warren and
and
glucuronolactone
h y d r o l y s i s of gymnemic a c i d .
as
sugar
Yackzan
constituents
f o r gymnemic a c i d based upon an u l t r a c e n t r i f u g e Stbcklin four
et a l ,
components,
characterised,
on
acid
technique.
r e p o r t e d i n 1967 t h e s e p a r a t i o n of gymnemic a c i d
gymnemic a c i d s although
these
A -A,. were
These components were n o t suggested
h e x a h y d r o x y o l e a n - 1 2 - e n e named gymnemagenin,
to
be
into
completely
D-glucuronides
acylated with various
of f o r m i c , a c e t i c , n - b u t y r i c , i s o v a l e r i c , and t i g l i c
2.
obtainable
a s s i g n e d a m o l e c u l a r w e i g h t o f a b o u t 600
of
a
combination
acids.
SAPONINS AND SAPOGENINS 2.1 2.1.1
Sapogenins Gymnemagenin and G y m n e s t r o g e n i n .
3 3 5 ° C , t h e major s a p o g e n i n from
G.
sylvestre
Gymnemagenin
(1),
mp
gymnemic a c i d m i x t u r e by / ? - g l u c u r o n i d a s e followed by a l k a l i n e h y d r o l y s i s . structure
was e l u c i d a t e d
328-
was o b t a i n e d by h y d r o l y s i s
of Its
t o be 3/?,16/?, 2 1 ^ , 22a;, 2 3 , 2 8 - h e x a h y d r o x y o l e a n - 1 2 - e n e
650 (1) mainly from the spectroscopic data of the sapogenin as well as i t s d e r i v a t i v e s .11-13 Treatment of gymnemagenin with acetone-CuSO, furnished two 14 diacetonides 2, mp 305-307°C and 3, mp 280-281°C. Rao and Sinsheimer obtained a third diacetonide 4, mp 301-303°C by treatment of 1 with acetone-H^SO, and 15 they also supported the s t r u c t u r e 1 of gymnemagenin. Liu et a l . confirmed the s t r u c t u r e of gymnemagenin by single c r y s t a l X-ray crystallographic analysis of the diacetonide 3. They also prepared the gymnemagenin hexaacetate 5, mp 296300°C by treatment of 1 with a c e t i c anhydride pyridine and N,N-dimethylaminopyridine (DMPA) as c a t a l y s t for lOh a t room temperature. Gymnestrogenin,
the 3y5,16y3 ,21^ ,23,28-pentahydroxyolean-12-en€
isolated from the leaves of G. s y i v e s t r e by Stbcklin. 289°C was also isolated by Sinsheimer and Rao
(6) was
This sapogenin, mp 288-
by hydrolysis of t h e i r gymnemic
acid D with ^-glucuronidase (Helix promatia) preparation. 2.1.2
Gymnemanol and 2 3 ~ h y d r o x y l o n g i s p i n o g e n i n .
Gymnemanol (7) 1Q
TO
mp 284-285°C has recently been isolated from the leaves by Mahato et al. ' alkaline
hydrolysis
followed
by
acid
hydrolysis
by
of the new saponins,
gymnemasins and its structure has been elucidated as 3/3,16/Q ,22oC,23,28-pentahydroxyolean-12-ene (7) by spectroscopic methods. A new type of non-acylated antisweet principles named gymnemasaponins 20 were isolated from the plant by Yoshikawa et al. These saponins are glycosides
of the aglycone, 23-hydroxylongispinogenin
(3/^,16/^ ,23,28-tetra-
hydroxyolean-12-ene)(8) which was previously isolated and characterised by 21 Mahato and Pal from Corchorus acutangulus* 2.1.3 Dammarane derivatives. Besides the isolation of oleanane type triterpene glycosides several new dammarane-type saponins designated gymnemasides together with the known glycosides, gypenosides were isolated from 22 the plant by Yoshikawa et al. A few more gypenosides have also been isolated from the leaves.
The gymnemasides and gypenosides are glycosides of dammarane
derivatives (9-15) . However, the gypenosides have previously been isolated from Makino. The new gymnemasides are glycosides of three
gynostemma
pentaphyllum
dammarane
derivatives, 19-oxo-3/? ,20S-dihydroxydammar-24-ene
,20S,25-pentahydroxydammar-23-ene
(9), 2o<: ,3/5" ,12y5
(10) and 25-hydroperoxy,2 oc ,3/^ ,12/? ,20S-
tetrahydroxydammar-23-ene (11) which were all characterised by spectroscopic methods. 2.2
Saponins
2.2.1 Isolation of crude saponins. Although
the saponins are
the dominating constituents of the leaves of G. svlvestre they are present as a very complex mixture. As already mentioned several workers have tried to isolate
the saponins
methods. A procedure
by the usual for isolation
solvent of
extraction and precipitation
the gymnemic
acids
reported
by
651
24 1
Gymnemagenin
T^o--'^"
CH2OH
6
8
-OH Gymnestrogenln
23-hydroxylongispinogenin
7
Gymnemanol
652
CHgOH
653 Kurihara
23
is described here.
The dried leaves of G. sylvestre
were immersed in water at 60°C for about
5h and filtered through a sheet of gauze. This was repeated several times. The pH of the combined extract was adjusted to 3.0 with 2N sulphuric acid. The precipitates thus formed settled at the bottom of the vessel and most of the supernatant was siphoned out. The precipitates were collected by centrifugation at
5000
rev/min., washed
with
water
and
dissolved
in ethanol.
Insoluble
materials were removed by centrifugation. The ethanol solution was concentrated under
reduced
pressure and
two parts of acetone per one part of alcohol
solution were mixed. Precipitates thus appeared were removed by centrifugation and
the
supernatant
greenish
residue
carbonate at
was
formed
evaporated
to dryness
which
extracted
the boiling
was
point. The gymnemic
under
reduced
several acids
times
pressure. A with
crystallised
diethyl from
the
solvent were further purified by chromatography over DEAE-sephadex A-25 column eluting with 95% ethanol. According to the method of isolation of crude gymnemic acid described by 24 Sinsheimer et al. , dry powdered leaves of G. sylvestre were wrapped in cheese cloth in bundles and were loosely packed in the side chamber of a continuous extractor. Defatting of the leaves was carried out by extraction with petroleum ether (30-60°C) for 18h. and then extracted with either water or 95% ethanol for 18h. The filtered ethanol extract was concentrated and petroleum ether was added until no more precipitation occurred. The precipitate was collected by centrifugation, dissolved in ethanol and plated on appropriate amount of silicic acid. The dry plated material was extracted in a continuous extractor, first with chloroform and then with ethyl acetate, each for 18h. Evaporation of the ethyl acetate extract to dryness afforded a light green residue which showed the presence of gymnemic acids when examined by TLC. 25 Yoshikawa et al. isolated the crude saponin from G. s-yivestrcras follows: The dried leaves were extracted with 60% EtOH at 60°C for two weeks. The EtOH extract obtained after evaporation of the solvent ijn vacuo was passed through an Amberlite XAD-2 column and eluted with MeOH. The MeOH eluate was psssed through Toyopearl HW-40 column saponin
was
further
(MeOH) to give the crude saponin. The crude
fractionated
by
chromatography
on
Servachrome
XAD-2
(elution with 40-70% MeOH). 19 According to the procedure adopted by Mahato et al. the crude saponin was Isolated as follows : The air dried leaves of G. sylvestre
were powdered,
defatted with petroleum ether (60-80°C) and extracted with hot 50% EtOH at 60°C. The
solvent
was
removed
under reduced
pressure and
the extract was
partitioned between water and £-butanol. The residue obtained from £-butanol layer after evaporating off the solvent was passed through Amberlite XAD-2
654 column and eluated with methanol which was further passed through sephadex-LH-20 and eluted with MeOH. The crude saponin was obtained by chromatography on silica gel column using CHCl^-MeOH (75:15) as the mobile phase. 2.2.2 separated
Isolation of Individual saponins. The into
pure
saponin
constituents
by
crude
Yoshikawa
saponin was 25 et al. by
chromatography on silica gel with AcOEt : MeOH : H^O (10:3:1), CHCl^ : MeOH : H^O (65 :35:10, lower layer) followed by HPLC on ODS column using 23-A0% CH^CN 15 as the mobile phase. Liu et al. chromatographed the crude saponin on an ODSRP 18 column using MeOH-H 0 1:1
>7:3
>1:0 to obtain 11 fractions. The
different fractions were then separated by recycling LC using MeOH
: 0.8%
(NH^)2C02 (pH 8) (65:35), MeOH : 0.25% KH^PO^ (pH 3) (2:1), MeOH
: 0.25%
KH2PO4 (pH 3) (57:43) and MeOH : 1.0% (NH^)2C0^ (3:2) as the mobile phases. 2.2.3 gymnemic
Analysis of saponins.
acids
Sinsheimer
and
by TLC on silica gel plates. They
Rao
employed
detected seven solvent
systems : (a) chloroform-formic acid-methanol (4:1:1) (aged for 3h at 27°C), (b)
chloroform-acetic
acid-methanol
(5:1:1),
(c)
methanol-t-butanol (4:1:1:1) (aged for 3h at ll^C), hydroxide-chloroform-t-butanol
chloroform-formic
acid-
(d) isopropanol-ammonium
(5:2:1:1), (e) isopropanol-ammonium hydroxide-
isoamyl alcohol (3:2:1), (f) isopropanol-ammonium hydroxide-diethyl carbonateisoamyl
alcohol-t-butanol
ketone-formic
acid-water
(3:2:2:1:1), (5:3:1:1).
and
Spray
(g) butyl reagents
formate-methyl used
for
ethyl
visualizing
components on chromatograms were : benzoyl chloride-sulphuric acid reagent, eerie
sulphate-sulphuric
acid
reagent
and
modified
Liebermann-Burchard
reagent. Precoated TLC plates GF254 (Merck) are frequently used and spots are detected by spraying 5% H SO, and heating at 100°C until colouration takes 22 place. Yoshikawa et al. observed that the saponin extracts from the leaves of G. syivestre the
when subjected to TLC on silica gel and sprayed with 30% H«SO,,
dammarane-type
saponins
stained
pink
and
violet,
and
oleanane-type
saponins stained blue and violet. The molecular masses of gymnemic acids and congeners were determined by 26 Imoto et al. by high performance liquid chromatography combined with atmospheric pressure ionisation mass spectrometry (API-MS). The crude saponin isolated from the leaves was chromatographed on octadecyl silica column and eluted with an aqueous methanol
solution containing ammonium acetate. The
fractions thus separated were directly introduced into an atmospheric pressure ionisation mass spectrometer connected with the liquid chromatograph by an interface
consisting
of a nebulizer and a vaporizer through a PTFE tube
(Hitachi, Japan). The vaporized sample and solvent molecules at 300°C were introduced into the ion source of the atmospheric pressure ionisation system.
655 The drift voltage of the spectrometer was set at 160V. Quasimolecular ions of gymnemic acids were detected as ammonium adduct ions and/or proton adduct ions. Molecular masses of
thirteen gymnemic acids and
five compounds not
containing glucuronic acid in the molecules were determined. Three pairs of geometrical isomers of gymnemic acids were also detected. Moreover, the acyl residues such as acetyl, tigloyl, 2-methylbutyroyl and benzoyl in gymnemic acids were identified by interpretation of the fragmentation patterns. 2.2.3
Structure Eludication.
investigated methods.
for
The
their
chemical
The
isolated
structures
by
pure
chemical
saponins
and
are
spectroscopic
molecular
masses are conveniently determined by fast-atom27 28 spectrometry ' (FAB-MS) using positive and/or negative
bombardment mass 1 13 H and C NMR spectroscopic analyses employing recent signal assignment
mode.
techniques are conducted for structure elucidation. The points of attachment of sugars and acyl groups are determined by the application of glycosylation 29-31 32 33 shift
and
experiments
acylation
are
shift
performed
to
'
rules. Acid
liberate
the
and
alkaline
sugars, acyl
hydrolysis
constituents
and
aglycones which are separately investigated for characterisation. The sugar and
acyl
constituents
are
identified
by
GC
analysis
of
their
suitable
derivatives and the aglycones are characterised by spectroscopic methods. /3 glucuronidase from
Helix
has also been used for the liberation of
promatia
gymnemagenin from deacylgymnemic acid. Acid hydrolysis of gymnemic acids was carried out by Yoshikawa et a l P in 5% H SO^ in 50% EtOH at 100°C for 3h. The reaction mixture was extracted with ether. From the organic layer the aglycone was
isolated
and
characterised
in the usual way. The aqueous layer was
neutralised with Amberlite IR-4 and evaporated in vacuo to dryness. The sugars were
identified
using
refractive
index detection and
chiral detection by
comparison with authentic samples. Alkaline hydrolysis was done in 50% 1,4dioxane and 10% KOH at 37°C for Ih. The reaction mixture was adjusted to pH A.O with 5% HCl and extracted with ethylene chloride. The organic acids were identified by HPLC by comparison with authentic samples.
25 The structure elucidation o-f gymnemic acid XII (27) by Yoshikawa et al. by a combination of chemical and spectroscopic methods may be mentioned as an example. The negative FAB-MS of the saponin showed the [M-H]
at m/z 967
suggesting its•molecular weight to be 968. It furnished gymnemagenin (1) as the aglycone and D-glucose and D-glucuronic acid in the ratio 1:1 as sugar 1 13 components on acid hydrolysis. H and C NMR spectra of 27 indicated the presence of one
/3 -glucopyranosyl unit, one
/?-glucuronopyranosyl unit, one
acetyl unit and one tigloyl unit. On alkaline hydrolysis it furnished acetic acid and tiglic acid as acyl components. The
H NMR spectrum of 27 displayed
two acylation shifts for the 21-H (+0.89 ppm) and 2S-^^
(+0.54 ppm) when
656 compared
with
deacylgymnemic
acid
(30).
A
long-range
selective
proton
decoupling experiment revealed t h a t 21-H coupled to carbonyl carbon of acid
and
presence
28-H« of
a
coupled tigloyl
to
carbonyl
group
at
carbon
C-21 and
of
acetic
acetyl
acid
tiglic
suggesting
group a t
C-28.
the
Cellulase
treatment of 27 afforded gymnemic acid I (16) as a prosapogenin. Comparison of 13 C NMR data of 27 with those of 16 disclosed C-3 of glucuronic acid as a
the
glycosylation s i t e in the former and thus the s t r u c t u r e of 27 was elucidated. 2.2.4
Chemical
Yoshikawa e t a l . ' gymnemic a c i d s VII
(22) contain
contains the
'
I-XII
isolated
Physical
i n pure s t a t e s
gymnemagenin
(1) as
Data.
t w e l v e a c i d i c s a p o n i n s named
the aglycone while
gymnestrogenin
gymnemic a c i d
(6).
V I I (GA-VII)
a g l y c o n e may be d e s i g n a t e d d i f f e r e n t l y VII)
and S p e c t r o s c o p i c
( 1 6 - 2 7 ) . A l l of t h e gymnemic a c i d s e x c e p t gymnemic a c i d
aglycone,
Yoshikawa's
Structures,
i s proposed t o avoid c o n f u s i o n .
As
gymnemic a c i d
suggested
containing
by
Liu
gymnestrerenin
et
VII al.
(6) as the
and t h e name g y m n e s t r o i c a c i d V I I (GT-
Maeda e t a l .
r e p o r t e d t h e i s o l a t i o n and
s t r u c t u r e e l u c i d a t i o n of c h r o m a t o g r a p h i c a l l y p u r e GA-I and GA-II which seemed 37 t o be i d e n t i c a l w i t h Yoshikawa's GA-IV and GA-III r e s p e c t i v e l y . Kiuchi e t a l . and Liu e t a l . two
reported
new gymnemic
However,
acids
to avoid
further which
confusion
t h e i s o l a t i o n and s t r u c t u r e
they
designatd
gymnemic
acids
e l u c i d a t i o n of VIII
and IX.
w i t h Yoshikawa's gymnemic a c i d V I I I and IX t h e s e
two s a p o n i n s may be c a l l e d gymnemic a c i d s X I I I ( 2 8 ) and XIV ( 2 9 ) r e s p e c t i v e l y . 25 34 35 The p h y s i c a l d a t a r e p o r t e d f o r gymnemic a c i d s I - X I I
(16-27)
'
mic a c i d s X I I I ( 2 8 ) and XIV (29)"^^ a r e : 16, mp 211-212''C, [ oC ] MeOH); 17, mp 212-213°C, 7.6°
l^]j^
[oC]^
I oC ]^ + 1 1 . 7 ° (c 1 . 1 ,
[oC ]^ + 9 . 6 ° (c 5 . 7 , MeOH); 2 3 , mp 185-187°C, [oC]^ +
2 1 . 5 ° ( c 3 . 5 , MeOH); 24, mp 194-196°C, 212°C,
+ 8 . 8 ° (c 5 . 4 , MeOH); 20, mp 2 0 2 -
[o^ ]^ + 2 . 2 ° (c 3 . 6 , MeOH); 2 1 , mp 225-226°C,
MeOH); 2 2 , mp 222-223°C,
and gymne-
+ 36.7°(c 2 . 4 ,
[ oC ]^ + 3 6 . 3 ° ( c 1 . 5 , MeOH); 1 8 , mp 218-219*0, [oC ]^ -^
(c 2 . 9 , MeOH); 19, mp 220-221°C,
203°C,
'
[oC]^ + 7 . 6 ° (c 1 . 8 , MeOH); 2 5 , mp 2 1 0 -
+ 1 4 . 9 ° (c 2 . 3 , MeOH); 26, mp 190-192°C,
I oC ]j^ + 1.7° (c 5 . 3 ,
MeOH); 2 7 , mp 209-211°C, [oC]^ + 1 1 . 7 ° ( c 3 . 6 , MeOH); 28, mp 218-220°C, [oC] + 1 7 . 3 ° (c 0 . 7 4 , MeOH); 29, mp 222-224°C, I oC ]j. -^ 1 1 . 4 ° (c 0 . 7 0 , MeOH). The ^H 13 and C NMR d a t a a r e shown i n T a b l e s 1 - 5 . Besides
the
non-acylated from
the
These sugar usual 31,
hot
are
gymnemic
saponins, aq.
5%
EtOH
glycosides
constituents 3-position mp
acids
of
appear
are
acylated
I-V
(31-35)
extract
of
the
saponins, have
leaves
at
23-
[ cC ]^
(c 2 . 6 , MeOH);
-^
and
28-positions
The p h y s i c a l 9.3°
(c
3.5,
3 3 , 203-205°C,
five
been of
23-hydroxylongispinogenin
of t h e a g l y c o n e .
184-185°C,
[ c?C ] ^ + 1.9°
which
gymnemasaponins
the (8)
instead
new
isolated 20 plant. and of
the the
d a t a of t h e s a p o n i n s a r e : MeOH);
32,
mp
190-192°C,
[ oC ]^ - 1 1 . 6 ° (c 1 . 1 , MeOH);
657
-0R4 CH2OR2
^2
Gymnemic acid II
^1 tig mba
18
Gymnemic acid III
mba
19
Gymnemic acid IV
20
Gymnemic acid V
21 23
Ac
•^3 H
H
Ac
H
H
H
H
H
tig
H
H
H
tig
H
H
tig
Gymnemic acid VI
tig
H
/5-glc
H
Gymnemic acid VIII
H
mba
H
H
24
Gymnemic acid IX
H
tig
H
H
25
Gymnemic acid X
H
Ac
H
H
26
Gymnemic acid XI
tig
tig
H
H
27
Gymnemic acid XII
tig
Ac
,/^-glc
H
28
Gymnemic acid XIII
mba
H
OG
H
29
Gymnemic acid XIV
tig
H
OG
H
30
Deacylgymnemic acid
H
H
H
H
16
Gymnemic acid I
17
tig;
H .CH. H3C, ^C = C , mba: C H ^ C H ^ - C - C O , H 0 CH3 TigloyI
(S)-2-Methylbutyryl
y^-glc .7^-Glucopyranosyl, OG!/?-orobino-2-Hexulopyranosyl
658 TABLE 1 •""H NMR d a t a
( 6 ) of gymnemic a c i d s I-VI ( 1 6 - 2 1 ) and g y m n e s t r o i c a c i d VII (22)
i n C-D^N H
16^^
17^^
18^
I9I5
20^5
3
4.32 4.34 4.34 .(dd,ll.5,4.5) (dd,12.0,4.5) (dd,11.0,4.0)
12
5.32 (brt)
5.34 (brt)
21^5
2235
5.38 (brt)
16
5.14 5.10 5.09 5.10 5.08 5.12 4.68 (dd,ll.5,5.0) (dd,10.0,5.0) (dd,11.0,5.0) (dd,ll.5,5.0) (dd,11.0,5.0) (dd,11.0,5.0)(dd,ll.0,5.0)
21
5.74 (d,10.5)
5.68 (d,10.5)
5.68 (d,11.0)
5.80 (d,11.0)
5.83 (d,11.5)
5.79 (d,10.5)
4.14 (dd,13.0,4.5)
22
4.59 (d,10.5)
4.54 (d,10.5)
4.95 (d,11.0)
5.08 (d,11.0)
6.35 (d,11.5)
5.03 (d,10.5)
2.08 (dd,13.0,13.0) 3.24 (dd,13.0,4.5)
23
3.71,4.37 3.70,4.36 3.69,4.35 3.68,4.32 3.70,4.34 3.72 (d each,11.0) (d each,11.0) (d each,11.0) (d each,11.0) (d each,11.0) (d,]0.5)
3.70,4.35 (d each,]1.0)
28
4.65,5.08 4.61,5.02 4.04,4.65 4.08,4.71 4.03,4.24 4.72 (d each,11.0) (d each,11.0) (d each,11.0) (d each,11.0) (d each,11.0) (d,11.0)
3.74,4.41 (d each,10.5)
Anoneric 5.29 (d,7.5)
5.27 (d,8.0)
5.22 (d,7.5)
5.21 (d,7.5)
5.23 (d,7.5)
5.26 (d,8.0) 5.33 (d,8.0)
Acyl group 1.64 (d,6.5)
0.98 (t,7.0)
0.99 (t,7.0)
1.66 (d,7.0)
1.48 (d,7.0)
1.63 (d,7.5)
1.91 (s)
1.26 (d,7.0)
1.23 (d,7.0)
1.88 (s)
1.62 (d,7
1.89
7.07 q,6.5)
1.54 (q,7.0)
1.52(m)
7.01 (q,7.0)
1.84 (s)
7.02 (q,7.5)
2.59 (sex,7.0)
182(m)
1.89 (s)
1.97 (s)
256(m)
7.03 (a.7
5.25 (d,7.5)
659 TABLE 2 •'"H NMR d a t a ( 6 ) of gymnemic a c i d s VIII-XIV ( 2 3 - 2 9 ) i n C D^N/D20 25 23^"^
H
25 2A^^
25 25"^^
25 26^^
25 IT^
15 28^
3
15 29^^^
4.28 4.28 (dd,11.0,4.0) (dd,ll.0,4.0)
12 16
5.35 (brt)
5.41 (bit)
5.06 5.08 5.05 5.14 5.13 5.08 5.10 (dd,12.4,5.0) (dd,12.4,5.0) (dd,12.4,5.0)(dd,12.4,5.0)(dd,12.4,5.0)(ddA1.0,5.0) (dd,ll.0,5.0)
21
4.04 (d,10.2)
22
4.46 (d,10.2)
23
3.70,4.35 3.70,4.37 3.70,4.36 3.70,4.35 3.71,4.34 3.71,4.32 3.68,4.29 (d each,11.0) (d each,11.0)(d each,10.3)(deach,10.2)(deach,11.0)(d each,11.0)(d each, 11.0)
28
4.67,5.09 4.66,5.09 4.62,5.15 4.65,5.20 4.63,5.05 4.05,4.67 4.08,4.69 (d each,11.0) (d each,11.0)(deach,11.0)(d each,11.0)(d each,11.0)(d each,11.0)(d each, 11.0)
Anomeric 5.24 (d,8.0)
4.07 (d,10.3) 4.57 (d,10.3)
5.26 (d,8.0)
4.03 (d,10.0)
5.80 (d,10.2)
5.78 (d,10.2)
5.69 (d,11.0)
5.76 (d,11.0)
4.45 (d,10.0)
4.65 (d,10.2)
4.60 (d,10.2)
4.95 (d,11.0)
5.00 (d,11.0)
5.25 (d,7.3)
Acyl group 1.94 (s)
5.26 (d,7.4)
5.23 (d,7.3)
5.20 (d,7.2)
5.19 (d,7.2)
5.31 (d,7.3)
5.28 (s)
5.26 (s)
tig (H21)
Acetyl
1.61 (d,7.0)
2.02 (s)
0.99 (t,7.0)
1.66 (d,7.0)
0.83 (t,7.0)
1.51 (d,7.0)
1.07 (d,7.0)
1.78 (s)
1.88 (s)
tig
1.23 (t,7.0)
1.89 (s)
1 44 (q,7.0)
6.97 (q,7.0)
7.05 (q,7.0)
1.61 (d,7.0)
1.52 (m)
7.02 (q,7.0)
1.67 (q,7.0)
t i g (H28)
1.88 (s)
1.82 (m)
2.39 (Six,7.0)
1.59 (q,7.0)
7.05 (q,7.0)
2.56 (m)
1.82 (s) 6.47 (q,7.0) 34, mp 201-203°C, \d. ] ^ - 1.1° (c 1.9, MeOH); 35, mp 186-188°C, [ CTC ]j^ - 6.2° (c 1.9, MeOH). The ^^C NMR data are shown In Table 6. Four new saponins, gymnemasins A-D (36-39) which are glucuronides of the 18 19 sapogenol, gymnemanol (7) have been isolated recently. ' The physical data of the saponins are : 36, mp 215-217°C, { d \^
^
15° (c 1.5, MeOH) : "^H NMR
8 (C D N) 0.88 (3H,s), 0.92 (3H,s), 0.94 (3H,s), 1.04 (3H,s), 1.14 (3H,s), 1.31 (3H,s), 1.62 (3H,d,J 7Hz), 1.88 (3H,s), 7.02 (lH,q,J 7Hz); ''•^C NMR (Table 7).
660 TABLE 3 13
spectral C NMF ( 1 ) i n C^D^N
Carbon
""7^
data
1625
C-1
38.9
39.0
C-2
27.7
26.3
(6 va l u e s )
of gymnemic ac ids I-V (16-20) and gymnemagehin
T ^ "7?^ 19^5 26.1
20^5 Carbon
16
17
18
19
20
38.7
38.7
38.8 g u - 1
106.3 106.3 106.1 105.5 106.1
26.0
26.0
2 6 . 0 gu-2
75.5 7 5 . 5 7 5 . 4 7 5 . 1 75.4
C-3
73.8
82.3
81.9
81.8
81.0
8 1 . 8 gu-3
7 8 . 1 7 8 . 1 7 8 . 1 78.2 7 8 . 1
C-A
42.9
43.7
43.5
42.6
42.6
42.7 gii-4
73.5 7 3 . 4 7 3 . 4 7 3 . 5 7 3 . 4
C-5
48.5
47.5
47.4
47.4
47.4 gu-5
77.8 7 8 . 0 7 7 . 8 77.4 77.6
C-6
18.5
18.2
18.0
18.1
1 8 . 1 gu-6
173.1 172.9 172.9 173.8 172.9
C-7
32.7
32.7
32.5
32.6
3 2 . 5 g-1
C-8
40.3
40.5
40.2
40.2
4 0 . 3 g-2
C-9
47.3
47.3
47.1
47.1
C-10
37.0
36.9
36.6
36.7
4 7 . 1 g-3 36.7 g-4
C-11
24.0
24.2
23.9
23.9
23.9 g-5
C-12
123.9
124.8
123.9
123.9
124.2 g-6
C-13
142.8
141.5
142.2
142.3
141.5 mba-1
176.6 176.6 168.2 167.6
C-IA
42.7
42.8
43.5
43.5
43.5 mba-2
42.1 42.0
129.7 128.9
C-15
36.0
36.4
36.4
36.3
36.2
36.8 mba-3
27.3
136.4 137.6
C-16
67.8
67.7
67.5
68.0
68.0
67.0 mba-4
12.1 12.0
C-17
46.6
45.9
45.6
47.0
47.1
48.0 mba-5
17.2
C-18
42.2
42.7
42.0
42.0
42.7
tig-1
168.5
C-19
46.7
45.3
46.2
46.2
45.9
tig-2
129.7
C-20
36.8
36.9
36.4
36.4
36.6
36.7
tiff-3
137.3
C-21
77.3
79.1
78.4
79.1
79.6
76.6
ti^4
12.7
C-22
73.3
71.7
71.4
71.2
71.2
74.6
tig-5
14.6
C-23
68.3
64.4
64.3
64.4
64.4
64.4 ac-1
171.4 170.9
C-2A
13.1
13.9
13.6
13.6
13.6
13.6 ac-2
21.1 20.7
C-25
16.1
16.5
16.2
16.2
16.2
tig-1^
167.8
C-26
17.1
17.3
17.0
17.0
16.9
tir2^
128.9
C-27
27.4
27.7
27.4
27.4
27.5 tig-3^
138.0
C-28
58.6
62.6
58.1
58.1
59.9
tig-4
12.2
C-29
30.4
29.6
29.6
29.6
29.2
tig-5
14.2
C-30
19.1
20.0
19.8
19.9
19.8
62.3
27.2
12.4
12.2
17.1 17.1
14.1
37, mp 221-222°C, [oC ]^ + 18.5° (c 1.2, MeOH) : -^H NMR 5 (C^D^N) 0.86 (3H,s), 0.90
(3H,s),
0.94
(3H,s), 1.04
(3H,s),
1.12
(3H,s), 1.30
(3H,s): "^^C NMR
(Table 7 ) ; 38, mp 212-214°C, [oC ]^ + 12.5° (c 1.0, MeOH), -^B. NMR 6 (C^D N) 0.88 (3H,s), 0.90 (3H,s), 0.93 (3H,s), 1.02 (3H,s), 1.28 (3H,s), 1.30 (3H,s), 1.64 (3H,s), 1.86 (3H,s), 7.01 (IHq, J 7Hz), ^^C NMR (Table 7); 39, mp 220-221°C,
661 TABLE 4 •^^C NMR speci t r a l d a t a ( 6 v a l u e s ) of gymnemic a c i d s V I ( 2 1 ) , V I I I ( 2 3 )•, I X ( 2 4 ) , g y m n e s t r o i c iacid VTI(22), deacylgymnemic a c i d (30) and g y m n e s t r o g e n i n ( 6 ) i n C3D3N Carbon
3*^
6^^ 38.7
2r'
^1^
23^^
JK
38.9
Carbon 30
21
22
23
24
38.6
gu-1
106,0 106.0 106.3 106.3 106.3
C-2
27.7
25.9
26.1
26.2
26.1
25.8
gu-2
75.3 74.3
75.5 75.4
75.1
C-3
73.4
82.0
81.9
82.1
81.6
816
gu-3
78.0 87.6
78.2
78.1
78.0
73.3
71.7
73.5 73.4
73.2
77.5
77.9
77.9
78.0
C-1
C-4
42.6
43.5
43.4
gu-4
C-5
47.4
47.4
47.3
gu-5
77.7
C-6
18.0
18.1
18.1
gu-6
172.7 172.1 172.9 173.0 173.0
C-7
32.6
32.6
32.3
g-1
106.0
C-8
40.2
40.3
40.3
g-2
75.6
C-9
47.1
47.2
47.2
g-3
78.8
C-10
36.7
36.6
36.6
g-4
71.6
C-U
23.8
24.0
23.7
g-5
78.3
C-12
123.9
124.0
124.0
g-6
62.5
C"13
142.7
141.9
141.9
mba-1
176.2 168.0
42.5
42.5
mba-2
41.8
C-15
36.8
35.9
36.3
36.9
36.2
36.0
mba-3
27.1 136.9
C-16
67.7
68.2
68.0
67.8
67.8
67.7
mba-4
11.9
14.1
C-17
43.7
46.5
47.2
43.7
45.5
45.6
niba-5
17.0
12.2
43.4
C-14
C-18
42.1
42.7
tig-1
168.2
C-19
46.6
46.2
46.2
tig-2
129.7
C-20
37.0
36.6
36.7
36.9
36.8
36.7
tig-3
136.5
C-21
72.7
77.2
79.7
72.9
76.9
76.7
tig-4
12.4
C-22
35.1
73.7
71.2
35.1
74.0
73.9
tig-5
14.2
C-23
67.9
64.4
64.0
64.5
64.4
64.0
ac-1
C-24
13.1
13.5
13.6
13.7
13.6
13.6
ac-2
C-25
16.2
16.3
16.2
C-26
17.0
17.2
17.2
C-27
27.4
27.5
27.5
C-28 C-29 C-30
68.5
43.0
62.3
62.4
30.3
30.2
30.2
18.9
19.0
18.9
58.5
58.1
68.5
129.2
oC ] + 8° (c 0.9, MeOH), -^H NMR 6 (C^D^N) 0.86 (3H,s), 0.89 (3H,s), 0.93 (3H,s), 1.04 (3H,s), 1.14 (3H,s), 1.30 (3H,s),
^^E NMR (Table 7 ) .
All the saponins mentioned so far contain aglycones which are oleanane22 type triterpenes. Yoshikawa et al. isolated from the leaves of G. sylvestre
662 TABLE 5 13 spectral C NMR
d a t a ( 6 v a l u e s ) of gymnemi c a c i d s X-XIV 1(25-29) i n C5D5N
1 ^ 1^ 1?^
Carbon "^5
26^^
C-1
38.8
38.8
38.8
38.6
C-2
26.1
26.1
26.1
26.0
C-3
81.9
81.9
81.7
C-A
43.5
43.6
C-5
A7.A
C-6
18.1
C-7 C-8 C-9
47.2
C-10
36.6
C-11
Ca rbon
25
26
27
28
38.6
gu-1
106.3
106.4
105.9
106.3 105.9
25.9
gu-2
75.5
75.5
74.2
72.1
71.8
82.1
81.8
gu-3
78.1
78.1
87.5
73.8
73.7
43.5
42.6
42.5
gu-4
73.4
73.5
71.7
69.6
69.5
47.4
47.4
47.3
47.2
gu-5
77.9
78.0
77.3
75.2
75.0
18.0
18.0
18.0
18.0
gu-6
172.9
172.9
172.9
171.5 172.2
32.6
32.5
32.5
32.5
32.4
g-1
105.8
97.0
96.8
40.3
40.3
40.2
40.2
40.2
g-2
75.6
93.8
93.7
47.2
47.1
47.1
47.1
g-3
78.2
79.7
79.4
36.7
36.5
36.6
36.6
g-4
71.6
69.5
69.3
24.0
24.0
24.0
23.8
23.8
g-5
78.8
79.4
78.9
C-12
124.2
124.2
124.6
123.9
123.9
g-6
62.3
62.8
62.5
C-13
141.9
141.4
141.2
142.2
142.2
mba-1
176.6
C-14
42.8
42.6
42.4
43.5
43.4
mba-2
42.0
C-15
36.1
36.4
36.3
36.2
35.9
niba-3
27.2
C-16
67.7
67.6
67.4
68.0
68.0
mba~4
12.0
C-17
45.3
46.1
46.7
47.0
47.0
mba-5
17.1
C-18
42.5
42.7
42.6
42.0
41.9
tig-1
168.2
168.5
168.5
C-19
46.2
45.8
45.7
46.2
46.2
tig-2
129.6
129.7
129.6
C-20
36.6
36.7
36.6
36.4
36.5
tig-3
137.4
137.3
136.8
C-21
76.8
78.9
79.0
79.0
79.6
tig-4
14.3
14.6
12.4
C-22
73.8
71.6
71.6
71.2
71.1
tig-5
12.6
12.7
14.2
C-23
64.4
64.3
63.8
64.1
64.0
ac-1
20.7
C-24
13.6
13.7
13.6
13.6
13.5
ac-2
171.C)
20.7 170.9
C-25
16.3
16.3
16.2
16.1
16.1
tig-1'
168.0
C-26
17.1
17.2
17.0
17.0
16.9
tig-2'
129.2
tig-3
137.0
C-27
27.5
27.5
27.4
27.4
27.4
C-28
62.6
62.5
62.4
58.1
58.0
tig-4'
14.3
tir5'
12.4
C-29
30.2
29.5
29.3
29.6
29.5
C-30
19.0
19.8
19.2
19.8
19.8
seven
new
together
dammarane-type
with
known
triterpene
saponins,
saponins, gypenoside XXVIII
(49)^^, LXII (50) and LXIII (51)^^
29
gymnemasides (47)^^
I-VII
XXXVII
UO-46)
(48)
, LV
The gypenosides were previously isolated
from the aerial parts of Gynostemma pentaphylJum
Makino. The other dammarane-
type
by
triterpene
saponins
which
were
isolated
Yoshikawa
et
al.
from
663 TABLE 6 13 C NMR s p e c t r a l d a t a l o n g i s p i n o g e n i n (8) i n Carbon 8
( 6 v a l u e s ) of gymnemasaponins I-V (31-35)^ and 23-hydroxy31
32
33
34
35
C-2
27. 6
27.7
27.4
27.4
27.4
27.4
C-3
73 4
73.4
72.3
72.3
72.2
72.1
C-A
42 8
42.9
42.9
42.9
42.0
43.0
C-15
36 7
37.0
36.9
36.9
36.9
36.9
C-16
66 6
66.3
66.1
66.3
66.2
66.3
C-17
41 0
41.3
41.3
41.4
41.3
41.4
C-23
68 .0
67.9
75.0
75.1
75.1
75.1
C-2A
13 .0
13.1
13.2
13.2
13.2
13.3
C-28
68 .9
78.0
78.0
77.9
78.0
77.9
105.2
105.2
105.3
105.4
g-2
75.3
75.3
75.2
75.2
g~3
78.7
78.7
78.7
78.6
g-4
71.8
71.9
71.7
71.6
C-23
sugars
g-1
g-5
78.4
78.4
77.2
77.2
g-6
63.0
63.0
70.2
70.2
g-1'
104.9
104.9
g-2'
74.6
74.6
g-3'
78.4
78.4
g-4'
71.7
71.6
g-5'
78.4
78.4
g-6'
62.8
62.7
105.9
C-28 s u g a r s g-1
105.8
105.8
105.7
105.9
g-2
75.0
75.0
75.1
75.1
75.2 78.5
g-3
78.7
78.7
78.6
78.7
g-A
71.7
71.7
71.6
71.7
71.6
g-3
78.6
78.7
77.3
78.6
77.3
g-6
62.8
62.8
69.8
62.8
69.8
g-l'
105.4
105.4
g-2'
74.8
74.8
g-3'
78.4
78.4
g-4'
71.4
71.3
g-5'
78.4
78.4
g-6'
62.8
62.7
664
CH2OH
22
Gymnemic acid VII
R2
1 31
Gymnemasaponin I
H
glc
32
Gymnemasaponin II
glc
glc
33
Gymnemasaponin III
glc
34
Gymnemasaponin IV
glc —
glc
glc
35
Gymnemasaponin V
glc
glc
glc —
glc ^ g l c glc
^-ORi
cf>o y
^
y
c ^
AiLoH
2^
h
h
35
Gymenmasin A
tig
glc
37
Gymnemasin B
H
38
Gymnemasin C
tig
H
39
Gymnemasin D
H
H
glc
665
R,0
40
Gymnemaside I
glc
"2 glc
41
Gymnemaside II
42
Gymnemaside III
— glc 2 ara — glc
glc
43
Gymnemaside IV
glc
1 6— x y l glc
'^l glc
44
Gymnemaside V
47
Gypenoside XXXVII
, 2 glc — glc 2 ara — glc
48
Gypenoside XXVIII
glc
— glc
rham—^glcOvJ
\ H 45
Gymnemaside VI
46
Gymnemaside VII
glc
glc —
xyl
glc —
xyl
H
666 TABLE 7 13 spectral data (6 values) of gymnemasin A (36), gymnemasin B (37) , gymneC NMR (38), gymnema!sin D (39) and gymnemanol (7) measured in C^DcN masin C Carbon
36
37
3^"^
39
7
Carbon
36
37
38
~^9
1
39.0
39.2
39.0
39.1
38.5 gu-1
105.3
105.3
105.4
105.5
2
26.2
26.1
26.2
26.2
27.5 gu-2
74.3
74.3
75.4
75.5
73.7 gu-3
87.2
87.3
78.2
78.3
3
82.0
82.1
82.1
82.2
4
43.5
43.5
43.4
43.5
42.8 gu-4
71.8^
71.8^
73.2
73.2
5
47.9
47.8
47.8
47.7
48.8 gu-5
77.5
77.5
77.8
77.9
6 7
18.5
18.6
18.7
18.7
18.7 gu-6
173.2
173.3
173.2
173.3
32.7
32.7
32.8
32.6
32.8 g-1
105.5
105.5
8
40.2
40.1
40.3
40.2
40.0 g-2
75.5
75.4
9
47.2
47.4
47.5
47.6
47.3 g-3
78.4^
78.4^
10
37.2
37.2
37.3
37.1
37.1 g-A
71.4^
71.4^
11
24.2
24.1
24.1
24.1
23.8 g-5
78.2^
78.3^
12
123.7
123.6
123.8
123.7
123.6 g-6
62.1
62.2
13
143.0
143.2
143.4
143.4
143.3 tig-1
167.9
14
42.5
42.6
42.7
43.8
42.7 tig-2
129.3
129.2
15
36.2
36.1
36.3
36.2
36.1 tig-3
137.2
137.3
16
67.6
67.6
67.7
67.8
67.7 tig-4
12.4
12.4
17
47.2
46.7
47.3
46.8
46.5 tig-5
14.2
14.1
18
42.6
43.4
42.5
43.4
43.5
19 20 21
46.8
46.7
46.9
46.8
46.6
31.2
31.4
31.3
31.5
31.4
41.8
43.5
41.8
43.4
43.6
22
71.4
70.5
71.5
70.5
70.3
23 24
64.3
64.4
64.4
64.4
68.2
13.6
13.5
13.7
13.6
13.4
25 26
16.3
16.3
16.4
16.4
16.2
17.2
17.1
17.2
17.1
17.0
27
27.2
27.2
27.3
27.2
27.1
28
60.2
59.3
60.3
59.4
59.7
28
60.2
59.3
60.3
59.4
59.7
29
33.4
33.3
33.4
33.5
33.5
30
24.8
24.7
24.9
24.8
24.9
a b ' May be interchanged within the same column. g=glucose; gu=glucuronic acid; tig=tiglovl'
167.8
667 TABLE 8 13 C NMR s p e c t r a l
data
( 5 v a l u e !s) of gymnemasldes I - I V ( 4 0 - 4 3 p in C5D5N
_
Carbon
40
41
42
C-1
33.5
33.4
33.5
C-2
27.7
27.7
C-3
87.7
87.7
C-4
40.1
C-5
Carbon
40
41
33.5
g-1
106,,9
104.9
106.9
27.6
27.7
g-2
75.,7
83.6
75.5
87.6
87.6
g-3
78..7
78.2
78.7
40.1
40.1
40.0
g-4
72..0
72.0
71.8
5A.7
54.6
54.9
54.6
g-5
78..4
78.3
78.4
C-6
17.7
17.7
17.7
17.7
g-6
63..0
C-7
34.7
34.7
34.7
34.7
g-1'
106.2
106.,1
C-8
40.5
40.4
40.4
40.4
g-2'
77.1
76.,5
42
62.9
63.0
C-9
52.9
52.7
52.8
52.8
g-3'
78.2
78.,2
C-10
52.8
52.9
52.8
52.9
g-4'
71.5
7 1 . ,6
C-11
22.5
22.5
22.5
22.5
g-5'
C-12
25.5
25.5
25.5
25.6
g-6'
C-13
42.4
42.3
42.3
42.4
a-1
C-14
50.5
50.5
50.5
50.5
a-2
81. .3
32.0
a-3
73.,5
78.2 62.7
43
78.,1 62.,6 104.,8
C-15
31.9
31.9
31.9
C-16
27.9
27.9
27.9
27.9
a-4
68..4
C-17
48.4
48.4
48.4
48.4
a-5
65..2
C-18
16.0
15.9
15.9
16.0
C-19
205.9
205.8
205.7
C-20
82.2
82.1
82.1
C-20 suga:rs g-1
98,.6
205.9
g-2
75..6
82.3
g-3
79,.1 71,.9
98..6
98.6
75.6
75..6
75.7
79.1
79..1
78.9
71.7
72,.0
71.8
98.6
C-21
21.7
21.7
21.7
21.3
g-4
C-22
40.3
40.3
40.1
40.3
g-5
77,.9
77,8
77,.8
76.6
23.2
g-6
63,.2
63.2
63,.2
70.4
C-23
23.3
23.3
23.3
C-24
126.1
126.1
126.1
126.2
x-1
105.9
C-25
130.8
130.7
130.8
130.7
x-2
74.8
C-26
25.9
25.8
25.8
25.9
x-3
78.0
C-27
17.9
17.9
17.9
18.0
x-4
71.1 67.0
C-28
26.7
26.6
26.7
26.7
x-5
C-29
16.8
16.7
16.7
16.8
r-1
17.2
17.2
17.2
r-2
C-30
17.3
r-3 r-4 r-5 r-6 g=glucose; a=arabinose; x=xylose; r=rhamnose
668 TABLE 9 13 (6 values) C NMR s p e c t r a l d a t a XXXVII (47)22 i n ^^D^N Carbon
44
45
C-1
33.5
48.3
~"46 48.4
of gymnemasides
47
33.4
Ca rbon
V - V I I ( 4 4 - 4 6 ) and g y p e n o s i d e
44
45
46
47
C-3 suga r s 105.0 g-1
C-2
27.7
68.7
68.8
27.6
g-2
83.6
C-3
87.8
83.7
83.7
87.7
g-3
78.2
C-4
40.1
39.9
40.0
40.1
g-4
71.8
C-5
54.7
56.5
56.5
54.7
g-5
78.3
C-6
17.8
18.8
18.8
17.8
g-6
62.8
C-7
34.8
35.0
35.1
34.7
g-1'
106.1
106.1
40.2
40.4
g-2'
77.1
76.4
C-8
40.5
40.1
C-9
52.8
50.3
50.3
52.8
g-3'
78.2
78.1
C-10
52.8
38.6
38.6
52.8
g-V
71.6
71.7
C-11
22.6
30.4
30.6
22.4
g-5'
78.1
78.1
C-12
25.7
70.5
70.3
25.6
g-6'
62.8
62.6
C-13
42.5
49.4
49.7
42.4
a-1
104.9
C-14
50.6
51.5
51.5
50.5
a-2
81.2
C-15
32.1
30.8
31.1
32.0
a-3
73.5
C-16
28.0
26.4
26.4
27.9
a-4
68.4
C-17
48.5
52.0
51.8
48.5
a-5
65.4
C-20 suga r s 98.7
98. 3
98.,2
98.7
g-2
75.5
75.,3
75. 0
75.5
g-3
79.1
79,,0
79.,0
78.9
21.2
g-A
71.7
7 1 . ,3
7 1 . ,6
71.7
40.1
g-5
76.6
76..7
76,,8
76.6
g-6
70.5
66..6
70,,0
70.5
x-1
105.9
105..6
105.9
130.8
x-2
74.8
74..8
74.5
25.1
25.9
x-3
77.9
78..0
77.9
25.4
18.0
x-4
71.1
71, .1
71.7
29.2
26.6
x-5
66.9
67..0
66.9
17.6
16.8
r-1
101,.5
17.5
17.2
r-2
72,.2
r-3
72 .9
C-18
15.9
16.1
16.1
15.9
g-1
C-19
205.9
17.2
17.3
206.1
C-20
82.3
83.3
83.2
82.3
C-21
21.4
23.0
23.1
C-22
40.1
39.9
38.0
C-23
23.3
122.7
126.8
23.2
C-2A
126.2
142.4
140.1
126.2
C-25
130.8
69.9
81.4
C-26
25.9
30.6
C-27
18.1
30.7
C-28
26.7
29.2
C-29
16.9
17.6
C-30
17.3
17.5
r-4
74 .5
r-5
69..3
r-6
18,
669
gIcO
1' X H
R|
^1 H
^2 H
49
Gypenoside LV
50
Gypenoslde LXII
glc
OH
51
Gypenoside LXIII
glc
H
R1 54
Gypenoside XLIII
glc
rham
55
Gypenoside XLV
H
rham
rham
56
57
Gypenoside LXXIV
CH2OH
Gypenoside XLVII
58
Gynosaponin TN-2
670 G. sylvestre
l e a v e s are gypenoside I I ( 5 2 ) , V (53)^^, XLIII ( 5 4 ) , XLV (55)^^,
XLVII (56)
, LXXIV (57) and gynosaponin TN-2 ( 5 8 ) 7
a l s o I s o l a t e d from
G,
pentaphyl
gymnemasides I-VII are : 40, mp 159-161°C, 5 (C^D^N) 0 . 9 1 , 0 . 9 7 , (lH,dd,
£
1.02,
£ 7.5Hz,H-l of g l u c o s e ) , [oC ] 1.68, 24),
+ 10.5°
1.A8,
the
1.69,
1.69 (3H, each s . Me x 7 ) ,
(lH,in,H-24),
10.31
(lH,s,H-19),
4.95
3.46
(lH,d,
5.06 (lH,d,J; 7.5Hz,H~l of g l u c o s e ) ; 4 1 , mp 212-214°C, 0.99,
1.03,
1.32,
1.48,
1.68 (3H, each s . Me x 7 ) , 3.37 ( l H , d d , £ 1 1 . 5 , 4 . 0 H z , H - 3 ) , 5.30 (lH,m,H1 0 . 3 0 ( l H , s , H - 1 9 ) . 4.91 ( l H , d , J 8.0Hz,H-1 of g l u c o s e ) , 5.05 ( l H , d , £ 8.0Hz,
(c 6 . 0 ,
5.35 ( l H , d , j ; 8.0Hz,H-1 of g l u c o s e ) ; 4 2 , mp 182-184°C, [o^l^ +
MeOH), "^H NMR 6 (C^D^N) 0 . 9 1 , 0 . 9 8 , 1 . 0 3 , 1 . 2 9 ,
(3H, each s .
Me x 7 ) ,
(lH,s,H"19),
4.95
glucose), (c 1 . 0 , each
5.30
were
H NMR data of
[ oC ]j^ + 2 3 . 7 ° (c 3 . 0 , MeOH), "^H NMR
(c 3 . 5 , MeOH), "^H NMR 6 (C^D^N) 0 . 9 1 ,
H-1 of g l u c o s e ) , 8.2°
1.37,
11.5,4.0Hz,H-3),
These gypenosides
lum . The mp, [ oC ]^ and
5.17
( l H , d , j ; 6.0Hz,H-1 of a r a b i n o s e ) ,
s.
5.05
7),
4.96
5.02
3.46
(lH,dd,£
(lH,d,J
11.5,4.0Hz,H-3),
8.0Hz,H-1
of
glucose),
( l H , d , j ; 8.0Hz,H-1 of g l u c o s e ) ;
1.48,
5.36 4.96
1.69
of
I ^ I p + 1^.5°
1.70,
1.72 (3H,
(lH,m,H-24),
10.32
( l H , d , j ; 8.0Hz,H-1
44, mp 187-189°C,
MeOH), •'"H NMR 6 (C^D^N) 0 . 9 0 , 0 . 9 9 , 1 . 0 4 , 1 . 3 1 , 1 . 4 8 ,
each s , Me x 7 ) , 3.34 ( l H , d d , j ; 1 1 . 5 , 4 . 0 H z , H - 3 ) ,
1.68,
( l H , d , j ; 7.5Hz,H-l
(lH,d,J_ 8.0Hz,H-1 of g l u c o s e ) ; 4 3 , mp 256-257°C,
Me x
xylose),
1.48,
5.30 (lH,m,H-24), 1 0 . 3 1
MeOH), -"-H NMR 6 (C^D^N) 0 . 9 1 , 0 . 9 7 , 1 . 0 2 , 1 . 3 8 ,
(lH,s,H-19),
(c 5 . 0 ,
3.37 ( l H , d d , £ 1 1 . 5 , 4 . 0 H z , H - 3 ) ,
of
[ c^ ]^ + 5.9°
1.70,
1.72 (3H,
5.36 ( l H , s , H - 2 4 ) , 1 0 . 3 0 ( l H , s ,
H-19), 4.89 ( l H , d , j ; 7.5Hz,H-l of g l u c o s e ) , 4.91 ( l H , d , j ; 7.5Hz,H-l of g l u c o s e ) , 4 . 9 8 ( l H , d , £ 7.5Hz,H-l of x y l o s e ) ,
5.31
( l H , d , £ 8.0Hz,H-1 of g l u c o s e ) ; 45, mp
188-189°C, [oC]^ + 8.7° (c 1 . 4 , MeOH), "'•H NMR 8 (C^D^N) 0 . 8 2 , 0 . 9 9 , 1 . 0 5 , 1.25,
1.57,
1.57
£ 15.5Hz,H-24), glucose),
5.40
(3H,
each
s,
Me x
7),
3.36
(lH,d,J
6.19 ( l H , d d d , £ 1 5 . 5 , 8 . 5 , 5 . 5 H z , H - 2 3 ) , (lH,brs,H-l
MeOH) ,''"H N M R 6 ( C ^ D ^ N )
of
0.87,
1.04,
1.09,
1.07,
6.07
(lH,d,
5.14 ( l H , d , j ; 7.5Hz,H-l
rhamnose); 46, 185-187°C, 0.99,
9Hz,H-3),
[ oC ]^ + 7.7°
1.26,
1.61,
(c
1.62
(3H,
each s . Me x 7 ) , 3 . 4 0 ( l H , d d , J 1 1 . 5 , 4 . 0 H z , H - 3 ) , 6.13 ( l H , d , J 1 5 . 5 H z , H - 2 4 ) , (lH,ddd,£ 15.5,8.0,4.5Hz,H-23),
4.96 ( l H , d , J 7.5Hz,H-l of x y l o s e ) ,
5.16
of 1.4,
6.20
(lH,d,
J 8.0Hz,H-1 of g l u c o s e ) .
3.
BIOLOGICAL ACTIVITY Gymnemic
acids
are
primarily
known
for
their
ability
to
inhibit
temporarily to taste sweet substances. However, these are also known to have therapeutic effects for gastroenteric disorders, diabetes and cough. Recent studies by several groups of workers demonstrated the inhibitory effects of gymnemic acids on various physiological functions.
671 3.1
Antisweet Activity
Maeda
et
al.
studied
the
antisweet
activity
of
gymnemic
acids
isolated from the leaves. The two gymnemic acids 1 and 2 thev characterized seem to be identical with gymnemic acids IV and III respectively. The antisweet activity of gymnemic acids was assayed as follows : A solution of gymnemic acid (5 ml) in O.OIM NaHCO^ was held in the mouth of each of the four subjects for 2 min. The solution was spat out and the mouth was rinsed with distilled water. The subjects were then asked to taste 10 sucrose solutions of strength from 0.1 to l.OM. The activity of a gymnemic acid solution was expressed as the maximum concentration of a sucrose solution whose sweetness was suppressed completely. Application of 1 mM
solutions of 1 and 2 to the mouth led to a complete
suppression of sweetness induced by 0.2M and O.AM sucrose respectively. Deacyl gymnemic acid showed no antisweet activity. Moreover, the difference between structures of the two gymnemic acids tested is only the presence or absence of a double bond in the acyl group. The authors suggested that the acyl groups might play an important role in generation of the antisweet activity. Similar conclusion was also drawn from the study of antisweet activity of ziziphin, an 45 antisweet principle isolated from Zizyphus jujuba . However, the results of 20 the study by Yoshikawa et al. on nonacylated antisweet principles from G. syJvestre suggested that the acyl groups only increase the antisweet activity rather than playing the essential role. A patent has been taken on a process for extracting the concentrated gymnemate from G. sylvestre
which involves (i) sterilising the dried leaves of
G. sylvestre with nitrogen gas at 120'^C, (ii) extracting the leaves with phosphate buffer (pH 7) at 60°C, (iii) removing fat ingredients with an organic solvent lighter than water, such as hexane, heptane and petroleum ether and (iv) removing chlorophyll with an organic solvent heavier than water, such as CHCl^, ethylene dichloride and CCl,. An apparatus for the extraction process • 46 has also been presented. The process for taste improvement of extracts from leaves of 47 G. syvestre has been reported. To reduce the bitterness and antisweet character of gymnemic acid, a mixture of starch and gymnemic acid was treated with
cyclomaltodextrin
disappeared
and
glucanotranferase.
the antisweet
activity
was
As
a
result,
weakened
15
the fold
bitterness in suitable
conditions. Addition of ^-cyclodextrin to gymnemic acid samples was effective in reducing the bitterness and antisweet activity. Encapsulation of G.
sylvestre
by mixing it with natural water soluble polymer and additional natural oil and 48 lecithins has also been patented. The encapsulated food is free from the bitter taste of Gymnema,
useful for the treatment of diabetes and is anti-
672 cariogenic. A patent has been taken on a method for reducing the bitter taste of Gymnema
extracts?" The extract which is useful for pharma-
sylvestre
ceuticals and low calorie foods is mixed with ^ 5 fold starch followed by treatment with cyclodextrin glucosyltransferase. Thus, bitter tasting crude EtOH extract of Gymnema sylvestre mixed with starch was dissolved in acetate buffer (pH 6) and heated with cyclodextrin glucosyltransferase at 54°C for 48 h. The extract showed a very weak bitter taste. Sweet taste sensation is believed to be induced by adsorption of sweet substances on the receptor protein in taste receptor membranes. Inspite of extensive studies by various investigators, the receptor mechanism of sweet substances is still not clear. A sweet taste suppressing peptide, gurmarin has recently been isolated from the leaves of G.
The complete amino
sylvestre.
acid sequence of gurmarin has been determined. It consists of 35 amino acid residues with an amino-terminal
pyroglutamyl
residue and has the molecular
weight of 4209. Gurmarin has no significant homology with other known proteins. A
recent
review by Kurihara
deals with structures and physiological
action of sweetness-masking substances such as gymnemic acids, ziziphin, heatstable sweet protein, mabinlins and taste converting protein from sourness to sweetness such as miraculins and curculins. The process of taste improvement of beverages containing sweetness with 52 high sweetness and gymnemic acids has been patented. It has been claimed that the tastes of the sweetness are improved by addition of gymnemic acids. The suppression of sweetness by gymnemic acids and the effects on glucose absorption
and 53 glucosyl-transferase has been reviewed.
3.2
in
the
small
intestine
on
glucan
Hypoglycemic and Antihyperglycemic
formation
by
bacterial
Activity
In a recent study it was demonstrated that oral administration of water soluble fraction of alcoholic extract of leaves of G. sylvestre led to marked lowering of blood glucose in normal, glucose-fed hyperglycemic, insulin-treated and
streptozotocin-lnduced
diabetic
rats.
The
results
revealed
that
the
maximum glucose suppression occurred after 2 h of treatment by the effective 54 dose of 500 mg/kg of the extract. The effect of gymnemic acid on the elevation of blood glucose concentration induced
with
Rang et al.
oral
sucrose
in
stretzotocin-diabetic
rats were
studied
by
The rats with streptozotocln induced diabetes mellltus and loaded
orally with 4g sucrose/kg were given one to four doses of 400 ng gymnemic acid/kg around the time of sucrose administration. They observed that gymnemic acids had dose-dependent hypoglycemic activity.
673 The effect of gymnemic acid on the elevation of blood glucose concentration induced with oral administration of sucrose in normal rats was investigated by Suh and Suh.
The results suggested that gymnemic acid had a suppressive
effect on blood
sugar level after sucrose administration. Thus the authors
opined that gymnemic acid might find application in the prevention of diabetes mellitus
and
obesity.
Investigation
on
the effects
of gymnemic
acid
and
pullulan on the oral sucrose tolerance in normal and diabetic rates revealed that these substances might suppress glucose absorption in the small intestine, leading to a suppression of insulin release from pancreas normally caused by the increase in blood glucose. Inhibitory effect of gymnemic acid on glucose 58 absorption in the rat was also investigated by Yoshioka. It was observed that the increase in blood glucose concentration after oral administration of 2g/kg of sucrose was suppressed by oral administration of the acid. Ikeuchi studied the effect of gymnemic acid in a large dose on the plasma 59 glucose concentration of rats. The hydrophilic and hydrophobic fractions were tested
separately.
An
increment
of
the
plasma
glucose
concentration
was
suppressed in normal rats 30 min after oral administration of the hydrophilic fraction (5 mg/kg body wt ) or the hydrophobic fraction (5 mg, 100 mg/kg body wt ) with the glucose solution
(1 g glucose/kg body wt). Ingestion of the
hydrophilic fraction (100 mg/kg body wt) alone increased the plasma glucose concentration,
but
that
of
hydrophobic
fraction
did
not
affect
it. Oral
administration of gymnemic acid (200, 500/kg body wt) raised the plasma glucose concentration and i.p. treatment of the acid (25 mg/kg body wt) also stimulated it with an increase in the glucagon, insulin, adrenaline, corticosterone and ACTH concentration. Thus it was apparent that some components of gymnemic acid affect
not
only
the adrenal
gland
but
also
the hypothalamus
and
/or the
pituitary gland.
3.3
Anticaries Activity
Gymnemic acid has been implicated for prevention of dental carles . decomposition of sugar and production of glucan by Streptococcus
/nutans
The
which
causes dental caries are prevented by gymnemic acid as a carlostatic agent. The plant
G. s^Jvestre
by S. mutans
can be used as carlostatic food. Plaque formation in
vitro
in the presence of sucrose was inhibited by gymnemic acid. G. syi-
vestre extract may be administered in the form of tea. Dentifrices containing gymnemic acids inhibit dental plaque formation. Gymnemic acids were mixed with CaCO^, glycerin, apatite, deodorant etc. to give a dentifrice containing ^200 ppm gymnemic acid.
674 3.4
Inhibition of Melanin
Formation
Preparation of skin cosmetics containing Gymnema sylvestre extract was 62 patented. The dried leaves of G. sylvestre were extracted with water and the extract concentrated and filtered. The inhibitory activity of the extract containing gymnemic acids on melanin formation was demonstrated. Antiallergic Activity
3.5
Inhibitory effects of pec tic substances on activated hyaluronidase and 63 histamine release from mast cells were reported. Pectic substances including those purified from G. sylvestre
inhibited histamine release from isolated rat
peretoneal mast cells, which had been induced by the antigen. The results suggest that pectic substances may have antiallergic activities. 3.6
Miscellaneous The
inhibitory
activity
of
hyaluronidase
was
found
in the aqueous
extract of G. sylvestre . The two active substances containing 73% D-galacturonic acid with the approximate molecular weights of A.O x 10 64 respectively were purified. An extract of G. sylvestre
and 2 x 10
leaves and purified gymnemic acid inhibited
glucose-stimulated gastric inhibitory peptide secretion in rats. et al.
Sinsheimer
reported isolation and antiviral activity of the gymnemic acids from
G. sylvestre. The strain of Ann Arbor 6/60 of Asian influenza virus was used for j ^ vitro tests and for jja vivo studies A/PR8 strain of influenza virus most highly adapted to mice was used. Both jji^ vitro and in vivo tests indicated significant antiviral activity from the gymnemic acids. In in vivo tests better activity was observed with i.p. administration than with s.c. treatment. 4. CONCLUSIONS Gymnema
sylvestre
which has long been well-known in India for its anti-
sweet property has now been found
to have several other useful medicinal
properties. The active principles of the plant are present as a very complex mixture
which
characterization.
created
limitation
Advances
in
their
in chromatographic
and
isolation
and
spectroscopic
chemical
techniques,
particularly in the last two decades, now permit the isolation and structural analysis of complex biologically active plant constituents which are present in too minute quantities to have been characterized previously. Several gymnemic acids and other bioactive triterpene glycosides and a sweet taste suppressing peptide, gurmarin have already been isolated. The plant has attracted much attention in recent years and more interesting chemicals are expected to be isolated
and
characterized.
discovered. Dammarane-type
Further uses of the plant are expected saponins present
to be
in this plant are structurally
675 similar
to
bioactive
ginseng
saponins.
As
such
ginseng-like
interesting
biological activities of this plant are anticipated. Moreover, occurrence of this species is not so much abundant. Tissue culture and genetic manipulation could provide new means for economic production of the plant and the chemicals it produces. The recent drift of people's preference for herbal medicines will encourage intensive activities in these potential areas.
REFERENCES 1
R.N. Chopra, Indigenous Drugs of India, 2nd ed., Art Press, Calcutta, India, 1958. 2 R.N. Chopra, S.L. Nayer and I.C. Chopra, Glossary of Indian Medicinal Plants, CSIR, New Delhi, 1956. 3 B.N. Sastri (Ed), The wealth of India, Raw Materials, Vol.IV, CSIR, New Delhi, pp. 275-277. 4 Falconer, Pharm. J. Trans. 7 (1847) 351. 5 D. Hooper, Pharm. J. 17 (1886-1887) 867. 6 D. Hooper, Chem. News, 59 (1889) 159. 7 L.E. Shore, J. Physiol. 13 (1892) 191. 8 R.M. Warren and C. Pfaffmann, J. Appl. Physiol. 14 (1959) 40. 9 C. Pfaffmann, Handbook of Physiology, Sec.l : Neurophysiology, Vol.1, American Physiological Society, Washington D.C., 1959, p.507. 10 K.S. Yackzan, Alabama J. Med. Sci., 3 (1966) 1. 11 W. Stocklin, E. Weiss and T. Reichstein, Helv. Chim. Acta, 50 (1967) 474490. 12 W. Stocklin, Helv. Chim. Acta, 50 (1967) 491-503. 13 W. Stocklin, Helv. Chim. Acta, 52 (1969) 365-370. 14 G.S, Rao and J.E. Sinsheimer, Chem. Commun., 1968, 1681. 15 H.M. Liu, F. Kiuchi and Y. Tsuda, Chem. Pharm. Bull., 40 (1992) 13661375. 16 W. Stocklin, Helv. Chem. Acta, 51 (1968) 1235-1242. 17 J.E. Sinsheimer and G.S. Rao, J. Pharm. Sci., 59 (1970) 629-632. 18 S.B. Mahato, N.P. Sahu, S.K. Sarkar and G. Podder, International Seminar on Traditional Medicine, Calcutta, November, 1992. 19 S.B. Mahato, N.P. Sahu, S.K. Sarkar and G. Podder, Phytochemistry, in press. 20 K. Yoshikawa, S. Arihara and K. Matsuura, Tetrahedron Lett., 32 (1991) 789-792. 21 S.B. Mahato and B.C. Pal, J. Chem. Soc. Perkin Trans 1, (1987) 629-634. 22 K. Yoshikawa, S. Arihara, K. Matsuura and T. Miyase, Phytochemistry, 31 (1992) 237-241. 23 Y. Kurihara, Life Sci., 8 (1969) 537-543. 24 J.E. Sinsheimer, G.S. Rao and H.M. Mcllhenny, J. Pharm. Sci., 59 (1970) 622-628. 25 K. Yoshikawa, M. Nakagawa, R. Yamamoto, S. Arihara and K. Matsuura, Chem. Pharm. Bull., 40 (1992) 1779-1782. 26 T. Imoto, F.M. Yamamoto, A. Miyasaka and H. Hatano, J. Chromatography, 557 (1991) 383-389. 27 D.H. Williams, C. Bradley, G. Bojesen, S. Santikam and L.C.E. Taylor, J. Am. Chem. Soc, 103 (1981) 5700-5704. 28. C. Fenselau, J. Nat. Prod., 47 (1984) 215-225. 29. S. Seo, Y. Tomita, K. Tori and Y. Yoshimura, J. Am, Chem. S o c , 100 (1978) 3331-3339. 30 R. Kasai, M. Okihara, J. Asakawa, K. Mizutani and 0. Tanaka, Tetrahedron, 35 (1979) 1427-1432.
676 S.B. Mahato, N.P. Sahu, A.N. Ganguly, R. Kasai and 0. Tanaka, Phytochemistry, 19 (1980) 2017-2020. Chem. Pharm. Bull. 26 H. Ishii, K. Tori, T. Tozyo and Y. Yoshlmura, 32 (1978) 674-677. S.B. Mahato and A. Kundu, Phytochemistry, in press. 33 34 K. Yoshlkawa, K. Amimoto, S. Arihara and K. Matsuura, Tetrahedron Lett. 30 (1989) 1103-1106. K. Yoshikawa, K. Amimoto, Arihara and K. Matsuura, Chem. Pharm. Bull., 35 37 (1989) 852-854. 30 (1989) 1547M. Maeda, T. Iwashita and Kurihara, Tetrahedron Lett 36 1550. 37 F. Kiuchi, H.M. Liu and Y. Tsuda, Chem. Pharm. Bull. 38 (1990) 23262328. Yoshikawa, T. Nakajima and M. Okuhira, Arihara, 38 T. Takemoto, S Yakugaku Zasshi, 104 (1984) 325-331. T. Takemoto, S. Arihara, K. Yoshikawa, T. Nakajima and M Okuhira, 39 Yakugaku Zasshi, 104 (1984) 939-945. K. Yoshikawa, T. Takemoto and S. Arihara, Yakugaku Zasshi, 106 (1986) 75840 763. T. Takemoto, S. Arihara, T. Nakajima and M. Okuhira, Yakugaku Zasshi, 103 41 (1983) 173-185. 42 T. Takemoto, S. Arihara, K. Yoshikawa, J. Kawasaki, T. Nakajima and M. Okuhira, Yakugaku Zasshi, 104 (1984) 1043-1049. 43 T. Takemoto, S. Arihara, K. Yoshikawa, K. Hino, T. Nakajima and M. Okuhira, Yakugaku Zasshi, 104 (1984) 1155-1162. 44 K. Yoshikawa, M. Mitake, T. Takemoto and S. Arihara, Yokugaku Zasshi, 107 (1987) 355-360. 45 Y. Kurihara, K. Ookubo, H. Tasaki, H. Kodama, Y. Akiyama, A. Yagi and B. Halpem, Tetrahedron, 44 (1988) 61-66. 46. B.Y. Hwant and S.Y. Choi, Dong Kook Pharmaceutical Company Ltd., European patent EP 406, 516, 09 Jan. 1991. 47 T. Nagaoka, H. Hane, H. Yamashita and I. Kenso, Seito Gijutsu Kenkyu Kaishi, 38 (1990) 61-70. 48 K. Numata, Japanese patent JP 03,130,051 03 Jun. 1991. 49 H. Hane and G. Kenmassa, Dainippon Sugar Co. Ltd., Japanese patent JP 6402,552 o6 Jan. 1989. 50 K. Kamei, R. Takano, A. Miyasaka, T. Imoto and S, Hara, J. Biochem. Tokyo, 111 (1992) 109-112. 51 Y. Kurihara, Kagaku to Seibutsu, 29 (1991) 531-536. 52 M. Okamoto, Y. Koike and M. Utena, Japanese patent JP 04,104,778, 07 Apr. 1992. 53 T. Imoto, Seibatsi Butsuri, 30 (1990) 146-150. 54 R.R. Chattopadhyay, C. Medda, S. Das, T.K. Basu and G. Podder, Fitoterapia LXIV (1993) 450-454. 55 J. Kang, H. Koh and T.K. Suh, Hanyang Vidae Haksulchi, 10 (1990) 587-601. 56. J.H. Suh and T.K. Suh, Hanyang Vidae Haksulchi, 9 (1989) 505-518. 57 Y. Kurata, Yonago Igaku zasshi, 38 (1987) 61-70. 58 S. Yoshioka, Yanago Igaku Zasshi, 37 (1986) 142-154. 59 H. Ikeuchi, Yonago Igaku Zasshi, 41 (1990) 414-431. 60 Y. Hiji, U.S. patent, US 4, 912,089, 27 Mar. 1990. 61 M. Hasomi, Japanese patent, JP 01,299,212, 01 Dec. 1989. 62 T. Horiuchi and H. Horiuchi, Japanese patent, JP 02,292,208, 03 De. 1990. 63 Y. Sawabe, K. Nakagami, S. Iwagami, S. Suzuki and H. Nakazawa, Biochim. Biophys. Acta, 1137 (1992) 274-278. 64 Y. Sawabe, S. Iwagami, Y. Maeda, K. Nakagami, S. Suzuki and H. Nakazawa, Eisei Kagaku , 36 (1990) 314-319. 65 T. Fushiki, A. Koiima, T. Imoto, K. Inoue and F. Sugimoto, J. Nutr. 122 (1992) 2367-2373. 66 J.E. Sinsheimer, G. Subba Roay, H.M. Mcllhenny, R.V. Smith, H.F. Massab and K.W. Cochran, Experientia, 24 (1968) 302-303. 31
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
677
Theory of the Origin, Function, and Evolution Secondary Metabolites Carsten Christophersen
1.
INTRODUCTION An overview of the structural theory of chemistry and the evolutionary theory of biology is given and brief outlines of the common areas of the two are sketched.
Living organisms are characterized by their abiHty to manufacture and handle metaboHtes, primary as well as secondary. At first glance the diversity and numbers of known secondary metabolites seem countless. And yet we know that they can be classified in relatively few chemical structural categories based on known or suspected biogenetic relationships. Another classification takes advantage of biological systematics and catalogues these compounds according to their distribution in taxons (chemotaxonomy). Only to a certain extent can these two groupings be deduced from each other. The two classification systems reflect the fundamental theories of chemistry and biology: the structural theory of chemistry and the evolutionary theory of biology. Modem structural theory predicts chemical parameters from detailed knowledge of chemical structures. At least in principle it is possible to predict reactivities, reaction pathways, thermodynamic relationships, kinetic parameters, and products for molecular reactions. This level of sophistication can only be attained for simple systems at present.
X^
NH
HO
yc \/ V
H'^^
OH
^H
Scheme 1. Structures of synthetic 3-azetidinol [1] and the antibiotic charamin [2] from the freshwater green alga Chara globularis.
From our own work studies of synthetic 3-azetidinol [1], part of the structure of the antibacterial agent charamin (4-azoniaspiro[3,3]heptane-2,6-diol, 2) from the green characean alga Chara globularis (1), serve to illustrate the kind of data and instrumentation available in modem investigations. The crystal and molecular stmc*Dedicated to Philip and Pia
678
ture was solved by X-ray techniques (2). The structural analysis in combination with ab initio calculations revealed the presence of strong intermolecular hydrogen bonding in the solid state. Solution data from ^H-NMR, IR, and Raman spectroscopy in the R(v) representation served, in combination with ab initio calculations, to establish the existence of a non-associated form, a conformer with intramolecular hydrogen bonding, and an intermolecularly hydrogen bonded chain-stacking arrangement (3-4). The results for the gas phase structures were substantiated by electron diffraction studies (5). Having thus established that the methods gave reliable information about the molecular characteristics of the systems treated, the analysis was extended to more elaborate models. Ab initio data, although of great help in interpreting the spectroscopic information, usually refer to gas phase conditions. In order to gain information on solution structures, which are the parameters of interest in a biochemical connection, ab initio calculations were performed on a model structure with 3-azetidinol binding to two molecules of water and two molecules of ammonia (3). The results were verified by assignment of the fundamental vibrations obtained from IR and Raman spectroscopy on aqueous solutions of 3-azetidinol. The results are in agreement with the occurrence of a hydrated, stacked, intermolecularly hydrogen-bonded chain. These examples serve to illustrate that although intricate biochemical systems are still beyond the scope of ab initio methods valuable information can nevertheless be gained from model studies. Even if the chemical analysis cannot be performed to the level of ab initio calculations at present, the use of chemical models may in many instances allow rather detailed analysis of biochemical problems. An example is the study of the interactions between urea and trimethylamine oxide (TMAO). TMAO is usually considered an indicator in marine organisms although it has been found in a few freshwater species of fish as well (6). In most elasmobranchs, where the concentration of urea is high, the concomitant high TMAO concentration is believed to counteract the denaturing effect of urea on the cellular proteins, especially the enzymes. The molecular interactions have proved difficult to study but recently the intermolecular forces in a TMAO-urea complex, as studied by X-ray diffraction techniques, have yielded information in the discussion of protein - urea - TMAO interactions (7). As the techniques advance, complicated situations can be comprehended in chemical terms. As a result increasingly complex molecular interactions can be subjected to deductive predictions based on chemical principles. The theory of evolution in principle describes the evolution of living organisms and explains this progression in terms of genetics, physiology and ecology. Based on principles of "struggle for survival'' and "survival of the fittest" it encodes the selection mechanism leading to contemporary life forms. With the appearance of Neo-Darwinism these principles attained a firm genetic basis. Later molecular genetics and molecular biology as a whole have served to give an ever more intricate picture of the mechanism of evolution at the molecular level.
679
Since these two theories describe the same reaUty, a synthesis should be possible. However, presumably because of the extreme complexity of the whole area only very limited fields have been included in attempts at such descriptions. Comparison of the two theories reveals that while chemistry deals with how interactions occur between molecules, evolution deals with the adaptive value of these interactions in the current ecosystem or physiological status of an organism. To unify the two theories it is therefore necessary to translate chemical interactions into the framework of specific physiological events. Considerations of this type lead to a third categorization for secondary metabolites, namely a classification based on biological activity. The biological activity known to be associated with some secondary metabolites can be indexed within a limited number of pharmacological schemes based on gross pharmacological activity. This classification is not identical to any of the others. For example, a certain biological activity associated with some members of one class in the chemical classification system is not necessarily present in other members of the same group or in closely related taxa. What then is the natural (functional) classification system, if any, for secondary metabolites? We submit that it is the interactions with biochemical receptors. Receptors can be attacked in a variety of ways ranging from potent activation to total destruction. As this is synonymous with different structural interaction between the receptor and the active compound, the latter may belong to various structural groups and still cause comparable pharmacological action (see e.g. Section 6.2, Scheme 19. Nicotinic toxins). 2.
SECONDARY VERSUS PRIMARY METABOLITES The concept of primary and secondary metabolites is outlined and it is concluded that an exact definition is not especially usefiil The coexistence of metabolites and the cellular machinery for their transformation is outlined by the concept of compartmentation (Appendix 1). Tertiary metabolites and xenochemicals are defined for the sake of completeness (Section 2.1) and the current theories of the function of secondary metabolites are briefly discussed (Section 2,2).
Metabolites are chemical compounds involved in metabolism. The primary metabolism degrades (catabolizes) and constructs (anabolizes) primary metabolites. Primary metabolism is the part of intermediary metabolism concerned with the immediate task of furnishing energy, building blocks, constituents and may be reserve metabolites of an organism. The remaining compounds are secondary metabolites and are anabolized in secondary metabolism. Usually the catabolism of secondary products occurs through primary metabolic pathways, which then provide the resulting primary metabolites for primary or secondary pathways. As a rule of thumb those metabolites with very narrow taxonomical occurrence (in one or a few species) are often secondary. There are, however, numerous exceptions to this rule (for some prominent examples see Appendix 11). The term "natural product" is often used synonymous with secondary
680
metabolite. Strictly speaking natural products are a concept of wider extent, which also encompasses transformed organic material (e.g. oil, coal, amber, etc.) as well as inorganic naturally occurring entities. Most often chiral metabolites are characterized by optical activity because of their origin in enzyme mediated reaction sequences. In fact, this rule is of such generality that the first example of a marine secondary metabolite of low enantiomeric purity (8-9) appeared as recently as in 1984. An exact definition of primary and secondary metabolites is not very useful since it is deemed to be artificial. Although primary metabolism is to some extent similar in all organisms there are variations, major (photosynthetic/non-photosynthetic - aerobic/anaerobic organisms) as well as minor (bromotyrosine in sponge skeletons, ceramide aminoethylphosphonate in sea anemones (10)). This means that a secondary metabolite for one organism may assume the role as a primary one for another. From a purely formal point of view the distinction is meaningless and best abandoned. For classification purposes it still serves to delineate a large body of biochemicals in a convenient compilation. As mentioned above the metabolites are chemicals involved in metabolism. As such they are presumably always products of or substrates for enzymatic reactions. The regulation of metabolism is to a large extent achieved by the regulation of enzyme availability and activity. However, considering the enormous complexity of the metabolic organization, even taking into account the often unique specificity of enzymatic transformations, how is the observed specificity of metabolism achie v'ed? One very important means to accomplish this is through compartmentation of the various reactions and reagents as described in Appendix 1. 2.1 Tertiary Metabolites and Xenochemicals There is a class of chemical compounds present in all living organisms, "tertiary metabolites", which scarcely deserve the name of metabolite. They are compounds formed in purely nonregulated chemical pathways. They are extremely difficult to pinpoint since it must be shown that they are not formed by enzymatic reactions. One such example is the unstable aldimine formed from the hemoglobin pchain and glucose and the more stable corresponding ketamine (named HbA^^,) formed by an Amadori rearrangement of the aldimine. By comparing these entities formed in vivo as well as in vitro the entire system including all rate constants and equilibrium constants was determined (11-16). From comparison of these kinetic and thermodynamic parameters with the corresponding values of model reactions it could be concluded that the formation of the products was not subject to enzymatic catalysis, not even by hemoglobin itself (11). Apparently these p-chain derivatives have no physiological function and seem to be formed and degraded governed simply by physicochemical rules without metabolic or otherwise regulatory intervention. Their concentration profiles are used to monitor metabolic control in diabetic patients (14). This is not an outstanding example, and many more chemicals of this type are
681
expected to occur in organisms. The last category of chemical compounds present in the living organism are chemicals not originating from metabolic processes (xenochemicals). Many of the latter class are synthetic compounds like agrochemicals such as insecticides (DDT) and herbicides (phenoxyacetic acids), synthetic drugs, PCB, phthalates, plastic, synthetic fibers, etc. Others are natural products neither primary nor secondary metabolites such as kerogen, the most abundant organic compound mixture on earth. Most of these compounds are metabolized or at least derivatized (often by hydroxylation effected by the cytochrome P450 system) by one organism or another. 2.2 Function of Secondary Metabolites - Current Theories Current theories on the function of secondary metabolites are sunmiarized by Haslam (17) and critically evaluated by Williams (18) in connection with the introduction of their theory. The latter authors convincingly argue that by far the most likely hypothesis is that "the secondary metabolites are a measure of the fitness of the organism to survive. The ability to synthesize an array of secondary products which may repel or attract other organisms has evolved as one facet of the organism's strategy for survival." Intuitively one would think that in an organism's spectrum of chemically related secondary metabolites the minor ones will have the least significance in the Darwinian struggle for survival. Actually some theories consider these minor metabolites as by-products of the enzymatic formation of the major metabolites. Such a view is, again intuitively, supported by the often very complex nature of certain metabolite patterns and the presently imperfect knowledge of the factors governing their expression and function. In some cases, however, more detailed investigations into these matters have appeared. In the case of pheromones it is occasionally observed that the individual components have none or comparatively low activity, while the composite is highly active (synergism) and highly species specific. In the fall cankerworm, Alsophila pometaria, a populational study within the United States of America showed constancy of the pattern of the pheromone mixture in spite of genetic vari19
17
15
13
11
8
5
Scheme 2. Structures of the three pheromones of the cankerwomi Alsophila pometaria.
682
ation. The female pheromone always has the composition of (3Z,6Z,9Z)nonadeca3,6,9-triene, (3Z,6Z,9Z,ll£)-nonadeca-3,6,9,ll-tetraene, and (3Z,6Z,9Z,llZ)-nonadeca3,6,9,11-tetraene in the ratio of 25:65:15 (19-20). Another case is the oviposition stimulants in the leaf of Citrus unshiu consisting of four flavonoids, four organic bases, a quaternary ammonium compound and inositol (Scheme 3). None of the components stimulated oviposition in Papilio xuthus alone regardless of the concentration and the combination must be entirely specific to the host plant (20-21). Consequently, in these cases it is possible for the plant to maintain the exact composition of the mixture and the interval of release. Furthennore, other organisms have apparently been evolutionarily selected to recognize a set series of metabolites.
The nature of the set of chemically related secondary metabolites for any single species argues against their origin in random metabolic enzyme mediated transformations. The exact cellular location of the synthesis of most natural products and hence the enzymatic activity at the site of synthesis is usually unknown. Comparison of systematically related species with chemically related sets of secondary metabolites serves to stress the difference in structures. Such diversity is unanticipated if analogous sets of enzymes were to work on similar sets of precursors. Furthermore, many obvious transformations, expected from established enzymatic activities, are not observed. rutinosyhO
OH
0
1: X = OH; Y = H 2:X=0Me:Y=0H
HO.
NHMe
TO:
\ 5
Scheme 3. Structures of some of the components of the complex oviposition stimulant signal of Citrus unshiu. Narirutin [1], hesperidin [2], rutin [3], vicenin-2 [4], S-hydroxy-Nw-methyltryptamine [5], adenosine [6], and N,N-dimethylprolinium ion [7].
683
In fact, the one feature most preventive to the appHcation of the enormous amount of natural product chemistry data in life sciences is that these data are nearly always determined on populations of organisms as opposed to individuals. Most often the populations are not even clones but represent genetically different specimens. As discussed later there may be large individual chemical variations in specimens of a given population. At present we are also badly in need of more specific information regarding the location (organ or organelle) of metabolites. The tertiary metabolites seem to be devoid of any special biological activity of their own. They may conceivably be rendered active by mutations affecting their precursors and in this way gain access to the main biochemical events. If this happens they must eventually be subjected to regulatory systems such as enzymatic intervention. Their introduction into metabolic processes and their new activity demands means of regulation other than purely kinetic and thermodynamic control. 3.
THEORY OF ADAPTIVE VARIANCE IN METABOLISM The theory is presented and assumed to be valid for all true types of metabolites, primary as well as secondary. The theory emphasizes the existence of receptors tuned to the metabolites. These concepts are shown to be compatible with currently accepted genetic mechanisms in procaryotes as well as eucaryotes (Appendix 3). The role and demands to natural products chemistry in order to produce results suited to distinguish between different predictions are discussed (Section 3,1). Since stimuli have such an important place in the theory the character of these are presented (Appendix 2).
The theory of adaptive variance in metabolism (22) states that: All metabolites, primary as well as secondary, aim at the survival of the population or individual, are expressed as the result of stimuli, and are targeted to interact with receptor systems. The theory implies that also minor metabolites have functions analogous to the major ones. Major as well as minor metabolites are expressed only when the parent organism is in some way challenged or stimulated (for a discussion on the nature of stimuli, see Appendix 2). In other words, an organism may produce completely different patterns of metabolites depending on the immediate situation. In this way the energy expenditure is confined to information storage at any time in the genetic material and production of the most efficient set of secondary metabolites. Intermediary metabolism with the extremely sophisticated regulation of expression and control of the participating agents offers ample documentation of this principle. In the present context we will mainly confine the discussion to secondary metabolites even if, as mentioned above, the distinction is often quite arbitrary. The secondary metabolites are proposed to aim at specific target receptors and to be optimized to recognize and interact with these receptors. Within an array of related natural products the different structural variations recognize different receptors
684 which may well constitute different developmental stages of the same target receptor type(s). Stimulus
Organism
Metabolites
M i
'=^5%,=_^^' 00( ^j — — y
J
to
I
Receptors
A„
• ^ = = ^ 6 ,
B„
L
t,
-^w^
—
y
-^ g J
t'n
Results
^ [ g c = > A,
=J> Xi
X„
= ^
Y,
Y,
c=>
Z,
Z,
L
t;
Figure 1. Model describing the response of an organism or cell to two different situations characterized by different stimuli. The response is different m the two situations depending on the initial challenge.
Viewed in this way there is nothing fortuitous in the complex pattern of secondary metabolites often expressed by a certain organism. The pattem is the result of a long selection process mirroring the evolution of receptors and reflecting a complementary image of part of the ecological environment interacting with the organism. However, if this is correct the expression of each substance must be thoroughly regulated and it is difficult to imagine this to occur without the existence of specific enzyme systems for even the minor metabolites. A crucial point in this whole endeavor is: is this model at all compatible with current models in genetics? The answer is yes as discussed in Appendix 3. 3.1 The Role of Natural Products Traditionally the major part of natural product chemistry has been centered around structure determination of metabolites from more or less random collections of plants. These data are only of limited value in this context where we wish to study the expression of certain genetic features more closely. The majority of natural product papers describe the isolation and identification of a few metabolites with rather imprecise information conceming absolute and relative amounts. En worse, as a rule, the identity of the raw material is only superficially described. The geographical location of the collection is often given roughly without the time of
685
collection or any description of the ecosystem. The chemical investigations are nearly always carried out on samples of many specimens sometimes gathered in different locations but the genetic set-up of these different specimens is unknown. The physiological state and ontogeny of the samples are usually also unknown. To illustrate this point an issue of 7. Nat. Prod. (Nov.-Dec, 1990, 53, number 6) randomly selected has 44 papers of which 14 deal with studies of previously collected samples, synthetic work, etc. leaving 30 papers dealing with studies of the composition of newly collected biological material. Of these 30 studies, approximately half (14) have grossly inaccurate descriptions of the geographical location of collection, ranging from no mention at all to locations such as Costa Rica, Peru, Lesotho, Bay of Naples, ditch near Naples etc. Only 14 papers state the month of collection while 17 record the year. The physiological state of the organisms is recorded in only three papers. Several papers fail to mention which anatomical part of the organism that was used. In short, only two papers supply approximately adequate information regarding the biological material studied. One reports an investigation of the roots of Thapsia villosa [2n=66(=6x)], at the stage of fruit ripening, collected in July 1988 six km south of Alter do Chao on road no. 245 (23). The time of day of the collection was omitted even though many plants are known to exhibit pronounced variations in metabolite pattern as a function of the time of day (24). In this connection it should be borne in mind that J. Nat. Prod, attaches special importance to the description and identification of the biological material investigated to such an extent that work based on inadequately described material is not accepted for publication at all. This lack of acceptable detailed information is curious since large variations in secondary metabolite content as a function of a considerable number of internal and extemal variables have been well documented for many years (see e.g. ref. 24). Taking these shortcomings into account, it comes as no great surprise that separate investigations of the same species often give quite different, and frequently conflicting, results. These discrepancies are, however, often obscured by the specific aims of the investigations. After an initial investigation of a certain species where the main metabolites are identified, succeeding investigations often concentrate on the minor metabolites ignoring quantitative variations in the main metabolite pattern. Since the investigations are usually carried out on different batches of organisms there is no way to reconstruct the true metabolite pattern. Accordingly an investigation of the secondary metabolite pattern of a given species may very well result in figures as given in the statistical distribution in Scheme 5. Even these statistical patterns are often not really significant since they depend on the method of analysis employed. As most investigations do not really aim at describing the quantitative occurrence of the metabolites investigated, the percentage of the compounds in the natural mixture is only given based on the amounts isolated for structure determination. Clearly, even in investigations of the quantitative composition of some biological material, the methods used imply a lower limit of detection and a certain variance. If the situation illustrated in Scheme 5 was taken
686
from a normal natural product chemistry publication, the presence of compound e Metabolite (%)
Specimen
I a b c d e Statistical distribution:
n
m
IV
V
100 0 0 0 0
0 100 0 0 0
20 20 20 20 20
10 0 50 30 10
0 10 40 40 10
a: 26%
b: 26%
c: 22%
d: 18% e: 8%
Scheme 5. Five specimens I, n, DI, IV, and V express varying amounts of five secondary metabolites a, b, c, d, and e. The statistical composition of the sample is given for comparison.
might easily have been overlooked. The figures calculated for a, b, c, and d would then have been 28, 28, 24, and 20% respectively. As is evident from this example, a statistical distribution of metabolites may represent large individual variations in metabolite patterns and the absence of any metabolite can only be stated in relative terms in accordance with the limitations of the method of analysis. Even though comparable chemical data are often used randomly in chemotaxonomical arguments, this variation should come as no great surprise considering the fact that such collections are often genetically different. Even when the population sampled is genetically homogenous (a clone) we submit that the individual variation may well be great depending on the immediate level of challenge for the different individuals of the population. Very few current investigations qualify to allow knowledge regarding such matters to be extracted. Therefore, we cannot in fact be certain whether the seemingly very complex pattem of secondary metabolites associated with the most well investigated organisms actually reflects a statistical mean of a population of organisms with widely different immediate production of secondary metabolites or an approximately homogenous population with regard to secondary metabolite production. Furthermore, as discussed in Appendix 1, the metabolites are often restricted to certain compartments (organelles) even at the cellular level and this trend is becoming even more pronounced for multicellular organisms. In colonial organisms, where the individuals are supposedly genetically identical, large quantitative and qualitative variations in secondary metabolite content have been demonstrated (see Appendix 7 for an example involving stony corals). Similar findings have been reported for algae, where e.g. the antibiotic activity varied in different parts of the thallus (25) as did the
687
concentration of lanosol (26). In higher organisms with true organs the variation in metabolite pattern is the rule. However, the vast majority of natural product investigations are carried out on extracts of whole organisms and accordingly offer no information regarding the location of the compounds in the organisms. It is an open question how many of the published natural products are actually present in the living organisms and how many are artifacts. In an investigation of the fruits of Crinum asiaticum (27) it was found that the pattern of alkaloids was different when the fruits were treated with anaesthetic agents (ether and lidocaine) prior to extraction. Apparently the alkaloids previously considered characteristic of the fruits were formed as a response to injury. The conclusion of considerations of this kind is that most of the previous and contemporary natural products investigations are unsuited to shed light on the questions addressed here. We are in need of detailed analyses of individuals in populations where the variables in the surroundings and in the organisms themselves are rigorously controlled and specified. Only in this way will any deeper understanding of the molecular basis of adaptation of individuals and populations advance. 4.
ADAPTIVE VARIANCE IN METABOLISM, EXAMPLES Literature studies reveal that the variation in secondary metabolite content is in accordance with potential challenging events. This is true over a very broad taxonomic spectrum, microorganisms (Appendix 4), algae (Section 4,1 and Appendix 5), bryozoans (Section 4,2 and Appendix 6j, corals (Appendix 7), and sponges (Appendix 8). Even if each example may be explained by different theories, a unified explanation is at present only offered by the present theory.
The following section will present some selected examples of investigations which support the theory. The selection has an overweight of marine investigations and examples coauthored by the present author. Even if some of these examples are not necessarily the most illustrative available, they have been selected simply because of my own familiarity with these investigations and because in retrospect the ideas presented here were matured during these studies. Further examples including microorganisms, other algae, other bryozoans, corals and sponges are presented in Appendix 4-8, respectively. A large body of evidence fits well into the present theory. However, on closer inspection many of these examples are evidently of a more complex nature. The phenomena referred to are those where a trait may well have originated in response to external stimuli but where the expression has become of a permanent character independent of the original elicitor. It is conceivable that many of the permanent metabolites originated in this way but proved of such value to the organism that eventually evolutionary selection circumvented the need for a specific environmental elicitor replacing it with some internal cellular transcriptional regulatory system. Some became dependent on some form of internal clock. Such systems often express
688
periodic events such as circadian rhythms, seasonal variations, reproductive rhythms, etc. 4.1 Algae Like higher plants many algae are known to exhibit seasonal variations in secondary metabolite pattern. British marine algae were found to exhibit four main patterns of antibiotic production throughout the year, namely uniform production during die year, maximum production during winter, during summer, or in spring (28). 4.1.1 Caulerpalean algae Even algae from geographical identical locations may elaborate different secondary metabolites as a function of their ecological context as illustrated by some caulerpalean algae (29).
AcO^
AcO..
OAc OAc AcO>.
OAc
OAc
Scheme 9. Structures of diterpenes from the green alga Penicillus dumetosus. The content is dependent on the environmental conditions. Dihydroudoteal [2].
Investigations of Caribbean Penicillus dumetosus growing in grassbed environments revealed a content of the triacetate 1. When growing in reef environments, the same species was found in all the cases investigated to elaborate dihydroudoteal [2] and minor amounts of 3 and 4 (Scheme 9). This observation correlates with the fact that reef habitats have much greater levels of herbivory than seagrass beds. Accordingly, in all cases investigated, reef populations of algae either showed different secondary metabolites, higher concentrations of major metabolites or greater variations in metabolites than grassbed populations. Furthermore, to mimic the grazing situation, a comparison was made between about 20 plants subjected to clipping for a period of 3-7 days and the same number left unharmed. The extractable matter rose from 0.2
689
to 0.44% of dry weight on clipping. The major metaboHte, the triacetate 1, rose from 20 to 40% of the extracts (from 0.04 to 0.18% of dry weight) on cHpping. A comparative study of four populations of caulerpalean algae, Halimeda tuna and Halimeda incrassata from the Florida Keys, Halimeda incrassata and Rhipocephalus phoenix from the Bahama Islands, convincingly demonstrated that, although single plant specimens were collected within meters from each other, concentrations of metabolites varied from 0 to >2% of algal dry weight. The present information allows no conclusion as to whether the plants devoid of secondary metabolites represent as yet unchallenged individuals, genetic variants, or examples of mimicry. It has been shown (29) that reef slope populations of Caulerpa racemosa elaborate secondary metabolites significantly avoided by herbivorous fish, while reef flat populations produce different and less deterrent metabolites. These observations all agree with the hypothesis that metabolites expressed by different populations (individuals) are the product of stimuli originating in the surrounding ecosystem. 4.2 Bryozoans The following description of investigations of natural products from Scandinavian specimens of the marine bryozoan Flustra foliacea is an example of a "classical" natural product study. Most of the pitfalls mentioned in Section 3.1 are amply demonstrated in this section. The exact geographical location of collection is not recorded since most of the material was delivered by commercial fishermen when they by chance happened to fmd larger amounts mixed in the catch. Thus a collection may represent individuals from several different locations and ecosystems. In principle the individuals could be widely genetically different and in different physiological states. Furthermore, since the were kept iced with the catch and only placed in a deep freezer on return of the trip, the time before processing could be very different. After extraction no specific effort was made to determine the exact concentration of the individual components and no effort was made to positively identify all known compounds in each collection. Some of the compounds isolated might even be artefacts as for example a racemic diterpene (30). The location of the secondary metabolites in the colony or the individual zooids has never been investigated. In addition many secondary metabolites isolated from bryozoans may conceivably originate with associated microorganisms (31). This symbiosis, however, does not affect the arguments presented here since the secondary metabolites still retain their function in the preservation of the species independent of their origin. For the purpose of this discussion host and symbiont perform as one intricate species. Even in spite of all the methodical shortcomings of these investigations as compared to the directives mentioned before they allow some inferences in the present context and serve to inspire the present work.
690
4.2.1 Flustriidae The marine bryozoan Flustrafoliacea collected in the North Sea has given rise to the isolation of a complicated mixture of alkaloids, twelve of which have so far been identified (32-36). The structures are depicted in Scheme 11: flustramine A (1)
3
R = -OH
.«?
I
R^ =
R2 =
H
II R^ = .
IV
R' = H
V
R^ =
m R^ = H R2 = ^ - Y /
Scheme 11. Alkaloids from the marine bryozoan Flustrafoliacea. Structures numbered in Arabic originate from specimens from the North Sea and those with Roman numbers are from Canadian specimens. Flustrabromine (11) and G-bromo-N^j-methyl-AT^j-fomiyl-tryptamine (10) were isolated as an intimate mixture of Z and E geometrical isomers. The fonnulas are not intended to depict absolute stereochemistry.
691
and B (2) (37-38), flustraminol A (4) and B (3) and flustramine C (5) (39), flustramide A (6) (40) and B (7) (35) and flustramine E (8) (36) are all based on the 6bromo-l,2,3,3a,8,8a-hexahydropyrrolo[2,3-fc]indole skeleton, while 6-bromo-A^^methyl-A^^j-formyltryptamine (10) (40) and flustrabromine (11) (41) are rather simple tryptamine derivatives. Flustrarine B (9) could be derived from the 6-bromol,2,3,3a,8,8a-hexahydropyrrolo[2,3-fc]indole skeleton by N oxidation followed by ring expansion. 7-Bromo-4-(2-ethoxyethyl)quinoline (12) (42) is the only example of a naturally occurring bromoquinoline. Investigations of the same species from New Brunswick and Nova Scotia have yielded five structurally identified alkaloids depicted in Scheme 11: Dihydroflustramine C (I) (43), flustramine D (II), isoflustramine D (III, tentatively identified as the minor component in a 65:35 mixture with II), dihydroflustramine C N-oxide (IV), and flustramine D TV-oxide (V) (44). Origin and Identity of Biological Material The Canadian material was collected in the Minas Basin, Nova Scotia and in the Bay of Fundy off the New Brunswick and Nova Scotia shores. The Scandinavian material was collected in the North Sea off the coast of Jutland and off Tjamo Marine Biological Station, Sweden. The Scandinavian samples are believed to be homogenous since the only mistakable species in this area is Securiflustra securifrons which contains alkaloids of a totally different nature (unpublished results). Alkaloid Content and Distribution In the Scandinavian collections flustramine A (1) and B (2) were always the main alkaloids and present in equal amounts (except in one study (36) where the alkaloids were isolated by Likens-Nickerson gas phase extraction, presumably giving rise to different amounts reflecting the different gas phase extraction efficiencies of the two compounds). In most collections the combined contents were around 7x10 % of dry weight. However, in one collection from Swedish waters the content was doubled (16x10"^%). One group of alkaloids was about a factor of ten less abundant: 6 (5xlO"^%), 8 (6x10-^%), 9 (4x10*^0), 10 (2x10'^%), 11 (6x10'^%), and another group around a factor 100 less abundant: 3 (8xlO"^%), 4 (6xl0"^%), 5 (3x10"^%), 7 (5-6x10'^%), 12 (trace). In the Canadian collection the main alkaloids were dihydroflustramine C (I) andflustramineD (II) present in 3x10"^ and 2x10'^% of wet weight, respectively (calculated from content of I and the ratio between I and II given in ref 44). These amounts are comparable to the ones found for the major flustramines. The N-oxides were present in very small amounts (not given), and III in trace amount (not given). From these data it seems that there may be a difference in the plethora of compounds present in the two populations of F. foliacea. In this connection it should be remembered that the minor alkaloids in the Scandinavian material originate from several different collections and thus may individually be characteristic of the
692
collection in question. Thus the seemingly very complicated mixture may merely reflect an average of many collections with few individual alkaloids. Alkaloids, Geographical Variation As yet not one single compound has been found to co-occur in the two populations. The five compounds from the Canadian collection arc all characterized by the presence of the inverted isoprene unit (2-methyl-3-buten-2-yl) in the 3a position. Of the tricyclic structures from the Scandinavian collection only three compounds (1, 5, and 6) encompass this configuration. Therefore, even if I and 8 only differ in this aspect they are clearly biogenetically different. Likewise, even if the A^-oxide of flustramine B (2-Nl-oxide) participates in an acid-base catalyzed equilibrium with flustrarine B (9) (45) analogous to the geneserine and geneseroline equilibria with hexahydropyrrolo[2,3-fe]indoleA^-oxidesandhexahydro-l,2-oxazino[5,6-fc]indoles(46) there is no possibility of transforming IV to 9. Biological Activity of the Extracts The Canadian collection exhibited strong activity against Bacillus subtilis, while the North Sea samples were below the limit considered active in our assay against the same species. For example I was claimed to be strongly active against Bacillus subtilis in the disc diffusion assay at 0.5 mg/disc without any inhibition zone given. Flustramine E (8) only showed faint activity in a similar assay using the same bacterium but using wells instead of filter paper discs (12 mm zone with 15 |ig/well and no inhibition with 1.5 jig/well). The apparent discrepancy may thus merely reflect different definitions of activity. Another explanation may be that different strains of B. subtilis were used. It is well known that different strains of B. subtilis may have widely different sensitivities towards antibiotics. The Scandinavian samples possess muscle relaxant activity in the electrically stimulated isolated guinea pig ileum assay (47) and antiviral activity against Herpes and influenza virus (48). Furthermore, the Scandinavian collections contain a mixture of simple monoterpenes with antifouling activity (49). Unfortunately, the muscle relaxing effect, the antifouling activity, and the antiviral activity of the Canadian population have not been reported. Based on these observations no safe conclusion can be drawn regarding differences in biological activity between the two sets of metabolites. The bryozoan Chartella papyracea from the same family, collected around Roscoff, France, contains a series of chartellines (50-52) and chartellamide A and B (53). A collection of this species (called Flustra papyracea) from the same area reportedly gave quite different alkaloids named papyraceabromines (54-55). The two main alkaloids exhibited wide-spectrum in vitro antibacterial activity. One of the latter named papyraceabromine-A was unstable and assigned the composition ^22^17^305^^4 ^^^^^ o^ ^^C- and ^H-NMR (C22H17), low resolution mass spectrometry (Br4) and elemental analysis (C22H17N3). The structurally different minor alka-
693
V s r
A : R^ = CLR2 = R^ = Br B:
R'
= CL R^= Br, R^ = H
C : R^ =^Cl. R 2 = R ^ = H
R^ = OMe, R^^R^^Br
A: R = H B: R= Br
Scheme 12. Alkaloids from Chartella papyracea. At left, chartelline A, B, C, and methoxy-dechlorochartelline A. At right chartellamide A and B.
loids papyraceabromine-B with five bromine atoms and papyraceabromine-C (monodebromo-B) are the more stable. Careful investigation of material recollected in the same area (Morlaix Bay) as Pietra's original collection only gave the known alkaloids (51). The data given by Pietra et al are not compatible with those of the chartellines, the main alkaloids of our samples, or our minor chartellamides. Furthermore antibacterial activity was absent in our crude alkaloid mixture. The solution of this problem will have to await future investigations but the results are comparable to the differences described for the geographically separate collections of Flustra foliacea (vide supra) and Amathia wilsoni (Appendix 6). 5.
RECEPTORS As the existence of specific receptors is as crucial to the theory as the variation in metabolite content, this section examines some classes of receptor families and their variability. Studies of their evolution {Chapter 5 and Appendix 10) demonstrate their ancient roots and their transformation in time into more and more sophisticated structures. Following the differentiations of receptors the secondary metabolites have experienced a concomitant evolution (Appendix 10). The metabolic pattern of extinct ancient organisms can be predicted from the rule of parsimony (Appendix 10). In addition the existence of unique structures with very wide systematic occurrence (Appendix 11) can be explained by the concept of similar receptor structures in the different receptor families sharing affinity to certain chemical entities.
Receptors are proteins, some with enzyme function, capable of interaction with other molecules to produce a specific change in properties and function. They range
694
in complexity from rather simple polypeptides to very complicated structures. Similarly the mechanism of signal transduction ranges from simple activation of a transmembrane enzyme to exceedingly intricate series of events including modulation of several intracellular messengers and proteins. The response to elicitors similarly ranges from slight modification of performance to signal transduction with elaborate biochemical and physiological consequences. The specificity of receptors shows large variations. In general intemal receptors related to primary metabolism tend to be highly specific (specialist receptors) whereas external receptors activated by natural products or physical stimuli tend to be less specific (generalist receptors), being activated by many different elicitors. An example of specific receptors is the vertebrate visual system with four different sensor cells, one tuned for night vision and three each tuned for detection of different colors by the presence of three different pigments. The generalist receptors are exemplified by the olfactory and taste receptors. These two systems, although both examples of generalist receptors, are very different. The olfactory receptors of man are able to perceive maybe as many as 32 primary odors, while the taste receptors can detect only four: sweet, bitter, salty and acid (56) of which sweet and bitter are closely related (57). While the olfactory receptors are usually activated by low molecular weight (16-300) apolar compounds, the taste receptors react too on a much broader (typically sweet 75 (glycine) to 21,000 (thaumatin)) array of predominantly polar compounds. For example molecules with spherical or quasi-spherical shape and with about 7 A diameter will give rise to a camphoraceous odor regardless of their chemical composition and stereochemical characteristics. In contrast taste modalities are usually evoked by much narrower structural constraints including demands for specific relative and absolute stereochemical configurations (58-59). An exception to these rough rules is the antennae of arthropods (60). As mentioned insects may detect and react to highly specific pheromone stimuli and there is also evidence suggesting that aquatic arthropods may have evolved very specific and sensitive taste modalities. It should be borne in mind that in order to penetrate the hydrophobic alkyl domain of the bilayer membrane molecules must be lipophilic. Therefore, generally hydrophilic stimulators must interact with cell surface receptors or must be actively transported through the membrane while the lipophilic messengers may exert their effect on the intracellular milieu directly. Three classes of cell-surface receptors are known at present. They are channel-linked receptors, catalytic receptors, and G-protein-linked receptors. The first are transmitter gated ion channels. The second are transmembrane enzymes activated directly by their ligands (most are tyrosine-specific protein kinases). The third indirectly activate or inactivate plasma bound enzymes or ion channels via a GTP-binding regulatory protein (G protein) activating one or more small intracellular messengers such as cyclic AMP (cAMP) or Ca^"^. These messengers in turn interact with other target proteins. A short treatise of G proteins is given in Appendix 9. Recently a G-protein-coupled cannabinoid receptor family was identified (61).
695
The cannabinoid-induced response was mediated through Gj and the relative potencies correlate with those of psychoactive cannabinoids. Stimulation of the receptor inhibits cAMP accumulation. This is a prime example on the identification of the receptor responsible for the physiological effect of a typical bioactive natural product. One of the structurally and otherwise best known complex receptors is the nicotinic acetylcholine receptor (nAChR). The nAChR functions as a neurotransmitter in higher organisms. It is the most completely characterized neurotransmitter and ion channel. More than 20 acetylcholine receptor subunits have been cloned from different species. The monomeric receptor protein, functioning as a non-specific ion channel, consists of five subunits composed of a, p, y, and 6 chains. The composition of nAChR has been determined as a2PYS (62). Several detailed structural models have been constructed for this receptor (63-65). Because of the prominence of the nAChR several concepts are illustrated exploiting this receptor (Section 6.2, Appendix 10). Apparently receptors have deep evolutionary roots since many may be traced from highly evolved animals through the phylogenetic system to the lowest procaryotic organisms. The function may well have changed through evolution but the structures are reminiscent of each other. In summary, receptor structure reflects the evolutionary history structurally as well as in the capacity as targets for metabolites from widely different species. 6.
METABOLITES AND RECEPTORS The interactions between metabolites and different receptors are demonstrated for the bryostatins (Section 6.1) and the nicotinic toxins (Section 6,2) including the charatoxins. It is demonstrated that single compounds can interact with different receptors and that different compounds may interact in different ways with a single receptor.
In most cases secondary metabolites have not been pharmacologically investigated and their potential physiological effect is concealed. In the majority of cases where they are known to exert a biological effect detailed descriptions have never been published. Even in the minority of cases, where more detailed knowledge is available the actual receptor influenced has seldom been identified. Even if the mechanism of interaction is known, we cannot be sure that the system studied is the system against which the signal is naturally directed. The causative agent of "Dogger Bank Itch", an occupational dermatitis elicited by the marine bryozoan Alcyonidium gelatinosum, is the dimethyl-(2-hydroxyethyl)sulfoxonium ion (66-69). This eczematous dermatitis is a type 4 reaction (cell mediated allergic reaction). The allergic reaction is certainly not the natural response to this unique toxin but it is the only known bioactivity. In the same way, bromo substituted derivatives of indole-3-acetic acid although they exhibit very potent auxin activity (70), cannot exercise this activity
696
in the deep water sponge Pseudosuberites hyalinus, from which they were isolated (71). Therefore, attempts at unraveling the interaction between most secondary metabolites and receptors, based on the current literature, must necessarily be conjectural and a target for modification as further data emerge. Furthermore the task is complicated by the fact that interactions with receptors may occur at sites different from the active site of the natural stimulant. If this is the case no structural resemblance to the natural elicitor need exist. Such situations are well documented for receptors such as the nAChR. Certain metabolites although generally considered secondary have a rather wide systematic occurrence. Examples and possible explanations are given in Appendix 11. 6.1 Bryostatins The bryostatins have been evaluated against the P388 lymphocytic leukemia system (72) giving in vitro ED5Q values ranging from 0.89 ^ig/ml to 1.8 10"^ |ig/ml representing a factor of 50,000 in potency. Protein kinases C (PKC) catalyze protein phosphorylations, a pivotal step in signal generation and transduction in cells. PKC's are widely distributed in tissue and organs (73). They constitute a family of enzymes with at least seven isoforms. Nearly 100 PKCs are known and a large number of receptors, growth factors and products of oncogenes are PKCs. In connection with the discussion of nAChR's it may be noted that PKC is involved in the regulation of this ion channel (74). Bryostatin 1 and 2 (75) bind to the PKC complex, activate partly purified PKC and enhance the acute release of prolactin as well as prolactin synthesis. Bryostatin 1 induces stretching in GH4C5 cells, while bryostatin 2 has little if any effect (76). Accordingly different activators may elicit different cellular responses by altering the substrate specificity or activating multiple forms of PKCs. In the human promyelocytic leukemia cell line HL-60, bryostatin 1 enhanced the phosphorylation of the same proteins as phorbol 12,13-dibutyrate. In addition bryostatin 1 in a concentration of 6 nM caused the appearance of two new phosphorylated proteins after 30-min. exposure (77) and directly stimulates bone marrow progenitor cells to form colonies and functionally activate neutrophils. Thus bryostatins are mimics of the multipotential granulocyte-macrophage stimulating factor (78). In addition to these effects the bryostatins have a series of other pharmacological activities which have been investigated in considerable detail (79-85). This example serves to illustrate the complexity and similarity of the array of receptors often associated with the biological activity of a set of secondary metabolites. It is noteworthy that the activity of bryostatins against the P388 system is so differentiated. Although this is hardly the "natural" receptor(s) at which the bryostatins are aimed, it does demonstrate a general principle, namely that in an array of
697
1 : R r (^
. R^ = H
2 : R = R' = H
Scheme 18. Bryostatin 1 [1] and bryostatin 2 [2] from the marine bryozoan Bugula neritina.
secondary metabolites one or a few are specifically optimized to recognize a certain developmental stage of a target receptor. P388 is a murine cell line and as such must have highly developed receptor systems. It could be significant that the highest activity against this system is exhibited by minor bryostatins (bryostatins 7 and 11). From the ecological context it would be anticipated that the main targets of the bryostatins were biologically less developed organisms. Consequently they would possess receptor targets against which the major bryostatins are optimized. In fact the target receptor could conceivably be the relevant protein kinase enzyme since this class is of general occurrence and known to interact with the bryostatins as mentioned above. A general class of agents against this type of receptor must consist of an array of structurally related compounds, since the enzymes have not only undergone modification in the phylogenetic evolution, but in addition also exist in isoforms. Furthermore, different effector mixtures may elicit different responses. The activity of this system is delicately tuned and regulated meaning that differential perturbations of the activity of some components may influence specific physiological mechanisms. 6.2 Nicotinic Toxins In the light of the evidence discussed under nAChR's it is not surprising that these have been the target for numerous natural products (Scheme 19). These include metabolites from bacteria (tetrodotoxin), cyanobacteria (anatoxin), dinoflagellates (saxitoxins and gonyautoxins), algae (charatoxins), annelids (nereistoxin), corals (lophotoxins, 86), molluscs (tetramine, 87-89), higher plants (all the curaremimetic
698
toxins such as nicotine, d-tubocurarine etc.), frogs (histrionicotoxins, 90), and snakes (bungarotoxins, coprotoxins, etc.). It seems safe to predict that many more will turn out to interact with these systems in even more subtle ways than realized at present. Investigations of the green characean alga Chara globularis served to identify two sulfur containing secondary metabolites, charatoxin I (ChTx I) and charatoxin II (ChTx II) (91-94). ChTx I acts on the nAChR of Torpedo and bee brain binding to another site than a-bungarotoxin and the related nereistoxin (Scheme 20) (95). ChTx I and II had equal toxicities against Musca domestica and Drosophila melanogaster, but ChTx I was twice as active as ChTx II against the weevil Sitophilus granarius (96). Conversely, ChTx I was 200 times less active than ChTx II in inhibiting photo-
0 0^NH-R3
0^ ^CHa ®J.
XM>=NH2
kH
CHO
Scheme 19. Some examples of nicotinic toxins. Anatoxin [1], saxitoxins [2], lophotoxin [3], and nicotine [4].
synthesis in the diatom Nitzschia palea (97). These results may reflect the difference in structure of the different receptors and illustrate the specificity of the interactions and the degree of sophistication involved.
699 S-S NMe2
S-S S-Me
3/S.^ $-Me
Scheme 20. Structures of nereistoxin (4-dimethylamino-l,2-
1.
PERSPECTIVES - CONSEQUENCES The variety of predictions based on the theory of adaptive variation in metabolism is illustrated within genetic research {Section 7,1), the production of natural products (Section 7,2), chemotaxonomy (Section 7,3), receptor studies (Section 7,4), and medicinal chemistry (Section 7,5). Contemporary procedures are subjected to criticism and reservations in the light of the present theory. Some of the examples are dealt with in this chapter.
7.1 Genetic Research A major challenge in molecular genetics is to unravel the events leading to genetic expression of one trait or another i.e. regulation of gene expression. The tools of modem natural products research may furnish novel and efficient approaches to the study of molecular genetics. Using the current technique of bioassay directed isolation, it is feasible to select any defined biological or biochemical activity provided a bioassay can be set up. In this way inducers of expression of genetic information may be identified. It is of course possible that the inducers identified in this way may actually be precursors of the real inducers formed by metabolic activation of these precursors. This consideration, however, is merely of academic interest since the activity sought for is identified. If desired, further investigations may easily be devised to settle this question. With access to the natural inducer and the modem organic chemical potential of modification and tagging, the process leading to gene regulation can be studied under nearly ideal conditions. In addition many, even extremely complex, natural products can be acquired isotopically enriched from biogenetic feeding experiments. Stmcture-activity studies may be performed to allow deeper comprehension of the molecular mechanism of deregulation and the actual stmctural changes following the interaction of inducer and repressor. One of the advantages of studying secondary metabolites is that they are not normally indispensable for the maintenance and propagation of the cell. As a consequence they can be totally absent in the cell prior to stimulation greatly facilitating the analytical procedures required. In other words the response must often be an on/off reaction in contrast to that to primary metabolites. Here the response is detected as a fluctuation of small concentrations which furthermore is intricately connected to the biochemical regulation of intermediary metabolism.
700
7.2 Commercial Natural Products In commercial exploitation of natural products it is crucial that the natural product or product mixture is optimally expressed in the crops. In cases where the producing organism is cultivated the protection against natural challenges may result in cessation of secondary products synthesis. Accordingly the most economic culture conditions are not always the otherwise optimal growth conditions. In fermentation technology the timing of appearance of the product, the rate of accumulation, and the dependence on nutrients and environmental factors are of utmost economic importance. The enormous amount of empirical information has only to a small extent been explained in terms of physiological and genetic mechanisms (98). It is well known that many organisms are retarded in physical growth in periods of intense secondary metabolite production. This phenomenon is presumably related to the energy metabolism being partly engaged in the production of the secondary metabolites leaving less energy for growth. The logical answer, from a productive point of view, is to artificially supply excess of nutrients to reduce the energetic restraint. In addition to this such cultures are often selected to continuously express the natural products in spite of the lack of challenges. Because of these artificial conditions it might be argued that such systems fall outside the present context. However, as the genetic set-up is still almost identical to the wild types, an optimum production of the desired products may conceivably be procured by varying growth conditions and stimulus.
Scheme 21. Structure of the A-factor.
In streptomycetes, the A-factor, 2-/50caproyl-3/?-hydroxymethyl-Y-butyrolactone, appears to activate streptomycin production, resistance, and sporulation. The compound is secreted by streptomycin producing strains of S.griseus and S. bikiniensis (99). This may then be an effector reflecting the physiological state of the cell and modulating the activity of a regulating protein. The genetic basis for regulation of this substance has been investigated (100). These predictions are most easily tested in cultures of microorganisms because of the short generation time and the ease of manipulation of culture parameters. In the light of the present hypothesis it is hardly surprising that many bacteria cultured under optimum growth conditions express no or insignificant amounts of secondary
701
products. At present this problem is tackled by empirical methods in the fermentation industry. Culture conditions developed by trial and error will often allow establishment of production cultures. This principle is used together with selection of strains optimized for production of the desired product. The present hypothesis anticipates a more scientific and rational approach by utilization of the natural elicitor to provoke expression of the product. At the same time this approach may have the added advantage of enhancing the product synthesis specifically with minimal expression of undesired by-products. 7.3 Chemotaxonomy Chemotaxonomy is based on the assumption that biological relationship is reflected in chemical relationship in the secondary metabolites. This is a reasonable hypothesis since related organisms must have related genomes and accordingly there is a high probability that at least some of the genes coding for secondary metabolites are preserved. As data have accumulated chemical characters have been widely included in the discussion of the systematic position of many organisms especially plants where the amount of data is now vast but also in groups such as sponges (101102). The message of the present hypothesis is, however, that the genetic potential of any given organism for biosynthesis is (almost) never fully expressed. Seemingly infinitesimal modifications in the ecological conditions may completely alter quantitatively as well as qualitatively the pattern of secondary metabolites expressed. Unfortunately chemotaxonomic work is usually performed on collections of organisms from only imperfectly characterized ecological habitats. Consequently the population studied may not express the optimum range of markers. In accordance with this view the absence of certain secondary metabolites can only be considered as tentative arguments until reasonable certainty is established that conditions stimulating their expression have been met. Another complicating factor is that some secondary metabolites have morphogen activity (103-104). Homarine isolated from tissue extracts of the hydroid Hydractinia echinata and the anthozoan Anthopleura when applied to whole animals prevents metamorphosis from larval to adult stages and alters the pattem of adult structures (104). Homarine and trigonelline, also present in the hydroid, prevent head and stolon formation when larvae are incubated in 10-20 \xM solutions of either. If the pyridine derivatives are applied as a pulse during metamorphosis they result in stolon formation by a large amount of the available tissue. Curiously enough both compounds are apparently present in mM concentrations in the tissue indicating that they must be compartmentalized. These results necessitate the conclusion that secondary metabolites may at least to some extent control the morphology of organisms. As much taxonomy is still based primarily on morphology, these phenomena may have relevance in certain taxonomical studies.
702
7.3.1 Chemotaxonomic markers - sponge systematics According to the view represented above the most reliable molecular candidates for taxonomic use are metabolites with permanent functions. Most primary metabolites, however, have functions so basic that they are ubiquitously present in all living cells (ATP, amino acids, etc). Some metabolites fortunately have vital functions even though they are at least to some extent species or taxon specific. Examples of the latter category include sterols of marine sponges. These sterols, which have been much investigated due to their high incidence of unusual structural features, are believed to be constituents of the bilayer membranes of sponge cell walls. Concurrently with the unique sterols many sponges contain unusual fatty acids in their phospholipids. Analysis of the sterol composition has yielded information resulting in what appears to be a consistent character in sponge systematics (101). Furthermore, sterol pattems have been shown to be sufficiently independent of seasonal, geographical and populational variation to allow identification of species (102). Undoubtedly these chemotaxonomic markers will give rise to revisions in sponge systematics as data accumulate. In addition, by use of the rule of parsimony (Appendix 10), it is possible, at least tentatively, to construct phylogenetic trees based on models of ancestors of the present sponge taxa. This may prove an interesting endeavour since little is known of sponge phylogeny, not even whether the phylum Porifera is monophyletic. Attempts to deduce taxonomic relationships based on content of secondary metabolites have only met with very limited success (105). Admittedly there are many pitfalls in this area as demonstrated by the present work and by the largely unexplored symbiotic associations between sponges and microorganisms. In spite of these^ obstacles careful analysis will often give a good impression as to whether a certain metabolite or class of metabolites are sponge derived or symbiont derived. Perhaps the lack of success merely reflects the formulation of questions since it is not at all evident that a correlation exists between position in the systematic hierarchy and the ability to synthesize certain secondary metabolites. It is much more likely that this information has cladistic relevance. If two species produce the same unique metabolite or even biochemically closely related unique metabolites we may anticipate taking into account all the provisions mentioned under the discussion of the rule of parsimony (see Appendix 10) that the two species originate from the same ancestor and we may construct a model of this ancestor with the ability to produce these metabolites. Naturally, these models must also be subjected to evaluation regarding all other important characters known. It is noteworthy that some sponges have peculiar membranes. Not only do they possess the highly unusual sterol composition referred to above but they also have a highly unique fatty acid composition in the membrane phospholipids. The fatty acids are long chain fatty acids (LCFA, C24 - C3Q, earlier referred to as demospongic acids) often polyunsaturated. In addition some sponges have branched chain fatty acids and some even have terpenoic acids in their membrane phospholipids. It has been suggested that the abundant terpenes in some sponges are membrane constituents. The
703
only other living organisms with documented isoprenoic membrane Hpids are the archaebacteria which do not synthesize fatty acids. Furthermore some sponges are known to contain unique nucleotides in analogy with the archaebacteria. Accordingly we cannot exclude that some sponges evolved from the archaebacterium kingdom. 7.3.2 Subspecies and chemical varieties Taking the Chilean study of Plocamium as an example (see Appendix 5), the chemistry of different populations might differentiate some subspecies or chemical varieties. This raises the question of chemical variations. Undoubtedly at present this category should be abandoned because of lack of data to substantiate the classification. After all it is a question of definition how and whether species should be subdivided. There is no doubt that at least some chemical variants have little divergence in the genetic material. The problems connected with the phenomenon of "red tide", the proliferation of dinoflagellates of the genus Alexandrium, are apparently increasing in European waters especially. Blooms of these organisms render molluscs unfit for human consumption owing to the accumulation of neurotoxins belonging to the saxitoxin group. These toxins are the causative agents of paralytic shellfish poisoning (PSP). The PSP phenomenon has serious implications for human health, economy, and in environmental connections. In several cases, however, the organism has been detected without the toxic manifestations associated. This has given rise to the concept of non-toxic clones. The explanation, however, may as well be that the populations in question do not express the paralytic shellfish toxins because of absence of the correct elicitor. Actually it has been shown recently that a clone of Alexandrium fundyense exhibits dramatic changes in toxin production depending on the growth conditions when grown in semi-continuous cultures (106). Under severe nitrogen- and phosphorus-limited conditions the toxin composition was dominated by one or two epimeric pairs while the toxin profiles became more heterogeneous at higher growth rates. These findings were thought to reflect differences in cell physiology and culture modes rather than inherent differences in toxin biosynthesis. Furthermore, it has recently been shown that cultures of Alexandrium ostenfeldii react to chemical elicitation by changes in toxin profile in some cases resulting in powerful enhancements of the toxicity towards higher organisms (107). The latter interpretation portrais a quite serious situation since the clones may commence toxin production without waming as a response to presently unknown stimuli. Recently a toxin producing bacterium, Moraxella sp., has been isolated and cultivated from Alexandrium tamarensis (108). Under ordinary culture conditions the bacterium produced only saxitoxin. Cuhivation under starvation conditions greatly enhanced toxin production even if the growth was severely restricted. Moreover, the main toxins produced under these conditions were gonyautoxin 1 and 4 identical to the major toxins of A. tamarensis. Clearly further investigations are warranted to clarify the precise aetiology. A series of related problems exist, diarrhetic shellfish poisoning (DSP), amnesic shellfish poisoning (ASP), neurotoxic
704
shellfish poisoning (NSP), ciguatera fish poisoning, and several with less firmly established aetiology.
7.4 Receptor studies Receptors have evolved in families. In cases where the secondary metabolite family is directed against different developmental stages of such receptors, worthwhile investigations concerning receptor structure, function and phylogeny can be devised. An example is given by the recent identification of the cannabinoid G-protein-coupled receptor. The description of this work is a good example of the techniques available in this line of research (61). 7.5 Medicinal Chemistry The aim of the medicinal chemist is to pursue receptor specific agents. In order to minimize physiological side effects the specificity must be high and interactions with other receptors as small as possible. Many natural products have performed satisfactorily under this set of conditions. Accordingly, after some years with predominantly synthetic product research, major ventures are now again initiated to identify new drug candidates from secondary products. In the intervening period the major effort has been devoted to screening of arbitrary synthetic products or to synthetic elaborations on naturally derived lead structures. In the first case success is almost purely serendipitous. The last approach at least is based on the ideas of structure-activity relationships and a great deal of empirical knowledge but the chances of discovering radically new activities are low. Taking the complexity of receptors and the lack of detailed structural knowledge into consideration this is to be expected. In the case of ChTX I mentioned above (Section 6.2) the toxin was synthesized (109) and a series of functional derivatives was prepared (96). However, none of these derivatives performed better than the natural products but most considerably worse! Another attempt to produce insecticidal agents by variations of ChTX I and II resulted in good synthetic methods for this type of compounds but no insecticidal activity (110-111). Perhaps because of experiences of this type the interest apparently is in the process of retuming to the natural product area once more in order to gain access to new lead structures. The strategy adopted is mostly large scale screening of extracts of organisms. This is not an optimized approach. Even if the screened organisms have the genetic ability to produce compounds of interest, there is no assurance that the compounds are expressed at the time of collection. A more rational approach would be to select, from knowledge of the receptors involved, organisms that would be likely to produce agents targeted at these receptors. This can often be done by careful ecological studies. A short cut to pinpointing organisms with active compounds is the utilization of ethnopharmacological information. A surprisingly high proportion of the organisms selected from such knowledge turn out to contain active compounds (57,112-117). Once an organism producing interesting activity has been identified, chemotaxo-
705
nomical considerations may often serve to select other organisms for study. The latter principle has the drawback, however, that new compounds encountered are most often variations over the same lead structure and at best produce only marginally better compounds. Stimulation of the selected test organisms and analysis of the family of receptor specific agents released as response may give good leads to the desired type of compound. Often the best lead is likely to be found among the minor metabolites since the first priority of most organisms must be interaction with organisms less developed than Homo sapiens. The main metabolites are therefore often directed against less developed representations of the target receptor. If large scale screenings are performed, all possible care should be taken to select the collection localities in such a way that the maximum evolutionary and ecological pressure is operating. In the case of marine sponges Green (118) showed that toxicity towards fish increased with decreasing latitude. On the other hand Bergquist and Bedford (119) found that a higher percentage of temperate (87%) as opposed to tropical marine sponges (58%) gave rise to antibiotically active extracts. The two situations may be radically different since potential predation by fish may be most prevalent in a tropical reef setting with large species diversity and nutritional limitations. The risk of bacterial infection may well turn out to be more predominant in temperate waters. In this connection it should be borne in mind that very little is known about aquatic bacterial populations. Only a tiny fraction of the bacterial diversity is known (120) and more than 80% of the extant microbes remain undiscovered (121). 7.5.1 Antiviral agents Among high priority areas of drug research antiviral therapy is prominent. The interest of the medicinal community is fueled not only by the progression of the AIDS syndrome but equally well by a range of spreading and uncontrolled virus infections such as Herpes-ll (genital Herpes) not to mention the common cold and recurring influenza epidemics. According to the theory of adaptive variance, the researcher should look for organisms challenged with viruses in order to identify antiviral agents. Preferably the organisms should be challenged by the target virus or a closely related one. Considering the species specificity of many virus this may prove impossible. The next best gamble is to select an organism challenged, either naturally or deliberately, with any virus capable of infecting the organism. It is quite possible that the aquatic environment and ecosystems are more dominated by the presence of virus than the terrestrial. This may be a function of the ease of dispersion as compared to earth and air. Suttle et al (122) have shown that by concentrating seawater, originally containing roughly 10^ to 10^ virus particles per ml, a medium unsuited for culturing phytoplankton was obtained. Owing to viral pathogens important marine primary producers such as diatoms, cryptophytes, prasinophytes and chroococcoid cyanobacteria cultures showed reduction in primary productivity by as much as 78%. Since the situation of the phytoplankton is of such nature as described
706
above, the remaining marine organisms are probably subject to comparable problems in their natural environments. However, as they seem to do quite well if left undisturbed they must have some inherent defence against viral infection. At least in some cases this defence is undoubtedly composed of secondary metabolites. As described above some sponges react to grafts by producing cytotoxic agents. These sponges would presumably react by viral challenge by producing antiviral compounds. This principle may be quite general: To induce agents directed against certain conditions the target organism must be stimulated with the appropriate stimuli characteristic of the challenging situation. Apparently many functions of secondary metabolites in lower organisms are replaced by the intricate and efficient performance of the immune system in higher organisms. However, recent results have revealed that the function of the immune system can be manipulated by the administration of certain secondary products as e.g. immunosuppressive and immunostimulating agents. There is a priori nothing strange in the existence of such activities. If organisms are to cope with the assault and defense systems of most competitors and predators it seems logical that the most developed system for handling extemal non-self agents cannot have been completely passed by. Especially organisms capable of parasitism must have developed strategies to circumvent the immune defense of their hosts. 8.
CONCLUSION
The existence of secondary metabolism is an insurance for propagation of the species in a complex and hostile environment. A certain amount of energy is invested in the genome and occasionally in reserve metabolites to be respectively expressed or employed in case of future emergencies. All metabolites participate in reactions aimed at the preservation of the individual or the species. The secondary metabolites have functions parallel to those exhibited by phytoalexins, only the origin and nature of the eliciting stimuli are diverse. The hypothesis is in harmony with current genetic mechanisms. Although most contemporary natural products investigations are too inaccurate to support the theory in detail, numerous analyses are in agreement with the predictions based on these principles. Receptors are highly conserved structures having evolved in families. It is thus not unexpected that secondary metabolites occur in arrays of related compounds in order to be capable of interacting with different evolutionary models of a target receptor. The immediate consequences in applied natural products research and biochemistry are legion. Most of the predictions lend themselves to experimental verification or falsification as for instance the notion that agents against certain conditions may be expressed by eliciting an appropriate target organism with the stimulus characteristic of the challenging situation. Admittedly the hypothesis is conjectural and the validity can only be assessed by its ability to successfully predict experimentally verifiable parameters within its domain. It is hoped that research based on a unified approach to the understanding of secondary
707
metabolism will be stimulated as a result of this hypothesis.
Appendix 1 Compartmentation Many enzymes are located in specific compartments such as mitochondria, cell wall, endoplasmatic reticulum, Golgi apparatus, and lysosomes restricting the possible substrates, effectors and other enzymes with which they interact. A well known example is the membrane bound enzymes which can only exert their effect on molecules located very near the respective membranes. This set-up allows different reactions which would otherwise compete to occur simultaneously in different parts of the cell. In this connection it should be remembered that access of metabolites to the different cellular compartments is usually intensively regulated and controlled. Unfortunately, in general, the definition and access-restrictions for the diverse cellular compartments are incompletely known at present. A good example of compartmentation, although not yet completely understood, is the apparent segregation of isoprenoid biosynthesis in plants (123). P-^'^C]Mevalonate was easily incorporated in squalene, phytosterols, P-amyrin and ubiquinone but not in the plastid constituents P-carotene, chlorophyll phytyl group or plastoquinone. The opposite result was obtained with ^^^€02. Another very enlightening example is the the accumulated evidence of the function of the plant vacuole as a muhifunctional compartment (124).
One important implication of the compartmentation, however, is the restrictions imposed on chemical compounds interacting with specific receptors. It is not sufficient for a potential bioactive substance to possess the shape and polarity demanded by the receptor. It must also be capable of entering the receptor envirormient. This may mean that the compound will have to penetrate the bilayer membrane and enter the appropriate compartment of the receptor. Throughout this journey the compound may well traverse areas of high enzymatic activity and thus be subject to metabolic transformations which may activate, deactivate or be without effect with regard to the influence on the target receptor. The possibility of compartmentation allows some organisms to sequester nutrient derived metabolites which would otherwise be metabolized or excreted as such or in slightly changed form. The detention of exogenous compounds may occur generally as for example lipophilic compounds (DDT, dioxins) in the lipophilic compartments of the cell or organism or in specific organs. The latter mechanism is especially well illustrated in marine nudibranchs where hundreds of investigations have served to identify specific nutrient-derived compounds. These compounds have a deterrent function towards would-be predators. Some crustaceans retain 2,6- and 2,4-dibromophenols from their diet (58-59). Whether any biological function is
708
associated with these compounds is unknown but it certainly makes the crustaceans unacceptable to some consumers. Shellfish concentrating ingested toxins (e.g. saxitoxins, brevetoxins, domoic acid) giving rise to paralytic, diarrhetic and anmesic shellfish poisoning etc. are prominent examples as are ciguatoxic fish.
Appendix 2 The Nature of Stimuli Stimuli or elicitors may be of physical origin, such as temperature, light intensity, salinity, pressure, mechanical activation (grazing or injury) or complex combinations of physical parameters. Nevertheless, the majority must be chemical consisting of primary or secondary metabolites and/or derivatives. Depending on the point of view and level of discussion, it might be argued that all elicitors are chemical, since even a physical stimulus must be translated into die chemical vocabulary to exert its effect. Some elicitors are directed against other species inhabiting the ecosystem e.g. a simple monoterpenoid mixture of the marine bryozoan Flustra foliacea apparently prevents fouling of the colony (49). Others are primary or secondary intracellular or intercellular messengers in a given organism e.g. Ca^"^ or hormones. Some are species specific e.g. sexual attractants of brown algae. Some monocarpic plants seem to produce "death hormones" transported from the fruits to the vegetative parts of the plant (125). The physiological action is termination of growth, activation of senescence, remobilization of nutrients and finally death. In the pea, Pisum sativum, 4-chloroindolacetic acid has been implicated as death hormone. Certain species of gorgonians have been observed to practice density regulation. Neighboring colonies seem to exchange chemical signals resulting in the dissolution of the smallest colony brought about by its own lytic enzymes (126). Neither the signalling system nor the mechanism enabling one tissue to force another to self-destruction has been explained. Presumably the skin of some holothurians disintegrates and dissolves by a related mechanism when disturbed (127).
Actually the literature abounds with more or less specific information on chemical messengers regulating the ecological relationship among organisms (128). Recent compilations of mediators of microbiological and algal origin cite abundant examples (129-130).
709
Appendix 3 Compatibility with Genetic Mechanisms Admittedly there are very few rationalized connections between evolution of the genome and of the phenotype (131). However, in the case of simple systems some evidence is available, hi bacteria the expression of the structural genes in an operon (structural genes plus operator genes) may, somewhat simplified, be envisaged as the inactivation of a repressor (binding to the operator gene and preventing transcription) by an inducer substance (of external or internal origin). This type of gene control is termed negative regulation since binding of the regulatory protein suppresses transcription. The repressor is continuously produced by a separate regulator gene. As regards the p-glucosidase system, lactose functions as the inducer binding to the repressor leaving the operon free for transcription of its several structure genes. Metabolism of lactose in Escherichia coli is accomplished by the lac operon where an operator regulates transcription of the genes lacZ, coding for p-galactosidase, and lacY, coding for lactose permease. In the cell p-galactosidase transforms lactose to glucose and galactose directly entering the energy-metabolism and into allolactose. The latter compound depresses the operator resulting in transcription of lacZ and lacY. From an E, coli strain where most of the lacZ gene had been deleted, a mutant capable of growing on lactose was isolated by providing an additive inducing the lacY gene. Interestingly enough these bacteria proved to have mutations in an entirely different operon (EBG, evolved P-glucosidase operon) distant from the lac operon. They were double mutants expressing a structural mutation of the ebg enzyme (different from p-galactosidase) enabling this to hydrolyse lactose. Normally, the transcription of the ebgA gene for this enzyme is repressed by the operator gene ebgR, not induced by lactose. Consequently the ability to grow on lactose requires a mutation in ebgR as well. Some ebgR mutants were constitutive while others were induced by lactose. Growth now depended on the permease-inducing additive. When selecting this mutant for its ability to grow on lactulose, further ebg enzyme mutants capable of utilization of lactose as well as lactulose (and lactobionate) were isolated. One of the mutants metabolized lactose to allolactose. Since the latter induces the lacY gtno an entire system of lactose utilization has evolved (131). While eukaryotes do not have operons, a model for eukaryotes of regulation of gene expression, although much more complex, is comparable in principle to the one described above. The structural genes are believed to be governed by adjacent receptor genes. The latter are activated by the product of an integrator gene which is tumed on by a sensor gene. This again requires activation by a stimulus which may be a hormone, an inducing factor released from another tissue or an extemal inducer. An example where the plant pathogenic fungus Nectria haematococca detoxifies the phytoalexins, pisatin, and maackiain, from the host serves to illustrate the principles involved (132).
710 Me-0
1
2
Scheme 4. Structures of the phytoalexins pisatin (1) and maackiain (2) from peas directed against the pathogenic fungus Nectria haematococca.
A family of genes codes for a family of P-450 cytochrome monooxygenase enzymes. One subfamily is pisatin demethylase. This enzyme is required for pathogenicity on pea, Pisum sativum, and the enzyme shows strong specificity for pisatin as its inducer and substrate. Another subfamily of enzymes effects the hydroxylation of maackiain, a phytoalexin from chickpea, Cicer arietinum, also a host of this fungus. The fungus may be pathogenic on both species of pea, one of them or neither. It is believed that these genes evolved from some ancestral genes conferring the ability to detoxify phytoalexins. During coevolution of pathogen and host, the genes evolved to express highly specific enzymes. In short, there is a common theme of responding to changes which may serve as a model of positive gene control in procaryotes as well as eucaryotes (133): 1. A signal, normally the intra- or extracellular concentration of a small molecule, is perceived by a sensor. 2. The signal is transmitted to a regulatory (activator) protein. 3. The conformation of the activator protein is changed by signal transduction, either covalent or noncovalent. 4. The conformationally changed activator protein binds to a specific DNA site. 5. This DNA-protein interaction catalyzes the binding or activity of RNA polymerase to facilitate transcription initiation at the promoter. Obviously, even if specific details of the generalized mechanisms of gene expression under negative as well as positive gene control may in some cases be omitted, and some details may imdoubtedly be much more complex, the mechanisms generally allow secondary products to assume the role as inducers of gene expression effecting the synthesis of specific chemicals. Furthermore, the mechanisms indicate many points where secondary metabolites could conceivably interact with the machinery to interfere with normal regulatory procedures. Even though, as mentioned, higher organisms have far more elaborate mechanisms for regulating the expression of genes there is no obvious reason why evolution following the paradigm of this simple model could not take place.
711
Appendix 4 Microorganisms The dependence of the expression of secondary metabolites on external parameters in microorganism cultures is well known in the fermentation industry. The rationale behind diis phenomenon has, however, apparently not been recognized resulting in empirical "trial and error" approaches to optimization of yield in production cultures (see Section 7.2). In these cases it should be borne in mind that the physical and chemical parameters are as a rule far removed from the natural conditions, and in this way the systems are artificial and not necessarily primarily governed by the Darwinian laws. In an investigation of aflatoxin production of the fungus, Aspergillus flavus, Zeringue and McCormick (134) showed that cotton leaf volatiles had a profound influence on the production of aflatoxin in the cultured microorganism. It was found that some components (C5-C9 alkenals) completely inhibited the growth of A. flavus, other stimulated the growth, while 3-methylbutanol and 3-methyl-2-butanol inhibited growth by 20% but increased aflatoxin Bj yield by 1.5 to 2-fold. Analogously, but to a much lesser degree, c/5-2-hexenol, myrcene and ocimene retarded growth but enhanced toxin production. Aflatoxin B| is the most potent carcinogenic known substance occurring naturally. Interestingly enough 3-methyl-2-butanol is one of the main characteristic odors of living colonies of A. flavus. As an illustration of the complexity of these interactions it may be mentioned that treating cotton leaves with cell-free extracts of ^4. flavus induces five phytoalexins (lacinilene C, lacinilene C 7-methyl ether, scopoletin, 2-hydroxy-7-methoxycadalene, and 2,7-dihydroxycadalene) in the leaves (135).
Scheme 6. Aflatoxin Bl, the most potent carcinogenic compound known.
712
J^&y
Me-Q
rXoH OH
6 Scheme 7. Five phytoalexins, lacinilene C [1], lacinilene C 7-methyl ether [2], scopoletin [3], 2hydroxy-7-methoxycadalene [4], and 2,7 dihydroxycadalene [5]) induced by A. flavus in cotton leaves. The antibiotic cortalcerone [6] is induced in the fungus Corticium caeruleum by various stimuli.
Agar cultures of the fungus Corticium caeruleum will only express the antibacterial cortalcerone (6, scheme 7) after activating treatment with toxic vapors or when being subjected to supraoptimal temperatures (136). Appendix 5 Cystoseiraceae The brown algal family Cystoseiraceae contains 15 genera including Cystoseira and Bifurcaria. Especially Cystoseira has been extensively studied for chemo
Scheme 8. Structures of eleganolone and elegandiol from the cystoseiracean alga Bifurcaria bifurcata.
taxonomic reasons (137). In the Mediterranean alga, Cystoseira elegans, the diterpenoids eleganolone, 2,3-epoxyeleganolone and elegandiol varied with a maximum content in April declining in May and absent in June and July (138).
713
The diterpenes are thus present in spring when the alga grows up and disappears before the alga vanishes in the summer. This is expected if the diterpenoids act as feeding deterrents. If the compounds are stored they must be degraded or released when the plant reaches a certain age. Interestingly enough C. elegans is the only Mediterranean Cystoseira species containing these diterpenoids (among eight analyzed). The Atlantic, perennial Bifurcaria bifurcata collected off Piriac (47^231^; 2^3 rW) at the French Atlantic Coast contains the same three compounds (139-140, in addition to 7,8-epoxyelegandiol (141) and exhibits a doubling of the eleganolone content from January (1.5% of dry weight) to July (3% of dry weight) (139). The same species collected off Roscoff (48^44?^; 4^0^) (142-143) gave eleganolone, elegandiol, and five new diterpenes. Perhaps less intriguing a collection off Oualidia (32^481^^; 8^57^), Morocco (144-145) gave 12-S-hydroxygeranylgeraniol, the corresponding 1 -oic acid and (2£,6£, 10£, 13£)-3,7,11,15-tetramethylhexadeca-2,6,10,13,15pentenol. Apparently the three different collections had all structurally different main compounds. Obviously the existence of three genetically different populations cannot be ruled out, however, this explanation seems less likely for the geographically neighboring French populations. Plocamium Geographical variations are well documented. The red alga Plocamium cartilagineum elaborates halogenated monoterpenes. Specimens from Cadiz, Spain and Elephant Island, Antarctic Peninsula have two cyclic compounds. Those from Isle of Wight, U. K., Monterey Bay, Califomia, Overton, S. Wales, U. K. and James Island, Antarctic Peninsula have five, one, three and four cyclic halogenated monoterpenes respectively. Specimens from the latter three locations in addition have one, three and two acyclic halogenated diterpenes, respectively, while specimens from the remaining three locations were devoid of acyclic compounds. Specimens from Covadonga Roadstead, Antarctic Peninsula, La Jolla, Califomia and Whidby Island, Washington contained one, twelve, and two halogenated acyclic monoterpenes, respectively, and none cyclic (146). The same authors found spectacular variations in cyclic halogenated monoterpenes from different populations of P. cartilagineum collected along the Chilean coast (1600 km's distance between the extreme locations), but were unable to detect acyclic compounds in any of these collections. Caulerpalean algae Another example is the content of caulerpicin in the green alga Caulerpa racemosa. Originally the toxin was isolated from a Philippine collection (147-148). The toxin was a mixture of sphinganine derived ceramides (149) with Cjg (32%), C20 (2%), C22 (6%), C24 (35%), and C26 (25%) saturated fatty acid residues. However, a sample of the same species collected off Sri Lanka contained a different mixture (150). The Sri Lankan sample was a mixture of A^-acylsphingosines (sphingosine= A4-
714 £-sphinganine) with approximately equal amounts of C14, Cj^, C22, C24 saturated and C22, C24 mono-unsaturated acids. Thus none of the compounds from the two collections were identical and only the C22 and C24 acyl groups were present in both samples. These results corroborate the "Theory of Micromolecular Evolution" maintaining that for a given taxon invasion of new geographic areas gives rise to modification of the relative importance of characteristic metabolic pathways (151).
Appendix 6 Amathia wilsoni The amathamides (Scheme 10), alkaloids from the Tasmanian marine bryozoan Amathia wilsoni, have been investigated as a function of geographical location of the collected material (152). Populations from the same location did not exhibit qualitative variations in alkaloid content. However, collections from different, geographically proximate locations gave widely varying alkaloid patterns (Table 1.). Only two collections (1 and 3) had an identical main alkaloid, while collection 4 did not give detectable amounts of the main alkaloids of the other samples. Collection 4 had two different alkaloids, collection 3 three, 2 and 5 four, and 1 five. If the amathamides C, D, E, and F of collection 5 are an optimized result of ax. evolutionary strategy for survival, then why does collection 4 only have amanthamide A and B? In other words, if the population from sample 4 is capable of producing a main component in 95% yield with only 5% minor compound, why then does the population from collection 1 then elaborate five different metabolites? Table 1. Variation in alkaloid composition (%) in y4. wilsoni from different locations.
Location
Amathamides A
1 2 3 4 5
6 11 0 95 0
B 1 0 0 5 0
c
D
E
F
63 6 48 0 6
5 0 17 0 26
23 44 35 0 25
0 39 0 0 44
715
R = R' = H
2. R = H
R= Br, R* = Me
5: R = Br
R = Br. R' = H
.Me Br. Br"
o
^ Me OH I
Br
Scheme 10. Structures of the amathamide alkaloids, amathamide A [1], B [2], C [3], D [4], E [5], and F [6] from the marine bryozoan Amathia wilsoni.
If all Specimens from the different locations have the genetic set-up for the production of all amathamides, different parts of the genome are apparently expressed in the geographically different populations. This is the most economic route to deal with varying levels and types of challenges. Miscellaneous bryozoans An analogous example may be found in the investigations of the bryozoan Pentapora fasciata. Extracts from collections at geographically diverse locations had activities against the growth of phytopathogenic fungi, e.g. Gleosporium fructigenum, ranging from potent in vitro inhibition to inactivity (collections from the East Pyrenean area) (153). The authors believe this to rule out bryozoan de novo synthesis and that it rather reflects the variation in epizoonts which are most difficult to remove from the material. The bryostatins, macrocyclic antineoplastic lactones, have been isolated from another common marine bryozoan, Bugula neritina (154). All 15 bryostatins known so far represent variations over the bryopyran skeleton. Different collections exhibited large variations, quantitative as well as qualitative, in bryostatin content. Bryostatin 4-7 without trace of 1-3 were obtained from some northeastern Gulf of Mexico collections. Some eastern Pacific collections had barely detectable amounts of bryostatin 4-7. Bryostatins have been found in other organisms as well, however in each case association with fi. neritina was demonstrated. The marine bryozoan Amathia convoluta contained bryostatin 4-6 and 8, while the marine sponge Lissodendoryx iso-
716
dactyalis showed the same bryostatins together with bryostatin A and B characteristic of this source. Bryostatin 4 and 5 isolated from the marine ascidian Aplidium californicum were beheved also to originate from associated B, neritina. Appendix 7 Corals In a study of bioactivity from Great Barrier Reef scleractinian corals Gunthorpe and Cameron (155) demonstrated four different activities (mouse toxicity, cytolysis of red blood cells, ichthyotoxicity and antibacterial activity) in 91% of the species (53 out of 58 species) investigated. However, the activities encountered varied considerably within different colonies of the same species. Except for negative correlations between antimicrobial activity and presence of immature gonads, and mouse toxicity versus high average monthly air temperature, no correlation could be established neither between the different activities nor between variation in specific activity. The present theory predicts the variation to be time dependant and to correlate with presence or abscence of challenging events such as predators or potential pathogenic microorganisms. In a subsequent study Gunthorpe and Cameron (156) foimd pronounced intracolonial toxicity variations in the corals. They concluded that intraspecific variations occurred between and within colonies: "Bioactivity within a given species is not a static phenomenon". The would-be elicitors have not been explored in these studies but no contemporary theory, except the present, offers any explanation for such apparently erratic variations in bioactivity. Appendix 8 Marine Sponges Blunt et al (157) noted that biological activity is often present in damaged or diseased sponges while healthy members of the same species are inactive. They convincingly demonstrated, in situ, that while grafts from the same clone were tolerated, grafts from specimens from other clones of the same species gave rise to toxin production and rejection of the graft. The species tested were Polymastia sp. and Latrunculia apialis. These species are believed to reproduce asexually and the propagules drift only short distances. In perfect harmony with these propositions it was observed that, in general, grafts from down-current or up-current specimens were tolerated while grafts of other specimens were rejected. In the case of L apicalis all grafts within 3 m of the donor were accepted and all from specimens at a distance of more than 6 m rejected. It was concluded that the clones could recognize self from non-self tissue and that these specimens were able to synthesize toxins. Interestingly enough it was found that L apialis showed enhanced cytotoxicity towards the P388 bioassay for the grafted tissue with the highest activity encountered in tissue near the graft, declining when more remote tissue was tested. The observations were inter-
717
preted to mean that these sponges, inhabiting one of the most stable ecosystems known, are "naive" with regard to previous encounters with competitors and predators and therefore have had no need to activate a chemical defence system. It may as well be fitted into the present hypothesis meaning that the sponges are not especially naive but that they were not challenged at the time of the initial sampling. They may well have expressed the cytotoxic material several times during their lifetimes and in this case may return to rest again after the grafts have been rejected.
HO'^O
Scheme 13. Structures of the sesterterpenes luffariellin A [1], luffariellin B [2], manoalide [3], secomanoalide [4] from Luffariella variabilis, thorectolide monoacetate [5] from Thorectandra excavatus.
In an investigation of the dictyoceratid Palauan marine sponge Luffariella variabilis (158) different specimens were found to contain different mixtures of antiinflammatory sesterterpenes. Of 410 specimens 5.4% contained only luffariellin A [1] and B [2], 86.8% only manoalide [3] and secomanoalide [4] and 7.8% a mixture of all four compounds. Manoalide [3] and secomanoalide [4] were earlier found in the same species (159-160) also from Palau. Clearly, disregarding the possibility of a symbiotic source of the metabolites, these sponges must have the genetic apparatus to express both sets of compounds depending on the immediate set of stimuli. The sesterterpenes manoalide monoacetate [3-25-acetate] and thorectolide monoacetate [5] were identified from another dictyoceratid sponge, Thorectandra excavatus (161)
718
collected near Darwin, Australia. Also in this case it was noted that some specimens gave manoalide monoacetate exclusively, others gave thorectolide monoacetate and still others mixtures of the two compounds. In the case of the dendroceratid sponge Dictyodendrilla cavernosa the sesquiterpene furan pallescensone was found in some specimens while others contained thediterpene lactone spongiane-16-one and the first acid anhydride to be isolated from a sponge, spongia-15,16-dione (162). The authors conclude that "the morphology of the species is more complex than previously supposed". Comparable results were found for a sponge of the genus Eurypon (163). In these sponge studies every care was taken to secure the identification of the material. The results obtained were unsuspected leading to additional concern to secure the correct taxonomy. Incidentally, the two investigations last mentioned were coauthored by P. R. Bergquist, one of the internationally leading sponge taxonomists and the material was collected in New Zealand waters, her domestic area. The risk of misidentification is thus minimal.
H 0-0
Me02C^A=AR 1
R
Me02C-Jwi)_Mg
R = (CH2)i5Me
2 : R = (CH2)i5Me
R = (CW2)i3^*^^^
3 : R = (CH2))i Me L R = {CH2)9/*^/ 5 R -- (CH2)9^"'**-^'''^
7 8 12
H 0-0
O-Me
6
R = lCH2)9^^v^^^/
9 10 11
R = (CH2)i3^^*/ R = (CH2)n/^'*.-''^-^ R = (CH2)i3-^^»v''^/
Scheme 14. Structures of cyclic peroxides from sponges. Chondrillin [1], plakorin [2], 7, 8, 9, and 10 are formally derived from C22 acids, 3, 4, and 5 from Cjg, and xestin A [11] and xestin B [12] from C24 straight chain fatty acids. Chondrillin [1] was isolated from a Great Barrier Reef sponge, Chondrilla sp. (164). A Plakortis sp. from Okinawan waters gave the stereoisomeric Z-compound, plakorin [2] (165). Plakortis lita from Okinawa in addition to chondrillin [1] gave four homologous compounds, 3-6, with the plakorin stereochemistry (166-167), while the same species from Truk Island m addition to chondrillin gave two compounds with chondrillin stereochemistry, 7 and 8, and three with plakorin stereochemistry, 2, 9, and 10 (168). A Xestospongia sp. from Fiji gave xestin A [11]
719 with plakorin sterecx:hemistry and xestin B [12] with chondrillin stereochemistry (169). All compounds exhibit optical activity attesting to their enzymatically assisted generation. All compounds seem to exhibit potent biological activity although a discrepancy exists regarding the cytotoxicity of chondrillin [1] in the P388 leukemia cell assay, where Sakemi et al (4) found an IC5Q of 5 ^g/ml whereas De Guzman et al (168) found no significant toxicity, ED50 > 10 |ig/ml. The same resuh was obtained by the latter authors for 10. The plakorin stereochemical configuration seems to confer the highest cytotoxicity in the P388 assay since 3, 4, 5, 6, and 11 exhibited IC5Q values of 0.05, 0.1, 0.05, 0.1, 0.3 ^g/ml and 12 3 ^ig/ml respectively. In addition xestin A [11] was strongly active at 5 |ig/ml against lung, colon, and mammary tumors where xestin B [12] was inactive in the same concentration. Plakorin itself is a potent activator of sarcoplasmic reticulum Ca^"^ATPase (165). Plakortis lita collected off Truk Island and Okinawa had only one compound in common, namely chondrillin [1]. The amounts were very different since the Okinawan sample had 1 as the main compound (0.19% of dry weight) while the Truk sample had 1 as the least abundant peroxyester (0.0008% of dry weight). The Okinawan C^g derived compounds 3, 4, 5, and C20 derived 6 are homologous to the Truk C22 derived compounds 2, 9, and 10, respectively, which are accompanied by the stereoisomeric 7 and 8. In the case of Xestospongia the C24 derived compound 11 is homologous to the Truk C22 derived 10 and the Okinawan Cjg and C20 derived 5 and 6, while 12 is homologous to 8 from the Truk sample. Furthermore, the activity in the P388 assay shows some astounding variations apart from the discrepancy concerning chondrillin [1], namely the fact that C22 derived 10 and the homologous Cjg derived 5 and C20 derived 6 exhibit a variation of at least a factor 100 in IC5Q (>10, 0.05, and 0.1 ng/ml respectively). Of course this apparent inconsistency may be coincidental and merely reflects that the effect is not evaluated against the receptor at which the products are aimed. The latter notion may fmd support in the fact that the main metabolite in the Okinawan sample is chondrillin [1], the least active compound in the P388 assay. Curiously enough these compounds are apparently not chemotaxonomically related since the plakinid sponges belong to the Homosclerophorida, Chondrilla to Hadromerida, and Xestospongia to Nepheliospongia. It is noteworthy that other plakinid sponges such as Plakortis halichondroides and Plakortis angulospiculatus (170) as well as a hadromerid sponge, Chondrosia collectrix, have cycHc peroxides derived from branched chain acids. Some of these compounds exhibit antimicrobial activity, ichthyotoxicity and cytotoxicity (166). Other examples of unique secondary metabolites or metabolite classes with wide taxonomic occurrence are discussed in Appendix 11. The common crust of bread sponge, Halichondria panicea, was found to contain large amounts of sulfur compounds, notably dimethyl trisulfide (171), to such an extent that the organism was considered hazardous to handle by the fishermen in the area. Samples of this species collected at other localities were found to contain trace amounts of the sulfur compounds.
Actually the literature contains several examples comparable to those described
720
above as pointed out by Christophersen and Jacobsen (172).
Appendix 9 G Proteins Some G-proteins (Gg) relay receptor activation to adenylate cyclase and thereby activate cAMP mediated reactions. In bacteria receptors and adenylate cyclase interact directly. Gg proteins consist of a heterotrimer of three polypeptides where one, the a chain (Gg^j^), binds and hydrolyses GTP and activates adenylate cyclase. Tight p chain and y chain complexes anchor Gg to the plasma membrane. It is believed that the complex releases the active a chain on activation. Another G-protein (Gj) contains the same py chains but a different a chain, G^^, This factor on receptor activation inhibits adenylate cyclase. Still another G-protein (Gp) serves to activate phospholipase C and thereby generates inositol triphosphate (InsP3) and diacylglycerol (DAG) from phosphatidylinositol-biphosphate (PIP2). InsP3 releases Ca^"*" from the calcium sequestering compartment thereby making this multipurpose ion available in the cytosol. The DAG produced in the hydrolysis of PIP2 can act as precursor for arachidonic acid, giving rise to the production of prostaglandins and other lipid signalling molecules, and it can activate protein kinase C (PKC, C because it is Ca^"^ dependent). This activation is transient since DAG is rapidly transformed, either by cleavage to arachidonic acid or by phosphorylation to phosphatidate. PKC transfers the terminal phosphate group from ATP to specific serine or threonine residues on target proteins to produce a variety of physiological alterations.
Appendix 10 Evolution of Receptors It is by now well documented that many key biochemical structures are surprisingly highly preserved during evolution. The two- subunit structure of ribosomes is universal and the stem-loop structures of rRNAs are extremely similar in all RNAs. Also the G-proteins (Appendix 9) seem to be evolutionary related containing similar subunit structure and amino acid sequence. The G-protein transducing from the vertebrate eye couples the reception of a photon by the rhodopsin molecule to the activation of a phosphodiesterase enzyme hydrolysing cyclic GMP. The transducing a subunit is about 65% identical in amino acid sequence to the Gj a subunit. Furthermore the receptors linked to G-proteins also seem to constitute a family of evolutionarily closely related receptors. They all consist of seven-pass transmembrane proteins and include such important receptors as the p-adrenergic receptors, the muscarinic acetylcholine receptor, many neuropeptide receptors, rhodopsin, and the cannabinoid receptor. Even bacteriorhodopsin and receptor proteins used by yeast belong in this
721
family which presumably arose early in evolution. A very important component in eucaryotic cells is the transcription factors. It is well established that the primary control of gene expression lies at the level of gene transcription (173). The transcription factors recognize and bind to regulatory sites in DNA. A limited number of families of site specific protems exist. To start transcription of protein-coding genes at an acceptable rate a transcription factor TFIID and RNA polymerase II with accessory proteins are necessary. TFIID specifically binds to the TATA box, usually located about 30 nucleotides upstream of the RNA start site of protein encoding genes. In addition specific activators are normally involved to accelerate transcription (174). Human functional TFIID has been cloned (175-177) and consists of 339 amino acids. The carboxy terminal 181 amino acids share 80% identity with the analogous protein in Saccharomyces cerevisiae, which is functionally replaceable in vitro (175). The budding yeast Saccharomyces cerevisiae and the fission yeast Schizosaccharomyces pombe have diverged several hundreds of millions of years ago (178) and yet the S. pombe TFIID is 93% identical to S. cerevisiae TFIID in a region consisting of a direct repeat (179). This evolutionary conserved core is found in all TFIID's known imtil now. Relative to human TFIID that of Drosophila is 88% conserved, Arabidopsis-l 83%, Arabidopsis-l 82% (177). The conserved structural motif is sufficient for binding as well as transcription promotion. In contrast, the A^-terminal is qualitatively and quantitatively different in different species. Even in the two TFIID's from the plant Arabidopsis thaliana, where about 200 amino acids share 93.5% identity, the A^-terminal sequences are only 75% conserved (180). There is circumstantial evidence that the TFIID itself is subjected to sophisticated regulatory systems (181). It is very likely that this general and important system is target for many secondary metabolites since interference with this system must have serious implications for the organism affected. Furthermore, even apart from the highly conserved core structural motif, the species specific variant A^terminal section lends itself to biochemical interference by reaction with endogenic as well as exogenic secondary metabolites. In the same way nAChR seems to represent an evolutionary highly conserved structure. This is demonstrated by the observation that nAChR's from so dissimilar organisms as the ray, Torpedo californica, the electric eel Electrophorus electricus, and mammalian muscle exhibit similar antigenic properties (182). Strong sequence homology exists in subunits from material from a mouse cell line (BC3H-1), Torpedo, and skeletal muscle nAChR although the mouse system apparently has a function closer to neonatal muscle nAChR (183). The latter observation indicates that the evolution of the nAChR's is reflected in the embryogenesis. The subunits exhibit a conspicuous homology among themselves, and may have diverged from the same primordial gene and accordingly show highly conserved sequences across species. Thus the mouse a subunit is 80% homologous with Torpedo, 86% with chicken, 95 with calf, and 96% with human a subunits (184). Also nAChR from vertebrates, although different, shows significant molecular similarities to nAChR from insects
722
(185). Organisms like the ciliated protozoan Paramecium^ baker's yeast Saccharomyces cerevisiae and the Gram-negative bacterium Escherichia coli have nonselective, mechanically gated cation-channels. The channel from yeast has conductance comparable to nAChR (186). The voltage sensitive sodium channels from eel electroplax have an amino acid composition remarkably similar to that of nAChR and the repeats from Drosophila sequences are closely comparable to the ones from eel and rat (187). The presence of acetylcholine has been well established in all mayor taxonomic groups of the Plant Kingdom and even though the AChR has not yet been isolated from plants strong acetylcholine binding activity has been demonstrated (188) and evidence for the presence of nicotinic-Iike as well as muscarinic-like AGiRs has been presented (189). In plants as in animals acetylcholine is synthesized by choline acetyltransferase and hydrolysed by cholinesterase (189-190). Although the physiology of acetylcholine in plants will have to await further studies, there is circumstantial evidence that it mediates phytochrome action (188). Acetylcholine has been shown to induce Na"*^ and K"*^ output from intact chloroplasts (191) and to stimulate gemiination of photoblastically positive seeds and inhibit gennination of negative ones in continuous white light (192).
Other ion-channel proteins and other basic receptors as well have long evolutionary histories. In a recent compilation of data on synaptic membrane proteins Betz (193) concludes that although the functional diversity of synaptic protein superfamilies largely arose by divergent evolution of common ancestral building blocks, the families themselves share significant sequence homology and have common ancestors early in phylogeny. Furthermore it was noted that there is a certain heterogeneity in channel properties depending on variable subunit composition giving rise to embryonic forms and adult isoforms. These variations offer different targets for secondary metabolites allowing even more subtle effects to ensue. Naturally the added aspect of species dependent variations in receptor structure allows a rigorous definition of target. The voltage sensitive calcium channels have recently been studied using conotoxins from marine cone snails. The o)-conotoxin GVIA is lethal to fish, amphibians, and birds but not to mammals, while MVIIA is only lethal to fish (194).
Evolution of Secondary Metabolites Phylogenetic evolution The present hypothesis assumes interaction between receptors and secondary metabolites. As a consequence receptor evolution reflects an evolutionary selection in secondary metabolites. The main lines in the chemical evolution of living systems, the RNA-world developing to a breakthrough organism, the ancestor of the progenote from which the three kingdoms, archaebacteria, eubacteria, and eucaryotes evolved (195) are established. Within each kingdom the biochemical evolution is still almost
723
terra incognita although much circumstantial evidence is apparent from contemporary chemotaxonomy. In the case of receptors, as discussed above, and other proteins, especially enzymes, the pedigrees are in the process of evaluation (see e.g. 196), but the low molecular weight substrates of these systems still await the emergence of usable working hypotheses to comprehend the available data. In order to reconstruct the paleontology of secondary metabolites information on past families of metabolites is needed. Although information on biochemical fossils is slowly emerging (197), the area is still largely dominated by speculations. The field where most work has been done and from which the most successful results have been gained so far is the lipophilic chemical fossils (198). In connection with crude oil investigations the pattems of sterols and derivatives have been carefully scrutinized. It has been possible to correlate types of geochemical intermediates with source organisms and to devise detailed pathways leading from secondary metabolites to geochemicals. In a few cases these pathways have been established in detail comparable with well investigated biogenetic pathways. Naturally the correlations are the easier and more reliable the less geochemical changes have occurred. However, even very ancient material may still yield interpretable information. Sterols are relatively stable organic structures, configurational isomerization reactions can in the absence of deep burial take longer than the Phanerozoic, viz. about 6(X) million years, to reach equilibrium. Furthermore, their presence in undegraded crude oil attests to the gentle physical conditions prevailing during their geological lives. In conclusion the sterols are presumably nearly optimal structures for geological preservation. Many other geolipids present in petroleum bear witness to the activity of living organisms in earlier geological eras. A complicating factor in crude oil investigations is the very low abundance of polar compounds. Oil deposits are never found at the location of generation. The oil has imdergone primary and secondary migration from the source rock to the reservoir (199-200). During these migrations, which can be of substantial distance, the geological formations act as a huge chromatographic column retaining the polar constituents (201). This obstacle can be overcome by investigating organic matter in sediments. As the sediments mature the kerogen (insoluble part of the organic material) assumes exceedingly complex hetero-oligomeric/polymeric structures. Although it is possible to degrade such structures, for example pyrolytically, the structural information gained in this process requires elaborate knowledge of the geological as well as the pyrolytic reactions employed. At present this knowledge is not detailed enough to allow unambiguous deductions. Geolipids that have been subjected to diagenetic and catagenetic transformations can be manipulated to yield useful information, while metagenesis degrades organic material to methane and graphite with total loss of structural information. In conclusion, even if ancient molecular structures can be elucidated, and have been, the processes of geological diagenesis, catagenesis and related mechanisms are still too incompletely known to allow definite conclusions. Consequently the deductions regarding the original molecular structures are at best tentative at present. However, as knowledge accumulates and
724
techniques advance it is inevitable that molecular paleontology will establish itself as an integral part of contemporary paleontology. At this stage studies of secondary products from extinct taxons and their evolution and dispersion in the living world may be achieved scientifically. The rule of parsimony Another method of gaining information about the evolution of extinct taxa is the use of the principle of parsimony which demands that no more causes or forces should be assumed than are necessary to account for the facts. The method involves the construction of a model of the extinct organism based on known or suspected descendants. Common traits displayed in each of these descendants are then assigned to the model progenitor. Successful use of the principle demands that the trait considered is not expressed as a result of lateral information transfer or convergent evolution. It must represent vertical inheritance and originate in the organism studied (rather than in symbionts or being accumulated from food etc.). Presumably this principle can be used on the pattern of metabolites from contemporary organisms to predict the metabolic set-up of extinct and extant progenitors.
Appendix 11 Taxonomic dispersion of unique structures A vexing problem in the dispersion of specific secondary metabolites between taxa is the appearance of unique compounds in very distantly related organisms.
Scheme 15. Structure of tetrodotoxin, the cause of human poisonings from ingestion of tetrodetoxic fishes "Fugu".
An example is the alkaloid tetrodotoxin which has been detected in several species of puffer fish (family Tetraodontidae), eggs of the Califomian newt, Taricha torosa, a goby (Gobius criniger), an octopus (Hapalochlaena maculosa), a frog (Atelopus chiriquiensis), five marine gastropods, namely, an ivory shell {Babylonia japonica), a trumpet shell (Charonia sauliae), Tutufa lissostoma, Zeuxis siquizorensis, and the lined moon shell Natica lineata (202), an arrow worm (phylum Chaeto-
725
gnatha), a starfish (Astropecten polyacanthus), ribbon worms (Cephalothrix linearis, Lineus fuscoviridis), a nemertean (Tubulanus punctatus, 203) and several other systematically unrelated groups. Tetrodotoxin synthesis was recently traced to bacteria of the genera Listonella {Vibrio), Alteromonas and Shewanella (31, 204). Even if the presence of this potent toxin in all these diverse taxons is the result of bacterial symbiosis, the genes coding for the biosynthetic pathway leading to tetrodotoxin production must be quite widespread in procaryotes. Owing to the absence of biosynthetic studies, the hypothesis that tetrodotoxin diversity is a result of convergent evolution cannot be completely ruled out but the hypothesis seems far fetched. Other comparable examples are known. The main alkaloid in a Nova Scotia collection of Flustra foliacea, dihydroflustramine C, is related to pseudophrynaminol from the skin of the myobatrachid burrowing frog Pseudophryne coriacea. Both classes of alkaloids are based on the 3a-prenyl pyrrolo[2,3-fe]indole skeleton (31).
OMe
R^R
R = H , R' = OH R = H , R' = OMe R = O H . R ' = OMe R = R' = OMe
Scheme 16. Pseudophrynamines from frogs of the genus Pseudophryne.
726
The pseudophrynamines are only known from seven species of the genus Pseudophryne in the family Myobatrachidae (205). However, almost all Pseudophryne species investigated also contained the well known dendrobatid alkaloids pumilotoxins in some populations they could not be detected. Pseudophrynamines were the dominant alkaloids in two species {P. guentheri and P. occidentalis) from Western Australia, while all five eastern species contained significant amounts of pumilotoxins as well. It is interesting to note that the pseudophrynamines vary like the bryozoan alkaloids and that they were the main alkaloids only in some populations of the same species. The debromoanalog of 6-bromo-A^jj-methyl-N^j-formyltryptamine (40) has been identified from the bark of the hallucinogenic plant Virola sebifera (Myristicaceae) (206).
R= -Me R=-Et
L: R =-CH =CH2 5 : R = -COMe
R=-CHOHMe
Scheme 17. Harnian [1], 1-ethyl-P-carboline [2], 5-l-(r-hydroxyethyl)-P-carboline [3], andpavettine [4] from the bryozoan Costaticella hastata. 1-Acetyl-P-carboline [5] are known from Arenaria kansuensis.
The p-carboline alkaloids harman [1] and pavettine [4] from the bryozoan Costaticella hastata (207) are known from numerous terrestrial plants. Still another bryozoan compound (S)-l-(r-hydroxyethyl)-p-carboline [3] is new but closely related to 1-ethyl-P-carboline [2] present in the bryozoan. It is also known from the roots of Hannoa klaineana (Simaroubaceae) (208) and 1-acetyl-p-carboline [5] from Arenaria kansuensis (Caryophyllaceae) (209). Incidentally, pavettine [4] was detected in only one collection of the bryozoan. In these examples, and many others not addressed here, nothing is known about the biosynthesis of the secondary metabolites. Accordingly it can not a priori be refuted that the biosynthetic pathways may differ and hence the chemical resemblances of the structures are only coincidental. Neither can it be excluded that the compounds are microbial metabolites. However, especially in the cases of the terrestrial plants mentioned above, this explanation seems unlikely. But even if the latter possibility is taken into account we are left with the need to explain how presumably
727
distantly related microorganisms have been selected to express nearly identical compounds. If convergent evolution is disregarded the genes coding for the enzymes necessary to effect these pathways must either be very old or very efficiently dispersed by lateral information transfer. In this connection it may be mentioned that migration of genetic information from mitochondria to the nucleus in Saccharomyces cerevisiae has been found to be a high frequency event (around 2x10'^ per cell per generation), while transfer in the opposite direction is at least 100,000 times less frequent (210). In addition other processes could lead to analogous results. In the major class of red alga, Floridophyceae, more than 15% of the genera are parasitic on other red algae, often closely related. Host specificity is in some cases so narrow that the parasite discriminates between different populations of the same species. In the case of the parasite Choreocolax and the distantly related host Polysiphonia transfer of nuclei from parasite to host has been demonstrated (211). This transfer confers new morphology and histology to the host cell attesting to the activity of the parasite genetic material. There is a possibility of an additional explanation for this wide taxonomic distribution of unique structures. In the case of tetrodotoxin the taxonomic diversity is remarkable ranging from highly developed teleosts through most of the animal kingdom to procaryotic bacteria. It is possible that the capacity for tetrodotoxin production is an extraordinarily ancient trait. As all the eucaryotic organisms in which tetrodotoxin has been detected have mitochondria, it cannot be ruled out that the genetic machinery for synthesis of this molecule originates from the procaryotic organisms that entered into symbiosis with an ancient eucaryotic organism many million years ago. Thus it is conceivable that tetrodotoxin is produced by the mitochondria or that the genes involved have been transferred to the host nucleus and now function as an integral part of the host genome. Obviously some of the contemporary procaryotes still retain this ancient genetic machinery. Tetrodotoxin has been detected in a variety of marine bacteria and even in Escherichia coli (212). This explanation may also account for the other examples discussed above. To further resolve considerations of this type details about the enzymology connected with secondary metabolite synthesis are essential. 9.
ACKNOWLEDGMENTS
I am indepted to professor P.J. Scheuer for introducing me to marine natural product chemistry (213-214), Dra. R. Encamacion for introducing me to traditional medicine (112, 215), Dr. U.W. Smitt to biologically active products from higher plants (216-217), Dr. S. Wium-Andersen to freshwater plants (218), Dr. O.S. Tendal to sponge systematics (102). Professor M. G. Ettlinger has contributed significantly to the final version of this paper by way of discussions and a very careful scrutiny of the whole manuscript. However most of all I am indebted to my colleagues Drs. U. Anthoni and P. H. Nielsen for their willingness to consider and reconsider the
728
reasoning resulting from my preoccupation with subjects with seemingly no relation to organic or natural products chemistry or to chemistry at all. Last but not least I am grateful to Ms Bente Karberg for idiomatic corrections and for preparing the manuscript for printing. 10.
REFERENCES
1
U. Anthoni, P.H. Nielsen, L. Smith-Hansen, S. Wium-Andersen, C. Christophersen, J. Org. Chem., 52 (1987) 694-695. M. Gajhede, U. Anthoni, C. Christophersen, P.H. Nielsen, Acta Cryst., B45 (1989) 562-566. U. Anthoni, C. Christophersen, P.H. Nielsen, D.H. Christensen, O.F. Nielsen, M. Gajhede, Spectrochim. Acta, 45A (1989) 1157-1164. U. Anthoni, D.H. Christensen, C. Christophersen, M. Gajhede, L. Henriksen, O.F. Nielsen, P.H. Nielsen, J. Molecular Structure, 220 (1990) 43-54. K. Hagen, H.V. Volden, U. Anthoni, C. Christophersen, M. Gajhede, P.H. Nielsen, J. Phys. Chem., 95 (1991) 1597-1600. U. Anthoni, T. Borresen, C. Christophersen, L. Gram, P.H. Nielsen, Comp. Biochem. Physiol., 97B (1990) 569-571. U. Anthoni, C. Christophersen, M. Gajhede, P.H. Nielsen, Structural Chemistry, 3 (1992) 121-128. A. Guerriero, M. D'Ambrosio, P. Traldi, F. Pietra, Naturwissenschaften, 71 (1984) 425-426. M. D'Ambrosio, A. Guerriero, F. Pietra, Helv. Chim. Acta, 67 (1984) 14841493. K.-A. Karlson and B.E. Samuelson, Biochem. Biophys. Acta, 337 (1974) 204213. H.B. Mortensen and C. Christophersen, Biochem. Biophys. Acta, 707 (1982) 154-163. H.B. Mortensen and C. Christophersen, Clin. Chim. Acta, 134 (1983) 317326. H.B. Mortensen and C. Christophersen, Diabete & Metabolisme (Paris), 9 (1983) 232-234. H.B. Mortensen, Aa. Volund, C. Christophersen, Clin. Chim. Acta, 136 (1984) 75-81. C. Christophersen, H.B. Mortensen, Acta Paediatr. Scand., 73 (1984) 855-856. P. Keil, H.B. Mortensen, C. Christophersen, Acta Chem. Scand., B39 (1985) 191-193. E. Haslam, Nat. Prod. Rep., 3 (1986) 217-249. D.H. WiUiams, M.J. Stone, P.R. Hauck, S.K. Rahman, J. Nat. Prod., 52 (1989) 1189-1208. C. Mitter and J.A. Klun, J. Chem. EcoL, 13 (1987) 1823-1831.
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
729
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
J.B. Harbome, Nat. Prod. Rep., 6 (1989) 85-109. R. Nishida, T. Ohsugi, S. Kokubo, H. Fukami, Experientia, 43 (1987) 342344. C. Christophersen, Comp. Biochem. Physiol., 98B (1991) 427-432. U.W. Smitt, C. Comett, A. Andersen, S.B. Christensen, P. Avato, J. Nat. Prod., 53 (1990) 1479-1484. H. Fliick, in: T. Swain (Ed.), Chemical Plant Taxonomy, Academic Press, London, 1963, pp. 167-186. I.S. Homsey and D. Hide, Br. phycol. J., 11 (1976) 175-181. D.W. Phillips and G.H.N. Towers, J. Exp. Mar. Biol. EcoL, 58 (1982) 285293. S. Ghosal, S.K. Singh, S.G. Unnikrishnan, Phytochemistry, 29 (1989) 805-811. I.S. Homsey and D. Hide, Br. Phycol. J., 11 (1976) 63-67. V.J. Paul and W. Fenical, in: P.J. Scheuer (Ed.), Bioorganic Marine Chemistry 1, Springer-Verlag, Berlin Heidelberg, 1987, pp. 1-29. P.B. Hoist, U. Anthoni, K. Bock, C. Christophersen, P.H. Nielsen, Acta Chem. Scand., 48 (1994) 765-768. U. Anthoni, P.H. Nielsen, M. Pereira, C. Christophersen, Comp. Biochem. Physiol, 96B (1990) 431-437. C. Christophersen in: P.J. Scheuer (Ed.), Marine Natural Products. Chemical and Biological Perspectives Vol. 5. Chapter 5, Academic Press, New York, 1983, pp. 259-285. C. Christophersen, Acta Chem. Scand., B39 (1985) 517-529. C. Christophersen in: A. Brossi (Ed.), The Alkaloids. Chemistry and Physiology, Vol. 24, Chapter 2, Academic Press, New York, 1985, pp. 25-111. P. Keil, E.G. Nielsen, U. Anthoni, C. Christophersen, Acta Chem. Scand., B40 (1986) 555-559. P.B. Hoist, U. Anthoni, C. Christophersen, P.H. Nielsen, J. Nat. Prod., 57 (1994) 997-1000. J.S. Carle, C. Christophersen, J. Am. Chem. Soc, 101 (1979) 4012-4013. J.S. Carle, C. Christophersen, J. Org. Chem., 45 (1980) 1586-1589. J.S. Carle, C. Christophersen, J. Org. Chem., 46 (1981) 3440-3443. P. Wulff, J.S. Carle, C. Christophersen, Comp. Biochem. Physiol., 7IB (1982) 523-524. P. Wulff, J.S. Carle, C. Christophersen, J. Chem. Soc. Perkin Trans. I (1981) 2895-2898. P. Wulff, J.S. Carle, C. Christophersen, Comp. Biochem. Physiol, 7 IB (1982) 525-526. J.L.C. Wright, J. Nat. Prod., 47 (1984) 893-895. M.V. Laycock, L.C. Wright, J.A. Findlay, A.D. Patil, Can. J. Chem., 64 (1986) 1312-1316. P.B. Hoist, U. Anthoni, C. Christophersen, P.H. Nielsen, J. Nat. Prod., 57
730
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
(1994) 1310-1312. Q.-S. Yu, HJ.C. Yeh, A. Brossi, J. Nat. Prod., 52 (1989) 332-336. T. Sjoblom, L. Bohlin, C. Christophersen, Acta Pharm. Suec, 20 (1983) 415419. L. Andersson, G. Lidgren, L. Bohlin, L. Magni, S. Ogren, L. Afzelius, Acta Pharm. Suec, 20 (1983) 401-414. C. Christophersen, J.S. Carle, Naturwissenschaften, 65 (1978) 440. L. Chevolot, A.-M. Chevolot, M. Gajhede, C. Larsen, U. Anthoni, C. Christophersen, J. Am. Chem. Soc., 107 (1985) 4542-4543. U. Anthoni, L. Chevolot, C. Larsen, P.H. Nielsen, C. Christophersen, J. Org. Chem., 52 (1987) 4709-4712. P.H. Nielsen, U. Anthoni, C. Christophersen, Acta Chem. Scand., B42 (1988) 489-491. U. Anthoni, K. Bock, L. Chevolot, C. Larsen, P.H. Nielsen, C. Christophersen, J. Org. Chem., 52 (1987) 5638-39. G. Guella, A. Guerriero, L Mancini, F. Pietra, unpublished results mentioned in ref. 55. F. Pietra, Gazz. Chim. Ital., 115 (1985) 443-485. E. Bignetti, A. Cavaggioni, R. Tirindelli in: A. Borsellino, L. Cervetto, V. Torre (Eds.), Sensory Transduction, Plenum Press, New York, 1990, pp. 6780. M. Gajhede, U. Anthoni, C. Christophersen, P.H. Nielsen, J. Cryst. Spectrosc. Res., 20 (1990) 165-171. U. Anthoni, C. Larsen, P.H. Nielsen, C. Christophersen, Biochem. Syst. Ecol., 18 (1990) 377-379. U. Anthoni, P.H. Nielsen, C. Christophersen, Dansk Vet. Tidsskr., 73 (1990) 495-499. K.-E. Kaissling in: A. Borsellino, L. Cervetto, V. Torre (Eds.), Sensory Transduction, Plenum Press, New York, 1990, pp. 81-97. L.A. Matsuda, S.J. Lolait, M.J. Brownstein, A.C. Young, T.L Bonner, Nature, 346 (1990) 561-564. W.S. Agnew, T. Claudio, F.J. Sigworth, in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. xiii-xvii. E.M. Kosower, Biochem. Biophys. Res. Commun., I l l (1983) 1022-1026. E.M. Kosower, FEBS Letters, 155 (1983) 245-247. J.R. Smythies, G. Kemp, J. Receptor Res., 9 (1989) 199-201. J.S. Carie, C. Christophersen, J. Am. Chem. Soc., 102 (1980) 5107. J.S. Carie, C. Christophersen, Bull. Soc. Chim. Belg., 89 (1980) 1087-1091. J.S. Carie, C. Christophersen, Toxicon, 20 (1982) 307-310. J.S. Carie, H. Thybo, C. Christophersen, Contact Dermatiris, 8 (1982) 43-47. T. Rasmussen, C. Christophersen, P.H. Nielsen, R. Rajagopal, J. Mar.
731
71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
87 88 89 90
91
Biotech., (in press). T. Rasmussen, J. Jensen, U. Anthoni, C. Christophersen, P.H. Nielsen, J. Nat. Prod., 56 (1993). 1553-1558. M.H.G. Munro, R.T. Luibrand, J.W. Blunt, in: PJ. Scheuer (Ed.), Bioorganic Marine Chemistry 1, Springer Verlag, Berlin Heidelberg, 1987, pp. 93-176. A. Gescher and I.L. Dale, Anti-Cancer Drug Design, 4 (1989) 93-105. R.L. Huganir, in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 147-164. G.R. Pettit, D.L. Herald, F. Gao, D. Sengupta, C.L. Herald, J. Org. Chem., 56 (1991) 1337-1340. J.S. Ramsdell, G.R. Pettit, A.H. Tashjian, Jr., J. Biol. Chem., 261 (1986) 17073-17080. B.S. Warren, Y. Kamano, G.R. Pettit, P.M. Blumberg, Cancer Research, 48 (1988) 5984-5988. W.S. May, S.J. Sharkis, A.H. Esa, V. Gebbia, A.S. Kraft, G.R. Pettit, L.L. Sensenbrenner, Proc. Natl. Acad. Sci. USA, 84 (1987) 8483-8487. H.G. Drexler, S.M. Gignac, R.A. Jones, C.S. Scott, G.R. Pettit, A.V. Hoffbrand, Blood, 74 (1989) 1747-1757. H.G. Drexler, S.M. Gignac, G.R. Pettit, A.V. Hoffbrand, Eur. J. Immunol., 20 (1990) 119-127. H. Hennings, P.M. Blumberg, G.R. Pettit, C.L. Herald, R. Shores, S.H. Yuspa, Carcinogenesis, 8 (1987) 1343-1346. A.D. Hess, M.K. Silanskis, A.H. Esa, G.R. Pettit, W.Stratfor May, J. Immunology, 141 (1988) 3263-3269. A.S. Kraft, F. William, G.R. Pettit, M.B. Lilly, Cancer Research, 49 (1989) 1287-1293. H. Mohr, G.R. Pettit, A. Plessing-Menze, Immunobiol., 175 (1987) 420-430. G. Trenn, G.R. Pettit, H. Takayama, J. Hu-Li, M.V. Sitkovsky, J. Immunology, 140 (1988) 433-439. P. Taylor, P. Culver, S. Abramson, L. Wasserman, T. Kline, W. Fenical, Biomedical Importance of Marine Organisms, Fautin, D. G., Califomia Academy of Sciences, 1988, pp. 109-114. U. Anthoni, L. Bohlin, C. Larsen, P. Nielsen, N.H. Nielsen, C. Christophersen, Toxicon, 27 (1989) 717-723. U. Anthoni, L. Bohlin, C. Larsen, P. Nielsen, N.H. Nielsen, C. Christophersen, Toxicon, 27 (1989) 707-716. U. Anthoni, C. Christophersen, P.H. Nielsen, J. Agric. Food Chem., 37 (1989) 705-707. B. Witkop and A. Brossi in: P. Krogsgaard-Larsen, S.B. Christensen, H. Kofod, (Eds.), Natural Products and Drug Development, Munksgaard, Copenhagen, 1984, pp. 283-300. U. Anthoni, C. Christophersen, J. 0gard Madsen, S. Wium-Andersen, N,
732
92 93
94 95
96 97 98 99
100
101 102 103 104 105
106 107 108 109 110 111 112 113 114
Jacobsen, Phytochemistry, 19 (1980) 1228-1229. C. Christophersen, Arch. Pharm. Chem., 94 (1987) 383-395. C. Christophersen, C.T. Pedersen, J. Becher (Eds.), Developments in the Organic Chemistry of Sulfur, Gordon and Breach, Science Publishers, New York, 1989, pp. 155-163. C. Christophersen, Phosphorus, Sulfur and Silica, 43 (1989) 155-163. M.E. Eldefrawi, I.M. Abalis, S.M. Sherby, A.T. Eldefrawi in: M.G. Ford, G.G. Lunt, R.C. Reay, P.N.R. Usherwood (Eds.), Neuropharmacology and Pesticide Action, Verlag Chemie and EUis Horwood, 1986, pp. 154-173. N. Jacobsen and L.-E.K. Pedersen, Pestic. Sci., 14 (1983) 90-97. S. Wium-Andersen, U. Anthoni, C. Christophersen, G. Houen, Oikos, 39 (1982) 187-190. J.F. Martin and A. Demain, Microbiol. Rew., 44 (1980) 230-251. A.S. Khokhlov, I.T. Tovarova, L.N. Borisova, S.A. Pliner, L.A. Shevchenko, E.Y. Komitskaya, N.S. Ivkina, LA. Rapoport, Dokl. Acad. Nauk SSSR, 177 (1967) 232-235. I.S. Hunter and S. Baumberg, in: S. Baumberg, I. Hunter, M. Rhodes (Eds.), Microbial Products: New Approaches, Cambridge University Press, Cambridge, 1989, pp. 121-162. P.R. Bergquist, W. Hofheinz, G. Oesterhelt, Biochem. Syst. Ecol., 8 (1980) 423-435. C. Christophersen, P.G. Nielsen, P.H. Nielsen, O.S. Tendal, Comp. Biochem. Physiol., to be pubHshed. S. Berking, Roux's Arch. Dev. Biol., 195 (1986) 33-38. S. Berking, Development, 99 (1987) 211-220. P.R. Bergquist, R.J. Wells in: P.J. Scheuer (Ed), Marine Natural Products, Chemical and Biological Perspectives., Academic Press, New York, 1983, vol. V, pp. 1-50. D.M. Anderson, D.M. Kulis, J.J. Sullivan, S. Hall, Toxicon, 28 (1990) 885893. H. Ravn, C.U. Schmidt, H. Sten, U. Anthoni, C. Christophersen, P.H. Nielsen, Comp. Biochem. Physiol., (1995), in press. M. Kodama, T. Ogata, S. Sakamoto, S. Sato, T. Honda, T. Miwatani, Toxicon, 28 (1990) 707-714. U. Anthoni, C. Christophersen, N. Jacobsen, A. Svendsen, Tetrahedron, 38 (1982) 2425-2427. L. Teuber and C, Christophersen, Acta Chem. Scand., B42 (1988) 629-634. L. Teuber. and C. Christophersen, Acta Chem. Scand., B42 (1988) 620-622. C. Christophersen, C, Larsen, R.D. Encamacion, The H. C. 0rsted Institute, Copenhagen, 1991, 1-41. C. Christophersen, C. Larsen, R.D. Encamacion, Danida Report (1989) 1-43. M. Gajhede, R.D. Encamacion, G.C. Leal, J.C. Patino, C. Christophersen, P.H.
733
115 116 117 118 119 120 121 122 123
124 125 126 127 128 129 130 131 132
133 134 135 136 137 138 139 140
Nielsen, Acta Cryst., C45 (1989) 2012-2014. R.D. Encamacion, S.G. Keer, P.H. Nielsen, C. Christophersen, J. Ethnopharmacology, 31 (1991) 43-48. R.D. Encamacion, N.A. Ochoa, U. Anthoni, C. Christophersen, P.H. Nielsen, J. Nat. Prod., 57 (1994) 1307-1309. S.E. Nielsen, U. Anthoni, C. Christophersen, C. Comett, Phytochemistry (1995), in press. G. Green, Marine Biology, 40 (1977) 207-215. P.R. Bergquist, J.J. Bedford, Marine Biology, 46 (1978) 215-221. S.J. Giovannoni, T.B. Britschgi, C.L. Moyer, K.G. Field, Nature, 345 (1990) 60-65. D.M. Ward, R. Weller, M.M. Bateson, Nature, 345 (1990) 63-65. C.A. Suttle, A.M. Chan, M.T. Cottrell, Nature, 347 (1990) 467-469. C.A. West, in: D.T. Dennis and D.H. Turpin (Eds.), Plant Physiology, Biochemistry and Molecular Biology. Longman Scientific and Technical, England, 1990, p. 353-369. M. Wink, J. Exp. Botany, 44, Supplement, (1993) 231-246. K.C. Engvild, Physiologia Plantarum, 77 (1989) 282-285. J. Theodor, Nature, 227 (1970) 690-692. O. Von Fiirth, Vergleichende chemische Physiologic der niederen Tiere Verlag von Gustav Fischer, Jena, 1903, pp. 453-457. J.B. Harbome, Introduction to Ecological Biochemistry, Third Edition, Academic Press, New York, 1988. M. Aubert, Mar. Poll. Bull, 21 (1990) 24-29. F.E. Round, The Ecology of Algae, Cambridge University Press, Cambridge, (1981). D.J. Futuyma, Evolutionary Biology, Second Edition, Sinauer Associates, Inc., Sunderland, Massachussetts, 1986. H. VanEtten, D. Matthews, P. Matthews, V. Miao, A. Maloney, D. Straney in: B.J.J. Lugtenberg (Ed.), Signal Molecules in Plants and Plant-Microbe Interactions, NATO ASI Series H: Cell Biology Vol. 36. Springer-Veriag, Beriin, 1989. S. Adhya and S. Garges, J. Biol. Chem., 265 (1990) 10797-10800. H.J. Zeringue Jr. and S.P. McCormick, Toxicon, 28 (1990) 445-448. H.J. Zeringue Jr., Phytochemistry, 29 (1990) 1789-1791. R. Baute, M.-A. Baute, G. Deffieux, M.-J. Filleau, Phytochemistry, 15 (1976) 1753-1755. M. Piattelli, New J. Chem., 14 (1990) 777-782. G. Combaut, L. Codomier, J. Teste, Phytochemistry, 20 (1981) 2036-2037. J.F. Biard, J.F., Verbiest, Y. Letoumeux, R. Floch, Planta Med., 40 (1980) 288-294. J.F. Biard, J.F. Verbiest, R. Floch, Y. Letoumeux, Tetrahedron Letters, 21
734
141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167
(1980) 1849-1852. G. Combaut, L. Piovetti, Phytochemistry, 22 (1983) 1787-1789. L. Hougaard, U. Anthoni, C. Christophersen, P.H. Nielsen, Phytochemistry, 30 (1991) 3049-3051. L. Hougaard, U. Anthoni, C. Christophersen, P.H. Nielsen, Tetrahedron Letters, 32 (1991) 3577-3578. R. Vails, B. Banaigs, C. Francisco, L. Codomier, A. Cave, Phytochemistry, 25 (1986) 751-752. L. Semmak, A. Zerzouf, R. Vails, B. Banaigs, G. Jeanty, C. Francisco, Phytochemistry, 27 (1988) 2347-2349. A. San-Martin and J. Rovirosa, Biochem. Syst. EcoL, 14 (1986) 459-461. M.S. Doty, G. Aguilar-Santos, Nature, 211 (1966) 990. M.S. Doty, G. Aguilar-Santos, Pac. Sci., 24 (1970) 351-355. P.G. Nielsen, J.S. Carle, C. Christophersen, Phytochemistry, 21 (1982) 16431645. M. Mahendran, S. Somasundaram, R.H. Thomson, Phytochemistry, 18 (1979) 1885-1886. O.R. Gottlieb, Micromolecular Evolution, Systematic and Ecology, An Essay into a Novel Botanical Discipline, Springer-Verlag, Berlin, 1982. A.J. Blackman, R.D. Green, Aust. J. Chem., 40 (1987) 1655-1662. R. Clauser, A. De Guerriero, F. Pietra, impublished results mentioned in ref. 55. M. Suffness, D.N. Newman, K. Snader, in: P.J. Scheuer (Ed.), Bioorganic Marine Chemistry 3, Springer-Verlag, Berlin Heidelberg, 1989, pp. 131-168. L. Gunthorpe and A.M. Cameron, Toxicon, 28 (1990) 1199-1219. L. Gunthorpe and A.M. Cameron, Toxicon, 28 (1990) 1221-1227. J. W. Blunt, M.H.G. Munroe, C.N. Battershill, B.R. Coop, J.D. Combs, N.B. Perry, M. Prinsep, A.M. Thompson, New J. Chem., 14 (1990) 761-775. M.R. Keman, D.J. Faulkner, R.S. Jacobs, J. Org. Chem., 52 (1987) 3081-3083. E.D. de Silva, P.J. Scheuer, Tetrahedron Letters, 21 (1980) 1611-1614. E.D. de Silva, P.J. Scheuer, Tetrahedron Letters, 22 (1981) 3147-3150. R.C Cambie, P.A. Craw, P.R. Bergquist, P. Karuso, J. Nat. Prod., 51 (1988) 331-334. M.R. Keman, R.C. Cambie, P.R. Bergquist, J. Nat. Prod., 53 (1990) 724-727. M.R. Keman, R.C. Cambie, P.R. Bergquist, J. Nat. Prod., 53 (1990) 13531356. R.J. Wells, Tetrahedron Letters, (1976) 2637-2638. T. Murayama, Y. Ohizumi, H. Nakamura, T. Sasaki, J. Kobayashi, Experientia, 45 (1989) 898-899. S. Sakemi, T. Higa, U. Anthoni, C. Christophersen, Tetrahedron, 43 (1987) 263-268. T. Higa, C. Christophersen, S. Sakemi, U. S. Patent 7, 731, (1988) 377, Mar.
735
168 169 170 171 172 173 174 175 176 177 178 179 180 181 182
183
184 185 186
187
188
15, 1988. F.S. De Guzman, F.J. Schmitz, J. Nat. Prod., 53 (1990) 926-931. E. Quinoa, E. Kho, L.V. Manes, P. Crews, G.J. Bakus, J. Org. Chem., 51 (1986) 4260-4264. S.P. Gunasekera, M. Gunasekera, G.P. Gunawardana, P. McCarthy, N. Burres, J. Nat. Prod., 53 (1990) 669-674. C. Christophersen, U. Anthoni, P.H. Nielsen, N. Jacobsen, O. Tendal, Biochem. Syst. EcoL, 17 (1989) 459-461. C. Christophersen, N. Jacobsen, R. S. C. Annual Reports, B (1979) 433-447. D.S. Latchman, Biochem. J., 270 (1990) 281-289. J. Darnell, H. Lodish, D. Baltimore, Molecular Cell Biology. Second Edition, Scientific American Books, W. H. Freeman and Company, New York, 1990. C.C. Kao, P.M. Lieberman, M.C. Schmidt, Q. Zhou, R. Pei, A.J. Berk, Science, 248 (1990) 1646-1650. M.G. Peterson, N. Tanese, B.F. Pugh, R. Tijan, Science, 248 (1990) 16251630. A. Hoffmann, E. Sinn, T. Yamamoto, J. Wang, A. Roy, M. Horikoshi, R.G. Roeder, Nature, 346 (1990) 387-390. B. Alberts, D. Bray, J. Lewis, M. Raff, K. Roberts, J.D. Watson, Molecular Biology of the Cell. Second Edition, Garland Publ., Inc., New York, 1990. J.D. Fikes, D.M. Becker, F. Winston, L. Guarente, Nature, 346 (1990) 291294. A. Gasch, A. Hoffmann, M. Horikoshi, R.G. Roeder, N.-H. Chua, Nature, 346 (1990) 390-394. M. Ptashne and A.A.F. Gann, Nature, 346 (1990) 329-331. E. Hawrot, K.L. Colson, T.L. Lenz, P.T. Wilson in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.) Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 165-196. S.M. Sine and J.H. Steinbach, in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 134-146. T.L. Lentz and P.T. Wilson, Intemat. Rev. Neurobiol., 29 (1988) 117-160. H. Breer, D.B. Sattelle, J. Insect Physiol., 33 (1987) 771-790. Y. Saimi and C. Kung, in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 1-14. W.S. Agnew, E.C. Cooper, W.M. James, S.A. Tomiko, R.L. Rosenberg, M.C. Emerick, A.M. Correa, J.Y. Zhou, in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 329-369. E. Hartmann and R. Gupta in: W.F. Boss, D.J. Morre (Eds.), Plant Biology 6, Second Messengers in Plant Growth and Development, Alan R. Liss, Inc.,
736
189
190 191 192 193 194
195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212
New York, 1989, pp. 257-287. A. Tretyn, M.E. Bossen, R.E. Kendrick, in: CM. Kassen, L.C. Loon, D. van Vreugdenhil (Eds.), Progress in Plant Growth Regulation 306, Kluwer Academic Publishers. The Netherlands, (1992). G.A. Miura and T.M. Shih, Physiol. Plant., 61 (1984) 417-421. V.V. Roshchina, Photosynthetica, 23 (1989) 197-206. A. Tretyn, J. Kopcewicz, E. Slesak, Biol. Plant., 30 (1988) 338-342. H. Betz, Biochemistry, 29 (1990) 3591-3599. L.J. Cruz, D.S. Johnson, J.S. Imperial, D. Griffin, G.W. LeCheminant, G.P. Miljanich, B.M. Olivera in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 417-430. S.A. Benner, A.D. Ellington in: H. Dugas (Ed.), Bioorganic Chemistry Frontiers, 1, Springer-Verlag, Berlin, 1990, pp. 3-70. S.A. Benner, A. Glasfeld, J.A. Piccirilli, Topics in Stereochemistry, 19 (1989) 127-207. B. Nagy, Naturwissenschaften, 69 (1982) 301-310. A.S. Mackenzie, S.C. Brassell, G. Eglinton, J.R. Maxwell, Science, 217 (1982) 491-504. J.M. Hunt, Petroleum Geochemistry and Geology. W. H. Freeman and Company, San Francisco, 1979. B.P. Tissot and D.H. Welte, Petroleum Formation and Occurrence, A New Approach to Oil and Gas Exploration. Springer-Verlag, Berlin, 1978. C. Christophersen, U. Anthoni, Sulfur Reports, 4 (1986) 365-456. D.F. Hwang, C.H. Chueh, S.S. Jeng, Toxicon, 28 (1990) 1133-1136. A.E. Ali, O. Arakawa, T, Noguchi, K. Miyazawa, Y. Shida, K. Hashimoto, Toxicon, 28 (1990) 1083-1093. U. Simidu, K. Kita-Tsukamoto, T. Yasumoto, M. Yotsu, Intemat. J. Syst. Bacteriol., 40 (1990) 331-336. J. W. Daly, H.M. Garraffo, L.K. Pannell, T.F. Spande, C. Severini, V. Erspamer, J. Nat. Prod., 53 (1990) 407-421. K. Kawanishi, Y. Uhara, Y. Hashimoto, Phytochemistry, 24 (1985) 1373-1375. A.J. Blackman, D. Matthews, C.K. Narkowicz, J. Nat. Prod., 50 (1987) 494496. L. Lumonadio and M. Vanhaelen, Phytochemistry, 23 (1984) 453-455. F.-E. Wu, K. Koike, T. Nikaido, Y. Sakamoto, T. Ohmoto, K. Ikeda, Chem. Pharm. Bull., 37 (1989) 1808-1809. P.E. Thorsness and T.D. Fox, Nature, 346 (1990) 376-379. L.J. Goff and A.W. Coleman, Proc. Natl. Acad. Sci. U. S. A., 81 (1984) 54205424. M.L. Tamplin, in: S. Hall and G. Strichartz (Eds.), Marine Toxins, Origin, Structure, and Molecular Pharmacology, American Chemical Society,
737
213 214 215 216 217 218
Washington, D. C, 1990, pp. 78-86. B.J. Burreson, C. Christophersen, P.J. Scheuer, J. Am. Chem. Soc., 97 (1975) 201-202. B.J. Burreson, C. Christophersen, P.J. Scheuer, Tetrahedron, 31 (1975) 20152018. R.D. Encamacion, G. Carrasco, M. Espinoza, U. Anthoni, P.H. Nielsen, C. Christophersen, J. Nat. Prod., 52 (1989) 248-251. S.B. Christensen, U. Rasmussen, C. Christophersen, Tetrahedron Letters, 21 (1980) 3829-3830. S.B. Christensen, I.K. Larsen, U. Rasmussen, C. Christophersen, J. Org. Chem., 47 (1982) 649-652. S. Wium-Andersen, K.H. Jorgensen, C. Christophersen, U. Anthoni, Arch. Hydrobiol., I l l (1987) 317-320.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
739
The Celastraceae from Latin America Chemistry and Biological Activity
O. Muhoz, A. Penaloza, A.G. Gonzalez, A.G. Ravelo, LL. Bazzocchi and N X . Alvarenga 'Universidad de Chile, Facultad de Ciencias, Casilla 653-Santiago, Chile. C.P.N.O. Antonio Gonzalez, Universidad de La Laguna, Carretera La Esperanza 2, La LagunaTenerife. Espafia.
1. INTRODUCTION The CELASTRACEAE family was last reviewed in 1978 (Brtining & Wagner) [1] and since then a great deal of new chemical and phannacological information has accumulated. Celastraceae species have a long tradition of use in medicine and folk agriculture, especially in Asia and Latin America but also in other continents and undoubtedly took on a new lease of life in the seventies when the MAYTANSINOIDS, compounds with exceptional antitumoral properties, were discovered [I]. Nonetheless, the maytansinoids have not been made into a useful drug form as they cause serious gastro-intestinal damage when applied to rats [2]. For some time now several research laboratories have been intensively researching this fiamily, inspired by its broad and varied botanical distribution, the interesting chemical nature of its secondary metabolites, the complexity of the biogenetic processes which produce them, and most of all by the different types of pharmacological action displayed by preparations of its constituents. In Latin America (Mexico, Central America, the Caribbean and South America) this study is particularly momentous due in tfie main to a socio-economic and cultural climate which has not in the past lent itself either to sound development or the rational exploitation of the resources of the various countries involved. In the course of these research programmes, cytotoxic quinones, polyester sesquiterj)enes and pyridine-sesquiterpene alkaloids with antifeedant and/or insecticidal properties have been isolated from Latin American species, m particular those of the Maytenus genus which is extensively used by rural communities and tribes in the Andes and the Amazon basin. Recently some sesquiterpene alkaloids with immunosuppressive activity and sesquiterpenes with antitumoral activity have also been described [3].
740 As a general rule, the biosynthesis of skeletons belonging to the Celastraceae family is extremely specific, the triterpene-quinones and P-dihydro-agaroftiran type skeleton sesquiterpenes from these species having a notably high degree of oxidation. The presence of triterpene-quinones indicates the biosynthetic specificity' of the Celastraceae family since these compounds are synthesized in the roots and are virtually exclusive to the family. The next few pages are an update on the state of the Celastraceae family in Latin America, detailing the different chemical structures found and the results of the studies of biological activity carried out since the last publication on the subject [1].
2. THE SYSTEMATICS OF THE CELASTRACEAE FAMILY AND THE LATIN AMERICAN GENERA
THE CELASTRACEAE FAMILY
The Celastraceae family consists of about 55 genera and 850 species. According to Takhtajan [4], Hippocrateaceae would be subordinate to the Celastraceae. Systematically the Celastraceae family is arranged hierarchically as follows: Division: Spermatophyta; Subdivision: Magnoliophyta (Angiosperms); Class: Magnolitae (Dicotyledons); Order: Celastrales; Family: Celastraceae The Celastraceae family is probably related phylogenetically with the Aquifoliaceae; the presence of glandular discs around the ovary and the bright coloured aril in the Celastraceae are the principal differences between the two families. The Celastraceae family is pantropically distributed with radiation towards temperate or temperate-cold climates. In other words, the Celastraceae are principally concentrated in the tropical and subtropical regions and to a lesser extent in the temperate zones of the world (Figure la). The family is better represented in Central America and the West Indies than m South America except for the Maytenus genus [5] (15 species in Peru and 15 in Venezuela). They are found growing as upright trees, bushes and lianas and almost invariably have resin ducts or cells in the bast of the stems and leaves. The leaves are simple, usually alternate and opposite; stipules small and deciduous or missing; the flowers are small, fasciculate, actinomorphic, forming cymes, very occasionally racemes or inflorescence. The flowers are in general bisexual and rarely polygamodioecious plants. The petals are freestanding or coneshaped. The fruits are berries, capsules, drupes or samaras. The seeds generally have a coloured aril and contain embryos with large cotyledons and a relatively oleaginous endosperm. The chromosomic numbers described are X = 8,12,14.
741 The Celastraceae genera are very diversified; some of die taxa widi most species are: Maytenus (225; tropical), Salacia (200; tropical), Euonomynus (176; Himalayas, China and Japan), Hippocratea (120; tropical. South America, Mexico and the South of the USA), Cassine (40; South Africa, Madagascar, tropical Asia and the Pacific), Celastrus (30; principally Asia, some in Australia and in tropical and temperate zones of America), Elaodendrum (16; tropical and subtropical), Pachystima (5; North America) and Gyminda (3; Central America, Mexico and Florida). Gentry [5] compared the woody taxa found in Africa and America and commented that tiae genus Maytenus was in taxonomic terms closer to other genera of the same family found on the same continent than to other species of Maytenus from the other side of Ae Atlantic; this resemblance pattern is observed also at the level of secondary metabolites. Although there is no clear explanation for the foregoing, paleobotanic studies have shown that the paleoflora of South America and Africa were very similar in the late Paleozoic era (340-240 My) [6]. Bearing in mind that the early stages of Continental Drift between Africa and South America occurred about 135 My ago it can be argued diat many populations widely distributed in Gondwana were separated as the resuh of the Continental Drift (Figure lb) wliich meant that the split populations would develop independently Aereafter. Although few examples of Ae Celastraceae family have been found and only recently, fossilized remains of Celastrus show that during the Tertiary Age this genus was unrestrictedly distributed throughout America [7] and Europe [8]. It has even been suggested that the present distribution pattern of some species of American Celastrus corresponds better to a dispersion centre in Asia than to (Mie in Central America [7]. The phylogenetic history of the Celastraceae family, - and of other widely distributed families,appears to run parallel with that of the Continental Drift. Modem chemomolecular studies may help to determine the phylogenetic relationships of its different members, as well as with other families of the Celastrales order. If the theory of the Continental Drift is true, it is not surprising that the phylogenetic relationships of the Celastraceae with odier plant families should date back a long time. New chemotaxonomical studies are constantly appearing which relate the Celastraceae with tfie Lamiaceae, for instance. The presence in both femilies of sesquiterpenes, diterpcnes and triterpenes of high and specificfrmctionalityconfirms such a relationship [9].
742
Figure la.- Present-day distribution of the species of the Celastraceae family. As can be seen, this family has developed for preference in die tropics in the New and Old World, and is extending towards temperate or temperate/cold climates in the southern hemisphere.
Figure lb.- In overall terms, the Continental Drift is theorized as shown in this scheme. Some 200 My ago, all Ae continents formed a single mass, known as Pangea (a), which later split into two, Laurasia (NorA America, Europe and Asia) to Ae north and Gondwana (the Antarctic, South America, India, Australia and New Zealand) to the south (b). Still later Gondwana disintegrated, with India breaking off first and fmally colliding with Asia, then South America litting off from Africa, New Zealand from Antarctica and last of all, Australia from Antarctica (c &d).
743 3. SESQUITERPENES The salient feature of the family has been its wealth of sesquiterpenes with abnost a hundred of these compounds being isolated and characterized chemically. This group of metabolites very common in Latin American species have the eudesmane basic skeleton, and originate the wellknown p-dihydro-agarofuran skeleton polyester sesquiterpenes which are esterified by a series of common organic acids - acetic, benzoic, 3-furoic, trans-cinnamic acid, etc. (Table I).
A. POLYESTER SESQUITERPENES Sesquiterpene esters based on the dihydro-agarofuran moiety occur mainly within the Celastraceae family.
The basic polyhydroxy skeleton vary according to the position, number and
configuration of the esterresiduesin the dihydro-p-agarofuran sesquiterpene. The interest generated by polyester sesquiterpenes from the Celastraceae has increased in line with the complexity of the substances isolated and the possibility of their being applied to combat insect plagues instead of synthetic insecticides. The complexity and increasing numbers of these sesquiterpenes makes it difficult to arrange them systematically. They can however, be treated as derivatives of a basic polyhydroxy skeleton and thereafter organized in sunpler series. Accordingly, 37 series of P-dihydro-agarofuran type sesquiterpenes have been proposed ranging from a skeleton with two hydroxyls (boariol) to one with nine (euonyminol and isoueonyminol series) (Figure 2).
Sesquiterpenes with two hydroxy groups
K
HO
Bovbt
Sesquiterpenes with three hydroxy groups OH
OH
OH
OH
OH
OH
HOi
^ - < Isocekxbicol
4{J-Hyd'«y-6-deo)y-oekxt3icol
744 Sesquiterpenes with four hydroxy groups CH2OH OH
Maikangunbi
OH
OH
OH 4 (J-Hyd-ocy-cebrbicol
OH
2p.4p-Dihydroy^<Jecjy<3eloft)iool
1 S-Hydrwy-celorbk^l
Sesquiterpenes with four hydroxy groups and one ketone group
2.3.13.1 STetra-deoxy-eMC3nlnol
Sesquiterpenes with five hydroxy groups CH2OH OH
Pentahydroxy-agarofu ran
OH 15-Hydnay-celapanol
OH
HO OH 4p- Hydros oeiapand
745 CH2OH
CHzOH
HO
OH
OH
2.3-DideGD^maytoi
3.4-Oidecsy-mayto(
OH
2.3.13.1 S-TetradeGKyisoeuoniminoi
OH
U-< OH 4-DeGD^magelland
Sesquiterpenes with five hydroxy and one ketone groups CH2OH
OH 3.4.13-TrideoDy«vonnd
Sesquiterpenes with six hydroxy groups CH2OH
HOSs^
746 CHaOH
OH
CH2OH
-
OH
ep-Hydroxy-pentahydroxy-agarofurano
2.3.13-Trideo)^oeuoniminol CH2OH
HO
OH
HO
2a .4p-DtTy(Jroxy.a^-celapanol
OH
2p .4p .Dihydroxy-S-epi-celapanol
Sesquiterpene with six hydroxy and one ketone groups CH2OH
r OH 3.13-Didecs^^-ey^3ninoi
747 Sesquiterpenes with seven hydroxy groups CH2OH
HO
CH2OH
^4-^
HO
OH
0.
OH 4^-Hydroxy-aiatol
May4oi
Sesquiterpene with eight hydroxy and one ketone groups CH2OH
CH2OH
Sesquiterpenes with nine hydroxy groups CH2OH
H2OH
isoeuonimnol Figure 2.- Polyhydroxy P-dihydro-agarofiiran skeleton sesquiterpenes
The structural elucidation of these sesquiterpenes, given the difficulty in determining the linking sites of the respective ester groups when more than three kinds of acid are involved as esters in the molecule, necessitates the use of nmr experiments (selective INEPT [10], Coloc [11] or HMBC 2D [12]), selective hydrolysis processes or X-ray crystallographic methods [13]. From comparison of the chemical shifts of similar compounds, it is not easy to establish the exact position of each ester.
748
Structures were deduced from spectroscopic studies, basically heteronuclear 'H-'^C correlations (HETCOR). The regiosubstitution characteristics were generally ascertained by using long-range correlation spectra with inverse detection, with a 2-D heteronuclear multiple bond connectivity techniques (HMBC) which located the substituents (benzoates, hydroxyls, acetates etc); ROESY and nOe experiments were used to complement information about conformational interaction and long-range relationships. The absolute configurations were determined in almost all cases by applying the exciton-chirality method to dibenzoyl derivatives. Total and/or partial hydrolysis of the esters and derivative preparations (acetates, benzoates, methoxyderivatives, etc.) provided further information. The 'H nmr spectra of these sesquiterp)enes show three well-defined absorption regions which are the main sources of analytical information: the high field region where signals for the methyl groups at C-12 and C-13 and, in the highest part of the spectrum, the characteristic methyl group at C-14 as doublet are all to be found. This latter is a useful diagnostic aid because this position tends to have a hydroxy substituent, and when this occurs, the doublet methyl signal disappears; secondly, the usual shift of the acetoxy groups is between 5 1.5 and 2.2 ppm; if the acetoxy shift is at 5 1.5, the relative position of the benzoyloxy group can to be determined as C-1 or C-9 on the decaline ring system since this group anisotropically shields the acyl groups in peri position to the ring. The H-7 equatorial proton appears as a doublet (J=3.0 Hz) between 6 2.5 and 3.0 depending on the relative position of the C-8 and C-6 protons with which it is normally coupled. The midspectrum between 5 3.0 and 6.5 is usually where the characteristic signals for the hydroxy acetate and/or benzoate geminal protons are to be seen together with those of the C-15 geminal protons which vary according to whether they are esterified or not. The shift and multiplicity of the signals usually provide information about Ae relative stereochemistry and regiochemistry of the substituents. The benzoate protons resonate in the low-field region between 7.0 and 8.0 ppm, making make it easy to distinguish them from the other ester groups of the p-dihydro-agarofuran system. Each of the signals exhibits the downfield chemical shift to be e3q>ected of the corresponding esterified OH group derivatives. Decoupling experiments and/or COSY, ROESY, etc. complete the information available. In general, the sesquiterpenes described to date have a stable configuration and conformation free from interconversions and normally determined by the configuration of the chiral centres C1, C-4, C-5, C-7 and C-10. This structural rigidity is also apparent in the stereochemistry and functionality of some stereocentres:
749 (a) C-1 and C-9 are usually esterified with C-1 regiosubstitution being a; (b) When there is substitution at C-6, it is always p; (c) If there are hydroxy group on C-4, it assumes a p equatorial position; (d) Stereochemistry and regiosubstimtion at C-2, C-3, C-8 and C-9 vary; (e) The oxo groups are usually at C-8 and less commonly at C-6. The following tables summarize the above data on the same lines as laid down by Wagner & BrUning[l].
TABLE 1. SESQUITERPENES ISOLATED FROM LATIN AMERICAN SPECIES 15
10
t L4^" V 13
1 Form
C-1
C-2
C-3
C-4
L 1
aOAc
aOAc
2H
H
pOAc
2
aOAc
aOAc
2H
H
3
aOAc
aOAc
2H
H
4
aOAc
aOAc
2H
5
aOAc
aOAc
2H
6
aOAc
aOAc
C.6
c-8
C-9
C-IS
Ref
2H
pOBz
OAc
14
pOAc
=o
pOBz
OAc
15
POAc
aOAc
POBz
OAc
15 1
H
POAc
aOAc
pOBz
OH
15
H
POAc
=o
pOBz
OH
15
2H
H
POAc
2H
pOBz
OH
14
7
aOAc
aOAc
2H
H
POAc
aOH
pOBz
OAc
15
1 8
aOAc
aOAc
2H
H
pOAc
aOAc
aOBz
OH
14
9
aOAc
aOAc
2H
H
POAc
aOH
pOBz
OH
15
10
aOAc
aOAc
2H
H
pOAc
aOH
aOBz
OH
14
11
aOAc
2H
2H
H
POAc
aOAc
pOBz
OAc
16
12
aOAc
2H
2H
H
POAc
2H
pOBz
OH
16
13
aOAc
2H
2H
H
pOAc
aOH
pOBz
OAc
16
14
aOBz
2H
2H
POH
POAc
pOAc
POBz
H
16
15
aOBz
POAc
POH
POH
pOAc
H
pOBz
H
17
16
aOBz
POAC
pOAc
pOH
pOAc
H
pOBz
H
17
17
aOBz
POH
POAc
pOH
POAc
H
pOBz
H
17
H
17
18
aOBz
pOAc
POH
pOH
H
H
pOBz
19
aOBz
pOAc
pOAc
pOH
H
H
pOBz
H
17
20
aOBz
pOAc
pOH
H
pOAc
H
pOBz
H
18,19
21
aOBz
POAc
POcA
pOH
pOBz
H
pOBz
H
17
22
aOBz
POAc
pOAc
pOH
H
pOAc
POBz
H
19
1
750
23
aOBz
pOAc
pOH
pOH
H
POAc
POBz
H
19
24
aOBz
pOAc
POAc
POH !
H
pOBz i
POBz
H
20
1 25
aOBz
aOAc
H
pOH
POAc
POAc
pOBz
H
20
26
aOBz
pOAc
H
pOH
pOBz
pOAc
pOBz
H
21
27
aOBz
POAc
H
pOH
POAc
pOAc
POBz
H
21 1
pOBz
H
20
28
aOBz
POAc
H
pOH
H
POBz
29
aOBz
pOAc
H
POH
H
H
POBz
H
20 1
30
aOBz
H
H
pOH
H
H
pOBz
H
19
: 31
aOBz
H
H
pOH
H
POBz
H
19
L32
aOBz
POAc
POH
POH
POH
H
POBz
H
33
aOBz
^ A c L POH
POH
POAc
H
POCinn
H
21 1 22 1
34
H
H
aOH
aOH
H
H
H
18,23 I
35
aOBz
2H
2H
pOH
POAc
2H
pOAc 1
H
24 1
36
aOBz
2H
2H
pOH
POAc
2H
POAc
OAc
24
1 37
aOBz
38
aOCinn
pOAc
H
2H
pOH
pOAc
pOH
aOAc
OAc
24
pOAc
2H
POH
POAc
2H
pOAc
H
24
2H '
39
aOBz
2H
2H
POH
POAc
pOAc
aOAc
H
24
40
aOBz
2H
2H
POH
pOH
H
pOAc
H
24
41
aCinn
POH
H
POH
pOH
H
pOAc
H
25
42
oEp-
pOAc
2H
pOH
pOH
2H
pOAc
H
25 1
2H
2H
pOH
pOH
2H
pOAc
H
25 1
2H
2H
pOH
pOAc
2H
pOAc
H
25 1
Cinn
43
aCinn + Ep-Cinn
44
aCinn + Ep-Cinn
45
aOBz
2H
2H
pOH
pOAc
pOAc
aOAc
OAc
26
46
aOBz
2H
2H
POH
pOAc
==o
aOAc
OAc
26 1
! 47
aOBz
2H
2H
POH
pOH
pOH
aOAc
OAc
26
! 48
aOBz
2H
2H
POH
POH
pOBz
aOAc
OAc
26 1
49
aOBz
2H
2H
POH
pOAc
POH
aOAc
OAc
26
50
aOBz
2H
2H
pOH
pOAc
2H
pOAc
OAc
26
51
aOH
aOBz
2H
POH
POAc
2H
POBz
OAc
27 1
52
aOAc
aOBz
2H
POH
POAc
2H
53
aOAc
aOBz
2H
pOH
POAc [ aOBz
54
aOBz
pOAc
2H
POH
POAc
55
aOBz
aOAc
2H
1 POH
POAc
56
aOAc
2H
2H
POH
POBz
1 57 _ aOAc L?8_._ ^ocOB^
2H
2H
POH
pOAc
__iH
! POH
pOAc
2H
pOBz
OAc
27
POBz
OAc
27
1 2H
pOBz
H
28
2H
pOBz
2H
pOAc
2H
1 pOBz
OBz
28
1. 2H
pOAc
OBz
28
i 28 1 OAc 1 28 H
751
L^i
aOAc
aOAc
2H
pOH
pOAc
=o
aOBz
OAc
i 60
aCinn
aOAc
2H
POH
POAc
=0
aOBz
H
29
61
aOAc
2H
2H
H
POAc
aOAc
pOFu
H
30
62
aOAc
2H
2H
pOH
POFu
2H
pOFu
H
30
2H
pOH
pOFu
aOAc
_pOFu
H
30
_63_
aOA£_ _ _ 2 H _
29
Ep-Cinn = Epoxycinnamate esters Cinn = Cinnamate ester 0-Fu = Furoate esters
The maytolins are an interesting instance of new sesquiterpenesfix>mthe Celastraceae characterized by the presence of a tetrahydro-oxepine nucleus. It would seem that these new types of skeleton are only biosynthesized by species of the MORTONIA genus, which consists of just four species, endemic to Mexico and the southern Unites States. The chemical study of three of these four species led to the isolation and characterization of eight new sesquiterpenes [3133](Figure3). OBz
OBz OBz
082 OBz
COOH 64 Mortonin A, R=H
66 Mortonin C
HO^i 67 Mortonin D
68 Mortonol A, R=H 69 Mortonol B, R= OAc
65 Mortonin B, R=OAc
Figure 3. Sesquiterpenes isolated fix)m the genus Mortonia
The structures proposed
for MORTONINS A and B are the first recorded example of a
natural product in which ring B of the eudesmane skeleton undergoes oxidative cleavage to the the y-lactone. The subsequent isolation and characterization of the di-ester ketone MORTONOL (68) from M. greggi suggests that this sesquiterpene might be the biogenetic precursor of the whole MORTONIN series (Figure 4).
752
64.66
67
Figure 4. Possible formation of Mortonins sesquiterpenes
Boariol [18,23] is another new sesquiterpene isolated from the Chilean species M boaria Mol. which does not conform to the classic model of the sesquiterpenes previously described, and is in fact the simplest of all the compounds recorded from the Celastraceae. 'H and ^^C nmr studies showed the presence of a secondary and a tertiary OH, the latter at C-4 but with the opposite configuration to the customary p-hydroxyl at this position. The application of the Horeau method and an X-ray diffraction study confirmed the absolute configuration of the compound [18,23]. The absence of substituents at C-1, another notable feature of this structure, casts doubis on the biogenetic theory for the P-dihydro-agarofuran sesquiterpenes from this family which presumes that such substituents are present in nature. The possibility that boariol (34) might be an artifact was ruled out on the basis of two data: several sesquiterpenes with the classic C-4 d-OH configuration have been isolated from M. boaria Mol., even some with C-3 substitution; no products were obtained with carbonyl groups at C-3 and without hydroxy groups at C-4 which could have been hydrated non-stereospecifically via enol formation [18,23]. Fig. 5.
753 B. SESQUITERPENE ALKALOIDS Sesquiterpene alkaloids have similar structures to polyester sesquiterpenes except that the hydroxy groups of the eudesmane basic skeleton are esterified by nicotinic acid and/or its derivatives.
Little has been published about sesquiterpene alkaloids from American species
which tend to be found in the roots of the plants (Table II).
TABLE 11. N4AYT0LIN-TYPE SESQUITERPENE ALKALOIDS 15 10
f. M>" *13
Form
C-1
C-2
C-3
C-4
C-6
70
aOBz
2H
2H
pOH
pONic
71
aOCinn
POH
2H
pOH
pONic
72
aOBz
2H
2H
POH
pONic
73
aOBz
2H
2H
pOH
pONic
74
aOAc
aONic
2H
POH
I 75
aONic
aONic
2H
pOH
76
aOAc
aONic
2H
77
aOAc
aONic
78
aONic
79
aOAc
C-9
j C-8
C-15
Ref
pOAc
H
1 34
2H
POAc
H
34
POAc
aOAc
H
34
pOH
aOAc
H
34
pONic
aONic
pOBz
OAc
35
pOBz
aONic
POBz
OAc
35
pOH
pOBz
aONic
pOBz
OAc
35
2H
pOH
pOAc
aONic
POBz
OAc
35
aONic
2H
pOH
pOAc
aOBz
pOBz
OAc
35
aOBz
2H
H
pOAc
2H
PONic
OH
36
2H
'
80
aOAc
aOBz
2H
H
pOAc
2H
PONic
OAc
36
81
aOAc
aONic
2H
(30H
POH
aONic
pOBz
OAc
37
82
aONic
aONic
2H
pOH
pOH
aONic
pOBz
OAc
37
83
aONic
aONic
2H
pOH
1 pOAc
aONic
\ pOBz
OAc
37
1 ^"^ 1 85
aONic
aOAc
pOH
POH
aONic
POBz
OAc
: 37
aOAc
1 aONic
L ^QAc
1 aONic
2H
1 2H
L_H
1 j
1 pOBz 1 OAc 1 37 __..
C. N4ACROCYCLIC SESQUITERPENE ALKALOIDS Celastraceae also elaborate other, more complex, alkaloids, also polyester sesquiterpenes, incorporating a macrocycle derived from an evonic, wilfordic, cassinic or other type p>Tidine dicarboxylic acid with an additional alkyl chain of the basic eudesmane cycle at C-3 and C-7 (Table III). Celastraceae alkaloids are well-documented for the European and Asian genera, particularly Catha, Celastrus, Euonymus and Trypterigium but are relatively rare among the Latin
754 American species. Except for a few from the Hippocratea, Peritassa and Orthosphenia genera, most new Celastraceae alkaloids have been obtained from species of Maytenus. As in the case of the polyester sesquiterpenes, structural elucidation has been based on 1H13C nmr correlations (HETCOR) and long range inverse detection (HMBC and HMQC). Relative configurations have been determined by the combined use of NOESY experiments. The absolute configuration of almost all the compounds was established by circular dichroism applications using the exciton chirality method in 1,2-dibenzoate systems. TABLE III. MACROCYCLIC ALKALOIDS FROM AMERICAN CELASTRACEAE (Wilfordate type)
!
Form
R'
R2
R3
COMPUESTO
Rcf|
86
OBz
OBz
OAc
EbcnifolincW-l
35
87
OBz
OBz
OH
Ebenifoline W.2
35
88
OBz
OAc
OAc
Euojaponine F
35
89
OAc
OAc
OAc
Euonine
35
1
90
OAc
OBz
Cangorinine W-I
36
1
91
OAc
OBz OBz
ONic
CingminiDe W-II
36
TABLE IV. MACROCYCLIC ALKALOIDS FROM AMERICAN CELASTRACEAE (Evoninoate type)
755
Rl
R^
R^
R4
R5
R7
R«
COMPUESTO
i ^^
OBz
OH
OH
OAc
OAc
OAc
H
OAc
EbcnifolincE-l
Rcf. 1 38 1
I 93
OB7
OAc
OH
OBz
OAc
OAc
H
OAc
Ebenifolinc E-2
38
OAc
OBz
OAc
H
OAc
Ebcnifolinc E-3
38
OAc
OAc
H
OAc
Ebenifolinc E-4
38
R6
94
OBz
OAc
OH
95
OBz
OAc
H
OAc
1^
OBz
OAc
OH
OH
OBz
OAc
H
OAc
Ebenifbline E-5
38
97
OBz
OH
OH
OBz
OAc
OAc
H
OAc
EuojaponineC
38
98
OBz
OAc
OH
OAc
OAc
OAc
H
OAc
Mayteine
38
99
OAc
OAc
OH
OAc
OAc
OAc
H
OAc
Euooymine
38
100
OAc OAc
OH
OBz
OAc
OAc
H
OAc
CangoriniE-I
39
101
OAc
OAc
OH
OAc
OBz
H
OAc OAc
Horridtne
40
I 102 OAc OAc
OH
OH
OAc
H
OH
OAc
Acanthotfaamine
41
103
OBz
CNMP
OH
OAc
OAc
OAc
H
OAc
Hippocrateine I
[AQ4_
OAc
CNMP
OH
OH
OAc
OAc
H
Mb
Hippocratcine n
42 J2
j
CNMP= 5 Carboxy-N-methylpiridonyl Mb= 2-Methylbutyroyl Orthosphenin (105) breaks the classical mould of the Celastraceae macrocyclic alkaloids described to date and is the only example of an evoninol nucleus with an oxo group at C-8 and residual cassinic acid. Its structure was ascertained by the spectroscopic methods mentioned above, hydrolysis and the preparation of derivatives [43]. Two new evoninate-type alkaloids have recently been described, peritassin A and B, obtained from species of the genus Peritassa. These structures are distinguished by the macrocyclic unit which consists of an evoninic acid isomer in which the pyridine ring of the dibasic acid is substituted at 4'-5' instead of the more usual substitution of evoninic acid at 2'-3' [44].
R«OAc PcritassineA R = OBz PeritassincB Figure 6. The Structures of Orthosphenin and Peritassin A and B
756 IV. DITERPENES In general, very little has been written about diterpenes from the Latin American Celastraceaeas these structures are not often found. Abietriene type diterpenes have been the general rule in the Celastraceae although the chemical study of the minor constituents of Orthosphenia mexicana and Rzedowskia tolantonguensis did enable pimarane type diterpenes to be isolated and chemically characterized [43,45] and the second of these species afforded a series of new diterpenes with an isopimarane skeleton, described for the first time in the Celastraceae. The structure of the diterpene 107 (C20H30O3) was established by spectroscopic methods and confirmed by x ray diffraction studies while, under nmr, the nor-diterpenes 109-113 proved to be structures with an exocyclic methylene and no carboxylic groups at C-4 and are assumed to be the result of an oxidative decarboxylation process as has occurred elsewhere. Orthosphenia mexicana yielded another new diterpene of the nor-isopimaradiene type (C19H28O3) related to the abovementioned products [43]; spectroscopic analysis and chemical trans-formations established its structure with a tertiary hydroxy 1, an a,p-unsaturated keto group and the presence of a typical vinyl system of the ABX type.
CH2OH
W^' Fig.
106 107 108
Rl (M COCHi CH2OH
R2
0 2H 2H
Ref 43 45 45
CHzOH
R2.„
^
Fig. 1 109
110 111 112 113
Rl 0 BOH aOH 0 POH
R2
H H H
(m OH
Ref
757 V. ALKALOIDS M. loesner Urb. and M buxifolia (A. Rich) Griseb collected on the island of Cuba have been extensively studied by H. Ripperger et al. [46-47] who isolated a series of new macrocyclic alkaloids of the spermidin type, commonly found in the Celastraceae family; the new alkaloids could be related to others akeady obtained by Kupchan's group. feHs o H
Fig
R
COMPOUND
Ref
114
OH
Mayfoline
46
115
OAc
N( 1 )-acetyl-N{ 1 )-deoxymayfoline
46
"T^ ^
^3^^..^^' OAc
Fig.
R
COMPOUND
Ref
116
C^rl] jK^H^^H^H-
Loesenerine
47
Q')T{^-Qr\\^-Q\\-(Z')r{^-
17.18-Didehydroloesenerine
47
CnHs-CHOH-C4H4-
16,17-Didehydroloesenerin-18-ol
47
i 117
118
VI. TRITERPENES A. TRITERPENES FROM THE AMERICAN CELASTRACEAE The triterpenoids hitherto described for the Celastraceae almost invariably belonged to the FRIEDO-OLEANE series (including methylene quinones and phenolic compounds), LUPANE, OLEANE, GLUTINANE AND TARAXERANE series. Characteristic of the family are the triterpene methylene quinones synthesized in the roots of the plant and considered as taxonomic indicators and the same holds true for the American species. To date about 12 different endemic species belonging to eight different genera have been studied and 26 new triterpenes have been described as well as new triterpenoid dimeric structures. As is usual, all the species studied have a broad range of biological activity probably due to the presence in most, of triterpene methylene quinones of known biological effect such as pristimerin, celastrol, tingenone, iguesterin [48] etc.
758 Particularly interesting has been the case of Orthosphenia mexicam which yielded five new triterpene methylene quinones with a new carbon skeleton, a greater degree of conjugation than hitherto reported, an extra 14-15 double bond and a rearranged methyl at C-15. Its structure was elucidated by a succession of chemical transformations, spectroscopic methods and X ray diffraction which determined the absolute configuration of this compound [49,50]. TABLE V. METHYLENE QUINONE TRITERPENES
Rl
R2
CO^Me
OH
CO^Me
OH
H
C07Me
OH
H
CO^Me
OH
H
CO^Me
H
H
R3
H
R4
R5
COMPOUND
Ref
0
Me
Netzahualcoyone
49
H? OH
Me
Netzahualcoyonol
50
Me
Netzahualcoyondiol
50
H? H-,
CO^Me
Netzahualcoyol
50
Me
Netzahualcoyene
50
Rl
R2
R3
COMPOUND
CO^^CH:,
H.
Pristimerin
50
H
H? 0
50
0
H7 OH
Tingenone
H
22-6-Hydroxy-tingenone
51
OH
0
H.
20-a-Hydroxy-tingenone
51
Celastrol
50
i COOH
JH2_
ik^
Ref
!
759
R2
R» CO,CH,
H
COMPOUND
Rcf,
Hz
Isopristtmerm in
52 1
0
Isotingcoooc m
. _52 . ]
TABLE VI. FRIEDELANE TYPE SKELETON TRITERPENES
[RJ
R3
R^
o o
H:
0
0
H:
H2 oOH H2 aOH 0
[CH,
0
H2
H?
CH^
0
H7
|CH,
o o o
CH, CH^ CH,
|ca>H
rco,cH^ [COjH
R2
H?
R5
R6
COMPOUND
Rcf
CH,
FricdcUmc-3,15-
53
1
CHi
15-a-Hydroxyfricdelin
53
1
CHi
15-a-Hydroxyfricdelffie-1 »3-iliooe
53
1
CH,
Friedeliii
53
1
H7
H2 0
CH,
Fncdetane-1,3-
53
]
% H2
H2.
CChCHx
3-Oxo-firiedooleaii-2S-oic metfayi ester
54
1
Maytenoic acid.
52-55
H7
H7
H?
CH,
H: fiOH 0
H2 H2
H2
CH,
Methyipopubiooate
55
H2
CH3
3^Hydroxy-2-oxofriedclan-20a-
55
H:
CH,
3-Oxofricdoolcan 20 a-hydroxy
52
H2
CH,OH
Canofilol
54
CH;
2a'Hydroxy populnonic ac.
51
cartrnxyik ac.
I CHt^OH 0
[CH^
0
1 CO^H __ _o
H: H2 H^ H: OH H^
ii2_
1
760 TABLE VII. FRIEDELANE TYPE SKELETON TRITERPENES
[RT" R2
R3
CX>MPOUND
Ref.
3-Hidroxy-2-oxofriedeliii-3-cne-20a-c«boxylic acid.
55 55
OH
CH,
H
OH
CH,
CH, 3-Hi(fax)xy-2-oxofiiedeian-3-aie-20a-indliyl caiboxyl
OH
CHO
H
Cangoronine
CH,
H
2-Oxofnedooiean-3-en-29-oic acid.
[H
52
J6
When the carboxy group at C-24 in cangorinin undergoes oxidative elimination, followed by oxidation of the A/B ring, pristimerin type triterpenes are obtained. The isolation (for the first time from a natural source) of both isopristimerin and isotingenone in the one plant bears out the biogenetic theory of the biomimetic conversion of pristimerin and tingenone, isopristimerin III and isotingenone III put forward some time ago by Monache et al [57].
TABLE VIII. FRIEDELANE TYPE SKELETON TRITERPENES
Rl CO:,H 1
H
R2
R3
H2 CH, 0 OH
COMPOUND
Ref
2,4(23)-Friedeladien-29-oic ac.
51
2,4(23>Friedeladicn'22B-hydroxy-21 -one
51
761 CHjOH
Figure 7. D:A-friedoolean-l-en-29-ol-3one (Ilicifolin) [52]
120
Figure 8. Transposed Friedo-olene Type Skeleton Triterpenes [58] 3-0X0-25(9-^8) abeo-fi4edoolean-(4)(23)-en-24->l-olida (119) y 3-oxo-25(9->8) abeo-J5nedoolean-(4)(23)-en-24-al (120) transposedfriedo-oleanetype skeletons isolated from Schaefferia cuneifolia
TABLE IX. OLEANE SKELETON TRITERPENES
762
R»
R2
=o
R3
R4
F5
R6
R7
R8
COMPOUND
ReJ
CH7OH
CH2OH
H
H
H
H
3-Oxo-28,29-
59
36,29-Dihydroxyolean-12-ene
59_
16,36,11 a-Trihydroxyolean-12-
60
12-ene OH
H
CH,
CH'^OH
H
H
H
H
OH
H
CH,
CH-,
OH
OH
H
H
cne OH
H
CH,
CH,
H
OH
H
H
36,11 a-EHhydroxyoIean-12-ene
60 J
OH
H
CH,
CHi
H
H
H
H
ft-Amyrm
60
OH
H
CH,
CH,
H
H
OH
H
36,15a-Dihydroxyolcan-12-eDC
61
OH
H
CH,
CO7H
H
H
H
H
Epikatonic ac.
55
CH,
CO-^H
H
H
H
H
Katononic ac.
55
=o OH
H
COOH
CH,
H
H
H
H
Oleanolic ac.
62
aOH
H
CH,
CO,H
H
H
H
aOH
Maytenfolic ac.
55
TABLE X. LUPANE SKELETON TRITERPENES
IRI i
R2 =0
R3
R4
R5
COMPOUND
Ref.
OH
CH,
H
3-Oxo-lup-20-en-30-ol
63,64
OH
H
OH
CH,
H
Lup-20-en-3B,30diol
63,64
OH
H
H
OH
H
Betulin
54,59,63
OH
H
H
CH,
H
Lupeol
54,60,62
OH
H
H
COOH
H
Betuiinic acid
62
H
CH,
aOH
3-Oxolup20-en-ll-ol
55
H
H
CH,
aOH
Li^)-cn-36,ll-diol
65
H
COOH
H
Betulonic acid
18
=0 OH =0
1
1
763 TABLE XI. OLEANANE A^« TYPE SKELETON TRITERPENES
Ri
R2
R3
R4
R5
COMPOUND
Ref.
OH
H
OH
H
H
Olean-18-ene-3B,166diol
66
OAc
H
OAc
H
H
3B, 16B-Diacctoxyolean-l 8-cnc
66
OH
H
H
16B-Hydroxyolean-18-en-3one
66
H
OH
H
16a-Hydroxyolean-18-en-3one
66
H
H
OH
3B, 11 a-Dihydroxyolean-18-ene
60
H
H
H
3-Oxo-olean-18-ene
51
=0
=o H
OH =0
TABLE XIL GLEANAlsfE A^ TYPE SKELETON TRITERPENES
Rl
R2
COMPOUND
Ref. J
OH
CH:,OH
36,29-Dihydroxyglutm-5-ene
52,60
OH
CH,
3B,Hydroxyglutin-5-ene
60
TABLE XIII. OLEANANE A^ 1'12 xypE SKELETON TRITERPENES
764
^\ H
R? ^^ COMPOUND OH H 3 &-Hydroxyolean-9( 11), 12-diene =0
H
Ref.
60 60
3-Oxo-olean-9( 11). 12-diene
TABLE XIV. OLEANANE TYPE SKELETON TRITERPENES (with a hemiacetal 24-hydroxy-3-Keto) ~ »C02H
RJ
R2
R3
OAc
H
H
3-Acetoxy-salaspenmc ac.
63
OH
OH H
Orthosphenic ac.
63
OH
OH OH 66-Hydroxyorthosphenic ac.
50
OH
H
52,59,63
H
COMPOUND
Salaspermico ac.
Ref.
Friedo-oleane triterpenes with a hemiacetal 24-hydroxy-3-oxo grouping are exclusive to the Celastraceae and several have been isolated from South American species. It has been proposed that these compounds may be intermediates in the biogenetic pathway to the Celastraceae triterpene quinones.
B. DIMERIC TRITERPENES Triterpene dimers have recently been reported and are generally characterized by the presence of modified triterpene quinone monomer units. To date, ten dimers based on tingenone and/or pristimerin units have been described, most of them from the Maytenus genus [67]. A possible biogenetic route to triterpene quinones with an extra double bond (e.g., netzahualcoyone and derivatives) may involve dehydrogenation and subsequent pristimerin regrouping. Indeed, (Table V) two epimeric di-triterpene quinone ethers umbelatin a and p (118,119) have been isolated from the roots of Rzedowskia tolantonguensis [68]. The structures of pristimerin and netzahualcoyone were ascertained by the standard spectral methods. The ^H
765 nmr spectrum of dimers showed two carbomethoxy groups, 12 angular methyls and six lowfield protons and correlation with those of pristimerin and related phenol, zeylasterone 2,3dimethyl ether (Tables XV and XVI) established the position of the C-23, C-25, C-26, C-28 and C-30 methyls. Treatment of pristimerin with 2,3-dichloro-5,6-dicyanoben2oquinone (DDQ) in dioxan afforded a mixture of four products separated by preparative tic. The least polar product was netzahualcoyone followed by the phenolic compound, 123; the other two products of medium polarity proved identical to the natural dimeric epimers 121 and 122 (Fig. 9). The behaviour of pristimerin in these conditions resembles the mechanistic theory of Barton about the formation of usnic acid [69]. These dimers, accordingly, could be synthesized by generating two radicals from a single precursor, in this case, pristimerin (Figure 10). These ideas have recently been bolstered by the discovery of another dimer, from M. umbelata, consisting of oxidated tingenone units. When tingenone was treated with DDQ in dioxan, various products were formed, two of which proved to be 124 and 125, confirming the proposed structures and the general way that methylene triterpene quinones act with DDQ [70]. Itokawa et al. [71] have recently isolated and characterized four new dimeric triterpenes from M. illicifolia (Paraguay), three of which have two units of pristimerin while the fourth (cangorosin B) has one pristimerin and one tingenone unit (Fig. 11). Magellanin [67], another example of a dimeric triterpene ether composed of modified pristimerin units (Fig. 12) was obtained from the roots of a Chilean species of the Maytenus genus. Its upper unit is modified pristimerin with an epoxide between C-3 and C-4; the structure of the lower is similar to that of cangorosin B. Sixty signals were observed in the ^^C nmr spectrum. The epoxide was confirmed by the chemical shift of C-3 and C-4 at 6 91.85 and 78.70 very like those described for Itokawa's dimer [71] and the presence of an aromatic A ring with a hydroxy 1 at C-3' and an ether bridge on C-2' could also be deduced from this sp)ectrum (Table XVII). Its ^H nmr spectrum unambiguously assigned the double bond between C-6' and C-T (5 6.63, dd, 1=2.66,10.20 Hz; 6 5.90, dd, J=2.48,10.20 Hz) and the H-l' proton (6 6.70, br s). The C-H couplings were determined by 2D heteronuclear inverse detection methods (HMQC and HMBC) (Tables XVIII and XIX) and its MS revealed fragments at m/z 464 (C30H40O4), 466 (C30H42O4), 480 and 450 corresponding to the two possible types of signal for cleavage around the ether group while lesserfragmentsreflected the typical rupture of pentacyclic triterpenes [67]. As with other dimers, magellanin may be the result of in vitro or in vivo auto-oxidation due to radical coupling [67].
766
+ 121 • 122
Netzahualcoyttno R«Ri«R2»H N«tz«hijatcoypn« R«OH; Ri •Ra «0
Fig. 9. Experimental Demostration oftfiePossible Biogenesis of Celastraceae Dimers
OOCH,
H-O
=o«' |DDQ
121 •
I>1 y;ir
0^
y ^
diketo radical
122
Fig. 10. Probable Formation of Dimers 121 and 122 by Radical Coi^ling
767
121 Ri=a-Me; R2=CCX)Me; R3=R4=H
Umbellatina
122 Ri=p-Me; R2=COOMe; R3=R4=H
Umbellatin p
124 Ri=a.Me; R2=H; R3=R4=0 125 Ri=a-Me; R2=H; R3=R4K) Fig. 11. Some Dimeric Triterpenesfromthe Celastraceae C00CH3
CO2CH3
C00CH3
Cangorosin B
Cangorosin A Atropcangorosin A Dihydroatropocangorosin A CO2CH3
CO2CH3
Mageilanin Fig. 12. DimersfromM. ilicifolia and M. magellanica
768 TABLE XV. ^^c NMR ( 50 MHz ) Data ( 6, CDCb, Chemical Shifts in ppm Relative to Me4Si) of Pristimerin and Ethers 121 and 122 Pristimerin
122*
121'
C
jC
c
C
119.0(d)
! 110.8(d)
114.9(d)
11.3(d)
2
178.4(s)
179.2(8)
173.4(8)
188.0(8)
115.3(d) 174.0(8)
i3
146. l(s)
171.3(8)
145.3(8)
171.5(8)
144.7(8)
1
u
117.0(8)
j 91.2(8)
j 124.0(8)
92.1(8)
|5
127.5(s)
128.5(8)
132.0(8)
127.7(8)
6
133.9(d)
129.0(d)
189.6(8)
126.7(d)
7
118.1(d)
117.4(d)
126.3(d)
116.2(d)
8
169.9(8)
164.5(8)
151.3(8)
161.4(8)
9
42.9(8)
38.8(8)
44.0(8)
38.2(8)
124.0(8) 130.1(8) 189.0(8) 126.2(d) 150.5(8) 41.9(8)
lio
164.7(8)
137.7(8)
151.3(8)
137.7(8)
151.0(8)
11
33.6(t)
33.0{t)
34.1(t)
33.0(t)
34.2(t)
[l2
29.7(t)
29.5(t)
29.7(t)
29.7(t)
|l3
39.4(8)
39.1(8)
39.3(8)
39.0(8)
[l4 lis
45.0(8)
44.5(s)
44.5(s)
44.7(8)
28.7(t)
28.7(t)
29.4(t)
28.7(t)
|l6
36.4{t)
36.5(t)
36.5(t)
36.4(t)
29.8(t) 39.9(8) 44.2(8) 28.6(t) 36.5(t)
|l7
30.6(8)
30.7(s)
30.7(s)
30.6(s)
30.6(s)
|l8
44.4(d)
44.8(d)
44.8(d)
44.5(d)
44.7(d)
|l9
30.9(t)
30.9(t)
31.0(t)
30.9(t)
31.0(t)
bo
40.4(s)
40.6(8)
40.7(s)
40.5(8)
40.5(8)
21
29.9 (t)
29.8(t)
30.0(t)
29.9(t)
29.9(t)
|22
34.8(t)
34.9(t)
35.1(t)
34.9(t)
34.9(t)
[23
10.2(q)
24.7(q)
13.3(q)
22.5(q)
12.8(q)
[25
38.3(q)
37.8(q)
40.2(q)
37.6(q)
37.7(q)
|26
1 21.6(q)
21.0(q)
22.5(q)
1 20.9(q)
S 22.2(q)
|27
18.3(q)
18.3{q)
18.6(q)
18.3(q)
18-6(q)
|28
31.6(q)
31.7(q)
31.7(q)
31.6(q)
! 31.6(q)
|29
178.7(8)
179.0(8)
1179.0(s)
179.0(s)
179.0(8)
[30
32.7(q)
33.0(q)
j 33.0(q)
32.7(q)
32.8(q)
[31
1 31.6(q)
L5L6(q)
! 51.8(q)
1 51.5(q)
1 51.5(q)
^ The values of the pairs C and C may be interchanged.
1 j 1 1 1 j 1 1 1
1 1 1 1 1
769 TABLE XVI. ^H NMR ( 200 MHz ) Data (5, CDCI3, for The Methyls. Zeylasterone Pristimerin
121 H*
H
122
2,3-DmiethyI ether
H
H*
23.Me
2.21
1.37 2.72
1.41
2.73
25-Me
1.48
1.48
1.57
1.47
1.58
1.60
26-Me
1.26
1.26
1.26
1.27
1.25
1.32
2.66
27-Me
1.10
1.05
1.08
1.06
1.08
1.12
|28.Me
1.18
1.16
1.16
1.16
1.16
1.18
30-Me
0.53
0.52 0.54
0.53
0.54
0.60
1
TABLE XVU. C NMR. (100 MHz) (6, CDCI3) Pristimerin
Magellanin
Pristimerin
Magellanin
C
c 1
C.28
31.6
31.5
31.8
C-1
119.0
115.6
108.0 1
C-29
178.7
178.9
179.3
C-2
178.4
191.1
140.0 1
C-30
32.7
32.8
32.2
C-3
146.1
91.8
C.31
51.6
51.6
51.6
C-4
117.0
78.7
C.5
127,5
130.7
C.6
133.9
126.3
C-7
118.1
116.3
C-8
169.9
160.5
137.6 122.4 125.0 124.0 129.1 45.5
C-9
42.9
41.6
38.2
1 1 1 1 1 1
TABLE XVm. H NMR. (200 MHz) Magellanin H-1
CUCD
C.D.
6.06 d
6.07 d
J(Hz) (1.16)
(1.44) 6.42dd
C-10
164.7
173.7
143.7
H-6
C-11
33.6
32.9
31.2
J(Hz) (1.16,6.30)
C-12
29.7
29.7
29.5
C-13
39.4
39.0
38.9
C-14
45.0
44.5
44.3
H-y
6.70 brs
7.04 brs
C-15
28.7
28J
28^
H-6'
6,63 dd
6.60 dd
|c-16
36.4
36.4
36.3
J(Hz) (2.66,1020)
(2.85,9.88) 1
C-17
30.6
30.6
30.4
H-T
5.48 dd
C-18
44.4
44.4
44.2
\^&) (2.48,10.20)
IH-7
6.32 dd 5.92 d*
J(Hz) (6.30)
5.90 dd*
(1.44,6.46) 5.32 d (6.46)
(2.58,9.88)
[c-19
30.9
31.0
30.5
|c-20
40.4
40.4
40.5
|c-21
29.9
30.0
29.8 !
TABLE XIX. Three-bond^H-^^C
|c-22
34.8
34.8
36.0
coupling (HMBC) in Magellanin
|c-23
10.2
22.5
10.9
H-I
C-3,C-5.C-9
|c-25
38.3
34.9
22.2
H-6
C-4,C-8,C-10
|c-26
21.6
22.2
17.0
H-7
C-5,C-9,C-14
|c-27
18.3
18.3
17.5
H-r C-3',C-15'
•overlapping signals
1
770 VII. MISCELLANEOUS A number of heterogeneous natural products have been isolated from American species including aromatic and phenolic compounds, flavonoids, catechins etc. The following table indicates the main studies on the subject.
TABLE XX. SOME NON-CLASSIFIED PRODUCTS FROM THE CELASTRACEAE COMPOUND
COMPOUND
Ref.
2,6-Diacetoxy-7-hidroxy-8-metoxychromone
61
4,5-Dihydroblumenol A
72
1
Blumenol A
72
(-H'-O-methyl-epigallocatechin
65
Ouratea proanthocyanidin A
65
Dulcitol
65
j
Epicatechin
18
1
5'-0-Methylgallocatechin
18
1
4-Hydroxybenzalddiyde
18
Femiginol
9,73-75]
Sakuranctin
9, 73-75 1
Vni. TRANSFORMATIONS The Tenerife group which is responsible for about 70% of all the research published on the Latin American Celastraceae has concentrated on the isolation and structural characterization of secondary metabolites; ahnost incidentally they have also developed various transformations and partial syntheses to test biogenetic theories in vitro and prove structural correlations by means of chemical transformations [76]. Thus, a succession of transformations showed fiiedoleane triterpenes with hemiacetal 24hydroxy-3-keto grouping to be possible key intermediates in the biogaietic pathway of the Celastraceae triterpene quinones, and a triterpene with a hemiacetal group in the remote C-24 position was synthesized from fiiedelin, as shown in the scheme [76] (Fig. 13).
771
FrtocMki
(20% yield based on lactone) i) Na BH4 in ether ii) IBDA/12, py/CH2Cl2, 100 W tunsgten filament iii) t-butyl chromate/ ether iv) LiAlH4 V) K2O + n-bu4N^Cr/THF + (NH4)6Mo7024H2/K2C03, H2O2
Fig. 13. Synthesis of a Friedelane Triterpene with a 24-Hydroxy-3-oxo-hemiacetal Group
XI.
BIOLOGICAL ACTIVITY
A. ANTIFEEDANT ASSAYS It has been known for some time that some polyester sesquiterpenes of the p-dihydroagaroftiran type such as those IBrom the Celastraceae deter various msects from feeding. In China the powdered root bark of Celastrus angulatus [77] has been sprayed on crops to protect them against insect attack. Chemical and biological analysis has shown that the powder is active against various species of insects including the cucumber beetle (Aulacohora femoralis chinensis\ the crucifer beetle {Colaphellus lowringi), the cabbage work (Pieris rapae) and migrant locusts {Locusta migratoria migratorioides and Locusta migratoria manilensis). Wilfordin, tryptofordin and celangulin (Fig. 14) are antifeedant compounds obtained from extracts of the Celastraceae species Maytentis rigida [78], Trypterigium wilfordii [79] and Celastrus angulatus [77, 80], respectively, and as some products isolated from South American species have similar structural characteristics, they too have been assayed.
772
The insects used for assay were fifth-instar larvae of Spodoptera littoralis Boisduval {Lepidoptera, Noctuidae) and the methodology used to determined the FR50 was that described in references [18, 81]. Antifeedant activity has been discerned in 16 sesquiterpenes obtained from five endemic Latin American species. The results are set out in Table XXII. All compounds were active at a dosage of lO^g/cm^ with 72 the most active with FR5o<0.5 at a dose of 0.1 |ig/cm^. It is difficult to compare these resuhs with those given in the literature for other substances using different testing procedures [77]. However, the use of triphenyl tin acetate has previously been studied under similar conditions to those described herein with a FR5o=0.37 when the assay dose was 10 fag/cm^ [18, 81]. Therefore, more than eight of the new products in principle appear to be more active than triphenyl tin acetate, which is the standard for antifeedant assays. Compound 15 was compared using other standard methods [77], and the remaining sesquiterpenes showed moderate antifeedant activity against Spodoptera littoralis in the test applied.
PAC
Aco y
OAc
OBz
QOm
,.OAc
AcO*'.
OAc Wilfordin
Celangulin
Triptofordm
Fig. 14. Structure of Some Antifeedant Compounds from the Celastraceae
773 TABLE XXI. ANTIFEEDANT ACTIVITY OF SOME SESQUITERPENES ON SPODOPTERA LITTORALIS
OH 34 Boariol
Form
C-1
C-2
C-3
C-4
C'6
C-8
C-9
C-15
Source
1
aOAc
aOAc
H
H
pOAc
H
pOBz
CH3
M.ch.
3
aOAc
aOAc
H
H
pOAc
aOAc
POBZ
CH2OAC
M.ch.
4
aOAc
aOAc
H
H
pOAc
aOAc
pOBz
CH2OH
M.ch.
6
aOAc
aOAc
H
H
pOAc
H
pOBz
CH2OH
M.ch.
8
M.b.
M.b.
aOAc
aOAc
H
H
pOAc
aOAc
aOBz
CH2OH
M.ch.
15
OBz
POAC
POH
pOH
POAc
H
pOBz
CH3
M.m
16
OBz
POAc
pOAc
pOH
pOAc
H
POBz
CH3
M.b.
17
aOBz
POH
pOAc
pOH
pOAc
H
pOBz
CH3
M.b.
18
aOBz
POAc
pOH
pOH
H
H
pOBz
CH3
M.b.
20
aOBz
POAc
pOH
H
pOAc
H
pOBz
CH3
M.b.
22
aOBz
POAc
pOAc
pOH
H
pOAc
POBz
CH3
M.m.
34
H
H
OH
aOH
H
H
H
H
39
aOBz
H
H
pOH
pOAc
pOAc
aOAc
CH3
O.m. !
70
aOBz
H
H
pOH
PNic
H
POAc
CH3
Z.t.
72
OBz
H
H
pOH
pNic
pOAc
aOAc
CH3
Z.t.
1 73
aOBz
H
H
pOH
PNic
pOH
aOAc
CH3
O.m.
M.in.: Maytenus magellanica Lam. M.ch.: Maytenus chuhutensis Sjeg. O.m.: Orthosphenia mexicana Stand Leg. R.t.:
Rzedowskia tolantonguensis Med.
M.b.: Maytenus boaria Mo I.
M.b.
774 TABLE XXII. FEEDING RATIOS OF TEST COMPOUNDS
Compound 1 3
Dose (jig/'cm)
FR,,±S.M.D.
10
0.19 ±0.06
1
0.54 ± 0.09
10
0.13 ±0.05
1
0.71 ±0.09
4
10
0.07 ±0.16
1
0.53 ±0.32
6
10
0.36 ±0.09
8
10
0.16±0.15
1
0.65 ±0.25
10
0.24±0.19
1
0.44 ±0.12
10
0.56 ±0.11
15 16 17
10
0.13 ±0.05
18
10
0.63 ±0.11
20
10
0.24 ±0.19
22
10
0.38 ±0.05
34
10
0.52 ±0.10
39
10
0.12 ±0.07
1
0.40 ±0.02
70
10
0.68 ±0.14
72
10
0.04 ±0.03
73
1
0.15 ±0.15
0.1
0.45 ±0.28
10
0.74 ±0.06
j
i
FCj^ The ACDT/ACD proportion when 50% of CD has been consumed SNfi) Standard mean deviation.
775 B. ANTIINFLAMMATORY AND ANTIPYRETIC ACTIVITY Anti-inflammatory and antipyretic properties have recently been ascribed to global and purified methanol extracts obtained from the aerial part of Maytenus boaria (MoL), a quite widespread tree species in the rural areas of Chile and Argentina. The sesquiterpene extract showed significant antipyretic activity on fully-grown albino New Zealand rabbits following a modified pyrogenic test protocol [82, 83]. The study of antipyretic activity of the global methanolic extract was carried out at dosages ranging from 400 mg/kg to 1200 mg/kg, 800 mg/kg for the infusion and 30mg/kg for the enriched fraction. Antiinflammatory activity was studied using a dosage of 500 mg/kg of the global methanolic extract and 50 mg/kg of the enriched fraction. Assays showed that these substances exercised a pharmacologically significant anti-inflammatory effect [82, 83].
C.
CYTOSTATIC
ACTIVITY
OF
P-DIHYDRO-AGAROFURAN
SKELETON
SESQUITERPENES The products I, 3 and 6 [14] were assayed on cultures of P815 tumoral (mouse mast cell) and 3T3-LI non-tumoral (mousefibroblast)cells which were cultured in Dulbecco medium modified by Eagle, supplemented with 10% newborn calf serum, following the colorimetric method of Mosmann (1983) [84] which is based on the capacity of live cells to transform a colourless substrate into a coloured one. The cells (106 cl/ml) were incubated at 37**C in CO2 atmosphere on 96-hole cell culture plates together with the products to be studied predissolved in ethanol at concentrations of 40, 20, 10, 5, 2.5, 1.2, 0.6, 0.3, 0.16 and 0.08 ^ig/ml. After 22 h. incubation MTT (3-(4,5-dimethylthiazol-2-yl)-2.5-diphenyltetrazolio
bromide) pre-dissolved in saline
phosphate buffer, was added, left to incubate fw 2 h., after which HCl 0.04 N in isc^)ropanol was added. The optical density was measured on an ELISA spectrophotometer at 600 nm wavelength and the LD50 results were given in ^g/ml. These fmdings are set out in Table XXIII, and it can be seen that product 1 shows selective cytotoxicity to tumoral cells; its LD50 to P815 cells was much lower than that registered for 3T3-L1 (> 20 jig/ml) which means Aat further testing is in order.
TABLE XXIII: CYSTOSTATIC ACTIVITY P815
3T3-L1
1
6.2
>20
3
37
>20
6
36
>20
Compound
776 ANTIVIRAL ACTIVITY STUDY To study the antiviral activity. Type 1 simplex herpes virus, KOS stock (HSV-Hg) and Indiana vesicular stomatitis virus (VSV) were used and HeLa cells were cultured in Eagle-modified Dulbecco medium (EMDM), supplemented with 10% foetal calf serum, on plates with 24 holes. Monolayers of these cells were infected with HSV-1 or VSV at 0.5 ufp/cell or 0.01 ufp/cell, respectively, and later the product to be assayed, pre-dissolved in DMSO, was added in concentrations of 10, 20, 50, 100 and 200 ng/ml. After 48 h incubation for HSV-1 and 24 h for VSV, at 37°C in CO2 atmosphere, the cytopathic CPE effect was measured on a phase-contrasting microscope [85]. Protein synthesis was determined in the non-infected cells used as control and the HSV-1 and/or VSV infected cells, in order to obtain quantitative data on the toxicity and antivkal activity of the compounds assayed. Thus, 0.5 free medium methionine and 1 ^Ci/ml of methinione 32S were added to the cells and incubated at 37**C for 1 h. Thereafter the medium was siphoned off and the cells were washed with saline phosphate buffer and precipitated with 5% trichloroacetic acid (TCA). After 5 min, die acid was eliminated and the cells were washed twice with 96% ethanol,dried under an infrared light and dissolved in NaOH O.IN with DSS. A liquid scintillation counter was used for the recount. The results showed no antiviral activity for the dihydro-Pagarofuran sesquiterpenes.
ANTIMRICROBIAL AND CYTOSTATIC ACTIVITY OF TRITERPENES AND DIMER TRITERPENICS At the outset of this research programme into the American Celastraceae, it was already known that the meAylene quinone triterpenes pristimerin, tingenone, iguesterin, isoiguesterin and celastrol possessed a variety of biological properties. Campanelli et al [86] have studied the action mechanism of tingenone, previously described as antitumoral [87,89], and shown that the compound interacted with the bacterial DNA increasing its MT, probably by formation of extra hydrogen bonds between a hydroxy group in tingenone and the DNA phosphate groups. Moreover, tingenone exercises a marked inhibitory effect on DNA, RNA and protein synthesis in mouse fibroblast [90] and HeLa cells [91] as well as antibacterial activity, with a MIC of between 0.5 and 10 ^g/ml on Gram (+) bacteria [92] and none on Gram (-) bacteria [93]. Pristimerin, a compound structurally related to tingenone shows similar biological activity, being cytostatic and antibiotic [91,93]. Iguesterin is cytostatic to HeLa cells, isoiguesterin affects leukaemia and shows a moderate cytostatic activity on KB cells [94]. Celastrol, too, is known to be antibacterial to Gram positive bacteria [93]. Another series of methylene quinone triterpenes with a new type
777
of structure and extended conjugation consists of compounds isolated from various Latin American
species
such
as
netzahualcoyene,
netzahualcoyone,
netzahualcoyonol
and
netzahualcoyondiol [95-98]. With the exception of netzahualcoyene, these compounds show antibacterial activity' against Gram (+) bacteria but are inactive on Gram negative, the most active being netzahualcoyone which has an MIC of 1.5-1.6 fig/ml on Staphylococcus aureus. The presence of groups C=0 and OH in Ring E is indispensable and a ketone group at C-22 enhances the efficiency of this activity. Analysis of the minimum inhibitory concentrations of netzahualcoyone against Staphylococcus aureus shows it to be more active than some of the antibiotics used in clinical practice (Table XXIV).
TABLE XXIV. MIC against S. aureus of Netzacualcoyone and some antibiotic of clinical use. Antibiotic
MIC(^g/ml)
Netzahualcoyone
1.5-1.62
Oxolinic acid
1.0-4.0
Pipemidic acid
9.2
Cefradine
6.4
Cefradoxyle
1.6-6.4
Cefaclor
1.6-6.4
Cefotaxime
3.2
1
Netzahualcoyone inhibits various cellular processes: oxidative phosphorylation and the transport of glucose in sensitive bacteria [96,97] and also exerts a strong cytostatic activity on cultured HeLa cells; this activity disappears when the ftmctional groups in the A ring of the molecule are blocked (Table XXV).
TABLE XXV. Cytostatic Activity of Netzahualcoyone and Related Compounds Againts HeLa Cells. Compound Netzahualcoyone
ID50 (^g/ml) 0.1 1
Netzahualcoyondiol
1
Netzahualcoyonol Netzahualcoyene
<1 <1
Dimetox>'-dihydro-Netzahualcoyone
70.0
Mercaptopurine
0.1
778 Triterpene dimers with modified quinone or phenol skeletons also exhibit antimicrobial activity [68] and have also been synthesized [68,70]. Table XXVI shows the MIC (minimum inhibitory concentration) of the different products tested on Gram positive and Gram negative bacteria the former proving to be more susceptible to the activity of the compounds except for product 125 which had no effect whatoever on either. Bacillus species proved more sensitive than Staphylococcus aureus and this seems to be a general trait of quinone triterpenes such as netzahualcoyone, tingenol and pristimerin which are all more active on B. subtilis than S. aureus. The most active compound was 122, its MIC on B. subtilis (12 ^ig/ml) being particulary interesting. The dimers were less active than the monomers.
TABLE XXVI. MIC (mg/ml) of Dimer Compounds Against Gram Negative and Gram Positive Bacteria Bacteria
121
122
124
125
S. aureus
>100
70-80
>45
>100
B. subtilis
23-20
2-1
5-2.5 >100
B. cereus
23^0
10-8
30-25 >100
Salmonella sp
-
-
>45
>100
E. coli
>50
>50
>45
>100
-Not assayed All assays were carried out in triplicate
ACKNOWLEDGEMENTS
We are indebted to AIETI, CICYT Projects FAR 91-0472 and SAP 92-1028-CO2-01 and the Commission of European Communities, Project CI* 10505 ES(JR), for subsidies; we also are obliged to CONAF, for the plant material.
779 REFERENCES 1.
R. Bruning and H. Wagner, Phytochemistry, 17 (1978) 1821-1858.
2.
G.M. Muguera and J.M. Ward, Cancer Treatm, Rep. 61, (1977) 1333.
3.
K. Ujita, T. Fujita, Y. Takaishi, H. Takuda, S. Nishino and A. Iwashima. The 39^ Annual Meeting of the Japanese Society of Pharmacognosy Tokyo, Abstract Papers (1992) p.58.
4.
A. Takhtajan, Outline of classification of flowering plants. The Bot. Revw., 46 (1980) 225359.
5.
A. Gentry, Comparative ecology of African and South American arid to subhumid ecosystems. In Goldblatt P. (ed), Biological relationships between Africa and South America. Yale University Press, New Haven and London (1993), p. 500-547.
6.
E. Romero, Comparative ecology of African and South American arid subhumid ecosystems. In Goldblatt P. (ed). Biological Relationships between Africa and South America. Yale University press, New Haven and London.
7.
D. Hou, Annals of the Missouri Botanical Garden. Revision of the Genus Celastrus, 42
8.
J. Muller, Fossil pollen records of extant Angiosperms, The Bot. Rew, 47 (1981) 1-142.
9.
A.G. Ravelo, J.G. Luis, CM. Gonzalez, E.A. Ferro, I.L. Bazzocchi, J. Jimenez, J.R. Herrera,
(1955)215.
LA. Jimenez, Z.E. Aguilar, M. Placencia and O. Mufioz, Rev. Latinoamer. Quim, 19/2 (1988)72. 10. A.G. Gonzalez, LA. Jimenez, A.G. Ravelo, J.G. Sazatomil, I.L. Bazzocchi, Tetrahedron, 49, 697-702(1993). 11. Yomg Q. Tu, J. Chem. Soc. Perkin Trano, 1, (1991) 425- 427. 12. Z. He, H. Wu, M. Niwa and Y. Hirata, J. Nat. Prod., 57 (1994) 305-307. 13. C.R. Smith, R.M. Miller, D. Weesleder, W.K. Rohwedder, N. Eickman and J. Clardy, J. Org. Chem., 41 (1976)3264-3269. 14. A.G. Gonzalez, M.P. Nufiez, A.G. Ravelo, J.G. Luis, LA. Jimenez, O. Mufioz, J. Nat. Prod., 53(1990)474-478. 15. A.G. Gonzalez, M.P. Nufiez, A.G. Ravelo, J.G. Luis, LA. Jimenez, J.T. Vasquez and 0. Mufioz, Heterocycles, 29 (1989) 2287-22%, 16. O. Mufioz, A.G. Gonzalez, A.G. Ravelo, J.G. Luis, J.T. V ^ u e z , M.P. Nufiez, LA. Jimenez, Phytochemistry, 29 (1990) 3225-3228. 17. A.G. Gonzalez, M.P. Nufiez, A.G. Ravelo,J.G. Sazatomil, J.T. Vdzquez, I.L. Bazzocchi, E.Q.Morales and O.M. Mufioz, J. Chem. Soc. Perkins Trans 1.(1992) 1437-1441. 18. O.M. Mufioz, R. Ruiz, A.G. Gonzalez, M.P. Nufiez, I A. Jimenez and A.G. Ravelo, Helv. Chim. Acta, 76 (1993) 2537-2543. 19. A.G. Gonzalez, M.P. Nufiez, I.L. Bazzocchi, A.G. Ravelo and A.I. Jimenez, Nat. Prod. Len., 2(1993)163-170. 20. A.G. Gonzalez, M.P. Nufiez, LA. Jimenez, A.G. Ravelo and I.L. Bazzocchi, J. Nat. Prod., 56 (1993)2114-2119.
780 21. A.G. Gonzalez, I.A. Jimenez, M.P. Nufiez, A.G. Ravelo, I.L. Bazzocchi, O.M. Muftoz and M.A. Aguilar, J. Chem. Ecol., 20 (1994) 823-830. 22. A.G. Gonzalez, I.A. Jimenez, I.L. Bazzocchi and A.G. Ravelo, Phyotchemistry, 35 (1994) 187-189. 23. A.G. Gonzalez, O.M. Mufioz, A.G. Ravelo, A. Crespo, LA. Jimenez, I.L. Bazzocchi, X. Solans, C. Ruiz and V. Romero, Tetrahedron Lett., 33 (1992) 1921-1924. 24.
A.G. Gonzalez, CM. Gonzalez, A.G. Ravelo, B.M. Fraga and X.A. Dominguez, Phytochemistry, 27 (1988) 473-477.
25. M. Jimenez, E. Garcia, L. Gardida and A. Lira-Rocha, Phytochemistry, 27 (1988) 22132217. 26. A.G. Gonzdlez, I.L. Bazzocchi, A.G. Ravelo, J.G. Luis and X.A. Dominguez, Heterocycles, 24(1986)3379-3384. 27. A.A. Sanchez, J. Cardenas and L. Rodriguez-Hahn, Phytochemistry, 26 (1987) 2361-2362. 28. A.G. Gonzalez, J.J. Mendoza, A.G. Ravelo, J.G. Luis, LA. Jimenez, J.T.Vdsquez, X.A. Dominguez, M. Chang, Rev. Latinoamer. (Juim., 20/2 (1989) 91-94. 29. W. Vichnewski, J.S. Prasad and W. Herz, Phytochemistrv\ 23 (1984) 1655-1657. 30. J. Becerra, L. Gaete, M. Silva, F. Bohhnann and J. Jakupovic, Phytochemistry, 26 (1987) 3073-3074. 31. L. Rodriguez-Hahn, L. Antunez, M. Martinez, A.A. Sanchez, B. Esquivel, M. Soriano-Garcia and A. Toscano, Phytochemistry, 25 (1986) 1655-1658. 32. L. Rodriguez-Hahn, M. Mora, M. Jimenez, R. Sancedo and E. Diaz, Phytochemistry, 20 (1981)2525-2528. 33. M. Martinez, A. Romo de Vivar, E. Diaz, M. Jimenez and L. Rodriguez-H. Phytochemistry, 21 (1982) 1335-1338. 34. A.G. Gonzalez, CM. Gon2alez, I.L. Bazzocchi, A.G. Ravelo, J.G. Luis and X.A. Dominguez, Phytochemistry, 26 (1987) 21332135. 35. H. Itokawa, 0. Shirota, K. Ichitusuka, H. Morita and K. Takeya, J.Nat. Prod., 56 (1993) 1479-1485. 36. G. Baudouin, F. Tillequin, M. Kock and H. Jacquemin. Heterocycles, 22 (1984) 2221-2226. 37. H. Itokawa, O. Shirota, H. Morita, K. Takeya and Y. litaka, J. Nat. Prod., 57 (1994) 460470. 38. H. Itokawa, O. Shirota and H. Morita, J. Chem. Soc, Peridn Trans I (1993) 1247-1254. 39. O. Shirota, H. Morita, K. Takeya and H. Itokawa,. Heterocycles, 38 (1994) 383-389. 40. A.G. Gonzdlez, E.A. Ferro and A.G. Ravelo, Heterocycles. 24 (1986) 1295-1299. 41. A.A. Sanchez, J. Cardenas, M. Soriano-Garcia, K. Toscano and L. Rodriguez Hahn, Phytochemistry, 25 (1986) 2647-2650. 42. R. Mata, F. Calzada, E. Diaz and R.A. Toscano, J. Nat. Prod., 53 (1990) 1212-1219. 43. A.G. Gonzalez, L. San Andres, A.G. Ravelo, J.G. Luis, LA. Jimenez and X.A. Dominguez, J. Nat. Prod., 52 (1989) 1338-1341. 44. J. Klass, W.F. Tinto, W.F. Reynolds and S. MacLean, J. Nat. Prod., 56 (1993) 946-948.
781 45. A.G. Gonzalez, I.L. Bazzocchi, J.G. Luis, A.G. Ravelo, B.M. Fraga, X.A. Dominguez and A. Perales, J. Chem. Res. (S) (1986) 442-443, J. Chem. Res.(M) (1986) 3642-3660. 46. M. Diaz and H. Ripperger, Phytochemistry, 21 (1982) 255-256. 47. A. Preiss, M. Diaz and H. Ripperger, Phviochemistr\-, 27 (1988) 589-593. 48. G.R.C.B. Gamlath, G.M.K.B. Gunaherath, and A.A.L. Gunatilaka in New Trends in Natural Products Chemistry Studies in Organic Chemistry. Vol. 26. A. Rahman and P.W. Le Quesne (Eds.) Elsevier Sc. Publ. Amsterdam. 1986. p. 109. 49. A.G. Gonzalez, B.M. Fraga, CM. Gonzalez, A.G. Ravelo, E. Ferro, X. Dominguez, M.A. Martinez, J. Fayos, A. Perales and M.L. Rodriguez, Tetrahedron Letters, 24 (1983) 30333036. 50. A.G. Gonzalez, CM. Gonzalez, E.A. Ferro, A.G. Ravelo and X, Dominguez, J. Chem. Research (S) (1988) 20-21, J. Chem. Res. (M) 273-287. 51. R. Estrada, J. Cardenas, R. Esquivel and L.R. Halin, Phytochemistry, 36 (1994) 747-751. 52. H. Itokawa, O.Shirota, H. Ikuta, H. Morita, K. Takeya and Y. Itaka, Phytochemistry, 30 (1991)3713-3716. 53. J. Klass, W.F. Tinto, S. Mc Lean and W. Reynolds, J. Nat. Prod., 55, (1992) 1626-1630. 54. A.G. Gonzalez, J.A. Amaro and J. Gutierrez, Rev. Latinoamer. Quim., 17/1/2
(1986) 56-
58. 55. J.R. De Sousa, G.D.F. Silva, J.L. Pedersoli and R. J. Alves, Phytochemistry, 29 (1990) 3259326L 56. A.G. Gonzdlez, J.G. Luis, L. San Andr6s, J.J. Mendoza and A.G. Ravelo, J. Nat. Prod., 54 (1991)585-587. 57. F.D. Monache, M. Pomponi, G.B. Marini-Berttolo and 0. Gon^alves de Lima, Anales de Quim., 70,(1974)1040-1044. 58. A.G. Gonzalez, J.J. Mendoza, A.G. Ravelo, J.G. Luis and X.A. Dominguez, Rev. Latinoamer. (Juim., 23/1, 22/4 (1992) 22-24. 59. A. G. Gonzalez, CM. Gonzalez, A.G. Ravelo, X. Dominguez and B.M. Fraga, J. Nat. Prod., 49,(1986)148-150. 60. A.G. Gonzalez, E.A. Ferro and A.G. Ravelo, Phvtochemistrv-, 26 (1987) 2785-2788. 61. A.G. Gonzalez, J.J. Mendoza, A.G. Ravelo, J.G. Luis and X. A. Dominguez, Rev. Latinoamer. Quim., 22/1 (1991) 3-5. 62. A.G. Gonzalez, E.A. Ferro and A.G. Ravelo, Rev. Latinoamer. Quim., 17/1-2, (1986) 51-53. 63. A.G. Gonzalez, LL. Bazzocchi, A.G. Ravelo, J.G. Luis, X.A. Dominguez, G. V^quez and G. Cano, Rev. Latinoamer. Quim., 18/2 (1987) 83-88. 64. A.G. Gonzalez, J.L. Bazzocchi, A.G. Ravelo, J.G. Luis and X.A. Dominguez, Rev. Latinoamer. Quim, 20/1 (1989) 17. 65. M. Furlan, M.A. de Alvarenga and G. Akisue, Rev. Latinoamer. Quim., 21/2 (1990) 72-74. 66. A. G. Gonzalez, J.J. Mendoza, A.G. Ravelo, J.G. Luis and V. Dominguez, J. Nat. Prod., 52, (1989)567-570. 67. A. Gonzalez, A. Crespo, A. Ravelo and O. Mufioz, Nat. Prod. Lett.,4 (1994) 165-169.
782 68. A.G. Gonziiez, J.J. Mendoza, J.G. Luis, A.G. Ravelo and I.L Bazzocchi, Tetrahedron Letters, 30 (1989) 863-866. 69. D.M. Barton, A.M. Deflorin and O. Edwards, J. Chem. Soc. (1959) 530-531. 70. A. Gonzalez, J.S. Jimenez, L.M. Moujir, A. Ravelo, J.G. Luis, L.Bazzocchi and A.M. Gutierrez, Tetrahedron, 48 (1992) 769-774. 71. H. Itokawa, O. Shirota, H. Morita, K. Takeya, N. Tomiska and A. Itai, Tetrahedron Letters, 31(1990)6881-6882. 72. A. Gonzalez, J.A. Guillermo, A.G Ravelo, J.A. Jimenez and H.P. Gupta, J. Nat. Prod., 57 (1994)400-402. 73. O. Mufloz, Quimica de la Flora de Chile. Ed. D.T.I. Universidad de Chile. Santiago 1992. p.p. 265. 74. L. Mandich, M. Bittner, M. Silva and C. Barros. Rev. Latinoamer. Quim 15 (1984) 80-81. 75. M.P. Nuflez, Ph.D.Thesis, Universidad de La Laguna. Spain 1991. 76. A.G. Gonzalez, R.L. Dorta, A.G. Ravelo and J.G. Luis,J. Chem. Research. (S) (1988) 150151. J. Chem. Research (M) (1988) 1228-1238. 77. N. Wakabayashi, W.J.Wu, R.M. Waters, R.E. Redfem, G.D. Mills, A.B. DeMilo and W. Lusby,. J. Nat. Prod., 51 (1988) 537-542. 78. F. Delle Monache, G.B. Marini Bettolo and E.A. Bemays, Z. Angew.Entomol., 97 (1984) 406. 79. K. Yamada, Y. Snhizuri and Y. Hirata, Tetrahedron, 34 (1978) 1915-1920. 80. J.Liu, Z. Jia, D.Wu, J. Zhou and Q. Wang, Phytochemistiy, 29 (1990) 2500-2503 81. A.G. Gonzdlez, LA. Jimenez, A.G. Ravelo, X. Belles and M.D. Piulachs, Biochem. System, and EcoL, 20 (1992) 311-315. 82. N. Backhouse, C. Delporte, R. Negrete, O. Mufloz and R.Ruiz, Int. J. Pharm., 32 (1994) 239. 83. C. Delporte, Ph.D Thesis. Universidad de Chile (1993). 84. T. Mosmam, Journal of Inmunological Methods, 65, (1983) 55. 85. M.E. Gonzalez, B. Alarcon and L. Carrasco, Antimicrobial Agents and Chemotherapy, 31/9 (1987)1388. 86. A.R. Campelli, M. D^Alagni and G.B. Marini-Bettolo, Febs Lett., 122 (1980) 256-260. 87. C.F. Santana, J.J. Afara and C.T. Cotias, Rev. Inst. Antibiotics, Recife, 11 (1971) 37-43. 88. C.F. Santana, C.T. Cotias, K.V. Pinto, M.W. Satiro, A.L. Lacerda and L.C. Moreina, Rev. Inst. Antibiotics Recife, 11 (1971) 61-67. 89. A.M. Melo, M. Lobo Jardin, C.F. Santana, Y, Lacet, J. Lobo Filho, O. Gon9alves de Lima and I. D'Alburquerque, Rev. Inst. Antibiotics Recife, 14 (1972) 9-12. 90. P.V. Angeletti and G.B. Marini-Bettolo, II Farmaco, 29 (1977) 569-571. 91. A.G. Gonzalez, V. Darias, J. Boada and G. Alonso, Planta Medica, 32 (1977) 282-286. 92. O. Gon^alves de Lima, I. D'Alburquerque, J. Sidney de Barros, L. Coelho, D.G. Martins, A.L. Lacerda and G.M. Maciei, Rev. Inst. Antibiotics Recife, 9 (1969) 17-25. 93. O. Gon9alves de Lima, E. Weigert, G.B. Marini-Bettolo, G.M. Maciei, J. Sidney de Barros and L. Coelho, Rev. Inst. Antibiotics Recife, 12 (1972) 13-18.
783
94. A.T. Sneden, J. Nat. Prod, 44 (1981) 503-506. 95. L. Moujir, A.M. Gutierrez-Navarro, A.G. Gonzalez, A.G. Ravelo and J.G. Luis, Biochem. System. EcoL, 18(1990)25-28. 96. L. Moujir. A.M. Gutierrez-Navarro, A.G. Gonzalez, A.G. Ravelo and J.G. Luis, Antimicrob. Agents Chemother., 31 (1991) 211-214. 97. L. Moujir, A.M. Gutierrez-Navarro, A.G. Gonzdlez, A.G. Ravelo and J. Jimenez, Biomedical letters, 46 (1991) 7-15. 98. A.G. Gonzalez, A.G. Ravelo, LL. Bazzocchi, J. Jimenez, CM. Gonzalez, J.G. Luis, E.A. Ferro, A.M. Gutierrez-Navarro, L. Moujir and L. de Las Heras, II Farmaco, 6 (1988) 501505.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 18 © 1996 Elsevier Science B.V. All rights reserved.
785
Structural Chemistry of Glycolipids from Fungi and Protozoa
Eliana B. Bergter and Maria Helena S. Villas Boas
Introduction
Glycosphingolipids (GSLs) are ubiquitous membrane components present In almost all living organisms found In nature.These molecules are known to be Involved in a variety of
functions, such
as
cell-cell
interactions,
differentiation,
oncogenesis
and
immunogenlcity. Glycosphingolipids have been isolated from most types of normal human cells, human neoplastic tissues, calf
brain, marine
invertebrates
and
metacestodes. Improvements In chromatographic procedures,the use of nuclear magnetic resonance spectroscopy (NMR) and FAB-MS spectrometry have led to the characterization of an increasing number of new glycosphingolipid structures. The combination of these sensitive and powerful techniques have made possible the identification of a variety of glycosphingolipid molecules expressed on trypanosomatids and pathogenic fungi. The aim of this chapter is to cover general aspects of glycolipid structural diversity and more specifically the structural characterization and possible biological functions of glycosphingolipids from fungi and protozoa. Glycosphingolipids consist of a ceramide (N-acylsphingosine) unit and one or more sugars linked glycosidically to the terminal primary hydroxyl group of sphingosine (Figure 1). H3C[CH2]2CH=CH-CH-CH-CH20-X
I
I
OH NHCO-R X - Sugar R - Fatty acid
786 Glycosphjngolipids show considerable structural variation in the fatty acid, sphingosine and carbohydrate components resulting in a great number of chemically distinct glycosphingolipids. The major types of sugars that are found are glucose, galactose, fucose, Nacetylgalactosamine,
N-acetylglucosamine,
N-acetylneuramlnic
acid
and
N-
glycolylneuraminic acid or its derivatives. Sphlngosines are a group of related long-chain aliphatic 2-amino-1,3-diols (long-chain bases-LCBs). Sphingosine (4-sphingenin; 2-D-amino-4-octadecene-1,3 diol) occurs most frequently in animal glycolipids.Others that are often found in glycollpids include dihydrosphingosine (sphinganine) and 4-eicosphingenine (C20). Phytosphingosines constitute the major base component of plants (1), protozoa (2) and fungi (3). The fatty acid components of glycollpids are usually more complex than the long chain basesThey range in complexity from C14 to C24 in chain length. Glycosphingolipids have been classified according to their carbohydrate structure rather than their ceramide moleties.They are classified as neutral glycollpids, sulfatides (sulphate-containing) and gangliosides (sialic acid-containing). Glycosphingolipids can residues, such as
also be grouped according to the number of carbohydrate
mono-, di-, tri-, tetra-, penta-, to poly-glycosyl ceramide.The
structures of all major basic types of glycollpids and the main sources of glycosphingolipids are listed In Table 1. Isolation and purification of glycosphingolipids Several methods for the isolation and purification of glycosphingolipids are described in the literature (4,5,6,7). Glycosphingolipids are isolated from cells or tissues using a mixture of chloroform/methanol 2:1 and 1:2 v/v and partitioned according to Folch et al (8).This method is useful for separating gangliosides and highly polar glycollpids from relatively nonpolar liplds.The lower phase contains the neutral glycollpids with relatively short carbohydrate chains along with contaminating neutral lipids and phospholiplds.The upper phase contains most of the gangliosides and glycollpids with longer carbohydrate chains. Glycoproteins and glycopeptldes may also be present in this phase.The higher polar glycollpids present in the upper-phase fraction can be separated from the nonolipid
contaminants
by
gel
filtration
on
chloroform/methanol/water (1:1:0.01 v/v) (9) or by dialysis.
Sephadex
LH-20
in
-
TABLE 1 Main basic glycolipid structures Series Lacto
Core Structure R~Galpl-4GlcNAcpl-3Gal~l-4Glcfll-l-Cer Galpl-3GlcNAcpl-3Galfll-4Glcp1-1-Cer R2.Galp1-4GlcNAcpl-3Galp1-4GlcNAcp1-3Galp1-4Glcl -Cer R2-Galpl-4GlcNAcpl
Common Name Lacto-neo-tetraosyl Lactotetraosyl Lacto-nor-hexaosyl
\
Galpl-GlcNAcpl-3Galpl-4Glcpl -Cer
Lacto-iso-octaosyl
3
I R4-Galpl-4GalNAcp1
Globo series
Ganglio series
R1-GalNAcpl-3Galal-4Galpl-1-Cer GalNAcPl-3Galal-3Galpl-4Glc[jl-l -Cer GalNAcal-3GalNAcD 1-3Galal-4GalD 1-1-Cer RZ
I
R3-Galp1-3GalNAcpl-4Galp1-4Glcpl-l-Cer R4
I
R4
I
Gala series lsoglobo series
Muco series
Globotetraosyl Globo-neo-tetraosyl Globopentaosyl
GalNAcpl-4Glalpl-4Glcpl-1 -Cer R4
I
GalNAcpl-4Galp1-BGalNAcpl-4Galpl-4Glcpl-1-Cer R2-Gala-4Galpl-l -Cer Gala1-3Galpl-4Glcpl-1 -Cer GalNAcpl-3Galal-3Galp1-4Glcpl-1-Cer R5
I
R2-Galpl-3Galpl-4Glcpl-1-Cer R2-Galel-3Galfll-3GaIp1-4Glcpl-l-Cer
Gangliotetraosyl Gangliotriaosyl Gangliopentaosyl Galabiosyl lsoglobotriaosyl lsoglobotetraosyl Mucotriaosyl Mucotetraosyl
Substitution sites in core structures: RI = 2-3Sialosy1, 2-6sialosy1, Galpl-3, Galal-3. fucosylal-2; Rz = Fucosylpl-2 or 2-3 sialosyl; R3 = 2-3Sialosyl. 2-8-sialosyl 2-3 sialosyl, fucosyl 1-2, or fucosyl 1-3; Rq = 2-3Sialosy1, 2-8 sialosyl 2-3 sialosyl; Rs = GalNAcal-3Galpl-6.
788 Chloroform/methanol mixtures can be replaced by phosphate buffer/tetrahydrofuran (1:8 v/v) in a procedure described by Tettamanti et al (10). The isolation and purification of neutral glycoljplds and gangliosldes using DEAE-silica gel, DEAE-Sephadex or DEAEcellulose is used frequently (11). Separation of total glycosphingolipid from lipid extract may be carried out as acetylated derivatives by column chromatography on Flohsil (12). Purification of individual glycolipids is usually performed by column chromatography using different silica-based matrices. Elution of the column with chloroform/methanol mixtures (gradient elution) gives better separations. However, fractionation using these procedures Is often poor and repeated rechromatography is required to isolate pure glycolipids.The recent use of porous silica, such as latrobeads, has eliminated many of these problems. Preparative thin-layer chromatography is useful for small amounts of samples. High-performance TLC on superfine (5|im) silica gel has been used to separate complex mixtures of gangliosldes and neutral glycolipids (13). Significant purification of certain components by HPLC has been reported by McCluer et al (14). Most of the more successful separations have involved derivatized glycolipids, allowing detection in the UV region. Characterization of glycosphingolipids The homogeneity of glycolipids separated by the various procedures already described should be determined by thin-layer chromatography or HPTLC in several solvent systems. Mixtures of chloroform/methanol/water in different ratios, such as 60:35:8 v/v or 55:45:10 v/v as well as those with added calcium (chloroform/methanol/0.02% CaCl2.2H20, 60:40:9 v/v) or ammonia (chloroform/methanol/15M
ammonia/water
60:35:1:7 v/v ) are particularly useful for the separation of gangliosldes.Neutral and acetylated glycolipids are separated by chloroform/methanol/water (60:25:4 v/v) and 1,2dichloromethane (DCE)/methanol/water (88:12:0.1 v/v) respectively. Specific detection reagents are used for the preliminary identification of glycolipids.The most useful reagents are the resorcinol-HCI spray for gangliosldes and the orcinolsulfuhc acid spray for neutral glycolipids and gangliosldes. The migration behavior of glycolipids on TLC in various solvents gives some indication of the number of sugar residues present in them.
789 Structure determination of glycolipids For the elucidation of the primary structure of an isolated glycosphingolipid, the following aspects have to be determined: (a) the sugar composition, (b) the anomeric configuration, (c) the conformation of the sugar ring, (d) the sites of glycosidic linkage (e) the sugar sequence, and (f) the ceramide structure. a) Sugar composition The preferred methods to determine the carbohydrate composition of glycolipids are based on gas chromatographic analysis of the individual sugar components such as methylglycosides or alditol acetate derivatives (15,16). Methanolysis in anhydrous methanolic HCI, followed by N-acetylation of amino sugars and sialic acid, and derivatlzation of the methyl glycosides formed to volatile trimethylsilyl ethers (TMS) or trifluoracetyl derivatives are widely used (17,18). b) Anomeric configuration and conformation of the sugar unit The anomeric configuration of a glycolipid has been determined by the use of specific exoglycosidases (19) and more recently by high-resolution proton nuclear magnetic resonance ( ' ' H - N M R ) spectroscopy (20-25). The anomeric configuration of a given sugar unit follows from its J^ 2 coupling constant, a value of ca 3-4 Hz indicating an a Isomer and that of ca 7-9 Hz a p isomer. The possibility of misldentiflcation between pglucose and p-galactose, P-N-acetylgalactosamine and N-acetylglucosamine and a-Nacetylgalactosamine,
a-galactose
and
a-fucose
units,
respectively,
can
be
unamblgously excluded by establishing the connectivities with the corresponding 2-H resonances, which have different characteristics. The "^Cl conformation of the sugar ring was rigorously established for galceramide and glucosylceramide and was assumed to apply to all D-sugar residues.
790
c) Site of glycosidlc linkage The sites of glycosidic linkage may be established directly by NOE. Although this procedure has been used for the estimation of the angles between sugar rings in synthetic tri- and tetrasaccharides of established structures (26), it can also be applied to the determination of unknown sequences and sites of glycosidic linkages {27)The determination of this site of glycosidic linkage by NOE is unequivocal,except where there is an equatorial proton at a vicinal carbon atom, such as H-4 in 3-glycosylated galactose residues. d) Sugar sequence The carbohydrate sequence can be studied by the partial hydrolysis of glycolipids and by the use of specific glycosidase, the most specific and useful method for sequential degradation of carbohydrates. The sugar sequence in glycolipids may be elucidated by mass spectrometry, using acetylated or permethylated glycolipids (28,29). Nev^r ionization methods such as FAB-MS and SIMS have been developed in which derlvatization is unnecessary. With the aid of FAB-MS, It has been possible for the first time to obtain data on such structural parameters as molecular mass, homogeneity, sequences, composition branching and linkage. This method has been successfully applied to glycolipid structure analysis (30,31). e) Glycosidic substitution Methylation is the most widely used method to determine the position of the glycosyl linkages.Permethylation of glycolipids can be done according to the methods of Hakomori (32) and Ciucanu and Kerek (33). The permethylated dehvatives are hydrolyzed and the products converted into partially methylated alditol acetates and identified by gas chromatography-mass spectrometry ( CG-MS) (34). f) Ceramide structure Fatty acids and the LCB components of ceramide can be analyzed by GC-MS after the glycolipids have been methanolyzed.
791 A combination of the methods of Gaver and Sweeley (35) and Carter and Gaver (17) has been widely used.Sphingosine is analyzed as TMS derivatives.The fatty acid methyl esters released after aqueous methanolic HCI treatment were analyzed by GC and the presence of a-hydroxy fatty acids was determined after trimethylsilylation.Long chain bases in sphingolipids were also analyzed by positive ion fast atom bombardment (36). Recent improvements in nuclear magnetic resonance spectroscopy together with mass spectrometry have brought about tremendous progress in the characterization of glycosphingolipid structures. ^ H-NMR analysis allows the elucidation of GSL structures, without the use of destructive methods and requires small amounts (nmole) of material. In addition to one dimensional '^ H-NMR,
other
methods
such
as
two-dimensional
^ H-NMR
shift
correlations
spectroscopy (COSY), two-dimensional nuclear Overhauser '^ H-NMR spectroscopy (NOESY)
and
homonuclear
two-dimensional
spin-echo
J-resolved
'^ H-NMR
spectroscopy. The Introduction of "^^C-NMR into the field of glycosphingolipid research should give useful information on the stereochemical conformation of molecules. This is of coniderable interest, as they most probably contribute to the immunological specificity of glycosphingolipids (37). Glycosphingolipids of protozoa In trypanosomatids, glycosphingolipids have been studied in several species of Trypanosoma, Leishmania and lower trypanosomatides.such as species of Chthidia. Since glycollplds may be important In host-parasite interactions (38), a detailed knowledge of the primary structure of these molecules isolated from trypanosomatids could provide insight into the function, biosynthesis and antigenicity of parasite carbohydrates. Glycosphingolipids ofCrithidia spp Glycosphingolipids with similar structures are found in Chthidia spp, which are considered to be lower trypanosomatids as their habitat is in invertebrate and since they do not have a cycle involving vertebrates (39). Neutral glycosphingolipids were isolated and characterized in Chthidia luciliae, C.oncopeiti and C.guilhermei (40). The major glycolipids of the three species of Chthidia detected by HPTLC chromatography contain
792 a
monohexosylceramide
unit.The
FAB-MS
spectrum
of
the
CMH
(ceramide
monohexoside) from C. luciliae and C. guilhermei revealed, in addition to ions at m/z 264 for the sphingosine and m/z 331 for the terminal hexose residue, a series of fragments derived from the molecular ion [M+H-60]+ between 936 and 1020. The ceramide moieties of the Crithidia glycosphingolipids have sphingosine as long-chain base and C20-C24
fatty
chromatography
acyl
groups
(saturated
of trifluoracetyl
or
derivatives
2-hydroxy of the
fatty
acids).
Gas-liquid
sugars from C.guilhermei
monohexosylceramide provided evidences for the heterogeneity of the sugar moiety with the presence of both galactosyl and glucosyl ceramides in a 5:1 ratio (Figure2). By a combination of HPTLC,GC-MS, FAB-MS, two molecular species were identified in C. guilhermei (Figure 3).
Oal
A
1 IJL
Retention time (min)
Figure 2 - Gas chromatogram of trifluoroacetyl derivatives of the sugars fronn Crithidia guilhermei monohexosylceramide (CMH).
793
Figure 3 - Proposed structures of CMHs from C. guilhermei (I) galactosylceramide (II) glucosylceramide.
Gtycosphingolipids of higher trypanosomatids Trypanosoma and Leishmania spp are higher trypanosomatids, residing during their life cycle in verterbrate as well as in invertebrate hosts. They have been extensively investigated, as they are responsible for a number of human diseases, the most important being Chaga's disease, sleeping sickness and kala-azar. These parasites have a highly glycosylated surface membrane. In the last few years, several reports describing the presence of various surface glycoconjugates apparently associated with the binding of parasites to mammalian cells, parasite virulence and the immunological responses of the host, have been published. Although
it
is
probable
that
glycolipids
represent
important
components
of
trypanosomatids, they have not so far been associated with cell surface reactivity related to phenomena such as host cell penetration, antibody and lectin binding, cell adherence and colony formation in vivo. Structural studies of these purified molecules constitute the first step in the understanding of their function.
794 GlycosphingoHpids of Leishmania spp Leishmania (L) amazonensis, the causative agent of human cutaneous leishmaniosis (CL), is a digenetic parasite with a life cycle alternating between extracellular motile promastigotes, in the gut of the sandfly vector (41,42) and intracellular nonmotile amastigotes, in the phagolysosomal compartment of mammalian macrophages (43). Neutral glycosphingolipids (GSLs) were isolated from amastigote forms of Leishmania amazonensis
with
isopropyl
alcohol/hexane/water
(55:20:25
v/v)
and
chloroform/methanol (2:1) and purified by Florisil, DEAE-SEphadex chromatography and HPLC on a latrobeads column (44). The GSL was characterized by FAB-MS (negative mode) of the permethylated compound, degradation with exoglycosidases and '^ H-NMR spectroscopy. The FAB-MS ( Figure 4) showed a predominant pseudomolecular ion [MH]at m/z 1,184, consistent with a GSL of a tetrasaccharide (Hex4) plus ceramide consisting of d 18:1 sphingosine linked to a 16:0 fatty acid.The lower migration region showed the presence of the corresponding ions for HexsCer (m/z 1,022), Hex2Cer (m/z 860), HexCer(m/z 536). GC-MS of the partially methylated aldltol acetates derived -
n
the GSL, resulted in the identification of the following derivatives: 2,3,4,6-tetra-O
"
Gal, 2,4,6-tri-O-Me-Gal, 2,3,6-tri-O-Me-Gal and 2,3,6-tri-O-Me-Glc in a molar proportion of 0.64:1.04:1.08:1.00, corresponding to a nonreducing Gal end unit, 3-0 and 4-0 substituted galactosyl residues and a 4-0-substituted glycosyl unit.
R e A I b a u t n i d V a e n c e
n^MT
lC«rr 536
/
uma
(CMH-HT 696 rCDH-MT ®fO ICTH-HT 1022
•^ > !•
m/k Figure 4- FAB-MS of native glycosphingolipid from Leishmania amazonensis
1164
795 The anomerjc configuration and linkage sequence of GSL were analyzed by ^H-NMR spectroscopy in dimethyl sulfoxide-d6/D20. The spectrum (Figure 5) of GSL displayed four anomehc resonances, one with a configurations (^Ji.2=3.4 Hz) and three with p configurations {^Ji.2= 7-8 Hz). The spectrum also showed the presence frans-vinyl signals of sphingosine at 5.36 and 5.54 ppm (R4 and R5 respectively) and a very weak triplet at 5.32 ppm representing the c/s-vinyl protons from unsaturated fatty acids. To confirm the glycan sequence, enzymatic degradations, using a-and p-galactosidases from bovine testis, were performed and monitored by HPTLC and ' ' H - N M R spectroscopy.The combination of data confirmed the complete structure of GSL as a novel globosehes structure: Gal P ( 1 ^ 3)Gal a(1 -^)Gal p(1 ->4) Glc 1 ^Cer, which has P(1 -^ 3)Gal substituting for the p (1-> 3) GalNac of globoside. These data reveal the existence of a high concentration of neutral GSLs in Lamazonensis amastigotes, in contrast v^th their virtual absence in promastigotes. It indicates clearly that a remarkable change in cell membrane glycoconjugate composition is associated with cell phase differentiation in Lamazonensis. Modulation of GSL biosynthesis and expression in amastigotes may be an important step correlated with their survival and proliferation in host macrophages.
QaV»1-»3Gat.cV4Gal|»V4Qlc/»V1Cer IV
R-5
R-4 ds
rnn
fiY) /
lU
il
I
R
I
M
-It-^llV5A0
5.20
'*'• ^-^^J^ 5J00 4.80
4J60
4.40
4^0
4X)0
PPM Figure 5 - ^H-NMR of a novel globoside from L amazonensis. The assignment of each resonance Is indicated by arable numbers for the positions of protons and roman numerals for sugar residues. R=frans-vinylprotons of sphingosine backbone. cis= c/s-vinyl protons of unsaturated fatty acid chains.
796 GiycosphingoUpids of Trypanosoma cruzi Trypanosoma cruzi, the causative agent of Chaga's disease (American trypanosomiasis) exists in three morphologically different forms related to the three different environments in which it lives. These three forms comprise amastigote, a dividing form found intracellularly in mammalian hosts, epimastigote, a multiplying form found in the vector's digestive tract and in culture, and the trypomastlgote a non-multiplying infective form, that occurs In the lumen of the rectum of the reduviid bug and is infective to the mammal. T cruzi multiplies discontinuously In the vertebrate host; the amastigote intracellular stages multiply by binary fission and they transform into nondividing trypomastigotes which emerge from tissues into the bloodstream, where they circulate for a certain period before penetrating cells and resuming their complex life cycle (45). Infection with T cruzi, results in the formation of chronic lesions In many tissues, Including muscle and the nervous system (46). The chronic pathological phase of Chaga's disease Is thought to be of autoimmune origin, due to the presence of crossreactive antigens betvy/een the parasite and the mammalian host or due to the failure of self-non self discrimination in the infected host (47,48,49). GiycosphingoUpids have been isolated from epi and trypomastlgote forms of Trypanosoma cruzi. The major neutral glycosphingolipids from 7. cruzi ceramide monoand dihexosides (CMH and CDH, respectively), have been purified and their structures elucidated using a combination of techniques I.e. column chromatography on latrobeads RS 2060, HPTLC and GC together with FAB-MS spectrometry and 500 MHz ' ' H N M R spectroscopy (50).The ceramide monohexoside fraction (CMH) which chromatographed as a double band, arising from the fatty acid heterogeneity, contains either glucose or galactose,
sphingosine
and
as
fatty
acyl
groups
mainly
C-24
saturated,
monounsaturated or 2-hydroxy fatty acids as fatty acyl groups (Table 2). The FAB-MS spectrum of the peracetylated CMH and CDH fraction from 7. cruzi contained
major
molecular ions [M+Na]+ at m/z 1220 and 1248 (Figure 6). Ions at m/z 331 and 289 together with m/z 619 are derived from the hexose-hexose resldue.The pattern of fragmentation of CDH is depicted in Scheme 1. The distribution of fatty acid chains was calculated from the relative intensities of the most intense [M+H-60]+ ion of the CDH fraction. The peracetylated CDH fraction was further analyzed by 1- and 2-dimenslonal NMR spectroscopy. The results shown in Table 3 are in complete agreement with the proposed structure of a lactosylceramide.
Table 2 - Relative distribution of fatty acid chain lengths of the ceramide monohexoside fractions CMH-CoHand CMH-Cn from T. cruzi as calculated from the (M + H 60)’ ions of the peracetylated compounds.
-
-
[M + H 601’ 848 850 8 76 878 904 906 908 918 920 932 934 936 946 948 960 962 964 974 976 978 990 992 1004 1006 1018 1020 1032 1034 1046
(Nz)
Chain length of n-fatty acid 16:l 16:O 18:l 18:O 20:l 20:o
CMH-Cn (X)
21 :1 21:o 22:l 22:o
2 5 3 10
23:l 23:O 24:l 24:O
3 8 6 16
251 25:O
7 7
3 15 3 6 2 4
Chain length of a-hydroxy
CMH-Co” (%)
16:l 16:O
8
18:l 18:O
5 5
20:l 20:o
3 3
21:l 21 :o 22: 1 22:o 23: 1 23:O 24: 1
4 5 1 8 tr. 9 8 20 5 8
24:O
251 25:O 26: 1
7
Based on the MS data only sphingosine is present in the ceramide moiety CMH-C,, ceramide monohexosides from T. cruzi with n fatty acids; CMH-Con, CMH with a-hydroxy fatty acids from T. cruzi; tf, trace
4 4
798
MASS PC* CHARCC
MASS PCR CHARGE
Figure 6 - Molecular ion region of the FAB-Mass spectra of peracetylated C M H - C Q H (A) and CDH (B) showing the [M + H - 60]* ions
-CMj-CO 289331
^
''^ ,
•H OAc , •V CH
Ac
Acorv p
Scheme 1
H ^ ^ \
HN OC M-^Na'^
-1248
M+H*
«1226
[Hd+H-eor -1166
<x
CH3
590 - - * —264 Acyl: -HOAc
799 Table 3 - Chemical shifts (ppm from internal tetramethylsilane) and coupling constants [Hz] for peracetylated lactosylceramide from T. cruzi in CDCI3 solution at 303 K as obtained by two-dimensional ^H-NMR spectroscopy.
^>i^°;^^yV^^/^ OC
H-1
ppm 4.49
J1.2
H-2
5.11
J2.3
H-3
4.96
^J3.4
H-4
5.35
^J4.5
H-5
3.88
J3.6; J5.6
H-6;6'
4.09;4.14
[Hz] [8.0] [10.5] [3.0] [1.0] [6.0;6.9] [11.4]
ppm 4.44
III
[Hz] [7.5]
4.87
[10.0]
5.19
[10.0]
3.80 3.59
CH2—R"
[10.0] [2.1;4.5]
4.08;4.51
[12.0]
ppm 3.90;3.52
H-1;1' •^1.1; Jl.2; J1.2
[Hz] [10.0;3.6;4.2]
4.30
H-2
m
J2.3
H-3
5.25
^J3.4
5.35
H-4 J4.5; J4.6
H.5
5.77
H-6;6»
2.02
H-N
5.62
JNH.2
H-aliph H3C(R';R"I
1.28-1.35 0.88
[7.5]t [15.0;0.9] [6.0]t
[9.0]
m, multiple!; t, triplet
The role of the monohexosylceramide isolated from epimastigote forms of T. cruzi, on the interaction of T. cmz/with heart muscle cells was studied by Vermelho et a! (51). The results show a large
increase in the number of infected cells when the highly
infective Dm 28c clone of T. cruzi and heart muscle cells were preincubated with the glycolipid before the interaction (Figure 7). This finding may be due to the uptake of the lipid by both parasites and host cells, as shown previously in fibroblasts (52) and tumor cell lines (53). The simultaneous addition of glycolipid and metacyclic trypomastigotes results in a decrease in the penetration of the parasite.
Competition
between
the
800
(%)
-I—Ti
Time (18 h)
Figure 7 - Effect of glycosphingolipid on the infectivlty of the Dm 28c clone of T. cruzi with heart muscle cells. Control (open column), heart muscle cells infected with T. cruzi, Dm 28c clone. Experiment 1 (suppled column), metacyclic fomfis, preincubated with glycoHpid for 30 min, following addition to the cell culture. Experiment 2 (solid column), heart muscle cells incubated with glycolipid for 30 mIn, before metacyclic addition. Experiment 3 (cross-hatched column), heart muscle cells incubated simultaneously with metacyclic fonms and glycolipid.
glycolipid and the protozoan for the receptor connobining site of the heart muscle cells or saturation of the receptors may be responsible. As glycosphingolipids were shown to be immunogenic (54) and parasite glycolipids as well as other glycoconjugates stimulate the host immune response (55), the reactivities of Chagasic patients sera and sera from rabbits hyperimmunized with epimastigote cells, were assessed using highly purified glycosphingolipid fractions from 7. cruzi epimastigote (56). A strong reactivity with GSL was obtained with T. cmzHmmunlzed rabbit. Reactivity with GSL was also obtained with human Chagasic sera.Compared to a group of normal individuals, the reactions of antibodies directed against lipid antigens were considerably increased in sera of patients with Chaga's disease. Recent studies by Avila and Rojas (55) showed the presence of elevated cerebroside antibody levels in chronic Chaga's disease. Chagasic sera did not differentiate between glycolipids with terminal p glucosyl or p galactosyl non-reducing units. They however discriminated between glucosylceramides with differences in the ceramide structure. These
results
suggest
that
antibody
801 recognition of cerebrosides involves either a complex epitope formed by the terminal sugar and elements from the ceramide moiety or, alternatively, the presence of the ceramide lipid allows an increased affinity of the antibody for a single carbohydrate unit. Although the glycolipids are recognized by Chagasic patients sera, the reactive antibodies are not specific since they
also reacted quite well with murine laminin
(Figure 8). Most of them seem polyreactive, strongly binding to murine laminin. The homologous reactivity with the parasite glycolipid v^s much less Intensive involving mainly IgM. These results and others with different antigens (57) give support to polyclonal B cell activation in Chaga's disease.
UMMMrfPURFEDAbs
la-PURFED Ate
Figure 8 - Reactivity of Chagasic semm antibodies purified on crude glycolipid on laminin Immunosorbents with both antigens. Antigen on ELISA plate: A and C, murine laminin at 100 ng/well; B and D, crude glycolipid at 10 |ig/we!l. IgG • ; IgM i .
Cross-reactive lipid antigens were isolated from epimastigote forms of 7. cruzi and from the mammalian brain with the monoclonal antibody VESP 6.2 (58) which had been raised against T. cruzi-re\aied trypanosomes T. dionisii and T. verpertillionis.Chemical reactions indicated that the sulphate group of the lipids is an important part of the epitope recognized by the Mab. The specificity of VESP 6.2 for these isolated lipid antigens was demonstrated by three different methods: a) high-performance thin layer chromatografy, b) solid phase radioimmunoassay and lysis of artificial liposomes (59).
802 A partially characterized sulfoglycosphingolipid
is also present
in 7. cruzi
trypomastigotes (60). Besides neutral and sulfoglycosphingolipids, sialoglycolipids have been characterized In trypomastlgote forms of 7. cruzi (61). Glycosphingolipids of Trypanosoma dionisii A monohexosylceramide from a 7 cruzi related trypanosome, 7 dionisii, has been Isolated and analyzed by HPTLC, gas liquid chromatography and FAB-MS (62). As sho\A^ in Figure 9, the GSL developed as a doublet band by TLC with the same migration as galactosylceramide from calf brain.
Figure 9 - High-performance thin-layer chromatography (HPTLC) of neutral glycosphlngollpld from 7 dionisii. Lane 1: Neutral glycosphingolipids from human erytrocytes and bovine brain. Lane 2: Monohexosylceramide from 7 dionisii. Runnlno solvent: CHCIa/MeOH/water, 60:25:4 v/v. Detection: orcinol-H2S04 reagent.
This GSL was methanolyzed and the methyl glycosides converted to their trifluoroacetyl derivatives for GLC analysis. The GLC chromatogram shows the carbohydrate composition of CMH, with galactose and glucose in the molar ratio of 1:1 (FIgurelO). The FAB-mass spectrum in positive mode provided molecular weight information and reflected the sample heterogeneity. The molecular ion region of monohexosylceramide of 7 dionisii is shown in Figure 11. Intense signals between 906 and 1020 derived from the molecular ion [M+ H-60]+ are shown.
803
.Oal
QIC
Retention time (min) Figure 10 - Gas chromatogram of trifluoroacetyl derivatives of the sugars from T. dionisii monohexosilceramides (CMH).
MH*-HOAc
Figure 11 - Molecular Ion region of the FAB* mass spectrum of peracetylated CMH from 7. dionisii, showing the [MH*-HOAc] Ions. The relative amounts of fatty acids calculated from the molecular Ion peak Intensities of CMH are shown in Table 4. The presence of sphingosine in the ceramide moiety was confirmed by a fragment at m/z 264.
804 Table 4 - Relative amounts of fatty acid chain lengths in the CMH from T. dionisii calculated to the intensities of the [MH - + - HOAc] ions in the FAB -i- mass spectrum of the peracetylated species. Fatty add 20:0 22:0 24:0 14:0-OH 16:0-OH 18:0-OH 20:0-OH
MH - HOAc m/z 906 934 962 936 964 992 1020
I. rel [%] 117 13.5 22.6 9.1 15.0 12.6 15.5
A similar monohexosylceramide has been previously described for 7. cruzi. Using a monoclonal antibody directed against 7. dionisii and 7 vespertilionis, Retry et al (58) demonstrated cross-reactivity between these parasites and 7 cruzi (59). Previous results on the mechanism of entry of 7 dionisii into non-phagocytic cells (63,64) and its subsequent intracelular development provided evidence for the close similarity between 7 dionisii and 7 cruzi. Consequently, 7 dionisii is useful as a nonpathogenic model for the study of Chaga's disease. Glycosphingolipid of Trypanosoma meaa 7 mega is a trypanosomatid of cold-blooded vertebrates isolated from the African toad, Bufo regularis. Like the trypanosomatids in general, 7 mega has two principal stages in its biological cycle. In its vertebrate host, the trypanosome multiplies in the trypomastigote form while the epimastigote (crithidia) form is present in the invertebrate host (65,66). A glycosphingolipid fraction from 7
mega (67) was isolated after
acetylation and further purified on a silicic acid column, as shown in Scheme 2. The carbohydrate components of the glycolipid were fucose and galactose in approximately equimolecular amounts. Fatty acids, forming amide group with the sphingosine bases, were a mixture of normal and a-hydroxy fatty acids. Normal C16:0,C18:0 and 2-hydroxy C18:0 were the predominant fatty acids (Figure 12, Table 5).
805 T. mega cells
I
Extraction with 200 vol. of chlorofonm-methanol (2.1 and 1.2 v/v] filtration lipid extract
residue
Acetylation Florisil chromatography 1) Hexane-dichloroethane (1:4) discarded 2)Dichloroethane-discarded 3)Dichloroethane-acetone (1:1) - glycolipids 4)Dichloroethane-methanol-water (2:8:1) - discarded Chromatography on silicic acid 1)Chloroform-methanol (1:1) - discarded 2)Chloroform-hexane (3:2) - discarded 3)Chlorofonfn-discarded 4)Chloroform-methanol (95:5) - glycolipid
Scheme 2 - Isolation and purification of the neutral glycolipid from T. mega.
loo
11
2
Uu
10 h2
XJJ4J
yiLudj
ul
1516
Figure 12 - Reconstructive ion chromatogram from GLC-MS experiment of fatty acid from T. mega glycolipid. Peak identification listed in Table 5.
806 Table 5 - Composition of fatty acids of the giycolipid of Trypanosoma mega.'
Peak mumber 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Fatty acid C14:0 CI 5:0 C16:0 CI7:0 CI8:0 CI8:0 C20:0 C22:0 C23:0 C22:0 C24:0 C23:0 C25:0 C24:0 C26:0 C25:0
RRT'^ Normal 0.81 0.90 1.00 1.07 1.16 1.25 1.30 1.44 1.51 1.53 1.57 1.59 1.63 1.65 1.69 1.71
a-hydroxy
5.0 1.5 49.0 1.0 33.0
-
8.0
-
1.0
-
1.0
-
3.0
-
0.5
1.0 2.5 1.5 4.0 0.5
0.5
Principal fragments (m/e) 74, 87,101, 129,143, 185, 199. 211, 213,242 74,87,101.129,143,157,185,225,227,256 74,87,101,129,143.157,185,239,241,270 74,87.101,129,143,185,253.255,284 74,87,101.129,143.185.267,269,298 73,89,103,129,159.327,371,386 74,87,101.129,143.185,295,297.326 74,87.101.129,143.185,323.325,354 74,87.101.129,143,185,337,339,368 73,89.103.129,159,389,399,427 74,87,101,129,143,185,351,353,382 73,89,103,129.159.397.441,456 74,87,101.129,143,185,365,367,396 73,89,103,129.159,411,455,470 74.87,101,129,143,185,379,381.410 73.89,103.129.159,425,469,484
^ Nomial fatty acids were analyzed as their methyl esters on an SE-54 column. Hydroxylated fatty acids were analyzed as their 0-TMS derivatives. Values are derived from GC-MS analysis. ^ Retention times relative to that of n CI6:0.
Glycosphingoliplds of fungi The nitrogen-containing lipid was described by Zellner (1911) and called "fungus cerebrin". The structure of the cerebrin was determined by Oda (68) as 2-amino-1,3,4tri-hydroxyoctadecane. Because of the close similarity with the animal sphingosine, it was called "phytosphingosine". Although glycosphingoliplds of animal tissue have been extensively studied (69,70), the structure and function of glycolipids of fungi are less well known. Glycosphingolipids are present in fungi of the most primitive class, i.e., Phycomycetes, as well as in the most complex class,namely, Basidiomycetes. Glycosphingolipids from Zygomycetes Cerebrosides have been isolated from Phycomyces blakeslearus (3), a fungus often found on animals dung, by extraction of mycelia with acetone and chloroform/methanol mixtures and purified on a silicic acid column, followed by a Florisll column. The bases obtained after hydrolysis were all phytosphingosine homologs ranging in length from C17 to C22.Palmitic,steahc,oleic, llnoleic and hydroxy palmitic acids were the major fatty
807 acids present in this glycolipid. The cerebrosides contained equal amounts of galactose and glucose. Gfycosphingolipids from Deuteromycetes Fungi apparently lacking a sexual phase (perfect stage) are commonly called imperfect fungi or Fungi Imperfecti (71). Monohexosylceramides have been isolated from Sporothrix schenckn,Fusicoccum amygdali, Fusarium lini, Fusarium solani, Aspergillus oryzae, A. fumigatus and A. versicolor (72,73,3,74,75,76). A glucocerebroside from the yeast form of the dimorphic human pathogen S. schenckii has been isolated and its components characterized by thin layer and gas liquid chromatography
and
mass
spectrometry
(72).
It
was
found
to
contain
glucose.sphingosine and a-hydroxy stearic acid (1:1:1). No role has been attributed to this compound In association with infection and cell surface reactivity. Balllo et al (73) isolated and characterized a N-2-hydroxy-3-trans-octadecenoyl-1-0-Dglucosyl-9-methyl-cis-4,8-sphlngadienine from Fusicoccum amygdali Del., a fungus responsible for almond and peach "canker". A similar structure was seen In Aspergillus oryzae, the Japanese yellow mold (75). Trans-unsaturated hydroxy fatty acids ( 2-hydroxyoctadec-3-enoic acid) found in F. amygdali and A, oryzae cerebrosides have not been detected in sphlngollpids of animals and plants. The structural characterization of the glycosphingoliplds (Figure 13) isolated from two pathogenic species of Aspergillus, the etiological agents of a number of different lung diseases including asperglHoma ( fungus ball) and Invasive aspergillosis, were carried out using high-resolution "^ D,2D-''H-NMR and ''^C-NMR spectroscopy and fast atom bombardment mass spectrometry (FAB-MS) (78). Thin layer chromatography of native and peracetylated glycosphingoliplds Isolated and purified from Aspergillus on silica gel and a latrobead column afforded fractions with a mobility corresponding to a monohexosylceramide, Figure 14 (A,B). The FAB-mass spectra of the peracetylated glycolipid from Aspergillus in the presence and in the absence of sodium acetate are shown in FigurelS. In the molecular ion region, intense signals were seen at m/z 946 [M+H-HOAc] and m/z 886 [M+H-2H0Ac]. Ions indicating a terminal hexose were present at m/z 331 [HexAc4] and m/z 229 and 169 (secondary Ions). The ceramlde
moiety
808
4 i . »
f
B M
m
n
Figure 13 - High-performance thin-layer chromatography of Aspergillus neutral glycosphlngolipids isolated by latrot)eads column cromatography. Lane 1: Folch lower phase. Lane 2: Glycosphingolipid fraction otJtained by silica gel chromatography. Lane 3: Chlorofomi fractions. Lane 4 Chloroform/methanol (95:5) fraction. Lanes 5-11: Chloroform/methanol (9:1) fractions. Lane 12-13 Chloroform/methanol (8:2) fractions. Lane 14: Methanol fraction. Solvent system chlorofomn/methanol/water (65:25:4 vA/). Detection: orcinol-H2S04 spray-reagent. Purified neutral glycosphingolipid fractions (Lanes 7-11) were combined.
B CMH CDH
CMH
i
CTH QbO
Li
a
Figure 14 - High-perfomiance thin-layer chromatography (HPTLC) of native (A) and peracetylated (B) neutral glycosphlngolipids from Aspergillus. (A) Lane 1: Neutral glycosphlngolipids from human erythrocytes and bovine brain. Lane 2: Glycosphingolipid from Aspergillus. Running solvent: chloroform/methanol/water (65:25:4 vA^). (B) Lane 1: Galactosylceramide from bovine brain. Lane 2: Glycosphingolipid from Aspergillus. Running solvent: 1,2 dichloroetha/acetone (8:2 v/v). Detection: orcinol-H2S04 reagent.
809 was represented by peaks at m/z 658 [CerAc2], m/z 598 [CerAc2-H0Ac] and m/z538 [CerAc2-2HOAc].
NMt H I CMMW
Figure 15 - FAB*-MS of the peracetylated glycosphjngolipid from A. fumigatus 2140. (A) (M+H*). (B) (M.Na').
The structure of the long chain base was deduced from the 1D and 2D-NMR spectra. The HH-COSY spectrum (Figure16) revealed the complete connectivity of the long chain base starting from proton H-1 until H-11, from proton H-2' until H-11, and from proton H2' until H-5' of the fatty acid. The chemical shift of 5.30 ppm (long chain base H-3) and 5.50 ppm (fatty acid H-2') Is typical for protons wnth acetylated OH-groups adjacent to the olefinic bonds. For the protons H-4 and H-5 of the long chain base, a coupling constant of 15.3 Hz and for protons H-3' and H-4' of the fatty acid 14.3 Hz, respectively, were observed. These are typical for frans-double bonds. The CH-COSY also confirmed the double bonds (Figure 17). The E-configuration of the double bond in position 8 of the long chain base was evident from the NOE experiment by irradiating
proton H-10.
The methyl group at position 9 of the base was determined by COLOC, as shown in Figure18. The sugar component of CMH from both A. fumigatus and A. vers/co/or was glucose. The (3 configuration of the sugar was evident from the coupling constant of H-1" (7.9 Hz) (Table 4).
no
Vj'w
* * .^^ 1
^x'\
• 4^5
r2
<
»//i * • •
NH
• t
4**
6"
-"a
• •
1
#
•
*
9a
6 ppm
I I I I I > I I I I I I I I I I I I I I t I I » TI t I I I I I I I I 1 t I > I I I I
ppm Figure 16 - 500 MHz HH-COSY of the peracetylated glucosylceramlde from A. fumigatus 2140 in CDCI3.
811
mmmmmmmm
ML
m
^ ^0|
^!
6^5'
16-
t^"
•2
-\'<'
4»iI 3*
4'|l' '"I"
'
I
120
I
'
ao
100
I
I
ao
40
I'""
20
Figure 17 -125/500 MHz CH-COSY of the peracetylated glucosylceramide from A. fumigatus in CDCb
" I I'iiilil 130
too
K
«
40
Figure 18 -125/500 MHz COLOC-section of the peracetylated glucosylceramide in CDCI3.
812 Table 6 - ^H-and ^^C-NMR chemical shifs [ppm] and coupling constants [Hz] for the peracetylated and native glucosylceramide from A. fumigatus NCPF 2140 and 2109 obtained by one and twodimensional NMR-spectroscopy.
c-
Galactosyl -ceramide
Glucosyl-ceramide peracetylated
Glc-Cer-native
No. peracetylated
"r
2" 3" 4" 5" 6a" 6b"
1a 1b 2 NH 3 4 5 6 7 8 9 9a 10 11
^JHH^^^
4.44 5.15 4.99 5.37 3.90 4.13 4.13
7.92" 10.53 3.444
NR NR NR NR
TTi 4.47 4.95 5.18 5.08 3.68 4.13 4.23
3.61 3.92 4.29 6.33 5.30 5.39 5.81 2.05 2.05 5.07
•"5:^^ 100.60 71.21 72.70 68.23 71.95 2.46a"/4.66b" 61.85 12.36b
"PH
67.15
3.68 4.05 3.93 7.46 4.07 5.40 5.68 1.99 1.99 5.05
3|
[Hz]
8.02 9.63 9.64 9.95
-
4.02710.4,b 4.72
9.I3 6.82 7.24 15.35
-
1.56 1.94 1.35
r
2' 3' 4' 5' 6'
5.50 5.50 5.87 2.05 1.35
Aliph
1.2-1.4 0.87
CH3
14.34
-
-
50.83
-
73.25 124.51 136.90 32.64* 27.47 123.00 136.31 16.06 39.79 28.1 168.48 74.46 123.00 138.12 32.41*
-
28.0-30.5 14.21
4.22 3.20 3.36 3.30 3.24 3.67 3.82
-
1.53 1.90 1.30
_ 4.42 5.44 5.79 1.99 1.31 1.2-1.4 0.82
The ^H-NMR chemical shifts of the ceramide part from the peracetylated glucosyl-and galactosylceramlde differ by < 0.02 ppm. NR, not resolved. * Assignments exchangeable.
The results show that the major GSL from species of Aspergillus has a distinctive ceramide structure, consisting of 9-methyM,8-sphlngadienine linked to a 2hydroxyoctadec-3-enoic acid (Figure 19). While the structure described here is similar to those from A. oryzae (19), F. amygdali (73) and some Basidiomycetes (77), it is the first report of this occurrence in species pathogenic to humans.
813
Figure 19 - Structure of sphingadienine from Aspergillus.
N-2'-hydroxyoctadec-3-enoyl-1-0-p-D-glucopyranosyl-9-methyl-4,8-
Glycosphingolipids of Basidiomycetes The Basidiomycetes is an important group of fungi which includes harmful species as well as useful ones. Some species cause plant diseases, such as stinking smut and black stem-rust of wheat. Several Basidiomycetes are significant in causing diseases of forest and shade trees and in destroying lumber. However, many Basidiomycetes are extremely valuable in nature, forming a mycorhizal relationship with both cultivated and noncultlvated
plants.
Neutral
glycosphingolipids,
whose
structures
have
been
determined in detail, are present in Schizophyllum commune and two Basidiomycetes species, Clitocybe geotropa and C. nebularis. A cerebroside was isolated from S. commune and purified by HPLC. Its structure was elucidated by mainly chemical methods of degradation followed by nuclear magnetic resonance and mass spectrometry of the different constituents (78). The glycolipid has the chemical structure of (4E,8E)-N-D-2'-hydroxypalmltoyl-1-0- (3-Dglucopyranosyl-9-methyl-4,8-sphingadienine. A similar structure has been reported by Karlsson et al. in the sea anemone Metridium senile (79). This molecule was capable of inducing fruiting body formation when tested on a dikaryotic strain of S.commune. In order to determine the part of the molecule responsible for the activity, the degraded products were tested.
814
^M-CM,-(C»V„-.CM,
CN
1
2
%r
3
^CM
4
B
"'^CM
6
7
^*C
•
9,10
»CM^),
11
Figure 20 - Structure of the glycosphingolipld from C. geotropa and C. nebularis.
Both the reduced periodate-oxidized cerebroslde and the ceramide were as active as the original cerebroside, suggesting that the sugar moiety is not important for activity. The sphingoid moiety of a cerebroside constitute an essential part of the activity and the fatty acid having a certain chain length and the 2-hydroxy fatty acids may be the functional structures of this glycolipid (80). Fruiting bodies of C. geotropa and C. nebularis from the large order Agaricales, contain 2'-hydroxyhexadecanoyl-1-0-p-D-glucopyranosyl-9-methyl-4,8-D-erythro-sphingadienine as the major neutral glycosphingolipld. The glycosphingolipids were extracted from both fungi with chloroform/methanol and the extracts were fractionated by silica gel column chromatography, with ethyl acetate/acetic acid/methanol/water, 40:3:3:2 v/v, as eluent. The structure was deduced by ''^ and '^-NMR spectroscopy, sugar and fatty acid analysis and FAB-MS (Figure 20). All these reports indicated that monohexosylceramides are ubiquitous components of fungi, as they occur in Zygomycetes, Deuteromycetes (Fungi Imperfecti) a^d Basidiomycetes (True fungi). The biological relevance of these molecules in fungi is,however, still unclear, with the exception of the glycolipid from S. commune that induce fruiting body formation. The exclusive preference for a branched long chain base in these glycolipids would indicate that it is important for a special function in the membrane, as suggested by Karlsson et al (79). The 9-methyl groups as well as the double bond in the fatty acid chain may therefore influence the membrane organization and polar region of membrane lipids. However, the physiological effects of glycosphingolipld conformation are not known.
815 REFERENCES 1 2 3 4 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
H.E.Carter, W.D.Celmer, W.E.N.Lands. K.LMueller and H.H. Tomizawa, J.Biol.Chem.,206 (1954) 613-623. H.E.Carter. R.C.Gaver and R.K.Yu, Biochim.Biophys.Res.Commun., 22(1966)316-320. B.Weiss and R.L.Stiller, Llpids,8 (1972) 25-30. R.W.Ledeen, R.K.Yu, Methods En2ymol.,83 (1982) 139-191. S.K.Kundu, S.K.Chakravarty, S.K.Roy and A.K.Roy, J.Chromat..170 (170)65-72. R .Rouser, G.Khtchevsky, A.Yamamoto, G.Simon, G.Galli and A.J.Bauman, Methods Enzymol., 14 (1969) 272-317. S.Hakomori, Methods Enzymol., 50 (1978) 207-211. J.Folch, M.Lees and G.H.SIoane Stanley , J.Biol.Chem., 226 (1957) 497-509. G.Tettamanti, F.Bonali, S.MarchesIni and V.Zamboti, Biochim.Biophys. Acta, 296(1973)160-170. H.Baumann, E.Nudelman, KWatanabe and S.Hakomori, Cancer Res., 39(1979)2637-2643. S.K.Kundu, in: A.J.Allen and E.C. Kisailus (Eds) Glycoconjugates, Plenum Press, New York, London, 1992 pp. 203-262. T.Saito and S.Hakomori, J.Lip.Res., 12 (1971) 257-259 J.N. Kanfer and S.Hakomori, in: D.J.Hanahan (ed), Handbook of Lipid Res.,Plenum Press, New York,London, 1983, pp 1-165. J.E.Evans and R.H.McCluer, Biochim.Biophys.Acta, 270 (1972) 565D.E.Vance and C.C.Sweeley, J.Lip.Res.,8 (1967) 621-630. R.Laine, W.J.Esselman, W.J. and C.C.Sweeley, 18 (1972) 159-167. H.E.Carter and R.C.Gaver, J.Lip.Res., 8 (1967) 391-395. S.Ando and Y.Yamakawa, J.Biochem., 70 (1971) 335-340. A.Kobata, Anal.Biochem., 100 (1979) 1-14. J.Dabrowski, P.Hanfland and H.Egge, Methods Enzymol., 83 (1982) 69-86. K.E.Falk, K.A.Karlsson and B.E.Samuelsson, Arch.Biochim.Biophys., 192(1979)164-176. S.Gasa, T.Mitsuyama and A.Makita,J.Lip.Res., 24 (1983) 174-182. A.Yamada, J.Dabrowski, P. Hanfland and H.Egge, Biochim.Biophys. Acta, 618(1980)473-479. T.A.W.Koerner, U.H.Prestegard, P.C.Demon and R.KYu, Biochemistry 22(1983)2687-2690. T.A.W.Koerner, U.H.Prestegard, P.C.Demon and R.K.Yu, Biochemistry 22(1983)2676-2687. R.U.Lemieux, K.Boch, L.T.J.Delbaere, S.Koto and V.S.Rao, Can.J. Chem., 58 (1965) 631-635. J.Dabrowski, P.Hanfland, H.Egge and U.Dabrowski, Arch.Biochim. Biophys., 210 (1981) 405-411. S.P.Markay and D.A.Wenger, Chem.Phys.Lipids, 12 (1974) 182-200. R.W.Ledeen, S.K.Kundu, H.C.Price and J.W.Fong, Chem.Phys.Lipids, 13(1974)429-446. H.Egge and J.Peter-Katalinic, Mass Spectrom.Rev., 6 (1987) 331-393. J.Peter-Katalinik and H.Egge, Methods Enzymol.,193 (1990) 713-733. S.Hakomori, Biochem.J.,55 (1964) 205-208.
816 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 59 60 61 62 63
I.Ciucanu and F.Kerek, Carb.Res., 131 (1984) 209-217. H.Bjorndall, B.Lindberg and S.Svensson, Car.Res.,5 (1967) 433-440. R.C.Gaver and C.C.Sweeley, J. Amer. Oil Chem. Soc.,42 (1965) 294-298. Y.Ohashi, M.Ywamori, T.Ogawa abd Y.Nagai, Biochemistry,26 (1987) 3990-3995. T.Taki, M.Kuroyanagi, H.Yoshida and S.Hande, J.Biochem., 111(1992) 614-617. E.Handman, M.T.McConville and J.W.Coding. Immunol.Today, 8 (1987) 181-185. R.B.McGhee and W.B.Cosgrove, Microbiol.Rev., 44 (1980) 140-170. R.B.Silva, A.L.M.Vazquez, M.F.F.Soares and E.Barreto Bergter, Proceedings of the XXIII Anual Meeting of the Biochemistry Society, 1994, pp 121. D.L.Sachs, Exp.Parasitol., 69 (1989) 100-103. K.P.Chang, in: D.J.Wyler (ed). Modern Parasite Biology, W.H.Freeman, New York, 1990, pp 79-90. K.P.Chang, Int.Rev.Citol., 14 (1983) 267-304. A.H.Straus, S.B.Levery, M.G.Jasiulionis. M.E.KSalyan, S.J.Steele, LR. Travassos, S.Hakomori and H.K.Takahashi, J.Biol.Chem., 268 (1993) 13723-13730. Z.Brener, Ann.Rev.Microbiol., 27 (1973) 347-383. D.losa, D.C.Massari and F.C.Dorsey, Am.Heart J., 122 (1991) 775-785. T.S.McCormick and E.C.Rowland, Exp.Parasitol., 69 (1989) 393-401. M.D.Sadigursky, B.von Kreuter, P.I.Ling and C.A.Santos-Buch,Circulation 80(1989)1269-1276. T.S.McCormick and E.C.Rowland, Exp.Parasitol., 11 (1993) 393-401. E.Barreto Bergter, A.B.Vermelho, R.Hartmann, G.Pohlentz, R.A.Klein, and H.Egge, Mol.Biochem.Parasitol., 51 (1992) 263-270. A.B.Vermelho, E.Barreto Bergter, M.C.Pereira, S.M.Silva and M.N.L. Meirelles, Biomedical Lett., 47 (1992) 113-123. V.Chigorno, M.Pitto, G.Cardace, D.Acquotti, G.Kirschner, S.Sonnino, R.Ghidoni and G.Tettamanti, Glycoconjugate J., 2 (1985) 279-287. K.Furukawa I.J.Thompson, H.Yamaguchi and K.LIoyd, J.Immunol.,142 848-854. M.A.J.Ferguson and S.W.Homans, Parasitol.lmmunol., 10 (1988) 465476. J.L.Avila and M.Rojas, Am.J.Trop.Med.Hyg., 43 (1990) 52-60. M.H.S.Villas-Boas, M.C.Silva, T.G.OIiveira, L.R.Travassos and E. Barreto Bergter. J.CIin.Lab.Anal., 8 (1994) 260-266. C.Unterkircher, S.Avrameas and T.Ternynck, J.CIin.Lab.Anal., 7 (1993) 60-69. KPetry, P. Voisin and T.Baltz, Acta Tropica, 44 (1987) 381-386. K.Petry, E.Nudelman, H.Eisen and S.Hakomori, Mol.Biochem.Parasitol., 30(1988)113-122. M.L.Uhhg, AS.Couto, R.M.Lederkremer, B.Zingales and W.CoHi,Carb. Res., 231 (1992)329-334. B.Zingales, C.Carniol,R.M.Lederkremer and W.Colli, Mol.Biochem. Parasite!., 26 (1987) 135-144. E.Barreto Bergter, M.H.Branquinha and A.B.Vermelho, Mem.Inst.Oswaldo Cruz, 89(1994)86. AJ.Liston and J.R.Baker, J.Protozool.,24 (1977) 41
817 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
H.S.Barbosa and M.N.L.Melrelles, J.Submicrosc.Cytol.Pathol., 25 (1993) 47-51. M.Steinert, Exp.Cell Res., 15 (1958) 560-569. M.Steinert and G.L.Bone, Nature, 178 (1956) 362 A.B.Vermelho, L.Hogge and E.Barreto Bergter, J.Protozool., 33 (1986) 208-213. T.Oda, J.Pharm.Soc.Japan, 72 (1952) 142-145. S.Hakomoh, Ann.Rev.Biochem., 50 (1981) 733-764. C.L.M. Stutts, C.C.Sweeley and B.A.Macher, Methods Enzymol., 179 167-214. C. J.AIexopoulos and C.W.Mims, in; John Willey & Sons (Eds), Introductory Mycology. Part 4: Division Amatigomycota, Willey, 1994, pp. 291307. D.B.S.Cardoso, J.Angluster, L.R.Travassos and C.S.AIviano, FEMS Lett, 43 (1987) 279-282. A.Ballio, C.G.Casinovi, M.Framondino, G.Marino, G.Nota and B. Santurbano, Biochim.Biophys.Acta, 573 (1979) 51-60. R.S.Duarte and E.Barreto Bergter. Proceedings of the II Jornada de Jniciagao Cientlfica, Rio de Janeiro.Brasil, 1995, pp 6. Y.Fujino and M.Ohnishi, Biochim.Biophys.Acta, 486 (1977) 161-171. M.H.S.Villas Boas, H.Egge, G.Pohlentz, R.Hartmann and E.Barreto Bergter, Chem.Phys. Lipids, 70 (1994) 11-19. M.Fogedal, H.Mickos and T.Norberg, Glycoconjugate J., 3 (1986) 233-237. G.Kawai and Y.lkeda, Biochim.Biophys.Acta, 754 (1983) 243-248. K.A.Karlsson, H.Leffler and B.E.Samuelsson, Biochim.Biophys.Acta, 574(1979)79-93. G.Kawai and Y.lkeda, Biochim.Biophys.Acta, 719 (1982) 612-618.
Acknowledgment - We gratefully acknowledge the assistance of Dr. Veronica M. Heam for critical reading of the manuscript.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 O 1996 Elsevier Science B.V. All rights reserved.
819
Potential Bioactive Conformations of Hormones of the Gastrin Family Luis Moroder and Jiirgen Lutz Max-Planck-Institut fur Biochemie, Am Klopferspitz 18a, D-82152 Martinsried Germany
Abstract: With gastrin and CCK as model peptide hormones lipoderivatization and thus, induced lipid afiinity was used to answer the question of whether peptide hormones, upon collision with target cells, accumulate on the membrane surface where a prefolding and preorientation is induced by the amphipathic character of the membrane bilayer structure. In model experimental systems the two lipo-peptides exhibited a differentiated behaviour in their mode of interaction with lipid bilayers and of display of the peptide headgroups at the water/lipid interphase. The CCK is structured in an amphipatic conformation with tendency to aggregate in domains, whereas gastrin is fully exposed to the bulk water phase in randomly coiled structure despite the anchorage to the lipid bilayer. A correlation of the conformational aspects with the biological properties strongly suggest for both peptides in the receptorbound state a very similar structure in the C-terminal portion of the molecule and a negatively charged flexible N-terminal arm involved in electrostatic interactions with receptor subsites, which in the case of the CCK-A receptor represent the ligand-discriminating factor. Independently of whether the peptide hormones are accumulated on the membrane surface, experiments with the lipo-derivatives of gastrin and CCK allowed to confirm that a membrane-bound pathway in the receptor recognition process with lateral penetration at the water/lipid interphase is indeed possible despite the bulky extracellular boundary domains of the receptors.
1.Introduction Molecular self-assembly is the spontaneous association of molecules into stable, structurally well defined aggregates held together by noncovalent bonds. It is ubiquitous in biological systems and determines information transfer at the supramolecular level in a large variety of complex biological systems of distinct functions. Molecular recognition underlies this process of self-assembly which requires only the information embodied in the shape, the surface properties and the deformation of a limited number of molecular entities. As it represents the
820
basis for signal transduction, understanding this process in terms of forces involved in the noncovalent interactions responsible for the specific association of complementary surfaces of molecules is a general concern. Molecules in solution have degrees of freedom of overall translation and rotation that become highly constrained when two molecules associate to form a stable complex. There is an entropic cost with this bimolecular interaction that is a consequence of degrees of freedom of motion lost. If the binding process involves capturing of flexible molecules such as peptide hormones from the aqueous extracellular phase by cell membrane-bound receptors, even intramolecular rotations about single bonds become strongly restricted in the bioactive conformation. The consequence is a further entropic penalty which has to be compensated by the enthalpy of binding. It has been proposed that the enthalpy/entropy compensations are largely a manisfestation of the unique properties of water (1). "High energy" water surrounding the unbound ligand contributes to a favorable enthalpy of binding when such water molecules are released to the bulk solvent. Additional enthalpic contributions to the binding derive from the shape (bioactive conformation)-dependent ligand/receptor association which is driven by van der Waals and hydrophobic interactions, and can be made even stronger and more specific by hydrogen bondiiig and electrostatic interactions. A fundamental question in the understanding of bjnding of peptide hormones to membrane-bound receptors is the event of recognition of the ligand by the receptor molecule. The classical approach considers a complementary fit of the ligand to its receptor in a "lock and key" manner (2). This raises the question of whether the flexible ligand has to fold prior to its binding to allow for a "complementary fit" or whether folding is induced in the binding process ("induced fit"). As isolated receptors were not available for decades, this question has addressed with model systems among which the ribonuclease (RNase) S ana S represented the most classical examples. RNase A is enzjmiatically cleaved by
Abbreviations: CCK. cholecystokinin; RNase. ribonuclease: G-protein, guanine nucleotide binding protein:
GPCR,
G-protein-coupled
receptor:
SDS
sodium
dodecylsulfate,
CTAH,
hexadecyltrJmethyl ammonium hydroxide: DMPC, di-myrlstpylphosphatidylcholine: DPPC, dlpalmitoylphosphatldylcholine: CMC, critical micellar concentration: SUV. small unilamellar vesicles: CD, circular dichroism: NMR nuclear magnetic resonance: hs-DC, high sensitivity differential scanning calorimetry: IR-ATR, infrared attenuated total reflecUon spectroscopy: NOE, nuclear Overhauser effect: MD, molecular dynamics: DMSO, dimethylsulfoxide: TFE, triiluoroethanol: for abbreviations of peptides see tables land 2, and fig. 11.
821
Fig, J.
Partition of peptide hormones between the extracellular aqueous phase and the cell membrane bilayer, and the resulting conformational equilibria at the water/lipid interphase.
subtilisin into two fragments, i.e. the 20-membered S-peptide and the 104membered S-protein, which reassociate in a 1:1 complex restoring a fully active enzyme (3). Whilst for the S-peptide no ordered structure could be detected in aqueous solution (4), it folds into an a-heBx in the S-protein-bound state (5) generating contacts and interactions with the S-protein counterpart identical to those observed in the x-ray structure of RNase A (6) and RNase S (7). This was well assessed by extensive structure-function studies on synthetic S-peptide analogs (8,9). The results of these studies strongly supported the mechanism of an "induced fit" in the molecular recognition of unfolded peptides by folded proteins, which has recently been further documented by the x-ray threedimensional structures of small peptide antigen/antibody (10) or peptide/human class II MHC protein complexes (11). Reinvestigation of the RNase S system at low temperatures (TC) allowed to detect the presence of a percentage of a-helical conformers in the S-peptide (12-14) and thus, to propose the alternative mechanism of the "complementary fit" even for this system, where binding of the folded S-peptide species to the S-protein represents the driving force for the observed conformational shift and thus, for the quantitative formation of the fully active RNase S'. As the RNase S system mimics efficiently the bimolecular
822 interaction between low mass peptide ligands and soluble receptors, it allowed precious information to be obtained about the specificity of molecular recognition and the prevailing contributions of aromatic interactions and hydrogen bonds to the binding affinity. With an increased knowledge about constitution and physical state of cell membranes, about structure and location of membrane-bound receptors and the two-dimensional mobility of the bilayer components as well as of their transbilayer movements (for recent reviews see ref. 15-22), a new scenario has been created of the cell surface as the first contact site of peptide hormones with target cells in their endocrine commimication pathway. Circulating hormones in a "firee" or "carrier boimd-state" collide with cell membrane surfaces and partition across the boundary separating the aqueous extracellular environment from the membrane bilayer. The resulting distribution determines the concentration of the hormones in each phase. This unspecific collisional event of the ligands with cell surfaces is statistically extremely more favored than the direct collision with the receptors, particularly in view of the actual distribution of receptors in the active state on the membrane bilayer. Based on this working assumption, Schwyzer has proposed an accumulation of the ligands on the cell surface and a prefolding and preorientation in the membrane environment as potential pathway of the hormone/receptor recognition process. This would compensate in successive steps the entropic penalty resulting from the loss of freedom of translationaJ and rotational motion of the fully flexible peptides in the bulk water phase (23-24). Moreover, the membrane-induced conformation and orientation of the peptides could also facilitate the peptide/receptor binding process as the stereochemical and functional requirements of the receptor binding sites are possibly better satisfied by the prefolded form than by the random coiled structure usually adopted by unrestrained peptides in aqueous solutions. This hypothetical pathway is schematically represented in fig. 1 and it foresees a two-dimensional migration of the hormone at the water/lipid interphase to the receptor as well as its hypothetical lateral penetration into the large receptor molecule to reach the binding cleft. The recent fast developments in moleculsir biology allowed for cloning and sequencing of a large number of membrane-bound receptors, the majority of which are composed by seven putative transmembrane segments connected at the two bilayer surfaces in sequence manner by more or less extended loops. As observed for most membrane-bound proteins, the non-degenerative purification and crystallisation of G-protein-coupled receptors (GPCR) are still problematic. At present the only possible alternative way of access to the spatial structure of these membrane-bound receptors is the molecular modelling using
823 bacteriorhodopsin (25) or more properly rhodopsin (26) as experimental models. Alt±iough there is no apparent sequence homology between bacteriorhodopsin and any GPCR, mammalian opsins which belong to the superfamlly of GPCRs and have sequence homology with their congeners, do possess the same activation process as bacteriorhodopsin. It seems therefore very likely that there is structural and functional homology betwen rhodopsins and the GPCRs. Thus, the choice of these proteins as templates for modelling experiments seems Justified and has found consensus. These MD simulations led to pictures of the GPCRs consisting of a bundle of seven transmembrane a-helices in a sequential anti-clockwise manner when observed from the extracellular phase (27-30). Their tight packing defines a central cleft whereby the extracellular loops are forming more or less bull^r domains at the bilayer surfaces as shown schematically in fig. 2.
Fig. 2.
Schematic representation of a receptor with the putative seven transmemhrane helices coupled to the peptide (H) and to its transduction system (G-protein, G).
824 Such structural models, however are hardly compatible with a lateral penetration of peptide ligands as required in a membrane-bound pathway ci »he hormone/receptor recognition and binding process. Experimental evidences for such mechanism are difficult to produce, since in the partition equilibriimi of hormones between the water and the lipid phase occupation of the receptor from the extracellular phase cannot be excluded. However, with an escape of the peptide ligands into the water phase and subsequent direct collision with the receptor binding site all the thermodynamic and conformational advantages of an accumulation at the membrane surface would be lost. Irrespective of the validity of the hypothetical membrane-bound pathway great efforts have been made in the last few years to determine preferred conformations of peptide hormones in media mimicking the more hydrophobic environments of membranes and/or receptor binding clefts. Our contributions in this field of research are confined to a pair of closely related gastrointestinal hormones, i.e. to gastrin and cholecystokinin (CCK), which were chosen as model bioactive peptides.
2. Gastrin and CCK
2. J. Biological Properties and Structure-Function
Relationship
Gastrin and CCK are biosynthesized by posttranslational processing of precr^rso:^ proteins which contain only one copy of the hormone sequence. Besicv:^ ^'^ sequence homology of the processed hormones, a pronounced heterogen .. v ^i size and state of sulfation of the tjnrosine phenolic function is characteristic for the circulating forms of gastrin and CCK. The biological properties, the distribution of the circulating forms and the structural features involved in the expression of the specific hormonal activities have been comprehensively reviewed (31-33). In the case of gastrin the dominant circulating form is little gastrin, a 17membered peptide originally isolated by Gregory and Tracy from porcine antral mucosa both in the sulfated and unsulfated form (34). Two additional gastrin peptides, circulating in minor amounts, were subsequently identified, the biggastrin with 34 residues and the minigastrin with 14 residues, both exhibiting identical hormonal potency as little gastrin (35,36). In the meantime it h a s been well established that the biological functions of gastrin are mediated by the gastrin/CCK-B receptor (37) and that the hormonal activity is exerted mainly in the digestive tract where the hormone stimulates secretion of gastric acid and of
825 electrolytes, and induces smooth muscle motility. Another important action of gastrin is the trophic effect, i.e. the stimulation of growth of acid secreting mucosa. Regarding CCK, upon the initial identification and isolation of a 33-membered peptide from porcine upper intestine, i.e. CCK-33 (38), this was found to represent the main circulating form in human gut (39). Subsequently, immunochemical studies using C-tenninal specific antibodies have revealed a wide range of circulating forms of CCK such as CCK-58, CCK-39, CCK-33, CCK25, CCK-18, CCK-8 as well as CCK-5 and CCK-4, the latter two deriving possibly from the enzymatic processing of gastrins, too. Many of these forms have been isolated, and their sequences were fully confirmed. The CCK peptides are produced in neuroendocrine cells of the upper intestine, but also in the central nervous system where the major circulating form is represented by CCK-8. In the periphery, the action of the CCK-peptides is mediated mainly by the CCK-A receptor (37) with high affinity for the sulfated forms. Among the numerous biological effects within the gut, CCK induces
Can be substituted by Leu, Nle. Mox
The peptide bond is essential for biological activity
Can not be replaced, essential functional group
SO Hay
Increases potency -^.R^^^^^^^^^^N
Indole moiety Important for biological activity
i/ H
NH2 -•— Essential for biological activity
Binding sites
Minimal structure for full biological activity
Fig. 3.
Schematic representation of the structure-Jimction correlations of the gastrin molecule as known from studies performed so far (for recent review see ref 33).
Conllguration at the Ca atom important for blnding to CCK-B receptors. Indole integrity
Increases
Crucial for
Conformational
nntrnrv
h4nAlnu tn P P K - A
rnlp
/
Does not alter
---
/
Can be
replaced by
QCHJ
Can be replaced by Me. Mox
\o H
v
Peptide bond essential for biological activity in
Essential for the biological activity
Minlmal structure for reconition by CCK-B
.
-
-
Mtnimal structure for recognition by CCK-A
Fig. 4.
Schematic representation of the structure-function correlations of CCK peptides as known from the studies performed sofar lfor recent review see ref. 331.
827
contraction of the gall bladder, stimulates the exocrine and endocrine pancreatic secretion, and inhibits gastric emptying. In the brain where the action of CCK is mediated predominatly by CCK-B receptors (37), this hormone is involved as neurotransmitter in satiety, hypothermia, analgesia, nociception, and in the regulation of cerebral blood flow. Extensive structure-function studies have been performed on gastrin and CCK by chemical modifications of the natural sequence and by synthesis of selected analogs in order to disclose the functional information encoded in the single residues of the linear peptides. The results are simimarized in figs. 3 and 4. Briefly, the common C-terminal tetrapeptide amide sequence represents the message site of this family of hormones as it possesses the full range of biological effects, particularly if mediated by the CCK-B receptor, although with remarkably lower potency (40-42). Replacements in this portion of the molecule without partial or total loss of receptor affinity and hormonal potency is allowed only for the methionine residue. It can be substituted by the more hydrophilic oxaanalogue methoxinine (43,44) and by the more hydrophobic carba-analogue norleucine (44-47); substitution with the side chain-branched leucine residue is allowed in the case of gastrin (44). but in CCK-peptides regarding peripheral actions it leads to partial loss of potency (46). Sulfation of the tyrosine in CCKpeptides was found to be essential for type-A receptor-mediated activities, whereby even the exact spacing of the tyrosine-O-sulfate from the C-terminal active site is apparently crucial since sulfated gastrin peptides are recognized by the CCK-A receptor as poorly as unsulfated gastrin or unsulfated CCK-peptides (48). Sulfation is not essential for recognition of both CCK- and gastrin-peptides by the gastrin/CCK-B receptor. Therefore, the physiological significance of gastrins occurring in the sulfated and unsulfated form remains intriguing, although even in this case sulfation of tyrosine is enhancing receptor binding affinities and hormonal potencies (48-51). Finally, regarding the minimal sequences of the two hormones capable of eliciting full biopotencies, these were found to correspond for CCK to the acylated CCK-7 (52) and for gastrin to the analog [Pyr^,Nlel5]-gastrin-[6-17] (53); see table 1 for the gastrin peptides utilized in this study.
2.2 Preferred Conformations of Gastrin in Solution and in Presence of Surfactant Micelles
Identification of the biologically relevant conformations of small unrestrained linear peptides is often more difficult than in the case of folded proteins. It is difficult to fmd suitable solvent systems that lend themselves to spectroscopic
00 h, 00
gastrin-17 qrI.-Gly-Pro-Trp-Leu-Glu-Glu-Glu-Glu-Glu-Ala-Tyr-Gly-Trp-Met-Asp-Phe-NH~ [Nle15]-gastrin17 qrI.-Gly-Pro-Trp-Leu-Glu-Glu-Glu-Glu-Glu-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH~
H-Gly-Pro-Trp-Leu-Glu-Glu-Glu-GluGlu-Ala-TyrGly-Trp-Met-AspPhe-NH2
gastrin-[% 171
125I-BH-[Nle15l-gastrin17 ~~~I-Desamino-Tyr-Gly-Pro-Trp-Leu-Glu-Glu-Glu-Glu-Glu-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH~
[Nle151-gastrin-15171 H-Leu-Glu-Glu-Glu-Glu-Glu-Ala-~-Gly-Trp-Nle-Asp-Phe-NH~ [F'yr6,Nle15]-gastrin-[6-171 Pyr-Glu-Glu-Glu-Glu-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH~
[l3~~7,Nle15]-gastrin-[ 7-171 Fyr-Glu-Glu-Glu-Ala-Tyr-Gly-Trp-Nle-Asp-PheNH2 (Pyr8,Nle15]-gastrin-[8171 €3~-Glu-Glu-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH2 m-Glu-Ala-Tyr-Gly-Trp-Nle-Asp- Phe-NH~
[Fyrg.Nle151-gastrin-[g-1 7 1
Fyr-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH2[~yrlo,~el5]-gast-[ 10-171 H-Trp-Met-Asp-Phe-NH2
gastrin-[14-171
Table. 1. Sequences of the gastrin peptides used in the studies discussed in this review and their abbreviations; (Pyr = pMrnglutamic acid)
829 measurements and at the same time impart on the peptide molecule a specific conformation relevant to its biological activity. Moreover, the overall flexibility of small peptides may lead to various interconverting conformers at conformational equilibriimi. Correspondingly, the structural information derived from spectroscopic experiments may reflect just an average of conformations, and many diagnostic features of ordered structure can be obscured by the presence of random coiled forms as well as by fluctuations around ordered ones. Several techniques and experimental conditions need, therefore, to be employed in such conformational studies in order to allow for decisive conclusions to be drawn. In the case of the gastrin peptides extensive conformationsd studies have been performed in aqueous and aqueous organic media. From these it has been concluded that water is not a structure-supporting environment for gastrin related peptides (54-57). Since the discovery that trifluoroethanol [TFE) is capable of inducing and stabilizing a-helical structure in oligopeptides (58), this solvent has extensively been used to detect preferred conformational states of bioactive peptides in conditions of reduced water activity, although the precise mechanism of stabilization of secondary structure elements in this solvent is still unknown. According to theoretical considerations that define TFE as an a-helix enhancing solvent, TFE acts only within the context of a preexisting helix-cofl equilibriimi. It interacts preferentially with the helical conformation of the peptide and, as a consequence, it shifts the conformational equilibrium towards the helical structure. On the other hand, it is well known that TFE is also an a-helix inducing agent, since helix propensity is not required for induction of such conformational state in a peptide. Because of the lower dielectric constant of aqueous TFE than that of water (59), it was proposed as proper miimicry of the microenvironments of cell membranes and of receptor binding clefts. The CD spectra of gastrin peptides related to the active site, i.e. to the C-terminal tetrapeptide, and of those elongated up to the heptapeptide in TFE indicate different degrees of rigidity of the tryptophan side chain as well as strong changes in the backbone conformation. Therefore the hypothesis of a constant geometry of the active site when the chain is extended in sequence mode at the N-terminus (60) as explanation of the very simflar biopotencies of these peptides could not be confirmed (54). On further stepwise elongation of the peptide chain up to the [Nlei^]-gastrin-[5-17] the CD spectra reveal a sigmoidal transition in terms of tendency for ordered conformations as well documented by fig. 5. This would indicate a cooperative effect of the chain length in folding process of gastrin peptides into preferred secondary structures (55). The shape and intensity of the CD spectrum of [Nle^5l-gastrin-[5-17] clearly reflects the presence of a-helix, pform and random coil. The intensity of the negative maxima and correspondingly, the content of a-helix is further enhanced by elongation of the peptide chain u p
830 to the little-gastrin sequence, i.e. to [Nle^^l-gastrin-l?. Interestingly, the biopotencies of this series of peptides (49) parallel the onset of ordered structure in TFE as shown in fig. 5, thus strongly supporting a possible biological implication of this increased conformational order. Finally ^H-NMR analysis of [NleiS]-gastrin-[5-171 allowed to identify its conformation in TFE (61,62). It consists of a hairpin structure with an N-terminal a-helix, a chain reversal at the
< 2. <
0.5
h
5 10 15 Number of residues in the chain
Fig, 5.
Relative variation of molar ellipticity values (Al0]/[0]toti at 216 (O) and 192 nm (A) as a Junction of chain length. Data of hormonal potency of the various gastrin peptides are reported for comparison (•).
sequence Ala-iyr-Gly-Ttp, followed by a C-terminal Sjo-helix. Using the CD and NMR data this preferred conformational state of [NlelSl-gastrin-[5-17] was modelled £ind it is shown in fig. 6; the compactness of this structure of the linear peptide becomes even more evident by a space filling representation. Support for the h)^othesis of an U-shaped structure, as potential bioactive conformation of gastrin, was obtained by replacing the glycine residue with alanine (63). CD and ^H-NMR analysis of this analogue in TFE confirmed the expected extension of the a-helical stretch towards the C-terminus and thus, the absence of a chain reversal. This conformational change was found to exert a significant impact on the bioactivities as the receptor binding affinity and hormonal potency were reduced to those intrinsic of the C-terminal tetrapeptide. It was therefore concluded that folding of gastrin in a hairpin mode stabilizes the bioactive structure of the active site with an concomitant 10-fold increase of the hormonal activity. Interestingly, sulfation of gastrin was found to affect only marginally the dichroic properties of gastrin in water and in TFE (64), a fact that
831
B Fig, 6.
Preferred conformation of [Nle^^]-gastrin-l5-l 7] in aqueous TFE. A) ribbon drawing of the backbone: B) space fUling representation of the molecule.
832
compares well with the hormonal potencies of sulfated and unsulfated gastrins. Conversely, C-terminal desamidation which is known to abolish completely functional binding to the receptor (42), was also without effect on the CD spectra, too (65). This observation clearly indicates that the C-terminal amide, although representing a crucial topochemical element at the level of the gastrin/receptor interaction, is not involved in stabilizing the bioactive conformation. Besides aqueous organic media as mimicry of membrane and/or receptor environments, even surfactant micelles have widely been used. In the case of gastrin an insertion of the highly charged peptide into SDS micelles does not occur at neutral pH values. However, at pH 2.2 where the side chain carboxyl functions cire protonated, insertion takes place as well assessed by fluorescence measurements, and the dichroic properties become very similar to those of gastrin in TFE (66). ^H-NMR analysis of [Nlel^]-gastrin-17 under these conditions, i.e. in SDS micelles at acidic pH values, allowed to confirm the structure proposed for gastrin in TFE. Again an U-shaped conformation was deduced with an a-helix spanning the sequence from tryptophan-4 to alanine-11 followed by a turn in the Ala-Tyr-Gly-Trp portion and by a SiQ-helical stretch in the C-terminal portion of the molecule (67). The unphysiological pH conditions used in these studies led us to examine the effect of differently charged micelles on the [Nlei5]-gastrin-17 conformation at neutral pH values (unpublished results). In the presence of the negatively charged sodium dodecylsulfate (SDS) and of the neutral surfactant octyl-p-glucopyranoside above the critical mlcellar concentration (CMC) an interaction of the gastrin molecule with the micelles could not be detected by fluorescence measurements; no blue shift of th^ tryptophan emission maximum referred to that in water was observed and ti Stem-Volmer quenching constant with KI was identical to that measured in aqueous solution. As shown in fig. 7, even the dichroic properties of the gastrin analog were not affected. However, in presence of the positively charged micelles of hexadecyltrimethylammonium hydroxide (CTAH) a strong blue shift of the fluorescence emission maximum by 7 nm was obtained, and the resulting CD spectrum strongly reminds that in TFE, although the intensities 6{ the negative maxima are reduced by about 30%. All data indicate a high preference of gastrin for an U-shaped ordered structure in different environments. This has also been demonstrated with immunological experiments. By selective linkage of gastrin-[2-171 to the single surface exposed cysteine residue-107 of iso-1-cytochrome c the peptide moiety is grafted to the surface of a protein where it may fold into a low-energy ordered structure possibly identical to its bioactive conformation at receptor level. In fact, the polyclonal antibodies raised against this conjugate in rabbits were found to cross-react with
833
5
0
-5
Q)
o
jy
-10
rH
-15
-20 1
195
i
205
1
1
1
215
1
225
1
1
235
1
1
245
X(nm)
Fig. 7.
CD spectra of [Nle^^hgastrin in 5 rnM phosphate buffer, 100 mM NaCl at a peptide/CTAH molar ratio of 1:58 (curve ij, peptide/octyl-ftglucopyranoside molar ratio of 1:1562 (curve 2), peptide/SDS molar ratio of 1:100 (curve 3). The spectrum of [Nle^^fgastrin in absence of detergent (curve 4) is reported as reference.
gastrin peptides with a rank order of binding affinities identical to that shown by the same set of peptides for the gastrin receptor (fig. 8). Again the plot of crossreactivities of the antibodies vs. chain length of the gastrin peptides parallels that of the preference of these peptides for ordered conformation (65). Identical immunoresponses were obtained with well defined conjugates of gastrin peptides to the human IgGl hinge fragment 225-232/225'-232' (68-70). In these conjugates the gastrin portion retains its ability to fold into the gastrincharacteristic conformation, and again antibodies were raised in guinea pigs which exhibited gastrin receptor-like specificity.
834
Pyr Gly Pro Tip Leu Glu Glu Glu Glu Glu Ala Tyr Gty Trp MetAsp Phe-NH,
1 2 3 4 5 6 7 8 9 10 11 1213 14 1516 17
Fig, 8,
Crossreacttvities of gastrin related peptides as determined by 50% inhibition of binding of gastrin to antigastrin antiserum V5 and expressed as percentage of the IC^Q of gastrin. The gastrin peptides analyzed are: gastnn'[14-17] (•), [Pyr^0^Nle^5].gastnn-[10'17] (0), fPyr^.Nle^^hgastrin19-17] (•), [Pyr8,Nlei5Kgastrin'l8'17] (O), [Pyr7,Nle^5].gastrin'[7'17] (A), [Pyr^,Nlei^}-gastnn-[6-17] (A), [Nle^^J'gastnn-[5-17] (•) and [Nle^^hgastrin (or gastrin) (Q),
2,3. Preferred Conformation of CCK in Solution and in Presence of Micelles
Surfactant
The specific interaction of the CCK hormone with two types of receptors, i.e. with CCK-A and CCK-B, could imply that the highly charged and flexible linear peptide is capable of adopting different bioactive conformations in the dynamic process of receptor recognition and binding. In this context conformational analyses have been performed on CCK-8 and related shorter fragments using CD, energy transfer fluorescence techniques, ^H-NMR and computational procedures (60,7177). The results underline the difficulty encountered in determiniiAg the conformational preferences of this molecule and thus, its potential bioacilve conformation. A broad spectrum of structures has been proposed which differ in
^
00
O)
u U
gC
CN
z
a; X3 (X
i-i
1
o o
o o
(N
CO
IC
|i^
o 7 PQ {sii ,5^
t_i
^ ta:J
13 O 7 {a:^
i-H
ti S
12; a;
^ O. a
2: b
PH
XJ 0.
a;
CO
0}
CO
835
CO
Z
cs
OH
a; XI CU
2:
cs
flJ
Cli
a; X! Pi
2
csi
3 ;:^
d
a
o;
4->
^
^
CO
E
01
g^
00
ou
CO
^
a
CO
i 1
OH
s
^ a
^
/^
Ui c;
9
§'*•'
c -S
! i-i a>
tio
tiX)
^2
i^ ii112
a
1 ^^^^ f
O
^
(N
P4
a; X3 Cli
-M
2
ffi ffi ffi
a; X\ 0.
^4->
S
^
1 ^
01
0
^>
ffi
CO
>.
CO
>>
0 K" ^up
9!_,
CO
a ^tdD
>^
a;
CO
OH
4I.
CO
a ^yio
^
cu ^
jt
s i
-
S
I
CD
CO
-s
.Si
CO
3
CO
•S
I
s;
§
o b o c o o o o 2
95 cb
12;
a
XI O,
^
•+->
a;
^' M
s
a;
a;
§
& ^
S 2
>.
(D
^< M
^
o
1 -J-J
a
^
s s s s s g i Oi s0
1 4-»
>. >. O
g-g-g-g^^^^
^ 1
9 o o ^
a;
-M
CO
s
E O
1
?
^
\ij
CO
o CO
^ a ^
i i ^
s CO1
^
.$S
^
ffi
CO
ffi
836 the spatial array of the side chains, but more importantly in the peptide backbone, too. Recently we succeeded to determine a preferred conformation of the fully active CCK-analog [Thr,Nle]-CCK-9 (see table 2) by ^H-NMR analysis in the cryomixture dimethylsulfoxide (DMSO)/water at 278 K . As shown in fig. 9, it consists of a ytum centered on the threonine and separated by the glycine residue from an ahelix involving the C-terminus, whilst the N-terminus is flexible and salt-bridged between the tyrosine-O-sulfate and the arginine guanido function (78-79). Again the space-filling model demonstrates a rather compact structure particularly in the C-terminus. The spatial structure of the CCK-9 analog strongly reminds in the C-terminal tetrapeptide portion, i.e. in the active site, the 3D-structure of gastrin. These findings fully agree with the pharmacological properties of CCK which is capable of interacting with the gastrin/CCK-B receptor with high binding affinity. But then its selective recognition by the CCK-A receptor has to be attributed exclusively to a specific electrostatic interaction of the tyrosine-Osulfate moiety with a receptor counterpart. Thereby the exact location of the hemi-sulfate ester moiety is apparently playing a decisive role since sulfated gastrin with the tyrosine-O-sulfate shifted by one position in the sequence is poorly recognized by this type A-receptor (48). As discussed in section 2.2., N-terminal extensions of the active site portion of the gastrin molecule leads to a remarkable stabilization of its potential bioactive conformation with concomitant increase of hormonal potency. This aspect has been analyzed in the case of CCK too, by using the series of peptides listed in table 2 (64). CD measurements revealed that N-terminal extensions of the CCK mrlecule in sequence mode are affecting only marginally the conformational St: s of the bioactive core. As expected from what was known for gastrin, the CD sp ra of the CCK-peptides in aqueous solution were consistent in shape and in isity with predominantly random coiled structures. However, titration of the aqueous solutions of the CCK-peptides with TFE was found to induce ordered conformation which is apparently stabilized at least to some extent by N-terminal elongation up to the undecapeptide. On the other hand, the biological properties of these CCK peptides, as determined in different assay systems, were insensitive towards an increase of the CCK-peptide size, thus definitely confirming that CCK8 and acylated CCK-7 represent the smallest fully active sizes of the CCK hormone regarding CCK-A receptor-mediated signal transduction (64). It seems, therefore, reasonable to conclude that, differently from gastrin, the biologically relevant conformational states of the bioactive core, i.e. of CCK-7, at least reirarding the receptor A-mediated activities, are not affected by the peptide size, bi "lat stabilization of this conformation in the larger size CCK-peptides derives m :y by weakened endgroup effects. The CD spectra recorded for this set of
837
B Fig. 9.
The preferred conformation of IThr.NlehCCK in DMSO/water. A) ribbon drawing of the peptide backbone; B) space fRing model
838
CCK-peptides at high TFE concentrations are very similar to those of gastrin peptides of similar size both regarding the overall pattern and the intensities of the maxima. The latter spectra were attributed to y-tums, and a folding of the Cterminal portion of gastrin into a helical structure in TFE had been confirmed by NMR analysis (67). The presence of the identical C-terminal pentapeptide in the CCK molecule makes it most reasonable to assume that the observed CD properties in aqueous TFE reflect a similar conformational state of this sequence portion in CCK, too. This would also compare well with the 3D structure determined for (Thr,Nle)-CCK-9 by NMR in aqueous DMSO since the relatively small conformational shift from a SIQ- to an a-helix could easily result from the strong a-helix inducing effect of TFE as solvent (80,81). In presence of negatively (SDS) and positively charged (CTAH) surfactant micelles a significant induction of ordered structure for the (Thr.Nle)-CCK-9 analog was detected by CD measurements (64), but the spectra were different from those in aqueous TFE. The location of the negative maxima are more consistent with ptype structures. In presence of the neutral octyl-P-D-glucop5n:anoside micelles again a transition from unorderd to ordered structure is induced confirming strong interactions with the uncharged micelles, and the resulting CD pattern compares well with that obtained in aqueous TFE. This would suggest that differently from what has been found for the homologous peptide gastrin, the CCK-peptides exhibit a pronounced tendency to assume various ordered conformations depending upon the physicochemical environment.
Antisera
CCK-12
CCK-10
gastrin
INlelSj-fiastrin-17
lNlelSj-gastnn-17
Anti-CCK-10
1.6x10-^2
1.7 X 10-^2
2.3 X 10'^2
1.5 X 10-12
3.0 X 10-12
6.0 X 10*12
1.8x10-^2
2.0 X 10"^2
2.6xl0'12
3.0X10-12
2.4 X 10*12
5.0 > 10-12
fiastim-(14-171
(2725) Anti-CCK-13 (2795)
Tables.
Specificity of anti-CCK antisera raised with CCK-10 and CCK-13 iso-lcytochrome c corgugates in guinea pigs as determined by competition assays using EUSA techniques. Crossreactivities of CCK- and gastrin peptides are expressed by the respective IC^Q values using CCK-12 as coated antigen.
Additional support for a folding of the CCK-peptides into a conformation similar to that of the C-terminal portion of gastrin with the N-terminal tail as flexible arm for a decisive electrostatic interaction with a receptor counterpart was obtained with immunological experiments (82,83). CCK-10, CCK-12 and CCK-13 were linked covalently to the cysteine residue 107 of iso-l-cytochrome c via the
839 maleimide/thiol reaction principle. These conjugates were used In immunization experiments in order to examine the effect of an increased spacing of the bioactive core of the CCK-molecule from the carrier molecule on the specificity of the antisera raised in guinea pigs. Antisera were obtained which were uncapable of discriminating CCK-peptides from the gastrin peptides. Irrespective of whether the C-terminus is recognized by the immune system in its preferred conformation or not, it represents the identical continuous epitope as present in gastrin (table 3).
2 A. Interaction of Gastrin and CCK with Lipid Bilayers
Both aqueous organic solvent mixtures and differently charged micelles can mimic only roughly the environment of natural cell membranes. In order to analyze in more appropriate model systems possible interactions of gastrin and CCK with cell membranes and to determine their conformational states in lipid bilayers, we have recently investigated in detailed manner this aspect using liposomes. The similarity betwen liposomes and natural membranes is extensively exploited both in vitro and in vivo because of the ability of liposomes to mimic the behaviour of natural membranes. Moreover, the value of liposomes as model membrane systems derives from the fact that they can be constructed with natural constituents. In our approach, we selected as model membranes those formed with the zwitterionic lipids di-myristoylphosphatidylcholine (DMPC) and di-palmitoylphosphatidylcholine (DPPC) as these lipids constitute the major components of most cell membranes. Moreover, in order to operate with a simple system, small unilamellar vesicles (SUVs) were used, i.e. with a diameter between 25 and 250 nm as resulting by rod-type sonication or by extrusion (51). A possible interaction of gastrin with the model membranes at peptide/lipid ratios of up to 1:100 has been investigated by monitoring conformational changes via CD spectroscopy, insertion of the aromatic chromophores of the peptides into more hydrophobic compartments of the bilayer via fluorescence measurements and by analyzing changes of the thermotropic behaviour of the lipid bilayer. All assays indicated the absence of a detectable interaction of the negatively charged gastrin molecule with the zwitterionic bilayer (84). This result was not unexpected in view of the lack of insertion of gastrin into negatively charged or neutral micelles at neutral pH values as reported in section 2.2, but it contrasts the findings of Schwyzer et al. (85) obtained by ATR-FTIR on lipid films with incorporated gastrin. The forced interaction of gastrin in lipid films which led to propose a perpendicular insertion of the C-terminal tail of gastrin in a-helical conformation into lipid bilayers, may be too artificial, since in our more natural two-phase model no interaction with the bilayer accompanied by conformational
840
transitions could be detected. Our findings disagree also with those obtained with pentagastrin, i.e. Boc-p-Ala-Trp-Met-Asp-Phe-NH2 (86-88). An insertion of this molecule into lipid bilayers has been well documented by various spectroscopic methods related to the tryptophan moiety, but it cannot be excluded that the observed lipid affinity is mainly dictated by the hydrophobic "non-gastrin" portion adjacent to this aromatic residue. A significantly different behaviour was observed for (Thr.Nle)-CCK-9 which interacts transiently with the lipid bilayer of DMPC SUVs according to highsensitivity differential scanning calorimetry (hs-DSC) measurements. It causes fusion of the vesicles and is then expelled because of the higher degree of order of the bilayer in larger vesicles (84). These findings raise the question of whether a similar physical phenomenon is implicated in the known fusion of neurotransmitter vesicles with the membrane of the synaptic junctions with concomitant release of the peptides (89.90), since CCK peptides are known to act as neurotransmitters. The observed fast expulsion of the CCK from the lipid bilayers and the absence of detectable insertion of gastrin into DMPC bilayers should not exclude interactions of these peptides with natural membranes, as these are known to be structured in domains of different lipid composition and of differentiated lipid bilayer packing. In order to overcome the problem of negligible interaction of the peptides with model bilayers and to force an insertion of the peptide hormones into lipid bilayers, lipo-derivatives of the peptides were synthesized. 3 . Lipophilic Derivatives of Gastrin and CCK
3.1, Design and
Synthesis
Nature is anchoring biomolecules to cell membranes with single transmembrane helices, helix bundles or sticky fingers. Among the lipophilic derivatives the most widely occurring forms are fatty-acylated amino groups, hydroxyl groups or cysteine thiol functions, isoprenylated cysteine thiol functions as well glypiated carboxyl functions. In order to assure optimal interdigitation of artificial sti-^ky fingers with membrane bilayers and tight entrapment of lipo-peptide derivai es in bilayer structures we have proposed the use of di-fattyacyl-glycerol moieties (91). These handles are very similar to the structure of the natmal glycosylphosphatidylinositol anchor shown in fig. 10. As the chirality of the glycerol moiety is only marginally affecting the packing of bilayers (92-94), the synthetic efforts were confined to the preparation of roc-di-fattyacyl-glycerol congeners suitable for a selective conjugation to peptides and proteins (95).
841
,-v.
Protein
. ?
o-p=o
0 1 6Mana1
±Qalo1-
±Gaia1
F\Q, 10.
^ 2Mana1
^2Gaia1
y
2Gata1
X .6 Mana1
4GicNH^1
6myt>-4no8itol1
I
o 1 o=p-o' 1
?
, ,
Structure of the glycosylinosttolphosphate structure used by nature to anchor proteins at the C-terminus to cell membranes.
Similarly to what nature is using, the thiol function has become a main target in conjugate chemistry as it allows selective crosslinking of molecules via disulfide or sulfide bonds exploiting mild thiol disulfide interchange or thiol addition reactions. For latter reaction type, the maleimide group as thiol acceptor (96) has found widespread applications. It was shown that this group is sufficiently stable under certain conditions of peptide synthesis to enable its incorporation at preselected peptide chain positions and thus, to represent an ideal anchor for subsequent covalent linkage of thiol-functionalized molecules (65,82,97-101). Correspondingly, symmetric and unsymmetric rac-1.2-di-fattyacyl-3mercaptoglycerol derivatives were synthesized in good overall yields following scheme 1 and 2. Both in the case of gastrin (53,97) and CCK (64,102) it was known that Nterminal modifications do not affect their bioactivity profile. Correspondingly, the N-terminus of gastrin and CCK was used for grafting the lipid moiety via the thiol/maleimide approach. For this purpose p-alanine was chosen as spacer of the maleimide group since the methylene moiety allows for sufficient flexibility without displacing too much the peptide chain from the double-tailed lipid. This fact was expected to allow more appropriate mimicry of natural lipids and thus, a better interdigitation with lipid bilayers.
842
r-\
HN-CQz-N
O
f-Bu-S-N -COfe-N
O O
,—OH I—OH
-
RC02H/DMAP/DCC r.L. 12 h
HO—1
1NNaOHA3ioxane r.L, 12h
HOH
92%
85-08%
I—S—S-^-Bu
-SH
L—S—S-f-Bu
O
X
BU3P/CF3CH2OH r.L. 12 h >95%
^SH
Scheme 1. Synthesis of symmetric l,2'di-fattyacyl'3'mercaptoglycerol derivatives
HO-n HO—
Trt-CIAoluene Py, 60'C. 20h
R'CO,H DMAP/DCC r.t, 12h
Trt-
HOH
69% •—S~S-f-Bu
95% L—S-S~f-Bu
R^O,H/DMAP DCC.r.t., 12h
ZnBr^CH,CU MeOH, r.t., 6min 85%
Trt~0-
*"v°i
85-90% -S-S~f-Bu
O
- S - S - • -Bu
Rl
V
R'^0^ L—S~S-f-Bu
BUjP/CFjCHjOH/ MeOf-Bu/HjO r.t., 12h >95% O
Scheme 2. Synthesis of derivatives.
^SH
imsymmetric
l,2-di-fattyacyl-3'mercaptoglycerol
843
AoH
DMF.r.t. 1.5h
H2N^-^OH
o
*
b HOSU/DCC 0°Ctor.t.4-24h
°"tir'^'p O
OH
OSu
p p
i
+
H-peptide
-*- [I
bsu
b
O
FT^O
R*^ " O H
+
N-(CH2)2-CO R^
DMF, r.L, 30 min 75-85%
,
• N-(CH2)2-COR,
-SH
Scheme 3. Synthesis reaction.
N-(CH2)2-COR,
of lipo-peptide derivatives via the maleimide thiol addition
O CH3-(CH2),TC-O-6H
O
CHz-S
v-/
-(CH2)2-ii-Peptide
Peptide: Arg-Asp-TyrCSOaH^Thr-Gly-Trp-Nle-Asp-Phe-NHj Gly-Pro-Trp-Leu-(Glu)5-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH2
Fig. 11.
Chemical structure ofDM-CCK and DM-gastrin.
[Thr.Nle]-CCK-9 [Nle^VHG-[2-17]
844 The maleoyl-P-alanine N-hydroxysuccinimide ester was conveniently prepared following the procedure of scheme 3 (101,102), and was then used to convert [Nlei5]-gastrin-[2-17] and [Thr,Nle]-CCK-9 into the reactive maleimido-derivatives for the final coupling of the l,2-diacyl-3-mercaptoglycerols to produce the lipopeptide derivatives DM-gastrin and DM-CCK shown in fig. 11 (51,104).
3.2. Physical Properties ofLipo-Gastrin and Lipo-CCK
By covalent attachment of the double-tailed lipid moieties to the N-termlni of [Nlel5].gastrin-[2-17] and [Thr,Nle]-CCK-9 the highly charged hydrophiUc peptides were transformed into the amphiphilic compounds DM-gastrin and DMCCK which were expected to aggregate in aqueous media into micelles or vesicles. Whilst sonication of DM-gastrin produces a polydispersed system of vesicles, extrusion generates a monodispersed population of vesicles of the diameter size of the filter used and of surprisingly high stability as determined by light scattering measurements (51). Moreover, by freeze-fracture electron microscopy only unilamellar bilayers were detected. Although this does not exclude the presence of a few multilamellar vesicles, formation of monolamellar vesicles should be preferred in view of the bulkiness of the headgroup and of the large hydration shell of the hydrophilic peptide headgroup. Interestingly, the surface of the lipogastrin vesicles appeared rough if compared to that of DMPC liposomes (105). In the case of DM-CCK sonication and extrusion, but surprisingly even simple vortexing leads to clearing of the aqueous solution indicating a significantly differentiated behaviour of the two homologous Upo-peptides (104). Taking into account the size of the headgroups in terms of peptide length an opposite effect was expected. Moreover, even after extrusion the lipo-CCK system rearranges into a polydispersed population of vesicles, the light scattering of which does not exclude the presence of micelles, too (104). The different behaviour of the gastrin and CCK derivatives can only be explained on the basis of a different interference of the two headgroups with the packing of the lipid tails into bilayer structures. Fluorescence quenching experiments with iodide were expected to give first indications in this context as both peptides contain in the C-terminal portion of the molecule a tryptophan residue, and gastrin an additional one near the lipid-grafting position. As shown in table 4, a blue shift of the fluorescence emission maximum was observed for both lipopeptide vesicles. This indicates an enhanced hydrophobic environment of the tryptophan residues referred to that of the unmodified peptides in aqueous solution (51,84). Similarly, the Stem-Volmer quenching constant [k^] was found to be remarkably lower for DM-gastrin than for the unmodified peptide. The experimental value was lower than that expected for monolamellar
845 Peptide [Nlel5]-gastrin-17 DM-gastrin [Thr,Nle]-CCK-9 1DM-CCK (vortexed) DM-CCK (sonicated)
A;ijnax(nin) 0 4-5 0 5-6 5-6
*^v
5.89 1.78 2.97 1.20 1.62
i
Table 4. Fluorescence properties and accessibility of the tryptophan residues of [Nle^^]-gastrin-17 and the lipo-gastrin derivative (DM-gastrin), of[Nle,Thr]CCK-9 and DM-CCK to iodide quenching; ^) the relative error on k^u amounts to 3%.
vesicles where the gastrin moiety should be statistically distributed on the inner and outer bllayer surface and thus, be accessible to the iodide quencher only to 50%. One tryptophan residue, most probably that located near the N-terminus. i.e. near the lipid moiety, must therefore be less accessible. As reported in table 4, similar results were obtained in the case of DM-CCK. The tryptophan residue of this molecule is located in the C-terminal portion, and an insertion of this residue into more hydrophobic compartments of the bilayer can occur only via chain reversal of the peptide headgroup. This differentiated behaviour of the two lipo-peptides is further confirmed by their dichroic properties in aqueous buffer. The CD spectra of [Nle^^l-gastrin-17] and DM-gastrin are shown in fig. 12. Whilst at low ionic strength interaction of the gastrin moiety with the bilayer structure allows for partial folding into ordered conformations, an increase of the ionic strength to more physiological values leads to exposure of the gastrin moiety to the bulk water in mainly random coiled structure (51). Thereby the shoulder around 215 nm could derive from P-type conformations involving aggregation, e.g. of the N-termini. Conversely, the CD spectrum of DM-CCK (fig. 13) even at high ionic strength reflects considerable contents of ordered structure of 7- and a-type turns (104) as present in the NMRstructure of [Thr,Nle]-CCK-9 in DMSO/water (79). Taking into account these indications as well as the ^H-NMR experiments on DM-CCK vesicles which allowed for an unambiguous assignment of an NOE between the indole group of the peptide moiety and the myristoyl-alkane chain of the lipid portion (84), MD simulations were performed in a CC^/water cell which mimics the cristalline L^phase of a phospholipid membrane (106). The results of these MD simulations showed that the y-tum at the threonine is maintaiined and that the C-terminus adopts an a-helical conformation with a superimposed p-sheet hydrogen bonding
846
JS
-10 h
196
206
ai6
225
Xtnm)
Fig, 12,
CD spectra of DM-gastrin (curve 1) cmd [Nle^^J-gastriri'l? (curve 2) in 5 mM phosphate buffer containing 100 rnM NaCl at pH 7,0.
Fig. 13.
CD spectra ofDM-CCK in 5 mM phosphate, 100 mM NaCl (pH 7.0 ) at a peptide concentration of 1.4-10-'^ M.
847 pattern (fig. 14). The helix axis of the C-terminal portion lies parallel to the interphase separating hydrophobic from hydrophilic residues, the only exception beeing the norleucine residue which points into the bulk water. The tryptophan is buried into the hydrophobic phase shielded by the phenylalanine
7]
/
^
Y
^
""^ vCV^^
^
A
%
\y
[y
Fig. 14. Stereoview of the energy-mintmized conformation of DM-CCK tn its averaged position in the biphasic water/CCU ceVL The tipper phase is water and the lower CCI4 .
residue. This hydrophobic clustering compares weU with the observed strong blue shift of the fluorescence emission maximum and with the reduced Stem-Volmer iodide-quenching constant described above. It also agrees with the NMR experiments where a distinct NOE was detected between the indole side chain of the tryptophan residue and the alkane chain of myrlstlc acid. Regarding the whole molecule, i.e. including the fatty acid chains, the chain reversal in the peptide moiety produces a conical shape which in view of the mismatch of the cross-areas between headgroup and fatty acid chains, if compared to those of DMPC. is expected to greaUy affect the order of the bilayers. This would explain the fast rearrangement of extruded DM-CCK vesicles into a polydispersed system. By analyzing the thermotropic behaviour of the two lipo-peptides with hs-DSC no chain melting transition could be detected above 5* C indicating that the vesicles
848 of DM-gastrin and DM-CCK are in the liquid state despite the double-tailing of the lipid moiety (104,107). This fluidification of the bilayers can be attributed to the presence of the hydrophilic and relatively large peptide chains that function as polar headgroups. It should not result from the racemic di-fattyacylthioglyceryl moiety as it is known that the configuration of the phospholipid is affecting only marginally the phase transition temperatures (92-94). The higher fluidity of the DM-CCK than that of DM-gastrin, as deduced from the easier clearing of the vesicle preparation, should derive from its conical shape, i.e. from the chain reversal with insertion of the C-terminus into the lipid bilayer. Conversely, in the case of DM-gastrin the peptide moiety is mostly exposed to the bulk water phase, thus allowing for a more ordered packing of the lipo-tails.
3.3. Interaction of Lipo-Gastrin and Lipo-CCK with Phospholipid Bilayers
A lipid transfer process from one vesicle population to another is related to an equilibrium that involves the redistribution of lipids between donor lipid bilayers and acceptor structures of non-equivalent chemical potential due to their differing lipid compositions. The mechanism by which spontaneous lipid transfer occurs between membranes has been extensively investigated in various independent laboratories (108-111). In general it has been well established that in intervesicular lipid transfer processes the relative fluidity of the donor vesicles is much more important than that of the acceptor bilayers (112). Therefore, the low phase transition temperature of the lipo-peptides was expected to facilitate their transfer to phosphatidylcholine vesicles as model of cell membranes. In fact, DMgastrin transfer to DPPC vesicles was found to proceed at high rates (107). Replacement of the myristoyl chain with the palmitoyl chain in the lipo-gastrin (51) reduces the rate of transfer in agreement with previous findings (110,112,113). Moreover, the experiments confirmed that the lipid transfer process is strongly favored by the fluid state of the acceptor vesicles as previously reported for similar experiments by Martin and MacDonald (114). Upon insertion of the DM-gastrin into DPPC SUVs the sharp peak at 40.5" C broadens to give a large peak of low intensity with unchanged phase transition temperature indicating a statistical insertion of the lipo-gastrin molecules into the phospholipid bilayer (107). The transfer of DM-CCK to DMPC SUVs occurs rapidly and quantitatively even below the phase transition temperature of DMPC (104). Differently from what was observed in the case of lipo-gastrin, the endotherm of the system DM-CCK/DMPC exhibits a broad peak at 24.97° C corresponding to the phase transition temperature of the DMPC bilayer with statistically inserted lipo-CCK molecules
849 and two additional peaks at 20.24" C and 18.42° C, respectively. These could correspond to differently enriched CCK-domains. In order to confirm this working assumption the effect of Ca^^ ions on the phase transition temperatures of this merged system was analyzed (104). Reduction of the electrostatic charge of the lipid head groups, as a result of Ca^* binding, is known to induce bilayers to condense increasing the packing density in the gel phase and. thus, to raise the phase transition temperature. Upon addition of Ca2+ to the system the overall pattern of the endotherm was retained, but the turbidity of the solution increased visibly as a result of charge neutralization and thus, liposome aggregation. By depleting the bilayer surface of water, Ca2+ was found to cause a parallel shift of the transition temperatures of the three peaks to higher values. In particular, an increase of the transition temperatures of 5" C was observed for the two peaks at lower temperature whereas only a 3 ' C temperature increase was observed for the DMPC-rich domain. This confirms the presence domain structures in the merged vesicles. It has been clear for many years from phase diagrams of simple lipid mixtures that lipids with different head groups or acyl chains mix non-randomily and form clusters (115,116); this has been recently confirmed by Ca2+ binding measurements (117). Taking into account the strong mismatch of the cross-areas of the headgroup and of the lipid portion of DM-CCK, its non-ideal mixing with DMPC, i.e. formation of differently enriched DM-CCK domains, becomes highly favored. The fact that a similar phenomenon was not observed for the lipo-gastrin, has to be attributed to a different structure of its peptide portion, i.e. to its protrusion into the bulk water without interference with the interdigitation of its lipid portion with the acceptor bilayer. A conformational study was performed on the lipo-peptides inserted into lipid bilayers by CD in order to prove this working assumption (104,107). To assure quantitative transfer of the lipo-peptides to the DMPC SUVs and thus, to operate in an homogeneous population of merged vesicles, dye leakage experiments were performed on DM-gastrin vesicles with entrapped carboxyfluorescein by adding increasing amounts of DPPC vesicles (107). At a lipo-gastrin/DPPC molar ratio of 1:50 quantitative dissolution of the DM-gastrin vesicles has occurred. As the lipopeptide transfer to DPPC vesicles was found to proceed at significantly lower rates than in the case of DMPC, this ratio as well as incubation above the phase transition temperature of DMPC were used to prepare the samples for the CD measurements. As shown in fig. 15, the overall CD pattern of DM-gastrin inserted into DMPC vesicles excludes the presence of ordered conformation at significant extents. By comparing the spectrum of DM-gastrin in DMPC vesicles with that of DM-gastrin (fig. 12) the negative shoulder around 215 nm is reduced to a weak shoulder. This spectrum reminds that of the parent gastrin molecule in aqueous buffer, thus suggesting that the gastrin moiety is exposed to the aqueous environment in mainly random coiled structure with minimal interactions with
850
Fig, 15. CD spectra ofDM-gastrin in 50 mM Tris adjusted to pH 7.0 with H3PO4 at a peptide/DMPC molar ratio of 1:100.
o
196
205
215
225
235
245
Xiran)
Fig. 16. CD spectra of a 1:25 molar mixture of vortexed DM-CCK vesicles and DMPC in 5 mM phosphate buffer, 100 mM NaCl (pH 7.0) after 12 h incubation at 30'C.
851 the phosphatidylcholine headgroups. Conversely, the CD spectrum of DM-CCK transferred to DMPC bilayers (fig. 16) is different from that of the CCK peptide headgroups in DM-CCK vesicles (see fig. 13). The red-shift of the negative maximum to 214 nm and the crossover point at 202 nm is consistent with peptide-peptide interactions in p-sheet type aggregates. This would fully agree with the formation of CCK-rich domains in which the CCKpeptide portion retains its chain reversal with insertion of the C-terminus into the bilayer structure and thus, the tryptophan residue into more hydrophobic compartments of the bilayer as confirmed by the additional blue shift of the fluorescence emission maximum. The chain reversal would lead to full exposure of the highly charged peptide portion Arg-Asp-Tyr(S04H) as head group to the water phase. This picture of the CCK-headgroup would also agree with that obtained by MD simulations in the CCl4/water box.
4 . Ca^-*- Binding of Gastrin and CCK in Membrane Environments
The different efi'ect of Ca2+ ions on the thermotropic behaviour of the CCK- and DMPC-rich domains in the fused DM-CCK/DMPC vesicles (see section 3.3) suggests a higher Ca2+ affinity of the CCK head groups than that of the phosphatidylcholine groups. Several peptide hormones have been shown to exhibit high Ca2+ affinity in membrane-mimetic conditions (118-122). This led to propose the Ca^+Zpeptide complexes as the bioactive states of the hormones (123). Previous findings that gastrin is capable of complexing up to three Ca2+ ions in TFE as monitored by CD changes in the near and far uv (118) led us to investigate into more details Ca^"^ binding to gastrin and CCK peptides in various media as well as their ionophoretic activities (124, 125). The affinity of peptides for cadcium can be determined by monitoring possible conformational changes induced by complexation of the metal ion or by measuring the energy transfer phosphorescence of Tb^"*". Lantanide ions are known to replace Ca2+ without causing structural modifications in proteins as both metal ions exhibit a strong propensity for oxygen donor groups, very similar ionic radii, lack of directionality in binding donor groups and an apparent variability in the coordination number (126). As shown in fig. 17, addition of Tb3+ to CCK leads to a decrease of the fluorescence emission maximum of the tryptophan residue with concomitant increase of the Tb^"*" phosphorescence which allows for titration of the peptide metal ion binding sites under various conditions. The results of these experiments are summarized in table 5. For [Thr,Nlel-CCK-9 neither Ca2+ nor Tb3+ binding could be detected in aqueous solution, whereas for gastrin a biphasic titration curve was obtained which
852
0.00
350
Fig. 17.
450
550
Fluorescence emission spectra ofDM-CCK (6 juM, 10 mM MOPS, pH 7,0, 30° C] in absence (curve o) and presence (curve b) ofTb^'^, Spectra were recorded with 284 nm exitation. Terbium emission was monitored at 549 nrru
exhibited a first plateau at a Tb^+/peptide molar ratio of 2 followed by a continuous increase to a maximum of six metal ions corresponding to the six carboxylate functions present in the gastrin molecule. By measuring CD changes in the near uv a Ca^+Zgastrin ratio of 1.5 was obtained indicating that Ca2+ binding is enhancing the rigidity of one of the two tryptophan side chains. In TFE the CCK-peptide shows affinity for one metal ion, whereas in the case of gastrin conformational changes are indicating a maximum of three and energy transfer phosphorescence a maximum of two metal ions bound to the peptide.
Peptide
[Nlel5]-gastrin-17 1 |Thr,Nlel-CCK-9
lOmM MOPS.
lOmM MOPS, pH 7.0; Ca2+
98% TFE Tb3+
98% TFE
pH 7.0; Tb3+ 2 (max. 6)
1.5
2 (or 3with CD)
3.0
-.-
-.-
1.0
1.0
Ca2+
DM-gastrin (SUVs, 0. IM
2.0
2.0
n.d.
n.d.
NaCl)
1.75
n.d
n.d.
n.d.
DM-CCK (SUVs: O.IM NaCl)
Table, 5. Metal ion binding affinities of gastrin and CCK and their lipo-derivatives as determined by CD (Ca^'^ binding) and luminescence (Tb^'^ binding).
853
B
Fig. 18, Low-energy calcium binding sites of [Nle^^hgastrin'15-17] (A) and IThr,Nle]-CCK'9 (B) as calculated with the GRID programme using the peptide coriformations determined by A/MR analysis in aqueous organic media.
854 As discussed in sections 2.2 and 2.3, for both gastrin (62) and CCK (79) the preferred conformational states in aqueous organic media have been determined by iH-NMR analysis. Using these 3D structures the potential Ca2+ binding sites have been calculated with the GRID program, and the results are shown in fig. 18. Both peptides show a high preference for a C-terminal binding site involving the side chain carboxyl function of the aspartic acid and the carbonyl group of glycine. In the case of gastrin a second low energy metal binding site was detected in the penta-glutamic acid sequence. In fact, gastrin has a preferrence for binding more than one metal ion whereas CCK tends to form a 1:1 complex. As shown in fig. 19, binding of one Ca2+ ion to the CCK peptide leads to a significant enhancement of the dichroic intensities and thus, to an increased content of ordered structure. This observation allows for a modelling of the 3D structure of the [Thr,Nlel-CCK/Ca2+ complex which is reported in fig. 20. Conversely, binding of Ca2+ to gastrin in TFE provokes an opposite effect (118), i.e. a collapse of the ordered conformation, as weU assessed by the changes in the dichroic properties in the far uv. In the aggregated states of the lipo-peptides. i.e. in the corresponding SUVs, the peptide headgroups of the inner layer should not be accessible for Ca^* ions and correspondingly metal ion/peptide molar ratios of 1.5 (or 1):1 for DM-gastrin and 0.5:1 for DM-CCK were theoretically expected. The experimental ratios were for both lipo-peptides significantly higher as reported in table 5. This can be
Fig. 19.
CD-spectra of IThr,Nle]-CCK-9 in aqueus TFE (—), in presence of 1 eq.
855 attributed to the fact that the vesicles are leaking or that transbilayer flip-flop is taking place at high rates. Both processes are possible because of the high fluidity of the vesicles. Additionally, in the case of DM-CCK intermolecular ion complexation involving the aspartic acid residue located in the N-termlnal portion has to occur in order to explain the high molar ratio of 1.75. Intermolecular Ca2+ complexation and thus aggregation of aspartic acid-containing peptides in lipid bilayers has been reported (127). Binding of Ca2+ to DM-CCK is again stabilizing the conformation of the peptide head group, whereas an opposite eflfect was observed in the case of DM-gastrin. Finally, to simulate the conditions of peptide hormones in a cell membrane-bound state, the effect of Ca2+ on the merged lipopeptide/DMPC vesicles was examined. At 1 to 2 mM Ca2+ concentration the CD patterns of both lipo-peptides were very similar to those of the peptide headgroups in the pure lipo-peptide vesicles, but the intensities of the dichroic bands were lower. Lower dichroic intensities have been attributed, e.g. in the case of a-helices, to the formation of bundles (128). Similarly, the lower intensities revealed in the present case can reasonably be attributed to a clustering of the lipo-peptides into domains as favored by intermolecular Ca^* complexation. These findings, besides confirming the experimental results of the microcalorimetric measurements, at least in the case of CCK, could be of biological relevance
Fig. 20,
Model of the energetically most favored Ca^^/IThr,Nlel'CCK'9 complex using for calcium a van der Waals radius of 1.95 A.
856 in terms of facilitating an accumulation of peptide (neuro)hormones at the cell membrane surface. Recent studies of Ananthanarayanan (122) have shown that various peptide hormones are capable of inducing Ca2+ influxes into phosphatidylcholine vesicles. The observed affinity of gastrin and CCK as well as of their lipoderivatives for Ca^^ in membrane mimicking environments led u s to examine rates of Ca2+ influxes induced by these hormones. Calcium ion fluxes mediated by a variety of channels and ionophores into liposomes and cells have been studied by loading the vesicles with indicator dyes like arsenazo III (129) or quin-2 (130). The significantly higher calcium affinity of the fluorescence indicator fura-2 (131) was a major advancement in the detection of Ca2+ concentrations in small cells and liposomes (132-134).
- |
1
j -
-J
I
I
r 3
I L Ca**
20
Fig. 21.
I
L_
40 60 Incubation time (mln)
80
Dependence of the Jluorescence ratio on incubation time after the addition of lThr,Nle]-CCK-9 (0), [Nle^^l-gastnn (A), DM-CCK (x ) and DM-gastrin (O). The buffer is used as blank (+).
Although [Nlel5]-gastrin-17 and [Thr,Nlel-CCK-9 are capable of binding calcium ions in TFE, they were unable to induce Ca2+ influxes into DMPC vesicles as shown in fig. 2 1 . This fully agrees with the findings from microcalorimetric and CD measurements which excluded major interactions of these peptides with the DMPC bilayer. Conversely, the induced lipid interaction of their lipophilic derivatives DM-gastrin and DM-CCK provoked ion influxes with full equilibration
857 of the system after more than Ih. The rate is similar to that observed for other peptides (122), but significantly lower than that of ionophores. It has recently been reported that interaction of peptides with lipid bilayers leads to strong perturbation of the fatty acid chain packing which markedly increases the rates of transbilayer flip-flop of lipid monomers (135). Therefore, the observed ion influxes should derive mainly from relatively fast flip-flop processes of the lipopeptides with bound csilcium ions, even if calcium ions are known to increase the packing of lipid bilayers. Both the rate of ion influxes and the most probable mechanism exclude that the observed effect is of physiological relevsmce. Moreover, such ionophoretic activity of bioactive peptides not mediated and restricted by the receptor recognition would represent an unspecific activity, irrespective of the cell encountered by the peptide hormone on its endocrine pathway. Thus, hormone-stimulated Ca2+ release from membrane pools has to occur mainly as a result of increased cytosolic inositol(tris)phosphate concentrations as induced by the specific hormone receptor interaction at the target cell (136,137).
5. Biological Properties of Lipo-Gastrin and Lipo-CCK
CCK and gastrin peptides are known to exert their physiological function via two receptor subtypes, i.e. the CCK-A receptor, mainly located in the pancreas and selective for the sulfated forms of CCK peptides, and the CCK-B receptor, widely distributed in the central nervous system and in the gastrointestinal tract, which recognizes CCK and gastrin peptides with similar affinities independently of their state of sulfation. These two receptor subt3T>es have been cloned and expressed in COS-7 cells (138,139). A comparison of the sequences of the CCK-A and CCK-B receptors showed 48% identity as expected for receptors within the same family (139). In analogy to other members of the GPCR superfamily seven putative transmembrane segments were identified which are connected by extra- and intracellular loops. Cysteines in the first and second extracellular domains are conserved in both receptors and may be involved in disulfide bridges as required for the stabilization of these domains. Moreover three potential asparagine-linked glycosylation sites are identified in the N-terminal domain. Great efforts are presently paid to identify the CCK and gastrin binding sites of these two receptors. The biological functions of the lipo-gastrin and lipo-CCK peptides were analyzed on these two receptors of known sequence using rat pancreatic acinar cells and the tumoral rat pancreatic acinar cell line AR42J. The CCK-A receptor has been thoroughly characterized in the rat pancreatic acinar cells (140-142) and in its
858 recombinant form expressed on COS-7 cells (139.142). Most of the receptors (80%) exist in the veiy-low-affmity state and only a small percentage (1%) in the high-affinity-state (142). The ability of the receptor to exist in three different states is an intrinsic property of this CCK-A receptor. The CCK-B receptor has been well characterized pharmacologically in gastric mucosal cells (143-146), in the recombinant form of transfected COS-7 cells (139) as well as in the AR42J cell line (147). In contrast to the normal rat pancreas the AR42J cells contain a majority (at least 80%) of high-affmity CCK-B/gastrin receptors and a minority of CCK-A receptors. Binding of DM-gastrin to CCK-B receptors on AR42J cells was determined in competition assays using as radioligand 125i-BH-[NlelS]-gastrin-[2-17] and compared to the parent [Nle 15]-gastrin-17 (51). As shown in fig. 22, under standard conditions the lipo-gastrin exhibits a 7-fold lower binding affinity than the immodified gastrin. Binding of DM-CCK to isolated rat acinar cells using l25i_ BH-[Thr,Nle]-CCK-9 as radioligand (fig. 23) led to very similar results, i.e. 5-fold lower affinity (84). Although N-terminal modification of gastrin and CCK should not affect the receptor recognition and binding capability, incorporation of the highly lipophilic di-fattyacyl moity leads to self-aggregations of the lipo-peptides in aqueous solutions at ionic strengths similar to those of the binding assay media (51,104). The mode of presentation of the lipo-hormones to the membranebound receptors is, therefore, completely different from that of the parent peptides. In fact, even if the characterization of the aggregational states of DMgastrin and DM-CCK has been performed at dilutions up to lO'^M, the stability of
100
-11
-10
-9
concentration. lQg(M)
Fig. 22. Receptor hinding affinities of [Nle^^J-gastrin-l? (•) and DM-gastrin (O) using AR4-2J membrane preparations and ^^^I-BH'[Nle^^]-gastrin-[2'17] as radioligand.
859 the vesicles suggests the persistence of aggregated states in form of vesicles or micelles even at the dilutions of the binding assays (lO''^ - I O - ^ ^ M ) . Radioligand displacement by DM-gastrin and DM-CCK occurs in parallel mode to that obtained with the parent gastrin and CCK peptide, respectively. This indicates that the binding of the lipo-peptides is directly proportional to their concentrations, irrespective of their aggregational state; moreover, even the binding mode is apparently not affected by the modification. Within the incubation time of the binding assay the transfer of the lipo-peptides to the cell membrane should be quantitative according to the results of the model experiments discussed in section 3.3. Therefore the higher IC50 values obtained for the lipo-peptides can either derive from an intrinsic lower affmity for the receptor binding sites or from the induced lipid interaction and thus, restricted two-dimensional migration to the receptor, since an escape of the lipo-peptides into the extracellular water phase is energetically highly unfavored. Support for the second hypothesis derives from following observations. By strengthening the interdigitation of the lipid tail with the membrane bilayer and thus, lowering the rate of diffusion, e.g. with a di-palmitoyl tail, the IC50 value is increased (51). As shown in the model experiments discussed in section 3.3, the interaction of DMCCK with the membrane bilayer leads to strong perturbations and thus, the
-12
Fig. 23.
-11
-10 -9 concentxatlon,
-8 log(M)
Receptor binding affinities of IThr,Nle]-CCK-9 (O) and DM-CCK (•) after 45 min incubation with isolated rat pancreatic acini using ^^^I-BHIThr,Nle]-CCK-9 as tracer.
860 interdigitation of its lipid tail with the bilayer should be weaker than in the case of DM-gastrin. In fact, the receptor affinitiy of DM-CCK is higher than that of DMgastrin. Finally, the estimated rates of two-dimensional diffusion of lipids in bilayers are about one order of magnitude lower than those of a threedimensional diffusion of peptides of this size in water (84); this compares well with the experimental values. Correspondingly, longer incubation periods of the lipo-peptides with the cells should anneal the observed differences. The data of fig. 24 show that this is really the case as almost identical binding affinities were obtained for DM-CCK and its parent peptide confirming that induced lipid interaction is lowering significantly the association rate. Analysis of the data of fig. 24 showed that at binding equilibriimi both high and low affinity receptors were occupied by the lipo-CCK as it is known to occur for CCK peptides (148-151). There was, however, a significant 3-fold lower affmity of DM-CCK as compared to [Thr,Nle]-CCK-9 for the low affinity binding sites. Regarding functional binding, i.e. final biological response, a good con: i.-^ion between binding and potency was observed for DM-gastrin with a siigi/ J ^ enhanced efficacy of the lipo-derivative compared to that of the parent gastrin hormone (107). In the case of DM-CCK again the CCK-typical up and down strokes of the dose-response curve related to amylase hypersecretion were obtained. Although the lipo-CCK showed the same efficacy as [Thr,Nle]-CCK-9 its potency was 100 times lower. Conversely, the potency of DM-CCK in stimulating increase of the IP3 concentration was reduced only by a factor 4 compared to the [Thr,Nlel-CCK-9, but again the efficacy of both ligands were identical.
-11
Fig. 24.
-10 -9 coxicentratlozi.
-8 log(M}
Receptor binding affinities of IThr,Nle]-CCK'9 (O) and DM-CCK (•) after 3 h incubation with isolated rat pancreatic acini using ^^^I-BH-VThr^Nle]CCK'9 as tracer.
861 CCK is generally believed to stimulate secretion of digestive enzymes from rat pancreatic acini by activating phospholipase C which hydrolj^es the membrane lipid phosphatidylinositol biphosphate with release of inositol(tris)phosphate (IP3) and diacylglycerol, increase of C5^osolic calcitim and activation of kinase C (148,150,152-154). These CCK actions are mediated by three states of the CCK receptor. Of these three CCK receptor states the high- and low-affinity states are involved in the competitive binding experiments as the very-low-affinity state cannot be identified by binding with the CCK-radioligand (141). The low-affinity state mediates the upstroke of the amylase dose-response curve and the veiy-lowafflnity state its downstroke. The biological data obtained with DM-CCK and [Thr,Nle]-CCK-9 indicate a similar amplification factor between binding to the functional low-affinity receptors and IP3 production for both peptides suggesting that the coupling of the G-protein (Gq) between receptor and phospholipase C was equivalent. However, the amplification factor between IP3 formation and amylase secretion was 27 times lower with DM-CCK than with IThr,Nle]-CCK-9. Therefore the difference in amplification between IP3 formation and the final biological response was responsible for the difference in amplification between binding and amylase secretion (a 35-fold factor between DM-CCK and unmodified CCK peptide). Stimulus secretion coupling in pancreatic acini involves several G-proteins (152). It is conceivable that DM-CCK occupies low-affinity receptors in such a way that they are poorly coupled to G-proteins different from the Gq, but contribute to the normal response. The experimental data concerning receptor binding affinities and final biological responses obtained with the lipo-derivatized gastrin and CCK aUow for interesting conclusions to be drawn. Despite the tight anchorage of the lipo-peptides to the membrane bilayers, binding to the receptor is occurring. This binding is timedependent and equilibration of the systems is obtained after longer periods of time than with the parent peptide hormones. This phenomenon can be correlated to the lower rates of two-dimensional migration of the lipo-peptides in cell membrane bilayers than those of the underivatized peptides. These are free to diffuse in three-dimensions to the receptor in the extracellular space or to float on the membrane surface with more or less pronounced interactions with the lipid headgroups or with more hydrophobic compartments. The generally accepted model of the spatial structure of the receptor, however, raises immediately the question of how a membrane-bound ligand can find its way to the receptor binding site. As an escape of the lipo-peptides into the extracellular water phase is energetically highly unfavored, the binding site (or sites) are reached by lateral penetration at the lipid/water interphase where the spectroscopic measurements in model systems are locating the lipo-CCK and
862 lipo-gastrin in a more or less conformationally ordered form. This would imply a remarkable flexibility and mobility of the boundary domains of the receptor, particularly, in view of the large size of the peptide ligand. A lateral penetration of the entire lipo-peptide molecule, i.e. including the lipid tail, is difficult to rationalize as the helix boundle of the receptor represents a tight assembly which precludes diffusion of membrane lipids into its core structure in order to maintain its three-dimensional assembly Consequently, it should therefore preclude also a penetration of the lipo-tail of DM-gastrin and DM-CCK, and the lipo-peptides should approach the receptor with the tail inserted into the lipid bilayer and then protrude into the binding cleft across the extracellular loops.
Peptide
PepUde:
Fig, 25.
^
^
,
Arg-Asp-TyT(S03H)-Thr-Gfy-Trp-Nle-Asp-Phe-NH2
rrhr.Nlel-CCK-9
Gly-Pro-Trp-Lcu-Glu-{Glu)4-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH2
INle ^ ^l-gastrln-17
Schematic representation of DM-CCK, showing the spacer (bold) and the lipo moiety inserted in the membrane.
In the case of the extensively studied CCK-A receptor, besides heterogeneity in its affinity states (139,142), heterogeneity in binding sites regarding agonists and antagonists have been detected (156,157). This fact could suggest a ligand binding process of a cascade-like dynamic for an optimal signal transduction, i.e. G-protein coupling. If these dynamics are hindered by the pivot-like restriction of the lipid tail, the binding can still occur at full extents, but its functionality may be impaired. In the present case, the different results obtained with DM-gastrin and DM-CCK in the functional binding could then be rationally explained by the different length of the spacer between site of anchorage to the bilayer and
863 bioactive core of the hormones. As shown in fig. 25, the fully active sequence of CCK is spaced from the lipid-tail by a dipeptide, whilst in DM-gastrin a pentapeptide chain is doing this job. This working assumption could give the tools for investigating an important open question of the hormone-mediated signal transduction, i.e. whether binding of the hormone to a first recognition site is followed by a switch to a second functional binding site located in inner compartments of the helix boundle, as proposed for the acetylcholine receptor (16, 159, 160).
6. Peptide and Non-Peptide Antagonists
The wide range of physiological responses which have been attributed to CCK-A and gastrin/CCK-B receptor-mediated hormonal messages have stimulated the search for agents which mimic or block the action of gastrin and CCK. As there were opportunities for drug discovery in the areas of analgesia (CCK-B antagonist), anxiety (CCK-B antagonist), drug dependency (CCK-B antagonist), memory (CCK-A agonist), parkinsonism (CCK-B antagonist) and psychosis (CCKA agonist), high affinity peptide and non-peptide antagonists which are capable of distinguishing between these two receptor subtypes have been developed (for recent reviews see ref. 33, 161-164). The availability of antagonists has offered new tools with which to explore the role of CCK and gastrin in periphery and brain. Among the antagonists only few peptide structures were found with deletion sequences, C-terminal variations of the parent peptide sequences or by backbone modifications. More successful were non-peptide antagonists based on benzodiazepine and peptoid structures (see fig. 26 for selected examples of CCK-A and CCK-B antagonists). High binding potencies and receptor selectivities have been realized with nonpeptide ligands which differ distinctly in their chemical structure from the natural ligands. Moreover, they differ structurally from one another, too. This fact is difficult to rationalize in terms of identical binding sites at the receptor level and of mimicry of the natural ligands in their bioactive conformation. Nevertheless based on the working assumption that agonists and antagonists should exhibit a correspondence among the pharmacophoric groups, attempts have been made to identify the bioactive conformations of gastrin and CCK by conformational energy analysis and selection of the low-energy conformers of acetylated CCK-7 and gastrin-(14-17) by their structural similarity with the two most studied antagonists L-364,718 (CCK-A antagonist) and L-365,260 (CCK-B antagonist) (165). This led to propose for CCK-B agonists an a-helical array in the C-terminus as determined experimentally in the studies discussed in sections 2.2 and 3.2 and a p-bend structure for the CCK-A receptor agonists. In the latter case
864 the p-bend induces the tyrosine and phenylalanine side chains to appoach one another with the tryptophan side chain pointing away from these two aromatic rings. This conformation differs significantly from that discussed in section 2.3 and 3.2. However, in the computed low-energy conformation of acetyl-CCK-7 with methionine-2 replaced by threonine as in the case of the fully active [Thr,Nle]CCK-9 analog, a sharp reverse turn between glycine and tryptophan is found, and the tyrosine and phenylalanine side chains lie far apart from one another in a manner more similar to the structure of the CCK analog in DMSO/water reported in section 2.3 (165). In this case a superimposition with the non-peptide antagonist L-364.718 does not allow to identify correspondence in the array of potential pharmacophoric groups.
CCK-A
CCK-B O
O
HO'
B
K^W^^ D
Fig. 26.
Structures of the CCK-A antagonists L364J18 (A) and lorglumide (B) and of the CCK-B antagonists L365,260 (C) and CI-988 (D).
In view of the most recent results which indicate distinct binding sites for agonists and antagonists on CCK-A and CCK-B receptors (156-158,166), this lack
865 of structural similarity is not surprising. However, at the same time an important method for verification of proposed bioactive conformations is lost. Surprising in this context was the finding that even peptide antagonists like H-Tyr(S03H)-NleGly-D-Trp-Nle-Asp-2-phenylethylester (167) occupy receptor binding sites which could not directly bind agonist ligands (157,158).
Conclusion
Double-tailed lipo-derivatization of the homologous peptide hormones allowed to bypass the problem of a partition of these peptides between the lipid and water phase of model lipid bilayers and of cell membranes, by shifting the equilibrium in great favour of the membrane-bound state. The main drawback of such an approach derives from the predetermined site of lipid anchorage. Therefore, the lipid moieties were grafted to the N-termini of two fully active gastrin and CCK analogs, as this portion of the peptide molecules is known not to be involved in the receptor recognition process. As expected from the resulting chemical constitution of the lipo-peptides this derivatization induces a spontaneous aggregation into (unilamellar) vesicles. By comparing the aggregational properties of the two lipo-peptides, the most striking observation was the pronounced effect of the peptide headgroups. In fact, replacing the sequence Gly-Pro-Trp-Leu(Glu)5-Ala-Tyr of [Nlel5]-gastrin-[2-171 with Arg-Asp-Tyr(S03H)-Thr of [NlcThrJCCK-9, but retaining the identical C-terminus Gly-Trp-Nle-Asp-Phe-NH2 substantial differences in the packing of the fatty acid chains and in the display of the peptide headgroups were revealed. Despite the sequence homology of the lipo-peptides a remarkably differentiated behaviour was also revealed in their mode of insertion and interaction with model bilayers and of their display at the water/lipid interphase. This would indicate that already at the level of the collisional events with the target cell membrane sequence-dependent characteristic properties are initiating the differentiation process for the selective receptor recognition process. Most striking were the findings that the CCK headgroup interacts with the lipid bilayer in an amphipathic helical array which strongly reminds that determined by NMR in aqueous DMSO. Thereby a pronounced tendency to cluster into domains was detected which was further enhanced in presence of calcium ions. This phenomen could be of physiological relevance as it should facilitate accumulation of the CCK hormone on the cell membrane with a preorientation and prefolding into bioactive conformations. Conversely, for the gastrin peptide even upon an induced lipid interaction a preferred ordered structure could not be detected. In all experimental models the
866 gastrin headgroup is exposed to the bulk water phase in randomly coiled structure, whereby even the presence of calcium is not enhancing the interaction of the peptide head group with the bilayer. All the data suggest that this peptide hormone may not be accumulated on the cell membrane since the zwitterionic lipids used in the model experiments represent the main constituent of natural membranes and acidic lipids present in cell membranes, too, are expected to prevent even more by electrostatic repulsion an interaction of the negatively charged peptide with the membrane. The effect of this repulsion was well assessed by comparing the behaviour of the gastrin in negatively charged micelles as mimicry of acidic cell membrane domains. Correspondingly, the preferred Ushaped conformation of gastrin as determined by NMR in TFE does certainly not represent its structure in membrane environments. The model experiments, however, do not exclude that this hairpin structure determined in aqueous organic media may be assumed by the gastrin peptide in the receptor-bound state. The very similar conformations of gastrin and CCK in the active site portion of the molecule correlates well with the biochemical properties in terms of identical recognition of both hormones by the CCK-B/gastrin receptors where the negative charges accumulated in the N-terminal, more or less flexible tail of CCK and gastrin may play a secondary role to increase the binding afBnity. For this purpose a specific conformation of the N-terminus seems not to be required, since in the case of the equally potent CCK peptide no preferred ordered structure could be detected in this portion of the molecule. The a-helical structure in the Nterminal portion of gastrin could therefore result from the TFE which is known t j be a strong a-helix inducing solvent (80,81). Then the increasing pc v 'ies of gastrin peptides in function of chain length would result from add lioixal electrostatic ligand/receptor interactions and the parallelism observed for the onset of a-helical conformation could be a fortuitous coincidence. On the other hand, the fully flexible N-terminus of CCK with an exact location of the tyrosineO-sulfate moiety plays a crucial role for the selective recognition by the CCK-A receptor. The whole body of data of these series of studies clearly reveals a surprisingly differentiated behaviour of the two homologous peptide hormones, with CCK showing affinity for membrane bilayers and gastrin not. However, independently of wether the peptide hormones are accumulated on the cell surface in their target cells, with the lipo-peptides it was definitely confirmed that a membranebound pathway in the mechanism of the hormone-receptor binding process as proposed by Schwyzer (23,24) is indeed possible. Thereby with the helical array of the CCK peptide parallel to the water/lipid interphase, the aromatic side chain ring of the phenylalanine residue is heading the structured molecule and could play a decisive role in the lateral penetration of the receptor assembly.
867 References 1.
2. 3. 4. 5. 6. 7. 8. 9.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
24. 25. 26. 27.
Lemieux, R.H., Delbaere, L.T.J., Beverbeek, H., Spohr, U. (1991) in: HostGuest Molecular Interaction from Chemistry to Biology, Ciba Foundation SjTiiposium, pp. 231-248, Wiley, Chichester. Fischer, E. (1894) Chem Bar. 27, 2985-2993. Richards, F.M. and Vithayathil, P.J. (1959) J. Biol Chem 234, 1459-1465. Tamburro, A.M., Scatturin, A., Rocchi, R., Marchiori, F., Borin, G. and Scoffone, E. (1968) FEBSLett. 1, 298-300. Rocchi, R. Borin, G., Marchiori, F., Moroder, L., Peggion, E., Scoffone, E., Crescenzi, V. and Quadrifoglio, F. (1972) Biochemistry 11, 50-57. Kartha, G., BeUo, J. and Harker, D. (1967) Nature 213, 862-865. Wyckoff, H.W., Hardmann, K.D., Allewell, N.M., Inagami, T., Johnson, I.N. and Richards, F.M. (1967) J. Biol Chem 242, 3984-3988. Moroder, L., Borin, G., Marchiori, F., Rocchi, R. and Scoffone, E. (1973) in: Peptides 1971 (Nesvadba, H., ed.) pp. 367-372, North-Holland, Amsterdam. Scoffone, E., Marchiori, F., Moroder, L., Rocchi, R. and Borin, G. (1973) in: Medicinal Chemistry III, Special Contributions (Pratesi, P., ed.) pp. 83-104, Butterworth, London. Garcia, K.C., Ronco, P.M., Verroust, P.J.. Brunger, A.T. and Amzel, L.M. (1992) Science 257, 502-507. Stem, L.J., Brown, J.H., Jardetzky T.S., Gorgia, J . C , Urban, R.G., Strominger, J.L. and Wiley, D.C. (1994) Nature 368, 215-221. Brown, J.E. and Klee, W.A. (1971) Biochemistry 10, 470-476. Bierzynski, A., Kim, P.S. and Baldwin, R.L. (1982) Proc, Natl Acad. Set USA 79, 2470-2474. Kim, P.S. and Baldwin, R.L. (1982) Nature 307. 329-334. OUveira, L., Paiva, A.C.M., Sander, C. and Vriend, G. (1994) TIPS 15, 170172. Wess, J. (1993) TIPS 14, 308-313. Hargrave, P.A. (1991) Curr. Opin, Struct Biol 1, 575-581. Burt, S.K., Hutchins, C.W. and Greer, J. (1991) Curr. Opin. Struct. Biol 1, 213-218. MerriU Jr., A.H. and Liotta, D.C. (1991) Curr. Opin. Struct. Biol 1, 516-521. Marsh, D. (1992) Curr. Opin. Struct. Biol 2, 497-502. Glaser, M. (1993) Curr. Opin. Struct. Biol 3, 475-481. Deveaux, P.F. (1993) Curr. Opin. Stmct. Biol 3, 489-494. Schwyzer, R.(1986) in: Natural Products and Biological Activities (Imura, H., Goto, T., Murachi, T. and Nakajima, T., eds.) pp. 197-207, Tokyo Press, Elsevier, Tokyo. Schwyzer, R. (1991) Bvopolymers 31, 785-792. Henderson, R., Baldwin, J.M., Ceska, T.A., Zemlin, F., Beckmann, E. and Downing, K.H. (1990) J. Mol Biol 213, 899-929. Schertler, G.F.X., Villa, C. and Henderson, R. (1993) Nature 362, 770-772. Baldwin, J. B. (1993) EMBOJ. 12, 1693-1703.
868
28. OUviera, L., Paiva, A.C.M. and Vriend, G. (1993) J. Comput. Aided Mol Design 7, 649-658. 29. Hibert, M.F. Trumpp-Kallmeyer,, S., Bruinvels, A. and Hoflack, J. (1991) Mol Pharmacol 40. 8-15. 30. Cronet, P., Sander, C. and Vriend, G. (1993) Prot Eng. 6, 59-64. 31. Dockray, G.L, Dimaline, R., Pauwels, S. & Varro. A. (1980) in: Peptide Hormones as Prohormones - Processing, Biological Activiies, Pharmacology (Martinez, J. ed.) pp. 245-284, Ellis Horword, Chichester. 32. Martinez, J. and Poitier, P. (1986) I7PS, 139-147. 33. Martinez, J. (1990) in: Comprehensive Medicinal Chemistry (Hansch, C , Sammer, P.G. & Taylor, J.B., eds.) Vol. 3, pp. 925-959, Pergamon Press, New York. 34. Gregory, R.A. and Tracy, H.J. (1964) Gut 5, 103-117. 35. Eysselin, V.E., MaxweU, V., Reedy, T., Wunsch, E. and Walsh, J.H. (1984) J. Clin. Invest 73, 1284-1290. 36. Carter, D.C., Taylor, I.L., Elashoff, J. and Grossman, M.I. (1979) Gut 20, 705-708. 37. Moran, T.H., Robinson, P.H., Goldrich, M.S. and McHugh, P.R. (1986) Brain Res. 362, 175-189. 38. Jorpes, J.E., Mutt, V. and Toczko, R. (1964) Acta Chem. Scand. 18, 24082410. 39. Mutt, V. (1980) in: Gastrointestinal Hormones (Glass, J.B.J., ed.) pp. 169221, Raven Press, New York. 40. Tracy, H.J. and Gregory, R.A. (1964) Nature 204, 935-938. 41. Morley, J.S. (1968) Proc. R. Soc. Ser, B 170, 97-111. 42. Jensen, R.T., Lemp, G.F. and Gardner, J.D. (1982) J. Biol Chem. 257, 55545559. 43. Gillessen, D., Trzeciak, A., Muller, R.K.M. and Studer, R. (1979) Int. J. Peptide Protein Res. 13, 130-136. 44. Wunsch, E., Moroder, L., Gillessen, D., Soerensen, U.B. and Bali, J.-P. (1982) Hoppe-Seyler's Z. Physiol Chem. 363, 665-669. 45. Kenner, G.W., Mendive, J.J. and Sheppard, R.C. (1968) J. Chem. Soc (C) 761764. 46. L. Moroder, L. Wilschowitz, M. Gemeiner, W. Gohring, S. Knof, R. Scharf, P. Thamm, J.D. Gardner, T.E. Solomon and E. Wunsch, E. (1981) HoppeSeylefs Z. Physiol Chem. 362, 929-942. 47. Moroder, L., Gohring, W., Nyfeler, R., Scharf, R., Thamm, P. and Wendlberger, G. (1983) Hoppe-SeylefsZ. Physiol Chem. 364, 157-171. 48. Huang, S.C, Yu, D.H., Wank, S.A., Manley, S., Gardner, J.D. and Jensen, R.T. (1989) Peptides 10, 785-789. 49. Wunsch, E., Moroder, L., Gohring, W., Borin, G., Calderan, A. and Bali, J.-P. (1986) FEBSLett. 206, 203-207. 50. Borin, G., Calderan, A., Ruzza, P., Moroder, L., Gohring, W., Bovermann, G. and Wunsch, E. (1987) Biol Chem. Hoppe-Seyler 368, 1363-1373. 51. Romano, R., Musiol, H.-J., Dufresne, M., Weyher, E. & Moroder, L. (1992) Biopolymers 32, 1545-1558.
869 52. Bodansky, M., Martinez J., Walker, M.D. Gardner, J.D., and Mutt, V. (1980) J. Med, CheiTL 23, 82-85. 53. G5hring, W., Moroder, L., Borin, G., Lobbia, A., Bali, J.-P. and Wunsch, E. (1984) Hoppe-Seylefs Z. Physiol 365, 83-94. 54. Peggion, E., Jaeger, E., Knof. S., Moroder, L. and Wunsch, E. (1981) Biopolymers 20, 633-652. 55. Peggion, E., Foffani, M.T., Wunsch, E., Moroder, L., Borin, G., Goodman, M. and Mammi, S. (1985) Biopolymers 24, 647-666. 56. Wu, C.C., Hachimori, A. and Yang, J.T. (1982) Biochemistry 21,4556-4562. 57. Torda, A.E., Baldwin, G.S. and Norton, R.S. (1985) Biochemistry 24, 17201727. 58. Goodman, M. and Listowsky, I. (1962) J. Am, Chem, Soc. 84, 3770-3771. 59. Urry, D:W. (1972) Biochim. Biophys, Acta 265, 115-168. 60. Pham van Chuong, P., Penke, B., de Castiglione, R. and Fromageot, P. (1979) in: Hormonal Receptors in Digestion and Nutrition (Rosselin, P., Fromageot, P. and Bonfils, S., eds.) pp. 33-44, Elsevier, North-Holland, Amsterdam. 61. Mammi, S., Goodman, M., Peggion, E., Foffani, M.T., Moroder, L. and Wunsch, E (1986) Int, J. Pept Protein Res. 27, 145-152. 62. Mammi, S., Mammi, N.J. and Peggion, E. (1988) Biochemistry 27, 13741379. 63. Mammi, S., Foffani, M.T., Peggion, E., Galleyrand, J.C., Bali, J.-P., Simonetti, M.. Gohring, W., Moroder, L. and Wunsch, E. (1989) Bkx:hemistry 28, 7182-7188. 64. Moroder, L., Romano, R., Weyher, E., Svoboda, M. and Christophe, J. (1993) Z. Naturforsch.4Sh. 1419-1430. 65. Moroder, L., Mourier, G., Dufresne, M., Bovermann, G., Gohring, W., Gemeiner, M. and Wunsch, E. (1985) BioL Chem. Hoppe-Seyler 368, 839848. 66. Mamimi, S., Mammi, N.J., Foffani, M.T., Peggion, E., Moroder, L. and Wunsch, E. (1987) Biopolymers 26, Sl-SlO. 67. Mammi, S. and Peggion, E. (1990) Biochemistry 29, 5265-5269. 68. Moroder, L., Bali, J.-P. and Kobayashi, Y. (1991) Biopolymers 3 1 , 595-604. 69. Moroder, L., Musiol, H.-J., Kocher, K., BaU, J.-P., Schneider, C.H., Guba, W., MuUer, G., Mierke, D. and Kessler, H. (1993) Eur. J. Biochem, 212, 325-333. 70. Moroder, L. and Wunsch, E. (1993) in: Gastrin (Walsh, J., ed.) pp. 187-194, Raven Press. New York. 71. Durieux, C., Belleney, J., Lallemand, J.Y., Roques, B.P. and Foumie-Zaluski, M.C. (1983) Biochem. Bvophys. Res. Commun. 114, 705-712. 72. Koizuka, I., Watari, H., Yanaihara, N., Nishina, Y., Shiga K. and Nagayama, K. (1984) Biomed. Res. Suppl 5, 161-168. 73. Foumie-Zaluski, M.C., Durieux, C., Lux, B., Belleney, J., Pham, P., Gerard D. and Roques, B.P. (1985) Biopolymers 24, 1663-1681. 74. Foumie-Zaluski, M.C., Belleney, J., Lux, B., Durieux, C., Gerard, D., Gacel, G., Maigret, B. and Roques, B.P. (1986) Biochemistry 25, 3778-3787. 75. Loomis, R.E., Lee, P.O. and Tseng, C.C. (1987) Biochem. Bvophys. Acta 911, 168-179.
870
76. Pattou. D., Maigret B.. Foumie-Zaluski. M.C. and Roques. B.P. (1991) Int. J. Peptide Protein Res. 37, 440-450. 77. Fang. S., Nikiforovich. G.V., Knapp, R.J., Jiao, D., Yamamura, H.I. and Hruby, V.J. (1992) in: Peptides, Chemistry and Biology (Smith, J.A. and Rivier, J.E. (eds.) pp. 142-143, ESCOM, Leiden. 78. Moroder, L., Weyher, E., D'Ursi, A., Bcone, D. and Temussi, P.A. (1992) ReguL Peptides 40, 213. 79. Moroder, L., D'Ursi. A., Bcone. D.. Amodeo, P. and Temussi, P.A. (1993) Biochem. Biophys. Res. Commun. 190, 741-746. 80. Jackson. M. and Mantsch. H.H. (1992) Biochim. Biophys. Acta 1118. 139143. 81. Sonnichsen, F.D.. van Eyk. J.E., Hodges. R.S., Sykes. B.D. (1992) Biochemistry 31, 8790-8798. 82. Moroder. L., Papini. A., Siglmuller, G.. Kocher. K.. Dorrer, E. and Schneider, C.H. (1992) Biol Chem. Hoppe-Seyler 373, 315-321. 83. Moroder, L. and Temussi, P.A. (1993) in: Peptides 1992 (Schneider, C.H. and Eberle, A.N., eds.) pp. 813-814, ESCOM, Leiden. 84. Moroder, L.. Romano. R.. Guba. W.. Mierke. D.F., Kessler, H.,Delporte. C , Winand, J. and Christophe, J. (1993) Biochemistry 32, 13551-13559. 85. Schwyzer, R., Kimura, S. and Erne, D. (1992) in: Peptides, Chemistry cmd Biology (Smith, J.A. and Rivier, J.E., eds.) pp. 168-170, ESCOM, Leiden. 86. Surewicz, W.K. & Epand, R.M. (1984) Biochemistry 23, 6072-6077. 87. Surewicz, W.K. & Epand, R:M: (1985) Biochemistry 24, 3135-3144. 88. Epand, R.M., Surewicz, W.K. & Yeagle, P. (1988) Chem. Phys. Lipids 49, 105110. 89. Brown, A.g. (1991) in: Nerve Cells and Neuronal Systems. Introduction in Newoscience (Brown, a.G., ed.) pp. 61-71, Springer Verlag. London. 90. Crawley, J.N. (1991) Trervds Pharmacol Set 12, 232-236. 91. Moroder, L. and Musiol, H.-J.(1990) in: Peptides, Chemistry, Structure and Biology (Rivier, J.E. and Marshall, G.R., eds.) pp. 811-812, ESCOM. Leiden. 92. Boyanov. A.I., Tenchov. B.G.. Koynova, R.D. and Koumanov, K.S. (1983) Biochim. Biophys. Acta 732, 711-713. 93. Eklund, K.K., Virtanen, J.A. and Kinnunen, P.K.J. (1984) Biochim. Biophys. Acta 793, 310-312. 94. Tenchov, E.G., Boyanov, A.I. and Koynova, R.D. (1984) Biochemistry 23. 3553-3558. 95. Moroder, L.. Musiol, J.-H. and Siglmuller, G. (1990) Synthesis 889-892. 96. Keller, O. and Rudinger, J. (1975) Helv. Chinh Acta 58, 531-541. 97. Wunsch, E., Moroder, L., Nyfeler, R., Kalbacher, H. and Gemeiner, M. (1985) Biol Chem. Hoppe-Seyler 366, 53-61. 98. von Grunigen, R., Siglmuller, G.. Papini, A., Kocher, K., Traving, B., G6hring, W. and Moroder, L. (1991) Biol Chem. Hoppe-Seyler 372, 163-172. 99. Papini. A., Rudolph, S., Siglmuller, G.. Musiol, H.-J., Gohring, W. and Moroder, L. (1992) Int. J. Peptide Protein Res. 39, 348-355. 100. Moroder. L.. Papini. A.. Siglmuller, G.. Kocher, K., Dorrer, E. and Schneider, C.H. (1992) Biol Chem Hoppe-Seyler 373, 315-321.
871 lOl.Gemeiner, M., Leidinger, E., Muller, I. and Moroder, L. (1992) Biol Cherru Hoppe-Seyler373. 1085-1094. 102.Winand, J., Poloczek, P., Delporte, C , Moroder, L., Svoboda, M. and Christophe. J. (1991) Biochim. Biophys, Acta 1080, 181-190. 103. Nielsen, O. and Buchardt, O. (1991) Synthesis 819-821. 104. Romano, R., Bayerl,, T.M. and Moroder, L. (1993) BiochiirL Biophys. Acta 1151, 111-119. 105. Bayerl, T.M., Werner, G.D. and Sackmann, E. (1989) Biochinh Biophys. Acta 984, 214-224. 106. Guba, W. and Kessler, H. (1994) J. Phys, Chem. 98, 23-27. 107. Romano, R., Dufresne, M., Frost, M.-C, Bali, J.-P., Bayerl, T.M. and Moroder, L. (1993) BiochiirL Biophys. Acta 1145, 235-242. 108. McLean, L.R. and Phillips, M.C. (1981) Biochemistry 20. 2893-2900. 109. Nichols, J.W. and Pagano, R.E. (1981) Biochemisiry 20, 2783-2789. llO.Thilo, L. (1977) Biochim. Biophys. Acta 469, 326-334. lll.Papahadjopoulos, D., Hui, S., Vail, W.J. and Poste, G. (1976) Biochim. Biophys. Acta 448, 245-264. 112.De Cuyper, M., Jordan, M. and Dangreau, H. (1983) Biochemistry 22, 415419. 113.Lentz, B.R., Carpenter, T.J. and Alford, D.R. (1987) Biochemistry 26, 53895397. 114. Martin, F.J. and MacDonald, R.C. (1976) Biochemistry 15, 321-327. 115.Mabrey, S. and Sturtevant, J.M. (1976) Proc. Natl. Acad. Set USA 73, 38623866. 116.Tenchov, B.G. (1985) Prog. Surf. Membr. Set 20, 273-340. 117. Huang, J., Swanson, J.E., Dibble, A.R.G., Hinderliter, A.K. and Feigenson, G.W. (1993) Biophys. J. 64, 413-425. 118.Peggion, E., Mammi, S., Palumbo, M., Moroder, L. and Wunsch, E. (1983) Biopolymers 22, 2443-2457. 119.Ananthanarayanan, V.S. and Orlicky, S. (1992) Biopolymers 32, 1765-1773. 120. Brimble, K.S. and Ananthanarayanan, V.S. (1992) Biochim. Biophys. Acta 1105,319-327. 121. Brimble, K.S. and Ananthanarayanan, V.S. (1993) Biochemistry 32, 16321640. 122. Ananthanarayanan, V.S. (1994) in: Peptides, Design, Synthesis and Biological Activity (Basava, C and Anantharamaiah, G.M., eds) pp. 223-234, Birkhauser Boston. 123. Ananthanarayanan, V.S. (1991) Biochem Cell Biol 69, 93-95. 124. Moroder, L., Lutz, J., Romano, R., Grams, F., Golbik, R. and Weyher, E. in: Peptides 1994 (Maia, H.S., ed.) ESCOM, Leiden, in press 125. Moroder, L. and Romano, R. (1994) Pure andAppl Chem 66, 2111-2114. 126. Brittain, H.G., Richardson, F.S. and Martin, R.B. (1976) J. Am Chem Soc. 98, 8255-8260. 127.0toda, K., Kimura, S. and Imanishi, Y. (1993) Bull Chem Soc. Jpn. 66, 1466-1471.
872 128.Woolley, G.A., Epand, R.M., Kerr, J.D., Sanson, M.S.P. and Wallace. B.A. (1994) Biochemistry 33, 6850-6858. 129.Blau, L., Stem, R.B. and Bittman, R. (1984) Biochinh Biophys. Acta 778, 219-223. 130.Ohki, K., Shunjl, N., Sagami, M. and Nozawa, Y. (1986) Chem. Phys. Lipids 39, 237-249. 131. Grienkiewicz, G., Poenie, M. and Tsien. R.Y. (1985) J. Biol Cheiru 260, 34403450. 132.Tsien, R.Y., Rink, T.J. and Poenie, M. (1985) CeU CalciumG, 145-157. 133. Blau, L. and Weissmann, G. (1988) Biochemistry 27, 5661-5666. 134.Berendes, R., Burger, A., Voges. D., Demanche, P. and Huber, R. (1993) FEBSLett. 317, 131-134. 135.Fattal, E., Nir, S., Parente, R.A., Szoka, F.C.Jr. (1994) Biochemistry 33, 6721-6731. 136. Irvine, R.F. (1992) FASEBJ. 6, 3085-3091. 137. Irvine, R.F. (1992) Curr. Opin. Cell Biol 4, 212-219. 138. Wank, S.A., Harkins, R., Jensen, R.T., Shapira, H., de Weerth, A. and Slattery, T. (1992) Proc. Natl Acad. Set USA 89, 3125-3129. 139. Wank, S.A., Pisegna, J.R. and de Weerth, A. (1992) Proc. Natl Acad. Set USA 89, 8691-8695. 140. Steigerwalt, R.W. and Williams, J.A. (1981) Endocrinol 109, 1746-1753. 141.Vinayek, R., Patto, R.J., Menozzi, D., Gregory, J., Mrozinski, J.E., Jensen, R.T. and Gardner, J.D. (1993) Biochim. Biophys. Acta 1176, 183-191. 142.Talkad, V.D., Fortune, K.P., PoUo, D.A., Shah, G.N., Wank, S.A. and Gardner, J.D. (1994) Proc. Natl Acad. Set USA 91, 1868-1872. 143.Takeuchi, K., Speir, G.R. and Johnson, L.R. (1979) Am. J. Physiol 237, E284-E294. 144.Takeuchi, K., Speir, G.R. and Johnson, L.R. (1980) Am. J. Physiol 239, G395-G399. 145. Magous, R. and BaU, J.-P. (1982) Eur. J. Pharmacol 82, 47-54. 146.Magous, R., Galle5n:and, J.-C, Leonard, A , Choquet. A. and Bali, J.-P. (1987) in: Gastrin and Cholecystoktntn. Chemistry, Physiology and Pharmacology (Bali, J.-P. and Martinez, J., eds.) pp. 153-158, Elsevier, Amsterdam. 147. Christophe, J. (1994) Am. J. Physiol 266, G963-G971. 148. Sankaran, H., Goldfme, I.D., Deveney, C.W., Wong, K.-Y. and Williams, J.A (1980) J. Biol Chem. 255, 1849-1853. 149.Wank, S.A., Jensen, R.T. and Gardner, J.D. (1988) Am. J. Physiol 255, G106-G112. 150. Menozzi, D., Vinayek, R., Jensen, R.T. and Gardner, J.D. (1991) J. Biol Chem. 266, 10385-10391. 151.Winand, J., Poloczek, P., Delporte, C., Moroder, L., Svoboda, M. and Christophe, J. (1991) Biochim. Biophys. Acta 1080, 181-190. 152. Hootman, S.R. and Williams, J.A. (1987) in: Physiology of the Gastrointestinal Tract (Johnson, L.R., ed.) 2nd ed., pp. 1129-1146, Raven Press, Nev^ York. 153. Gardner. J.D. and Jensen, R.T. (1987) in: Physiology of the Gastrointestinal Tract (Johnson, L.R., ed.) 2nd Ed. pp. 1109-1127, Raven Press, New York.
873
154.Matozaki, T. and WiUiams, J.A. (1989) J. Biol Chem, 264, 14729-14734. 155.Winand, J.. Delporte, C , Poloczek, P., Cantraine, F., Dehaye, J.-P. and Christophe. J. (1992) Second Messengers and Phosphoproteins 13, 173-186. 156.Beinbom, M., Lee, Y.-M., McBride, E.W., Quinn, S.M. and Kopin, A.S. (1993) Nature 362, 348-350. 157. Silvente Poirot, S., Hadjiivanova, C , Escrieut, C , Dufresne, M.. Martinez, J., Vaysse, N. and Fourmy, D. (1993) Ew. J. Biochem. 212, 529-538. 158. Silvente Poirot, S., Escrieut, C , Dufresne, M., Martinez, J., Bouisson, M., Vaysse, N. and Fourmy, D. (1994) Mol Pharmacol 45, 599-607. 159.Trumpp-Kallmeyer, Hoflack, J., Bruinvels, A. and Hibert, M., F. (1992) J. Med, Ghent 35, 3448-3462. 160. Hibert, M., F., Trumpp-Kallmeyer, Hoflack, J . and Bruinvels, A. (1993) TIPS 14, 7-12. 161.Freidinger, R.M. (1989) Med. Res, Rev, 9, 271-290. 162. Bock, M.G. (1991) Drugs of the Future 16, 631-640. 163. Woodruff, G.N. and Hughes, J. (1991) Anna. Rev, Pharmacol Toxicol 31, 469-501. 164.Harro, J., Vasar, E. and Bradwejn, J. (1993) TIPS 14, 244-249. 165.Pincus, M.R., Carty, R.P., Chen, J., Lubowsky, J., Avitable, M., Shah, D., Scheraga, H.A. and Murphy, R.B. (1987) Proc, Natl Acad, Set USA 84, 48214825. 166. Chang, R.S.L., Lotti, V.J., Chen, T.B. and Kunkel, K.A. (1986) Mol Pharmacol 30, 212-217. 167.Lignon, M.F., Galas, M.C., Rodriguez, M., Laur, J., Aumelas, A. and Martinez, J. (1987) J, Biol Chem. 262, 7226-7231.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
875
When Two Steroids are Better than One : The Dimeric Steroid-Pyrazine Marine Alkaloids A. Ganesan INTRODUCTION This chapter reviews the history of a remarkable family of natural products first isolated from the marine tube-inhabiting invertebrate worm Cephalodiscus gilchristi. C. gilchristi, found in the temperate Southern Hemisphere, is often attached to bryozoans and sponges. This tiny worm (- 5 mm in tube colonies) can exist independently of the coenicium (worm tube), and exposure to predators during such moments may have aided the evolution of chemical defence mechanisms. In 1972, Pettit's group collected a sample of C. gilchristi at a depth of approximately 20 m in the Indian Ocean off southeast Africa. In 1974, methanol and aqueous extracts of C. gilchristi were confirmed active in the American National Cancer Institute's primary antitumour assay at that time, the murine lymphocytic leukemia P388 (PS system), with 32-41 % life extension at 25-37.5 mg/kg. Identification of the active components was a slow process hampered by the limited material, and required over ten years effort with recollections of fresh sample. Success was achieved with a 1981 batch of 166 kg (wet weight, including worm tubes) of C. gilchristi, from which PS bioassay-guided solvent partitioning^ (Scheme 1) yielded active dichloromethane and carbon tetrachloride fractions. 166 kg wet weight
CH2Cl2-MeOH extraction
H2O fraction
C H2CI2 fraction
hexane 9:lMeOH-H20 fraction fraction
CCI4 fraction PS active, 42 g
4:1 MeOH-H20 fraction
CH2CI2 fraction PS active, 28 g
3:2 MeOH-H.O fraction
Scheme 1.. Solvent partitioning scheme for the extraction of cephalostatins. The active fractions were further purified by extensive column chromatographic and HPLC separations (for details of the protocol, see next section) to obtain 138.8 mg (8x10'^^ % yield) of a pure compound, cephalostatin 1, mp 326 °C dec. Evidence from TLC staining, NMR, and mass
876 spectrometry indicated a steroidal alkaloid but the complete structural elucidation required X-ray analysis of crystals carefully grown from a pyridine solvate. As with many other compounds of marine origin, the structure of cephalostatin 1 (Figure 1) is unprecedented and has no analogy with terrestrial natural products. The compound is an unsymmetrical steroid dimer linked at the C-2,3 positions (for ease of reference, steroid numbering is used throughout; dimers are numbered from C-1 for the "right" half and C-T for the "left" half) by an aromatic pyrazine ring. Oxygenation of the side chain results in a spiroketal ring system as in many saponin aglycones. Two other noteworthy features are the presence of a C-14,15 alkene in both halves, and the oxidation of the C-18' angular methyl to a hydroxy methyl group.
HO
O1 P
Figure 1. Cephalostatin 1.
ISOLATION OF OTHER CEPHALOSTATINS Cephalostatin 1 has an ED50 of 10"^-10"^ M-g/ml against the PS cell line and is in fact among the most potent compounds ever recorded in the NCI's antitumour screening program. Over the next few years, the Pettit group solved the structures of several other cephalostatins purified from the 1981 batch of C. gilchristi. Compared to the parent compound cephalostatin 1, cephalostatins 2-4^ (Figure 2) are hydroxylated at the C-9' ring juncture of the left half. Furthermore, cephalostatin 3 contains an additional methyl group in the side chain, while in cephalostatin 4 the C-14'-15' alkene is epoxidized. These three cephalostatins have similar PS activity to cephalostatin 1. The isolation"^ of cephalostatins 5 and 6 (Scheme 2) is typical of the careful and lengthy separations performed for these compounds by the Pettit group. In these two cephalostatins, ring C of the left half is aromatized. The PS activity of cephalostatins 5 and 6 is dramatically reduced to an ED50 of lO'"^ |Xg/ml, possibly due to the flattening out of the steroid and the loss of C-D ring stereochemistry. Cephalostatin 6 can be derived from cephalostatin 4 by a plausible sequence of events. First, dehydration at C-9' would generate an enone (a similar enone was later isolated from C gilchristi, see cephalostatin 14, vide infra). The epoxide ring can then undergo nucleophilic ring opening at C14' by participation of the enone to yield a dienone. Finally, a retro-aldol reaction at C-18' would complete the aromatization. Alternatively, the angular hydroxymethyl group may be lost via oxidative fragmentation similar to that observed in the biosynthesis of cholesterol from lanosterol. Cephalostatin 5 is presumably related to the epoxy derivative of cephalostatin 3 by the same reaction sequence.
877 CH2CI2 fraction 28 g, from Scheme 1 Silica gel, 1:1 hexane-EtOAc to 4.5:4.5:1 EtOAc-MeOH-HjO
5.8 g (>10)
9.3 g
«io-')
9.8 g (0.3)
Sephadex LH-60 10:10:1 hexane-CH2Cl2-MeOH 6.53 g (10-') Silica gel, gradient 10:10:0 to 10:10:4 hexane-EtOAc-MeOH
1.15 g in two fractions (10-')
402 mg (~ 10-') Silica gel, gradient 10:10:1 to 10:10:4 hexane-EtOAc-MeOH
212 mg in three fractions
171 mg
(lo-'-io-^)
1. C-18 reverse phase HPLC, 1:1 MeOH-HjO to MeOH 2. Silica normal phase HPLC, 30:70:0 to 30:70:10 hexane-EtOAc-MeOH
127 mg in nine fractions (10-^-10-^)
10 mg (10-") 1.Sephadex LH-20 4:5:1 hexane-CH2Cl2-MeOH 2. HPLC, C-18, 1:1 MeOH-HzO to MeOH 3. Sephadex LH-20 MeOH
Cephalostalin 5, 5.5 mg (4 x 10-^)
Cephalostatin 6, 3.0 mg (2 x 10-')
Scheme 2. Detailed chromatographic separation scheme for cephalostatins 5 and 6. Numbers in brackets refer to PS bioassay activity, in )Lig/ml.
878
*OH
'^
Cephalostatin 2 R = H Cephalostatin 3 R = Me
R
HO
O ^
O^
OH O 1
O-^
Figure 2. Cephalostatins 2-6. Cephalostatins 7-9 (Figure 3) exhibit further structural variation in the left half. In cephalostatin 7, the C-18' angular methyl is no longer oxidized, and the side chain now forms a 6/5 spiroketal. This spiroketal can be derived from the right half's 5/5 spiroketal by ring opening followed by recyclization and removal of the C-23 alcohol. Thus, the two halves of cephalostatin 7 are almost identical. Cephalostatins 8 and 9 retain the C-18' hydroxymethyl group. Compared to the parent cephalostatin 1, cephalostatin 8 has one more methyl group and is less oxygenated in the side chain. Meanwhile, cephalostatin 9 is really the hemiketal form of cephalostatin 1. By now, the NCI had switched from the PS bioassay to a panel of 60 human solid tumour cell lines. Against this panel, cephalostatins 7-9 are of similar potency to cephalostatin 1-4, with TI50 values of 10"^-10'^° molar against a number of the cell lines. The cephalostatins also display a characteristic panel graph, and exhibit one of the most extreme cases of differential cytotoxicity encountered in the NCI assay. Cephalostatin 1, for example, has GI50 values in these cell lines ranging from 6x10'^ to 2.5x10''' molar.
879
HO /
OH O \
O^
H^ OH O ^
oJ^^'^OH O
Cephalostatin 8
He ^4^^^^0H OH 0 1 9 ^
Cephalostatin 9
Figure 3. Cephalostatins 7-9. In the last two years, Pettit's group has reported the structure of eight more cephalostatins isolated from a 450 kg collection of C gilchristi made in 1990. Cephalostatins 10 and 11^ (Figure 4) are the C-1 and C-l' methoxy derivatives of cephalostatin 2 respectively. Cephalostatin 10 thus represents the first member with a structural change in the right half, which has remained constant in cephalostatins 1 through 9. In their cytotoxic effects, both these cephalostatins are of similar activity to cephalostatin 1. The disymmetry of the cephalostatin halves can be considered a biosynthetic puzzle - was the enzymatic machinery selectively fusing two non-identical steroids only, or was it differentiating the two halves of a homodimer? The latter possibility sounds more likely, and is supported by the similarity between the two halves of cephalostatin 7 {vide supra). The structures^ of cephalostatins 12 and 13 (Figure 4) provide further corroborating evidence. For the first time, the two halves of cephalostatin 12 are identical and correspond to the right half of earlier cephalostatins. Cephalostatin 13 differs from the symmetrical dimer only by C-l' hydroxylation. Interestingly, these compounds were found in the more polar n-butanol fraction during solvent partitioning, while previous cephalostatins were retained in the dichloromethane layer. The symmetrical compounds are also much less active against the NCI panel; while cephalostatin 1 has a mean panel GI50 of 1 nmolar, cephalostatin 12 and 13 were 400 nmolar and >1 fimolar respectively. The implications of this are unclear. Possibly, the increased polarity of the left half in these compounds is responsible for the decreased activity.
Cephalostatin 10 R = OMe, R' = H Cephalostatin 11 R = H, R' = OMe OH O 1 V^
)^0 HO-K\^'
i OH OH Cephalostatin 12 R = H Cephalostatin 13 R = OH
Figure 4. Cephalostatins 10-13 Cephalostatins 14 and 15 (Figure 5) are related to cephalostatins 2 and 3 respectively by aepoxidation at C-14'-15', dehydration of the C-9' alcohol, and hydroxylation at C-8'. Cephalostatins 14 and 15 display reduced activity, with mean panel GI50 of 100 and 68 nmolar respectively, perhaps due to the epoxide orientation - with steroidal bufadienolides, the p-epoxides are more cytotoxic.
OH O 1
Cephalostatin 14 Cephalostatin 15
p-^
R =H R = Me
Figures. Cephalostatins 14 and 15. The most recent cephalostatins to be reported are cephalostatin 16 and 17^ (Figure 6). Cephalostatin 16 is composed of the left half of cephalostatin 2 coupled to the left half of cephalostatin 7. Cephalostatin 17 also contains the left half of cephalostatin 2, while the other half is identical to the typical right half of cephalostatins 1-11 except for one less hydroxyl group. The
mean panel values for cephalostatins 16 and 17, at 1 nmolar and 4 nmolar respectively, are comparable to cephalostatin 1.
n- o ^ ^ f ' M l l
OH O c
Jij^3c
11 ^X-/
' ii
THT
N
H
J
X''///
•0
Cephalostatin 16
OH 0 1
pOH
Cephalostatin 17 Figure 6. Cephalostatins 16 and 17. In general, the cephalostatins isolated recently are in relatively minor abundance compared to cephalostatin 1 (for example, only 3.8 mg, 8x10*^ % yield, of cephalostatin 17 was obtained), and of lower activity except for cephalostatins 16 and 17. The Pettit group has also detected*^ other new cephalostatins in very small quantities (approximate yield of 10"^ %) with promising activity against brain cancer xenografts.
THE RITTERAZINES It remains to be seen what structural surprises are in store with future compounds isolated from C gilchristi, and also if similar compounds are produced by Antarctic members of the Cephalodiscus genus. Recently, exciting developments have been disclosed^ ^ from an unexpected quarter by Fusetani's group, working with the tunicate Ritterella tokioka Kott 1992 collected off the coast of Japan. The lipophilic extract of the tunicate showed promising cytotoxicity. PS bioassaydirected fractionation of 5.5 kg of tunicate yielded 2.9 mg of ritterazine A, with an ED50 of 10'^ |lg/ml in the PS assay. The structure of ritterazine A (Figure 7) was solved based on ^H and ^^C NMR data, and it bears an obvious resemblance to the cephalostatins. Prior to the extraction of the tunicate, colonies were first washed free of macroepibionts and sands. No attached hemichordates were observed.
882
HOI Figure?. RitterazineA. The left half of ritterazine A is identical to that of cephalostatin 7, except for an additional C7' hydroxyl group. The right half comprises a rearranged steroid nucleus. A reasonable biogenetic pathway for the right half is shown in Scheme 3 (In this and later schemes, hydrogens at steroid trans ring-junctures and portions of the steroid nucleus that are not participating in the reaction are usually omitted). Protonation of the C-14,15 alkene is followed by a 1,2-Wagner-Meerwein shift and trapping of the resulting carbocation by water to give the observed skeleton. The intermediate carbocation can also be derived by an alternative mechanism via protonation at C-15, pinacol-like rearrangement, and Prins reaction between the aldehyde and alkene.
ritterazine A
Scheme 3. Possible biogenesis of the ritterazine A right half. Subsequently, two other ritterazines with an unrearranged steroid nucleus as in the cephalostatins were isolated^^ from the same collection of R. tokioka (Figure 8). Ritterazine B has the same left half as ritterazine A. The right half is comparable to that of cephalostatin 1, except for the absence of hydroxylation at C-17, C-23, and C-26. Ritterazine B is also the first of these dimeric steroids where C-14 is not part of an alkene or epoxide. Instead, a p-hydrogen giving rise to cis C-D ring fusion is present. In ritterazine C, one half is identical to the right half of ritterazine B, while the other half is identical to the right half of cephalostatin 1 (this half is chosen as the "right" half in Figure 8 to emphasize this relationship) except for additional C-7 hydroxylation. Ritterazines B and C have an IC50 in the PS assay of 1.8x10"'* |ig/ml and 9.4x10"^ |ig/ml respectively.
883
\ OH OH
Ritterazine B
OH O 1
^/"s^r-
H
^\/N^^,.-\
THT
1
i^S^i ^ ^
0 V
Hj •
0
OH
i
O^
N
^ ^
Ritterazine C
Figure 8. Ritterazines B and C. Further work with an 8.2 kg collection of the tunicate led to the characterization'^ of ten new ritterazines (Figure 9). Ritterazine D is the C-22 epimer of ritterazine A. Ritterazine E has one additional methyl group compared to ritterazine D. The biological activity of these two compounds is similar to that of ritterazine A. In ritterazines F and G, C-22 is again epimeric to the configuration observed in ritterazine B; ritterazine G has the additional modification of C-14,15 unsaturation. In the PS bioassay, ritterazines F and G have IC50 values of T.SxlO'"^ M^g/ml - the highest activity seen among the ritterazines apart from ritterazine B (note that this is still much lower than the activity of cephalostatin 1). Ritterazines H and I form another pair of C-22 epimers, with C-12 now at the ketone oxidation state and biological activity reduced by approximately twenty-fold. The presence of both 5/5 spiroketal diastereomers for a number of ritterazines implies that epimerization at C-22 is not energetically prohibitive. Interestingly, such spiroketal epimers have yet to be observed among the cephalostatins. Ritterazines J-M are most closely related to the left half of cephalostatin 7. Ritterazine K is the symmetrical dimer formed from this half, while ritterazine J has one additional hydroxyl group. Ritterazines L and M are a pair of C-22 spiroketal epimers in which the C-17 hydroxyl group is lost. These ritterazines all have IC50 values around 10'^ fig/ml in the PS assay. The biological profile of the ritterazines, with ritterazines B, F, and G the most potent, seem to indicate the importance of the 5/5 and 6/5 spiroketals, while the CI4,15 alkene is not crucial. Oxidation at C-12 to a ketone results in decreased activity, as does the rearrangement of the steroid nucleus as in ritterazines A, D, and E. However, further data on related compounds will be needed before these hypotheses can be confirmed. Meanwhile, the difference in nomenclature between the cephalostatin and ritterazine families is rather confusing. Regardless of their origin, these compounds clearly belong together and perhaps they should be reclassified under a single family of steroidal alkaloids.
884
HO'
i OH OH
RitterazineD R = H Ritterazine E R = Me
Ritterazine F Rj = OH, R2 = R3 = H HOI
\ OH OH Ritterazine G Rj = OH, R2 = H, C14,15 alkene Ritterazine H Rj = R2 = O, R3 = H Ritterazine I
Rj = R2 = O, R3 = H, C-22 epimer OH Rol
f' OH
H R i = Ro
HOL
^ OH OH
Ritterazine K R j = H R2 = Ritterazine L Rj = R2 =H Ritterazine M R^ = R2 = H, C-22 epimer
Figure 9. Ritterazines D-M. The isolation of the ritterazines from a phylum unrelated to the hemichordates leaves the true source of these dimers unclear. With certain natural products isolated from marine macroorganisn it is now established'"^ that they are actually produced by symbiotic microflora, and this may be case here. If a microorganism that produces these dimers can be identified and grown in laborai. conditions, it raises the exciting prospect of obtaining these highly potent steroids in greater quantities by large-scale fermentation. The biosynthesis of these compounds seems to occur in two phases: (1) coupling of two steroids via a pyrazine linker, and (2) relatively unselective oxidation at various positions. Some of these compounds are related to others by simple processes: the hydration of cephalostatin 1 to its hemiketal form cephalostatin 9; the dehydration of cephalostatin 2 to an enone, which in turn may be an intermediate to cephalostatin 6; the skeletal rearrangement of ritterazine B to ritterazine A, which may be acid catalyzed; and the pairs of ritterazines epimeric at C-22. One can speculate whether such reactions are non-enzymatic transformations occurring in the organism or even during the isolation procedure.
885
SYNTHETIC STEROID-PYRAZINE DIMERS VIA a-AMINO KETONES The intriguing structure of these steroids coupled with their potent biological activity and limited availability makes them an attractive challenge for the synthetic organic chemist. One of the key features of any attempted synthesis is the central heterocyclic ring. The classical method of pyrazine synthesis involves the dimerization of a-amino ketones. A steroidal example relevant to the cephalostatin problem was reported^^ as early as 1968 (Scheme 4): androstanolone (1) was converted to a-oximino ketone 2, which was hydrogenated to afford the hydrochloride salt of a-amino ketone 3. The salt was neutralized to the free base, and condensed in situ to yield symmetrical dimer 4 in modest overall yields.
l.H2,Pd-C,HCl 2. aq. Na2C03
Scheme 4. First preparation of a pyrazine linked steroid dimer. Subsequent to the isolation of the cephalostatins, there was renewed interest in such steroid dimers. An improved procedure for the preparation of 4 was reported^^ by the Fuchs group (Scheme 5), in which the intermediate a-amino ketone was produced in higher yield by the sequence of bromination with phenyltrimethylammonium bromide, displacement by azide to give a-azido ketone 5, and hydrogenation (an identical route was independently developed^^ in the Heathcock group). OH
OH l.PTAB 2. [(Me2N)2CNH2] N3 j ^ •
O
71 %
H2, Pd-C - ^ Dimer 4 41 %
Scheme 5. Synthesis of 4 via azido ketone 5. The Fuchs group also prepared dimers 6-11 (Figure 10) by analogous procedures. These compounds were tested by the MTT cytotoxicity method in a panel of five human tumour cell lines. The dimer 6 with the cholestanyl side chain had little activity, with an ED50 > 100 |ig/ml. However, all the other dimers displayed some degree of cytotoxicity. For example, dimer 4 had an ED50 of 7 |ig/ml against the colon adenocarcinoma cell line HT-29, and values around 30 jig/ml against the others.
886 Although the activity is low, it is interesting given the huge simphfication in structure compared to the natural products. The results suggest that more sophisticated cephalostatin analogues with improved biological activity can be prepared synthetically. Such compounds would be more accessible than the scarce natural products; moreover, their cytotoxicity can be modulated, whereas the natural products may be too toxic for direct therapeutic use. Dimers 4 and 10 were also tested in two murine epithelial tumour xenografts transformed by mutations in the low-molecular weight guanine nucleotide binding protein Ras. Ras is an important protein in cellular signal transduction, cycling between an inactive GDP-bound state and an active GTP-bound form. The protein also has intrinsic GTPase enzymatic capability, thus preventing permanent activation. In the mutated Ras, GTP hydrolysis is greatly diminished, leading to signals that cause cell proliferation. At the maximally tolerated dose of 150 mg/kg/day, 4 inhibited the tumours by 28 and 59 %, without any deaths due to cytotoxicity.
Common pyrazine and A-B ring core of dimers 6-11.
HO
9
10
11
Figure 10. Symmetrical dimers prepared from azido ketones by Fuchs and coworkers. While preparing large quantities of 4 for animal testing, the Fuchs group isolated 5-10 % of a byproduct, identified as its CI-azido derivative. Control studies determined that this compound was produced by the presence of excess azide during the displacement reaction, and the suggested mechanism is shown in Scheme 6. Thus, azide acting as a base^^ enolizes a-azido ketone 5, which then loses nitrogen to yield a-amino enone 12. This process can also be recreated by treating 5 with DBU. Dimerization of 12 followed by reaction with azide completes the process. It is possible that similar generation of an electrophilic centre adjacent to the pyrazine occurs during cephalostatin biosynthesis, as some of the natural products also show C-1 substitution. Interestingly, dimer 13 has increased cytotoxicity compared to 4 and 6-11, with an ED50 in the 0.2-0.4 M-g/ml range against the same tumour cell lines. Presumably, this increase is a nonspecific effect caused by the presence of the toxic azide functional group.
887
"Xt H,N
I
azide
dimerization
:
^""m^
OH
*
13
Scheme 6. Proposed mechanism for formation of the C-1 azido dimer 13. Further improvement was achieved^^ in the a-amino ketone dimerization process by Smith and Heathcock. Cholestanone (14) was converted to its 2a-azido derivative 15. The azide (Scheme 7) was reacted with aqueous triphenylphosphine, rather than reduction by hydrogenation, and the crude dihydropyrazine dimers aromatized by air oxidation in the presence of p-toluenesulfonic acid. The product (6) was then isolated in high yield by simple filtration. The triphenylphosphine reaction proceeds via an imino phosphorane, which in principle can undergo dimerization by a Staudingerlike reaction. However, no dimerization was observed under anhydrous conditions, implying that hydrolysis of the phosophorane ylide to the a-amino ketone occurs first. The rate of phosphorane hydrolysis in NMR experiments was relatively slow (several hours), and it may be possible to produce unsymmetrical dimers by trapping the a-amino ketone as it is being formed by a more reactive keto steroid.
PPh,
""H C oXX
PhgP
x^:^
.o
[/.• "2°. H,N,,^^4/
' • ^
i;4^ ^ rvt^^^x) ^ N - ^ ^
87% overall
:
*
.
^
.
Dimer 6
Scheme 7. Preparation of pyrazine dimers using triphenylphosphine for azide reduction.
UNSYMMETRICAL PYRAZINES An obvious disadvantage of a-amino ketone dimerizations is their unsuitability for crosscoupling. The alternative of condensing a 1,2-dicarbonyl component with a 1,2-amine is also inappropriate for most cephalostatins and ritterazines, which are unsymmetrical both from right to left and top to bottom. It is possible that unsymmetrical condensation could be carried out in a stepwise manner using two a-amino ketones protected in different ways. However, this would
decrease the efficiency of pyrazine formation, which would be a late step in a total synthesis with highly functionalized and precious monomers. In the Heathcock group, a number of solutions to this problem were explored. One discovery^^ was the hetero-Diels-Alder-like reaction between oxadiazinones and enamines (Scheme 8). Although this reaction has the required regiospecificity and takes place under very mild conditions, it proved impossible to prepare an oxadiazinone fused to a cycloalkane, as required in a cephalostatin synthesis. -80 ^C to RT H3C -CO2, -pyrrolidine y-^Y>^
H3C N
I
I
+
• K' \
O
Ph
80%
N Ph
"3
I \ --"3
Scheme 8. Synthesis of an unsymmetrical pyrazine via a Diels-Alder like reaction. Later, ^^ Smith and Heathcock studied dimerizations with a-amino oximes as one of the components. Amino oxime 16, derived from 15 in two steps (Scheme 9), was reacted with epoxy acetate 17. The product mixture contained "trans" pyrazine dimer 6 and its A^-oxide derivative 18 in low yield, together with 2a-acetoxycholestanone (19). However, the expected "cis" dimer 20 was produced in too small a quantity for isolation. Heating 16 alone yielded 6 together with a trace of A^oxide 18. Epoxy acetate 17 was also unstable to the reaction conditions, slowly rearranging to 2pacetoxycholestanone (21) which then epimerized to the equatorial acetate 19. Heating amino oxime 16 together with 2p-acetoxycholestanone gave dimers 6 and 18 with the latter predominating, as observed in the initial experiment. Furthermore, the ratio of 6 to 18 increased with higher initial concentrations of 21, suggesting that dimer 18 was the product of cross-coupling. I.NH2OH 2. Ph3P, H2O O
15
90%
":xt OH
1. AC2O O
16 + 17
2. dimethyldioxirane ^ O , 14
74%
O Ac
17
toluene, 85 ^C. 24 h • •
N 6,5
«
tt txxi: N 18, 10 %
Ac
O
19
20, not detected Scheme 9. Dimerization of the cholestanyl amino oxime 16 and epoxy acetate 17.
For improved yields, an analogue of amino oxime 16 less prone to self-dimerization was needed. Towards this end, the 0-methyl derivative 22 was prepared, which undergoes complete dimerization only at 140 ^C. Heating acetoxy ketone 21 with 22 at 85 °C gave dimer 6 in 3.5 % yield after 1 day, together with unidentified compounds which may be intermediates in its formation. Extending the reaction time to 14 days increased the yield to 23 % (Scheme 10). A protocol was worked out involving initial heating at 85-90 ^C during which the acetoxy ketone and amino oxime ether preferentially react with each other, followed by heating at 140 °C to complete the process. Acetoxy ketone 19 was also isomerized to the 3(3-acetoxy-2-one 23, which then dimerizes with 22 to give "cis" pyrazine 20.
140 ^C, 24 h OMe
22
87%
Dimer 6
Ac 22, 85 ^C, 14 d >. Dimer 6 21
23%
Ac
:rt; 19
Me4N0Ac
•:ii
35%
Ac
22, 90 ^C, 5 d 20%
23
tt
N Dimer 20
Scheme 10. Dimerizations involving amino oxime ether 22. A complication in these reactions is the similar properties for "trans" dimer 6 and the "cis" dimer 20. The only noticeable difference in the NMR spectrum is the shift of one ^H resonance by 0.01 ppm, and a shift in the pyrazine ^^C resonances by about 0.1 ppm, while the optical rotations are identical ([ajo = +82^). However, the compounds have very different solubilities - trituration of the mixture with ethanol yields the trans dimer upon filtration, while the crude cis dimer is obtained by evaporation of the filtrate. This separation procedure enabled purification of the mixture obtained from reaction of 2a,3a-diaminocholestane (obtained by borane reduction of 22) and 2,3diketocholestane (obtained by oxidation of 3-cholestanone with potassium t-butoxide and oxygen) (Figure 11). H2N,, H2N
'tt O'
110QC,24h^ Trans Dimer 6, 28 % + Cis Dimer 20, 31 %
Figure 11. Pyrazine dimers via reaction between 1,2-diamines and 1,2-diketones. The utility of the amino oxime ether - acetoxy ketone combination was shown using two different steroid monomers. Androstanolone was converted to the 2p,17p-diacetate 24 (Scheme 11), which was then reacted with amino oxime ether 22 to afford unsymmetrical dimer 25 in reasonable yield. In this dimerization, replacement of 24 by its 2a-acetoxy epimer gave similar results, while the 2a-bromoketone instead led to messy reaction mixtures. Reactions between 22 and 2,3-epoxy acetates were also tried, which produced an approximately 1:1 ratio of cis and trans pyrazine dimers.
890 Compound 25 was also hydrolyzed to give the corresponding dimer with the free C-17 hydroxy! group.
r ' j A 2.dimethyldioxirane Ac ^ [ \ amino ^"^' oxime ether 22 ^^^xvsl^As^^x-^^ 3. refluxing toluene ^ O^^x-Vsfxx^.x^'^^ ^^^''C, 24 h; 140 ^C 24 h O^^^--^^
58%
O ^ ^ - ^ ^
.
^
^^^
^
Scheme 11. Smith and Heathcock's synthesis of an unsymmetrical pyrazine. Starting from hecogenin acetate (26), an inexpensive commercially available sapogenin whose side chain resembles that of the cephalostatins, keto alcohol 27 was prepared (Scheme 12). This was carried forward to acetoxy ketone 28, the recrystallization step in this sequence being necessary as the MoOPD oxidation yields an inseparable mixture of the 2-hydroxy-3-one, 3-hydroxy2-one, and 4-hydroxy-3-one. Reaction of 28 with amino oxime ether 22 under the usual conditions provided 29 % of unsymmetrical dimer 29. The acetate group in 29 was also hydrolyzed to give a more crystalline dimer. Dimers 6, 25, and 29, together with the deacetoxy derivatives of 25 and 29, were all submitted for testing in the NCI's solid tumour panel. None of the dimers were sufficiently active to warrant further investigation. l.NaBH.
1. IDA; MoOPD 2. AC2O 3. recrystallization 38%
28
Scheme 12. Synthesis of a unsymmetrical dimer with a spiroketal side chain. Another route to unsymmetrical steroidal pyrazines is the likely biosynthetic process of differentiation of a symmetrical dimer, as accomplished^^ by Winterfeldt's group. Starting from
891 hecogenin acetate, compound 30 containing the C-14,15 alkene present in the cephalostatins was synthesized by a literature procedure^^ involving the interesting sequence of photochemical isomerization and Prins reaction (Scheme 13). O
"'-. Q^>^'''
-^
0 I]
\*o
•
0
.,<^
"r*^
80%
hecogenin acetate, 26
lumihecogenin acetate
AcOH 45%
Scheme 13. Introduction of the C-14,15 alkene into hecogenin acetate. Compound 30 was carried forward to a-bromoketone 31 which was displaced by azide to yield oc-amino enone 32 directly in good yield (Scheme 14). The formation of 32 is probably due to the use of excess azide (as discussed in Scheme 6, vide supra).
Scheme 14. Direct formation of an enamino ketone by azide displacement. The amino enone was hydrogenated to directly afford symmetrical pyrazine dimer 33 (Scheme 15) in 64 % yield. The dimer was enolized with an excess (3.2 equivalents) of base, and trapped with pivaloyl chloride to give a statistical 1:2 mixture of the bis-enol pivalate and the unsymmetrical monopivalate 34. The two compounds were readily separable by column chromatography. The bis-pivalate can be recycled by hydrolysis back to the ketone, while reduction of monopivalate 34 followed by hydrolysis yielded hydroxy ketone 35, in which the C-ring oxygenation pattem is similar to that in cephalostatin 1. It would be interesting to see if 35 can undergo acid catalyzed rearrangement to the ritterazine skeleton, although such experiments may be complicated by the known high sensitivity of ritterazine A to acid.
892
KHMDS; pivaloyl chloride 40%
l.NaBH4 2. aq. KOH 78%
^^
V^^
o*\U o i
N' 35
o
Scheme 15. Synthesis of an unsymmetrical pyrazine dimer by Winterfeldt and coworkers.
THE FUCHS SYNTHESIS OF TETRAHYDROCEPHALOSTATIN 12 The successful synthesis of symmetrical and unsymmetrical steroid-pyrazine dimers outlined in the previous sections has solved one of the key issues for a total synthesis of the cephalostatins. However, the preparation of suitably functionalized steroid monomers is still a daunting challenge, and it is Fuchs and coworkers who have accomplished the most progress in this endeavour. In 1994, Jeong and Fuchs reported^^ the preparation of a homodimer corresponding to the right half of cephalostatin 1, except for the C-14,15 alkene. In fact, this dimer corresponds to the saturated version of cephalostatin 12, but the existence of this symmetrical cephalostatin had yet to be reported at the time of this work. Diacetate 36 (Scheme 16), prepared from hecogenin acetate 26 in two steps, was subjected to a modified Marker sapogenin ring-opening procedure^"* to give dihydrofuran 37. The ester was hydrolyzed, and the free alcohol dehydrated by the sequence of tosylate formation, displacement by selenide, and selenoxide elimination to yield diene 38. This compound already contains all the necessary carbons for the cephalostatin right half, thus obviating the need for any carbon-carbon bond forming reactions.
893
Py.HCl, (Cl2CHCO)20 refluxing xylene, 30 min O 37 l.aq. Na2C03 2. p-TsCl 3.PhSeSePh,NaBH 4. m-CPBA 41 %
Scheme 16. Jeong and Fuchs' synthesis of diene 38 from hecogenin acetate. Treatment of 38 with phenyl methyl sulfoxide activated^^ by trifluoroacetic anhydride produced the separable C-23 diastereomeric trifluoroacetates in a ratio of 2.2:1 (Scheme 17). Following hydrolysis, the correct diastereomer (stereochemistry was assigned by X-ray analysis) for the cephalostatin synthesis was protected as its silyl ether 39, while the wrong epimer was recycled by Mitsunobu inversion in 80 % yield. Stereospecific osmylation of 39 was performed with the use of Corey's chiral diamine,^^ and gave the required diastereomer 40 in a ratio of 7.7:1. CF
PhSOMe, TFAA
lO
• lO
Ph
38
Ph
HA^'^^OH 1. separation 2. protection •
^
OTBDPS
•
OTBDPS
98%
Scheme 17. Hydroxylation of the steroid side chain of 38. Spiroketalization of 40 under acid catalysis was unproductive. Upon prior deprotection of the silyl ether to give triol 41, the spiroketalization was successfully accomplished, providing a mixture of the 5/5 and 6/5 spiroketals 42 and 43 respectively (Scheme 18). While the 6/5 spiroketal 43 could be quantitatively isomerized to 42 under more vigorous conditions, both these spiroketals unfortunately have the "unnatural" configuration at C-20. Presumably, electrophilic attack on the
894 alkene takes place from the less hindered a-face to give the P-methyl group. Under equilibrating conditions, epimerization at this position to the more favourable a-methyl orientation as in the natural products can be anticipated, but was not observed.
OH
Scheme 18. Formation of 5/5 and 6/5 spiroketals. The inability to equilibrate C-20 required a detour involving electrophilic spiroketalization via bromination. In this case, only the 5/5 spiroketal was formed from 41, but reduction of the halogen with triphenyltin hydride only gave 42, the undesired C-20 epimer. However, the corresponding spiroketalization product of the silyl ether 40 was reduced by triphenyltin hydride with a diastereomeric ratio of 4.2:1 in favour of 44 with the desired C-20 stereochemistry (Scheme 19). -OH
-OH Ph3SnH OTBDPS82%(+17%C-20epin£r):
O
OTBDPS
Scheme 19. Adjustment of C-20 stereochemistry via reductive dehalogenation. The free alcohol in 44 was protected as the silyl ether, and the less hindered C-3 acetate selectively hydrolysed and subjected to Jones oxidation (Scheme 20). The ketone was a-brominated to give 45 in 76 % yield, together with 14 % of the C-26 desilylated monobromide and 7 % of the 2,2-dibrominated product. >'^0H
Ac.
l.TBDMSCl 2. aq. KHCO3
O
^
V'^^OTBDMS O ^
3. CrOo O
OTBDPS
Br —/y
4.PTAB •
61%
O 45
Scheme 20. Preparation of an a-bromo ketone intermediate for dimerization. In the final stages of the synthesis (Scheme 21), the dimerization protocol developed earlier (Scheme 5, vide supra) was followed, with the slight modification that triphenyltin hydride was used for reduction of the azide functional group. Cyclization of the a-aminoketone with PPTS afforded dimer 46 in 79 % yield, accompanied by 17 % of deazidoketone monomer. Removal of the
895 protecting groups yielded tetrahydrocephalostatin 12 (47). Unfortunately, the biological activity of this compound was not disclosed.
Br
45
l.[(Me2N)2CNH2]N3 2. Ph3SnH 3. PPTS
deprotection
73%
94%
46
OH
">. O"
d^'"'
1
U'^^OH OH
H0^^^
Scheme 21. The synthesis of tetrahydrocephalostatin 12.
A COMMON INTERMEDIATE FOR CEPHALOSTATIN 7 MONOMERS The next results reported by the Fuchs group concerned cephalostatin 7. The ability to form the 5/5 spiroketal of the right half and the 6/5 spiroketal of the left half from a common advanced intermediate would be advantageous from a practical standpoint. The feasibility of this approach can be seen in the work on tetrahydrocephalostatin 12 (Scheme 18, vide supra). A more detailed study^^ was carried out of the spiroketalization reaction. Compounds obtained by C-23 deoxygenation of intermediates in the tetrahydrocephalostatin 12 synthesis were used, with various permutations differing in their C-20 stereochemistry and also the presence or absence of benzylation at C-26. These compounds were subjected to a series of acid-catalyzed cyclizations. In the case of the C-26 alcohol, eight products are possible (Scheme 22). Depending on the starting point, four of these products were observed upon overnight treatment with PPTS at room temperature (structural assignments were aided by X-ray analysis of two of the compounds). However, when heated overnight at 83 °C, the product mixtures collapse to only one 5/5 spiroketal and one 6/5 spiroketal in a 1.5:1 ratio. Molecular mechanics calculations performed using the CAChe v3.5 program are consistent with the experimental observations. The two products detected in the high-temperature equilibration are only approximately 0.5 kcal/mol apart, while the six other possibilities are significantly higher in energy, by at least 4 kcal/mol. With the C-26 benzyl derivative, four different 5/5 spiroketals are possible (Scheme 22). All of these were observed in room temperature cyclizations, while equilibration at higher temperatures gave a single 5/5 spiroketal. Again, this is in agreement with calculations, where one product is at least 4 kcal/mol lower in energy than the others.
896
[ R = H,
55.4 J
Scheme 22. Possible products of spiroketalization experiments, together with their energies in kcal/mol by molecular mechanics calculations. Highlighted diastereomers indicate those of lowest energy. Molecular mechanics calculations were also performed on compounds with the C-14,15 alkene and the C-17 alcohol present as in the natural product (Figure 12). The results indicate that preparation of the left half's 6/5 spiroketal is best achieved before deoxygenation at C-23, as the 6/5 spiroketal is then 2.5 kcal/mol lower in energy than the 5/5 spiroketal. In the deoxygenated series, both 5/5 and 6/5 spiroketals are within 1 kcal/mol of each other. R R
R = H, 55.2 R = OH, 59.6
:H, 54.6 : OH, 57.1
Figure 12. Molecular mechanics calculations on spiroketals. Numbers indicate energies in kcal/mol. Subsequent to this study, the preparation of a suitable common intermediate for cephalostatin 7 was reported^^ by Kim and Fuchs. The need to introduce the C-14,15 alkene and the C-17 alcohol necessitated a substantially different synthetic route from that earlier employed for tetrahydrocephalostatin 12. Marker side chain degradation of hecogenin acetate was followed by allylic bromination to yield 48 (Scheme 23). This compound was epoxidized and the halogen reductively cleaved with concomitant epoxide ring opening using a zinc/copper couple to give intermediate 49.
897 The alcohol was protected as its TMS ether, and the C-15,16 alkene stereospecifically dihydroxylated to give compound 50. The diol was then converted to its cyclic sulfate derivative according to the Sharpless protocol.^^ Attempted base-catalyzed elimination of the sulfate to introduce the C-14,15 alkene was plagued by side-reactions involving epoxide formation by displacement of the sulfate by the adjacent TMS ether, perhaps aided by enolization of the methyl ketone. Instead, displacement of the sulfate by iodide ion occurred uneventfully to provide 51 as its tetrabutylammonium salt. l.DIBAL 2. AC2O 3. Py.HCl, AC2O I.H2O2, OH" 2. Zn/Cul,ultrasound Br 74 < l.SOCl2,Et3N OAc
l.TMSOTf OH 2. OSO4
<^
2. NaI04, RUCI3 3. BU4NI
90'
' . I
.49
OH
89%
••50
Scheme 23. Introduction of the C-17 alcohol into a precursor for cephalostatin 7. The successful formation of the C-14,15 alkene was accomplished by oxidation of 51 with mCPBA (Scheme 24), probably via an iodoso intermediate which undergoes syn-elimination of hypoiodous acid and hydrolysis of the sulfate to yield 52. The oxidative elimination is a procedure originally developed^^ by Reich with obvious potential for complex molecule synthesis, although it is presently less popular than related sequences involving sulfoxides or selenoxides.
^ OTMS l.m-CPBA _ 2. cat. H^SO.
OSO,
^
1
93% 52 Scheme 24. Introduction of the C-14,15 alkene by the Reich protocol. With the alkene in place, the E-ring dihydrofuran was constructed by an intramolecular Wadsworth-Emmons reaction to yield ester 53 (Scheme 25). Adjustment of the oxidation state afforded aldehyde 54, and completion of the synthesis involved extension of the side-chain by addition of methallylstannane. Various conditions for this reaction were investigated, with the best results obtained with the use of 5 M lithium perchlorate^'. This gave a 1.3:1 mixture of separable alcohols 55 and 56 in quantitative yield. The production of both diastereomers in nearly equal
proportions is not a disadvantage in this case, as the compounds are suitable for processing into the two halves of cephalostatin 7 separately (see next section).
OAP'
OTMS (EtO)2PaCHN2CD2Et OH Rh2(OAc)4^ refiuxing benzene
TMSi OAcn I
CD2Et
TMS
l.IiBH4 2.Mn02 3.AC2O
O.p'OEt OIMS/ ^OEt ^ QO^ NaH
OPsLn L ^CHO O
1
72%
methallyl stannane. 5MLiaO. >95%
54 SchmB 25. Preparation of a common advanced intermediate fcff cephalostatin 7.
THE FUCHS SYNTHESIS OF THE TWO HALVES OF CEPHALOSTATIN 7 The right half of cephalostatin 7 was synthesized"^^ in an analogous manner to that of tetrahydrocephalostatin 12. The silyl ether of 55 was dihydroxylated to give a pair of inseparable diastereomers in a 4:1 ratio, the major product 57 corresponding to the natural C-25 stereochemistry (Scheme 26). However, no reaction was observed with conditions used for spiroketalization (Scheme 22, vide supra), while harsher conditions led to undesired products. As in the earlier work, the brominated derivative 58 of the desired 5/5 spiroketal was formed by treatment with NBS. A diastereomeric byproduct was also obtained, occurring from cyclization of the minor epimer of 57.
Q
OTBDPS OSO4, Corey diamine
HO-\^^'^OH TMSi OAc; OTBDPS
95% -OH ''III
NBS
O
OTBDPS
77% 58, + 15 % of C-25 epimer Scheme 26. Spiroketalization of the cephalostatin 7 right half.
57, + C-25 epimer in 4:1 ratio
899 Removal of the halogen in 58 using triphenyltin hydride as in the tetrahydrocephalostatin 12 synthesis led to complex product mixtures, possibly due to the bulky TMS ether at C-17 which can hinder quenching of the radical as well as form a reactive site for radical fragmentation. Instead, treatment of 58 with chromium[II] diacetate in the presence of a thiol gave 15 % of the undesired C20 a-methyl compound and 60 % recovered starting material, in a very slow reaction (48 hours). These conditions were based on analogy^"^ to Barton's chromium[11]-mediated halohydrin reductions. The speed of the reduction was improved by adding ethylenediamine; however, the product was the C-20 exo-methylene compound. Better results were obtained with chromium[n] chloride, provided a large excess of thiol (~ 100 molar equivalents) was used, suggesting that the thiol not only serves as the source of the hydrogen but also as an activating ligand for chromium[n]. The optimum conditions (Scheme 27) involved removal of the TMS protecting group prior to the reduction, and yielded 68 % of the desired compound 59 together with 19 % of the C-20 epimer.
TMS^Br
"OH
vN^^CM
\ VO OAc) ^ OTBraS
5eq.Cra2, '"/( 100eq.PtSH O CfTEDPS 68%
-CM QAC ^\
9 o amrx^
58 Sdieii£27. SteaeoselectiverenDvaloftheC-20halogai The preparation^"^ of the left half of cephalostatin 7 was more straightforward. Compound 56, the minor product of the stannane addition (Scheme 25, vide supra) was deoxygenated by Barton's xanthate procedure (Scheme 28) to give 60, which was stereoselectively dihydroxylated ,f35 using the Sharpless AD-mix-a reagent to yield 61 and its inseparable C-25 epimer in a 2.5:1 ratio.
OH l.NaH;CS2;MeI 2. Ph3SnH 87%
• 61, +C-25 epimer
Scheme 28. Elaboration of the left half of cephalostatin 7. Treatment of the mixture of 61 and its epimer with (+)-camphorsulfonic acid yielded an inseparable mixture of three spiroketals, which were purified by the sequence of silylation, chromatographic separation, and desilylation (Scheme 29). The major product 62 corresponds to the desired left half, while 63 is the alternative 5/5 spiroketal product. The formation of 62 and 63 in nearly equal amounts is consistent with molecular mechanics calculations, which indicate the two compounds are within 0.1 kcal/mol of each other. The third product, 64, is derived from the C-25 epimer of 61. This epimer can also produce a 6/5 spiroketal, although it was not detected (molecular mechanics calculations predict the 6,5 spiroketal to be 0.8 kcal/mol less stable than 64). The structures of all three products were established by X-ray analysis. Interestingly, in all three products,
900 the desilylation step also hydrolyzed the C-12 acetate, which is normally less reactive than the C-3 acetate. The accelerated hydrolysis at C-12 is probably due to neighbouring group participation by the C-17 alcohol.
61 + epimer
l.CSA 2. silylation 3. separation 4. desilylation
'OH
•.
62,'31% overall
63,21 % overall
64," 19% overall
Scheme 29. Spiroketalization of the left half of cephalostatin 7.
SYNTHESIS OF CEPHALOSTATINS 7,12 AND RITTERAZINE K While this manuscript was in preparation, the Fuchs group completed^^ the total synthesis of cephalostatins 7. Right half ketone 65 (Scheme 30), obtained from 59 {vide supra) was brominated and displaced by tertramethylguanidinum (TMGA) azide to afford 66. In methylene chloride, the azide displacement was plagued by formation of the a-amino enone (Scheme 6, vide supra) while the use of nitromethane^^ as solvent resulted in a quantitative yield. Similarly, left half ketone 67, obtained from 62, was brominated and displaced by TMGA azide in nitromethane to give 68. The MTM protecting group was lost during the bromination step, but bromination of an unprotected ketone gave a byproduct with a rearranged 5/5 spiroketal. TMS
OTBDPS
l.PTAB 2. TMGA-N3 72%
66
' OMTM l.PTAB 2. TMGA-No 72% 68 (with MTM group lost) Scheme 30. Preparation of a-azido ketones for the two halves of cephalostatin 7. The final step involved reduction of azido ketones 66 and 68 to amino ketones followed by statistical pyrazine formation. In this reaction, the product of cross-coupling is a protected form of
901 cephalostatin 7, while self-dimerization yields precursors to the symmetrical natural products cephalostatin 12 and ritterazine K. It is possible that such coupling occurs biosynthetically. If the rates of coupling are assumed to be similar, the fact that cephalostatin 12 (product of right half dimerization) is isolated in 10-fold higher quantity than cephalostatin 7 (cross-coupling product) from C gilchristi implies that the right half monomer is present in greater concentration. There may be a downstream biosynthetic step in which the symmetrical cephalostatin 12 or ritterazine K are converted to the more abundant unsymmetrical cephalostatin 1. In the laboratory, a 1:1 mixture of 66 and 68 was reduced with an excess of ethanolic NaHTe (Scheme 31). Silica gel was then added as a mild acid catalyst, and the reaction mixture stirred under exposure to air to complete the aromatization. Chromatography afforded protected cephalostatin 7 in 35 % yield, while the cephalostatin 12 and ritterazine K precursors were produced in 14 % and 23 % yield respectively. The deazido derivatives of 66 and 68 were also found in 36 % and 15 % respectively. As this side-reaction perturbs the initial 1:1 ratio of reactants, it helps explain why the pyrazine product distribution is not strictly 1:2:1 but biased towards formation of ritterazine K. Each of the three protected pyrazines was individually treated with excess tetrabutylammonium fluoride to give the first synthetic samples of cephalostatin 7, cephalostatin 12, and ritterazine K. 1. NaHTe 2. Si02, air 3.TBAF
66 + 68
H^ nJ^^'^OH OH O 1 V^ '// Cephalostatin 12, 11 % v'^^T^^
IHT
^*V,^—^"JS.
HO
,
/^T'
C'h '^OH
^#^ M
TO^v^^O
JL>.
N
'
HT OH
OH O 1
O-^ Cephalostatin 7, 28 %
H 1^
IHT /t^>r^
^ 0'
HT*
r '^ .y^ J ^ OH OH
HO /i
Jk>>
"^ N'
OH O 1
I lO ^
7' OH Ritterazine K, 18 '
Scheme 31. Total synthesis of cephalostatin 7, cephalostatin 12, and ritterazine K.
902 Samples of the synthetic compounds were provided to Professor Pettit's group, who confirmed the identity of cephalostatins 7 and 12 based on NMR and chromatographic comparison. If this mechanism for pyrazine formation occurs in C. gilchristi, one would expect it to also produce ritterazine K. A search among the currently unidentified residual Cephalodiscus extracts revealed a substance with identical chromatographic profile to ritterazine K. However, it was present in only microgram quantities, and its identity could not be confirmed by NMR.
THE FUCHS APPROACH TO DIHYRDOCEPHALOSTATIN 1 The Fuchs group has also prepared a monomer for the left half of dihydrocephalostatin 1. The dihydro compound was chosen in order to investigate the spiroketalization process and also the importance of the C-14,15 alkene for biological activity. The route begins^^ with keto aldehyde 69 (Scheme 32), available in 60 % yield from Marker degradation of hecogenin. This was reduced to keto alcohol 70, and the C-18 methyl group functionalized by Meystre's hypoiodite method,^^ after which Jones oxidation provided lactone 71. The C-3 acetate was hydrolyzed, and the free alcohol reprotected as a silyl ether, followed by reduction to give triol 72. Regioselective carbenoid insertion into a neopentyl alcohol set the stage an intramolecular Wadsworth-Emmons reaction (see Scheme 25, vide supra), after which the oxidation state was adjusted to yield 73. l.HOCH2CH20H,PPTS 2. NaBH4, ^^^^3 3.H2,Pt02
d^
OH
4. PPTS 59%
70
l.Pb(OAc)4,l2 2. H2Cr04 56%
l.Et02CC(N2)PO(OEt)2, Rh2(OAc)4 2. H2Cr04 3.NaH 4. LiAlH4 5. TFAA, DMSO, Et3N
I.KHCO3 2. TBDPSCl 3. LiAlH.
—
68%
72
73%
CHO
TBDSO Scheme 32. Preparation of a keto aldehyde precursor for dihydrocephalostatin 1.
•
903 Compound 73 was reacted with methallylstannane to give a separable pair of diastereomeric alcohols (Scheme 33) in a 1:2.7 ratio favouring the desired product. The unwanted diastereomer was recycled in 79 % yield by Mitsunobu inversion. The alcohol was then benzylated to afford intermediate 74, after which reduction of the C-12 ketone yielded a 1:9 ratio of a - and p-C-12 epimeric alcohols. The alcohols were osmylated and subjected to periodate cleavage to provide 75. This was reacted with methyl Grignard, followed by acid catalyzed cyclization with (+)camphorsulfonic acid. Three spiroketal products 76, 77, and 78 were isolated in a 1:15:1 ratio.
OBn CHO y\
1. methallyl stannane, BF-^.Et^O 2. BnBr, NaH
90%
l.MeMgBr 2. CSA
Q HO
78% 76
''^ 77
Scheme 33. Elaboration of 5/5 spiroketal of dihydrocephalostatin 1. Attempts at producing X-ray quality crystals of the spiroketals were unsuccessful. The compounds were treated with fluoride to effect C-3 deprotection, and nOe effects used to assign C-12 stereochemistry. Oxidation of the diols gave ketones 79 and 80 (Scheme 34), whose structures were determined by 2D-NMR experiments.
l.TBAF 2. H2Cr04 76 + 77 + 78
Scheme 34. Equilibration of the 5/5 spiroketal.
904 According to molecular mechanics calculations, the C-23 benzyl group favours equatorial attack of the alcohol on the oxonium ion intermediate during spiroketalization, hence explaining the kinetic preference for product 77. The calculations also revealed that 79 and 80 should be less stable than their diastereomeric spiroketals by approximately 2 kcal/mol. Indeed, heating 79 with camphorsulfonic acid gave a new ketone 81, while similar treatment of 80 produced 82 (Scheme 34). Upon extended reaction times, 81 was converted to 30 % of 82 (estimated to be approximately 3 kcal/mol lower in energy) along with decomposition products. This last reaction requires epimerization at C-20 through an oxonium ion-enol ether equilibration. Deprotection of the benzyl group in diketone 82 yielded 83 (Scheme 35), a dihydrocephalostatin 1 left half intermediate suitable for coupling with the right half.
BnO
H2, Pd-C 100 % O 82 83 Scheme 35. Preparation of a dihydrocephalostatin 1 left half monomer.
SUMMARY AND FUTURE PROSPECTS The family of dimeric steroid-pyrazine alkaloids isolated from Cephalodiscus and Ritteria now stands at thirty members. There are undoubtedly other examples of this group of steroidal alkaloids that have yet to be discovered, and it is probable that members common to both sources will be found. On the synthetic front, the discovery of these alkaloids has sparked interest in the construction of unsymmetrical pyrazines, and the methods developed will be useful in other settings as well. The Fuchs group has successfully accomplished landmark syntheses of tetrahydrocephalostatin 12, cephalostatin 7, cephalostatin 12, and ritterazine K and is clearly close to a synthesis of dihydrocephalostatin 1. These efforts have added significantly to the areas of steroid and spiroketal chemistry. The ability to make unnatural cephalostatins will greatly aid our understanding of the biological potency of these compounds. For example, one half could be kept identical to a natural product, while varying the other with synthetic steroids prepared from commercially available materials. In this sense, the cephalostatins provide a unique opportunity for such experiments, as the steroid skeleton is a readily accessible and well understood scaffold in which the effects of particular substituents can be determined. The situation here is in stark contrast to certain other potent biologically active compounds such as taxol (which promotes microtubule assembly) and bryostatin (a protein kinase C inhibitor). In the latter cases, construction of the skeletal framework is a formidable enterprise, hindering further dissection of the biological activity. The availability of larger quantities of biologically active material, whether from natural or synthetic sources, will unravel the site of action of these steroids, about which nothing is known at present. These alkaloids do not contain functional groups commonly assisted with antitumour agents e.g. alkylation and Michael acceptor sites, intercalators, redox-active quinonoid groups, and radical generators. The range of biological activity among the various natural products contains some tantalizing structure/activity clues which are difficult to fully decipher.
905 Soon after the 1988 communication on cephalostatin 1, it was predicted"^^ that the compound acts on the cell membrane. Steroids are components of eukaryotic cell membranes, where they incorporate into one half of the phospholipid bilayer and provide rigidity. Taking into account the dimeric nature of the cephalostatins, these steroids may now traverse the full length of the bilayer (for example, cephalostatin 1 is 30 A x 9 A x 5 A) and adversely affect its properties. A number of other highly oxygenated marine natural products (e.g. brevetoxin, palytoxin) are also membrane active agents. Fuchs has made two suggestions on the cephalostatins' mechanism of action. First,^^ the oxygenated functional groups of these compounds may form a network of hydrogen bond donors and acceptors that interacts with a specific enzyme target. More recently,"'^ Fuchs has implicated the C14,15 alkene as being important in biological activity, and there is some evidence"^^ supporting this hypothesis. He points out that the C-14,15 alkene may be susceptible to electrophilic attack in vivo, either by protonation or epoxidation, followed by a rearrangement similar to that postulated in Scheme 3 (vide supra) which would generate a number of reactive centres in the molecule. It will be interesting to see if the CD-ring skeletal rearrangement can be achieved in the laboratory, and if the reactive intermediates can alkylate DNA, for example. Another potentially reactive site in these molecules is the spiroketal ring system. Perhaps this undergoes ring opening to generate reactive carbonyl and free alcohol groups. Finally, returning to the title of this chapter, the necessity for these compounds to be dimeric remains unclear. It is not known if any of the advanced synthetic mono-steroid intermediates prepared also exhibit high cytotoxicity. Furthermore, whether the central pyrazine ring is simply a linker or serves some additional function is also a mystery. This could be tested by examining the biological activity of unnatural dimers with other linkages e.g. benzene rings, other heterocyclic systems, or even acyclic tethers.
ACKNOWLEDGEMENTS I wish to thank Professors George Pettit and Philip Fuchs for kindly providing preprints of references 9, 36, 38, and 40, and for their comments on the manuscript.
REFERENCES ' Pettit, G. R.; Kamano, Y.; Aoyagi, R.; Herald, C. L.; Doubek, D. L.; Schmidt, J. M.; Rudloe, J. J. Tetrahedron 1985, 41, 985-994. ^Pettit, G. R.; Inoue, M.; Kamano, Y.; Herald, D. L.; Arm, C; Dufresne, C ; Christie, N. D.; Schmidt, J. M.; Doubek, D. L.; Krupa, T. S. J. Am. Chem. Soc. 1988,110, 2006-2007. ^Pettit, G. R.; Inoue, M.; Kamano, Y.; Dufresne, C ; Christie, N.; Niven, M. L.; Herald, D. L.; /. Chem. Soc, Chem. Commun. 1988, 865-867. Also see J. Chem. Soc, Chem. Commun. 1988, 1440 for correction of a minor typographical error in the structures. '^Pettit, G. R.; Kamano, Y.; Dufresne, C ; Inoue, M.; Christie, N.; Schmidt, J. M.; Doubek, D. L.; Krupa, T. S. Can. J. Chem. 1989, 67, 1509-1513. ^Pettit, G. R.; Kamano, Y.; Inoue, M.; Dufresne, C ; Boyd, M. R.; Herald, C. L.; Schmidt, J. M.; Doubek, D. L.; Christie, N. D. J. Org. Chem. 1992, 57, 429-431. ^Pettit, G. R.; Xu, J.; Williams, M. D.; Christie, N. D.; Doubek, D. L.; Schmidt, J. M.; Boyd, M. R. J. Nat. Prod. 1994, 57, 52-63. ^Pettit, G. R.; Ichihara, Y.; Xu, J.; Boyd, M. R.; Williams, M. D. Bioorg. & Med. Chem. Lett. 1994, ^,1507-1512. ^Pettit, G. R.; Xu, J.; Ichihara, Y.; Williams, M. D.; Boyd, M. R. Can. J. Chem. 1994, 72, 22602267.
906 ^Pettit, G. R.; Xu, J.-P.; Schmidt, J. M. Bioorg. & Med. Chem. Lett. 1995, 5, 2027-2032. '^Pettit, G. R. Pure&Appl. Chem. 1994, 66, 2271-2281. ^' Fukuzawa, S.; Matsunaga, S.; Fusetani, N. J. Org. Chem. 1994, 59, 6164-6166. ^^Fukuzawa, S.; Matsunaga, S.; Fusetani, N. 7. Or^. Chem. 1995, 60, 608-614. ^^Fukuzawa, S.; Matsunaga, S.; Fusetani, N. Tetrahedron 1995, 57, 6707-6716. '"^For example, see: Kobayashi, J.; Ishibashi, M. Chem. Rev. 1993, 95, 1753-1769. ^^(a) Ohta, G.; Koshi, K.; Obata, K. Chem. Pharm. Bull. 1968, 7(5, 1487-1497. (b) Smith, H. E.; Hicks, A. A. J. Org. Chem. 1971, 36, 3659-3668. ^^Pan, Y.; Merriman, R. L.; Tanzer, L. R.; Fuchs, P. L. Bioorg. & Med. Chem. Lett. 1992, 2, 967972. '^Ganesan, A. Ph.D. dissertation. University of California-Berkeley, 1992. '^Edwards, O. E.; Purushothaman, K. K. Can. J. Chem. 1964, 42, 712-716. ^^ (a) Smith, S. C ; Heathcock, C. H. J. Org. Chem. 1992, 57, 6379-6380. (b) Full paper: Heathcock, C. H.; Smith, S. C. 7. Org. Chem. 1994, 59, 6828-6839. ^^Ganesan, A.; Heathcock, C. H. J. Org. Chem. 1993, 58, 6155-6157. ^' Kramer, A.; Ullmann, U.; Winterfeldt, E. J. Chem. Soc, Perkin Trans. 11993, 2865-2867. ^^Welzel, P.; Janssen, B.; Duddeck, H. LeibigsAnn. Chem. 1981, 546-564. ^^ Jeong, J. U.; Fuchs, P. L. J. Am. Chem. Soc. 1994,116, 773-774. ^^Marker, R. E.; Wagner, R. B.; Ulshafer, P. R.; Wittbecker, E. L.; Goldsmith, D. P. J.; Ruof, C. H. J. Am. Chem. Soc. 1947, 69, 2167-2230. For modifications, see: (a) Dauben, W. G.; Fonken, G. J. J. Am. Chem. Soc. 1954, 76, 4618-4619. (b) Micovic, I. V.; Ivanovic, M. D.; Piatak, D. M. Synthesis 1990,591-592. ^^(a) Suryawanshi, S. N.; Fuchs, P. L. J. Org. Chem. 1986, 57, 902-921. (b) Jain, S.; Shukla, K.; Mukhopadhyay, A.; Suryawanshi, S. N.; Bhakuni, D. S. Synth. Commun. 1990, 20, 1315-1320. ^^ Corey, E. J.; Da Silva Jardine, P.; Virgil, S.; Yuen, P.-W.; Connell, R. D. J. Am. Chem. Soc. 1989, 777,9243-9244. ^^ Jeong, J. U.; Fuchs, P. L. Tetrahedron Lett. 1994, 35, 5385-5388. ^^Kim, S.; Fuchs, P. L. Tetrahedron Lett. 1994, 35, 7163-7166. ^^Sharpless, K. B.; Gao, Y. J. Am. Chem. Soc. 1988, 770, 7538-7539. ^° Reich, H. J.; Peake, S. L. J. Am. Chem. Soc. 1978,100,4888-4889. ^' Henry, Jr. K. J.; Grieco, P. A.; Jagoe, C. T. Tetrahedron Lett. 1992, 33, 1817-1820. ^^ Kim, S.; Sutton, S. C ; Fuchs, P. L. Tetrahedron Lett. 1995, 36, 2427-2430. "Barton, D. H. R.; Basu, N. K.; Hesse, R. H.; Morehouse, F. S.; Pechet, M. M. J. Am. Chem. Soc. 1966,88,3016-3021. ^^ Jeong, J. U.; Fuchs, P. L. Tetrahedron Lett. 1995, 36, 2431-2434. ^^ Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.; Hartung, J.; Jeong, K.-S.; Kwong, H.-L.; Morikawa, K.; Wang, Z.-M.; Xu, D.; Zhang, X.-L. J. Org. Chem. 1992, 57, 2768-2771 36 ^^ Jeong, J. U.; Sutton, S. C ; Kim, S.; Fuchs, P. L. J. Am. Chem. Soc. 1995, 777, 10157-10158. 37 Li, C ; Shih, T.-L.; Jeong, J. U.; Arasappan, A.; Fuchs, P. L. Tetrahedron Lett. 1994, 35, 26452646. ^^ Bhandaru, S.; Fuchs, P. L. Tetrahedron Lett. 1995,46, 8347-8350. Heusler, K.; Wieland, P.; Meystre, C. H. Organic Synthesis 1965,45, 57-63. "^^ Bhandaru, S.; Fuchs, P. L. Tetrahedron Lett. 1995, 46, 8351-8354. "^^ Ganesan, A.; Heathcock, C. H. Chemtracts-Org. Chem. 1988, 7,311-312. Unpublished results from E. Winterfeldt's group, quoted in reference 19(b).
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 O 1996 Elsevier Science B.V. All rights reserved.
907
Human IgGl Hinge-Fragment as a Core Structure for Immunogens Luis Moroder, Gerd Hiibener and Manfred Gemeiner 1) Max-Planck-Institut fur Blochemie, 8 2 1 5 2 Martinsried. Germany; 2) Veterinar- Medizinische Universitat, Wien, Austria
Abstract: After the early discovery that synthetic fragments of proteins if suitably presented to the immune system, are capable of eliciting antibody responses cross-reactive with the parent proteins, the concept of synthetic vaccines has become a main target in immunology. Today there are wellestablished tenets that generation of the humoral response requires cooperation and communication between B and T cells and that small-sized antigens induce immunity only when appropriately presented for recognition by both the B-lymphocytes and T-helper cells. As immunological responsiveness is dependent on the HLA genotype of an individual, genetic restriction presents an additional burden for the design of fully synthetic immunogens as mimetics of surface regions of pathogens, but which nonetheless give raise to antibodies capable of neutralizing the microorganism and abolishing its infectivity. In the context of knowledge in the field our contribution is reviewed by comparing the intensity and specificity of the immune responses to a self-antigen, i.e. to the hormone gastrin, in various approaches as antigen/protein conjugates, built-in immunoadjuvanticity, liposomal preparations and finally with well defined constructs in which the hinge segment of human IgGl with its peculiar structural and functional properties was exploited. The conservative character of the hinge as pivot in the dynamics of the immunoglobulins, and its apparently universal role as recognition site in the interplay of immunologically relevant molecules was found to induce surprisingly high immunogenicity even in the case of selfantigens indicating that such hinge-peptide/antigen constructs as interesting molecules for studying the processes of immunity at cellular level.
1. Introduction With t h e development of vaccines medicine gained the ability to control, eliminate, or even eradicate selected diseases. The classical vaccines consisting mostly of killed or live a t t e n u a t e d microbial agents or their isolated c o m p o n e n t s have been highly successful. Smallpox h a s been eradicated, and viral d i s e a s e s like measles, mumps,
poliomyelitis,
rubella
and
yellow fever
rarely
occur
in
developed
countries. Similar s u c c e s s h a s been achieved with bacterial d i s e a s e s s u c h as diphteria, t e t a n u s , tubercolosis and whooping cough, a n d equally successful were vaccines in veterinary medicine. However, m a n y d i s e a s e s r e m a i n for which no
908 vaccines exist, e.g. malaria, herpes and t h e autoimmunodeficiency
syndrome
(aids). Despite t h e success, c u r r e n t p r o c e d u r e s of vaccine p r e p a r a t i o n p r e s e n t serious s h o r t c o m i n g s s u c h a s whether a particular viral vaccine p r e p a r a t i o n is completely killed or sufficiently a t t e n u a t e d , t h e genetic variation of v i r u s e s , or the difficulty in p r e p a r i n g enough material for vaccine production a s well a s the h a z a r d to p e r s o n n e l and environment w h e n working with large a m o u n t s of p a t h o g e n s , b u t also the low t e m p e r a t u r e s required for storage a n d t r a n s p o r t (1,2). With t h e existing vaccines immunization p r o g r a m m e s t h a t c a n r e a c h every child at c o s t s acceptable for every country are very difficult to be realized. The C h i l d r e n ' s Vaccine Initiative envisages the ideal vaccine of t h e future to b e safe, heat-stable, and advances
in
technology,
effective w h e n administered orally early in life (3). Recent
the
field
peptide
of molecular synthesis,
and
cellular
polysaccharide
immunology, chemistry
recombinant
and
microbial
p a t h o g e n e s i s m a y provide t h e clues for the rational design of vaccines of the future t h a t generate a d e q u a t e a n d s u s t a i n e d protective i m m u n e r e s p o n s e s even w h e n n a t u r a l disease fails to do so (1). Although the molecular b a s i s of immunological recognition r e m a i n s unclear,
a wealth
immunocompetent
of information
has
recently
been
accumulated
molecules and their interactions in p r o d u c i n g
largely on
the
successful
antibody r e s p o n s e s . With t h i s improved knowledge considerable efforts are now exerted t o w a r d s t h e development of new vaccines (1,4,5). One of t h e strategies used for this purpose is r e c o m b i n a n t DNA technology for the p r o d u c t i o n of pathogen s u b u n i t s , i.e. proteins, in either bacterial, yeast or animal cells, or even live vaccines by introducing relevant genes into t h e genome of vaccinia virus. In vitro expression of immunogenic proteins of p a t h o g e n s a s a m e t h o d for producing new vaccines h a s been often a n d unexpectedly u n s u c c e s s f u l (1). This indicates t h a t a protein isolated from a micro-organism is rarely a s i m m u n o g e n i c as t h e same protein w h e n it is a c o n s t i t u e n t p a r t of the Abbreviations: IgGl, immunoglobulin Gl; KLH, keyhole limpet hemocyanin; BSA, bovine serum albumin; RNase A. ribonuclease A; LDH, lactate dehydrogenase; MHC, major histocompatibility complex; LHRH, luteneizing hormone releasing hormone; CCK, cholecystokinin; VIP, vasoactive intestinal peptide; TASP, template-assembled synthetic proteins; MAPS, multiple antigen presenting system; SUV. small unilamellar vesicle; DCC, dicyclohexylcarbodiimide; HOSu, N-hydroxysuccinimide; Mal>, maleimido; TFA, trifluoroacetic acid; TFE, trifluoroethanol; DMF. dimethylformamide; DMSO, dimethylsulfoxide; CD, circular dichroism; NMR nuclear magnetic resonance spectroscopy; ELISA, enzyme-linked immunosorbent assay; MD, molecular dynamics; Mox, methoxinine (oxa-analogue of methionine); P-Ala. 3-aminopropionic acid; all amino acids are of L-configuration unless stated otherwise; abbreviations for amino acids and derivatives are according to the lUPAC-IUB nomenclature.
909 parent organism probably as a result of the altered environment. Moreover, the subunit approach, where success was more fortuitous than rational, does not provide much insight into the mechanism of immune responses at molecular level which, however, could be obtained from the use of synthetic peptides as antigenic determinants of microbial proteins (1). This alternative approach towards the development of potential vaccines is based on rather simple fully synthetic structures which are mimicking small regions in the pathogen, but can nonetheless give raise to antibodies capable of neutralizing the microorganism and abolishing its infectivity (1,5). The concept of synthetic vaccines based on relatively small peptides has been proposed already in the early 1970s (6,7). Actually, Anderer (8,9) made the very important discovery that a short tryptic fragment of the protein of tobacco mosaic virus and its synthetic replicate elicit antibodies capable of neutralizing the infectivity of the virus. But it was not until Sela and coworkers (10) demonstrated that a synthetic fragment of the coat protein of Ms2 bacteriophage leads to antibodies that cross-react with the virus particle, that the concept of synthetic vaccines received the due attention. Since these early pioneering studies great progresses were achieved in the better understanding of the structural requirements for an effective immunization, i.e. for the ability to provoke antibody responses and sensitized cells of appropriate specificity that lead to protection, for built-in immunoadjuvanticity and possibly for a bypass to the genetic restriction and antigenic competition. Our contributions to the field were exclusively devoted to the chemical approach, and are reviewed here in the context of some central issues in the field as induction of immunogenicity into fully synthetic constructs and of immunological memory. The the gastrointestinal hormone gastrin was selected as antigen throughout the study to allow for a more precise comparative analysis of the various approaches used to induce immune responses to this self-antigen.
2. Immunogens on Peptide/Carrier Basis In terms of immunological properties proteins or related fragments exhibit essentially two distinct characteristics which do not necessarily coincide. The first, immunogenicity. is the capacity to elicit an immune response, manifested either by antibody production (humoral immune response) or by cell-mediated immunity. The second. (B-cell) antigenicity, refers to the capacity to be recognized in a specific manner by immunocompetent cells or by an antibody. Macromolecular antigens, mostly proteins and glycoproteins, usually express in their structure a multitude of possible antigenic determinants, or epitopes, that
910 dictate the specificity of the immune response. However, only a limited number of potential epitopes are involved in the induction of immunogenicity. It is well established that immunizations with short peptides generally result in low levels of antibodies and only a few natural (11) and synthetic peptides (12-20) were sufficiently immunogenic to fulfil the requirements of a fully competent immunogen. Therefore conjugation of peptides to synthetic or natural (protein) carriers for macromolecularization is required to elicit strong immune responses. The carriers are typically proteins like keyhole limpet hemocyanin (KLH), tetanus toxin, bovine serum albumin (BSA), ovalbumin, chicken immunoglobulin and thyroglobulin, all being potent immunogens which apart from providing polyvalent presentation of the antigenic peptide as B-cell epitope, also function to provide Tcell specific epitopes. Today there are well-established tenets in immunology that generation of the humoral immune response requires cooperation and communication between B and T cells (21) and that small-sized antigens induce immunity only when covalently attached to proteins that can be recognized upon processing by T-cells (22-24). Moreover, immunological responsiveness is dependent on the HLA genotype of an individual (25), which has now been correlated with the ability of the individual's MHC molecules to bind fragments of the immunogen for presentation to the T-cells (26). The rules that govern the binding of peptides (fragments of the immunogen) to class I and class II MHC (27-29), the 3Dstructures of the two classes of MHC molecules (30-36) and the sequence motifs characteristic for recognition of T-cell epitopes by the MHC molecules of T-helper and cytotoxic T-cells have been disclosed (37-41).
2.1. Conventional Peptide/Protein
Conjugates
Generally, peptides are coupled to the carriers by homobifunctional or heterobifunctional reagents (45-46) among which the most commonly used are carbodiimides, particularly in the water soluble form, glutaraldehyde, isocyanates and diimidoesters. This ill-defined chemistry leads to extensive inter- and intramolecular crosslinking of the component parts of the conjugate as a result of lack of selectivity of the reactions exploited with the reagents. The presence of various reactive functional groups both in the antigenic peptide and in the protein carrier generates chemical modifications and crosslinkings at different positions of the peptide chain and carrier surface. Correspondingly, the peptide antigen may be presented to the immune system in heterogeneous form regarding its chemical structure, its carrier-surface environment and thus, possibly also its spatial array in the conjugate. Heterogeneous epitopes both in terms of chemical and spatial
911 structure are formed which generally induce heterogeneous responses (47,48). Additionally, epitope suppression and destruction, and generation of new dominant epitopes related to the cross-linking reagents are obvious consequences of this methodology (49-51). Using peptides of the gastrin hormone family as model antigens these uncontrolled effects were fully confirmed in our laboratory (48,52). In fact, gastrin/BSA conjugates prepared in conventional manner by the water-soluble carbodiimide led in animal-dependent manner to antibodies of differentiated specificity in full agreement with results from other laboratories where immunization experiments were usually performed on a larger number of animals to obtain in a "trial and error" manner antisera of the desired specificity (53-55).
2.2. Selective Polyvalent Peptide/Protein
Conjugates
Aware of these serious shortcomings resulting from the poor selectivity of the conjugation procedures, we have proposed the use of a more defined chemistry based on a reactive anchor group introduced at well defined positions of the synthetic antigen and suitable for a selective conjugation to a protein (47,48,56,57). Similarly to what nature is using in some of the posttranslational processings of proteins, thiol functions accessible on the surface of proteins lend themselves as ideal targets for conjugate chemistry since these reactive groups allow for highly selective crosslinking of molecules via disulfide- or thioether-bond formation exploiting mild thiol interchange or thiol addition reactions. For latter type reaction Keller and Rudinger (58) introduced the maleimide group as highly reactive thiol acceptor on the basis of previous studies on the reactivity of thiols with maleimide (59,60). Whilst other laboratories contemporarily proposed maleimidated carrier proteins as reaction partners for cysteine-containing peptides (61), our efforts in the field were focussed on the synthetic accessibility of peptide derivatives containing the reactive maleimide handle at positions of the peptide chain preselected in view of a maximum retention of the antigenic structure (Fig. 1). The maleimido-protein approach which today represents the most frequently used selective conjugation technique is based on maleimido-benzoylated proteins and synthetic peptides containing at their N- or C-termini an additional cysteine residue (61). This method, however, bears two main difficulties: i) The maleimido-function which is introduced via reaction of the carrier protein, generally KLH, with m-maleimidobenzoyl-N-hydroxysuccinimide ester, leads to a bulky aromatic anchor group on the protein surface with can generate strong
912 new artificial immunodeterminants in the conjugation site of the antigenic peptide. ii) The solid phase synthesis of peptides with C-terminal cysteine residues is accompanied by tedious side reactions (62,63) and peptides with N-terminal cysteine exhibit high tendency for oxidative dimerization (64). As bypass to these side reactions at least regarding N-terminal cysteine the use of N<^-acetylcysteine (64) or mercaptoacetic acid (65) as N-terminus has been proposed. The problem, however, is more easily bypassed according to our experience by N- or C-terminal elongation of the antigenic peptides with a Gly-Cys and Cys-Gly dipeptide, respectively (unpublished results).
( CARRIER ]—SH
V
+
[[
N—j SPACER^I
CARRIER — S — r A J [^^ N - - | SPACER|--[
ANTIGEN
ANTIGEN
|
|
Fig. 1. Schematic representation of the thiol/maleimide approach for the monoand polyconjugation of peptide antigens to carrier proteins.
As illustrated in Fig. 1, our approach is based on the maleimide group linked to the synthetic peptide and the thiol group placed on the protein partner. For this purpose sufficient stability of the maleimide function in the course of the peptide synthesis as well as in the final deprotection and purification step is required. A detailed analysis of these critical aspects revealed full stability of the maleimide function under the normal conditions of peptide synthesis (56,57). On the other hand, thiolated carrier proteins are readily prepared by reduction of cystinecontaining proteins, by mercaptosuccinylation of carrier proteins (66,67) or by the choice of natural mono- or poly-cysteine-containing proteins (68). This type of conjugation has been examined in our laboratory by using gastrin as antigen. The related maleoyl-gastrin derivative was synthetized according to the scheme outlined in Fig. 2 (57). A relatively short p-alanyl-spacer was incorporated
913 into the N0'-maleoyl-P-alanyl-gastrin-12-171 derivative as it was expected to space sufficiently the gastrin molecule as antigen from the conjugation site and at the same time to reduce to a minimum the non-peptidic structure as artificial component and thus, as a potentially new immunodominant epitope.
O OH
H2N
HOSli/DCC oM" lo r.t., 4-2411
0>^ / O .
/ N OH
+ H-Gly-Pro-Trp-Leu-(Glu) 5-Ala-Tyr-Gly-Trp-X-Asp-Phe-NH 2
X = Met, Nle, Mox
P
O II N-CH2—CH2—C—Gly-Pro-Trp-Leu-(Glu)5-Ala-Tyr-Gly-Trp-X-Asp-Phe-NH2
II O
Fig. 2.
Synthetic route for the preparation of N^-maleoyl-fi-alanyl-gastrin-[2-17]; the methionine residue in the gastrin sequence can be replaced by norleucine or methoxinine without effect on the biological and antigenic properties of gastrin (69).
RNase A was selected as the poly-cystine-protein and, upon reduction of the four disulfide bridges with mercaptanes, it was reacted with N^^-maleoyl-P-alanylgastrin-[2-17] in different molar ratios as shown in Fig. 3, in order to analyze the effect of mono- and poly\^alent presentation of the antigenic peptide to the immunocompetent cells. Thereby the remaining free cysteine residues were
914 allowed to reoxidize and size exclusion chromatography was used to isolate monomeric scrambled RNase/gastrin adducts from oligomeric forms. Immunization experiments with the three conjugates in rabbits led to almost identical results in terms of anti-gastrin antibody titers. The immune responses
SH HS-
SH
SH
red. RNase A
I—SH
+ Mai > B-Ala-gastrin-[2-l7]
1—\—r SH
SH
SH
1) 1. equiv.; 2. equiv., excess 2) oxidation RNase A / gastrin |
ratios: (1:1); (1:2); (1:6)
(scrambled)
Fig. 3.
Synthesis ratios.
of "RNase A"/gastrin
conjugates at different
carrier/antigen
were as strong as those induced with conventional BSA/gastrin conjugates containing 30 to 40 copies of the antigenic peptide and prepared by unspecific crosslinking with carbodiimide or glutaraldehyde (52,68). However, the specificity of the antisera was found to vary distinctly from animal to animal indicating that, despite the site-specific attachment of the antigen to the protein, the problem of environmental heterogeneity of the antigen moiety in its presentation to the immunocompetent cells, as expected from the chemistry used, was not resolved by this approach. The identical problem may arise when the method of maleimidoproteins is applied or the alternative approach based on statistically mercaptosuccinylated carrier proteins.
2.3. Selective Monovalent Peptide/Protein Conjugates A significant improvement in terms of uniformity of the immune responses in various experimental animals was expected from the use of an overall well-defined peptide/protein conjugate. Cytochromes are known to exhibit a highly conservative character both in terms of primary and tertiary structures (71), and
915 correspondingly this class of proteins is poorly immunogenic, but it can be rendered more immunogenic via oligomerization with cross-linking reagents (72). Extensive immunological studies on cytochromes allowed the identification of efficient T-cell epitopes (73,74) which should elicit the cognate T-cell help if cytochromes are used as carriers. For this purpose we selected iso-1-cytochrome c as this protein is known to contain a single cysteine residue located in the penultimate position of the sequence (75) which according to the observed facile oxidation to dimers (76) was expected to be fully exposed on the surface of the globular protein for conjugation reactions. Moreover, superimposition of the sequence of iso-1-cytochrome c with the known 3D-structures of cytochromes (77,78) clearly confirmed a surface position of the C-terminal end of the molecule.
+
Mai > B-Ala-gastnn-[2-l7]
Fig. 4. Synthesis of the strcuturally well defined monovalent c/gastrin-l2-17l conjugate.
iso-l-cytochrome-
Conjugation of N»-maleoyl-P-alanyl-gastrin-[2-17] to iso-1-cytochrome c (Fig. 4) was expected to produce a unique population of immunogens both regarding site-specific grafting of the antigen and the environment of the antigen on the protein surface for recognition by the B-cell surface-immunoglobulins if a preferred gastrin/protein-surface interaction was taking place. The experimental results of the immunizations performed in rabbits (68), guinea pigs (79) and C57BL/6 mice (unpublished results) confirmed this working assumption. Besides
916 the high titers of anti-gastrin antisera in these animals (only in Balb/c mice a weak response was observed) which were comparable to those obtained with the gastrin/RNase and gastrin/BSA immunogens, a monoclonal-like homogeneity in the immune specificity was achieved as well assessed by the cross-reactivities of the various polyclonal antisera with a large set of gastrin-related peptides of different chain length and of critical amino acid substitutions both regarding the conformational and hormonal properties of the antigen (for the S3aithetic gastrinpeptides used in these immunological studies see Table 1). The gastrin molecule is a linear, relatively short polypeptide chain without structural restraints which in aqueous solution is fully randomly coiled (80), but capable of assuming preferred ordered conformations in aqueous organic media and in surfactant micelles as well assessed by extensive conformational studies (80-83). The biological relevance of the proposed bioactive conformation has been confirmed (81,84). Modifications in the N-terminal part of the gastrin molecule are known not to affect both biopotency (85) and antigenicity of the hormone (67,86,87). Therefore, the selective grafting of the gastrin molecule to a defined position of the carrier protein was performed at its N-terminus, and antibodies of receptor-like specificity were obtained and this not in a fortuitous manner, but constantly in all experimental animals. As a consequence of these findings two alternative immune recognition processes can be envisaged: i) The unmodified gastrin sequence linked selectively to a defined surface position of the carrier retains sufficiently free conformational space to fold upon interaction with the B-cell surface immunoglobulins, into a low-energy ordered structure similar to its bioactive conformation, thus triggering immune responses of receptor-like specificities. ii) Upon linkage of the antigen molecule to the carrier, the protein environment induces a folding of the gastrin peptide into its sequence-encoded preferred secondary structure, thereby presenting the antigen to the B-cell surface receptors as conformational epitope. Support for this explanation was obtained by CD measurements (88). Although the CD contributions of the carrier protein are fully dominant, the difference CD spectra indicated an increase of the dichroic intensities in the 222 nm region as index of enhanced a-helical content as expected for a folded form of the gastrin moiety. The results obtained with gastrin were confirmed by similar experiments performed with CCK peptides as antigens (89,90). Similarly to the CCK-B/gastrin receptor which is unable to discriminate between gastrin and CCK (91), antibodies raised against structurally defined iso-1-cytochrome/CCK-peptide adducts were recognizing both gastrin and CCK with identical affinities suggesting again expression of a conformational epitope. In fact, a comparison of the preferred
gastrin Pyr-Gly-Pro-Trp-Leu-Glu-Glu-Glu-Glu-Glu-Ala-Tyr-Gly-Trp-Met-~p-Phe-NH~ [Nlel5]-gastrin Pyr-Gly-Pro-Trp-Leu-Glu-Glu-Glu-GIu-Glu-~a-Tyr-G~y-Trp-Nle-Asp-Phe-NH~ gastrin-[2-17] H-Gly-Pro-Trp-Leu-G1u-Glu-Glu-Glu-Glu-~a-Tyr-Gly-Trp-Met-Asp-Phe-NH~ ~oxl5]-gastrin-[2-17] H-Gly-Pro-Trp-Leu-GIu-Glu-Glu-Glu-GIu-~a-Tyr-Gly-Trp-Mox-Asp-Phe-~~ [Nlel5]-gastrin-[5- 171 H-Leu-Glu-Glu-Glu-GIu-Glu-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH~
Pyr-Glu-Glu-Glu-Glu-Ala-Tyr-Gly-Trp-Me-Asp-Phe-NH~ py1-6,Me~~]-gastrin-[6-17] Pyr-Glu-Glu-Glu-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH2
[Pyr7,Nlel5]-gastrin-[7-17]
Pyr-Glu-Glu-Ala-Tyr-Gl y-Trp-Nle-Asp-Phe-NH2
pfl,Me15]-gastrin-[8-17]
P yr-Glu-Ala-T yr-Gl y-Trp-Ne-Asp-Phe-NH2
[Pyr9,Nlel5]-gastrin-[9-171
Pyr-Ala-Tyr-Gl y-Trp-Me-Asp-Phe-NH2
pyrlO,Nlels]-gastrin-[ 10-171
Pyr-Leu-Gly-Pro-Gln-Gly-Pro-Pro-His-Leu-Val-Ala-~p-Pro-OHbig-gastrin-[1-141 H-Trp-Met-Asp-Phe-NH2
Table I . Sequences ofgastrin and big-gastrin peptides used in the studies reviewed; Pyr
gastrin-[ 14-17]
= pyroglutamic
acid
918 conformations of gastrin (82,83) and CCK (92.93) in aqueous organic media indicates almost identity in the spatial array of the active site region of the two hormones. The data collected in these experiments allow interesting general conclusions to be drawn. It is known that B-cells recognize antigens in their native form in solution via specific receptors on the B-cell surface in a highly conformationdependent fashion involving residues which are aligned in a sequence portion (continuous epitopes) or may be far apart in the primary structure of larger antigens (proteins), but proximal in the 3D-structure of the folded antigen (discontinuous epitopes) (94,95). Although experimental evidences suggest that most of the antigenic determinants (B-cell epitopes) of proteins consist of discontinuous epitopes, synthetic peptides have been successfully applied to raise specific anti-protein antibodies (96). The practical consequence of these findings is that surprisingly often linear peptides are mimicking conformational antigenic sites of the parent proteins or portions of these by folding into similar structures if sufficient conformational information is encoded in their sequence (97.98). The required conformational transition of the peptide into its sequence-dependent preferred structure is certainly facilitated if a maximum of free conformational space is retained by the antigenic peptide on the carrier surface and if the peptide is placed in an unique proteic environment.
2 A, Non-Covalent Peptide/Protein Complexes Recently an interesting alternative to the peptide/protein conjugates has been presented, consisting of a hydrophobic complexation of peptides, particularly of peptides with a hydrophobic tail, to proteosomes (99,100). Proteosomes are purified preparations of meningococcal outer membrane proteins which interact hydrophobically to form vesicles of 60 to 100 nm in diameter (101). Proteosomes are components of group B meningococcal vaccines where a polysaccharide is hydrophobically complexed to the proteosomes by way of a lipid moiety (102). The proteosomes besides providing the required T-cell help for peptide immunogenicity act also as potent immunoadjuvants probably as a result of their liposome-like aggregational state. Lipo-peptides containing a fatty acid or a hydrophobic peptide sequence as anchor are forming tight complexes with these proteosomes and the complexes were found to be highly Immunogenic without adjuvants. A similarly Interesting non-covalent complex could be envisioned with the streptavldln or avldln/blotin complex because of its extremely low dissociation constant of 10"^^ M. This system has been extensively exploited in immunological techniques, particularly to improve sensitivity and specificity of immunoassays
919 (103). Avidin and streptavldln are tetrameric proteins capable of binding up to four molecules of biotin or biotlnylated peptides. The high binding constant is only slightly affected by conjugation of the biotin via its carboiQd terminus to molecules of different sizes if siafflciently large and non-bulky^ spacers are used. We have carefully examined the system in order to assess the required spacing of a biotlnylated antigenic peptide for optimal recognition by antibodies even in the streptavidin-complexed state (87). As shown in Fig. 5. by assuring suflBcient spacing between the biotin headgroup and the conformational epitope of gastrin12-17] full recognition of the antigen by anti-gastrln antibodies of receptor-like specificity is retained. This would strongly recommend the use of this tetravalent complex schematically outlined in Fig. 6, as promising immimogen since all four Uganded antigens are presented in the identical protein environment. Moreover, streptavidin has been shown to contain efficient T-cell epitopes (104) which would
1:100000
1:10000' 1:10001:300 Serum Dilution
Fig. 5. Anti-gastrin antibody binding capacity of biotinylated gastrin captured by polystyrene-adsorbed streptavidin: biotinyl-cysteamine/N^'maleoyh/i' alanyl'[Nle^^]'gastrin'l2-17] adduct (O—O) and biotinyl'NH'lCH^4'C0'GlyPrO'[Nle^^]'gcLStnn'l2'l 7] (A-A).
920 make this molecule an efficient immunogen. Additionally the 3D-structure of the streptavldin/biotin complex is known (105) and modelling of the attached antigencoiald possibly allow for a better understanding of the conformational effects of protein surfaces on the attached peptide antigens. As even lO^-fold excesses of biotin failed to displace the biotlnylated gastrin (87) in full agreement with the difficult retrieval of biotinyl-ligands from affinity columns (106), the ubiquitous presence of biotin in physiological fluids should not displace the antigenfromthe immunogenic complex. To our knowledge this system has not yet been applied in immunization experiments.
antigen "^
Fig. 6. Schematic representation of the tetravalent streptavidin/biotinyl-antigen complex as potential immunogen.
3. Synthetic Peptide/Carrier Conjugates 3. J. Multiple Antigen Presenting Systems Poly-a-amino-acids and random copolymers as poly-(DX-Ala)-poly-(L-Lys) as well as branched poly-(L-Lys) polymers like poly-lL-Lys-(D,L-Alain)l or poly-[L-Lys-(LeuiD.L-Alam)! have been used as fully synthetic carriers for peptide antigens (10.107111). These are generally less immunogenic than proteins as well documented in a comparative study using the C-terminal tetrapeptide of gastrin as antigen and poly-(L-Lys) or poly-(L-Glu) as carriers (112). Recently a structurally better defined branched S3nithetic carrier based on a dendritic matrix of lysyl residues has been proposed as multiple antigen presenting system (MAPS) (113.114). This approch allows for amplifying the antigen presentation 4- to 8-fold and thus, to attain macromolecularizatlon of low-weight peptide antigens. It poses, however, great
921 synthetic problems because of induced chain aggregation in the course of the solid phase synthesis and of very difficult purification and analytical characterization. These intrinsic difficulties could possibly be overcome more easily by functionalized l3^ine matrices suitable for selective linkage of presyntheslzed antigenic peptides (115.116). Moreover, using p-alanyMysine as building unit nearly symmetrical core matrices can be obtained which could possibly represent a more flexible scaffold in peptide imimunogens (117). Upoderivatization of this scaffold besides allowing embedment of the multiple antigen constructs into phospholipld-based liposomes as promising formulation for s)nithetic vaccines, was found to induce even C5^otoxic T-l)niiphocytes and thus, to trigger both humoral and cell-mediated immune responses (117-119). Thereby the nature of the lipophilic tail has not necessarily to mimic the trls-palmitoyl-Sglyceryl-cysteine moiety of the lipo-protein of E. coli (see chapter 3.2), but can rely on simple fatty-acylated lysine residues (117). Generally, the MAPS approach was found to significantly augment the immunogenicity of peptides (120-123), but their use for raising anti-peptide antibodies capable of crossreacting efficiently with the cognate native protein appears to be much more limited (124). This could possibly result from strong interchain interactions, which in turn may prevent the onset of sequence-specific conformational properties or induce conformations dictated by the system-facilitated aggregational phenomena. Unfortunately, only limited conformational studies were performed so far on these synthetic constructs to assess the conformational space of the single antigenic moieties in these boundled constructs (125). Besides the more or less branched poly-a-amino-acid and oligo-a-amino-acid carriers, chemically modified biopolymers such as dextranes, or entirely S3aithetic unsoluble and water-soluble resins such as polystyrene and chemically linked acrylic polymers were applied as supports for antigenic peptides in immunization experiments (126-133). Interestingly, with peptide/polyoxyethylene (132) and peptide/water-soluble polydimethylacryl-amide resin conjugates (130,131) optimal titers of antibodies were obtained even against short peptide hormones, i.e. self-antigens like substance P and K, bradykinin and LHRH. Conversely by selective linkage of maleoyl-gastrin to thiol-functionalized polyoxyethylene or thiopropyl-Sepharose we were unable to raise detectable antibody titers against gastrin in rabbits (48,68). Our results are more consistent with the known immunosuppressive activity, particularly, of polyoxyethylene.
922 3.2. Synthetic Non-CovcHent Peptide/Carrier Complexes Liposomes are also known for their Immunological adjuvant action and have therefore been extensively used as vehicles for enhancing both humoral and cellmediated inmiune responses of associated antigens (134.135). Since liposomes are made up from synthetic or natural components which are by themselves nonimmunogenlc. non-toxic and biodegradable, their use for the preparation of synthetic peptide vaccines appears highly promising. The structural versatility of liposomes has permitted tailoring of the system towards optimal adjuvanticity which was achieved by appropriate choice of the vesicle composition, surface charge, vesicle size, by the mode of association of the antigen and by the liposomal fluidity. The intensity of the immune response can further be enhanced by the incorporation of adjuvants such as lipid A or its non-toxic derivative monophosphoryl lipid A into the vesicle bilayer (135). Many studies have been performed with liposome-associated protein antigens whereby simple entrapment, surface adsorption and anchorage to the bilayer via lipo-derivatization has been extensively investigated (136). On the contrary little work has been carried out with synthetic peptides where usually one or more lipidic amino acids are incorporated or conjugation of the peptides to phospholipids is used for their anchorage to the liposomes (137-141). A veiy detailed study has recently been reported with a synthetic hexapeptide antig .. which does not contain a T-cell epitope as required for T-cell help (142.143). Tl a results of this study showed that SUV preparations containing surface-bound peptides and as immunoadjuvant monophosphoryl lipid A were capable of eliciting long lasting immune responses against the peptide antigen. Monophosphoryl-lipid A can be embedded into liposomes that carry the antigen without loosing its adjuvant activity (145,146) and its presence in the bilayers was found to be essential for the observed immune response. An impressive variety of natural and synthetic compounds have been recognized as efficient immunostimulants, i.e. as immunoadjuvants (147-149). Despite the diversity in chemical structure, surface activity seems to constitute one of the common physicochemical properties of most of the adjuvants which makes these amphipathic substances capable of accumulating at membrane/water interphase. In analogy to the glycosylinositolphosphate used by nature to anchor proteins on the cell membrane surface we have proposed the use of l,2-di-fattyacyl-3mercaptoglycerol as "sticky finger" for the lipo-derivatization of peptides via the malelmido/thiol capture reaction in order to assure their tight embedment into lipid bilayers (150). Adopting this procedure new adjuvant compounds were designed which are readUy incorporated as integral parts into lipid bilayers or might constitute by themselves the bilayer membrane of liposomal preparations of
923 antigens (151). On the basis of this concept we have synthetized a series of 1,2-difattyacyl-3-mercaptoglycerols and linked them as shown in Fig. 7, to the dipeptide unit of the muramyl-dipeptide (150.151). Tlie muramyl dipeptide, i.e. N-acetylmuramyl-L-alanyl-D-isoglutamine. was recognized as the smallest adjuvant-active unit of Mycobacteria cell walls (152.153).
X -SH
o
DMF,r.t.30n
I
75-85%
N-(CH2)2-CORi
R = palmitoyl myristoyl retinoyl
Rl=.NH-(CH2)2-COOH -L-Ala-D-isoglutamine
Fig. 7, Synthesis oflipidic compounds as potential imnumoadfuvants. The compounds listed in Fig. 7 as well as the lipophilic muramyl-tripeptide derivative N-acetyl-muramyl-L-alanyl-D-isoglutaminyl-N^-palmitoyl-L-lysyl-S-tertbutylthio-cysteamlne. shown in Fig. 11. were analyzed for the their immunostimulating potencies. All these compounds exhibited activities in the range of those of conventional immunoadjuvants whereby the most active was found to be the lipo-muramyl-tripeptlde derivative in full agreement with the
924 Immunoadjuvantlcity derivatives (153).
reported
previously
for
similar
muramyl-tripeptide
Again using gastrin as model antigenic peptide we have attempted the liposome approach using these synthetic lipo-adjuvants and the llpo-gastrin derivative shown in Fig. 8. for vesicle preparations. The double-tailed llpo-gastrin was obtained as well characterized adduct via the thlol/maleimlde reaction (154). The highly charged hydrophlllc gastrin molecule is transformed by this procedure into a sequential amphathlc compound which, as expected from the chemical structure, aggregates In aqueous solution to form upon sonlcatlon or extrusion a monodlspersed population of vesicles of surprisingly high stability. In presence of dl-myristoylphosphatldylchollne vesicles at lipid to Hpo-peptlde molar ratios > 50
X CH3-(CH2)i2'^0-CH2
CH3-(CH2)i2''''^0—CH
I
N—(CH2)2-C~R
R = Gly-Pro-Trp-Lcu-(Glu)5-AIa-Tyr-Gly-Trp-Nle-Asp-Phc-NH2
Fig. 8. Chemical structure of the lipophilic l,2-dimiiristoyl-3'mercatoglycerol/Mdl>fi 'alanyl'[Nle^^]'gastrin-l2-17] adduct a quantitative transfer of the llpo-gastrin to the phospholipid bilayers is taking place as well assessed by differential scanning calorlmetry, whereby Incorporation of the llpo-peptlde is not affecting the phase transition temperature of the resulting SUVs (155). In these vesicle preparations the gastrin headgroups are exposed to the bulk water in random coil structure according to CD and fluorescence measurements (155) and most interestingly, retain their full affinity for the gastrin receptors (154). Similar full exposure of the gastrin antigen for recognition by the surface-immunoglobulins of B-cells was expected in liposomes prepared directly with the llpo-gastrin and the lipo-adjuvant compounds reported
925 in Fig. 7. This recognition by the B-cells should then lead to a fast internalization of the lipo-derivatlzed antigen and according to Friede et al. (143) to a strong immune response against this small peptide antigen even in absence of an efi&cient T-cell epitope. Immunization experiments with these lipo-adjuvant/lipo-peptide vesicles schematically represented in Fig. 9. were performed in C57BL/6 mice and compared with the effect of gastrin by itself emulsified in complete Preud's adjuvant. As already observed in rabbits and guinea pigs unmodified gastrin as self-antigen was completely non-immunogenic; however, even the immune response to the liposomal preparations was found to be weak and in some animals no anti-gastrin antibody titers were detectable. Moreover, only IgM type responses were detected without switch to IgG even after boostering (Gemelner. M., Dorrer, E. and Moroder. L.. unpublished results). This observation apparently contrasts the strong and long lasting immune response induced by the liposome approach for small peptides lacking T-ceU epitopes where besides IgM at least a partial switch to IgG was observed (143). The negative results with the liposomal gastrin preparation might derivefiromthe fact that gastrin itself is self-antigen even in mice because of the high interspecies sequence homology of this gastrointestinal hormone as well evidentiated by the alignment of the known gastrin sequences shown in Fig. 10. Moreover, replacement of monophosphoiyl-
Fig. 9. Schematic representation of the bilayers of lipo-gastrin/lipo-adjuvant vesicles.
926 lipid A by the newly synthetlzed adjuvants could possibly play a decisive role, although the high density of negative charges of the gastrin containing vesicles was expected to facilitate phagocytosis by macrophages (156.157).
Man Pig Cat Dog Cow Sheep Rat Guinea pig
<E <E <E <E <E <E <E <E
G P W L E E E E E A YGWMDP-NH2 G PWME E E E E A YGWMDF-NH2 G P W L E E E E A A Y G WM D F-NH2 G P W M E E E E A A Y G W M D F-NH2 G P W V E E E E A A Y G WM D F-NH2 G P W V E E E E A A Y G WM D F-NH2 R P P M E E E E A A Y G W M D F-NH2 G P W A - E E E A A Y G W M D F-NH2
J ^ . JO. Sequences of the hormon gastrin of mammalian origin.
4. Synthetic Immunogenic Constructs 4. i. Antigens with Built-in Immunoadjuvanticity It is well established in immunology that lipo-derlvatlzation of antigens is significantly enhancing the immunogenlcity of peptides and proteins (158-161). Moreover, it has been demonstrated that lipo-derivatlzation allows for induction of MHC class-I restricted cytotoxic and MHC class II mediated humoral immune responses bypassing the prerequisite of adjuvants for efficient immunizations (162), particularly, if peptides are covalently linked to N-palmitoyl-S-[2.3bis(palmitoyloxy)propyll-L-Cys-Ser or related moieties (163-166) which represent the N-terminal lipidic component of lipoprotein A of E. colt (167-169). A similar effect has been reported if peptides, even if self-antigens such as the hormone LHRH. are linked covalently to other well established immunoadjuvants. e.g. to the muramyl-dipeptide. In fact, with a N-acetylmuramyl-dipeptide/LHRH construct immunological castration of male mice was achieved (170-173). The identical effect has recently been obtained with a N-acetylmuramyldipeptide/LHRH construct linked to a solubilized polymer support (131). We have attempted this approach with the self-antigen gastrin. In order to exploit the
927 selective thlol/maleimlde reaction a thlol-functionallzed lipo-derlvatized Nacetylmuramyl-tripeptide was synthetized as schematically outlined in Fig. 11 (174); for the synthesis of N-acetylmuramyl-L-alan)d-D-iso-glutaminyl-S-tert-butylcysteamine a similar route was used. Thereby the N-acetylmuramyl moiety was linked to the peptide component without protection of the carbohydrate by using
Fmoc-Lys(Boc)-NH-(CH2)2-S-StBu --2SL
Fmoc-Lys-NH-(CH2)2-S-StBu CH3
H-Lys.NH-(CH2)2-S-StBu ^pipendine CH3-{CH2)i4-CO
i
Fmoc-Lvs-NH-(CH2)2-S-StBu CH3-(CH2)i4-CO
Nps-Ala-D-Glu-NH2
DCC/HONSi
Nps-Ala-D-Glu-NH2 LLys-NH.(CH2)2-S-StBu
-i^U-
I
H-Ala-I>Glu-NH2 LLys-NH.(CH2)2-S-StBu
I
CH3-(CH2)i4-CO
CH3-(CH2)i4-CO PH2OH
DCC/HONSu CH3CH
NHAc
COOH
CH3CH
NHAc
C0-Ala-D-Glu-NH2 l-Lys-NH-(CH2)2-S-StBu
I
CH3-(CH2)i4-CO
Fig, 11. Synthetic route for N-acetylmuramyl'L-alcmyl-l>iso-glutajninyU^^ palmitoyV'lysyl'S'tert'butyl-cysteamtne, N-hydro3Qrsuccinimide as additive; this procedure greatiy facilitated the synthetic accessibility of the N-acetylmuramyl compounds. The adjuvant molecules were then deprotected at the thiol function by reduction with tii-butylphosphlne and
928 reacted with the malelmldo-gastrin derivative to yield the well defined adjuvant/gastrin conjugates shown in Fig. 12 (175). Upon immunization of rabbits with these conjugates no immime response could be observed at least in terms of detectable anti-gastrin antibodies. The question arises of whether the immunological castration induced with the covalent N-acetylmuramyl-dipeptide/LHRH adduct derives from an immune response to the hormone or fi-om occupancy of the LHRH receptor containing cells with immunoadjuvant moieties and correspondingly, firom an immune response to these cells.
A)
CH3CH ^fHAc C0-Ala-D-Glu-NH2
O
NH-(CH2)2-S-t--\ I N-(CH2)2-CO.[Moxl5].HG-[2.l7]
B) CH2OH )H HO I CH3CH
NHAc
C0-Ala-D-Glu.NH2 /9 LLys-NH-(CH2)2-S-r^ I N-(CH2)2-CO-[Moxl5]-HG-[2-17] CH3-(CH2)i4-CO "-^ O
Fig, 12, Chemical structures of the (A) N'Ocetymuramyl'dipeptide'Cystecanine/ and IB) lipo-N'aceytlrrairamyl'tripeptide'Cysteamine/[MQx^^]-gc^ 12-17] adduct
929 4,2. T- and B-Cell Epitope Chimeras One of the main disadvantages of the peptlde/carrier conjugate approach for antipeptlde immunizations Is the possible carrier-induced B-cell epitope suppression (176-178). the expression of new antlgen-imspecific immime epitopes related to the crossllnklng structures (49) and the reduced conformational space of the antigenic peptides on the carrier surface (177.179.180). Moreover, in view of the development of synthetic vaccines the choice of the carrier is extremely difficult. Taking into account the current knowledge about the mechanisms of Immune responses which require for both humoral and cell-mediated immimity the cooperation and communication between B and T-cells. it was suggested that Band T-cell epitopes need to be linked in order for T-cells to provide cognate help for B-cell activation and antibody production. On the basis of these concepts there have been numerous reports showing the successful construction of synthetic immunogens by the simple combination of well-defined T-cell determinants and B-cell epitopes. For this purpose different approaches have been used as linear polymerization of peptides (181-183). copol)niierization of B-cell and T-cell determinants by bifunctlonal crossllnklng reagents (184). or by collnear synthesis of B-cell epitopes with "natural" or with "foreign" T-cell epitopes, whereby a surprisingly strong effect on the specificity of the antibodies was found to derive from the orientation of the epitopes in the synthetic constructs (185-191). Moreover, it has recently been demonstrated that covalent linkage of the B- and Tcell epitopes does not represent an essential prerequisite for a good humoral response (192.193). The conformational preferences of the B-cell determinants in these chimeric linear peptides may play a critical role in the process of Induction of peptide antibodies of predetermined specificity, particularly if crossreactlvities with the parent protein is the goal of the immunization. These aspects have found only recently the due attention (194-199). The experiences gained in this field and from the de novo design of miniproteins opened new rational synthetic approaches for the synthesis of conformationally stabilized immunogens. In a long-term study Kaumaya (200) investigated this conformational aspect via the synthesis of topographic detennlnants by preserving maximal sequence homology to the native protein antigenic site in order to retain the functional specificity, and by introducing artificial mutations to facilitate a folding of the synthetic construct into a conformation mimicking the antigenic protein surface. For the conformational stabilization of the constructs secondary structural motifs known for proteins such as a-helices. p-sheets. p-tums and loops of the antigenic sites of the C4 isoenzyme of LDH as model protein were engineered to fit into known stable supersecondary structural motifs as ap. pap, papa or 4-a-helical bundles.
930 The Immunogeniclty of these constructs and the specificity of the antibody responses confirmed the validity of this promising approach (199.201.202). Besides the reconstruction of these natural subdomains of proteins as Immunogens it has been proposed to use synthetic templates for the design of conformationally stabilized fuUy synthetic miniproteins as potential immunogens has also been proposed. This concept foresees built-in folding devices for the induction of protein-like folding units, i.e. the employment of polyfimctionalized synthetic scaffolds on which peptides of specific conformational preferences are linked in a manner to stabilize via interchain interactions three-dimensional arrays of well-defined functionality (203). This template-assembled synthetic protein fFASP) approach, pionieered by Mutter (204). allowed for interesting progresses even in immunology (205). As alternative to these fully, more or less rigid synthetic templates used in the de novo design of proteins, we have introduced the concept of a natural template and for this purpose we have chosen the hinge segment of immunoglobulins (206,207). 5. Hinge-Peptide Based Fully Synthetic Immunogens 5.1. Hinge Segment oflgGl As known from x-ray crystallographlc studies on several antibody molecules and related fragments the two heavy and two light chains of immunoglobulins are
switch peptides Paratop
N-Termini
C-Tennini Fig. 13. Schematic drawing of the IgGl structure.
931 folded into domains which are arranged in pairs interacting by non-covalent forces except the CH2 domain of the Fc portion (208): interchain disulfide bridges provide further stability to these complex molecules (Fig. 13). In human IgGl the two heavy chains are linked by two proximal disulfide bridges in a portion of the molecule which connects the two Fab arms with the Fc segment. Because of the unique spatial structure and functional properties of this portion of the molecule it has been named the hinge segment. This segment can be divided into three regions: the upper hinge, the core and the lower hinge as shown in Hg. 14 (209210).
F(ab22
^
^ fak
.FcgJ ^^
E&
H-Glu-Ppo-Lys-Ser-Cys-Asp-Lys-Thi^Hb-Thr-Cys-Pro-Pi^Cyf-PitHAJa-Pro-Glu-Leu-Leu-Gly-Gly-Pn^ H-Glu-Pro-Lys-Ser-Cys-Asp-Lys-Thr-Hb-Thr-Cyi-Pi^PnHCys-Pro-Ala-Pro-Glu-Leu-Leu-Gly-Gly-Pro-OH 216 225 v-1 , ' 23« core
^
^^ upper hinge
^.^ middle hinge
.
^
lower hinge
genetic hinge
Fig. 14. The amino acid sequence of the human IgGl hinge segment with indication of the enzymatic cleavage sites and the resulting fragments.
The upper hinge includes the amino acids from the end of the CHI domain to the first residue in the hinge that restricts free motion (the first cysteine that forms an interheavy-chain bridge). In the x-ray structure this upper hinge folds into a oneturn helix with little inherent stability and full solvent accessibility (Fig. 15). In fact, the crystal structure of the trypsin generated Fc(t) (211) and comparative NMR analysis of various enzymatic hinge fragments of IgGl (212.213) as well as of i3C-labelled mouse IgG2a (214) indicate a high degree of flexibility for this portion of the heavy chain which correlates well with the known segmental flexibility responsible for the functional movements of the Fab arms. Interactions between the CHI domains and the upper hinge are absent in the SD-structure and thus, the spatial structure of the Fab fragments is fully retained upon enzymatic removal of the hinge segment. The rigid core hinge contains the two interheavychain disulfide bridges and it is folded into two parallel disulflde-linked poly-(Pro)II heUces (Fig. 15).
932
F^. 15. Stereoview of the SD-stmcture of the upper and core hinge as determined by x-ray analysis. The lower hinge connects the rigid core to the CH2 domain. The crystal structure of Fc(t) has shown that the two CH3 domains pair tightly in lateral contacts whereas no interaction takes place between the two CH2 domains (211). This would suggest an extended and to some degree rigid structure for the lower hinge. NMR analysis of the enzymatic IgGl fragments (213) led to the conclusion that the structure of the lower hinge, i.e. of the segment Pro230-Leu^^. in the doublestranded hinge fragment Lys222-Leu234^ m F(ab')2 and Fc(t) is essentially the same as that in the intact IgGl; thereby an extension of the poly-(Pro)-II helix from the core us. the lower hinge was suggested. More recent ^^c-NMR analysis are more supportive for high degree of flexibility of this lower hinge at least in mouse IgG2a (214). From these data it was concluded that the cyclic portion of the hinge segment although located in an overall flexible region, acts like a pivot for the functional movements of the Fab arms and Fc portion (215). Because of these peculiar properties a synthetic replicate could function as as suitable template in synthetic more or less constrained miniproteins as protein surface mimetics and thus, as potential immunogens. Essential premise for such an application was sufficient stability of the parallel disulfide linked cyclic peptide towards disproportion ation into monomers, antiparallel dimers as well as oligomers under the acid or basic conditions required in vairlous synthetic steps as well as under the physiological conditions of in vivo biological and immunological experiments. On the other hand, sufiiciently fast enzymatic processing is also required in the antigen presenting cells too, if an immunogenic construct is the goal.
933 5.1. Synthesis and Conformation of the IgGl hinge fragment 225-232/225'-232' Taking into account the results of the NMR analysis on enzymatic hinge fragments by Ito and Arata (212) and the fact that preliminary synthetic studies on the cyclic bis-cystinyl-octapeptide [H-Cys-Pro-Pro-Cys-OH]2 were not as satisfying as expected, an IgGl hinge-fragment was selected with N- and particularly. C-terminal exocyclic extensions as shown in Fig. 16 (216) in order to avoid in first instance the known proximal endgroup effects on the stability of biscystinyl structures (217). H-Thr-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH
I
I
H-Thr-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH
Fig. 16. Hinge fragment 225'232/225''232'ofhumanIgGl.
The synthesis of the bis-cystinyl-octapeptide dimer in parallel alignment as present in the native human IgGl was achieved by the use of a selective cysteineNps-Thr(tBu)-Cys(Acni>Pro-Pn)-Cys(StBu)-Pn>-Ala-Pn>-OH 1 tributylphosphine H-Thr(tBu)-Cys(Acin)-Pn>-PrQ-Cy$-PrtHAIa-Pro-OH
I
j Boc-N=N-Boc H-Thr-OH Boc-N-^fH-Boc +1
H-Thr(tBu)-Cy$(Acm)-Pro-Pro-Cys-PrcHAIa-Pro-OH H-Thr(tBu)-Cy$(Acm)-Pro-Pro-Cys-Pr<>-Ala-Pro-OH 1) I2 / 80 % acetic acid 2)TFA H-Thr-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH H-Thr-Cys-Pro-Pro-Cys-Pro-AIa-Pro-OH
+ H-Thr(tBu)-Cys-Pro-Pro-Cys(Acin)-Pro-Ala-Pr(>-OH
H-Thr(tBu)-Cys(Acm)-Pro-Pro-Cys-Pro-AIa-Pro-OH HO-Pn>-Ala-PrQ-Cys(Acm>-Pro-Pro-Cy$-Thr(tBu)-H 1)12/80% acetic acid 2)TFA H-Thr-Cy»-Pro-Pro-Cys-Pro-Ala-Pro-OH HO-Pro-Ala-Pro-Cys-Pn>-Pn>-Cys-Thr-H
Fig. 17. Synthesis of the hinge-peptide 225-232/225''232' in parallel and antiparallel alignment by selectiDe disulfide bridging methods.
934 pairing strategy based on the two S-tert-butylthio and the S-acetamido cysteine protecting groups. As outlined in fig. 17, the choice of this protection scheme allowed for a selective two-step disulfide bridging of the bis-cysteinyl-monomer to the cyclic structure (216). Upon reductive cleavage of the disulfide-type protecting group with tributylphosphine (218) and mild oxidative interchain disulfide formation with azo-dicarboxylic acid di-tert-butyl ester (219). the second disulfide bridge was generated by the iodine method in acetic acid. Exploiting the concept of high dilution in order to avoid oligomerization almost exclusively the desired parallel dimer was obtained. A similar route was applied for the preparation of the antiparallel dimer. In the iodine-oxidation step a surprisingly high percentage of parallel dimer was obtained, the formation of which could only be explained by an attack of the intermediate sulfenyl-iodide on the existing disulfide. Although this type of reaction is known from sulfur chemistry (220), to our knowledge it has not yet been observed as side reaction in peptide chemistry. Its occurrence in the present case at well-detectable extents can rationally be explained only if the hinge-peptide exhibits a high inherent preference for a parallel alignment. In order to analyze this aspect we have studied into details the oxidation of the monomeric bis-cysteinyl-peptide as well as the disproportionation of the antiparallel into the parallel dimer. Object was to determine the thermodynamically most favored product distribution and thus, the stability of the parallel hinge peptide in view of its use as a core structure for the assembly of multichain constructs (221). Studies on air oxidation of the bis-cysteinyl-peptides H-Cys-(Gly)n-Cys-OH with n = 0 to 15, have clearly shown that the nature of the oxidation products is largely dictated by the probability of collision of the thiol groups when n > 4 (222-225). Thus, under conditions of high dilution (10"3 to lO'^ M) intrachain disulfide bridging with formation of > 20-membered rings was in all cases predominant. Conversely, for n < 3 the peptide chain is not sufficiently flexible and simple statistical theory cannot be applied anymore. Besides intrachain-bridged monomers, formation of dimers and oligomers is expected to occur. Thereby sequence-dependent conformational preferences can play a decise role, particularly if oxidation is performed in aqueous media where thiol disulfide exchange reactions is shifting the product distribution towards the thermodynamically most favored structure. This has recently been fully confirmed in a detailed study on peptides containing the Cys-X-Y-Cys sequence and related to the active sites of thiol-protein oxidoreductases (226.227). In the heavy-chain hinge fragment 225-232 the intervening sequence between the two cysteine residues is Pro-Pro and the native structure of this portion of the IgGl molecule consists of a parallel dimer. By estimating for the Cys-Pro-Pro-Cys portion the propensity for chain reversal, i.e. for p-tum formation, according to
935 H-Thr(tBu)-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH CD oxidatian
H H
J '
(2)
j 1
OH OH
HO—I H '
(3)
1 '
H OH
H-
(4)
-OH
+ oligomers
Fig. 18, Air-oxidation of the bis-cysteinyl-octapeptide expected products.
225-232 of human IgGl and
Chou and Fasman (228) a p-bend potential of = 1.9x10*4 was derived which is significantly higher than the average potential of 5.5x10"^. Nevertheless airoxidation of the bis-cysteinyl-octapeptide 1 of Fig. 18 at pH 6.8 under exclusion of heavy metal catalysis and at a concentration of 3x10"^ M led to the surprising product distribution of 90:8:2 molar ratios for the compoimds 2:3:4 (221). Only in absence of water and thus, by preventing thiol disulfide exchange reactions, oxidation with azodicarboxylic acid di-tert~butyl ester was found to generate the statistically more or less expected ratios of 12:10:78 for 2:3:4. Addition of thioredoxin to the oxidation experiment in aqueous solution was found to significantly enhance the rate of air-oxidation, but without any effect on the final product distribution. These results as well as the observation that the antiparallel hinge dimer is converted again to a product distribution of 90:8:2 for 2:3:4 when incubated with the bis-cysteinyl-peptide 1, clearly confirmed the thermodynamic preference for the parallel form of the hinge-peptide (221). Formation of the correct disulfide pattern in proteins occurs concomitantly with acquisition of the correct folded form and it is driven by the thermodynamic stability of the native 3D structure. In the initial stages of protein folding processes thermodynamically stable local structures may play an important role (229-232). Thereby short range interactions are essentially implicated to promote stable core structures around which the rest of the protein chain will fold. These sequence-specific short range interactions may suffice for folding of isolated protein fragments into stable native-like structures as weU demonstrated with the bovine pancreatic trypsin inhibitor mono-cystinyl fragment 20-33/43-58 (233).
936 Similarly, sequence-specific information must be the driving force for the observed predominant parallel alignment of the hinge-peptide 225-232/225'-232' In aqueous solution. Despite the relatively small size of this proteinfiragment.the parallel alignment corresponds to an energetically highly favored structure and may represent a stable subdomain which could play an important role as
Ftg. 19. CD spectra of [H'ThrltBu)-CyS'PrO'Pr(>Cys-Pro-Ala'PrO'OH]2 (curve 1), [HThr(tBu}'Cys(Acw)'Pro-Pro-CyS'PrO'Ala'Pro-OH]2 (curve 2) and H-TtuitBu}' CyS'ProPrcyC^S'PrO'Ala'Pro-OH (curve 3). The CD spectrum of the unprotected hinge-peptide 225-232/225''232', Le. [H-Thr-Cys-Pro-Pro-CysPro-Ala-PrO'OH]2. is identical to curve 1. "chain folding initiation structure" (234) in the assembly of the immunoglobulins. This assumption is supported by the observation that the initial step in covalent interchain crossllnking of immunoglobulins is the disulfide formation between a nascent heavy chsdn and a completed one (235). In order to confirm the protein subdomaln-llke character of the hinge-peptide 225-232/225'-232' its preferred conformation was analyzed by CD and vl spectroscopy (88.236) and compared to its 3D-structure in the native proteiii. a«? mentioned above formation of the correct disulfide pattern in this proteLu fragment was expected to occur as consequence of a thermodynamically highly favored conformation. The CD spectra of the monomeric bis-cysteinyl-compound H-Thr(tBu)-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH. of the dlmerlc mono-cystinyl-
937 compound [H-Thr(tBu)-Cys(Acm)-Pro-Pro-Cys-Pro-Ala-Pro-Pro-OH]2 as intermediate of the cysteine pairing process and of the parallel dimers [HThr(tBu)-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH]2 and [H-Thr-Cys-Pro-Pro-O^s-Pro-AlaPro-OH]2 in aqueous solution at pH 7.0 are reported in Fig. 19. A comparison of these CD spectra indicate a gradual blue shift of the negative maximum with an abrupt increase of the CD intensity by a factor of 2 upon ring closure. Similarly, the weak positive maximum around 225-230 nm appears only upon formation of the second disulfide bridge. The CD spectrum of the parallel dimer with its strong negative maximum at 198 nm and the weak broad positive extremum at 228 nm is strongly supportive of a poly-(Pro)-II helical conformation (237). From comparison of the CD spectra of the synthetic precursors with that of the blscystinyl-compounds no clear evidence was obtained for the existence at the conformational equilibria of correctly prefolded conformers of the linear or intermediate mono-cystlnyl compound as driving force for the preferred parallel alignment in the oxidative folding of this protein subdomaln. However. CD spectroscopy is not sufiBclently sensitive to detect small populations of conformers suitable for favoring this process. Nevertheless, the observation that the Thr sidechain protecting group was without any detectable effect on the CD pattern was useful, since for the construction of complex molecules on the hinge scaffold this protection is necessary. These findings were further confirmed by reoxldatlon experiments on the deprotected llnesir bis-cystelnyl-octapeptlde which led to the identical results as described above for the threonlne-protected derivative (unpublished results). 2D-NMR measurements on the synthetic hinge peptide 225-232/225'-232' in DMSO combined with MD calculations allowed to define more precisely the preferred conformation of this double-stranded protein fragment (236). As shown in Fig. 20. the core portion consists of two parallel poly-(Pro)-II helices connected by the disulfide bridges of identical chirallty as in the x-ray structure of the Kolprotein (215), i.e. left-handed for Cys226/Cys226' {^^^ = -100.5') and right-handed for Cys229/cys229' (5^gg = 91-), These findings fully confirm the interpretation of the CD spectra. Superimposition of the NMR and x-ray structure of the cyclic portion gives a rms value of 1.07 A which indicates that the 3D-structure of the core hinge-peptide in solution is practically Identical to that of the parent protein in crystals. The two strands are well separated without interchain interactions due to the extended side chain conformations of the cystines which act as spacers. Nonetheless the mobility of the core structure is strongly limited as a result of the restricted conformational space of the prolines. The occurrence of short range interactions are generally believed to be responsible for the onset of local structure. These Interactions which could favor
938 the observed correct oxidative folding of the hinge-peptide are absent and therefore, this role is possibly exerted by the C-terminal extensions. The flexibility of the exocyclic portion in N-terminal direction towards the Fab. which is responsible for the movements of the Fab arms in the intact protein, cannot be
Fig, 20. Stereoview of the core hinge SD-structure as determined by NMR analysis of the synthetic hinge-peptide 225-232/225''232' in DMSO sobxtion. analyzed in the synthetic fragment due to the shortness of the extension. Although in direction Fc the mobility is increased, the preference for an extended conformation is retained; an extension of the double helical fold beyond the disulfide bridge could not be observed: Similarly, interchain interactions in this exocyclic portion were not detected in the NMR experiments. Because of the high stability of the hinge-peptide structure and its rigidity In the core portion, but flexibility of the exocyclic parts, this structural unit fulfils the requirements of a template for the construction of synthetic mlnlproteins as potential inmiunogens (206.207).
5.2. Synthesis and Conformation of Hinge-Peptide/Gastrin Chimeras Gastrin as endogeneous regulatory peptide is a self-antigen against which induction of Immune responses should be more difficult than against foreign antigens. It Is. however, well established that conjugation of self-antigens to immunogenic proteins generates the desired antibody responses. In the case of
939 gastrin this has been well documente using both conventional polyvalent (42.53) and the selective monovalent iso-1-cytochrome c/gastrin conjugate (67). Less successful were the attempts to Induce an immune response against this hormone by simple coupling of the gastrin molecule to lipophilic adjuvant molecules (175) or by its lipo-derivatization (unpublished results). Among mammalians the known gastrin sequences exhibit a highly conservative character as shown in Rg. 10. with an identical C-termlnal portion in all known sequences. Thus, interspecies sequence vsiriatlons should not be responsible for the observed immune responses in the experimental animals whilst such mutations were found to represent the main immunodeterminants in the highly conservative family of C3^ochrome c (238). Despite this expected immunological inertness of endogenous hormones, gastrin was selected as model antigen in our studies on fully s)nithetic inmiunogens for the following reasons: i) As a hormone gastrin has the ability to fold into a defined bioactive conformation at receptor level for its specific signal transduction. Conformational studies on gastrin in membrane and receptor mimicking environments led to the proposal of a potential bioactive structure of this hormone and to identify its biological relevance (82.83,239). ii) Immunological studies on gastrin/carrier conjugates clearly confirmed expression of a conformational B-cell epitope with antibody responses of gastrin receptor-like specificity, if sufficiently free conformational space is retained by the antigen in the conjugate to assume its preferred secondary structure (68). The main task of our model studies was to assess the biological and conformational properties of various hinge-peptide/gastrin constructs and to correlate these with the specificity of the antibody responses if an immune response ccould be elicited by such simple chimeric compounds. The presence of tryptophan residues in the gastrin sequence severely impairs selective disulfide bridging of linear bis-cysteinyl-hinge-octapeptide/gastrin chimeras by the approach outlined in Fig. 17 for the hinge-peptide 225-232/225'232'. since the intermediately formed sulfenyl-iodide is known to react with the indole function of tryptophan to produce thio-indole derivatives as noxious side products (240). On the other hand, synthetic steps on double-stranded cystine peptides are difficult to control analytically. Therefore, air-oxidation was chosen for the production of parallel hinge-peptide/gastrin constructs in view of the excellent results obtained with the hinge-peptide itself (221) and on the assumption that the subdomain character is retained in the colinear
940
hlnge/gastrln peptides despite the strong sequence dlflferences if compared to those adjacent to the hinge in the IgGl molecule. As outlined in Fig. 21. the hlnge-peptlde compounds extended N-termlnally with
2xa-OH+
H-Thr(tBu)-CY$ uhCYS 1 H-Thr(tBi
Cvs cvs I
OH
H-Thr(tBu)-Cvs
OH
HO
— CCyk yi
Cy,s
-OH
Cys-Thr(tBu)-H ^-I
M.A. xa-Thr(tBu)-C);s
Cvs
2xa-0H +
OH
xa-Thr(tBiu ) - c y s — C y s — OH
M.A.
xa-Thr(tBu)-CyjS
Cys
HO
Cyi.Thr(tBu).xa
Cyl
l)HCI/2-meihyUndole 2) TFA/2-methylinclole/anisolc 3) chromatography
I 1 '
OH
DHCl^-mclhylindolc 2) TFA/2-mcthylindolc/ anisole 3) chromatography
xb-
Cys
Cys
OH
xb-
-Cvs
Cys-
-OH
xb-
Cys
Cys
OH
HO-
-Cys—Cys-
-xb
J
I main product
minor side product
air oxidation xb-
-Cys-
Cys-
-OH
t tributylphosphinc xb-
- Cys(StBu)
Cys(StBu)
OH
4 1) HCl/2-mcthylindolc 2) TFA/2-methylindolc/anisole 3) chromatography xa-Thr(tBu)-Cys(StBu)
Cys(StBu)
OH
Cys(StBu)
OH
A T M.A. xa-OH +
H.Thr(tBu)-Cys(StBu)
xa = Nps-Lcu-[GIu(OtBu)l 5-Ala-Tyr(tBu)-Gly-Trp-NIe-Asp{0tBu)-Phcxb = H-Lcu-lGlulj-Ala-Tyr-Gly-Trp-NIe-Asp-Phc-
Fig.21. Scheme for the selective synthesis of the lN\e^^]-gastrinrl5'17]/hingepeptide compound in parallel and antiparallel alignment and of the parallel compound by air-oxidation.
941
gastrin in parallel and antiparallel alignment were synthesized by direct acylatlon of the side-chain protected double-chain hinge-peptlde 225-232/225'-232' followed by deprotectlon and purification (241). EfBlclent scavengers like mercaptanes were avoided in the deprotectlon step because of a possible scrambling of the disulfide bridges; nevertheless the compounds could be isolated as structurally well defined reference substances for the analytical control of the air-oxidation experiments. For latter reaction the fully deprotected colinear gastrln/bls-cysteinyl-hinge-peptlde was incubated at pH 6.8 and 10'^ M concentration under air-ojgrgen. With the help of the parallel and antiparallel reference compounds hplc allowed to establish a surprisingly high preference (>90%) for formation of the desired parallel dlmer which could then be isolated in good yields as structurally well characterized compound.
Fmoc-Thr(tBu)-Cy$—Cys—OH
I
I
+ 2H-ya
Finoc-Thr(tBu)-Cys— Cys—OH
I M.A. Fmoc-Thr(tBu)-Cys--Cvs—ya Finoc-Thr(tBu)-Cys--Cys—ya 1) TFA/2-methylindoIe/anisole 2) pipcridinc 3) chromatography H
Cys—Cys—yb
H
Cys—Cys—yb
f
H
l) air oxidation 2) chromatography Cys—Cys—yb 1) TFA/anisole/l,2-€thandithioIe 2) pipcridinc 3) iribuiylphosphine
Fmoc-Thr(tBu)-Cys(StBu)-Cys(StBu)—ya A DCC/HOSu Fmoc-Thr(tBu)-Cys(StBu)-Cys(StBu)—OH
+ H-ya
ya = -Gly-Pro-Trp-Lcu-(Glu(OtBu)] 5.AIa-Tyr(tBu)-Gly-Trp-Nlc-Asp(OtBu)-Phe-NH2 yb = -Gly-Pro-Trp-Leu-IGIu] 5-Ala-Tyr-Gly-Trp-Nlc-Asp-Phc-NH2
Fig. 22. Scheme for the synthesis of hinge-peptide/[Nle^^]-gcistrin'[2'17] in paraUel alignment.
942 Following a similar route the effect of C-termlnal extensions of the hlnge-peptlde with (Nlei5]-gastrln-I5-17] on the product distribution in air-oxidation experiments was also analyzed, and again full retention of the high preference for the parallel alignment was observed. Thus, a relatively facile synthetic access to the hlnge-peptlde/gastrln constructs was uncovered. As potential Immunogens for gastrin besides the hlnge-peptide/[Nlel5]-gastrln-[517] compound also the related INlel5j-gastrtn-[2-17J adduct was synthesized (79) In order to analyze the effect of a larger spacing between bloactlve core of the hormone and scaffold in terms of free conformational space and thus, of biological and immunological properties. For the synthesis of this chimeric compound the N-terminal threonine residue was protected as Fmoc derivative as outlined in Fig. 22 for two specific reasons: I) Acidolytic deprotectlon of the side chain-protected hlnge-peptlde 225232/225'-232' was found to be accompanied by formation of a side product at extents up to 6% (216). Detailed studies on this side product and its orlgine indicated that the N-terminal threonine residues, unexpectedly, undergo N«trifluoroacetylation by exposure to trifluoroacetlc acid even in presence of water to hydrolyse trifluoroacetc anhydride as potential contaminant, via intra- and/or intermolecular 0->N shift of the Intermediately formed Otrifluoroacetyl-threonine derivatives (242). II) It was known from previous studies that llpophillzation of antigens enhances their immunogenlcity (160). The bulky and highly hydrophobic fluorenylmethyl group at two adjacent positions of the molecule was expected to possibly play this role in the case immunizations with the underivatized compound would fail. Previous immunological studies on big-gastrin/BSA conjugates (big gastrin = gastrin-34 which contains as C-terminal portion gastrin-17) have shown that besides the dominant antigenic site located in the C-terminal part of the molecule, a second antigenic determinant is centered around the Pro-Pro-His sequence in the N-terminal portion of the molecule (243). In order to analyze the feasibility of bivalent immunogens with the hinge-template, the sequence 1-14 of big-gastrin was incorporated at the N-termini and the sequence 2-17 of gastrin (= sequence 19-34 of big-gastrin as Nle^^-analogue) at the C-termini of the hinge-peptide (79). Thereby the synthetic route via the double-ch2Lin hinge-peptide served again to produce the reference substance in parallel alignment (Fig. 23). The linear fully protected octatriacontapeptide was assembled by successive fragment condensation steps. Subsequent removal of the acid-labile side-chain protecting groups was followed by reductive cleavage of the
943
Fmoc-Thr(tBu)-Cys— Cys—ya Fmoc-Thr(tBu)-Cys— Cys—ya 1 piperidine H-Thr(tBu)-Cys—Cys—ya 2 xa-OH
+
I
I
H-Thr(tBu)-Cy$—Cys—ya j M.A.
xa-Thr(tBu)-Cys—Cys —ya xa-Thr(tBu)-Cys—Cys—ya I l)TFA/2-mcthylindole/amsoIc ^ 2) chromatography xb
Cys—Cys—yb
xb
Cys—Cys—yb k 1) air oxidation I 2) chromatography
xb
Cys— Cys—yb A 1) TFA/anisolc/l,2-cthandithiol I 2) tribuiylphosphinc
xa-Thr(tBu)-Cys(StBu)—Cys(StBu) — ya I DCC/HOSumOOBt xa-OH + H-Thr(tBu)-Cys(StBu)—Cys(StBu) —ya T piperidine Fmoc-Thr(tBu)-Cys(StBu)—Cys(StBu) — ya
xa = Pyr-Lcu-Gly-Pro.Gln-Gly-Pro-Pro-His-Lcu-Val-AIa-Asp(OtBu)-Proxb == Pyr-Leu-Gly-Pro-Gln-Gly-Pro-Pro-His-Leu-Val-AIa-Asp-Proya = -Gly-Pro-Trp-Leu-(Glu(OtBu)) 5.Ala-Tyr(tBu)-Gly-Trp-Nlc-Asp(OtBu)-Phe-NH2 yb = -Gly-Pro-Trp-Lcu-[GluI ^-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NHj
Fig. 23, Synthetic route for big-gastrin-ll'MJ/hinge-peptide/ir^^^J-gc^ S-tert-butylthlo groups, and the resulting bis-cysteinyl-peptide was then exposed to air-oxidation. As expected from the model studies, the parallel dimer was formed almost exclusively, despite the large size of the compound which as 76-
944 membered construct represents a de novo designed synthetic protein of remarkable size.
I
I
« ^
260 X(nm)
Fig, 24. CD spectra of the htnge-peptide (curve 1), [Nle^^ygastnn-[2'17] (curve 2) and hinge'peptide/[Nle^5]'gastnn-l2'17] (curve 3) tn aqueous solutiorh The computed spectrum (curve 4) was obtained by addition of the CD curves 1 and 2. The CD spectrum of hinge-peptide/lNle^^J-gastrin-lS-l?] is almost superimposable to that of curve 3.
The gastrin receptor-like specificities of the antibodies raised with the monovalent iso-l-cytochrome-c/gastrin conjugate (68) raised the question of whether the gastrin molecule folds on the surface of the carrier protein into its preferred bioactive structure as a result of the protein environment and thus is recognized as a preformed conformational epitope by the B-cell surface immunoglobulins or whether the folding of the gastrin moiety occurs upon interaction with the B-cell receptors. This important question could not be answered definitely by the CD measurements on the cytochrome conjugate (88). However, more precise information was expected from the low mass and structurally defined hingepeptide/gastrin chimeras. The extension of the double-chain hinge-peptide 225-232/225'-232' at the two Ctermini with the INle 15].gastrin sequences 5-17 and 2-17 leads to CD spectra (Fig. 24) in aqueous solution which are practically identical to those computed by
945 linearly combining the dichroic contributions of the two components (88). Similarly, an additional extension of the hinge-peptide/INle^5j.gastrin-[2-171 at the two N-termini with the big-gastrln sequence 1-14 leads to a CD spectrum in water that corresponds to the computed spectrum. In these spectra clearly the dominant contribution to the dichrolsm results from the double-stranded poly(Pro)-II conformation of the hinge-peptide core. From extensive conformational analysis of gastrin-17. big-gastrin and their related fragments by CD techniques it was concluded that in aqueous solution these peptides are essentially unordered (80). According to the dichroic properties of the chimeric compounds the gastrin-related components and the hinge portion retain their own conformational properties in water without reciprocal interferences and conformational restrictions. This fact could accoimt for the surprising observation
o
0
Fig. 25. Comparison of the CD spectra of hinge'peptide/lNle^^l'gastrin'12'17] (curve 1) and hinge-peptide/INle^^hgastrin-lS'l 7] (curve 2) in 95% TFE. that both N- and C-terminal extensions of the hinge sequence 225-232 with IgGlunrelated peptide sequences does not affect the intrinsic ability of the hingepeptide portion to fold correctly by alr-oxldation in aqueous solution into the parallel dimeric alignment.
946 Changes in solvent polarity, as obtained with aqueous TFE. have often been used to mimic the water-limited environment of biologically active peptides while interacting with target molecules such as their receptors. CD and NMR studies on gastrin and gastrin-related peptides and the corresponding biological data led to the assumption that the structure assumed by this hormone in TFE is of biological relevance (83,84). A conformational model was proposed consisting of two helical segments at the chain ends stabilized by mutual interactions in a Ushape mode. The CD spectrum of the hinge-peptide/[Nlei5]-gastrln-[5-171 in 95% TFE was found to be identical to that computed by the sum of the spectra of the constituent parts (Fig. 25). The CD spectrum of the hinge-peptide/[Nlei5]-gastrin[2-17]» however, dlifers from the computed spectrum by a significantly weaker dichroism in the 210-230 nm range with appearance of a broad negative band of weak intensity around 230-235 nm. The dichroism at 208 nm is retained suggesting helical structures as present in the gastrin molecule in TFE. The difference spectrum between the experimental and computed spectrum reminds the classical B-spectrum for type II p-tums (244). The sequence difference between the gastrin-[5-171 and gastrin-[2-17] chimeras is the extension of the exocyclic hinge portion Pro-Ala-Pro by the gastrin tripeptide Gly-Pro-Trp which leads to a sequence capable of folding into a poly-(Pro)-II helix, but also of high ptum potential. The CD spectra do not allow for a definite differentiation of these two possible conformational states of the spacer, although the difference CD spectrum is more indicative of a p-tum local conformation. The bivalent construct containing both the gastrin and the big-gastrin antigenic determinants exhibits dichroic properties in TFE which reflect in the N-terminal extensions the presence of p-tum type ordered structures (88) as already suggested by the CD studies performed on the synthetic big-gastrin fragment 1-20 (80). The results of the CD measurements strongly suggest that in the three hingepeptide/gastrin chimeras the gastrin components retain largely their ability to fold into gastrin-characteristic conformational states both in water and TFE, although quantitative evaluations cannot be drawn because of the limited accuracy of this analytical technique. Therefore more precise NMR analysis was performed and related data were used for molecular dynamics calculations (245). The final 3D-structures resulting from the molecular dynamics simulations are reported in Figs. 26 and 27. The most significant and surprising result was the loss of symmetry in both the gastrin-[5-171 and gastrin-l2-171 constructs regarding the gastrin moieties whereas the hinge-peptide core structure remaines fully constrained in its cyclic portion as already observed for the hinge-peptide
947 225-232/225'-232' itself (235). As shown In Fig. 26, in the gastrin-l2-17] dlmer at least one chain remains largely accessible in Its hairpin structure, whereas the second chain folds more tightly over its N-terminus and the spacer between gastrin and the cyclic portion of the hlnge-peptlde. This indicates that in the dimeric construct only one gastrin chain retains sufficiently ifree space to fold into its preferred conformational state and thus, to be recognized by the gastrin receptors or B-cell surface immunoglobulins.
C'terminus
N'terminus
Fig. 26. Stereoview of the starting conformation of hing€-peptide/[Nle^^]'gastTin-l2' l 7] (top) as modelledfromthe conformational states of the component parts in TFE and after 50 ps simulation (bottom). In the case of the shortened gastrin construct, i.e. hlnge-peptide/lNlel^J-gastrln[5-17], strong interferences of the peptide chains, particularly in their central portion, are observed which allow Just the C-termlnal part of one gastrin molecule to protrude into free space (Fig. 27). As a result of the packing of the two gastrin chains the accessibility and the conformational space is greatly restricted and should allow recognition of just the C-termlnus of one of the two gastrin moieties by target receptor molecules. For the molecular d3niamlcs simulations the
948 conformation of the single components In TFE was used to buUd the starting structures since according to the CD spectra these conformations are more or less retained in the chimeric compounds in TFE. The partial coUaps of these structures in the simulations confirm once more that TFE is a too strong a-helix inducer (246.247). The results of these conformational studies clearly reveal the disadvantages of template-assembled synthetic constructs in terms of interchain interferences. If defined secondary structural elements like ap, aa. PaP have to be stabilized by direct anchorage on the template, successful application of this approach has been repetedly reported (203). More globular ordered structures as in the present case need more conformational space which can be created by larger spacers. In view of this result it is not surprising that multiple attachment of protein fragments on a branched lysine carrier may trigger immune responses uncapable of recognizing the parent proteins (123).
C-terminus
N-terminus
i
Fig, 27. Stereoview of the starting conformation of hinge-peptide/lNle^^l-gastrin-fd' 17] (top) as modeUedfrom the conformational states of the conqx>nent parts in TFE and after 50 ps stmidation (bottom)
949 5.3, Bioacttvities of the Hinge-Peptide/Gastrin Chimeras The hlnge-peptide/gastrln chimeras contain the fully active [Nlei5]-gastrln-[2-17] and [Nlei5j-gastrln-[5-17] sequences (85) dimerlzed at their N-termini via the bridging sequence Pro-Ala-Pro-Cys Cys-Pro-Ala-Pro which according to the NMR studies on the hinge-peptide is restricted in its mobility, but in an extended secondary structure that should assure more conformational free space than present in most of the literature known dimeric hormone preparations (248-253). The gastric acid stimulatory potency of the dimeric gastrin constructs were compared with those of the parent gastrin analogs [Nlei^j-gastrin and [Nle^S]gastrin-[5-17] (245). As shown in Table 2, on molar basis the gastrin-[2-171 dimer was found to be as potent as the parent hormone: however, the hingepeptlde/gastrin-[5-171 dimer exhibited only a 7-fold reduced potency which is comparable to those of C-tenninal gastrin peptides, i.e. of the tetra- and
Gastrin compound
ED50 (pmol/kg)
IC50 (nm)
[Nlel5]-gastrin or [Nle 15]-gastrin-[5-17]
26.5 ± 10.0
0.4
hinge-peptide/[Nle 15].gastrin-[2-17]
23.5 ± 4.50
0.5
hinge-peptide/ [Nle 15] -gastrin-[5-17]
179.5 ± 10.0
30
hinge-peptide/des-amido-[Nle 15]-gastrin-[5-17]
-.-
-.-
[Nle 15]-gastrin-[5-17] /hinge-peptide
-.-
-.-
1
Table. 2, ED^gfa^ ^ ^^^ stimulation of gastric acid secretion in rats by gastrin and hinge-peptide/gastrin chimeras and their receptor binding affinities in parietal cells isolated from rabbit gastric fundus.
heptagastrin (254). The dimers lacking the C-terminal amide were inactive as expected from previous structure function studies on gastrin peptides since this structural element is known to represent a crucial factor for the hormonal activity of gastrin (254). These des-amidated compounds were also devoid of any antagonistic activity. This was further confirmed by their inability to displace the receptor-bound radiolabelled ligand from isolated parietal cells (see Table 2). In this receptor binding assay the gastrin-15-17) dimer exhibited again a remarkably lower affinity than the parent hormone, i.e. 1%, whereas the gastrin-[2-17] dimer
950 was recognized on a molar base with Identical affinity as the monomerlc form. In terms of gastrin equivalents the dlmerlc construct exhibits half the potency and receptor affinity of the monomerlc gastrin. This raises the question of whether both gastrin components are half as potent as the monomer due to mutual Interferences, e.g. sterlcal hindrance or reduced conformational space for optimal receptor Interaction, or whether In the dlmerlc construct only one gastrin moiety remains fully accessible with the second one restricted In Its conformational space, thus preventing a flip-flop-type receptor Interaction as one of the possible mechanisms responsible for the enhanced activities observed for several peptide hormone dlmers (253). Latter working hypothesis Is supported by the conformational studies discussed above which could also explain the low potency and receptor affinity of the gastrln-[5-17] dlmer. If the restricted conformational space Is responsible for the observed biological properties a similar pattern was expected from Immunization experiments where again the dlmerlc compounds have to be recognized by the B-cell surface receptors In a more or less accessible form.
5,3 Immunogenicities of the Hinge-Peptide/Gastrin Chimeras Standsird protocols were used for Immunizing guinea pigs with the hingepeptide/gastrin compounds and with the monovalent lso-1-cytochrome-c/gastrin conjugate as reference (245). The titers of anti-gastrln antisera were determined by standard ELISA procedures and are listed in Table 3. Significant anti-gastrin antisera titers were obtained which were comparable to those raised by gastrin/protein conjugates indicating that the synthetic constructs behave as fully competent Immunogens. Interestingly the immune response against the hinge-peptide portion was found to be very weak or none. In order to avoid misleading results deriving from nonadsorption of the hinge-peptide or from inadequate exposure of this peptide on the polystyrene surface a hinge-peptideA^P-[4-18] construct was used as antigen in the ELISA. With the chimeras containing N-terminally the antigenic peptides and the sole threonine residue as spacer very weak crossreactivities with the hingepeptide antigen were detected possibly resulting from the generation of an overlapping, although weak new B-cell epitope. This was confirmed by the antisera raised against the [Nlel5]-gastrin-[5-17]/hinge-peptide which did not crossreact with intact human IgGl although weakly recognizing the hinge-peptide antigen. Unmodified gastrin was found to be non-immunogenic in guinea pigs as previously already observed in rabbits and mice where even conjugation of gastrin-l2-17] with the muramyl-dipeptide or a lipophilic muramyl-tripeptlde
951
Immunogen
Anti-gastxin antibody titers* i
hlnge-peptide/[Nlel5]-gastrin-[2-17]
3.64
hinge-peptlde/[Nlel5)-gastrin-[5-17]
4.02
hlnge-peptide/des-amldo-INle 15].gastrln-[5-17)
3.45
[Nle 15j .gastrln-[5-17)/hlnge-peptlde
3.20
big-gastrin-11-14] /hinge-peptide/ [Nle ^ ^j -gastrln-I2-17]
3.55
[Nle^Sj.gastrin iso-l-cytochrome-c/gastrin-[2-17J
-.3.37
Table 3. Immune responses in guinea pigs to hinge-peptide/gastrin chimeras and to iso-l'Cytochrome-c/gastrin for comparison; *) the titers are averages of the antisera of three animals.
derivative failed to induce immunogenicity (175). Consequently, the observed immunogenlcity of the hinge-peptide/gastrin chimeras has to arise primarily from the hinge-peptide portion, although the function of this moiety at cellular level is not yet clear. A possible pathway could derive from reaction of the disulfide bonds of the synthetic chimeras with cysteine-containing endogenous proteins which would present the hinge-peptide/gastrin as macromolecularized antigen to the immune competent cells. The stability of the hinge disulfides which should be in the order of that of intact IgGl, makes this hypothetical pathway rather unlikely. Moreover, pulse/chase experiments on human Epstein-Barr virus-transformed B-cells have shown that the hinge-peptide/[Nlel5]-gastrin-[2-17J compound is internalized and bound to MHC class II from which it is released in acidic media in its intact form as monitored by a comparative hplc analysis with a fluorescent probe of the hinge-peptide/gastrin compound (245). The recovery of the intact dimer upon internalization indicates a remarkable stability of the bis-cystinyl structure in the in vitro experiments, although quantitative estimations are difficult in such pulse/chase experiments. On the other hand, the high degree of paralleUsm between the receptor binding assay in vitro and the gastric acid stimulation potency determined in vivo excludes that reduction, disproportionation or thiol
952 disulfide exchange reactions with serum or cellular proteins is occurring at extents capable of sensibly affecting the results.
Fig. 28, Stereoview of the MHC class U protein with the bound hinge-peptide 225232/225''232'as resulting from modelling experiments and energy mintmization of the complex. A) Lateral view; B) view along the binding cleft. The MHC class II molecules bind T-cell epitopes apparently with less defined restriction regarding the peptide size than the MHC class I molecules and as known from the x-ray crystallographic studies preferentially in an extended
953 conformation (43,44). In this context the hlnge-peptlde portion In the gastrin constructs Is folded in an extended poly-(Pro)-II conformation, but as a dlsulfidellnked dlmer. The pulse/chase experiments and the observed immunoglobulin Isotype switch to IgG would suggest that the hlnge-peptlde portion Is playing the role of an efficient T-cell epitope. Modelling experiments performed with the use of the x-ray structure of the MHC class II protein DRB1*0101 (30.31) showed an excellent fit of the double-stranded hlnge-peptlde with a surprisingly good overlapping of the peptide backbone of one strand with that of the bound and cocrystalllzed Influenza virus peptide (Fig. 28). Additionally, the second strand of the hlnge-peptlde is only minimally protruding out of the binding pocket whereby the double stranded molecule apparently occupies most of the binding pocket with contacts to the bottom and lateral helices in Van der Waals dlstancles. In crystals of the IgGl Kol-protein the hinge region is in an extensive close contact with the hypervariable segments of a second molecule leading to well-defined and strong densities for all the residues Involved. As the loss of surface area of the hypervariable segments in this contact is 1314 A^ the observed Interaction in the crystals has been classified as "antigen-like" binding of the hinge segment to the antigen-binding site of the IgGl (215). Moreover, it is well known that natural IgGanti-F(ab')2 antibodies belong to the immune repertoire of healthy individuals and represent potent immunoregulatory molecules (255-257). It has been suggested that some of the £intl-F(ab')2 antibodies recognize conserved domains of the F(ab')2 portion of IgGl (255.258). Recently, with the help of the synthetic hlngepeptide 225-237/225'-237' which contains an extended lower hinge in comparison to the hinge peptide used for the construction of the gastrin chimeras, it has been demonstrated that these IgG-antl-F(ab')2 antibodies bind strongly to this double-chain hinge fragment as a conformational epitope expressed or stabilized by extension of the lower hinge sequence to position 237 (259). CD spectroscopy allowed to detected in this extended hinge fragment a stabilized exocyclic P-tum which could possibly contribute to the expression of the epitope. Since all molecules Involved in Immune regulation, e.g. surface B-cell receptors, MHC molecules, Fcy-receptors, rheuma factors, are characterized by a similar structural assembly, the hinge segment could possibly represent an universal type epitope and thus, be Involved in various pathways as recently demonstrated in the case of the Fey receptors (260) and rheuma factors (259). In cell culture experiments the lgG-anti-F(ab')2 antibodies are known to suppress B-cell proliferation (257). Correspondingly, treatment with the hingepeptide/gastrin chimeras could block these IgG-anti-(F(ab')2 antibodies in the serum and as a consequence, the B-cell Immune response would be stimulated. Alternatively, as reported by Roosnek and Lanzavecchia (261). rheuma factors-
954 producing cells exert a particular antigen-presenting function. Immunization with an antigen leads to production of antibodies which upon complexing the circulating antigen are recognized by the rheuma factor producing B-cells. The immune complex is internalized, processed and fragments of the antigen are presented to the T-cells which then exert their immunostimulating function. As the anti-F(ab')2 antibodies which belong to the class of rheuma factors, recognize the hinge-peptide there are anti-hinge B-cells capable of recognizing and internalizing the hinge-peptide/gastrin chimeras and thus, of presenting these molecules or parts of them to T-cells with subsequent immunostlmulation. Finally, injected hinge-peptide/gastrin constructs are recognized by the anti-hinge antibodies, the Fc of which binds to the Fcy-receptor of macrophages. The complex is internalized, processed and the antigen presented to the T-cells.
216
225
Human yl E P K S C D K T H T Mouse Y l V P R D C G Guinea p i g y l Q S W G H T Rabbit y A P S T C S K P M
232
CPP CKPCI C P P C I P C
238
C P A P E L L G G P CTVP B V S C G A P Z L L G G P P P P E L L G G P
Fig. 29. Comparison of the hinge sequence of human IgGl with those of guinea pig, rabbit and mouse. The hinge-peptide sequence used for the design and synthesis of the gastrin chimeras is related to human IgGl and inmiunizations were performed in guinea pigs. An alignment of the genetic hinge from different species is reported in Fig. 29. The lower hinge of human, guinea pig and rabbit IgGl shows the same size and very similar sequences, whereas in the mouse IgGl this hinge portion is significantly shorter. The high degree of sequence homology between the human and guinea pig hinge may possibly be responsible for the strong immune responses in guinea pigs if this is induced by one of the mechanisms discussed above. Supportive for this hypothesis is the observation that in Balb/c and C57BL/6 mice the hinge-peptide/gastrin chimeras were found to trigger no or very weak immune responses (Gemeiner, M. and Moroder, L., unpublls ;c:d results). Specificity of the Immune Response to Hinge-Peptide/Gastrin Chimeras The specificity of the antisera raised against the various hinge-peptide gastrin constructs was determined using a whole set of gastrin-related peptides (245).
955 Both the hlnge-peptlde/[Nlei5]-gastrln-[2-171 and the big-gastrin-[l-141/hlngepeptlde/INlei5]-gastrln-[2-17I generated antibodies of monoclonal-type character In all animals regarding specificity vs. the gastrin antigen. As shown in Fig. 30, the C-termlnal gastrin peptide amides 14-17 and 13-17 are recognized to negligible extents. Similarly, the complementary sequence 1-13 and its constituent subfragments 1-5 and 4-13 are not competitive ligands. Expression of an immunodeterminant sequential epitope in the antigen gastrin can therefore be excluded. Extension of the C-termlnal gastrin peptides vs. the characteristic pentaglutamic acid sequence produces a sharp transition of the binding affinities upon incorporation of four glutamic acid residues to reach a maximum value at the level of the biologically fully active [Nlei5]-gastrin-[5-171 and [Nlei^j-gastrin
IC«,[%]
Pyr Gly Pro Trp Leu Glu Glu Glu Glu Glu Ala lyr Gljr Trp Met Asp Phe-NH,
1 2 3 4 5 6 7 8 9 10 11 1213 14 1516 17
Fig. 30. Crossreactivities of gastrin related peptides as determined by 50% inhibition of binding of gastrin to antigastrin antibodies raised against hinge'peptide/[Nle^^]-gastrin'l2-17] in guinea pigs, and expressed as percentage of the IC50 of gastrin. The gastrin peptides analyzed are: gastrin-[14-17] (^), lPyr^0^j^l5Ugastnn-ll0'17] (0), IPyT^^Nle^^j-gastrin-lQ17] (•). [Pyi^,Nle^5lgastnn-[8'17] (O), [Pyr7,Nle^5].gastnn-l7-17] (A), [Pyf^.N]e^^]-gastrin-[6-17] (A). [Nle^5J-gastrin'[5'17] (•) and [Nle^^l-gastrin (or gastrin) (Q).
sequences. Since the biologically inactive des-amido-[Nlel5]-gastrin-l5-17) behaves as a very weak llgand, the experimental data indicate that the gastrin moiety in the synthetic construct is recognized by the B-cell surface immunoglobulins as a
956 conformational epitope similar to the bioactive structure of the hormone at receptor level. This hypothesis is further supported by a comparison of the parallelism between bioactivity. gastrin receptor affinity and onset of conformational order shown in Fig. 31 and by the competitive affinity of the gastrin peptides for the antl-gastrln antibodies. Moreover, the correlation of the crossreactivles of these antibodies with gastrin analogs of differentiated blopotencies is surprisingly good. This allows to conclude that with the hlngepeptide/gastrin constructs in which the bioactive core sequence of the hormone, i.e. 6-17, is sufficiently spaced from the template to allow for at least one gastrin moiety to fold into its preferred structure, antibodies are obtained of gastrin receptor-like specificity.
5 10 15 Number of residues In the chain
Fig, 31. Relative variation of molar ellipticity values (Al]/[Q]ioi) at 216 (0) and 192 run (A) as a fimction of chain length. Data of hormonal potency of the various gastrin peptides are reported for comparison (•).
Conversely, Immunizations performed with the hinge-peptide/[Nlel5]-gastrin-l517] compound clearly revealed a restricted accessibility of the shortened gastrin moieties. In fact, the antibodies raised against this immunogen completely differ in their specificity from those described above. All crossreactivity data obtained with the gastrin related peptides indicate expression of an epitope related to the C-terminal portion of the molecule and most probably of sequential character, as amino acid replacements not seen by the gastrin receptor in this portion of the molecule are strongly affecting the binding affinity for the antisera. These data are in full agreement with the conformational properties of this chimeric construct which clearly indicated strong interchain interferences with restriction of the
957 conformational space and with free accessibility limited to the C-terminal portion of the gastrin moiety. Conformational studies on shortened gastrin peptides lead to the conclusion that their extension to the critical length of gastrin 7-17 is necessary to allow for the hairpin conformation to be assumed. Therefore free presentation of just the C-terminal portion of gastrin to the B-cell surface iipmunoglobulins allows for recognition of this part only as sequential epitope. Very similar antibodies were obtained with the hinge-peptide/des-amido-INle^^j. gastrin-[5-17] dimer. The dimeric gastrin compound [Nlel5l-gastrin-[5-171/hinge-peptide in which the gastrin molecule is linked C-terminally to the hinge-peptide and thus, in a biologically inactive form as confirmed in the bioassays, generates an antibody population capable of recognizing only the N-terminus of the gastrin-[5-17] antigen (245). The antibodies crossreact with gastrin peptides containing the Nterminal portion Leu-Glu-Glu independently of whether the amino function is free or acylated. Gastrin-peptides lacking this portion are not anymore recognized indicating again the expression of a sequential epitope restricted to this short portion which in analogy to what was observed for the hinge-peptide/[Nlei^)gastrin-[5-17] probably represents the only segment fully accessible whereas the rest seems to be shielded by interchain interferences. Interestingly, immunizations with the bivalent immunogen big-gastrin-[l14]/hinge-peptide/[NlelSl-gastrin-[2-17I led to immune responses containing high titers of antigastrin antibodies of gastrin receptor-like specificity and low titers of antibodies directed against the N-terminal antigen. These antibodies are unable to recognize fragments of the big-gastrin-[l-141 sequence like the peptides 1-6 and 714 (245). The peptide 7-14 contains at its N-terminus the peculiar sequence ProPro-His which according to the conformational studies discussed above is probably involved in a P-tum type conformation. Endgroup effects can prevent the onset of this secondary structure element in the fragment 7-14 and thus, the crossreactivity with the antibodies unless this lack of binding affinity results from charge effects. The results again confirm that even the second antigenic site retained sufficient conformational free space to express its sequence specific conformational epitope. The results of the immunization experiments clearly underline the difficulty of presenting antigens to the immune system in a manner that intrinsic conformational properties can fully evolve and trigger the immune responses. Conjugation of peptides to proteins is known to suppress expression of epitopes and to provoke immune responses capable of recognizing the peptide but not its intrinsic conformational epitope as present on the protein surface. Bundled antigens on multiple antigen presenting systems are characterized by similar
958 shortcomings since interchain interactions may lead to restricted conformational space as well documented in the model studies with the hinge-peptide constructs. Therefore in the de novo design of synthetic immunogens particular attention should paid to the conformational analysis if better insights into the mechanism of immune recognition is the goal.
Conclusion Cell-surface receptors can be divided into three classes, depending on whether information is transmitted by allosteric conformational changes, by receptor dimerlzation or by receptor aggregation (262). The signal transduction of the hormone gastrin occurs upon interaction with a membrane-bound receptor with seven putative transmembrane helices (263) where allosteric conformational changes are most probably transmitting the signal, although clustering of hormone receptors may not be excluded and could possibly be responsible for the higher potencies observed for some hormone dimers. Recently a detailed investigation of bradykinin antagonist dimers has clearly demonstrated a decisive role of the size of the spacer used in the dimerlzation (253), a fact which could support contemporary interaction with two receptors. Dimerization of regulatory peptides has repeatedly been used in the search for superactive agonists and antagonists, but with contradictory results in terms of the resulting hormonal potencies. Detailed studies on the conformational effects of these dimerizations as potential factors responsible for the unpredictable biopotencies of dimers of peptide hormones have not been reported so far. In the case of the hingepeptide/gastrin dimers the conformational analysis clearly revealed a strong correlation of the differentiated biopotencies with different degrees of interchain interactions of the gastrin moieties and correspondingly, with their accessibility for recognition by the receptors. Aggregation-activated receptors are frequently encountered in the immune system (264-266). The members of this receptor class bear short cytoplasmatic domains which act to bind and recruit other cellular factors following the aggregation of their extracellular domains. Aggregation of these surface receptors, e.g. surfaceimmunoglobulins of B-cells. could be favored by multiple presentation of the antigens (142). Clustering of these receptors would lead to differentiation of the Bcells into plasma cells and to their proliferation (267.268). The dimerization of the antigen gastrin in the hinge-peptlde constructs hardly allows for such concomitant capping of two B-cell receptors as the differentiated immune response to the hinge-peptlde/[Nlei5]-gastrin-[5-171 and hinge-peptide/lNlel^jgastrin-l2-17] is more in agreement with the conformational preferences resulting
959 from the strong intrachain interferences which allow for recognition and binding of solely one gastrin chain. Nonetheless strong immune responses were obtained with these dimerlc constructs, and these have then to be attributed to the peculiar properties of the hinge-peptlde portion. In view of the multiple functions of the hinge segment in the immunoglobulin molecules it seems proper to interpret the experimentally observed strong immune responses in terms of specific recognition of the hinge portion by immimoglobulln-related molecules involved in the mechanisms of immtmlty. In this sense the hlnge-peptide could play the role of an universal recognition site, since molecular modelling indicates a surprisingly good fit of this double-chain peptide even into the binding cleft of MHC class II proteins. For this reason constructs with the hlnge-peptide could represent Interesting tools for Investigating at molecular level the intricated mechanisms Involved in Immunology. References 1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12) 13) 14) 15) 16) 17)
Brown, F. (1994) Phil Trans. R Soc. Lond. B 344, 213-219. Rabinovich, N.R., Mclnnes, P., Klein, D.L. and HaU, B.F. (1994) Science 265, 1401-1404. La Montagne, J.R. and Rabinovich. N.R (1994) Int. J. Technol Assess. Hecdth Care 10, 7. Amon, R. (1986) TIBS H , 521-524. Sela, M. and Amon, R. (1992) Vaccine 10, 991-999. Amon, R. (1972) in: Immunity and Viral and Rickettsial Diseases (Kohn, A. and Klingberg, A.M., eds.) pp. 209-222, Plenum Press, New York. Sela, M. (1974) BuIL Inst. Pasteur 72. 73-86. Anderer, F.A. (1963) Z. Naturforsch. 18, 1010-1014. Anderer, F.A. (1963) Biochim. Bvophys. Acta 7 1 , 246-248. Langbeheim. H., Amon, R. and Sela, M. (1976) Prcx:. Natl Acad. Set USA 7 3 , 4636-4640. Senyk. G., Williams, E.B., Niteckl, D.E. and Goodman, J.W. (1971) J. Exp. Med. 133, 1294-1308. MiUch. D.R., McLachlan, A., Chisari, F.V. and Thornton, G.B. (1987) Nature 329, 547-549. DlMarchi, R., Brooke. G., Gale. C . CrackneU. V.. Doel. T. and Mowat. N. (1986) Science 232. 639-641. Francis. M.J.. Fry. CM.. Rowlands. D.J.. Brown. F.. Bittle. J.L.. Houghten. R.A. and Lemer. R.A. (1985) J. GerL Virol 66. 2347-2354. Francis. M.J.. Fry. CM.. Rowlands. D.J.. Bittle. L.. Houghten, R.A.. Lemer. R.A. and Brown, F. (1987) Immunology, 61.1-6. Antonl. G.. Mariani, M.. Presentlni. R.. Lafata. M.. Neri. P.. Bracci. L. and CianfrlgUa (1985) Mol Immunol 28. 1237-1241. Emlni. E.A., Bradford, A.J. and Wimmer, E. (1983) Nature 304, 699-703.
960 18) Shapira. M,. Jibson. M., Muller. G. and Amon. R. (1984) Proc, Natl Acad. Set l/SA 8 1 . 2461-2465. 19) Young. C.R and Atassi. M.Z. (1982) iTramm. CommuTL 11. 9-16. 20) Mariani. M.. Bracci. L., Presentlni. R . Nucci. D.. Nerl. P. and Antonl. G. (1987) MoL JmmunoL 24. 297-303. 21) Claman. H.. Chaperon, E. and Trlplett. R. (1966) Soc. Exp. Biol Med. 122. 1167-1171. 22) Landsteiner. K. (1946) in: The Specificity of Iwimmological Reactions, Harvard University Press. 1946. 23) Mitchison. NJL (1971) Eur. J. Immimol 1. 18-27. 24) Katz. D.H. (1980) Adv. Immun. 29. 137-207. 25) Benacerraf. B. and McDevltt. H. (1972) Science 175. 273-279. 26) Guillet. J.-G.. Lai. M.-Z.. Brlner. T.J.. Buus. S.. Sette. A.. Grey. H.M.. Smith. J.A. and Gefter. M.L. (1987) Science 235, 865-870. 27) Rosenthal. A.S. (1978) Immunol Rev. 40. 136-152. 28) Berzofsky. J.A.. Cease. K.B.. Comette. J.L.. Spouge. J.L.. Margalit. H.. Berkower. I.J.. Good. M.F.. Hiller. L.H. and DeUsi. C. (1987) Immunol Rev. 98. 9-52. 29) Benacerraf. B. (1978) J. Immunol 120.1809-1812. 30) Brown. J.. Jardesky. T.. Gorga. J.. Stem. L.. Urban. R . Strominger. J. and Wiley, D. (1993) Nature 364, 33-39. 31) Stem. L., Brown, J.. Jardesky. T., Gorga, J., Urban. R , Strominger. J. and Wiley, D. (1994) Nature 368, 215-221. 32) Bjorkman, P., Saper, M., Samraoui, B., Bennett. W., Strominger, J. and Wiley. D. (1987) Nature 329. 506-511. 33) Bjorkman. P., Saper, M., Samraoui, B., Bennett, W., Strominger, J. and Wiley, D. (1987) Nature 332, 512-518. 34) Saper, M., Bjorkman, P. and Wiley D. (1991) J. Mol Biol 219, 277-319. 35) Fremont, D.. Matsumura, M., Stura, E., Peterson, P. and Wilson, I. (1992) Science 257, 919-927. 36) Matsumura, M., Fremont, D., Peterson, P. and Wilson, I. (1992) Science 257, 927-934. 37) Rammensee, H.-G., Falk, K. and Rotzschke, O. (1993) CUJT. Immunol 5, 3544. 38) Rotzschke, O. and Falk, K. (1991) Immunol Today 12, 447-455. 39) Engelhard. V.H. (1994) Curr. Immunol 6. 13-23. 40) Chicz, RM. and Urban, R G . (1994) Immunol Today 15, 155-160. 41) EUiott, T. (1991) Immunol Today 12, 386-388. 42) SinigagUa, F. and Hammer, J. (1994) Curr. Immunol 6, 52-56. 43) Rothbard, J.B. (994) Curr. Biol 4, 653-655. 44) Cresswell, P. (1994) Curr. Biol 4, 541-543. 45) Das, M. and Fox, F. (1979) Rev. Biophys. Bioeng. 8, 165-193. 46) Han, K.-K., Richard, C. and Delacourte, A. (1984) Int. J. Biochem 16, 129145. 47) Moroder, L., Gemeiner, M., Kalbacher, H., Nyfeler, R and Wiinsch, E. (1983) in: Peptides 1982 (Blaha, K. and Malon. P., eds.) pp. 667-672. Walter de Gruyter, Berlin, New York.
961
48) Moroder. L.. Nyfeler, R.. Gemeiner, M.. Kalbacher, H. and Wunsch, E. (1983) Biopolymers 22, 481-486. 49) Brland. J.P.. Muller, S. and Van Regenmortel. M.H.V. (1985) J. Immunol Meth, 78, 59-69. 50) Van Regenmortel. M.H.V., Briand, J.P., MuUer, S. and Plaue, S. (1988) in: Synthetic Polypeptides as Antigens (Burdon, R H . and van Knlttenberg, P.H., eds.) pp. 95-216, Elsevier, Amsterdam. 51) Reichlin. M. (1980) MetK Enzymol 70, 159-165. 52) Geiger, R., Moroder, L., and Wunsch, E. (1984) in: Peptides 1984 (Ragnarsson. U., ed.) pp. 451-456. Almqvist and Wicksell Int.. Stockholm. 53) Rehfeld. J.F. (1978) J. Biol Chem 253. 4016-4021. 54) Yanalhara, Ch., Yanahaira, N., Shimlzu, F., Sato, H., Uehata, S. and Imagawa, K. (1980) Biomed, Res. 1, 242-247. 55) Turkelson, CM., Dale, W.E., Reidelberger, R. and Solomon. T.E. (1986) ReguL Peptides 15. 205-217. 56) Wunsch, E., Moroder, L., Nyfeler, R.. Kalbacher. H. and Gemeiner. M. (1985) Biol Chenh Hpppe-Seyler 368. 855-861. 57) Moroder, L.. Tzougraki. Ch.. Gohrlng. W.. Mourier. G.. Musiol. H.-J. and Wunsch. E. (1987) Biol Chenh Hoppe-Seyler 368, 855-861. 58) KeUer, O. and Rudinger, J. (1975) Helv. Chtm Acta 58, 531-541. 59) Marrian, D.H. (1949) J. Chenh Soc. 1515-1516. 60) Wold, F. (1967) Meth. Enzymol 11, 617-640. 61) Green, N., Alexander, H. Olson, A., Alexander, S., Shinnick, T.M, Sutcliffe, J.G. and Lemer, R.A. (1982) CeU28, 477-487. 62) Eritja, R., Ziehler-Martin, J.P.. Walker. P.A., Lee, T.D., Legesse, K., Albericcio, F. and Kaplan, B.E. (1987) Tetrahedron 43, 2675-2680. 63) Batz, H.-G., Hubner. C , Kerscher. L.. Klein. C. and Kunz. H. (1989) in: Peptides 1988 (Jung, G. and Bayer. E., eds.) pp. 754-756, Walter de Gruyter, Berlin. New York. 64) Routhe. M.R., Bowman, G.J., Gangitano, J.M., Lambing, J.L., Pratt, S.M. and Hopp, T.P. (1994) in: Peptides - Chemistry, Structure and Biology (Hodges, R.S. and Smith, J.A., eds.) pp. 159-160. ESCOM, Leiden. 65) Munson, M.C., Lebl, M., Slaninova, J. and Barany, G. (1993) Peptide, Res. 6, 155-156. 66) motz, I.M. and Heiney. R.E. (1962) Arch. Biochenh Biophys. 96. 605-612. 67) Moroder, L., Mourier, G., Bovermann, G., Dufresne, M., Gohring, W., Gemeiner. M. and Wunsch. E. (1987) Biol Chem. Hoppe-Seyler 368, 849-853. 68) Moroder, L., Mourier, G., Dufresne. M.. Bovermann. G.. Gohring. W., Gemeiner, M. and Wunsch, E. (1987) Biol Chenh Hoppe-Seyler 368, 839-848. 69) Wunsch. E.. Moroder. L.. Gillessen. D. and Soerensen. U.B. (1982) HoppeSeyler's Z. Physiol Chenh 363. 665-669. 70) Krois, D.. Dufresne, M.. Neumann, H., Moroder, L. and Wunsch, E. (1990) Biol Chenh Hoppe-Seyler 371, 43-48. 71) Dickerson, R.E. and Timkovich, R.T. (1975) in: The Enzymes (Boyer, P.D., ed.) pp. 397-547, Academic, Press, New York. 72) Reichlin. M., NisonofT, A. and MargoUash, E. (1970) J. Biol Chenh 245, 947954.
962 73) Schwartz. R.H., Fox. B.S.. FVaga. E.. Chen. C. and Singh B. (1985) J. Immunol 135, 3028-3033. 74) Shell. J.M.. Schell. T.D.. Shepherd. S.E.. Kllmo. G.F.. Kioschos. J.M. and Paterson. Y. (1994) Eur. J. Immunol 24. 2141-2149. 75) Narita. K. and Tltanl. K. (1969) J. Biochem. 65. 259-267. 76) Motonaga. K.. Misaka, E.. Nakajlma. E.. Ueda. S. and Nakanishl. K. (1965) J. Biochem. 57. 22-28. 77] Dickerson. R.E., Takano. T.. Elsenberg. D.. Kallal. O.B., Samson. L.. Cooper. A. and MargoUash. E. (1971) J. Biol Chem. 246. 1511-1535. 78) Salemme. F.R.. Freer. S.T.. Xuong. N.H.. Alden. RA. and Kraut. J . (1973) J. Biol Chem. 218. 3910-3921. 79) Wunsch. E.. Moroder, L.. Hubener. G.. Musiol. H.-J., von Grunigen. R.. Gohrlng. W.. Scharf. R. and Schneider. C.H. (1991) Int. J. Peptide Protein Res, 37. 90-102. 80) Peggion. E.. Jaeger. E.. Knof. S.. Moroder. L. and Wunsch. E. (198 ) Biopolymers 20, 633-652. 81) Peggion. E., Foffani. M.T., Wunsch, E., Moroder. L.. Borin. G.. Goodma . vf and Mammi, S. (1985) Biopolymers 24, 647-666. 82) Mammi, S., Mammi, N.J. and Peggion, E. (1988) Biochemistry 27, 1374-i .79. 83) Mammi, S. and Peggion, E. (1990) Biochemistry 29, 5265-5269. 84) Mammi, S., Foffani, M.T., Peggion. E.. Galleyrand. J . C . Bali. J.P.. Simonetti. M., Gohring. W.. Moroder, L. and Wunsch, E. (1989) Biochemistry 28, 71827188. 85) Gohring, W., Moroder, L., Borin, G., Lobbia, A., Bali, J.-P. and Wunsch, E. (1984) Hoppe-Seyler's Z. Physiol Chem. 365, 83-94. 86) Moroder, L., Bovermann. G., Mourier, G., G6hring, W., Gemeiner, M. and Wunsch, E. (1987) BioL Chem. Hoppe-Seyler 368, 831-838. 87) von Grunigen, R., Siglmuller, G., Papini, A., K6cher, K.. Traving, B., Gohring. W. and Moroder, L. (1991) BioL Chem. Hoppe-Seyler 372, 163-172. 88) Moroder, L., Bali, J.-P. and Kobayashi, Y. (1991) Biopolymers 31, 595-604. 89) Moroder, L., Papini, A., Siglmuller, G., Kocher, K., Dorrer, E. and Schneider, C.H. (1993) BioL Chem. Hoppe-Seyler 373, 315-321. 90) Moroder, L. and Temussi, P.-A. (1993) in: Peptides 1992 (Schneider, C.H. and Eberle, A.N., eds.) pp. 813-814, ESCOM, Leiden. 91) Martinez, J. (1990) in: Comprehensive Medicinal Chemistry (Hansch, C , Sammer, P.G. and Taylor, J.B., eds.) Vol. 3. pp. 925-959, Pergamon Press, New York. 92) Moroder, L., Romano, R., Weyher, E., Svoboda, M. and Christophe, J . (1993) Z. Naturforsch. 48b. 1419-1430. 93) Moroder, L., D'Ursi, A , Picone, D., Amodeo, P. and Temussi, P.A. (1993) Biochem. Biophys. Res. Commun. 190, 741-746. 94) Shinnik, T.M., SutcUffe, J . C , Green, N. and Lemer, R.A. (1983) Ann. Rev. Microbiol 37, 425-426. 95) Barlow, D.J., Edwards, M.S. and Thornton, J.M. (1986) Nature 322, 747-748. 96) Lemer, R.A. (1984) Adv. Immunol 36, 1-4. 97) Van Regenmortel, M.H.V. (1987) 77BS 12, 237-240. 98) Van Regenmortel, M.H.V. (1989) Immunol Today 10, 266-272.
963
99) Lowell, G.H., Smith, L.F., Seid, R.C. and Zollinger, W.D. (1988) J. Exp, Med. 167, 658-663. 100) Lowell,G.H., Ballou, W.R., Smith, L.F., Wirtz, RA., Zollinger, W.D. and Hockmeyer, W.T. (1988) Science 240, 800-802. lODPrasch. C.E. and Peppier, M.S. (1982) Infect. Immun. 37, 271-280. 102)Gotschllch, E.G., Eraser. B«A.. Nlshlmura, O., Robbins, J.B. and Liu, T.-Y. (1981) J. Biol Chem. 256, 8915-8921. 103) Wichek, M. and Bayer, E.A. (1988) Anal Biochem. 174, 1-32. 104) Wichek, M. and Bayer, E.A. (1984) Immunol Today 5, 39-43. 105) Weber, P.C., Ohlendorf, D.H.. Wendoloski. J . J . and Salenmie, F.R. (1989) Science 243, 85-88. 106)Hofmann, K., Titus. G., MontibeUer. JJL and Finn. M.F. (1982) Biochemistry 21, 978-984. 107) Shapira. M.. Gibson. M.. MuUer. G.. Amon. R and Sela. M. (1985) Proc. Natl Acad. Set USA 81. 2461-2465. 108) Audibert. F.. JoUvet. M.. Chedid. L.. Amon. R and Sela. M. (1982) Proc. Natl Acad. Set USA 79. 5042-5046. 109) Jacob. C O . . Amon. R. and Sela, M. (1986) Immunol Lett. 14. 43-48. 110)Hudecz. F.. Hilbert. A., Mezo. G. et al. (1993) Peptide Res. 6. 263-271. l l D H u d e c z . F.. Rajnavolgyi. E.. Price. M.R and Szekerke. M. (1992) in: Peptides 1992 (Schneider. C.H. and Eberle. A.N.. eds.) pp. 869-870. ESCOM Leiden. 112)Jaffe. B.M.. Newton. W.T. and McGuigan. J.E. (1970) Immunochemistry 7. 715-725. 113)Posnett. D.N.. McGrath. H. and Tarn. J.P. (1988) J. Biol Chem 263. 17191725. 114) Tarn. J.P. (1988) Proc. Natl Acad. Set USA 85. 5409-5413. 115)DriJfhout. J.W. and Bloemhoff. W. (1991) Int. J. Peptide Portein Res. 37. 2732. 116) Spetzler. J.C. and Tarn. J.P. (1995) Int. J. Peptide Protein Res. 45. 78-85. 117) Huang, W.. NardeUi. B. and Tarn, J.P. (1994) Mol Immunol 3 1 . 1191-1199. 118)Defoort. J.-P.. Nardelli. B.. Huang. W.. Ho. D.D. and Tarn. J.P. (1992) Proc. Natl Acad. Set USA 89. 3879-3883. 119) NardeUi. B.. Defoort. J.-P.. Huang. W. and Tarn. J.P. (1992) AIDS Res. Hum. Retroviruses S, 1405-1407. 120) Tarn. J.P. and Lu, Y.A. (1989) Proc. Natl Acad. Set USA 86. 9084-9088. 121)Tarn. J.P., Clavijo, P., Lu. Y.A.. Nussenzweig, V.. Nussenzweig. R. and Zavala, F. (1990) J. Exp. Med. 171. 299-306. 122)Troalen. F.. Razafindratsita. A.. Puisieux. A.. Voeltzel. T.. Bohuon. C , Bellet. D. and Bidart. J.M. (1990) Mol Immunol 27. 363-368. 123) Marguerite. M.. Bossus, M.. Mazingue. C . Wolowczuk, I.. Gras-Masse. H.. Tartar. A.. Capron. A. and Auriault. C. (1992) Mol Immunol 29. 793-800. 124)Briand. J.-P.. Barin. C . Van Regenmortel. M.H.V. and MuUer. S. (1992) J. JmnumoL MetK 156. 255-265. 125)Esposito. G.. Fogolari. F.. Viglino. P.. Cattarinussi, S.. De Magistris. M.T.. Chiappinelli. L. and Pessi, A. (1993) Eur. J. Biochem 217. 171-187. 126)Modrow. S.. H5flacher. R , Mertz, H. and Wolff. J. (1989) J. Immunol Meth. 118. 1-7.
964
127)Chong. P.. Sydor. M.. Wu. E. and Klein, M. (1992) MoL Immunol 29, 443446. 128) Fischer. P.M.. Comis. A. and Howden. M.E. (1989) J. Immunol Meth. 118. 119-123. 129)Kanda. P.. Kennedy, R.C. and Aparrow. J.T. (1991) Int. J. Peptide Protein Res. 3 8 . 385-391. 130)Goddard. P.. McMurray. J.S.. Sheppard. R.C. and Emson. P. (1988) J. Chem. Soc. Chem. Comnum. 1025-1027. 131) Hegel. M.. Pichova. D., Minarlk. P. and Sheppard. R.C. (1990) in: Peptides 1990 (Giralt, E. and Andreu, D.. eds.) pp. 837-838. ESCOM. Leiden. 132)Rapp, W.. Zhang, L.. Beck-Sickinger. A.G.. Deres. K., WiesmuUer. K.H.. Jung, G. and Bayer. E. (1990) in: Peptides 1990 (Giralt. E. and Andreu. D.. eds.) pp. 849-850. ESCOM. Leiden. 133)Butz. S.. Rawer. S.. Rapp. W. and Birsner. U. (1994) Peptide Res. 7. 20-23. 134) Gregoriadis. G. (1990) Immunol Today 11. 89-97. 135) Alving. C.R. (1989) J. Immunol Meth. 140. 1-13. 136)Ostro. M.J. (1987) Liposomes from Biophysics to Therapeutics, M. Dekker. New York. 137) Goodman-Snitkoff. G.. Eisele, L.E., Heimer, E.P.. Felix, A.M., Anderson, T.T., Fuerst. T.R. and Mannino. R.J. (1990) Vaccine 8, 257-262. 138) Goodman-Snitkoff, G., Good, M.F., Berzofsl^. J.A. and Mannino, R.J. (1991) J. Immunol 147. 410-415. 139)Watari. E., Dietzschold, B.. Szokan. G. and Heber-Katz, E. (1987) J. Exp. Med. 165. 459-470. 140) Dal Monte, P.R. and Szoka, F.C. (1989) J. Immunol 142, 1437-1443. 141) Dal Monte, P.R. and Szoka, F.C. (1989) Vaccine 7. 401-408. 142)Frisch. B.. Muller, S., Briand, J.P., Regenmortel, M.H.V. and Schuber, F. (1991) Eur. J. Immunol 21, 185-193. 143)Friede, M., Muller, S., Briand, J.P., Regenmortel, M.H.V. and Schuber, F. (1993) MoL Imnumol 30, 539-547. 145) Yasuda, T., Kanegasaki, S., Tsumita. T., Tadakuma. T., Ikewaki, N., Homma, J.Y., Inage, M., Shiba, T. and Kusumoto, S. (1984) Eur. J. Biochem 140, 245248. 146) Alving. C.R. and Richardson, E.G. (1984) Rev. Infect. Dis. 6. 493-496. 147) Lederer. E. (1980) J. Med. Chem. 23, 819-825. 148)Baschang, G. (1989) Tetrahedron 45, 6331-6360. 149) Georgiev, V.S. (1990) TIPS 11, 373-378. 150)Moroder, L., Musiol, H.-J. and Siglmuller, G. (1990) Synthesis 10, 889-892. 151)Gemeiner, M„ Leidinger, E., Miller, I. and Moroder, L. (1992) Biol Chem Hoppe-Seyler 373, 1085-1094. 152)Ellouz, F., Adam, R., Ciorbaru, R. and Lederer, E. (1974) Biochem. Biuj Res. Commun. 59, 1317-1325. 153)Matsumoto, K.. Otani, T.. Une, T.. Osada. Y., Ogawa, H. and Azuma. I. (1983) Jn/ect. J/Ti/n. 39, 1029-1040. 154) Romano, R., Musiol, H.-J.. Weyher, E., Dufresne, M. and Moroder. L. (1992) Biopolymers 32, 1545-1558.
965 155) Romano, R., Dufresne. M., Prost. M.-C. Ball. J.-P., Bayerl. T.M. and Moroder. L. (1993) Biochim Biophys. Acta 1145, 235-242. 156)Machy, P. and Leserman, L.D. (1983) Biochinh Biophys. Acta 730, 313-320. 157) Heath, T.D., Lopez, N.G. and Papahadjopoulos, D. (1985) Biochim. Biophys, Acta 820, 74-84. 158) Rude, E., Meyer-Dellus. M. and Gundelach, M.-L. (1971) Eur, J. Immunol 1, 113-123. 159) Stark. J.M., Locke, J. and Healtley. R.V. (1980) Immunology. 39, 345-352. 160)Hopp, T.P. (1984) Mat Immunol 1, 13-16. 161)Rouaix, F., Grass-Masse, H., Mazingue, C Diesis, E.. Ridel. P.R.. Estaquler. J., Capron, A., Tartar, A. and Auriault. C. (1994) Vaccine 12, 1209-1206. 162) Martinon, F., Grass-Mass, H., Boutlllon, C., Chlrat, F., Deprez, B., Gulllet, J.-G., Gomard, E., Tartar, A., Levy, J.-P. (1992) J. Immunol 149, 3416-3422. 163)Deres, K., Schlld, H., Wiesmuller, K.-H., Jung. G. and Rammensee H.G. (1989) Nature 342, 561-564. 164)Schild, H., Norda, M., Deres, K, Falk, K., RStzschke, O., WiesmuUer, K.-H., Jung, G. and Rammensee, H.-G. (1991) J, Exp, Med, 174, 1665-1668. 165)Seifert, R., Schultz, G., Richter-Freund, M., Metzger, J., Wiesmuller, K.-H., Jung, G., Bessler, W.G. and Hauschildt, S. (1990) Biochem. J. 267, 795-802. 166) Metzger, J., Jung, G., Bessler, W.G., Hoffmann. P., Streaker, M., Ueberknecht. A. and Schmidt, U. (1991) J. Med, Chem. 34, 1969-1974. 167) Wiesmuller, K.-H., Besslerr, W. and Jung, G. (1983) Hoppe-Seyler's Z. Physiol Chem, 364, 593-606. 168) Bessler, W.G., Suhr, B., Buhrlng. H.-J., Muller, D.P., Wiesmuller, K.-H., Becker, G. and Jung, G. (1985) Immunbiol 170, 239-244. 169) Jung, G., Wiesmuller, K.-H.. Becker, G., Buhring, H.-J. and Bessler, W.G. (1985) Angew, Chem. 97, 883-885. 170)Reichert. C.M., Carelli, C , JoUvet, M., Audibert, F.. Lefrancier, P. and Chedid, L. (1980) MoL Immunol 17, 357-363. 171) Carelli, C , Audibert, F., GaiUard, J. and Chedid, L. (1982) Proc, Natl Acad. Set USA 79, 5392-5395. 172) Carelli, C . Ralamboranto, L.. Audibert. F., Gaillard, J.. Briquelet. N., Dray, F., Fafeur, V.. Haour, F. and Chedid. L. (1985) Int, J. Imnumpharmacol 7, 215-224. 173)Hosmalin. A., Carelli, C , Gaillaerd, J., Lefrancier, P., Drobeco, H., Leclerc, C , Amar, D., Audibert, F. and Chedid, L. (1987) Clin, Immun. Imnvunpath, 45, 447-460. 174)Moroder, L., Gobbo, M., Becker, G. and Wunsch, E. (1989) BioL Chem, Hoppe-Seyler 370. 365-375. 175)Moroder, L.. Dufresne, M., Gohring, W., Wunsch, E., Leidinger, E. and Gemeiner, M. (1989) Biol Chem. Hoppe-Seyler 370. 1209-1214. 176)Schutze, M.P., Leclerc, C , JoUvet, M., Audibert, F. and Chedid, L. (1985) J. Immunol 135. 2319-2322. 177)Dryberg. T. and Oldstone, M.B.A. (1986) J. Exp, Med, 164. 1344-1349. 178)Schutze, M.P.. Deraud. E., Przewlocki, G. and Leclerc, C. (1989) J. Immunol 142, 2635-2640. 179) Amon, R.. Teicher, E. and Scheraga. H.A. (1974) J. Mol Biol 90, 403-407.
966 180)Niman. H.L.. Houghten. R.A.. Walker. L.E.. Relsfeld, RA.. Wilson. IJL. Hogle. J.M. and Lemer. R.A. (1983) Proc. Natl Acad. Set USA 80. 4949-4953. 181) Jacob. C O . . Amon. R and Sela. M. (1985) AfoL Immunol 22. 1333-1339. 182)Borras-Cuesta. F.. Fedon. Y. and Petlt-Camurdan. A. (1988) Eur. J. Immunol 18. 199-202. 183) Broekhuljsen. M.P.. Van Rijin. J.M.M:. Blom. A.J.M.. Pouwels. P.H.. EngerValk. B.E.. Brown. F. and Francis. M.J. (1987) J. Geru Virol 6 8 . 3137-3143. 184)Leclerc. C . Przewlocki. G.. Schutze. M.P. and Chedid. L. (1987) Eur. J. Immunol 17. 269-273. 185)BoiTas-Cuesta. F.. Petit-Camurdan. A. and Fedon. Y. (1987) Eur. J. Immunol 17. 1213-1215. 186)Francis.M.J.. Hastings. G.Z.. Syred. A.D.. McGinn. B., Brown. F. and Rowlands. D.J. (1987) Nature 330. 168-170. 187) Cox. J.H.. Ivanyi. J.. Young. D.B.. Lamb. J . R . Syred. A.D. and Francis, M.J. (1988) Eur. J. Immunol 18. 2015-2019. 188) Partidos. C . Stanley. C. and Steward. M. (1992) Eur. J. Immunol 22. 26752680. 189) Partidos. C . Stanley. C. and Steward. M. (1992) Mol Immunol 29. 651-658o 190) Good. M.F.. Maloy. W.L.. Lunde. M.N.. Morglalt. H.. Comette. J.L.. Smith. G.L.. Moss. B.. Miller. L.M. and Berzofsky. J.A. (1987) Science 235. 10591062. 191)Rouaix. F.. Gras-Masse. H.. Mazingue. C . Ridel. P.-R. Diesis. E.. Marguerite, M.. Estaquier. J.. Capron, A.. Tartar. A. and Auriault. C. (1994) Immunopharmacol 28, 137-143. 192) Good. M.F.. Pombo. D.. Lunde. M.N.. Maloy. W.L.. Halenbeck. R . Koths. K.. Miller. L.H. and Berzofsky. J.A. (1988) J. Immunol 141. 972-977. 193) Partidos. CD.. Obeid, O.E. and Steward. M.W. (1992) Immunology 77. 262266. 194)Richman. S.J. and Reese. RT. (1988) Proc. Natl Acad. Set USA 85. 16621666. 195) Gras-Masse. H.. Jolivet. M.. Drobecq. H.. Aubert. J.P.. Beachey. E.H.. Audibert. F.. Chedid. L. and Tartar. A. (1988) Mol Immunol 25. 673-678. 196)Dorow. D.S.. Shi, P.-T.. Carbone. F.R, Minasian, E., Todd, P.E.E. and Leach, S.J. (1985) Mol Immunol 22. 1255-1265. 197) Dyson, H.J., Cross, K.J., Houghten. R.A.. Wilson. I.A.. Wright. P.E. and Lemer, R.A. (1986) Nature 318. 480-483. 198) Dyson, H.J.. Ranee. M.. Houghten, RA., Wright, P.E. and Lemer. RA. (1988) J. Mol Biol 201. 201-217. 199)Kaumaya, P.T.P., Bemdt, K.D.. Heidom. D.B.. Trwhella. J.. Kezdy. F.J. and Goldberg, E. (1990) Biochemistry 29. 13-23. 200)Kaumaya. P.T.P.. Kobs-Conrad. S.. Digeorge. N.M. and Stevens. V.C. (1994) in: Peptides - Design, Synthesis and Biological Activity (Basawa. A. and Anantharamaiah. G.M.. eds.) pp. 133-164. Birkh§user. Boston. 201) Kaumaya. P.T.P.. VanBuskirk, A., Goldberg, E. and Pierce, S.K. (1992) J. Biol Chem. 267, 6338-6346. 202) Kobs-Conrad, S.. Lee, H., DiGeorge, A.M. and Kaumaya, P.T.P. (1993) J. BioL Chem. 268. 25285-25295. 203)Tuchscherer, G. and Mutter. M. (1995) J. Peptide Set 1, 3-10.
967 204) Mutter. M. and Villeumler (1989) Angew. Chem. Int. Ed. Engl 28. 535-554. 205)Tuchscherer. G.. Servls. C . Corradln. G.. Bliim. U.. Rivier. J . and Mutter. M. (1992) Protein Set 1. 1377-1386. 206)Moroder. L.. Bovermann. G. and Wunsch, E. (1988) in: Peptide Chemistry 1987 (Shiba. T. and Sakaklbara. S.. eds.) pp. 759-764. Protein Research Foundation. Osaka. 207) Bovermann. G.. Moroder. L. and Wunsch. E. (1989) in: Peptides 1988 (Jung. G. and Bayer. E.. eds.) pp. 748-750. Walter de Gruyter, Berlin. New. York. 208) Marquart. M. and Deisenhofer. J. (1982) Immunol Today 3, 160-166. 209) Burton. D.R (1985) Mol Imnumol 22. 161-206. 210)Feinstein. A.. Richardson. N.E. and Taussig. M.J. (1986) Imnumol Today 7. 169-174. 211) Deisenhofer. J. (1981) Biochemistry 20, 2361-2370. 212) Ito. W. and Arata Y. (1985) Biochemistry 24. 6467-6474. 213) Endo, S. and Arata. Y. (1985) Biochemistry 24. 1561-1568. 214) Kim. H.-H. Matsunaga. C.. Yoshlmo, A.. Kato. K, and Arata. Y. (1994) J. Mol Biol 236. 300-309. 215) Marquart. M.. Deisenhofer. J.. Huber. R. and Palm. W. (1980) J. Mol Biol 141. 369-391. 216) Wunsch. E.. Moroder. L.. G6hring-Romani. S.. Musiol. H.-J.. G6hrlng. W. and Bovermann. G. (1988) Int. J. Peptide Protein Res. 32. 368-383. 217) Zhang. R. and Snyder. G.H. (1989) J. Biol Chem. 264. 18472-18478. 218) Moroder. L.. Gemeiner. M.. GShring. W., Jaeger. E.. Thamm. P. and Wunsch, E. (1981) Biopolymers 20, 17-37. 219)Mukaijama. T. and Takahashi. K. (1968) Tetrahedron Lett. 5907-5909. 220) Helmkamp. G.K.. Cassey. H.N.. Olsen, B.A. and Pettitt. D.J. (1965) J. Org. Chem. 30. 933-935. 221) Moroder. L.. Hubener. G.. Gohring-Romani. S.. G6hring, W.. Musiol, H.-J. and Wunsch. E. (1990) Tetrahedron 46, 3305-3314. 222)Heaton. G.S.. I^don, H.N. and Schofield, J.A. (1956) J. Chem. Soc (C) 31573168. 223) Large. D.G.. Rydon. H.N. and Schofield. J A . (1961) J. Chem. Soc. (C) 17491752. 224)Yarvis, D., I^don. H.N. and Schofield, J.A. (1961) J. Chem. Soc. (C) 17521765. 225) Hardy, P.M.. Ridge, B..R^don. H.N. and dos S.P. Serrao. F.O. (1971) J. Chem. Soc. (C) 1722-1731. 226)Siedler. F.. Rudolph-Bohner, S., Doi. M.. Musiol, H.-J. and Moroder. L. (1993) Biochemistry 32, 7488-7495. 227)Siedler. F., Quarzago, D., Rudolph-Bohner, S. and Moroder, L. (1994) Biopolymers 34, 1563-1572. 228) Chou, P.Y. and Fasman, G.D. (1977) J. Mol Biol 115, 135-175. 229) Scheraga, H.A. (1980) in: Protein Folding (Jaenicke, E., ed.) pp. 261-286. Elsevier. Amsterdam. 230)Wetlaufer. D.B. (1981) Adv. Protein Chem. 34, 61-92. 231) Kim, P.S. and Baldwin, R.L. (1982) AnniL Rev. Biochem. 51, 459-489. 232) Montelione, G.T. and Scheraga, H.A. (1989) Ace. Chem. Res. 22, 70-76.
968
233) Oas. T.G. and Kim. P.S. (1988) Nature 336. 42-48. 234) Montelione. G.T.. Arnold. E.. Meinwald. Y.C.. Stlmson. E.R.. Denton. J.B.. Huang. S.-G., Clardy. J. and Scheraga. H.A. (1984) J. Am. Chent Soc. 106, 7946-7958. 235) Bergmann. L.W. and Kuehl. W.M. (1979) J. Biol Chem. 254. 5690-5694. 236)Kessler. H.. Mronga. S.. MuUer, G.. Moroder. L. and Huber. R (1991) Biopolymers 31. 1189-1204. 237) Sreerama. N. and Woody. R.W. (1994) Biochemistry 33. 10022-10025. 238)Relchlln. M. (1975) Adv. Immunol 20. 71-123. 239) Moroder. L. and Lutz. J. (1995) in: Chemistry of Natural Products (Atta-urRahman, ed.). Elsevier. Amsterdam, in press. 240) Sieber. P.. Kamber. B.. Riniker. B. and Rittel. W. (1980) Helu. ChinL Acta 63. 2358-2363. 241)Wunsch. E.. Moroder. L.. G6hrlng-Romanl, S.. Musiol. H.-J.. Gohring. W. and Scharf. R. (1991) Int, J. Peptide Protein Res, 37. 61-71. 242)Hubener. G.. Gohring, W.. Musiol. H.-J. and Moroder, L. (1992) Peptide Res. 5. 287-292. 243)Dockray. G.J.. Gregory. R.A.. Hood. L.E. and Hunkapillar. M.W. (1979) Bioorg. Chem. 8. 465-470. 244)Woody. RW. (1985) in: The Peptides (Hruby. V.. ed.) Vol. 7. pp. 15-114. Academic Press. Orlando. Fl. 245) Moroder. L.. Musiol. H.-J.. K6cher, K.. Bali. J.P., Schneider. C.H.. Guba, W.. Muller, G., Mierke. D.F. and Kessler. H. (1993) Eur. J. Biochem. 212. 325333. 246) Jackson, M. and Mantsch. H.H. (1992) Biochim. Biophys. Acta 1118, 139143. 247) Sonnichsen, F.D., van Eyk. J.E., Hodges, R.S. and Sykes, B.D. (1992) Biochemistry 31, 8790-8798. 248)Kondo, M.. Kodama, H.. Costa. T. and Shimohigashl. Y. (1986) Int. J. Peptide PorteinRes. 27. 153-159. 249) Kodama, H., Shimohigashi. Y., Costa, T. and Kondo, M. (1988) Int. J. Peptide Protein Res. 32, 41-46. 250) Stolz, M.B. and Fauchere, J.-L. (1988) Helv. Chint Acta 7 1 , 1421-1428. 251) Sakaguchi. K., Kodama, H., Matsumoto, H., Yoshida, M., Takano, Y., Kamiya, H., Waki, M. and Shimohigashi, Y. (1989) Peptide Res. 2, 345-351. 252) Stepinski, Y., Zajaczkowski, I., Kazem-Bek, D., Temeriusz, A., Lipkowski. A. and Tam, S.W, (1991) Int. J. Peptide Protein Res. 38, 588-592. 253) Cheronis, J . C , Whalley, E.T., Nguyen, K.T., Eubanks, S.R., Allen, L.G.. Duggan, M.J., Loy, S.D., Bonham, K.A. and Blodgett, J.K. (1992) J. Med. Chem. 35, 1563-1572. 254) Moroder, L. and Wunsch, E. (1987) in: Gastrin and Cholecystokintn Chemistry, Physiology and Pharmacology (Bali, J.-P. and Martinez, J., eds.) pp. 21-32, Elsevier, Amsterdam. 255) Birdsall, H.H. and Rossen, R.D. (1984) J. Immunol 133, 1257-1264. 256) Manheimer, A. and Bona, C. (1985) Eur. J. Immunol 15, 718-722. 257) Susal, C , Guo, Z., Temess, P., Padanyi, A., Petranyi, G. and Opelz, G. (1990) Transplant. Proc. 22, 1893-1894.
969 258)Temess. P.. Marx, U.. Sandilands. G.. Roelcke. D.. Welschof. M. and Opelz, G. (1993) Clin. Exp. Immunol 933. 253-258. 259)Temess. P.. Kohl, I.. Hubener, G.. Battlstuta. R . Moroder. L.. Welschof, M.. Dufter. C . Finger. M.. Hain. C . Jung. M. and Opelz. G. (1995) J. Immunol, in press. 260) Lund. J.. Winter. G., Jones, P.T.. Pound. J.D.. Tanaka. T.. Walker. M.R. Artymluk. P.J.. Arata. Y., Burton. D.R. JeflFeris. R and Woof. J.M. (1991) J. Immunol 147. 2657-2662. 261)Roosneck. E. and Lanzavecchla. A. (1991) J. Exp. Med 173. 487-489. 262) Seed. B. (1994) Chemistry and Biology (1994) 1. 125-129. 263) Wank. SA..Pisegna. J . R and de Weerth. A. (1992) Proc. Natl Acad. Set USA 89. 8691-8695. 264)Reth. M. (1992) Arum. Rev. Immunol 10. 97-101. 265) Gambler, J.C.. Pleiman. CM. and Clark, M.R (1994) AnntL Rev. Immunol 12. 457-486. 266) Ravetch. J.V. and Kinet. J.P. (1991) Annu. Rev. Immunol 9. 457-492. 267) Watts. C. and Davidson. H.W. (1988) EMBOJ. 7. 1937-1945. 268) DeFranco, A.L. (1987) Anna. Rev. Cell Biol 3, 143-178.
This Page Intentionally Left Blank
Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.
971
C-NMR Spectroscopy of Coumarins and their Derivatives : A Comprehensive Review
B. Mikhova and Helmut Duddeck
1.
INTRODUCTION Coumarins constitute an important class in the realm of natural products with significant biolo-
gical activity (1). Although many books and articles have appeared since 1970 containing ^^C NMR data of various classes of natural products, only a very few of them deal with coumarins derivatives, and these mainly cover the literature of the 1970s only (2,3). Since that time, however, the data of many more coumarins have been published and NMR spectroscopy has seen a revolution. Thus, we believe that it is time to update the earlier reviews. Since the eariy 1970s, ^^C NMR spectroscopy has developed into one of the most valuable tools for structure elucidation of organic compounds and natural products because the ^^c NMR spectrum is a fmgerprint of a given compound. Moreover, ^^c data of a derivative not yet reported can often be extrapolated from the chemical shifts of compounds with related structural features. Nevertheless, it is still mandatory to perform a safe ^^C signal assignment of unknown molecules in order to avoid misinterpretations which may lead to erroneous conclusions. Therefore, we present a brief overview of NMR methods (section 2.1) which can be divided into two parts: a) The classical procedures have already been summarized by us before (3); nevertheless we include some of them here for the sake of completeness, b) New one- and two-dimensional NMR experiments have been designed during the 1980s which make some of the classical methods obsolete. The data in this review have been compiled in a data base using MDL ISIS-Base. Literature has been covered until spring 1995.
972 2.
METHODS OF i^C SIGNAL ASSIGNMENTS In general, ^^C NMR spectra are recorded under proton broad-band decoupling in order to
avoid the severe signal overlap which can easily occur because of the large one-bond carbonhydrogen coupling constants ^J(C,H) = 120-250 Hz. This procedure results in a breakdown of all signal splittings due to such couphngs. Owing to the low natural abundance of the ^^C isotope (ca. 1.1%), ^-^C NMR signals appear as narrow singlets if no further NMR-active nuclei with high natural abundance (e.g. ^^F or ^ip) are present in the molecule. This spectral simplification, however, produces a serious drawback in signal assignments since valuable coupling information is destroyed. Thus, a variety of assignment methods have been developed some of which are introduced in this chapter. There are five main areas: experimental NMR techniques, coupling constants, solvent effects, presence of auxiliaries and derivatization.
2.1
Experimental NMR Techniques
In the 1970s these techniques consisted mainly of ^H decoupling methods, the most prominent ones of which were the ^H broad-band (BB) decoupling and the single-frequency off-resonance decoupling (SFORD) techniques. In SFORD, the decoupler frequency is positioned outside the ^H resonance range (off-resonance). Thus, all carbon-hydrogen couplings are reduced to such an extent that only the largest coupling constants, namely ^J(C,H), give rise to small residual splittings, from which the number of hydrogens adjacent to the respective carbon atoms can be read directly; singlets correspond to quaternary carbons, doublets to methine, triplets to methylene and quartets to methyl groups. Nevertheless, in these spectra, signal overlap and second-order effects sometimes prohibit unambiguous interpretations in unfavourable cases. After 1980, NMR spectroscopy saw a revolution due to the introduction of new one- and twodimensional experiments, new superconducting magnets providing magnetic fields up to 17.6 T (^Hresonance frequency: 750 MHz) and an enormous progress in computer technology. New techniques, such as the recording of J-coupled spin echoes (Attached Proton Test), and INEPT and DEPT have been developed to circumvent these difficulties (4). By executing INEPT or DEPT involving polarization transfer from ^H to ^^C, the information about the number of adjacent hydrogens is also no longer reflected in residual signal splittings as in SFORD but in signal phases and intensities; CH and CH3 signals appear as positive and those of CH2 as negative singlets (Fig. Ic). Alternatively, it is possible to suppress all signals except those of CH (Fig. lb). Therefore, a comparison of these two DEPT spectra with the ^H BB-decoupled spectrum (Fig. la) allows an unambiguous assignment of all four sorts of CHn fragments (n = 0-3).
973
9-0-^0
b •'-'^
i II*' i>A hi»M' wi*i
OCH3
//5
150
10
iOO
50
Fig. 1: (a) ^H BB-decoupled ^^c NMR spectrum of dihydrobergapten (G5-1) and two DEPT spectra (b and c). Signals of quarternary carbon atoms are often of very low intensity owing to saturation effects (long spin-lattice relaxation times, TO and/or low nuclear Overhauser enhancements (NOE). Hence, they might easily be hidden below much more intense signals corresponding to hydrogen-bearing carbon atoms. In such cases, it is advisable to perform a J-coupled spin echo experiment with a delay time of l/[2-^J(C,H)] leading to zero-intensities of protonated carbon signals and leaving only signals of quartemay carbons. Fig. 2b shows that C-10 of hemiarin, which is hidden below the intense signals of C-3 and C-6 in the ^H BB-decoupled spectrum (Fig. 2a), appears clearly in the APT spectrum whereas httie residual peaks remain for C-3 (left, negative) and C-6 (right, positive) due to small deviations from the one-bond ^-"^C^H coupHng constant used for the delay setting.
974
9^0
' O
10
OCH3
7 2 9
10
—I
1
150
1—
100
50
Fig. 2: ^H BB-decoupled (a) and APT spectrum (b) of 7-methoxycoumarin (B7-4, hemiarin)
Correlations between ^H and ^^C nuclei via scalar couplings can be shown by two-dimensional COSY spectra (4-6). This is exemplified by a heteronuclear (^H,i3C) COSY spectrum of 7-methoxycoumarin (B7-4, hemiarin) shown in Fig. 3a (bottom). The cross-peaks connect the signals of hydrogen and carbon atoms directly attached. Cross-peaks indicating couplings over more than one bond can be detected by long-range variants of ^H^^^c COSY experiments or by COLOC (Fig. 3b) (7). It is interesting to note that C-7 bearing the methoxy group is identified by a COLOC peak due to a three-bond coupling between C7 and the methoxyl protons. In addition, cross-peaks proving couplings between C-2/H-4, C-9/H-4, C-5/H-4, C-7/H-5, C-9/H-5, C-4/H-5, C-8/H-6, C-9/H-8, and C-6/H-8 can be identified; the coupling of C-10 with H-3 is not visible in this scale due to overlap by the cross-peak representing U(C-3,H-3).
975
OCH,
10
2 9
OCH,
©
1
5
5('H)
i
4
CHjO^Y'^^^^
0 • ' •
cb ?
i
1
•
(P
®
J
® ^
OCH,
O 1
5('H)
i
<^
5
©
*
4 160
140
120
100
80
60
^
5(i3C)
Fig. 3: (a) iH,^3c COSY and (b) COLOC spectra of 7-methoxycoumarin (B7-4, hemiarin); the encircled signals in the COLOC spectrum correspond to one-bond ^U.^^C correlations already depicted in the ^H,^3C COSY spectrum.
976 If long-range heteronuclear couplings to only one single proton are required, it may be sufficient to record one-dimensional selective irradiation experiments, such as selective INEPT (8). In recent years inverse techniques (HMQC, HMBC etc.) have been developed providing analogous spectra as shown in Fig. 3 which, however, require much less material and can often be performed in a much shorter spectrometer time (9).
b
Jii'
vr-—1(V-
OCH,
7.0
Fig. 4:
6. 5
6. 0
5-0
4.!
4.Q
-i—r-
(a) ^H NMR spectrum of 7-methoxycoumarin (B7'.4, hemiarin); (b) NOE-difference spectrum with the methoxyl protons irradiated.
For heteronuclear correlations it is often mandatory to have an unequivocal ^H signal assignment. Hydrogen atom distances in space can be monitored by various methods based on homonuclear NOE experiments. Among them are the one-dimensional NOE-difference technique (Fig. 4) and two-dimensional experiments such as NOESY and ROESY (10). Fig. 4b shows the NOEdifference spectrum of 7-methoxycoumarin (B7-4, hemiarin) with the methoxyl protons irradiated, displaying responses of the H-6 and H-8 atoms which are close in space with respect to the methoxyl group rotating around the C^-0 bond. Heteronuclear NOE experiments irradiating protons and observing ^-^C nuclei have proven to be useful as well (10, 11).
977 2.2
Coupling constants
Carbon-hydrogen coupling constants are useful aids for peak assignments in the ^^C NMR spectroscopy of coumarin derivatives. A simple train of consecutive FID recordings without decoupling is, however, disadvantageous. The gated-decoupling method (Fig. 5) is, rather, recommended. Here, delays are inserted between successive recordings during which the decoupler is switched on (12). This technique yields a considerable signal-to-noise ratio improvement, because the NOE is partly regained. Further, larger pulse durations (up to 90° pulses) are allowed which, at least in part, compensate the loss of time during the decoupling delays.
CH,0
b~JLi OCH3
150
100
50
Fig. 5: (a) Broad-band and (b) gated decoupled ^-C NMR spectra of 7-methoxycoumarin (B7-4, hemiarin), (c) expanded section of (b).
978 This experiment is not only useful for determining ^^C^H coupling constants but is an alternative for identifying quartemary carbon atoms overlapped by other signals of carbons with hydrogen atoms attached (see Fig. 5c). Peak assignments by observing ^^C,^H coupling constants (13) have been described in various pubhcations dealing with the ^^C NMR of coumarins (2, 14-30). Fig. 6 shows some typical one-bond coupling constants [^J(C,H), in Hz] in coumarin (A-1) (21) and psoralen (F-1) (23).
162.9 164.3 164.3 r ^ J ^ ^ ^ ^ r " - " ^ 172.4
164 3 ^ ^ ^ A v ^ 165.3
Fig. 6:
One-bond coupling constants [^J(C,H), in Hz] in coumarin (A-1) and psoralen (F-1).
In coumarin itself all ^J(C,H)-values are within a small range (about 163-165.5 Hz) except for C-3, which has the only value larger than 170 Hz. This was used to differentiate between C-3 and C-8 (21). The same pattern was found for a variety of derivatives; for example, it was helpful in assigning C-3 and C-6 in some aminocoumarins (28); only for coumarins bearing a 4-hydroxy group is ^J(C-3,H-3) considerably smaller (15, 19, 24). In furanocoumarins, the largest U(C,H)-values appear in the furan moiety, that for C-2' even exceeding 200 Hz (23, 26, 27, 31, 32). Carbon-hydrogen couplings via more than one bond have also been investigated intensively, and have proved to be very useful for signal correlations (17, 19-21, 23). Three-bond couplings, ^J(C,H) for example, depend on the geometry of the molecular moiety; they are generally larger in transoid than in cisoid orientations. Thus, the assignment of the C-2 and COOH signals of 3-carboxycoumarin (B3-13) was accomplished: 3J(C^H'^) = 10.3 Hz {transoid); 3J(COOH,H^) = 5.4 Hz (cisoid) (20, 21), as well as the assignment of the C-2 and C-a signals of some carbonyl and aromatic substituents at C-3 (28). Even if the °J(C,H) values (n > 1) cannot be determined exactly owing to second-order effects, the signal shapes often give helpful hints (33). Giinther et al (34) established a fingerprint rule for orr/io-disubstituted benzenes, by which a- and |J-methine carbon signals can be distinguished. This rule can be employed successfully for the distinction of the C-5 and C-6 signals (16) and of the C-6 and C-8 signals (28) of coumarin derivatives. Coupling pattern differences of the carbonyl carbon signals due to long-range couplings represent a helpful technique for the differentiation between benzocoumarins and benzochromones (30).
979 The influences on the ^^C^H couplings due the introduction of a methyl group in methyl-angelicins and other annulated furanocoumarins were highlighted (26, 27). The corresponding influence of a methoxy substituent in coumarins is discussed (29), and on this basis the previous assignment of C-2 and C-10 as well as C-2 and C-5 for the natural product citropten (C57-4) has been reversed. In one instance the method of biosynthetic labelling (35) was applied to synthesize ^^C-enriched aflatoxin Bi, which contains a coumarin moiety (36-39). In these papers several assignments were assisted by the enhanced ^^C signal intensities and by one-bond ^^C^^C coupling constants, U(C,C).
2.3
Solvent effects
The i^C shielding is not very sensitive to solvent changes. Although coumarins contain a polar lactone residue, the ^^C chemical shifts of coumarin and its derivatives remain constant within about ±1 ppm when the deuteriochloroform solvent is replaced by deuterated dimethyl sulphoxide (23, 40), despite the different complexing abilities of these two solvents. Thus, '^C chemical shifts appear to be rather insensitive to solvent changes and only if small substituent effects on the ^^C chemical shifts (SCS) are to be discussed is it advisable to record all spectra in identical solvents. Basically, this also holds for protic organic solvents. Shght alterations of ^J(C,H)-values in the order of 3-5 Hz might occur on solvent changes (23). Much more pronounced effects were reported by Sojka (15) who compared the ^-^C chemical shifts of a number of coumarin derivatives in chloroform and in 96% sulphuric acid and by Yufit et al. (74) for some 4,7-diaminocoumarins in 10-25% and 40% sulphuric acid solutions. For coumarin itself the differences are: C-2, 13.2; C-3, -5.3; C-4, 16.5; C-5, 4.3; C-6, 7.1; C-7, 8.0; C-8, 3.4; C-9, 0.6; and C-IO, 3.7 ppm, when the data of the chloroform spectrum are subtracted from those in sulphuric acid (15). Even the 'J(C,H) values are sensitive, and vary up to 16 Hz. These dramatic effects are explained by protonation of the carbonyl group and by considering a different balance of mesomeric forms of the molecule. They are not constant in their magnitude, however, when coumarin is substituted in different positions. 2.4
Signal shifts in the presence of auxiliaries
Another possibiUty of producing explicable signal displacements is the addition of complexing reagents. Bose et al. (41) reported that titanium tetrachloride in deuteriochloroform can be used as a shift reagent in ^H and ^^C NMR spectroscopy, and applied this method to coumarin and some angular furanocoumarins (42). The use of lanthanide shift reagents (LSR) to simplify NMR spectra and for structure determinations was recognized in the eariy 1970s (43). Although it was shown that contact contributions are generally not neghgible in ^^C NMR (44, 45), this compHcation can be overcome by using ytterbium
980 complexes (44, 46). In this case, contact shifts are essentially restricted to carbon atoms directly bonded to the complexation site.
2.5
Derivatization
Another method is to compare the ^^C chemical shifts of a given compound with those of a derivative easy to prepare. Although this procedure is somewhat obsolete after the advent of modem one- and two-dimensional NMR techniques we include it due to its importance in the early years of coumarin ^^C NMR spectroscopy. Structural information can be gained from signal displacements by derivatization which may be realized either in situ by adding a given reagent to the substrate solution in the NMR tube, or by a separate chemical reaction prior to the measurement. The reaction of hydroxylated compounds (alcohols or phenols) and amines with trichloroacetyl isocyanate (TAI) (47) to form urethanes and ureas, respectively (48), can be performed within the NMR tube. This method, however, is restricted to compounds containing those functionalities and has not been applied widely. Derivatization prior to the NMR experiment may be very valuable, e.g. when one or more hydrogen atoms can be replaced by deuterium at specific positions (49). Thereby, for example, a methine signal is split into a 1:1:1 triplet due to one-bond carbon-deuterium coupling (deuterium spin quantum number I = 1, cf. signal of deuteriochloroform). Further, its total intensity is decreased by less efficient spin-lattice relaxation and NOE. Thus, in practice, the signal of a deuteriated "•^"' i ahnost disappears in the spectrum. For adjacent carbons slight line broadenings [°J(C,D), n -^ ... and isotope shifts of a few tenths of a ppm have to be expected (50). Selective deuteriation of coumarins has been used in some instances. 3-Deuterio- and 4-deuterio-coumarin have been investigated for signal assignments as well as 3-deuterio-4-hydroxycoumarin and 6,7-dideuterio-4-hydroxycoumarin (19, 47, 51, 52). In the latter case the C-5 signal was unambiguously identified and distinguished from that of C-6. Selective deuteriation, however, is depreciated since it is often, synthetically, a laborious procedure. The preparation of coumarinic thionolactones and comparison of their ^^c NMR data with those of the parent compounds, has been used as a tool for spectral assignments, at least for the pyrone ring carbons (23, 40, 53, 54).
981 3.
13C CHEMICAL SHIFTS AND SUBSTITUENT EFFECTS 3.1
General
Numerous publications have appeared for calculating ^-^C chemical shifts by ab initio and semiempirical MO methods (55-58) and correlating the experimental shielding data with the physicochemical parameters and structural properties of the molecules (59, 60). There have also been attempts to predict the ^^C chemical shifts of coumarins. The first to do so was Sojka (15), who found a fairly good correlation between the carbon shifts of coumarin and its protonated derivative with the n charge densities calculated by the CNDO/2 method (61). Shortly afterwards, Giinther et al (14) reported that substituent effects (SCS) in various mono- and dimethoxycoumarins correlate weU with the HMO atom-atom polarizabihties, TCy (62): SCS = KTtij [K(OMe) = 80.13] Furthermore, Giinther et al. (14) found a result similar to Sojka's (15) when correlating coumarin shifts with ;: charge densities calculated by the Hiickel MO method (63). This, however, fails for methoxylated coumarins (14) demonstrating that simple charge density - shift relationships are not generally applicable rehably. Ernst (20) reported a linear relationship between P-methyl substiment effects in unhindered methylated coumarins and the K bond order, Pjc, of the C-a—C-P bond calculated by the INDO MO method (64): pSCS = 13.49 - 19.80 P;c In a more rigorous way, Grigor and Webb (65) reproduced the ^^C shieldings of coumarin and some mono- and dimethoxycoumarins by refmed INDO MO calculations and found that, in addition to atom-atom polarizabilities (14) and n bond orders (20), other factors such as excitation energies and electron-nucleus distances, (r'^)2p, play an important role in the determination of the ^^C chemical shifts of these compounds. The effect of the nature of the substituent in the pyrone ring on the electron structures of 3amino- (B3-11), 3-hydroxy- (B3-3) and 3-carboxycoumarin (B3-13) was evaluated on the basis of quantum-chemical (CNDO/2) calculations and ^^C NMR data (66). The character and degree of the relationship between the chemical shifts and the electron densities on the carbon atoms were established by means of regression analysis. Two studies correlate substituent-induced chemical shifts (SCS) with substituent parameters. Gottlieb et al (33) investigated the SCS on the pyrone ring carbons C-2, C-3, C-4, C-9 and C-10 in 6- and 7-substituted coumarins, and found good correlations of the ring-junction atoms C-9 and C-10 with the Hammett constants G^ and of C-2 and C-4 with Om, and, separately, with Cp. More-
982 over, the correlation of the C-3 chemical shifts with c^ is excellent. This behaviour is strongly reminiscent of related data for substituted styrenes (67) so that Gottlieb et al concluded that the SCS are transmitted essentially via the -CH=CH-CO- moiety, the lactone group insulating the alternative transmission pathway.
Table 1.
Substituent effects (SCS) in various monosubstituted coumarins
Substituent
C-2
C-3
C-4
C-5
C-6
C-7
C-8
C.9
3-Br 3-Cl 3-OH 3-NH2 3-NMe2 3-Me 3-COOH 4-OH 4-OMe 4-Me 5-OMe 5-Me 6-Br 6-Cl 6-OH 6-OMe 6-OAc 6-NH2 6-NO2 6-Me 6-CHO 6-COOH 6-CN 7-Br 7-Cl 7-OH 7-OMe 7-OAc 7-NH2 7-NO2 7-Me 7-COOH 8-OH 8-OMe 8-Me
-4.1 -3.9 -1.9 -1.6 -2.4 1.7 -3.2 2.5 2.1 0.1 0.7 0.2 -1.0 -0.8 -0.3 -0.3 1.3 2.4 -0.4 0.5 0.3 0.8 -0.2 -1.0 0.6 0.3 0.4 1.3 3.2 -0 3 0.5 1.2 -0.4 -1.0 0.5
-5.4 4.4 25.3 16.9 20.9 9.3 1.8 -25.1 -26.4 -1.3 -1.8 -0.5 1.1 1.2 -0.3 0.7 1.0 0.0 2.4 0.1 1.5 0.9 2.3 0.2 0.1 -4.9 -3.7 -0.4 -6.8 37 -1.0 1.6 -0.3 0.5 -0.1
0.4 -2.9 -28.6 -18.2 -27.1 -4.4 4.9 22.4 22.6 8.7 -4.8 -3.2 -1.8 -1.70.2 -1.0 0.41.0 -0.2 -0.2 0.4 0.7 -0.4 -1.2 0.1 0.7 -0.3 0.6 1.6 -1.0 -0.2 0.2 0.9 -0.6 0.2
-1.2 -0.2 -1.8 7 -2.7 -1.1 2.2 -4.7 -5.3 -3.5 28.4 8.2 1.8 1.3 -15.7 -18.3 7.3 -15.9 -3.7 -0.3 2.9 2.6 5.2 0.5 1.2 1.5 0.6 1.2 1.4 1.4 -0.5 0.5 -9.7 -9.1 -2.5
0.1 0.5 0.1 0.1 -0.7 -0.1 0.5 -0.9 -0.7 -0.2 -19.0 1.3 -7.7 5.0 29.3 31.5 22.8 20.0 20.2 9.7 9.0 3.4 -15.5 3.2 1.1 -11.1 -12.3 -5.4 -11.8 -4.9 1.2 1.5 0.0 -0.5 -0.4
-0.2 0.0 -4.3 -7.0 -5.2 -1.4 2.6 0.3 0.4 -0.1 0.8 -0.2 2.4 -0.4 -12.1 -12.8 -5.8 -11.3 -4.8 1.0 0.9 1.8 3.6 -6.3 6.3 29.8 30.8 22.0 20.7 16.4 11.3 2.8 -13.4 -18.2 1.4
-0.4 -0.3 -0.8 -0.9 -1.5 -0.1 -0.2 -0.3 0.1 0.5 -7.2 -1.7 1.9 1.6 0.5 1.3 1.7 1.1 2.0 0.1 1.9 0.9 2.3 3.4 0.9 -13.9 -15.8 -5.7 -15.9 -3.7 0.5 2.1 28.3 30.8 9.9
-1.3 -1.7 -4.7 -5.9 -4.7 -0.7 0.7 -0.2 -0.8 -0.4 1.5 0.7 -1.2 -1.8 -7.1 -5.5 -2.2 -6.7 4.0 -1.7 4.2 3.1 2.9 0.2 0.6 1.8 1.7 1.0 2.7 0.2 0.3 0.0 -11.5 -9.1 -1.5
C-IO 0.0 0.0 1.9 3.1 1.5 0.8 -0.7 -2.7 -3.3 1.2 -9.0 -1.1 1.4 0.8 0.3 0.3 1.0 0.9 0.7 -0.2 0.9 0.2 2.1 -1.2 -0.9 -7.3 -6.5 -1.6 -8.5 5.3 -2.3 3.8 0.9 0.5 -0.2
Rabaron et al (24) described a three-parameter correlation of the ^^C chemical shifts of various 3-substituted 4-hydroxy- and 4-hydroxy-7-methoxy-coumarins with J, % and Q (68): SCS = a J -hb
:R
+ cQ -H d
983 The use of Swain and Lupton's parameters ^ and !R (69) alone does not yield satisfactory results (24). Although the aforementioned calculations and correlations are of great merit by allowing deep insight into charge densities and SCS transmission mechanisms, practical applications are hampered by their inherent restrictions and limitations. Comparisons of SCS in coumarins with those in related molecular systems, such as substituted naphthalenes, also give interesting results which can be used diagnostically. In the following sections, the SCS of various substituted coumarins (Table 1) are discussed along these lines.
3.2
ipsola Effects
In general, the a SCS are similar to those of corresponding 1- and 2-substituted naphthalenes (20, 49, 70-72). There are only a few exceptions. For example, the a-methyl effect in 7-methylcoumarin is considerably larger than for all other isomers (11.2 vs. 8.3-9.7 ppm) (20). Despite the similarity of the a-hydroxy effects in 4-hydroxycoumarin and 1-naphthol (22.4 and 23.4 ppm, respectively), the corresponding methoxy effects are quite different (22.6 and 27.6 ppm, respectively). In 6-cyano- and 7-nitrocoumarin the a SCS are smaller by ca. 3 ppm than in 2 cyano- and 2-nitronaphthalene. Again, there is no satisfactory explanation.
3.3
orthol^ Effects
By analogy with naphthalenes (72), the p SCS in coumarins depend strongly on the positions of the substituents and the affected carbons: for 4-, 5- and 8-substituted coumarins the substituents shield the neighbouring methine carbons to a greater extent than the quartemary carbons; for 6- and 7-substituted coumarins the differences of the (3 effects are smaller than in the corresponding naphthalenes. The averages of the two p SCS for a certain substituent in both systems, however, are approximately the same, showing that the p SCS reflect a subtle balance of canonical forms and electric field influences (72). This is demonstrated in Fig. 7 for 4-hydroxycoumarin: The electron density at C-3 is higher than at C-10, because the canonical structure A is much more favoured than B. The latter does not retain the n electron sextet of the benzene ring (72) and, further, it is tetraionic. The possible a-pyrone-y-pyrone tautomerism (52, 73), however, is not involved in this discrepancy between the coumarinic and naphthalenic system, since methoxy derivatives reveal similar effects.
984
17.0
Fig. 7.
-25.1
P SCS in 1-naphthol and 4-hydroxycoumarin.
Abnormal P effects exist for substituents in the 3-position. This can be interpreted in terms of intramolecular interactions (electronic, steric and/or by hydrogen bridging) between the substituents and the neighbouring carbonyl group. Such interaction effects are discussed below. It has already been noted that p-methyl effects at ortho carbons can be correlated to n bond orders (20).
3.4
meta/y.r,u Effects
Substituent effects at mera-positioned carbons are small, in agreement with those of naphthalene derivatives. Methyl and carboxyl SCS are negligible, whereas those of hetero substituents are generally deshielding (up to 2.5 ppm). An exception has been reported if the yanti carbon is C-2 (carbonyl). The effect of a C-4—^NEt2 group on C-2 in C47-3 (as compared to B7-37) is +1.7 ppm (74). Some authors claimed that yanti effects of a 3-(l,l-dimethylallyl) group on C-10 can be used to distinguish between C-3 alkylated linear dihydrofurano- from dihydropyranocoumarins (75a). No such effect is observed in the corresponding angular derivatives. This fact is proposed as spectroscopic criterion for distinguishing between these isomers (75b). 3.5
y.yn Effects
This type of molecular arrangement, which leads to the well known ("steric") diamagnetic y SCS (59, 72), is represented only in a few cases among the available data for monosubstituted coumarins: 4-X-C-5 (X = Me, OH, OMe) and 5-Me-C-4. As expected, the SCS values are -3.2 to -5.3 ppm. Additionally, some ysyn SCS can be estimated from the spectra of some di- and tri-substituted coumarins. The effect of the 4-phenyl group on C-5 in 4-phenyl-7-hydroxycoumarin, when compared with 7-hydroxycoumarin, is only -1.5 to -2.0 ppm if intramolecular interaction is permissibly neglected. The low value may be a consequence of the anisotropy of the phenyl group; this, however, is not
985 observed in aliphatic molecules such as 2-phenyladamantane (76). Likewise, a 5-methoxyl Ysy,, SCS of -5 to -6 ppm at C-4 may be deduced from the data for C57-4, C58-1 and D578-8. 3.6
para/b Effects
As has been shown by Ernst (71, 72) for naphthalene derivatives, para SCS can amount up to ±10 ppm, and can be correlated linearly with total charge density changes calculated by INDO MO methods. This also seems to hold for 6- and 7-substituted coumarins, because fairly good correlations exist between respective para effects in both molecular systems. Exceptions only occur for 3-carboxycoumarin and 3-hydroxycoumarin, which are probably due to intramolecular hydrogen bridging. Analogous observations are noted in connection with ortho effects. Ysyn and para (5) effects of alkyl and 0-alkyl substituents in linear furanocoumarins can be used to distinguish between the substituent's position, C-5 or C-8 (77). Moreover, such effects of substituents at C-5 allow a differentiation between linear and angular furanocoumarins (77). 3.7
Long-range effects
Long-range effects differ in some cases from those of naphthalenes (20, 72), as demonstrated in Fig. 8.
X = COOH X = OH Fig. 8.
+2.5 -2.1
+0.8 -0.3
+1.6 -4.9
Long-range SCS on carbon atoms marked by "•" ; naphthalenes compared with coumarins.
Apparently, in 6-substituted coumarins the effects cannot be transmitted to the carbonyl group, since the corresponding canonical form is highly unfavoured. On the other hand, in 7-hydroxycoumarin (B7-3, umbelliferone) the canonical form depicted in Fig. 9 is of exceptional importance:
HO"^ \ ^ Figure 9.
-Q'
"O
Canonical form of 7-hydroxycoumarin (B7-3, umbelliferone)
A similar value is observed by comparing the C-3 chemical shifts of 4-methylumbelliferone (C47-6, 5 = 110.4) (19) and 4-methylcoumarin (B4.12, 5 = 115.1) (20). An analogous comparison of 6(C-3) of C47-12 (15) and B4-12 (20) gives an SCS of -7.1 ppm for the diethylamino substituent, which is an even stronger electron-donating function than a hydroxy group (71b, 72).
986 Other long-range SCS are small, or even negligible, within the limit of experimental error and reproducibility.
4.
EFFECTS OF INTRAMOLECULAR INTER ACTION ON SUBSTITUENT EFFECTS It is a well known and often reported fact that individual SCS in molecules with more than one
substituent are additive, unless there is an intramolecular interaction between them. This is also mentioned in a number of ^^C NMR smdies on coumarinic compounds (18, 19, 24, 78, 79). A systematic investigation of the data in this review confirms these findings (see Table 2). Table 2.
Non-additivity effects (5cxp - 5cai)* in di- and trisubstituted coumarins.
Substituents 3-Br, 4-OH 3-Cl, 4-OH 3-Me, 4-OH 3-COOR, 4-OH 3-Br, 6-Br 4-OMe, 5-Me 4-Me, 6-OH 4-OH, 7-OMe 4-OMe, 7-OMe 4-Me, 7-OH 4-Me, 7-Me 5-Me, 6-Me 5-Me, 7-OMe 5-Me, 8-Me 6-OH, 7-OH 6-OMe, 7-OMe 6-Me, 7-Me 6-Me, 8-Me 7-OH, 8-OH 7-OMe, 8-OMe 3-Cl, 4-OH, 7-OMe 3-Me, 4-OH, 7-OMe 4-OMe, 5-Me, 7-OMe 4-OMe, 5-Me, 8-OMe 4-Me, 6-OH, 7-OH 4-Me, 6-Me, 7-Me 4-Me, 7-OH, 8-OH 4-Me, 7-OH, 8-Me
C34-1 C34-2 €34-11 €34-14 €36-1 €45-2 €46-8 €47-1 €47-2 €47-6 €47-13 €56-1 €57-6 €58-4 €67-1 €67-5 €67-61 €68-1 €78-4 €78-6 D347-1 D347-3 0457-2 D458.1 D467-2 D467-12 D478-1 D478.4
€-2
€-3
€-4
€-5
€-6
€-7
€8
€-9
€-10
-0.3 -0.4 -1.4 -3.1 0.4 -0.1 -0.3 -0.9 -0.3 -0.4 0.4 0.0 -0.5 -0.4 1.0 0.2 0.2 -1.6 0.8 -0.1 0.3 -1.2 -0.4 0.1 0.3 0.0 0.3 -0.5
3.5 3.2 -0.1 1.2 -0.1 1.9 -0.4 1.1 1.0 0.2 0.1 0.0 -0.4 -0.7 0.8 0.1 0.2 -1.4 0.5 0.3 4.4 0.8 1.6 0.2 0.7 0.0 0.6 0.0
-4.1 -2.6 -1.9 1.1 1.3 5.8 1.2 0.4 0.3 0.2 0.6 0.7 -0.3 -0.8 0.4 0.5 0.4 -0.8 0.2 0.4 -2.1 -0.9 6.6 6.8 -0.2 0.7 2.0 -0.1
1.3 0.3 0.8 -1.1 0.9 6.1 0.5 0.4 0.2 0.1 0.4 -1.8 -0.2 -1.0 -1.0 -2.4 0.9 -0.7 -0.5 2.8 0.9 1.3 6.6 5.6 -0.8 1.2 -0.7 -1.1
0.6 0.4 0.3 0.5 0.6 2.5 -1.2 0.5 0.2 -0.2 0.2 -2.6 0.0 -0.9 0.6 2.6 -1.9 -1.1 -0.3 -3.3 0.6 0.4 2.7 1.9 0.4 -2.0 -0.5 -0.9
0.8 0.4 0.6 0.8 0.2 -0.5 0.1 0.2 -0.4 -0.3 0.2 1.1 -0.7 -1.0 1.1 3.0 -1.9 -0.9 1.8 11.0 0.0 0.5 -1.2 -0.3 0.8 -2.1 1.5 -3.8
0.6 0.5 0.0 0.6 0.3 0.3 -0.3 0.3 -0.6 -0.6 0.0 -0.5 -0.5 -1.8 0.2 -2.0 0.6 -1.8 1.8 4.9 0.6 0.0 -0.4 -0.1 -0.6 0.3 1.2 -2.9
-0.7 -0.8 -1.2 -0.7 0.2 0.9 -0.1 0.2 -0.3 -0.3 0.1 1.9 1.0 0.3 0.5 -0.1 0.3 -1.1 0.0 1.6 -0.7 -1.2 0.8 -0.2 -0.2 0.1 0.1 -0.7
0.2 -0.1 -0.5 -1.1 -0.4 0.0 -0.2 -0.4 -0.5 -0.5 0.1
5cxp, experimental chemical shift; 5caic, chemical shift calculated assuming additivity of individual SCS. Experimental value not reported.
b
-0.1 -1.0 -0.4 -1.4 0.6 -1.0 0.3 0.9 -0.5 0.9 -0.3 0.0 -1.1 0.4 -0.3 -0.3
987 Non-additivity (NA) effects occur only when functional groups within the molecule interact electronically, sterically, by hydrogen bridging or by other mechanisms. One case has already been discussed in the previous section - the effects of substituents in the 3-position are altered by the influence of the neighbouring lactone group. On the whole, one has to allow for NA effects if the substituents are in close proximity.
4.1
Substituents at vicinal carbon atoms
The C-3 signals of 3-bromo- and 3-chloro-4-hydroxycoumarins (C34-1, C34-2 and D347-1) appear at higher, and those of C-4 at lower frequencies than expected by assuming additivity. In methylhydroxy derivatives C34-11, D347-3 and D478-1 only the hydroxylated carbons are affected; and only slight deviations from additivity, if at all, are observed for dimethyl (C56-1, C67-61 and D467-12) and dihydroxy derivatives (C67-1, C78-4, 0467-2 and D478-1). Apparently, these interaction effects are mainly of electronic rather than steric origin. In dimethoxy compounds C67-5 and C78-6, however, clear NA effects are observed at the substituted and the neighbouring unsubstituted carbons. The finding that the methine signals in the a position to methoxylated carbons (C-5/C-8 in C67-5 and C-6 in C78-6) feature negative NA effects, suggests that the conformational behaviour of the methoxy groups is sterically perturbed. The methyl groups are forced outwards, increasing their diamagnetic Ysyn effects (see Fig. 10).
Fig. 10.
Steric perturbation of methoxyl groups in C67-5.
Such NA effects increasingly lead to problems in signal assignments with the number of substituents becoming higher, for example in 6,7,8-trioxygenated coumarins (80). Carbon atoms of the substituents themselves are affected as well (section 5)
4.2
Substituents in peri position
This molecular arrangement is present in coumarins with substituents simultaneously in the 4-and 5-positions. For example, non-additivity effects can be evaluated for two derivatives (D457-2 and D458-1) bearing 4-methoxy and 5-methyl groups. They are distinctly positive at the substituted carbons' signals (+5.6 to +6.8 ppm), and the neighbouring atoms C-3 and C-6 are also affected.
988 whereas the individual SCS are additive for the quaternary carbons C-9 and C-10. The magnitudes and signs of NA effects at these six atoms, however, seem to be strongly dependent on the nature of the substituents. This can be guessed by inspecting dimethyl- (81, 82), diamino- (71b) and dihalonaphthalene (71c). The origin is apparently severe steric substituent interaction and molecular distortion (83).
4.3
Highly substituted coumarins
It stands to reason that with an increase in the number of substituents attached to the basic coumarin molecule there is an increase in the NA effects, and these become increasingly confusing. For example, the individual SCS of a given group in, for example, a furanocoumarin may be quite different from that in coumarin itself or in benzene. Thus, spectral interpretations which are based solely upon such SCS comparisons should be regarded with caution, and a rigorous signal assignment by modem multipulse NMR techniques is mandatory. Indeed, misassignments by neglecting this fact have appeared in the literature, and it is mentioned in section 6 that it was not possible to correct all of them from the tables 7-13. Fig. 11 demonstrates such different 8-methoxy SCS values (54). -6.9
-15.5
30.8^0^^0 OMe
Fig. 11.
Methoxy SCS on the benzenoic carbon atoms in benzene, coumarin (A-1) and psoralen (F-1).
This is even more drastic for bromo substituents (54); see Fig. 12. Comparing the ^^C chemical shifts of imperatorin and its tribromo derivative, one obtains 5-bromo SCS which, in part, differ enormously from those of bromobenzene; the a effect, for example, is 20 ppm larger.
Fi gure 12. Bromo SCS on the benzenoic carbon atoms in bromobenzene and the tribromofuranocoumarin derivative F58-1.
989 5.
13C CHEMICAL SHIFTS OF METHYL, METHOXYL AND SOME OTHER CARBON ATOMS As expected, ^^C chemical shifts within substituents are quite uniform unless the carbon atoms
are close to the coumarin system. In particular, carbon atoms directly attached to the ring system (a) display some dependence on their relative position. To some extent, this can be observed for P-positioned carbon atoms as well. These effects are demonstrated for methyl and methoxyl substituents (Table 3 and Table 4, respectively) as representative examples, as well as for some other atoms which provide valuable hints concerning the stereochemistry of ring-annulated coumarins.
5.1
Methyl substituents
In general, resonances of methyl carbons appear at 5 « 17, if they are attached to C-3, C-4 or C-5. At position 6 or 7, however, the values are larger (8 = 19-21) because, in contrast to those at C-3 through C-5, there is no syn-periplanar atom with respect to these methyl carbons which experiences a shielding y-syn effect (60, 84). On the other hand, methyl groups at C-8 show an even stronger shielding (5 = 14-16) than those at C-3 to C-5. Apparently, the endocyclic oxygen atom is more effective in this mechanism than a methine group. These chemical shifts, however, may be changed significantly if further substituents are closeby. Especially, substiments on neighbouring carbons atoms of the coumarin system may strongly shield the methyl carbons. Here again, methyl carbon at 3 or 8 position are particularly sensitive. On the other hand, substituents in peri position deshield a methyl carbon due to a h-syn effect (60, 85), an influence which is exceptionally strong (4-5 ppm). Substituents further away do not have effects larger that a few tenths of a ppm. Some typical examples for the abovementioned effects are shown in Fig. 13.
O
o
O
o 5 = 15.3
6=18.2
HO'
"Y"
"O^
CH ^ 3
Fig. 13.
5 = 8.0
"O
"^
5 = 22.7
^O'
"O
CH3
OCH3
Y O OCH, ^
Methyl chemical shifts depending on neighbouring substituents
O
990 Table 3.
Methyl resonances in coumarins (selection of representative examples) C-3
B3-14 B4-12 B5-2 B6-8 B7-39 B8-3 C34-11 C38-2 C46-3 C46-7 C47-6 C47-13 C57-6 C58-4 C67-61 C68.1 C68.2 D347-3 D347-4 D378-2 D457.2 D457-3 D457-8 D458.1 D467.2 D467-8 D467-11 D467-12 D478-1 D478-4 E3457-1 E3457-2 E3458-1 £3467-1 £3467-2 £4567-1 £4678-1 £-34578-1
C-4
C-5
C-6
C-7
C-8
17.1 18.5 18.2 20.7 21.7 15.3 9.8 18.8
22.4 21.1 18.4 18.0 18.6
21.6 18.4 14.3
17.0 19.2 19.6
20.3 14.1 15.6
9.5 14.9 8.0 23.4 23.3 21.8
20.7 22.7
18.1 16.3 16.9 19.4
18.3 18.6 18.3 18.4
20.1 8.0
22.1
10.5 19.4
22.6 16.4 15.1 16.0 20.5 10.4
9.5 22.0
Other subsdtuents 4-OH dithiocoumarin 4-OEt 6-Cl 7-OH 7-OMe 6-Et 4-OH, 7-OMe 3-CH2CH2COOEt, 7-OH 4-CF3, 7-OH 4,7-[OMe]2 5J-[OH]2 4-CF3 4,8-[OMe]2 6,7-[OH]2 4-Pr\ 7-OH 7-NEt 7,8-[OH]2 7-OH 4,7-[OMe]2 3-CH2COOEt, 5,7-[OH]2 3,4,8-[OMe]3 4-Pr\ 3,7-[OMe]2 3-CH2COOEt, 6,7-[OH]2 4-SMe, 6-OMe, 7-OH 6-CH2CH=CH2, 7-OAc 4J,8-[OMe]3
It is interesting to see that the 8-values of the methyl resonances in 3,8-dimethyl-dithiocoumarin (C38-2) are clearly larger [CH3-(C-3): 5 = 18.8 and CH3-(C-8): 5 = 22.4] than those expected for the corresponding coumarin itself.
991 The Y and 5 effects discussed above are essentially additive as shown in Fig. 14:
5 = 14.9
CH,
HO
1 A^i^ - 0 ^ 0
5 = 19.4 CH3
r ^ T"^^"
^k:^^^
HO'"
^R
0 ^ 0
R = CH2-CH2-COOEt Fig. 14.
^^C chemical shifts of methyl carbons experiencing y and 6 effects from neighbourijig substituents
Methyl resonances of a number of methylated furano-annulated coumarins have been reported (26, 27). The chemical shifts of methyl groups attached to the coumarinic rings correspond to the values in Table 3. Those, however, at C-2' are 8 « 14 (5 ~ 11.5 if there is another methyl group at C-30. Chemical shifts of methyl carbons at C-3' are 5 « 9.5 (5 = 8 if there is another methyl group at C-20. 5.2
Methoxyl substituents
Since the methyl carbon atoms of methoxyl groups are in P position with respect to the coumarin system, their ^^C chemical shifts are very imiform being between 6 = 55 and 56.5 irregardless of their position. Only steric crowding affords larger 5-values (5 = 60-62); i.e. , if the methoxyl group is flanked by two ortho and/or peri oriented groups. This rule holds also if one or both orthoneighbours is/are part of annulated rings, as for example in furano- or pyranocoumarins. Some typical instances are collected in Fig. 15.
5=55.5
^O^
"O
HO OCH, ^ 6 = 61.2
Fig. 15.
Methoxyl chemical shifts depending on neighbouring substituents
992 Table 4.
Methoxyl resonances in coumarins (selection of representative examples); values marked by "*" may be interchanged. C-4
B4-3 BS-l B6.4 B7-4 B8-2 €34-22 C37-3 C38-1 C47-1 C47-2 C47-7 C57-4 €57-6 €58-1 C58-3 C67-2 C67.4 C67-5 C67-6 C67-28 C78-5 C78-6 €78-8 C78-9 C78-16 D347.3 D357-1 D367.1 D378-3 D457-2 D457-6 D457-10 D457.il D457.12 D458.1 D467-3 D467.4 D567-7 D578.1 D678-4 E3457-1 E3457-3 E3467-3 £3478-2 E4568-1 E5678.1 E.34578-1
C-5
C-6
C-7
C-8
Other substituents
56.3 56.1 55.5 55.6 56.4 61.9 55.7 55.6 55.8 55.6 55.5 55.9 55.4
56.1 56.1
5-Me 56.8 56.8
55.8 56.3 55.2 56.2 56.1
56.2 56.3 56.2 56.3
60.6 61.6 61.3
55.9 55.8 56.7 56.3 55.8
56.4 56.2 55.4
61.0
55.9 55.4 55.9*
55.6 55.6*
56.0
55.8 56.4* 56.3 56.1
55.9*
56.3* 56.3* 60.9 56.3
60.5 55.8 56.5
61.7 55.1 55.7 56.5 56.1
61.3
56.1
61.2
61.4 60.3 60.4
3-CH(Ph)-CH2Ac 3-Ph 3-NMe2 4-OH 4-Me
5-CH2CH=C(Me)2 6-OH 7-OH 7-OCH2CH=C(Me)2 6-CH=CH-C(Me)20Me 7-OH 8-OCH2CH=C(Me)2 7-OCH2CH=C(Me)2 8-CH2-CH2-Pr' 3-Me, 4-OH 3-Ar, 7-OH 3-C(Me2)-CH=CH2 3-C(Me)2-CH=CH2 5-Me 4,7-Me2 4-Ph, 7-OH 4-Ph, 5-OH 4-Ph 5-Me 4-Me, 6-OH 4-Me 5-CH=CH-Ac 8-OAc 7-OCH2-CH=CMe2 3,5-Me2 3-CH2COOEt, 4-Me 3-CH2COOEt, 4-Me 3-CH2COOEt, 4-Me 4-SMe, 5-Me, 7-OH 5-CHCH=CMe2, 7,8-[OH] 3,5-Me2
993 5.3
Regio- and stereochemistry of pvranocoumarins
A simple diagnostic method has been proposed to distinguish regioisomers of natural benzodioxane lignoids (Fig. 16) with different aryl (Ar) and alkyl (R) substituents (07'8'-l to 07'8'-13), based on small, but systematic differences in ^^c chemical shifts of C-7 and C-8 of the coumarin moiety, provided that both isomers are available (86).
o
y
o
o
Ar^7N<:^
o' R-r^
R Fig. 16.
Ar Benzodioxane lignoids 07'8'-l to 07'8'-13
The two geminal methyl groups attached to C-2' in dihydropyranocoumarins allow stereochemical assignments. In the angular 3',4'-diols (N3'4'-l and N3'4'-26) the difference of the ^^C chemical shifts of these methyl groups is larger than 3 ppm in the cis- and smaller than 2 ppm in the rr^n^-isomer. This seems to be a rule which holds also for 3',4'-dialkoxy substiments (87).
N3'4'.l
N3'4'.26 OH HO. H3C-
H3C
O L3'4'.l
Fig. 17.
^O
H3C^
O' ^ - -
^O'
O
L3'4'-2
Angular and linear 3',4'-disubstituted dihydropyranocoumarins
In the case of linear dihydropyranocoumarins, however, these trends are opposite, i.e. the difference is up to 7 ppm in trans- but negligible in cw-isomers as shown for the pair L3'4'-l and L3'4'-2 (Fig. 17). Apparently, effects of conformational preference and mobility are responsible for these observations.
994 It should be noted that a secure differentiation of the C2'-methyl signals from those of other methyl groups which might exist in the molecule, is a prerequisite for an application of this rule.
Ackno wled gements B. M. thanks the Deutsche Forschungsgemeinschaft for two fellowships at the University of Hannover in 1994 and 1995 in preparation of this article. The authors gratefully acknowledge valuable assistance of Carsten Duddeck in establishing Tables 5-13, This work was supported by the Fonds der Chemischen Industrie.
References to sections 1-5 1 2
3 4 5 6 7 8
9 10 11 12 13 14 15 16 17 18 19 20 21
R. D. H. Murray, J. M^ndez and S. A. Brown, The Natural Coumarins, John Wiley & Sons Ltd., Chichester, 1982, and references cited therein. E. Wenkert, B. L. Buckwalter, I. R. Burfitt, M. J. Gasic, H. E. Gottlieb, E. W. Hagaman, F. M. Schell, P. M. Wovkulich and A. Zheleva, in: G. C. Levy (Ed), Topics in Carbon-13 NMR Spectroscopy, Vol. 2, Wiley, New York, 1976, pp. 81-121. H. Duddeck and M. Kaiser, Org. Magn. Reson., 20 (1982) 55-72. R. Benn and H. Giinther, Angew. Chem., Int, Ed. Engl, 22 (1983) 350-380. R. R. Ernst, G. Bodenhausen and A. Wokaun, Principles of Nuclear Magnetic Resonance in One and Two Dimensions, Clarendon Press, Oxford 1987. W. R. Croasmun and R. M. K. Carlson (Eds), Two-Dimensional NMR Spectroscopy. Applications/or Chemists and Biochemists, 2nd edition, VCH Publishers, New York, 1994. H. Kessler, C. Griesinger, J. Zarbock and H. Loosli, J. Magn. Reson., 57 (1984) 331-336. (a) A. Bax and R. Freeman, J. Amer. Chem. Soc, 104 (1982) 1099-1100; (b) A. Bax, J. Magn. Reson., 57 (1984) 314-318; (c) J. A. Laakso, E. D. Narske, J. B. Gloer, D. T. Wicklow and P. F. Dowd, J. Nat. Prod., 57 (1994) 128-133; (d) L.-J. Lin, L.-Z. Lin, N. Ruangrungsi and G. A. CordeU, Phytochemistry, 34 (1993) 825-830. C. Griesinger, H. Schwalbe, J. Schleucher and M. Sattler, in: ref. 6, pp. 457-580. D. Neuhaus and M. P. WiUiams, The Nuclear Overhauser Effect in Structural and Conformational Analysis, VCH Publishers, New York, 1989. J. Redondo, F. Sanchez-Ferrando, M. Vails and A. Virgili, Magn. Reson. Chem., 26 (1988) 511-517. R. Freeman, /. Chem. Phys., 53 (1970) 457-458; O. A. Gansow and W. Schittenhelm, J. Am. Chem. Soc, 93 (1971) 4294-4295. Ref. 1, page 45. H. Gunther, J. Prestien and P. Joseph-Nathan, Org. Magn. Reson., 1 (1975) 339-344. S. A. Sojka, J. Org. Chem., 40 (1975) 1175-1178. N. J. Cussans and T. N. Huckerby, Tetrahedron Lett., (1975) 2445-2446. N. J. Cussans and T. N. Huckerby, Tetrahedron, 31 (1975) 2587-2590. N. J. Cussans and T. N. Huckerby, Tetrahedron, 31 (1975) 2591-2594. N. J. Cussans and T. N. Huckerby, Tetrahedron, 31 (1975) 2719-2726. L. Ernst, J. Magn. Reson., 21 (1976) 241-246. C.-J. Chang, H. G. Floss and W. Steck, J. Org. Chem., 42 (1977) 1337-1340.
995 22 23 24 25 26 27 28
K. K. Chan, D. D. Giannini, A. H. Cain, J. D. Roberts, W. Porter and W. F. Trager, Tetrahedron. 33 (1977) 899-906. M. H. A. Elgamal, N. H. Elewa, E. A. M. Elkhrisy and H. Duddeck, Phy to chemistry, 18 (1979) 139-143. A. Rabaron, J.-R. Didry, B. S. Kirkiacharian and M. M. Plat, Org. Magn. Reson., 12 (1979) 284-288. A. Patra and A. K. Mitra, Org. Magn. Reson., 17 (1981) 222-224. R. Rodighiero, P. Manzini, G. Pastorini and A. Guiotto, J. Heterocycl Chem., 21 (1984) 235240. A. Guiotto, P. Manzini, A. Chilin, G. Pastorini and R. Rodighiero, J. Heterocycl. Chem., 22 (1985) 649-656. M. A. Kirpichenok, I. I. Granberg, L. K. Denisov and L. M. Melnikova, Izv. Timiryazewsk.
S-kh.Akad., 3 29 30 31 32 33 34
35
36 37 38 39 40 41
42 43
44 45
46
{l9S5)ll2-m.
A. S. Osborne, Magn. Reson. Chem., 27 (1989) 348-354. A. S. Osborne, Monatsh. Chem., 115 (1984) 749-756. A. Z. Abyshev, V. P. Zmeikov and I. P. Sidorova, Khim. Prir. Soedin., (1982) 301-307. A. Patra, S. K. Panda, K. C. Majumdar, A. T. Khan and S. Saha, Magn. Reson. Chem., 29 (1991)631-644. H. E. Gottlieb, R. A. de Lima and F. delle Monache, J. Chem. Soc, Perkin Trans. 2, (1979) 435-437. (a) H. Gunther, H. Schmickler and G. JikeU, J. Magn. Reson., 11 (1973) 344-351; (b) G. JikeH, W. Herrig and H. Gunther, J. Am. Chem. Soc, 96 (1974) 323-324; (c) P. Granger and M. Maugras, Chem. Phys. Lett, 24 (1974) 331-334. (a) A. G. Mclnnes J. A. Walters, J. L. C. Wright and L. C. Vining, in: G. C. Levy (Ed), Topics in Carhon-13 NMR Spectroscopy, Vol. 2, WHey, New York, (1976), pp. 125-178; (b) T. J. Simpson, Chem. Soc. Rev., 4 (1975) 497-522. D. P. H. Hsieh, J. N. Seiber, C. A. Reece, D. L. FitzeU, S. L. Yang, J. L Dalezios, G. N. La Mar, D. L. Budd and E. Motell, Tetrahedron, 31 (1975) 661-663. P. S. Steyn, R. Vleggaar, P. L. Wessels and D. B. Scott, J. Chem. Soc, Chem. Commun., (1975) 193-195. K. G. R. Pachler, P. S. Steyn, R. Vleggaar, P. L. Wessels and D. B. Scott, J. Chem.. Soc, Perkin Trans. 1, (1976) 1182-1189. R. H. Cox and R. J. Cole, J. Org. Chem., 42 (1977) 112-114. H. Duddeck, M. H. A. Elgamal, F. K. Abd Elhady and N. M. M. Shalaby, Org. Magn. Reson., 14(1980)256-257. (a) A. K. Bose, P. R. Srinivasan and G. Trainor, J. Am. Chem. Soc, 96 (1974) 3670-3671; (b) A. K. Bose and P. R. Srinivasan, J. Magn. Reson., 15 (1974) 592-593; (c) A. K. Bose and P. R. Srinivasan, Tetrahedron Lett., (1975) 1571-1574. A. K. Bose, H. Fujiwara, V. S. Kamat, G. K. Trivedi and S. C. Bhattacharyya, Tetrahedron, 35(1979) 13-16. (a) C. C. Hinkley, J. Am. Chem. Soc, 91 (1969) 5160-5162; (b) J. K. M. Sanders and D. H. WiUiams, Nature (London), 240 (1972) 385-390; (c) R. E. Sievers (Ed.), NMR Shift Reagents, Academic Press, New York, 1973; (d) B. C. Mayo, Chem. Soc Rev., 2 (1973) 49-74; (e) A. F. Cockerill, G. L. O. Davies, R. C. Harden and D. M. Rackham, Chem. Rev., 73 (1973) 553588; (f) O. Hofer, in: E. L. Eliel and N. L. Allinger (Eds), Topics in Stereochemistry, Vol. 9, Wiley, New York, 1976, pp. 111-197. O. A. Gansow, P. A. Loeffler, R. E. Davis, M. R. Willcott III and R. E. Lenkinski, J. Am. Chem. Soc, 95 (1973) 3389-3390. (a) J. W. ApSimon, H. Beierbeck and J. K. Saunders, Can. J. Chem., 51 (1973) 3874-3881; (b) K. Tori, Y. Yoshimura, M. Kainosho and K. Ajisaka, Tetrahedron Lett., (1973) 31273130. J. Reuben, J. Magn. Reson., 11 (1973) 103-104.
996 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
(a) V. W. Goodlett, Anal. Chem.. 37 (1965) 431-432; (b) L R. Trehan, C. Monder and A. K. Bose, Tetrahedron Lett, (1968) 67-69. A. K. Bose and P. R. Srinivasan, Tetrahedron, 31 (1975) 3025-3029. W. Kitching, M. Bullpitt, D. Doddrell and W. Adcock, Org. Magn. Reson., 6 (1974) 289-292, and references cited therein. (a) O. A. Subbotin, P. I. Zakharov, V. A. Zagorevskii and D. A. Zykov, Khim. Prir. Soedin., (1975) 476-479; (b) S. Berger, in: P. Diehl, E. Ruck, H. Giinther, R. Kosfeld and J. Selig (Eds), NMR - Basic Principles and Progress, Vol. 20, Springer, Berlin, Heidelberg, New York, 1990, pp. 1-30. K. K. Chan, D. D. Giannini, A. H. Cain, J. D. Roberts, W. Porter and W. F. Trager, Tetrahedron, 33 (1977) 899-906. M. Vanhaelen and R. Vanhaelen-Fastre, Pharnu Acta Helv., 51 (1976) 307-321. M. E. Brokke and B. E. Christensen, J. Org. Chem., 23 (1958) 589-596. M. H. A. Elgamal, N. M. Elewa, E. A. M. Elkhrisy and H. Duddeck, Symposium-Papers, 11th lUPAC International Symposium on the Chemistry of Natural Products, Vol. 2, 1978, pp. 271-274. For annual reviews see: Nuclear Magnetic Resonance, Specialists Periodical Reports, Chemical Society, London. R. Ditchfield and P. S. Ellis, in: G. C. Levy (Ed), Topics in Carbon-13 NMR Spectroscopy, Vol. 1, Wiley, New York, 1974, pp. 1-51. G. E. Maciel, in: G. C. Levy (Ed), Topics in Carbon-13 NMR Spectroscopy, Vol. 1, Wiley, New York, 1974, pp. 53-77. W. Kutzelnigg, U. Heischer and M. Schindler, in: P. Diehl, E. Fluck, H. Giinther, R. Kosfeld and J. Selig (Eds), NMR - Basic Principles and Progress, Vol. 23, Springer, Berlin, Heidelberg, New York, 1990, pp. 167-262. (a) J. B. Stothers, Carbon-13 NMR Spectroscopy, Academic Press, New York, London, 1972; (b) H.-O. Kalinowski, S. Berger and S. Braun, ^^C NMR-Spektroskopie, Thieme, Stuttgart, 1984. (a) N. K. Wilson and J. B. Stothers, in: E. L. Eliel and N. L. Allinger (Eds), Topics in Stereochemistry, Vol. 8, Wiley, New York, 1974, 1-158; (b) H. Duddeck, in: E. L. EUel, S. H. Wilen and N. L. Allinger (Eds), Topics in Stereochemistry, Vol. 16, Wiley, New York, 1986, 219324. J. A. Pople and D. L. Beveridge, Approximate Molecular Orbital Theory, McGraw-Hill, New York, 1970. D. J. Sardella, J. Am. Chem Soc, 95 (1973) 3809-3811. E. Heilbronner and H. Bock, Das HMO-Modell, Verlag Chemie, Weinheim, 1968-1970. J. A. Pople and D. L. Beveridge, Approximate Molecular Orbital Theory, McGraw-Hill, New York, 1970. G. I. Grigor and G. A. Webb, Org. Magn. Reson., 9 (1977) 477-479. L. M. Ryzhenko, V. I. Labunskaya, B. F. Ryzhenko and A. D. Shebaldowa, Khim. Geterosikl. Soedin. 12 (1988), 1611-1614 [Chem. Heterocyclic Comp. (Consultant Bureau), 12 (1988) 1330-1333]. (a) D. A. R. Happer, Aust J. Chem., 29 (1976) 2607-2614; (b) D. A. R. Happer, S. M. McKerrow and A. L. Wilkinson, AW5t J. Chem., 30 (1977) 1715-1725. T. Schaefer, F. Hruska and H. M. Hutton, Can. J. Chem., 45 (1967) 3143-3151. C. G. Swain and E. C. Lupton, J. Am. Chem. Soc, 90 (1968) 4328-4337. P. R. Wells, D. P. Arnold and D. Doddrell, J. Chem. Soc, Perkin Trans. 1, (1974) 1745-1749. (a) L. Ernst, Chem. Ber., 108 (1975) 2030-2039; (b) L. Ernst, Z. Naturforsch., Teil B 30 (1975) 794-799; (c) L. Ernst, J. Magn. Reson., 20 (1975) 544-553. L. Ernst, J. Magn. Reson., 22 (1976) 279-287. B. S. Kirkiacharian, A. Rabaron and M. Plat, CR. Acad. ScL, Ser. C, 284 (1977) 697-700. D. S. Yufit, M. A. Kirpichenok, Y. T. Struchkov, L. A. Karandasova and I. L Gandberg, Bull Acad. ScL USSR, Chem. ScL, (1991) 702-710.
997 75
76 77 78 79 80 81 82 83 84
85 86 87
(a) F. A. Macias, G. M. Massanet, F. Rodriguez-Luis and J. Salva, Magn. Re son. Chem., 27 (1989) 105-701 \ (b) F. A. Macias, R. Hemandez-Galan, G. M. Massanet, F. Rodriguez-Luis, M. Vasquez and J. Salva, Magn. Reson. Chem., 28 (1990) 732-735. H. Duddeck, F. HoUowood, A. Karim and M. A. McKervey, J. Chem. Soc, Perkin Trans. 2, (1979) 360-365. F. A. Macias, G. M. Massanet, F. Rodriguez-Luis and J. Salva, Magn. Re son. Chem., 28 (1990)219-222. P. Joseph-Nathan, J. Hidalgo and D. Abramo-Bruno, Phytochemistry, 17 (1978) 583-584. R. D. Lapper, Tetrahedron Lett., (1974) 4293-4296. A. D. Vdovin, E. K. Batirov, A. D. Matkarimov, M. R. Yagudaew and V. M. Malikov, Khim. Prir. Soedin., {19^1)196-199. A. J. Jones, T. D. Alger, D. M. Grant and W. M. Litchman, J. Am. Chem.. Soc, 92 (1970) 2386-2394. D. Doddrell and P. R. Wells, J. Chem. Soc, Perkin Trans. 2, (1973) 1333-1336. A. Greenberg and J. F. Liebman, Strained Organic Molecules, Academic Press, New York, 1978, pp. 150-152. (a) D. M. Grant and E. G. Paul, J. Am. Chem. Soc, 86 (1964) 2984-2990; (b) W. R. Woolfenden and D. M. Grant, J. Am. Chem. Soc, 88 (1966) 1496-1502; (c) D. M. Grant and B. V. Cheney, J. Am. Chem. Soc, 89 (1967) 5315-5318. (a) S. H. Grover, J. P. Guthrie, J. B. Stothers and C. T. Tan, / Magn. Reson., 10 (1973) 227230; (b) S. H. Grover and J. B. Stothers, Can. J. Chem., 52 (1975) 870-878. A. Amoldi, A. Amone and L. Merlini, Heterocycles, 22 (1984) 1537-1544. F. A. Macias, G. M. Massanet, F. Rodriguez-Luis, J. Salva and F. R. Fronczek, Magn. Reson. C/zm., 27 (1989) 653-658.
998 6.
TABLES OF ^^C CHEMICAL SHIFTS OF COUMARIN AND ITS DERIVATIVES Tables 5-13 contain ^^C chemical shifts of the coumarin moiety and some other atoms in the
molecular backbones of 876 coumarin derivatives. They are arranged according to the substitution patterns and basic molecular systems compiled in Fig. 18 below. Therefore, compound chiffres have been composed in such a way that the basic system (initial letter) and the substitution pattern (numbers after the letter) can be read directly; the number after the hyphen indicates the hierarchical order which is given by the nature of that atom in a substituent which is direcdy connected to the coumarin moiety; sequencing according to the Cahn-Ingold-Prelog rule. For example, D348-1 is a trisubstituted coumarin (D) carrying substituents at C-3, C-4 and C-8 (348; 3-Br, 4-OH, 8-Me). Since bromine has the top position of all substituents at C-3 in all 3,4,8-trisubstituted coumarins listed, the compounds is the first entry (-1). The only exception from this rule appears in di- and tricoumarins where the numbers after the letter indicate the positions at which the two (three) coumarin systems are connected. Coumarin, deuterated species and chalcogen analogues Monosubstituted coumarins Disubstituted coumarins Trisubstituted coumarins Tetra- and pentasubstituted coumarins
F:
Linear furanocoumarins
G:
2,3 '-Dihydro derivatives of F
H:
Angular furanocoumarins
I:
2,3'-Dihydro derivatives of H
J:
Coumarins with other annulated five-membered rings
Fig. 18.
2
^
Molecular systems in Tables 5-13
999
O 7^9-0
2^0
K:
Linear pyranocoumarins
L:
3',4'-Dihydro derivatives of K
M:
Angular pyranocoumarins
N:
3 ',4 '-Dihydro derivatives of M
O:
Coumarins with other annulated six-membered rings
P:
Dicoumarins directly connected by one bond
'
3
0.^/0
O P38'-i
o
P66'-i
o P88V
P68'
Continuation of Fig. 18.
O
1000
Q:
Dicoumarins connected by more than one bond
O
O
Q33'.i
Continuation of Fig. 18.
1001 OMe
MeO O
MeO AcO OAc
Q88'.i
Q88-i
Q3''3 '-i
R:
Dicoumarins connected by rings
AcO
HO' y
^o' ^0
OH
O
O
^ ^
O
^^Y AcO OAc
OMe MeO
0
0
Continuation of Fig. 18.
1002
Tricoumarins
Continuation of Fig. 18.
The following abbreviations have been used for substituents: Me = methyl; Et = ethyl; Pr = propyl; Bu = butyl; Pent = pentyl; Hex = hexyl; Oct = octyl; Ph = phenyl; Ac = acetyl; apio/s apiofuranoside; galp = galacto-pyranoside; glcp = gluco-pyranoside; rhamps rhamno-pyranosyl, xylp = xylo-pyranosyl. Abbreviations for solvents are: A = acetone; B = benzene; C s chloroform; CCl = carbon tetrachloride; D = dimethyl sulphoxide; DX = dioxane; M = methanol; P = pyridine; TFA = trifluoro acetic acid; ? = solvent unknown. Most solvents have been used in deuterated form. All ^^c chemical shifts are given in ppm relative to teti-amethylsilane (8 = 0). Open entries mean that chemical shifts have not been reported in the original papers. Values which might be interchanged are marked by "*", "•"' or "^", but only if their chemical shift difference in larger than 1 ppm. If spectra exist in the literature which were recorded in different solvents, the data set obtained in deuteriochloroform has been chosen for the tables.
1003
Many papers reporting chiral coumarin derivatives do not give clear information about the chirality and the enantiomeric purity. Therefore, it is stressed that in all cases the stereochemical notation of the structures in Tables 6-13 indicate only relative, not absolute configurations.
CAVEAT: It is strongly advisable to take data of highly substituted compounds with great caution. The data sets may contain inconsistencies in the signal positions and sequences indicating misassignments. This occurs particularly in tri- and tetrasubstituted coumarins where assigning merely based on analogy with similar structures and/or derivatives with fewer substituents is very dangerous. Unfortimately, however, it was not possible to correct all these mistakes because of missing experimental evidences. In such cases the authors decided to report the original data sequences. In a few instances where two divergent data sets have been published in different papers both data sets are listed.
0 0 P
Table 5.
'F chemical shifts of coumarin, deutcriated species and chalcogen analogues 5
A-1 A-2 A-3 A-4 A-5 A-6 A-7 A-8
*
Substituents
C-2
C-3
C-4
C-5
C-6
C-7
C-8
C-9
C-10
Soh.
Reference
- (cournarin)
160.4 158.6 158.5 185.4 198.0 209.0 188.4 193.5
116.4
143.6 142.2 142.5 143.7 134.5 131.4 144.2 145.4
128.1 127.6 127.7 130.0 127.8 130.3 133.0 135.0
124.4 123.8 123.9 124.2 125.5 123.4 126.0 127.0
131.8 131.2 131.2 131.6 132.2 134.4 129.4 129.7
116.4 116.4 116.4 126.5 116.8 127.7 128.2 135.7
153.9 154.0 153.9 137.7 156.7 140.3 137.0 125.0
118.8 118.7 118.9 126.2 120.5 128.0 126.9 130.8
C CCI CCI C C C
1-19 I1
3-ZH 4-*H I-thio 2-thiono 1-thio-2-thiono I-seleno 1 -telluro
166.3 126.0 129.7 136.0 122.8 125.0
? ?
11
12,19 10,12,13 10,13 19 19
Tahlc 6.
I T chemical shifts of rnonosuhstitutcd cournarins
Substituents
B3- 1 03-2 83-3 B3-4 B3-5 B3-6 83-7 B3-8
3-Br 3-C1 3-OH 3-OCH2-CH=CH2 3-OCH2-C(Me)=CH2 3-0CH2-CH=CMe2 3-0
0
3-OCH2-CH=CH-Ph(trans)
X =H X =OMe
B3-9 83-10 B3-11 B3-12 B3-13 B3-14 B3-15 B3-16
3-NH2 3-NMez 3-COOH 3-Me 3-Et 3-PI"
C-2
C-3
C-4
C-5
C-6
C-7
C-8
C-9
C-10
SO~V. Reference
156.3 156.5 158.5 156.6 157.2 157.4
111.0 120.8 141.8 142.4 143.2 143.5
144.0 140.7 115.0 115.1 115.8 116.0
126.9 127.9 126.3 126.0 126.3 126.2
124.5 124.9 124.5 123.9 124.4 124.4
131.6 131.8 127.5 127.7 128.2 128.1
116.0 116.1 115.6 113.7 113.7 113.5
152.6 152.2 149.2 148.7 149.3 149.4
118.8 118.8 120.7 118.9 119.4 119.6
C/D D C C C
4 20 21 22 22 22
157.5
142.1
115.6
126.1
124.2
127.9
115.3
149.1
119.4
C
22
157.3
143.3
116.1
126.4
124.5
128.1
114.4
149.5
119.5
C
22
159.8 159.9
142.9 143.1
120.8 120.9
128.6 128.8
126.2 126.4
130.5 130.7
117.0 117.2
151.3 151.5
120.9 121.0
M M
36 36
158.8 158.0 157.2 162.1 165.6 161.0
133.3
125.4 116.5 148.5 139.2 137.3 138.3
108.0(?) 124.5 125.4 123.7 130.3 124.9 127.0 124.3 127.0 124.1 127.0 124.0
124.8 126.6 134.4 130.4 130.3 130.3
11.5.5 114.9 116.2 116.3 116.3 116.3
148.0 149.2 154.6 153.2 152.4 153.0
121.9 120.3 118.1 119.6 119.5 119.5
1)
23 24 6,25 5,26
137.3 118.2 125.7 130.9 129.8
C
C D C C
C
B3-17 B4-1 B4-2 B4-3 B4-4 B4-5 B4-6 B4-7 B4-8
3-Pr' 4-OH 4-0- (oxido) 4-OMe 4-OEt 4-OCHz-CH=CHz 4-OCH2-C(Me)=CH2 4-OCH2-CH=CMe2 4-OPent"
B4-9 B4-10 B4-11 B4-12 B4-13 B4-14 B5-1 B5-2 B6-1 B6-2 B6-3 B6-4 B6-5 B6-6 B6-7 B6-8 B6-9 B6-10 B6-11 B7-1 B7-2
4-OHex" 4-OCHz-CH=CH-Ph (trans) 4-Me 4-PS' 4-Ph 5-OMe 5-Me 6-Br 6-C1 6-OH 6-OMe 6-OAc 6-NH2 6-NO2 6-Me 6-C=N 6-CHO 6-COOH 7-Br 7-c1
116.5 1 16.7 116.5 116.5 1 16.6 116.7
152.4 153.7 161.6 153.1 153.3 153.2 153.2 151.5 153.3
119.5 116.1 104.7 115.5 115.8 115.5 1 15.6 115.9 115.8
C C D C C C C C C
37 3,7,21,27-29 30 7,21,31.35 32 33,34 35 33 32
132.0
116.3
153.2
115.9
C
35
132.2 132.1 131.7 131.4 131.7 132.6 131.6 134.2 131.4 1 19.7 119.0* 126.0 120.5 127.0 132.8 135.4 132.7 133.6 125.5 138.1
116.7 116.5 116.9 117.1 117.1 109.2 14.7 18.3 18.0 16.9 17.7* 18.1 17.5 18.4 16.5 18.7 18.3 17.3 19.8 17.3
152.7 153.2 153.5 153.6 154.0 155.4 154.6 152.7 152.1 146.8 148.4 151.7 147.2 157.9 152.2 156.9 158.1 157.0 154.1 154.5
115.8 115.5 120.0 119.2 118.8 109.8 117.7 120.2 119.6 119.1 119.1 119.8 119.7 119.5 118.6 120.9 119.7 119.0 117.6 117.9
C C C C C C C C C D CICCI CIM CIM CIM C CIM CIM CM C CIM
32 35 5,26 37 37 38 5.39 4 3.9 9,2 1 2,9 9 9 9 5,9,14,26 9 9 9 4 9
165.4 162.9 166.6 162.5 162.5 162.5 162.5 160.5 162.5
135.8 91.3 87.1 90.0 90.3 90.8 90.8 90.8 90.3
138.3 160.0 177.2 166.2 165.8 164.9 165.0 164.4 165.6
127.2 123.3
123.8 123.5
130.3 132.1
116.3 116.1
122.8 122.8 122.8 122.8 123.0 122.8
123.7 123.7 123.7 123.7 123.6 123.7
132.2 132.2 130.5 132.2 132.1 132.2
162.8
90.4
164.3
123.0
123.5
162.5 162.5 160.5 160.8 160.5 161.1 160.6 159.4 159.6 160.1 160.1 161.7 162.8 160.0 160.9 160.2 160.7 161.2 159.4 161.1
90.3 90.7 115.1 113.8 115.0 114.6 15.9 17.5 17.6 16.1 17.1* 17.4 16.4 18.8 16.5 18.7 17.9 17.3 16.6 16.5
165.6 165.0 152.3 155.8 155.5 138.8 140.4 141.8 141.9 143.8 142.6 144.0 144.6 143.4 143.4 143.2 144.0 144.3 142.4 143.8
123.3 123.0 124.6 124.2 126.8 156.5 136.3 129.9 126.8 1 12.4 109.8 120.8 112.2 124.4 127.8 133.3 131.0 130.7 128.6 129.3
123.7 123.6 124.2 124.0 124.0 105.4 125.7 116.7 129.4 153.7 155.9 147.2 144.4 144.6 134.1 108.9 133.4 127.8 127.6 125.5
0
0
o\
160.7 160.8 161.7 162.7 160.4 158.8 160.1 160.0
1 1 1.5 112.7 116.0 1 12.7 112.4 113.7 11 1.9 113.2
144.3 143.3 144.2 144.8 142.9 143.3 142.7 143.3
129.6 128.7 129.3 129.4 128.4 128.2 128.1 128.0
1 13.3 112.1 119.1 113.6 1 12.5 117.1 112.1 117.2
161.6 162.6 153.8 162.8 161.2 16 1 .o 161.2 161.4
102.5 100.6 1 10.7 101.6 101.2 107.0 100.6 106.8
155.7 155.6 154.9 156.1 155.2 155.0 154.9 155.2
117.2 112.9 112.1 113.5 111.6 113.5
160.8
112.7
143.2
128.9
113.1
161.7
101.7
155.7
160.3 160.2
113.2 116.7
144.2 142.8
129.5 128.6
113.8 118.3
160.3 154.7
103.4 110.3
161.8
113.0
143.4
122.9
113.1
161.2
160.7
112.5
143.0
128.3
112.7
162.1
113.1
143.5
127.4
B7-17
172.6
117.5
135.2
B7-18
162.5
113.3
143.7
B7-3 B7-4 B7-5 B7-6 B7-7 B7-8 B7-9 B7-10
7-OH 7-OMe 7-OAc 7-OEt 7-OCH2-CHXHz 7-OC(Me2)-C=CH 7-OCH2-CH=CMe2 7-OC(Mez)-CH=CH2
1 1 1.5 1 12.3
D
c
3,6,9,18,21 2,8.9,1 I , 14 9 9 16 40 6 41
112.9
C
42
155.1 153.3
113.4 116.1
D C
21 17
101.5
155.8
112.6
C
43
161.7
101.3
155.4
112.1
C
16,44,45
113.2
161.3
101.7
155.9
112.7
C
46
120.0
104.4
151.5
135.9
151.7
118.2
C
47
128.9
113.5
161.5
101.8
156.0
112.7
C
48
C/CI C/M C/M
c
C C
7-00
B7-11 87-12 B7-13
7-O-p-D-gICp 7-OCO-CHzPh
B7-14 B7-15 B7-16
7-0
a
&
7-0
+ 0 0 -4
7-0
B7-19
011
160.1
112.2
144.0
129.2
112.7
161.5
101.3
1552
112.2
1)
16
161.8
113.0
143.2
128.6
112.9
160.9
101.2
155.8
112.5
C
49
161.2
112.8
143.3
128.5
113.1
162.0
101 4
155.8
112.3
C
4030
161.3
112.9
143.4
128.6
113.1
162.0
101.4
155.8
112.4
C
SO
161.2
112.9
143.4
128.6
113.2
162.0
101.5
155.8
112.4
C
SO
161.3
112.9
143.4
128.7
113.2
162.1
101.6
155.9
112.4
C
161.3
113.1
143.5
128.6
112.8
162.8
101.1
155.9
112.2
C
7-(\
B7-20
O
W
B7-21 OH
B7-22
/
/
7 - 0 A
B7-23
Of 1 I
B7-24
I
/
0-7
OH
I
/
SO
B7-26
HO
7-0,
HO ''.'
B7-27
161.0
112.6
143.3
128.6
112.6
162.0
101.2
155.6
112.3
C
1
161.8
112.5
143.2
128.5
112.5
160.8
101.0
155.4
112.1
C
49
162.0
112.7
143.5
128.7
113.2
161.3
101.4
155.6
112.3
C
51
160.7
112.6
143.0
128.4
112.7
161.7
101.0
155.5
112.1
C
1
162.2
113.1
143.4
128.7
113.1
161.2
101.4
155.9
112.5
C
49
161.2
112.9
143.3
128.7
113.0
161.9
101.5
155.7
112.4
C
so
_ -
'- 01I
OH
B7-28
B7-29
AcO J &
B7-30 4)Ac
B7-31
-
0
%
-
B7-32
.
€10
162.2
113.1
143.4
128.7
113.0
161.2
101.4
155.9
112.4
C
49
162.0
112.9
143.3
128.7
112.9
161.1
101.3
155.8
112.5
C
49
B7-34
162.0
113.1
143.3
128.8
113.0
161.1
101.4
155.9
112.6
C
49
B7-35
161.9
112.8
143.3
128.5
113.0
161.1
101.4
155.7
112.3
C
51
161.2
113.0
143.4
128.8
113.2
161.8
101.7
155.9
112.7
C
54
"OAc
7-o\.
-
.
HO""
B7-33
'- OAC
B7-37
161.0
113.5
143.3
128.9
113.0
161.4
101.4
155.9
112.9
C
54
B7-38
161.0
113.5
143.3
129.0
112.9
161.2
101.4
155.9
113.0
C
54
B7-39
161.3
112.9
143.5
128.8
113.2
162.3
101.8
155.9
112.4
C
54
B7-40
161.4
112.9
143.4
128.7
113.2
162.4
101.4
156.2
112.4
C
54
163.6 162.0 160.1 160.9 161.6
109.6 108.8 120.1 115.4 118.0
145.2 143.6 142.7 143.4 143.8
129.5 128.7 129.5 127.6 128.6
112.6 108.5 119.5 125.6 126.0
152.5 150.5 148.2 143.1 134.6
100.5 97.2 112.7 116.9 118.5
156.6 156.5 154.1 154.2 154.0
110.3 108.1 124.1 116.5 122.6
CIM C C/M C CIM
9 67 9 5,9,26,39 9
B7-41 87-42 B7-43 B7-44 B7-45
7-NHz 7-NEt2 7-NOz 7-MC 7-COOH
160.5
113.6
143.9
129.6
112.7
144.3
106.9
151.6
113.6
C
53
161.3
113.0
143.4
128.7
113.3
162.0
101.5
155.8
112.5
C
54
161.2
113.0
143.5
128.8
113.2
162.0
101.6
155.9
112.5
C
54
161.0
113.6
143.3
129.0
112.9
161.2
101.4
155.9
113.0
C
54
144.7 147.2 126.3
142.4 144.8 152.4
119.7 llY.3 118.6
D
21 2,38 5.26
,0-7 _ .. .._.
B7-49 BS-1
B8-2 B8-3
0
8-OH 8-OMe 8-Me
160.0 159.4 160.9
116.1
116.9 116.3
144.5 143.0 143.8
118.4 119.0
125.6
124.4 123.9 124.0
118.4 113.6 133.2
C/CI C
Table 7.
I
T
chemical shifts of disuhstitutcd coumarins
Su bstituents C34-1 C34-2 C34-3 C34-4 C34-5 C34-6 C34.7 C34-8 c34-9 C34-10 C34-11 C34-12 C34-13 C34-14 C34-15 C34-16
C34-18
3-Br, 4-OH 3-CI,4-OH 3-OPh, 4-OH 3-OH, ~-CH~-C(CI)=CHI 3-OH. 4-CH*-CH=CH2 %OH, 4- CH~-C(MC)=CH~ 3-OH, 4- C(Me2)-CH=CH2 3-OH. 4
0
3-OH, 4-CH(Ph)-CH=CH2 3-N02,4-0H 3-Me, 4-OH ~ - A c4-OH , 3-CHz-CH=CHZ, 4-OH 3-COOEt. 4-OH 3-CH2-C(Me)=CH2,4-OH 3-Ph, 4-OH
3-CHzPh, 4-OH
C-2
C-3
C-4
C-5
C-6
C-7
C-8
C-9
C-10
Solv.
Reference
158.5 158.6 158.4 159.8 158.5 159.9 161.3
89.4 98.9 119.9 133.5 138.4 137.7 137.9
162.3 160.5 155.4 121.8 125.0 122.9 129.1
123.4 123.4 123.5 125.1 124.7 124.7 127.3
124.2 124.4 124.3 124.0 124.1 124.2 123.4
132.7 132.5 131.7 128.6 127.9 128.1 128.4
116.3 116.3 116.2 116.8 116.2 116.5 116.8
151.7 151.2 150.9 148.8 148.8 148.7 148.3
116.3 116.0 116.5 120.3 120.7 120.9 120.3
D D D C D C C
7,27 27 27 55 22 22 22
160.0
137.1
129.6
125.0
124.3
127.9
116.8
148.8
120.3
C
22
158.3 156.1 163.2 159.9 161.5 156.5 161.6 161.9
138.7 121.2 100.5 101.4 102.5 94.3 101.8 106.2
125.9 163.9 159.7 177.8 164.6 172.0 163.9 160.3
124.8 124.3 123.0 124.7 123.1 124.4 122.9 123.8
124.0 125.6 123.7 125.1 123.9 124.5 123.8 123.8
127.4 134.2 131.3 136.6 131.7 135.5 131.8 132.1
116.2 116.6 116.0 116.8 116.6 116.5 116.3 116.1
148.6 152.4 151.8 154.1 152.2 153.7 152.3 152.3
119.5 117.8 116.4 114.7 115.9 114.3 115.6 116.5
D D D D C D C D
22 27 27,28 27 35 27 35 27,28
160.8
107.2
163.0
122.7
123.6
131.5
116.1
152.3
115.9
C
35
162.9
104.4
160.6
123.3
123.7
131.6
116.1
152.1
116.3
1)
27,28
I
0
'Y
-
~-CH(P~)-CH=CH~,~-OAC 3-CH(Ph)-Et, 4-OH 3-CH(Ph)-CH2Ac,4-0- Na' 3-CH(Ph)-CH2Ac,4-OMe
C34-19 C34-20 C34-21 C34-22
C34-23 C34-24 C34-25 C34-26 C34-27
3
C34-28
3
C34-29
R=H,X=F R = H, X = OMe R = Me, X = H R=X=Ph
*
,4-0-
,4OH
4
I
C34-30 C34-31
3
C34-32
3u
/
/
/
/
/
''-0H
/
CHo
9
4"H
152.2 152.2 153.5 153.0
116.0 116.1 122.3 116.8
C D IIX C
35 7,56 7,56
177.0 177.1 177.1 176.9 176.8
154.1 154.1 154.3 154.0 154.1
118.3 118.2 118.6 118.6 118.4
L)
D 1) D D
30 30 30 30 30
109.2
177.1
153.2
118.3
D
30
160.7
104.0
163.6
123.2
124.0
131.9
116.6
152.6
115.0
C
57
159.8
105.0
164.2
123.0
123.8
131.4
116.5
152.2
115.9
C
57
160.9
103.4
163.8
122.7
123.0
131.4
116.2
152.2
115.9
C
58
160.7
103.0
163.8
122.6* 123.8* 131.6
116.4
152.4
115.8
C
59
155.4 161.6 167.9 162.3
121.3 107.9 103.4 119.8
161.1 160.7 175.7 164.2
159.8 159.9 160.2 158.3 156.8
106.8 107.2 107.3 106.9 107.8
159.4
122.8 123.4 126.6 126.7
124.1 123.7 123.9 123.5
131.8 131.7 131.6 131.4
116.6 116.3 116.7 116.6
7
0,
P
OAc
'
/
/
,4OH
162.0
119.3
154.5
122.3
131.6
116.8
152.1
116.2
C
60
c34-34
160.7
103.8
163.8
122.7* 123.7* 131.4
116.3
152.3
116.0
C
59
c34-35
160.6
103.5
163.7
122.6* 123.7* 131.4
116.3
152.3
116.0
C
59
C34-36
161.9
119.4
154.3
122.7* 124.3* 131.3
116.5
151.9
116.2
C
59
160.8
103.2
163.6
122.7* 123.6* 131.5
116.3
152.4
116.0
C
59
161.0
102.9
163.1
122.7'
123.8* 131.6
116.4
152.0
116.0
C
59
161.8
119.4
154.4
122.2* 124.1* 131.3
116.4
151.9
116.1
C
59
155.7 157.2 157.9 159.1 160.5 161.0 159.2
112.0 138.7 138.3 136.1 124.7 116.1 131.2 132.7 132.0
143.5 113.5 114.4 119.0 139.8 138.1 138.6 137.6 137.7
129.6 121.1 127.7 126.7 128.7 128.8 129.0 128.0 127.9
118.2 116.6 117.2 100.3 100.3 100.8 102.6 106.6 106.5
151.6 152.9 148.4 151.4 155.2 154.0 154.7 154.3 154.0
116.3 121.6 116.8 114.1 113.2
D
4 24 24 24 61 62 41 41 41
c34-33
3-
124.3
,4-OH c34-37
/
/
OH
/
3
A
.4-OH C34-38 d3
/
I
c34-39 - 3 C36-1 C36-2 C37-1 C37-2 c37-3 c37-4 c37-5 C37-6 c37-7
/
1
/
OAc
/
/
,
.4-OAc OH
3 6 B r2 3-NMe2,6-N02 3-NMc2,7-Br 3-NMe2, 7-OMc 3-Ph. 7-OMe 3-CH=CH-C(Me2)0H (trans), 7-OMe 3-C(Me2)-CH=CH2.7-OH 158.1 3-C(Me2)C S H , 7-O-C(Me2)CH=CH2 3-C(Me2)-CH=CH2,7-O-C(Me2)CH=CH* 159.8
120.9 144.4 129.4 112.3 112.6 113.1 113.4 117.1 117.3
134.2 121.6 122.7 159.7 162.5 162.2 161.1 158.1 158.7
112.8 114.3 113.8
C C C
C
C C
C C
0 VI
1
C37-8 3-C(McJ)-CH=CHz,7-OMe
159.9
131.7
137.6
128.5
112.2
162.0
100.0
54.9
112.9
C
45,41
-
2 m
c37-9 C37-10 C37-11
X=NH X = NMe
x=s
wMe
C38-1 3-NMe2,8-OMe
\
C38-2 C45-1 C45-2 C45-3 C46-1 C46-2 C46-3 C46-4 C46-5 C46-6 C46-7 C46-8 C47-1 C47-2 C47-3 C47-4 C47-5 C47-6 C47-7
s
161.9 159.8 160.7
108.2 111.4 113.1
142.5 146.0 141.8
130.5 130.0 130.7
110.0 109.4 109.9
151.9 151.8 152.0
97.0 97.4 97.3
56.8 157.7 157.1
108.9 C/CCI 108.8 C/CCI 108.9 C/CCI
64 64 64
157.6
137.6
116.7
117.2
123.7
109.1
146.1
138.8
121.1
C
24
208.7
141.0
130.5
129.0
126.9
135.0
31.3
140.8
129.3
C
12
161.8 162.6 161.3 162.1 163.0 162.5 162.2 162.5 163.3 159.9 159.9 162.4 162.6 163.7 164.4 163.2 158.6 161.1
91.4 90.0 93.0 91.1 90.0 90.3 90.3 90.3 115.0 116.0 1 14.4 88.7 87.3 90.3 80.1 92.3 108.4 111.6
168.8 169.4 166.6 165.6 166.4 165.6 165.6 165.6 154.3 151.1 153.7 166.1 166.2 160.0 152.7 161.7 152.9 152.6
137.2 137.1 137.1 122.8 122.7 122.8 122.8 122.8 127.4 124.1 109.4 124.4 123.6 126.0 121.3 125.5 124.3 125.4
127.1 127.5 127.8 133.0 133.6 123.7 123.7 123.7 121.5 129.6 152.3 1 1 1.7 I 11.6 107.5 107.8 107.8 111.1 112.0
131.6 131.5 132.0 133.4 133.4 133.8 133.8 133.9 135.2 131.5 119.7 163.1 162.6 149.8 150.4 150.0 159.3 162.4
14.8 15.1 14.8 16.1 16.4 11 6.6 116.6 1 16.6 118.9 118.4 117.1 100.6 100.1 98.0 98.2 97.7 100.4 100.7
155.1 154.7 154.4 151.7 151.5 152.9 152.9 1.52.9 151.6 151.9 146.3 155.6 154.5 156.6 155.9 156.4 151.3 155.1
114.4 D 114.4 C 113.8 D 115.5 '? 114.3 C 115.8 C 115.8 C 115.8 C 118.0 C/TFA 121.1 C 120.1 D 109.2 D 108.5 C 104.9 C 102.5 C 104.0 C 110.1 113.3 C
s
Me
4-OH, 5-Me 4-OMe, 5-Me 4-O-P-I)-glcp,5-Me 4-OH, 6-Me 4-OMe, 6-Me 4-OEt, 6-Me 4-OPent", 6-Me 4-00ct", 6-Me 4-Me, 6-Br 4-Me, 6-C1 4-Me, 6-OH 4-OH. 7-OMe 4,7-[OMeIz 4,7-[NEtzJz 4-NHexCY", 7-NEt2 4-morpholino, 7-NEtz 4-Me, 7-OH 4-Me, 7-OMe
65 65 65 65 65 32 32 32 66 66 18
27 14 67 67 67 3.2 1 38,66,68
160.4 160.0 160.3 160.4 161.6 161.4 160.5 159.6 159.7 160.3 160.9 160.1 162.2 162.9 161.1
114.5
108.1 108.0 114.2 111.1 108.6 108.2 1 10.4 111.7 110.5 91.0 89.8 116.0
151.9 153.3 153.4 152.2 152.5 152.7 142.3 141.5 141.5 155.8 155.8 156.0 166.0 166.8 140.9
125.4 126.3 126.4 124.9 125.0 124.5 126.7 126.3 126.1 127.9 127.8 127.7 120.9 120.6 134.2
117.8 113.9 113.6 1 10.6 108.0 125.6 1 14.4 109.3 109.2 113.1 112.1 117.9 123.1 123.4 132.8
154.2 160.0 160.3 151.4 1SO. 1 143.2 162.3 150.9 151.1 161.6 162.6 153.0 133.4 133.6 133.7
110.4 103.8 103.5 99.1 97.1 117.4 103.6 98.4 98.0 102.9 101.0 114.3 125.1 126.2 114.3
153.1 154.4 154.5 155.1 155.5 153.9 156.5 157.3 157.2 155.4 155.6 154.6 152.0 151.7 153.4
114.3 110.1 CICCI 108.4 C 117.8 C 106.3 CIM 103.0 CICCI 102.6 CICCI 110.8 CID 112.4 C 116.6 C 115.5 C 115.4 C C
162.5
110.4
139.1
154.5
97.7
162.0
98.4
156.7
103.3
C
51
162.7 161.9
110.1 113.0
139.7 137.4
154.9 147.7
93.7 106.7
162.5 160.6
98.9 99.5
156.6 156.0
103.6 106.3
C C
51 51
161.81 111.0 C57-4 5,7-[OMe]? 161.3 110.4 C57-5 ~-CH~-CH=C(MC)-CH~-CH~-CH=CMC?, 7-OMe 160.5 11 1.8 C57-6 S-Me, 7-OMe 159.4 114.7 C58-1 5,8-[OMe]? 160.2 114.7 C58-2 5-OMe, 8-OCHz-CH=CMez 160.2 114.7 C58-3 5-CH2-CH=CMe2,8-OMe 159.7 114.5 C58-4 5,8-MC? 161.4 112.0 C67-1 6,7-[OH]z 160.6 112.5 C67-2 6-OH, 7-OMe
139.0 138.7
157.3 156.0
94.9 95.5
164.1 163.5
93.1 92.5
157.1 156.5
104.2 104.0
C
C
2,1838 6
139.8 138.2 138.7 139.0 139.8 144.5 144.1
136.7 149.5 149.8 148.9 132.8 112.9 1 1 1.8
113.4 103.7 104.1 105.4 124.4 143.2 143.3
161.7 114.9 117.5 114.7 132.0 150.6 151.6
98.4 141.3 140.2 141.1 122.8 103.2 99.9
155.9 144.7 145.3 144.6 152.0 149.1 148.2
I 1 1.1 C 110.2 CCVC 110.4 C 110.7 C 116.5 C 11 1.4 D 11 1.5 1)
C47-8 4-Me, 7-OAc C47-9 4-Me, 7-O-a-r~-glcp C47-10 4-Me, 7-O-p-11-glcp C47-11 4-Me, 7-NHz C47-12 ~ - M c7-NEt2 , C47-13 4,7-Me2 C47-14 4-CF3.7-OH C47-15 4-CF3, 7-NH2 C47-16 4-CF3.7-NEt2 C47-17 4-Ph, 7-OH C47-18 4-Ph, 7-OMe C47-19 4-Ph, 7-OAc C48-1 4-OH, 8-Me C48-2 4-OMe, 8-Me C56-1 5.6-Me?
118.1 1 1 1.7 1 1 1.9
C
D D
68 21 21 64 3.69 39 70 64 64 61,71 61 61 65 65 39
5-ox,7-0 C57-1 C57-2 c57-3
x = H , R' = O H , R? = H
X = H, K' = OH, R2 = OH X = Ac, R' = OAc, R2 = OAc
R,
14 2 72 72 26 6,18,2 1 73.74
-
0 4
160.6 160.2 160.7 161.3 162.1
112.7 112.5 113.5 113.0 114.2
144.0 142.3 142.8 143.3 144.1
113.4 107.0 108.0 107.8 108.1
143.4 143.2 146.2 146.4 146.9
148.6 149.2 152.8 151.9 152.2
103.2 102.5 99.9 100.8 101.7
147.7 149.8 150.0 149.8 150.2
113.0 110.5 11 1.2 111.1 112.4
D C C C C
75 73.74.76 2.18 77 78,79
161.0
112.9
143.1
107.8
116.2
151.6
100.9
149.3
111.1
D
73
160.6 160.8 160.8 161.8 160.8
112.3 114.6 114.2 114.8 114.4
144.2 143.5 143.4 143.3 143.5
109.7 111.0 110.1 110.4 110.3
146.0 147.5 147.9 147.6 147.1
149.8 151.3 151.3 150.4 151.3
103.1 105.9 104.6 106.1 104.7
148.9 149.8 150.1 149.4 150.1
113.3 113.9 113.1 114.1 113.4
D P P P P
76.80 80 80 80 83
~ 6 7 - 1 4 6-0Me. 7-0
160.5
112.5
144.3
109.0
146.1
151.6
101.1
149.3
111.1
D
81
C67-15 6-OMe, 7-o-p-D-glCp-(6tl)-p-D-apiof C67-16 6-OMe, 7-0-p-D-glcp(6tI)-P-D-apiof pentaacetate C67-17 6-OMe, 7-0-P-1,-glcp-(6tI)-p-D-apiof hexaacetate C67-18 6-0-p- D-gICp, 7-OH C67-19 6-0-p- D-gICp, 7-OMe C67-20 6-O-p-D-glc~-(6tl)-~-D-apiof, 7-OMe
160.6 160.7
112.4 115.1
144.2 142.8
109.8 109.8
146.0 147.3
149.8 149.1
103.2 107.1
149.0 149.5
113.4 114.0
D C
76.82 76
160.6
115.0
142.7
109.7
147.3
148.9
105.8
149.2
114.1
C
76
160.5 160.5 164.1
112.1 112.8 112.0
144.4 144.4 146.4
115.0 113.1 116.4
142.6 143.3 153.0
151.4 152.8 145.1
103.1 100.3 104.7
150.5 150.0 153.0
110.8 111.3 112.0
D D M
1,18,21,83 74 83
160.1
113.1
143.1
108.0
146.5
149.6
101.0
151.7
111.1
C
84
161.9
1!3.3
143.9
128.2
128.9
161.3
99.0
155.0
112.4
C
85
C67-3 C67-4 C67-5 C67-6 C67-7
6-OH, 7-O-p-r,-glcp-(6tI)-P-D-apiof 6-OMe, 7-OH 6,7-[OMel2 6-0Me, 7-OCHz-CH=CMe2 6-OMe, 7-OCHz-CHOH-C(Me2)OH 6-0H, 7-0+!
C67-8 C67-9 C67-10 C67-11 C67-12 C67-13
OH
6-OMe, 7-O-a-D-gkp 6-OMe, 7-(2’-OAc)-O-a-~-glcp 6-OMe. 7-(6’-OAc)-O-a-D-glcp 6-OMe. 7-[2‘,6’-(oA~)~O-a-I~-glcp] 6-OMe, 7-[3’,6’-(0A~)~O-a-l~-glcp]
f, O-p-D-g’cp
7-OMe C67-22 6-CH2CHlC(Mel)OH, 7-OMe
E! 00
C67-23 6-CH2-CH=CMe2,7-OH
162.0
12.7
143.7
128.4
129.0
158.5
103.3
153.5
12.8
C
86
C67-24 6-o
161.6
13.1
143.9
126.9
132.1
158.7
105.3
154.6
12.4
C
87
157.4
13.9
143.3
130.3
128.2
160.8
99.6
156.7
12.1
C
77.88
159.9
114.6
143.1
131.3
123.3
162.0
99.8
158.7
112.7
C
77
160.5
113.9
143.3
130.4
122.3
160.7
99.6
157.3
112.2
C
88
162.5 162.5 161.2 163.5 160.7
113.8 113.4 112.5 114.1 114.5
143.8 144.2 143.2 145.8 142.8
125.0 125.6 129.4 131.3 130.4
124.2 124.0 125.2 128.9 124.1
160.4 160.3 160.6 160.1 157.0
99.0 99.3 98.4 103.9 96.8
155.7 155.5 154.4 155.5 154.4
12.6 12.6 11.7 14.8 13.4
C C C M C
85 85 89 90 90
163.8
114.1
146.0
131.2
128.9
160.2
104.2
55.5
114.8
M
90
160.7 C67-34 6-CH2-CHOAc-C(Me2)0H, 7-0-fl-~-glcp-(6+1)-P-r>-apiofhexaacetate
113.4
142.6
130.3
124.1
157.0
96.5
54.4
114.6
C
90
162.9
1 11.2
144.6
127.8
126.4
158.9
102.7
153.8
1 1 1.8
C
44
162.2
112.3
144.1
128.4
125.6
158.5
103.1
154.1
112.2
C
91
,7-OH
C67-25 6-CO-CH=CMe2.7-OMe 0
C67-26
C67-27 C67-28 C67-29 C67-30 C67-31 C67-32
6
,
0
,7-OMe
,7-OMe
6 OAc
6-CH=CH-C(Me2)0Me(trans), 7-OMe 6-CH=CH-C(Me2)OEt(trans), 7-OMe 6-CHzCHOH-C(Me*)OH,7-OMe 6-CH2CHOH-C(Me2)OH,7-0-P-~-glcp 6-CH2CHOAc-C(Me2)OH, 7-0-P-D-glcp tetraacetate
C67-33 6-CHz-CHOH-C(Me2)OH, 7-0-P-D-glcp-(6+ l)-P-D-apiof
,7-OH
C67-36
01I
e
0 \o
0
N
0
C67-38
6
q,
162.2
112.2
144.1
128.3
125.6
158.5
103.1
154.1
112.2
C
91
162.4
112.2
144.2
128.2
125.6
158.4
103.1
154.1
112.2
C
91
159.7
113.2
143.8
129.5
119.7
160.5
100.1
156.2
112.0
C
92
161.0
113.0
143.5
128.3
124.6
160.2
98.9
155.2
112.0
C
93
160.9
113.6
143.3
126.0
128.2
160.4
99.1
155.5
112.1
C
93
161.0
113.0
143.6
126.2
128.3
158.7
98.8
155.1
1 1 1.9
C
93
160.9
112.3
143.5
128.2
126.8
160.2
98.5
154.8
111.5
C
93
0
6
C67-39
C67-41
7-OMe
O
n
6%1^
C67-42
C67-43
I
6 C67-44
vy
. 7-OMe
vou
161.3
112.9
143.6
126.5
127.2
159.1
98.5
155.2
11 1.8
C
93
161.3
112.8
143.6
126.5
127.8
159.0
98.4
155.0
11 1.8
C
93
160.2 161.4 164.4
112.4 114.2 114.0
143.9 143.7 143.0
126.6 124.9 130.6
128.0 134.3 110.3
159.3 159.7 160.0
98.2 100.6 104.8
155.2 159.7 158.9
111.6 117.7 1 1 1.9
A C/M C
94 40 40
163.3
113.2
143.4
127.3
127.1
161.3
98.9
155.9
112.2
C
95
qTe
7-0Me
C67-45
O d 0
C67-46 C67-47 C67-48
~-COOH,7 - c o - p i 6-COOEt, 7-OH
0
Meorno ow ,7-OMe
6 -Me
C67-49
0
xo C67-50 C67-51
Me0
X=H X=Me
OH
,7-OMe
161.5
113.1
143.8
127.3
129.1
160.7
99.2
154.7
112.1
C
96
161.2
113.1
143.6
127.3
129.1
160.8
99.2
154.8
112.0
C
97
-
0 N N
M X 0e.7*
C67-52 C67-53
\
N
R,O
H
C67-55 C67-56 C67-57 C67-58
OR3
X = OH, KI = R3 = Me, RZ= H X = OMe, RI = R3 = Me, RZ = H
xM o-e7* \ C67-54
/
N
Rlo
/
113.6
145.6
131.5
111.6
147.3
99.7
155.8
112.9
C
97
112.2
144.4
128.4
130.5
160.9
98.4
152.0
111.5
C
97
161.8
113.5
145.6
130.0
111.7
162.3
99.6
155.8
112.8
A
98
161.9 161.3 161.3 161.8
112.2 113.0 113.8 112.3
144.4 143.9 144.6 144.3
128.3 128.6 129.2 128.4
130.8 129.4 130.5 130.3
160.9 159.8 160.3 160.9
98.4 99.0 98.9 98.4
154.4 154.4 154.9 154.3
112.2 112.3 112.2 111.8
C C C C
96 96 97 99
161.9
112.4
144.2
128.6
130.6
160.9
98.4
1.54.3
111.5
A
98
161.3
113.2
145.0
129.7
131.0
159.4
99.2
155.4
113.3
A
98
161.6 159.8 161.3 162.7 160.4
115.7 115.0 116.2 112.8 114.0
143.6 142.8 143.9 145.8 143.0
128.2 124.6 124.2 112.9 129.0
133.4 132.6 140.0 130.4 109.0
142.2 133.3 133.3 163.1 160.7
117.6 124.6 126.0 74.2 76.8
152.8 149.6 150.7 156.8 155.1
116.9 117.4 118.5 113.8 113.8
C C C M C
39 26 5 40
161.8
OR3
Me
X=H,Ri =Rz=K3=H X = OH, R I = Me, K2 = R 1 = H X = OH, RI = K3 = Me, Rz = H X = OMe, K I = Me, Rz = R3 = H X = OMe, RI = R3 = Me, Rz = H
0
C67-59 C67-60 C67-61 C68-1 C68-2 C78-1 C78-2
X I = OH, Xz = H X I = OMe, X2= OH
6,7-Mez 6,8-Mez 6-Et. 8-Me 7-OH, 8-1 7-OCHzPh, 8-1
100
159.3 161.1 160.0 159.7 160.1 160.5 160.6 160.0
114.5 111.7 I 1 1.4 113.5* 113.5 113.3 113.3 113.4
143.0 145.4 144.7 143.1 144.8 143.6 143.6 143.9
114.5 119.4 123.6 122.4 124.2 122.7 122.6 127.2
128.0 113.0 113.4 108.3* 111.7 108.3 110.0 114.9
160.4 150.0 153.9 155.4 153.4 156.0 154.9 159.5
81.3 132.6 134.2 136.3 131.4 134.9 136.5 108.4
155.0 144.2 148.1 148.1 144.8 148.4 148.6 153.3
C 114.5 D 112.7 u 112.2 113.7 C/CCI D 112.3 C 113.6 113.6 C 112.7 C
18 72 72 46
159.7
112.6
144.4
123.3
110.4
154.3
135.4
147.3
112.6
?
101
160.3
113.3
143.3
122.2
109.8
154.6
137.1
147.9
113.5
?
101
C78-13
160.3
113.5
144.5
123.8
110.6
155.9
136.7
148.6
114.3
A
43
C78-14
160.3
113.4
143.5
122.7
110.1
154.2
136.3
147.9
114.0
C
43
C78-15 C78-16 C78-17 C78-18 C78-19 C78-20 C78-21 C78-22 C78-23 C78-24
160.1 7-O-P-D-gkp, 8-OH 161.3 7-OMe. 8-CH2-CH2-Pr' 161.3 7-OMe, 8-CH2-CH2-C(Me2)OH 160.7 7-OMe, 8-CH2-C(Me2)-CH2-CH0 162.1 7-OH, 8-CH2-CH=CMe2 160.9 7-OMe, 8-CH2-CH=CMe2 160.4 7-OAc, 8-CH2-CH=CMe2 7-((3(Mez)-C=CH, 8-CH2-CH=CMe2 160.3 160.6 7-0H, X-CHl-CH=C(Me)CH20H(E) 7-OH, 8-CH2-CH=C(Me)-O-p-r)-glcp ( E ) 163.8
113.4 112.8 112.6 112.9 112.1 112.4 119.0 115.3 111.0 1 1 1.8
144.7 143.7 143.7 143.6 144.5 143.5 143.3 143.2 144.9 146.4
118.3 125.9 126.1 127.3 126.5 126.0 125.7 126.0 126.8 127.8
114.5 107.1 107.2 107.2 113.2 107.1 115.8 113.2 112.3 113.5
148.2 160.3 160.1 160.3 158.5 159.9 161.8 161.2 158.8 160.4
134.0 119.3 118.6 114.7 115.0 117.4 116.8 116.0 114.5 116.0
144.7 152.8 152.7
112.3 112.8 112.7 113.9 112.6 112.6 111.8 112.2 111.4 113.2
D C C C C
18 89 89 89 40 1.6,4 1.44 41 41 103 103
C78-3 C78-4 C78-5 C78-6 C78-7 C78-8 C78-9 C78-10
7-OC(Mel)-C&H, 8-1 7,8-[OHl2 7-OH, 8-OMe 7 3 -[ OMelz 7-OH, 8-0-p- D-glcp 7-OMe. 8-OCH2-CH=CMe2 7-OCH2-CH=CMe2,8-OMe 7-OCH2-CH=CMe2,8-CH2CO-Pr'
C78-11
7
C78-12
7
&
,8-OMe
a
, 8-OMe
153.1 152.4 151.4 152.5 153.1 154.6
C
C C D M/C
40 21 74
2
1
0 h) I*,
C78-25 C78-26 C78-27 C78-28 C78-29 C78-30 C78-31 C78-32
7-O-p-D-gICp, 8-CH2-CH=C(Me)CH20H( E ) 7-OMe. 8-CHI-CHOH-C(Me)=CH2 7-OMe. 8-CH2-CHOAc-C(Me)=CH2 7-OH, 8-CH2-CHOH-C(Me2)( )H 7-OMe, 8-CH2-CHOH-C(Me2)0H ~ - O M C~-CH~-CHOAC-C(MQ)OH , 7-OAc, 8-CH2-CHOAc-C(Me2)0H 7-OAc, 8-CH2-CHOAc-C(Me2)OAc
163.2 161.1 160.9 163.8 163.8 160.6 161.3 161.0
113.0 113.1 113.1 111.9 112.7 112.2 113.5 113.4
146.0 143.8 143.7 146.5 143.9 143.6 144.2 144.0
128.0 127.0 127.2 128.2 126.8 126.8 127.4 127.4
115.4 107.3 107.2 114.2 107.4 106.8 112.2 112.3
159.7 160.6 159.0 161.3 160.5 160.2 158.8 158.8
119.5 115.0 114.4 115.5 115.8 114.1 125.7 126.7
153.9 153.5 153.7 155.1 153.3 152.8 153.7 153.5
1 14.0 112.2 112.8 113.4 1 13.0 112.2 112.3 115.8
M C C M C C C C
103 40 40,240 104 40,89,104 89 40,104 40
160.7
112.7
143.5
127.0
107.2
160.5
113.9
153.1
112.6
C
89
160.7
113.2
143.7
127.0
107.5
161.0
115.0
153.4
112.4
C
40
7-OH. 8-CH2-CHOH-C(Me2)-O-p-D-glcp 163.9
111.9
146.5
128.3
114.3
155.1
115.5
161.1
113.5
M
104
C78-33 7-OMe, 8 +O
C78-34 C78-35
o f -
C78-36
7-OAc, 159.9 8-CH2-CHOAc-C(Me2)-O-f3-bglcptetraacetate
115.5
143.2
126.3
118.8
152.1* 119.3
153.5* 116.3
C
104
C78-37
7-OMe, 8-CH2-CO-Pr'
161.0
113.0
143.8
127.5
107.2
160.4
113.0
153.2
112.0
C
40,105
C78-38
7-OCH2-CHOH-C(Me2)Cl,8-CH2-CO-Pr' 160.8
113.4
143.8
127.8
108.6
159.6
113.4
153.2
112.0
C
105
C78-39
7-OMe, 8-CHOH-CHOH-C(Mc)=CHz(threo)
160.3
113.4
143.8
128.4
108.0
160.6
116.6
153.4
113.3
C
106
161.6
113.0
143.2
128.8
107.4
160.0
112.2
153.6
112.5
C
89,106
C78-40
X C78-41
7-OMe, 8
160.5
113.2
143.2
129.1
108.2
160.8
109.6
154.4
112.4
C
106
C78-42
7-OH, ~-CHOH-CHOH-C(MC~)OH
163.3
1 1 1.9
145.3
129.0
1 1 1.7
160.8
117.8
153.0
117.0
A
100
C78-43
7-OCH2Ph. ~-CHOH-CHOH-C(MC~)OH 160.1
113.5
143.8
129.0
109.5
159.5
117.4
152.9
113.6
C
100
160.1
113.3
143.5
129.2
108.2
160.9
113.1
153.5
112.2
106
159.7
114.1
143.0
134.1
108.2
163.3
114.1
156.9
113.3
c c
106
160.9
113.3
144.9
127.2
109.2
159.3
71.6(?) 152.7
113.2
C
100
160.9
112.9
143.6
128.5
107.5
159.4
113.1
153.0
113.0
C
106,107
160.4
113.2
143.6
129.8
108.0
160.7
107.0
152.9
113.0
C
92
160.0
113.0
144.9
128.3
113.4
160.0
109.7
152.8
1 1 1.0
D
99
C78-44
7-OMe, 8
*
C78-45
7-OMe, 8-CHO
C78-46
7-OCH2Ph, 8-CH=CH-C(Me2)OH(trans)
C78-47
7-OMe, 8-CH(CHO)=CMe2
C78-48
9
C78-49
A& 7-OH, 8 '
9" OMe
0 Me
Table 8.
I
F
chemical shifts of trisuhstituted coumarins $
X
4
u
Substituents
D345-1 D345-2 D345-3 D345-4 D346-1 D346-2 D346-3 D347-1 D347-2 D347-3 D347-4 D347-5 D347-6 D347-7 D347-8 D347-9 D347-10 D347-11 D348-1 D348-2 D348-3
3-Br, 4-OH, 5-Me 3-OC(Me)=CH2,4-OH, 5-Me 3-NOZ,4-OH, 5-Me 3-CHzPh, 4,5-[OH]2 3-Br, 4-OH, 6-Me 3-OC(Me)=CH2,4-OH, 6-Me 3-NOZ,4-OH, 6-Me 3-C1,4-OH, 7-OMe 3-OPh, 4-OH, 7-OMe 3-Me, 4-OH, 7-OMe 3-CH2CH2COOEt,4-Me, 7-OH 3-CH2CH2COOEt,4-Me, 7-OMe 3-CHzCH2COOEt,4-Me, 7-OAc 3-CHzCOOEt,4-Me, 7-OH 3-CH2COOEt,4-Me, 7-OMe 3-CH2COOEt,4-Me, 7-OAc 3-CH2Ph,4-OH, 7-OMe 3-Ph, 4-OH, 7-OMe 3-Br, 4-OH, 8-Me 3-OAc. 4-OH. 8-Me 3-N02. 4-OH, 8-Me
C-2
C-3
C-4
C-5
C-6
C-7
C-8
C-9
158.1 158.8 156.8 162.5 158.5 158.5 157.0 158.9 159.0 163.8 160.6 161.7 160.2 160.9 160.7 161.1 163.3 162.2 158.4 158.5 156.5
89.3 101.0 121.8 102.4 89.3 101.0 121.3 96.4 118.3 97.7 115.1 121.4 122.1 115.0 115.8 126.0 101.8 103.7 89.0 100.7 121.3
164.5 181.7 169.6 162.6 162.0 177.6 165.2 160.7 156.2 160.4 149.5 147.4 146.7 145.6 139.3 148.4 160.9 160.6 162.4 177.6 164.5
137.1 140.6 139.9 154.6 122.8 124.1 125.3 124.6 124.8 124.1 126.8 12.5.7 125.4 126.8 126.5 125.6 124.5 124.9 121.1 122.3 123.4
127.8 127.5 126.9 110.2 133.4 133.9 133.3 1 12.3 112.4 111.5 112.6 112.2 117.9 1 13.0 112.1 119.3 111.6 11 1.7 123.8 123.7 123.9
131.7 135.3 131.7 132.1 133.6 137.2 134.3 162.9 162.5 162.0 160.6 161.9 161.1 160.7 161.8 153.0 162.3 162.7 133.7 137.0 135.1
114.7 1 14.7 114.8 107.9 116.0 116.2 116.4 100.6 100.7 100.2 101.9 100.6 110.0 102.0 100.4 110.2 100.5 100.4 125.7 125.5 125.7
153.1 155.4 154.0 153.1 149.7 152.2 150.7 153.0 152.9 153.5 153.4 153.8 152.6 153.4 153.2 152.7 153.9 154.2 150.0 152.1 150.9
C-10
Soh.
114.4 D 113.6 D 1 19.0 D D 103.8 115.3 D 114.1 D 119.2 D 109.1 D D 109.6 D 109.7 112.1 C or D C 114.0 C 124.1 112.1 C or D 113 1 C C 1iX.I 109.5 D I) 109.6 D 115.6 D 114.1 D 118.3
Reference
65 65 65 7 65 65 65 27 27 27 68 68 68 68 68 68 27 27 65 65 65
uclH A S-OCH,, 7-OH
163.6
121.9
138.6
158.6
96.4
163.0
95.5
156.8
104.8
C
108
163.5
121.3
139.0
158.9
96.6
163.5
95.8
157.1
104.9
C
108
3-C(Me2)-CH=CH2,,6,7-IOMe12 160.1 132.2 ~-C(MC~)-CH=CH~. 6-CHzCH0, 160.0 132.5 7-OMe 3-C(Me2)-CH=CH2,6-CH2-CH=CMe2, D367-3 7-OMe 160.2 131.5 3-C(Me2)-CHO,6-CHO.7-OH 164.2 170.0 D367-4 D367-5 3-CH=CH-Pr' (trans), 6-CH2-CH=CMe2, 161.1 131.0 7-OH 3,6,8-Bn 154.9 112.8 D368-1 3-NMe2,6,8-C12 157.0 138.7 D368-2 3-C(Me2)-CH=CH2,,7,8-IOMe12 158.8 131.6 D378-1 3-C(Mez)-CH=CH2.,7-OH, D378-2 8-CH2-CH=CMe2 160.1 132.6 3-C(Mc2)-CH=CH2,7-OMe, D378-3 8-CH2-CH=CMe2 160.3 132.9 3-C(Me2)-CH=CH2,7-OAc, D378-4 8-CH2-CH=CMe? 160.2 133.1 3-C(Me2)-CH=CHz,7-OC(Me&C=CH, D378-5 ~-CH~-CH=CMQ 160.2 132.0 4-OH, 5-Me, 6-CI 160.5 92.1 D456-1 4,7-[OH]?, 5-Me 162.2 88.2 D457-1 4,7-[OMe]2. 5-Me 162.7 87.4 D457-2 4-Me, 5,7-[OH]2 161.0 108.8 D457-3 4-Me, 5.7-IOMeJ z 160.3 1 1 1.4 D457-4
137.3 137.4
108.3 129.6
146.3 130.0
149.2 159.8
99.5 98.5
152.2 154.8
1 1 1.8 112.6
C C
45 109
137.9 134.4
127.3 138.3
127.1 136.0
159.9 164.2
97.8 104.8
153.6 164.2
112.3 118.4
D C
45 86
138.6 143.1
125.4 121.6 129.3 108.4
157.5 136.2 126.6 146.6
102.5 116.4 121.4
112.6 109.9 123.3 113.7
C C
86 4 27
135.1
153.2 148.4 143.8 154.2
C
137.6
128.1 129.2 123.2 122.6
110
138.4
126.4
112.9
157.4
114.7
153.6
112.9
C
41
137.9
125.8
107.1
157.2
117.1
152.5
113.2
C
41
137.5
125.5
118.6
159.4
117.2
150.4
112.3
C
41
137.9 167.5 169.1 169.3 156.4 154.7
124.9 133.8 138.6 138.2 160.0 159.5
114.6 129.6 115.6 115.4 99.0 95.5
159.3 131.9 160.3 161.6 157.8 163.1
114.3 115.7 100.5 98.6 94.4 93.6
155.7 153.5 157.1 156.3 154.9 154.3
112.1 114.9 106.6 107.6 102.0 105.1
C D
41 65 65 14 68 38.68
D357-1
5-OCH,, 7-OH
D357-2 D367-1 D367-2
113.3
1)
1)
C D C
1
0 t4
J.
156.0 155.5 155.1 1.54.9 157.0 155.5 155.8 155.3 155.2 155.1 155.1 154.4
158.7 159.2 158.2 155.9 135.4 157.0 158.3 156.9 158.0 156.9 156.9 157.9 157.9 156.8 158.2 156.9 156.7 158.1 158.1 158.1 157.9
96.4 108.2 107.5 98.9 130.8 99.1 96.4 98.0 95.0 99.1 99.1 95.7 95.8 92.9 95.8 99.3 99.0 95 .4 95.9 95.9 95.9
162.5 155.5 143.4 162.2 143.4 161.6 162.1 162.8 163.2 161.2 161.3 162.8 163.0 159.8 163.0 161.2 161.3 163.0 163.0 163.2 163.0
94.0 100.7 110.4 95.6 116.1 94.8 95.7 92.9 93.9 94.5 94.5 93.6 93.7 98.2 93.7 94.6 94.5 93.3 93.8 93.8 93.8
157.3 148.6 155.6 157.2 155.7 156.7 156.5 156.7 156.5 156.6 156.5 156.4 156.4 156.1 156.7 156.7 156.5 156.5 156.5 1.56.6 156.3
104.9 C 112.2 C 108.5 C 97.7 CIM 110.6 CIM 100.5 I> 101.3 I> 101.7 D I) 102.5 100.5 D 100.5 D 102.6 D 103.3 D 102.1 CID 103.0 I> 100.8 D 100.5 D 102.3 D D 102.8 102.7 D D 102.8
112.3
158.3
158.1
98.1
163.2
101.1
157.3
104.5
M
114
163.0
113.2
158.0
158.0
96.7
165.0
100.0
157.2
105.5
M
114
159.5
112.2
156.3
155.4
98.5
162.7
95.2
56.0
103.3
D
115
p - ~ - x y l o p7-OMe , 162.9
113.3
158.1
158.1
96.6
165.0
100.6
57.1
105.8
M
114
112.3 90.2 91.0
158.3 169.2 164.3
158.1 127.5 122.3
98.7 126.4 128.3
163.3 113.5 140.3
101.1 145.7 118.3
57.3 44.5 51.8
105.1
M
114.9 114.9
D
114 31 65
D457-5 4-Me, 5,7-[OEt], D457-6 4-Me, 5,7-[OA~]2 D457-7 4,7-Me2, 5-OMe D457-8 4-CE, 5,7-[OH]2 D457-9 4-CF3,5.7-Me2 D457-10 4-Ph, 5,7-[OH]2 D457-11 4-Ph, S-OMe, 7-OH D457-12 4-Ph, 5-OH, 7-OMe D457-13 4-Ph. 5,7-[OMeI2 D457-14 4-@-OH-Ph),5,7-[OH]2 D457-15 4-@-OMe-Ph),5,7-[OH12 D457-16 4-@-OH-Ph),5,7-[OMeI2 D457-17 4-@-OMe-Ph),5,7-[OMc]2 ~457-184-(o,m‘-[OH]2-Ph), 5-OH, 7-OMe D457-19 4-(m,p-[OH]2-Ph),5,7-[OMc]z D457-20 4-(ni-OMe, 4’-OH-Ph), 5,7-[OH]z -P~), D457-21 ~ - ( / w O - C H ~ - O - ~ ’5,7-[OH]2 D457-22 4-(m-OMe,4‘-OH-Ph), 5,7-[OMe]2 D457-23 4-(ni-OH, 4’-OMe-Ph), 5,7-[OMe]2 OMeI2-Ph),5,7-[OMeI2 D457-24 4-(n~p-[ D457-25 4-(ni-O-CH2-0-4’-Ph).5,7-[OMeI2 D457-26 D457-27 4-(m,p-[ OH]z-Ph), 5-O-P-D-gICp, 7-OMe
D457-28 4-(m,p-[OH]2-Ph),S-O-P-D-galp, 7-OMe
I 1 1.4 1 12.2 113.6 109.9 1 16.4
161.5 160.3 161.3 161.3 158.4 159.8 159.6 159.7 159.4 159.8 159.7 159.4 159.4 162.6 159.7 159.8 159.8 159.6 159.6 159.6 159.2
110.8 111.2 1 1 1.8 109.5 109.7 1 11.2 1 1 1.5 110.6 11 1.3 109.7 109.8 1 1 1.1 111.2 1 10.7 11 1.6
163.6
110.1
154.9 15 1.6 154.5 141.8 140.5 155.9 155.4 155.7 155.1
38 68 38 70 70 111,112 112 112 112 112 112 112 112 113 112 112 112
112 112 112 112
D457-29 4-(m,p-[OH]~-Ph), s-O-p-D-glcp-(6+
D457-30 4-(m.p-[OHlz-Ph), D458-1 D467-1
S-O-P-D-glCp-(6t apiof, 7-OH 4,8-[OMe]2, 5-Me 4-OH, 6-C1,7-Me
163.5 161.8 161.4
c
-Is
m
112.2 115.1 I 1 1.1 I 1 1.3 107.3 107.9
153.0 152.2 152.4 151.2 142.5 135.5 160.5 153.2
109.6 109.0 10.5.3 118.8 109.0 11 1.3 126.1 124.9
142.8 143.2 146.2 138.7 143.2 148.3 122.9 123.5
150.2 151.2 152.7 151.5 151.1 143.6 162.5 162.2
102.9 99.8 100.0 112.2 103.5 11 1.3 102.6 98.8
148.0 147.4 149.4 144.6 149.9 110.5 154.7 151.8
111.9 112.2 112.1 118.2 105.9 151.9 111.1 111.6
C C/M D P C
160.1
131.7
137.8
124.1
127.0
159.6
98.1
153.9
112.4
C
4,6-Me2, 7-NEt 4,6,7-Me2 4-Me, 7,8-[OHI2 4-Me,7,8-[OMejz 4-Me, 7,8-[OAcI2 4.8-Mez, 7-OH 4-CF3, 7,8-[OH]2 4-CF3, 7-OH, 8-Me
161.9 161.6 160.7 160.6 159.2 160.8 160.3 160.0
109.1 114.2 110.5 112.4 114.8 110.1 110.7 110.9
152.6 152.6 154.1 147.8 145.3 153.1 142.7 140.7
125.0 125.0 115.7 119.5 121.7 122.5 116.5 123.3
118.4 133.1 112.6 108.1 118.6 111.8 113.2 112.9
149.6 141.9 149.6 155.4 151.9 159.1 150.3 160.6
96.1 117.8 132.5 136.3 130.5 111.1 132.5 113.1
154.8 152.2 143.6 152.6 146.8 153.1 143.7 154.3
109.5 C/CCI 117.9 C 113.3 C 114.6 C orD C 114.9 112.0 C 106.6 C/M 106.5 C/M
64 39 21 68 68 117 70 70
D567-1
5-OH, 6-OMe, 7-OCH2-CHOH-C(MeZ)OH
161.3
111.6
138.8
146.1
131.8
154.7
93.0
151.6
103.2
C
79
D567-2
5,6-[OMeI2 ~-OCH~-CHOH-C(MC~)OH
168.2
114.1
137.4
160.4
138.3
155.4
99.1
150.9
107.3
C
79
D567-3
5-OAC,6-OMe, ~-OCH~-CHOAC-C(M~~)OH
161.1
112.9
138.6
146.9
136.6
155.9
96.8
151.4
107.8
C
79
D567-4 D567-5
5,7-(OMe]2,6-O-P-~-galp 5,7-[OMeIz,6-CH=CH-Ac (trans)
164.8
109.4 111.4
142.3 138.5
146.2
139.2
158.0
98.2
151.9
112.5
M C
120 121
D567-6
5,7-[OH]z,6-P-D-glcp
160.4 154.0
109.2 107.9
139.4 139.0
153.7 159.9
102.2 108.2
154.9 159.9
94.6 93.8
159.9 154.4
108.6 104.2
D D
119
D567-7
5-CH=CH-Ac (trans), 6,7-[OMe]z
113.9
140.2
C
121
D467-2 D467-3 D467-4 D467-5 D467-6 D467-7 D467-8 D467-9
4-Me, 6,7-[OH12 4-Me, 6-OH, 7-OMe 4-Me,6,7-[OMel2 ~ - M c6,7-(OAc]2 , 4-CF3, 6,7-[OH]2 4-Ph. 6-OH. 7-OMe 4-Pr'. 6-Me, 7-OH 4-Pr', 6-Me, 7-OMe
160.8 159.9 161.4 160.1 161.0 160.3 161.7 160.6
1 10.6 1 10.9
D467-10
4-C(Mez)-CH=CH2,6-CHzCH20H, 7-OMe
D467-11 D467-12 D478-1 D478-2 D478-3 D478-4 D478-5 D478-6
101.4
11
D
c
21 74 74.68 74 70 111
116 116 109
118
0 t d
W
-
164.3 155.5 161.3 161.4 161.3
139.2 124.4 109.8 107.6 108.5
155.7 147.8 154.3 153.5 154.8
110.5 102.3 104.3 103.8 104.2
C
155.5 155.2 155.0
100.5 95.3 90.6 90.5 90.6
C C C
40 I22 106 106 106.12 1
138.8
147.3
90.3
155.7
113.2
54.2
107.0
C
40
1 10.9
138.8
155.8
90.3
161.4
107.3
47.3
113.2
C
40
161.1
1 10.5
138.6
155.7
90.2
161.2
106.0
153.9
103.5
C
6,106
161.2
111.0
138.8
155.8
90.4
161.2
107.5
154.0
104.2
C
106
D578-10 5,7-[OMe12,
161.4
110.7
138.7
156.8
90.3
159.3
107.4
151.2
104.0
C
40
D578-11 5,7-OMe, 8-CH2CO-Pr'
161.8
110.9
138.8
156.5
90.4
161.7
103.8
156.2
104.1
C
40,106
D578-12 5,7-[OHl2,
160.1
110.1
140.0
152.5
108.3
155.4
98.9
155.1
103.3
A
124
132.0 132.3 137.4 143.0
139.1 138.3 138.8 141.8
112.0 110.0 114.3 114.4
C
121 125 78 78 77
D578-1 D578-2 D578-3 D578-4 D578-5
5,7-[OMeI2.8-OAc 5-OH, 7-OMe. 8-0-a-n-glcp 5,7-[OMeI2,8-CH2CH2-Pr' 5,7-[OMe12,8-CH=CH-Pr' 5,7-[OMeI2,8-CH=CHAc
163.5 160.1 161.4 161.0 161.2
113.9 109.8 110.9 1 10.9 110.8
138.5 139.5 138.6 138.8 138.9
D578-6
5-OMe, 7-OH, 8-CH2-CHOH-C(Me)=CH2
161.4
1 10.8
D578-7
5,7-[OMel2, 8-CHz-CHOH-C(Me)=CH2
161.2
D578-8 D578-9
5,7-[OMe]2, 8-CH2-CHOH-C(Me2)0H
146.4 150. 1
8-CH2-CHOAc-C(Me2)OEt
8-CH2-CH=C(Me)-CH2CH2CH=CMe2
1)
D578-13 D678-1 D678-2 D678-3 D678-4
5-CH=CH-Ac (trans), 7,8-[OMe12 6,7,8-[OHl3 6-OMe, 7,8-[OH12 6-OMe,7-OCH2-CH=CMe2,8-OH 6,8-[OMeI2.7-OCH2-CH=CMe2
161.4 159.8 160.2 160.6
114.4 112.0 111.4 114.4 115.1
139.0 144.5 144.9 144.7 143.5
103.5 100.0 100.2 103.6
106.4 143.5 144.7 150.1 144.9
137.5 138.8 139.0 150.7
D678-5
6-OMe, 7-OCH2-CH=CMe-CH20H( E ) , 8-OH
159.3 159.3
113.8 113.8
143.8 143.8
99.8 99.8
149.2 143.8
137.9* 137.9 140.4* 137.9
140.4* 113.8 137.9* 113.8
D 1)
78 128
6-OMe, 7-o-p-D-glcp, 8-OH
160.0
115.0
144.4
100.9
149.4
138.2
138.4
D
78
D678-6
137.2
115.0
?
D D C
0 b2
0
6.8-[ OMe],
D678-7
160.4
115.1
143.3
103.8
50.7
145.8
141.9
143.1
114.4
123
D678-8
160.5
115.3
143.5
103.8
50.7
145.1
141.9
143.2
114.5
123
I
160.5
115.2
143.5
103.9
150.7
145.3
141.9
143.1
114.5
123
X =Ac
160.4
115.3
143.4
104.2
150.6
145.4
141.7
143.4
114.6
123 123
160.4
115.2
143.3
104.1
150.5
145.4
141.7
143.2
1143
C
123
160.4 160.3
114.9 115.3
143.4 143.4
104.1 104.2
150.5 150.6
145.5 145.6
141.6 141.7
143.2 143.4
114.3 114.5
C C
123 123
1
D678-9
&
D678-11
D678-12
I
6.8-IOMcI2
I 1 0 '>'
7-01 -
D678- 13 D678-14
XO
.
6,8-[OMe12
X =H X = Ac
D678-15
D678-16 D678-17
D678-18
I10
XO
1 1 0 w 6,8-[OMe],
X =H X =Ac
160.4
11.;.
i
143.3
104.3
150.6
145.4
141.7
143.4
114.6
C
126
160.5 160.4
115.1 115.3
143.3 143.3
104.1 104.2
150.7 150.6
145.6 146.0
141.7 141.6
143.4 143.3
114.5 114.5
C C
123 123.126
160.3
115.0
143.4
104.0
150.5
145.6
141.5
143.0
114.4
C
123.126
Table 9.
I
?C chemical shifts of tetra- and penlasubstituted coumarins
Substituents
E3457-1 E3457-2 E3457-3 E3457-4 E3457-5 E3457-6 E3457-7 E3458-1 E3467-1 E3467-2 E3467-3 E3467-4 E3467-5 E3467-6 E3467-7 E3478-1 E3478-2 E3478-3 E3478-4 E3478-5 E3478-6 E3678-1
3,S-Mez. 4,7-[OMel2 3-CH,COOEt, 4-Me, 5,7-[OH]2 3-CH2COOEt, 4-Me, 5,7-[OMe]z 3-CH2COOEt. 4-Me, 5,7-[OAc]2 3-CH2CH,COOEt, 4-Me, 5,7-[OH]2 3-CH,CH,COOEt, 4-Me, 5,7-[OMeI2 3-CH2CH2COOEt. 4-Me, 5,7-[OAc]2 3,4,8-[OMeI3, 5-Me 3-[OMeJ2,4-Pr', 6-Me, 7-OH 3-CH,COOEt, 4-Me, 6,7-[OH]2 3-CHZCOOEt. 4-Me, 6,7-[OMelz 3-CH,COOEt, 4-Me, 6,7-[OAc]2 3-CH2CH2COOEt,4-Me, 6.7-IOH12 3-CH2CH2COOEt,4-Me, 6,7-[OMe], 3-CH2CH2COOEt,4-Me, 6,7-[OAc]2 3-CH2COOEt, 4-Me, 7.8-10812 3-CH2COOEt. 4-Me, 7,8-[OMel2 3-CH2COOEt, 4-Me, 7,8-[OAc]2 3CHzCH,COOEt, 4-Me, 7,8-[OH]2 3-CH2CH2COOEt,4-Me, 7,8-[OMe]2 3-CH2CH2COOEt,4-Me, 7,8-[OAc]2 3-C(Me2)-CH=CH2,6,7,8-[OHI1
c-2
c-3
c-4
c-5
C-6
c-7
164.3 160.7 158.9 160.1 160.5 159.9 159.3 159.4 158.8 161.2 161.9 150.8 161.0 161.8 160.0 160.8 161.0 159.2 160.6 161.4 159.7 158.5
109.9 113.0 115.4 120.5 117.4 120.2 124.7 130.5 139.1 115.1 116.9 119.5 119.5 121.9 124.9 112.8 1 16.5 119.2 119.3 121.7 121.7 129.2
166.8 151.2
137.1 160.4 162.0 153.6 159.9 161.9 152.6 127.3 126.4 109.8 105.7 119.1 109.1 105.7 118.7 115.6 1 19.6 121.9 1 15.4 119.7 121.7 99.9
115.8 99.4 95.8 113.8 99.3 95.7 1 14.4 126.9 123.0 142.3 146.8 138.8 142.7 146.1 138.7 112.2 108.0 1 18.6 112.0 108.1 118.3 144.7
161.2 157.7 161.4 151.6 157.4 161.4
151.1
147.2 149.7 149.5 145.5 157.6 151.5 146.3 148.8 1SO. 1 146.1 147.2 146.4 142.0 146.4 145.6 141.9 147.1 145.0 145.0
c-8
c-9
c-10
Soh.
Reference
98.8 94.2 93.2 108.2 94.1 93. I 151.1 108.0 112.0 145.5 158.9 102.6 149.6 102.4 152.4 100.0 150.1 112.1 149.2 102.3 99.9 152.0 149.8 1 1 1.7 149.9 131.9 154.7 135.8 148.4 130.3 148.3 137.9 154.6 136.0 146.7 130.0 141.4* 131.8
155.4 154.8 1.55.2 147.9 154.7 155.3 147.7 141.0 146.2 149.1 148.2 147.7 147.6 147.9 143.9 148.9 149.0 144.8 148.3 146.5 144.2 138.4*
115.8 102.2 105.3 111.9 102.3 105.4 11 1.8 1 16.9 111.5 111.7 112.9 118.6 11 1.8 113.0 119.1 1 14.9 114.8 119.7 113.1
C D C C D C C C P D C C D C C D C C D C C
127 68 68 68 68 68 68 31 116 68 68 68 68 68 68 68 68 68 68 68 68 128
115.3
119.3 110.0
D
1
E: w
E4567-1
4-SMe, 5-Me, 6-OMc. 7-OH
159.3
104.3
160.1
129.6
143.1
151.9
102.2
151.5
1 1 1.9
C
129
E4568-1
4-Pf'. 5-011. 6-CH2-CH=CMe2, 8-COCH2Pr'
159.4
110.1
159.0
158.8
102.3
156.3
104.5
165.4
109.9
C
130.13 1
E4568-2
4-Pf', 5-OH, 6-CH2-CH=CMe2, 8-COCH(Me)Et
159.4
110.1
159.0
158.8
102.3
156.1
104.2
165.4
109.9
C
1 30,132
E4578-1
4-Q-OMe-Ph), 5,7,8-[OMcJ3
160.4
112.8
156.8
156.0
94.0
154.5* 132.8
149.9* 103.9
A
133
E4578-2
4-(p-OMe-Ph), 5,7-[OMe] 2, 8-OEt
160.4
112.8
157.4
156.8
94.1
154.6
132.4
150.7
103.0
C
133
4-(mVp-[OMe12-Ph), 5,7-[OMelz, 8-OH
161.6 162.8
111.1 112.6
144.8 158.7
157.1 152.9
93.4 95.9
150.8 152.5
127.6 129.6
150.2 145.2
103.4 105.4
D 1)
134 135
E4578-4
4-(m-OH,p-OMe-Ph), 5.7-[OMe12, 8-OH
160.6 162.8
111.4 112.5
145.6 158.4
156.3 152.9
94.5 94.7
151.9 152.2
128.1 129.5
150.8 145.2
103.3 104.2
D D
134 136
E4578-5
4-(m-OH,p-OMe-Ph), 5,7,8-[OMe]3
159.3
111.2
155.3
153.6
94.0
155.3
129.8
144.6
102.8
D
134
E4578-6
4-(ni-OCHzO-p-Ph),5,7,8-[OMe]?
160.2
153.8
155.1
92.7
155.8
129.7
145.9
102.3
C
134
E4678-1
4,8-Me2, 6-CH2CH=CH2,7-OAc
160.6
114.3
152.3
122.8
1 19.7
1S0.4
117.9
151.1
1283
C
137
E5678-1
S-CHz-CH=CMe2,6-OMc, 7,8-[OH]2
159.3
1 11.0
142.7
121.9
142.4
141.4
130.8
140.7
108.6
D
128
E5678-2
5-CH2-CH=CMe-CH20H( E ) , 6-OMe, 7,8-[OH]2
159.3
1 1 1.0
142.7
121.7
142.4
141.4
130.9
140.5
108.6
D
128
E34578-1 3,5-Me2, 4,7,8-[OMe13
163.6
110.2
166.5
130.7
112.1
153.6
134.4
147.5
110.6
C
127
E45678-1 4-Ph, 5,7-[OH12,6-COCH2Pr', 8-CHz-CHXMe2
159.4
112.7
154.5
156.7
100.9
163.2
107.3
159.2
108.1
C
130.138
E45678-2 4-Ph. 5,7-[OHI2,6-COCH(Me)Et, 8-CH2-CH=CMe2
159.4
112.7
163.2* 154.5* 100.9
156.9* 109.3
159.2
103.1
-
E4578 3
139
Tahle 10. I T chemical shifts of linear (F) and angular furanocoumarins (H), their 2’,3‘-dihydro derivatives (G and I) and of coumarins with other annulated five-rnembered rings (J) 5
Su bstituents F-1 F-2 F-3 F-4 F3-1
- (linear furanocoumarin)
2-thiono 1’-thio 1 ’-seleno 3-C(Me2)-CH=CH2
c-2 161.1 197.4 159.8 160.6 159.2
4
c-3 c - 4 114.7 127.0 114.9 115.7 132.4
144.2 136.6 142.8 143.7 144.9
c-5 c - 6 120.0 120.4 121.3 123.5 119.1
125.0 125.7 135.4 138.8 124.0
c-7
5-OH 5-OMe
F5-3 2-thion0, 5-OMe F5-4 S-OCH~-CH=CMC~
F5-6 5-OCH*-CHOH-C(Me)=CHz F5-7 ~-OCH~-CHOH-C(MC~)CI F5-8 5-OCH2-CHOAc-C(Me2)CI
c-9 c-10 c-2’ c-3’ Solv.
156.6 99.9 152.2 115.6 156.0 99.1 153.2 116.8 142.2 108.8 149.7 115.7 144.8 113.0 150.4 116.6 155.2 98.2 150.2 115.3
8
147.0 148.6 126.5 129.4 146.0
,
Reference
106.6 106.7 122.4 126.7 105.9
11
10
C C C
140 140 4434, 110,141
99.2 150.7 117.7 146.4 106.2
C
110
160.1 110.8 139.5 147.7 112.4 156.8 90.9 152.6 103.7 144.7 104.6 160.3 112.8 139.4 149.6 113.0 158.5 94.0 152.7 106.7 145.0 105.3
I) C
197.2 125.3 131.2 149.4 113.1 157.9 92.7 154.2 107.4 146.4 105.6 160.7 112.0 139.0 148.4 113.7 157.6 93.7 152.1 106.9 144.4 104.7
I) C
142 8,10,44 45,14 I , I43 8.10 77,142
160.3 112.3 138.5 147.8 113.5 157.4
93.8 151.9 106.6 144.8 104.1
C
1
161.1 112.9 139.2 148.4 114.2 158.0 161.2 112.5 139.0 148.8 113.8 158.8 161.0 113.1 138.9 148.2 113.7 150.1
94.7 152.6 107.3 145.2 104.7 94.3 152.7 107.5 145.4 104.4 94.5 152.6 106.8 145.2 104.7
C C C
143 143 143
162.9 126.6 138.0 118.7 124.5 155.5 F5-1 F5-2
c-8
C
8,10,44
%
L
F5-9 F5-10 F5-11 FS-12
5-OCH2-CHOH-C(Me2)0H 5-OCH2-CHOH-C(Me2)OMe 5-OCH2-CHOAc-C(Me2)OAc 5-OCH2-COPr'
FS-13
5-0
v
4 Q
160.4 161.2 160.8 160.5
111.8 112.7 113.2 112.7
139.9 139.4 139.0 138.9
149.1 148.8 148.8 147.5
113.2 113.4 113.2 113.2
157.6 1.58.1 158.1 157.5
93.0 94.4 94.5 94.5
152.1 152.5 152.3 152.1
106.4 107.2 106.9 107.0
145.4 145.0 145.2 145.1
105.5 104.9 104.7 103.8
C C C C
44 143 143 1
161.0 113.0 139.0 148.5 113.5 158.2
99.4 152.7 106.9 145.1 104.7
C
144
161.3 112.4 139.5 148.4 112.8 158.1
93.9 152.4 106.4 145.0 104.9
C
145
D C D
5 - 0 4 0
F5-14 F8-1 F8-2 F8-3 F8-4 F8-S F8-6 F8-7 F8-8 F8-9
160.0 160.4 196.6 160.2 161.2 160.1 160.1 8-OCH2-CHOH-C(Me2)-O-P-D-glcp 160.0 8-OCH2-COPr' 160.0
F8-10
"
F8-11
8-OH 8-OMe 2-thiono, 8-OMe 8-OCH2-CH=CMe2 ~-OCH~-CHOH-C(M~)=CHZ 8-OCH2-CHOH-C(Me2)OH 8-OCH2-CHOAc-C(Mez)OH
O
W
113.7 114.5 127.1 114.2 114.4 113.5 114.8 114.2 113.3
145.3 144.4 136.8 144.3 144.9 144.2 144.2 145.4 144.3
110.1 113.1 113.4 113.2 115.6 116.4 113.7 113.8 114.8
125.2 126.2 127.1 125.7 126.8 126.0 125.9 125.9 126.0
145.3 147.6 147.0 148.3 148.7 146.7 147.9 147.2 146.6
130.1 132.7 131.6 131.2 132.4 131.5 131.3 131.6 131.2
139.8 142.9 144.8 143.5 143.1 143.2 143.2 142.7 142.7
116.2 116.5 117.9 116.2 116.9 114.6 116.5 116.5 116.5
147.2 146.6 148.5 146.4 147.4 147.6 146.7 147.9 146.6
106.9 106.8 107.0 106.5 107.4 106.7 106.8 107.1 106.8
C
1,8,10 1,8,10,143 10 1.8 143 146 103 147 143
160.8 115.0 144.6 113.2 126.1 148.9 131.6 143.7* 116.7 146.9' 107.0
C
48
159.7 114.8 145.2 113.6 125.0 148.4 129.2 144.4* 116.6 146.8' 106.9
C
147,148
C
C C C D
s
F2 ‘-1 F2’-2 F58-1 F58-2 F58-3
2’-Pr‘ 2-thiono, 2‘-Pr‘ 5- B r, 8-OCH2-CHB r-C(Me2)B r 5-OMe. 8-OH 5 ,8-(( )Me)>
161.1 197.2 159.0 159.7 160.5
114.1 126.7 115.8 112.2 112.8
144.2 136.7 142.1 139.6 139.5
F58-4 F58-5
5-OMe, 8-OAc 5-OMe, 8-OCH2-CHOH-C(Me2)C1
159.7 161.1
12.6 139.1 146.7 113.6 149.9 118.1 13.8 140.3
F58-6
5-OMe, 8-OCH2-CH(O-p-~-glcp)-C( Me2)OH
118.8 119.2 127.7 141.0 144.4
126.5 127.0 115.5 114.7 114.9
156.3 99.1 155.7 98.6 146.4 130.6 146.9 125.3 149.9 128.3
167.3 99.5 167.4 99.9 147.1 107.3 146.0 105.1 145.3 105.3
C C C C C
43.8 106.9 145.1 105.3
C C
8,lO 10 1SO 44,141 8,10,44, 45, I 10 44,141 105
159.4
12.5 139.6 144.1 114.4 149.2 126.4 142.9 106.8 146.2 105.6
D
149
F58-7
5-OMe, 159.5 8-OCH2-CHOH-C(Me2)-O-P-~-glcp
12.5 139.6 144.1 114.5 149.6 126.8 143.2 106.9 146.2 105.4
1)
149
F58-8
5-OCH2-CHOH-C(Me)=CH2,8-OMc
162.1 112.5 140.3 143.1 114.3 148.9 127.4 146.5 107.1 145.4 105.0
C
151
F58-9
5-OCH2-CHOH-C(Me2)-O-~-D-glcp, 159.6 112.4 140.2 143.8 115.8 149.0 127.5 142.8 107.9 146.3 105.5
1)
149
160.4 114.8 144.3 131.0 126.0 148.3 131.5 148.6 116.6 146.7 106.8
C
152
160.3 113.1 139.5 143.6 115.3 149.8 128.6 143.1 108.1 145.3 105.0
C
145
CIM C C C
143 143
51.5 53.6 43.7 39.6 43.7
115.0 116.5 106.3 107.1 107.7
8-OMe
5-OMe, 8
-
O
e
o
F58-10
5-o+ F58-11
’
OH.
8-OMe
109.9 112.3 114.3 110.1
156.1 111.4 150.2 104.6 143.4 156.7 114.2 150.8 107.8 144.2 153.5 99.8 151.5 114.7 152.5 155.0 112.8 154.6 104.4 146.1
F58-12 S-OH, ~-C(MC~)-CH=CH~ F58-13 5-OMe. 8-C(Me2)-CH=CH2 F2’3‘-1 2’-Pr‘, 3’-OMe 5-OH, ~ - C ( M C ~ ) - C ~ ~ = C H ~ , F582’-1 2’-C(Me)=CH2
162.3 160.7 160.9 159.9
F482’-1
161.6 112.9 155.7 112.3 125.6 153.4 116.2 149.1 109.3 157.6 102.9
4,8,2’-Mej
140.5 139.4 144.0 139.9
146.0 142.6 116.5 150.7
113.5 115.0 121.6 114.2
103.2 104.0 136.3 100.6
C
1
84 8
-
8 4
+ W 0 00
G3-1 G5-1 G2’-1 G2‘-2 G2’-3 G2’-4 G2’-5 G2’-6 G2’-7 G32’-1 G32’-2 G52’-1 G82’-1 G82’-2 G82’-3 G82’-4 G82’-5
3-C(MeZ)Et 5-OMC 2’-C(Me)=CHz 2 ‘-C(Mez)OH 2‘-C(Me2)0COC(Me)=CHMe (Z) 2 ‘-C(Me)OCO-CH=CMez 2’-C(Me2)O-P-D-glcp 2 ’-CH2-CH=CMez 2‘-C(Me2)-O-p-D-glcp-(1 t6)-apiof 3-C(Me2)-CH=CH2,2’-C(Mez)OH 3-C(Mez)-CH=CH2,2’-C(Me2)OAc SOH, 2’-C(Me)=CHz 8-OMe. 2 ’-C(Me)=CHZ 8-OH, 2’-C(Mez)OH %OH, 2’-C(Me2)0SO
162.6 161.5 162.0 162.2 161.1 161.0 160.4 161.3 160.3 160.0 159.6 161.3 161.1 160.2 160.3 160.1 160.3
113.0 110.4 111.6 112.7 112.6 112.5 112.1 112.8 112.2 112.9 112.8 104.4 113.3 112.7 113.0 111.4 112.9
89.1 72.4 87.8 91.3 88.7 90.0 87.6 89.7 90.8 88.2 88.1 87.1 90.5 89.4 91.3 89.8
28.6 28.3 33.4 29.5 29.6 29.3 27.8 33.6 29.0 29.5 29.7 32.0 31.5 29.4 29.8 28.9 29.6
C C C C C C D C D C C A C D D D D
86 8 143 1,8,1S3-155 153 1 154 150 156 84 110 143,157 103 154,158 158 45,154 154
G82’-6
%OH, 2 ’-C(Mez)O-(6“OMe)-p-I)-glcp
161.3 111.0 143.7 114.0 125.2 151.0 128.1 145.0 113.0
89.5
29.6
D
154
G2’3’-1 G2’3’-2 G2’3’-3 G2’3’-4
160.4 160.6 (trans)-2’-C(Men)O-p-D-glcp, 3’-OH 160.1 160.6 (cis)-2’-C(Mez)O-P-D-glcp,3’-OH
112.8 114.3 112.6 112.9
91.9 91.8 97.0 91.9
71.0 73.2 70.0 77.5
D D D D
159 159 156 160
G352’-1
3-C(Mez)-CHOH-CH20H,5-OMe, 2’-C(Mez)OH
162.3 129.4 139.2 123.7 125.0 159.8
95.8 153.8 112.5
90.7
28.8
D
109
G352‘-2
3-C(Me2)-CHOAc-CH20Ac, 5-OMe, 162.7 127.8 139.3 123.4 124.9 160.0 2’-C(Me2)0H
97.1 154.7 112.7
91.0
29.5
C
109
(cis)-2’-C(Mez)OH, 3’-OH (cis)-2’-C(Me2)0H, 3’-OAc
130.7 138.6 110.5‘ 139.2 1 1 1.8 143.6 1 1 1.6 144.3 112.0 143.8 111.8 143.5 111.2 144.6 112.2 143.7 111.9 144.4 130.5 137.9 130.4 137.7 110.2 139.8 107.3 137.3’ 110.8 144.9’ 110.9 145.1’ 112.9 147.8’ 111.0 143.7’
123.1 124.1 152.7 105.9 123.4 125.5 123.6 125.6 123.4 124.5 123.1 124.4 123.8 125.6 123.4 124.6 123.8 125.4 123.1 124.5 122.8 123.7 151.2 109.0 115.9* 113.6 113.9* 125.2 114.0* 125.2 117.3* 127.0 114.0* 125.5
111.9 144.8 125.7 128.5 114.2 143.9 127.4 124.4 111.5 144.5 125.2 127.9 111.8 144.9 125.7 128.6
160.2 165.5 162.0 163.7 163.4 163.1 163.0 163.5 162.9 162.1 161.9 165.2 160.5 151.3 151.2 152.7 151.1
97.0 154.5 92.9 156.6 97.5 155.3 97.7 155.6 97.6 155.7 97.6 15’5.4 96.7 155.0 98.0 156.1 96.7 154.9 96.8 154.4 96.7 154.2 90.6 157.8 126.9* 146.4’ 128.0* 143.6’ 128.1* 143.8’ 126.6* 145.9’ 128.1* 145.1’
162.8 97.8 156.2 163.9 96.9 158.1 163.0 97.0 155.9 162.4 97.7 156.1
H-1 H3- 1 H3-2 H3-3 H3-4 H3-5 H3-6 H4-1 H5-1 H5-2 H6-1 H6-2 H2’-1 H2’-2 H2’-3 H2’-4 H2’-5 H3’-1 H34-1 H32’-1 H32’-2 H32’-3 H32’-4 H32’-5 H32’-6 H33’-1 H46-1 H42’-1 H43’-1 H56-1 H56-2 H52’-1 H62’-1
- (angular furanocoumann)
3-Me 1 ’-thiono, 3-COOMe 1 ’-thiono. 3-COOEt I ’-thiono, 3-CONH2 1 ‘-thiono, 3-CSNH2 1 ’-thiono, 3-CsN 4-Me 5-OMe 5-Me 6-Me 6-OMe 2’-Me 2-P$ 2’-C(Me2)OH 2‘-C(Me2)Ph 2’-C(Me2)(p-OMe-Ph) 3’-Me 3,4-Me2 3.2’-Me2 1 ’-thio, 3-COOEt,2’-Br 1 ’-thio, 3-COOEt, 2’-Me 1‘-thio, 3-CONH2,2‘-Me 1 ’-thio, 3-CSNH2, 2’-Me 1’-thio, K I N , 2’-Me 3,3‘-Mez 4.6-Me2 4,2 ’-Me2 4,3’-Me2 5,6-[OMeI2 5,6-Me2 5,2 ’-Me2 6,2’-Me2
160.2 162.2 155.5 155.7 160.1 159.0 156.7 160.7 160.8 160.6 161.1 160.1 161.0 160.9 160.9 160.7 160.7 160.4 161.8 162.3 155.2 155.6 160.0 159.0 156.6 162.1 161.1 160.8 160.4 159.5 160.9 161.0 161.2
114.5 123.1 113.2 113.0 114.1 1 14.1 100.2 112.8 112.0 113.5 114.0 114.1 113.8 113.7 113.8 113.8 113.8 113.7 119.7 122.7 113.8 113.2 114.0 114.0 100.0 122.8 112.7 112.7 112.5 113.7 113.2 113.0 113.5
144.5 140.2 119.1 119.1 119.5 119.6 119.8 153.6 139.7 141.3 144.5 145.3 144.6 144.6 144.5 144.5 144.6 144.4 147.0 140.4 118.6 118.7 119.0 119.2 119.4 140.3 153.5 153.5 153.3 140.0 141.9 141.5 144.5
123.9 123.0 149.7 149.5 148.4 150.2 154.0 120.5 154.1 132.2 123.7 104.8 122.7 122.7 123.4 123.0 122.9 123.7 120.3 121.8 149.0 149.0 148.5 150.2 153.9 122.9 120.4 119.3 120.3 142.6 129.5 130.9 122.6
108.8 108.5 124.7 124.7 124.8 124.9 124.2 108.3 90.4 109.6 119.1 142.5 108.1 108.2 108.6 108.0 108.1 108.7 108.0 107.9 125.4 124.0 124.1 124.2 123.6 108.5 118.6 107.7 108.2 134.6 117.4 109.1 118.4
157.3 156.6 145.7 145.5 145.5 145.5 146.6 157.2 157.8 157.3 156.7 146.1 157.1 157.0 157.0 157.2 157.2 157.9 156.3 156.3 146.1 145.2 144.9 145.2 146.3 157.1 156.2 157.1 157.9 149.1 156.7 157.0 156.0
116.9 116.6 115.6 115.6 118.3 122.2 114.6 116.9 109.9 115.1 116.3 117.7 118.3 118.0 117.7 117.8 117.8 117.4 116.5 117.9 116.7 115.8 117.7 122.1 114.5 117.2 116.3 118.5 117.4 113.4 114.0 116.4 117.5
148.5 147.3 150.0 150.1 150.0 149.4 149.8 147.9 148.6 149.1 146.9 142.3 147.6 147.8 147.9 148.0 147.9 149.8 147.0 146.6 149.2 149.8 148.7 146.4 147.2 149.4 144.2 147.2 148.1 146.0
113.5 114.1 127.4 127.3 127.8 127.4 127.7 114.5 105.7 112.7 113.6 113.8 113.4 113.3 113.4 113.4 113.2 113.4 114.9 114.0 127.3 127.7 127.8 127.8 128.1 114.0 114.6 114.5 114.3 109.0 112.8 112.5 113.3
145.9 145.6 129.7 129.7 129.5 129.8 130.4 145.8 144.2 145.0 145.6 147.4 156.8 166.3 164.7 167.1 167.4 142.3 145.4 156.4 116.9 145.2 143.5 143.6 144.1 142.0 145.5 156.7 142.2 146.9 144.8 155.8 156.4
104.0 103.9 119.7 119.7 119.7 119.7 119.7 104.3 103.8 104.0 104.4 104.2 99.9 97.2 98.0 98.9 98.6 115.9 104.1 99.8 122.8 117.5 117.4 117.5 117.4 115.8 104.6 100.2 116.2 103.8 104.3 99.8 100.1
C C D D D D D C C C C D C C C C C C C C D D D D D C C C C D C C C
161 162 163 163 163 163 163 162 149 162 162 149 162 161 161 161 161 162 162 162 163 163 163 163 163 162 162 162 162 149 162 162 162
-
502 H63’-1 H2’3’-1 H342’- 1 H343’-1 H462’-1 H463’-1 H562’-1 H 563’- 1
6,3‘-Mez 2‘,3’-Mez 3,4,2’-Me3 3,4,3’-MCz 4,6,2’-MQ 4,6,3’-Mez 5,6,2’-Mez 5,6,3’-Me>
161.0 113.5 144.5 1233 119.0 156.8 116.7 148.2 113.3 160.9 113.3 144.6 122.4 108.0 156.2 118.4 148.8 113.3 162.1 119.4 147.2 119.0 107.4 156.1 118.0 145.2 114.9 161.6 19.4 146.9 120.1 108.0 156.9 116.9 147.4 114.8 161.3 12.4 153.6 119.3 118.0 156.0 117.7 145.6 114.4 161.1 12.5 153.5 120.3 118.5 156.8 116.8 147.7 114.4 161.1 12.7 141.9 128.3 116.7 156.3 115.2 146.2 112.5 160.8 12.7 141.8 129.2 117.2 157.1 114.4 148.5 112.5
12’-1 I2’-2
2 ‘-C(Me2)OH 2’-C(Me2)0CO-C(Me)=CHMe(Z) 5-OMe, 2 ’-C(Me2)OI-I 5-OMe, 2’-C(Mez)0CH2-CH=CMe2 2’-C(Me2)OAc,3’-OCOEt 2 ’-C(MeZ)OAc, 3 ’-OCOC(Me)Et
161.0 161.0 161.8 161.6 159.7 159.4
I2’3’-3
2’-C(Me2)0COCH2Pr‘, 3 ’-KO-C( Me)=CHMe (Z)
160.4 113.7 111.1 131.2 108.1 164.3 113.5 152.4 113.9
12’3’-4
2’-C(Mez)WO-C(Me)=CHMe ( E ) , 3’-OCO-CH(Me)Et
159.5 113.2 143.4 131.3 107.6 163.6
12’3’-5
2’-C(Me2)OCO-C(Me)=CHMc(Z), 3’-OCO-CH(Me)=CHMe (2)
152’-1
I52’-2 12’3’-1 I2’3’-2
Fo
142.0 151.7 156.3 141.9 156.3 141.9 155.4 141.2
116.2 109.8 100.0 116.0 100.4 116.5 100.0 116.1
C C C C C C C C
162 162 162 162 162 162 162 162
108.6 91.4 113.5 89.3 102.5 65.9 102.4 65.9 113.4 88.2 113.3 88.2
27.6 27.6 27.1 26.1 68.6 68.3
C C C C C C
41 143,145 166 166 167 167
88.9
68.5
C
168
81.1 151.8 113.3
88.3
68.3
C
167
160.4 113.7 144.1 131.2 108.1 164.3 113.1 152.4 113.8
89.2
68.5
C
168
160.7 112.9 144.1 129.7 107.1 162.4 113.7 151.8 113.8 113.9
48.8
C
169
12.3 112.1 108.5 111.3 113.3 113.1
144.0 144.0 141.5 138.7 143.4 143.3
128.7 128.9 154.7 154.8 131.3 131.3
106.7 106.7 93.3 93.2 107.6 107.5
163.7 164.0 145.6 115.7 163.6 163.5
114.0 113.0 131.7 131.6 81.0 80.8
151.3 151.2 151.5 151.5 151.9 151.7
0
I2 ’3’-6
0
0
12'3'-7
523-2 523-3
47.9
C
170.171
152.4 136.4 134.7 123.9 124.4 129.5 117.3 152.3 116.2 162.2 102.7
C
55
152.4 136.5 133.1 125.0 124.7 129.7 117.4 152.4 116.2 160.4 113.3
C
55
152.3 136.5 132.9 124.0 124.7 129.6 117.5 152.4 116.2 160.2 113.4 152.4 136.5 133.1 124.8 124.6 129.5 117.4 152.4 116.4 160.0 113.9 152.5 136.5 133.3 124.9 124.6 129.6 117.5 152.5 116.4 160.0 114.1
C
C C
5s 55 55
1.533 117.5 114.7 126.1 124.3 129.6 117.0 1.51.2 120.3 131.6 132.5
D
I73
162.5 128.0 127.9 122.9 124.5 128.0 116.3 151.1 118.8
D
22
155.2 117.4 177.1 161.6
X = H, Y = C1
X = CI, Y = H X = Me, Y = H
523-4 J.23-5
X = Y =Me
J.23-6
&)
90.9 165.8 107.9 153.0 104.0 113.6
89.2
41.6
-
g I
523-8
523-9
\
di3 \
523-10 323-1 1 523-12 523-13
0
~
55
163.2 113.7 155.9 122.6 114.5 154.0 155.9 157.4 104.1 158.2 113.9
D
174
33.7 39.6 43.6 39.0
C C C C
35 3s 33
160.9 113.5 151.6 150.2 123.9 135.2 112.9 152.0 105.3 144.9 105.3 160.8 114.7 136.8 150.5 123.0 136.4 112.5 151.2 103.6 141.9 116.9 161.0 113.2 151.6 149.5 125.4 134.0 112.2 151.0 104.7 155.5 101.1
C C C
176 176 176
160.5 160.8 159.6 160.1
101.5 101.2 110.2 107.4
166.1 165.3 164.5 165.3
122.4 122.6 122.4 122.7
123.5 123.6 123.3 123.8
131.9 132.0 131.7 132.3
116.6 116.8 116.6 116.8
154.7 154.9 154.5 154.9
112.4 112.8 112.6 116.8
83.9 92.7 92.3 90.9
35
\
0
356-1 556-2 556-3
C
0
RI = Me, R2 = H, R3 = H RI = R2 = Me, = H R, = R2 = Rs= Me RI = H, Rz = Ph, R3 = Me
%
158.2 11 1.6 1.56.6 120.4 124.3 130.0 117.2 152.3 112.8 155.6 104.0
R l = Me, Rz = H R I = H, R2 = Me RI = Rz = Me
0
556-4 556-5 556-6
556-7 556-8
556-9 556-10 556-11 556-12 556-13
RI = R z = R 3 = & = H KI = Rz = Me, R, = & = H R I = R3 = Me, R2 = & = H K I = R3 = H, RZ = & = Me RI = H, Rz = R3 = & = Me
161.0 161.3 161.0 161.1 161.3
116.2 114.0 114.5 114.6 114.1
140.1 153.1 153.1 139.5 139.7
124.5 122.6 125.1 123.4 124.8
151.0 150.8 151.1 150.9 149.0
114.9 126.7 114.1 126.7 125.7
113.4 114.5 112.3 113.8 112.5
151.0 151.2 150.7 151.3 151.2
111.5 11 1.3 112.4 110.4 110.0
147.5 146.5 158.0 143.9 153.5
104.6 107.5 103.5 116.3 110.7
C C C C C
176 176 176 176 176
Ri = COOMe KI = COOEt Ki = CONHz RI = CSNHz Rj =GIN
155.9 155.9 160.6 158.8
112.0 112.0 113.0 112.7 100.9
145.3 145.2 144.2 144.7 150.4
116.6 116.9 118.8 123.3 114.4
137.2 137.1 137.2 137.0 137.1
113.2 113.2 113.2 113.0 113.5
128.8 128.8 128.4 128.3 129.8
153.6 153.5 153.2 152.5 153.4
135.4 135.3 135.8 135.6 136.2
132.4 132.4 132.3 132.5 133.0
121.6 121.7 121.5 121.1 121.7
D D D D D
163 163 163 163 163
C C
177 178
C C
176 176
?d 0
567-1 567-2
0 R=Me R = CHz-CH=CMe2
+
161.4 111.7 138.9 138.0 131.7 151.5 161.2 111.6 139.0 137.3 132.4 151.6
92.4 152.6 106.6 101.8 92.4 152.6 107.4 101.9
\
0
0
R2
567-3 567-4
RI =Me, Rz = H Ri = H, Kz = Me
161.0 113.3 1.52.5 106.7 151.1 132.6 104.9 149.5 115.7 159.8 102.8 161.5 114.6 144.1 108.0 150.3 134.5 105.6 150.3 114.8 155.3 110.6
g
RzwRI 0
0
Ph -4-0 Ph
578-1 578-2 578-3 578-4
RI = RZ= H R, = CH2-CH=CMe2,R2 = H RI = H, R2 = CH2-CH=CMe2 R, = H, R2 = CH2-CHOH-C(Me2)OH
113.5 143.9 121.8 105.7 150.9 119.2 138.6 120.8 105.5 149.8 113.2 144.0 121.5 119.3 149.0 113.3 123.0 118.2 149.4
159.7 159.8 160.0 159.8 160.1 160.1
114.5 113.6 113.4 113.1 113.8 112.6
133.5 132.4 132.8 132.9
143.7 139.5 139.6 139.4
115.1 115.9 114.8 114.9
119.7 125.8 126.4 126.2
141.1 140.4 141.2 139.5 139.8 140.5
140.3 139.5 138.5 137.6 139.7 137.8
115.0 114.1 116.0 113.7 114.4 115.1
147.6 158.8 147.3 158.0 154.4 158.4
C C C C
loo 100 100 100
C C C C C C
I76 176 176 176 176 176
0
R4
R,
578-5 578-6 578-7 578-8 578-9
159.8 161.0 160.1 160.0
RI = R 2 = R 3 = & = H R, = R2 = & = H, R3 = Me R1 = R2 = Me, R3 = & = H RZ = Rt = Me, R I = & = H R , = Rz = H R3 = & =Me
144.4 144.5 153.5 144.2 144.7 153.6
122.2 121.9 118.4 120.9 121.6 118.0
116.9 115.9 126.5 125.5 114.6 125.5
131.8 133.5 131.6 133.0 134.9 133.3
107.4 103.5 106.1 102.0 111.0 102.3
a. p
Tahle 1 1. I ?Cchemical shifts of linear (K) and angular pyranocoumarins (M), their 3’,4’-dihydro derivatives (L and N) and of coumarins with other annulated six-membered rings (0) 5
R
4
0
/ 4 ’
M W4’ M
K K
Substituents
3’
C-2
C-3
C-4
C-5
C-6
C-7
C-8
160.6 160.0 161.0 160.3 160.6 1613
112.5 131.0 112.4 112.8 112.9 112.5
143.1 137.6 138.5 143.4 143.5 144.0
120.4 120.1 157.6 120.8 121.0 115.4
118.2 118.0 107.4 118.9 118.7 119.6
156.4 154.9 152.9 148.9 149.6 156.5
103.9 103.3 100.8 135.1 134.1 113.5
C-9 C-10
C-2’ C-3‘ C-4’ Soh.
Reference
124.6 124.0 115.8 118.9 119.0 127.1
D C C C C C
84 179 180 44 157 41
159.4 131.0 138.7 115.8 120.3 161.9 116.0 150.2 112.0
77.5 130.6 127.1
C
41
161.0 110.2 139.2 150.5 102.0 157.4 113.7 150.9 103.9
78.1 128.0 115.7
C
181
5-OH H ~ I ~160.8 , 128.5 134.2 147.0 106.4 155.1 115.2 153.2 104.2 K358-1 ~ , ~ - [ C ( M ~ Z ) - C H = C
79.0 129.3 115.8
C
181
159.6 129.1 133.0 150.1 103.8 156.5 112.9 150.2 104.0
77.8 127.8 115.9
C
181
161.3 111.9 164.1 151.5 112.2 153.6 113.3 150.5 104.5
77.1 130.6 115.6
?
182
K-1 K3-1 K5-1 K8-1 K8-2 K8-3 K38-1
K58-1
- (linear pyranocoumarins)
3-C(Me&CH=CHz 5-OMe 8-OMe 8-OCH2-CH=CMe2 8-CH2-CH=CMe2 ~-C~MC?)-CH=CH,. 8-CHz-CH=CMe2 5-OH, 8 M
155.0 154.5 155.7 147.9 148.7 146.9
12.3 12.8 11.3 14.3 12.9 12.1
77.5 77.5 77.5 77.5 76.9 77.5
130.8 129.2 130.6 130.9 130.9 130.4
K358-2 3-C(Me)2-CH=CH2,5-OH
K3458-1 3-@-OH-Ph),4,5-[OMe]Z, 8-CHzCH2Pr‘
I
77.1 130.6 115.6
?
182
112.2 112.3 112.6 112.6 112.9 113.4
75.8 78.4 76.4 76.6 76.7 77.9
32.4 67.6 69.1 69.0 70.1 69.3
21.9 30.4 27.7 27.7 27.9 30.8
C C C C C C
103 44,142,184 104 184 86 86
160.0 132.3 137.3 128.3 114.7 154.8 104.0 153.4 113.6
77.9
76.1
28.7
C
185
160.0 97.1 154.7 105.2 157.3 103.7 155.4 112.8 156.5 103.6 154.9 112.7 156.1 104.8 155.2 13.3
86.6 80.2 81.0 77.4
46.4 71.7 75.4 75.0
28.1 65.0 68.5 71.2
C P C
185 155 155 187
K3458-2 3-Q-OH-Ph), 4,5-[OMeJZ, 8-C€fz-CH=CMez
161.3 111.9 164.1 151.5 112.2 153.6 113.3 150.5 104.5
L-1 L3’-1 L3’-2 L3’-3 L3’-4 L33’-1
161.5 160.5 160.9 160.9 161.2 160.0
L33’-2
3 ’-OH 3’-OCO-CH,Pi 3 ’-OCO-CH=CMe2 3’-OCO-C(Me)=CJ!Me (Z) 3-C(Me2)-CH=CH2,3‘-OH
3-C(Me2)-CH=CH2, 3 ’-01@-SO,Me)-Ph]
112.8 112.3 113.1 113.0 113.4 132.0
143.3 144.2 142.9 142.9 143.2 137.4
138.4 143.7 144.4 143.1
128.2 129.4 128.1 128.5 128.7 128.7
153.7 129.2 129.0 129.5
118.4 118.1 115.8 115.9 115.9 116.0
112.5 122.5 123.8 117.8
lS7.7 156.6 156.2 156.3 156.5 155.5
104.6 103.2 104.3 104.4 104.7 103.9
154.0 153.6 154.1 154.1 154.5 153.3
L54’-1 L3’4’-1 (cisJ-3’,4’-[ OH12 L3’4’-2 (rrarzs)-3’,4’-[ OH12 L3‘4’-3 (tmns)-3’-OAc, 4’-OMe
162.1 160.8 161.4 160.7
L3‘4‘-4 (trans)-3‘-OH, 4’-OCO-CH=CMe2
160.0 113.4 143.3 128.3 117.9 156.4 104.6 155.1
13.0
79.9
73.4
70.3
C
155
’-OCO-CH=CMe2, L3’4’-5 (trans)-_? 4’-OH
159.8 113.5 143.3 128.1 121.1 155.4 104.4 155.0
13.2
78.4
75.7
67.4
C
155
(Z), L3 ’4 ’-6 (tmns)-3’-OCO-C(Me)=CHMe 4’-OH 160.9 113.7 143.2 128.0 121.0 155.2 104.5 155.1 113.2
78.3
76.7
67.7
C
155
160.7 113.8 143.1 129.0 117.1 156.1 104.8 155.3 113.3
77.8
72.7
66.3
C
187
77.9
72.0
67.5
C
155
L3’4‘-7 L3’4’-8
(tmns)-3’-OAc, 4’-OCO-CH=CMe2
(trans)-3’-OCO-C(Me)=CHMc Q,
4’-OAc
10.9 12.8 13.1 13.6
160.5 113.8 142.9 128.8 116.7 156.0 104.8 155.3 113.2
P
B
( ~ ~ . ~ ~ . ~ ) - ~ ’ - O C O - C ( M C )(Z), =CHMC 4’-OCO-CHzPr‘ 160.4 113.9 143.1 128.8 117.2 156.1 104.7 155.4 113.5
78.2
72.1
67.3
C
187
L3’4’-10 (rrans)-3‘,4’-IOCO-CI-I=CMe212 160.7 113.6 143.1 129.2 117.1 156.2 104.7 155.2 113.2
77.9
71.2
66.1
C
187,153
L3’4’-11 (cis)-3’,4’-IOCO-C(Me)=CHMe ( E ) ] 2 160.4 113.3 142.2 127.2 116.0 156.1 104.1 154.9 112.7
77.9
68.7
64.3
C
1
160.6 113.7 143.1 127.5 115.3 156.3 104.5 155.3 113.1
78.0
68.9
66.1
C
1
160.3 113.7 143.0 126.8 114.6 155.9 104.5 155.2 113.1
77.7
69.9
65.9
C
1
159.5 109.3 139.7 152.1 106.8 157.2 114.7
76.5
37.6
25.2
1)
99
L3’4’-9
1.3 ’4’- 12 (cis)-3’-OCO-C(Me)=CHMe ( E ) ,
&
4’-0
0
L3’4’-13
(cis)-
Me0
L584 ’-1
OH
103.6
L583’-2
LS2’4’-1
161.5 110.5 140.6 151.5 107.5 159.3 115.9 153.9 104.7
77.6
38.8
23.5
A
189
161.5 110.9 138.9 151.4 104.1 160.2
96.1 156.3 103.1
74.1
34.9
62.3
C
188
162.0 110.4 138.2 154.2 110.2 158.5
97.2 154.7 105.9
86.5
47.4
36.2
C
185
I
0
R=OH
L53’4‘-2
R = CH2-CO-CHSMe2
162.1 111.1 138.8 152.4 111.5 159.9
99.0 155.1 104.2
8S.2
47.4
30.4
C
124
L53’4’-3
R =CH2-CH=CH-C(Me2)0H 162.1 111.1 138.8 152.5 110.8 160.6 (El
98.2 155.2 104.3
85.6
47.1
34.3
C
124
L53’4’-4
R = CH2-CH=CH-C(Me2)00H( E ) 162.0 111.1 138.8 152.4 110.8 160.7
99.0 155.3 104.3
8.5.6
47.1
34.2
C
I24
LS3’4’-S
R = CH2-CHI-CH(O0H)-CMe2(00H)-C(Me)=CH2 162.1 111.1 138.7 152.5 110.4 161.0
98.9 155.2 104.2
86.1
47.1
26.0
C
124
L53’4’-6
R = CH2-CHOH-CHOH-C(Me2)0H 162.0
11.1 138.7 152.3 110.8 160.3
99.0 155.1 104.2
85.5
47.3
28.1
C
124
L53’4’-7
R = CH2-CO-CHOH-Pt
161.9
11.2 138.6 152.4 112.0 159.5
98.9 154.9 104.2
84.6
48.1
36.5
C
124
161.7
11.6 138.3 151.8 111.1 159.8
99.4 155.1 104.2
85.2
46.9
29.5
C
124
R=
L53’4’-8
4
+
L3584’-1
3,8-[C(Me2)-CH=CHzI2. 159.7 129.8 132.5 158.8 104.0 159.3 113.9 158.9 103.5 5-OH, 4’-ketO 0
R P
79.8
47.8 198.3
C
190
158.3 127.5 133.7 151.7 106.8 156.3 114.0 151.9 103.1 159.2 130.8 133.0 154.5 104.4 156.3 119.9 152.8 105.7 159.6 129.3 133.0 150.6 103.2 155.9 115.9 152.9 103.5
76.3 77.5 77.5
37.6 38.1 37.9
25.1 25.4 25.1
D C C
191 191 191
162.3 111.1 139.1 151.5 107.2 157.9 161.4 110.8 139.9 154.6 110.6 158.4 161.3 110.7 139.8 154.7 110.6 158.5
84.7 85.3 86.2
37.4 38.3 37.0
35.7 36.2 36.5
C P P
I88 188 188
OH
OMe
/
0 ‘
0
0
/
L3584’-2 Ki = R2 H L3584’-3 RI = H, R2 = Me L3584’-4 R, = R2 = Me
x, XI
yfi 0
‘
0
L53’4’5’-1 Xi = X2 = H L53’4’5’-2 X I = OH, X2 = H L53’4’5’-3 XI = H, Xz = OH
0
99.1 154.7 103.2 98.4 155.3 104.8 98.3 155.2 104.7
L
0 P W
160.4 112.2 143.5 127.5 114.6 155.9 108.8 149.8 112.2 159.7 131.5 138.1 127.6 115.3 155.5 108.9 149.4 113.0
77.2 130.4 113.1 77.2 130.6 113.2
C C
44 41
M2'-1
R = CH2-CH2-CH=CMe2
161.3
12.7 144.1 128.0 113.5 156.9 109.3 150.3 112.7
80.3 129.9 115.7
C
192
M2'-2
R = CH~-CH=CH-C(MC~)OOH (E) 161.2
12.9 144.1 128.2 113.4 156.7 109.5 150.3 112.8
79.8 129.4 116.1
C
192
161.5 110.8 139.0 150.3 106.4 150.3 102.5 152.2 103.3
78.2 127.9 115.4
C
84,62
159.3 112.5 154.6 158.0 101.3 164.3 102.0 156.2 107.0
79.8 126.1 115.3
C
194
M3456- 1 3-Q-OMe-Ph). 4,S-[0MeJz, 6-CH2CH2Pr'
161.2 112.1 163.9 154.9 107.8 153.9 119.9 147.4 103.8
77.4 129.7 114.3
?
182
M3456-2 3-(p-OMe-Ph), 4,5-[OMeJ2, 6-CHzCH=CMeZ
161.2 112.1 163.9 154.9 107.8 153.9 119.9 147.4 103.8
77.4 129.7 114.3
?
182
161.0 111.5 143.6 125.9 114.1 157.2 108.8 152.8 111.2 161.4 112.4 143.9 126.1 114.4 157.0 109.6 153.1 112.0
75.1 78.0
31.1 26.4
16.1 15.8
C C
44 103
161.3 160.4 160.2 160.7 160.6
78.1 77.3 77.3 76.9 85.0
685 71.3 72.7 71.8 71.6
24.6 23.5 21.8 2.5.0 24.8
C I)
183,41,158
1)
183 183
-
MS6-1 M4S6-1
4-Ph, 5-OH, 6-COCH2Pr'
R
N-1 N-2 N3'-1 N3'-2 N3'-3 N4'-1 N4'-2
p0 R=Me
R = CH2OH 3'-OH 3 '-0SO3-K' 3'-0-P-D-gl~p 4'-OH 4'-O-p-r)-gIcp
112.5 111.8 111.6 112.7 112.4
143.9 144.8 144.7 144.9 144.9
126.7 126.9 126.8 126.6 127.0
114.2 113.7 113.6 114.2 113.1
156.4 156.2 155.9 159.2 159.2
107.4 107.5 107.6 110.1 11 1.1
153.6 153.2 152.9 153.8 153.9
112.2 1 1 1.7 11 1.7 110.9 1 1 1.6
I) I)
8
I
(angularpyranocoumarin) 3-C(Mez)-CH=CH>
M-1 M3-1
158
183
c
N36-1 N33’-1 N3’4’-1
3,6-C12 3-C(Me2)-CH=CH2,3’-OH (cis)-3’,4’-[OH]z,
154.2 118.4 138.7 125.0 119.3 152.4 110.5 149.7 111.1 160.3 131.2 138.4 126.8 114.2 155.5 116.8 149.4 112.9 160.2 11 1.7 144.5 128.8 113.8 155.7 111.2 153.8 111.7
76.5 78.0 78.8
31.0 68.7 71.1
16.8 26.0 60.1
N3’4’-2 N3’4’-3 N3’4’-4
(cis)-3’-OH, 4’-OMe (cis)-3’-OH, 4’-OEt (cis)-3’-OSO3-K+,4’-OH,
160.6 112.6 143.7 128.9 114.6 159.7 109.6 156.7 112.3 160.7 112.6 143.8 128.8 114.6 156.8 109.7 154.8 112.4 159.9 111.8 144.5 128.9 113.8 155.4 111.2 153.8 111.8
78.8 78.8 77.6
71.0 70.3 76.5
70.6 69.6 58.8
D
44 41 15 1 , I 83, 196,153 195 196 IS8
N3’4 ’-5 (cis)- 3 ’-OH,
159.9 1145 143.3 129.3 113.0 157.0 107.3* 154.3 112.3* 78.6
71.6
63.3
C
198
N3’4’-6
160.0 114.4 143.1 129.2 113.4 156.8 107.8* 154.4 112.6* 78.7
70.6
60.2
c
198
71.6
62.9
C
195,198,199
70.6
59.4
c
195,I98
160.6 112.5 143.9 128.7 114.5 156.1 110.7* 156.1 112.3* 77.8
71.6
60.2
C
195,198,151
N3’4’-10 (cis)-3’-OCO-CH=CMe2,4’-OAc 1.59.8 113.2 143.1 129.1 114.4 156.4 107.5*154.0 112.5* 77.6
70.7
59.6
C
195,198
N3’4’-ll (cis)-3’-OCO-C(Me)=CHMe(Z), 4 ‘-OEt 160.7 112.7 143.4 128.4 114.3 156.1 108.2* 154.8 112.3* 76.6
70.6
69.4
c
153,200
69.9
69.5
C
196
C C C
c
4‘-OCO-C(Me)=CHMe (Z) (cis)-3’-OAc,
4’-OCO-C(Me)=CHMe (2) N3’4’-7
(cis)-3’-OH,4’-OCO-CH=CMe2
N3’4’-8
(cis)-3’-OAc, 4’-OCO-CH=CMe2
N3’4’-9
(cis)-3’-OCO-CH=CMe2,4‘-OH
1.59.6 113.0 143.3 129.2 114.6 157.0 107.2’ 154.4 112.4’ 78.7
159.8 113.2 143.2 129.1 114.3 156.6 107.4” 153.9 112.5* 71.3
N3’4’-12 (cis)-3’-OCO-CH=CMC~. 4’-OEt
160.8 112.4 143.6 128.5 114.6 1.56.2 108.6* 155.0 112.8* 79.7
-
8
I
I
z
h)
N3’4’-13 (cis)-3’-OCO-C(Me)=CHMeQ, 4’-OAc,
160.4 113.4 143.8 129.8 115.0 154.6 107.7* 157.2 113.0* 77.6
153,200-202
70.8
60.5
C
159.8 114.4 143.2 129.1 113.3 156.9 l07.5* 154.5 112.5* 78.2
70.0
60.8
C
198
N3’4’-15 ( ~ ~ s ) - ~ ’ , ~ ’ - [ O C O - C H ~ P ~159.7 ‘ J ~ 113.2 143.2 129.3 114.4 156.6 107.3* 154.0 112.5* 72.2
70.2
60.4
C
165,195
N3’4’-16 (cis)-3’-OCO-CH2Pr’,
70.2
60.2
C
195
N3’4’-14 (cis)-3’-OCO-C(Me)=CHMe(2). 4’-CXOPr‘,
159.7 113.2 143.1 129.2 114.4 156.6 107.3* 154.0 112.5* 72.2
4 ’-OCO-C(Me)=CHMe (2)
N3’4’-17 (cis)-3’-OCO-C(Me)=CHMe(Z), 4’-OCO-CH2Pr‘,
159.7 113.2 143.1 129.2 114.4 156.6 107.4* 154.0 112.5* 77.4
70.5
60.4
C
153,195,200
N3‘4’-18 (cis)-3’-OCO-CH=CMe2,
159.8 113.2 143.1 129.1 114.4 156.7 107.6* 154.1 112.6* 77.5
70.4
59.6
C
196
59.9
C
44,153, I96 198,200
4’-OC0-CH2Pr1
N3’4’-19 (cis)-3’,4 ’-[ OCO-C(Me)=CHMe (Z)]2
11.8 144.8 129.0
14.1 156.2 109.0* 150.8 111.9, 77.1
70.1
159.8
13.3 143.1 129.1
14.3 156.8 107.8* 154.1 112.5* 77.5
70.5
59.6
C
153
N3’4’-21 (cis)-3’,4’-[OCO-CH=CMe212 159.8
13.2 143.2 129.0
14.4 156.8 107.6* 154.1 112.5* 77.3
69.5
59.8
C
I96
160.1 114.6 143.4 129.2 113.1 156.9 107.4* 154.1 112.6* 77.5
70.2
59.9
C
203
160.1 114.5 143.2 129.1 113.2 156.9 107.9* 154.1 112.6* 77.8 4’-OCO-CH=CH-(Wans)-@-OH-Ph),
70.4
60.0
C
203
78.2
58.1
C
195
160.3
N3 ’4’40 (cis)-3’-OCO-C(Me)=CHMe (Z), 4’-OCO-CH=CMe2,
N3’4’-22 (cis)-3’-OCO-C(Me)=CHMe(Z), 4’-&OkH=CH(trak)-’(4‘’-OH-Ph)
N3’4’-23 (cis)-3’-OCO-CH=CMe2,
N3’4’-24 (cis)-3’-o-p-D-gkp,4‘-OH
160.3 11 1.9 144.6 129.1 113.8 155.5 110.6* 154.0 112.0* 77.6
N3’4’-25
OH
N3’4’-26 (trans)-3’,4’-[ OH12,
159.0 113.2 144.4 131.6 115.4 157.7 109.9* 155.4 114.0* 80.4
72.9
76.7
A
I89
161.5 112.0 144.4 128.4 114.8 156.4 11 1.8’ 154.3 112.5* 79.5
74.8
66.4
C
151,183, 196,197
78.4 78.4 76.4 77.4 77.3
74.1 72.7 70.9 72.0 73.9
70.7 71.4 70.3 63.9 63.2
C C C C C
44,195 196 195 195 151.195
160.5 113.8 143.8 129.6 114.9 154.9 107.1* 157.2 112.9* 77.6 159.9 113.2 143.2 129.1 114.4 156.6 107.5* 154.1 112.5* 77.6
71.6 70.2
63.7 59.6
C C
202 165
160.0 113.2 143.2 129.0 114.4 156.8 107.6* 154.1 112.6* 77.7
69.4
59.8
C
165
160.3 1 1 1.7 144.6 128.4 114.0 155.8 110.3* 154.3 11 1.8* 77.7
80.6
62.1
C
183
76.4 184.3
C
151,153 200.204
N3’4‘-27 (trans)-3’-OH, 4’-OMe, 160.8 N3’4’-28 (trans)-3’-OH, 4’-OEt, 160.9 N3’4’-29 (truns)-3‘-OAc,4’-OMe, 160.7 N3’4’-30 (tr~ns)-3 ’,4’-[OAcJ*, 159.9 N3’4’-31 ( ~ ~ u ~ . ~ ) - ~ ’ - O C O -4‘-OH C H ~ P ~160.6 ‘, N3‘4’-32 (trans)-3’-OCO-C(Me)= CHMe (0, 4‘-OAc,
N3’4’-33 (c~s)-~‘-<XO-CH~P~‘, 4’-OCO-CH=CMe2,
112.8 112.8 112.8 113.2 1 12.6
143.6 143.6 143.6 143.3 143.9
128.6 128.5 128.6 129.2 128.5
114.6 114.6 114.5 114.5 114.8
156.1 156.3 155.9 156.5 156.0
108.9* 155.1 109.3* 155.1 108.1* 154.8 106.8* 154.5 110.8* 154.5
112.6* 112.6* 112.4* 112.6* 112.4*
N3’4’-34 (C;~)-~’,~’-(OCO-CH=CM~~)~ N3’4’-35 (trans)-3’-O-P-l~-glcp,4’-OH
N3’4’-36 3’-(R)-OCO-C(Me)=CHMe(Z), 159.4 113.9 143.0 134.6 114.8 161.8 108.1* 153.5 112.7* 82.3 4 ’-keto
& \
034-1 034-2 034-3 034-4 034-5
0 0 R=H R = CHZO-Ph R = CH2O-(+CI-Ph) R = CH,O-@-CI-Ph) R = CH20-(o.p-Mel-Ph)
156.6 156.7 156.7 156.6 156.7
137.2 139.9 140.0 140.0 139.9
122.8 125.3 125.3 125.1 125.4
121.9 125.0 124.7 124.7 125.1
124.4 124.7 125.1 124.9 124.6
128.8 128.8 128.9 128.9 128.7
116.6 117.3 117.4 117.4 117.2
149.7 149.5 149.6 149.6 149.5
116.9 117.0 117.1 117.0 117.2
65.8 64.9 6S.O 649 64.9
126.7 125.9 125.7 126.7 125.3
117.8 130.3 129.9 130.1 130.7
C C C C C
55 55 55 55 55
160.6 159.7 159.9 159.9 160.0
101.5 101.4 101.2 101.1 100.9
160.1 161.7 161.7 161.9 161.8
122.6 123.0 123.1 123.1 123.1
124.0 124.0 124.1 124.1 124.1
132.2 132.5 132.6 132.7 132.7
116.7 116.4 116.5 116.5 116.5
153.2 153.2 153.2 153.2 153.2
114.9 114.7 114.7 114.7 114.7
67.3 67.2 67.4 67.4 67.5
117.0 113.0 112.9 113.1 113.4
119.1 129.4 128.7 128.5 127.6
C C C C C
55 55 55 55 55
161.8 100.0 159.4 115.5 122.6 123.1 131.9 153.2 116.7
83.1 117.2 125.0
C
63
034- 12
163.1 100.4 158.9 122.2 123.7 131.2 116.5 152.3 115.9
79.8
17.4
C
102
034-13
160.9
99.9 159.0 122.6 123.7 132.1 115.4 153.2 116.6
83.1 125.0 117.3
C
59
160.5
99.4 1S8.9 122.6 123.9 132.1 116.8 153.2 117.3
83.2 125.0 117.3
C
59
161.0
99.9 159.0 122.6 123.4 132.1 115.4 153.0 116.7
83.2 123.9 117.2
C
59
&k?\
034-6 034-7 034-8 034-9 034-10
0
R=H
0
R = CHZO-Ph
R = CH?O-(o-CI-Ph) R = CHzO-(o,p-Cl-Ph) R = CHZO-(o-NOl-Ph)
d: \
034-1 1
0
0
31.4
R=
034- 14 R=
034- 15
034-16
034-17 034-18
R=
C
59
C
C
57 57
C
57
163.1 100.8 158.9 122.4 123.6 131.1 116.6 152.5 116.1
C
57
99.2 158.0 122.4 123.7 131.2 116.4 152.0 116.0 99.9 158.2 122.6 123.7 131.2 116.4 152.3 115.9 99.4 157.8 122.6 123.7 131.2 116.5 152.3 115.9
C C
57 57
160.7 100.0 159.0 122.6 123.6 132.1 115.5 153.2 116.8
rn 0
0
R I = Et, R2 = H R I = El, Rz = Me
CH,OAc
R , = Mc, R, = 034-19
163.2 99.8 159.1 122.3 123.6 131.2 116.5 152.3 115.9 163.1 100.8 159.9 122.3 123.6 131.2 116.4 152.4 116.2
&
163.1 100.4 158.9 122.2 123.7 131.2 1 16.5 152.3 115.9
R, =Me, K,
=
83.2 125.1 117.3
CH20Ac
034-20
034-21 034-22 034-23
K, = R z = H RI = Me, R l = Pr' R I = Pr', R2 = M e
-OA0
163.5 163.7 163.7
034-24
3
163.2
99.4 158.0 122.4 123.5 131.1 116.4 152.3 115.9
C
57
163.6
99.0 160.7 122.2 123.6 131.6 116.5 152.6 115.7
C
57
163.1 100.2 159.2 122.3 131.1 131.1 116.5 152.5 116.5
C
57
162.0 100.0 156.0 122.3 124.1 132.0 116.7 152.6 114.6
C
57
186.2 127.0 132.2 212.5 134.5 130.2
C C
12 12
\
034-26
0
0
"
H
034-27
034-28 034-29
X=S
136.0 128.5 135.0 127.9
K
034-31
= Pent"
160.0 114.8 150.3 152.6 105.7 142.2 108.3 155.6 112.4 161.0 113.0 149.5 152.5 106.2 147.4 107.4 155.8 112.0
C C
206 206
160.9 100.5 153.7 150.5
99.8
I>
I 15
93.3 149.6 103.3
C
201
OMe
96.4 163.0
93.1 154.5
AcO
056-1
056-2
Me0
Me0
161.2 112.3 137.6 139.5 128.9 152.3
R
R = CH=CH-C(Me2)00H (Z) 056-3 K = CH=CH-C(Me2)00H ( E )
161.1 116.5 138.5 149.4 110.5 155.4 112.5 152.2 106.4 161.1 116.6 138.5 149.3 110.9 157.2 111.3 152.7 106.3
71.6 129.4 113.2 71.4 129.3 112.8
C C
62 62
95.7 155.4 103.2 91.8 156.2 103.6
80.0 126.0 117.2 79.9 126.2 117.2
C C
209 209
I10
056-4 056-5
K = CH=CH-C(Me2)0H K = CH=CH-C(Me2)00H
162.8 110.2 139.6 151.3 106.3 156.4 161.6 111.5 138.6 150.6 106.6 158.5
-I
056-6
160.7 110.7 151.7 154.7 106.1 153.0 104.0 153.3 106.0
056-7
1.59.7 113.3 1.51.9 154.4 I 1 1.9 153.6 114.1 152.6 108.0
056-8
158.0 112.6 155.3 156.4 105.7 161.5 103.8 155.9 102.4
056-9
Rzfm
160.3 116.7 154.3 130.4 131.5 133.8 117.9 154.9 114.6
0 '
0
R,
067- I 067-2
067-3
77.6 116.1 130.9
C
52
C
52
C
164
C
172
C C
175 186
C
193
0
R , = OMe, R2 = C(Me)=CH2 K, = H, RZ = C(Me2)OH
160.5 114.5 143.4 108.9 141.1 140.3 142.5 139.2 112.7 161.3 114.4 143.0 114.1 140.8 146.8 104.8 112.8
160.9 114.4 142.7 114.3 139.9 146.7 105.0 149.3 113.2
75.6 78.7
67.6 65.4
067-4
160.2 112.1 138.1 i6n.s
101,s161.3
97.4 162.1 103.5
067-5
162.6 130.3 138.2 122.9 124.9 160.2
96.8 154.4 112.7
75.3
74.2
162.4 130.3 137.9 123.0 124.3 159.8
96.7 154.4 112.7
84.0
87.2
C
188
32.0
C
197
30.5
C
197
C/CCI C/CCI
64 64
C C
142,205.208 210
'I()$/,,.
067-6
_o
\
R,
'"
U 067-7 067-8
078-1 078-2
Kl = H, K2 = Me Kl = COOEt, K2 = H
X=H X =OMc
162.2 107.8 i s m 121.4 117.6 145.4 106.4 152.7 108.6 157.5 108.5 148.7 126.8 118.7 148.1 107.7 153.5 105.9
160.9 114.1 143.8 100.2 145.8 139.0 132.4 136.8 111.6 160.6 113.2 138.9 128.8 138.0 141.2 142.4 139.7 107.1
65.9 65.8
79.0 78.9
xlm 0
078-3 078-4 078-5 078-6
078-7
'
0
80.2* 75.6 77.8* 75.5*
P C P C
21 1 21 1 212 212
160.2 114.2 143.5 101.0 145.8 137.0 133.1 145.7 111.8
77.0* 75.0*
C
213
160.7 160.6 160.8 160.7 160.4 160.8
81.5* 81.1* 79.9* 79.7* 76.7*
RI = R2 = H, X I = OMe, X2 = H 160.7 RI = K2 = Ac, X I = OMC,X2 = H 160.4 R I = R ~ = H , X I = H , X ~ = O M C160.4 KI= RZ= Ac, X I = H, X2 = OMe 160.2
Me0
113.8 114.3 113.6 113.9
144.4 143.5 144.3 143.8
101.2 100.8 119.8 119.9
146.2 145.7 113.2 113.9
138.1 136.3 147.6 146.5
133.2 132.1 138.4 133.3
139.4 140.7 149.2 150.6
111.8 11 1.8 113.6 113.7
McoP
y f - 0 RlO OR*
77.1* 76.0 79.9* 77.2*
0Me 0Ac
0
078-8 078-9 078-10 078-11 078-12 078-13
0
0
-.
XI
RI = X I = X2 = H, R2 = Me RI = XI = X2 = H, R2 = AC RI = X2 = H, R2 = Me, X I = O H R I = X2 = H, R2 = M e , X I = OEt RI = X I = OAc, R2 = Me, X2 = H R1 = X I = H, R2 = Me, Xz = OMe
113.9 114.2 113.6 113.3 114.4 113.7
144.4 143.6 144.5 144.5 143.5 144.5
101.3 100.6 101.1 101.0 100.5 101.0
146.3 146.0 146.3 146.3 145.8 146.3
138.5 137.7 138.4 138.2 136.9 138.1
132.6 132.4 133.0 132.0 131.7 132.5
139.3 140.9 139.3 139.2 140.8 139.2
112.0 111.9 11 1.9 112.2 111.9 I 1 1.9
74.5* 74.2* 77.5* 77.3* 75.1*
214 214 21 1 125 21 1 215
0
p 0
0
LNJ
078- 14 078-15 078-16 078- 17 078- 18
078- 19
078-20
R
R = Ph
K = P-Cl-Ph
R = o-OMe-Ph R = m-Me-Ph K = CH2-Ph R = CHZ-CH2-Ph
162.9 162.8 162.8 162.8 162.9 161.8
118.2 118.5 119.5 120.1 117.4 117.1
C C C C C C
216 216 216 216 216 216
141.9 130.9
C
12
113.9 152.2 125.9 100.6 144.4 123.7 148.7 114.1 152.1 125.5 100.4 143.8 120.9 147.9 113.8 151.8 124.2 101.2 142.7 121.6 147.9 114.2 52.2 125.9 99.8 142.9 122.7 149.0 113.8 52.2 125.4 100.5 144.6 121.5 150.3 113.9 51.8 124.8 100.2 143.6 120.8 148.5
206.4 136.6
31.3
Table 12.
I T chemical shifts of dicournarins directly connected by one bond (P), over more than one bond (Q), by rings (R). In cases of unsymmetrical dicournarins the second data row corresponds to the coumarin moiety with dashed atoms numbers
C-2
Substituents
&
C-3
C-4
C-5
C-6
C-7
C-8
C-9 C-10
othercarbons
Soh.
Reference
C
217
\
0
0
0
P38’-1
4,4’,7‘-[OMe]l, 6,6‘-Mez, 7-OH
164.0 162.3
77.2 170.3 139.3 116.2 159.4 101.1 155.9 109.3 87.7 168.2 138.5 111.5 160.3 109.3 153.9 108.4
P38’-2
4,4’,7,7’-[OMel4, 5-CHzOH, 5’-Me
161.1 161.4
97.9 166.7 144.0 110.6 162.2 98.8 155.1 107.2 87.6 169.3 139.1 111.3 159.6 107.7 153.4 107.7
D
218
”% \
0
0
P66’-1
4,4’- [OMe]?, 5,s ’-Me2, 7,7’-[ OH]*
166.0
87.5 172.6 139.3 124.4 160.5 101.6 157.3 108.3
M
219
P66’-2
4,4’,7’-[OMe]3, S,S’-Mez, 7-OH
162.4 162.4
88.1 170.5 138.3 123.9 161.4 97.8 157.4 108.8 87.7 170.5 138.3 123.2 159.1 101.0 156.8 108.2
M
2 17,219
P66’-3
4,4’,7,7‘-[OMe14, 5,5’-Mez
163.0
87.9 170.0 137.2 123.4 160.1
C
219
97.4 156.2 108.1
0
D
220
C
220
86.5 169.6 136.8 119.0 158.8 100.2 155.3 106.2 86.8 169.6 137.3 115.5 158.0 109.3 153.7 106.2
D
22 1
161.4 161.6
87.3 169.4 137.1 118.6 158.6 100.3 155.3 106.2 86.9 169.3 137.7 110.7 158.9 11 1.3 152.7 107.6
D
22 1
4,4‘,7,7’-[OMe14, 5,5’-Me?
161.5 161.3
87.4 169.3 136.7 119.4 160.0 97.6 155.9 107.4 87.7 169.2 137.9 111.4 158.8 1 10.4 152.6 107.7
D
22 1
7,7’-[OH]2
160.9 111.3 145.4 129.4 113.0 159.9 107.3
53.8 I I 1.6
11
222
PS8’-2
7.7’4 OMeI2
161.0 113.0 143.1 130.4 107.3 159.5
53.0 112.7
C
223
PS8’-3
7-OH, ~’-O-CC-I,-~~C~I~I~160.7 11 1.3 145.1 129.6 112.9 159.7 106.6 53.6 1 1 1.4 160.5 113.2 145.4 129.6 111.6 157.4 110.2 153.0 113.7
D
222
P88‘-4
4,4‘-[OMeJ2,5.5’-Me2, 7,7’-[OH 12
169.4 161.4
86.3 153.7 136.8 115.3 158.3 105.9 161.4 105.5 86.3 169.4 136.8 111.3 159.4 108.4 153.3 107.4
1)
rj
224 22 1
PSS’-5
4,4’,7,7’-[OMej4,5,5’-Me2
169.6 162.8
87.7 153.3 138.3 111.3 159.4 108.4 162.8 107.4 87.7 169.4 138.3 111.3 159.4 108.4 153.3 107.4
C C
224 221
P68’-1
7-OMe, 7’-OH
160.2 112.5 144.1 128.6 118.2 160.6 99.4 155.3 11 1.4 160.1 112.7 144.7 131.1 111.0 159.0 1 1 1.2 152.9 11 1.9
P68’-2
7,7’-[0Me],
160.7 113.1 143.6 128.6 117.9 161.2 99.6 156.0 113.0 160.2 113.4 143.5 130.8 107.7 160.8 112.3 152.8 113.8
P68’-3
4,4’,7,7’-[ OMeI4,5,5’-Me2, 7-[OHl2
163.6 161.7
P68’-4
4,4‘,7’-[OMe]l, 5,S’-Me2, 7-OH
P68’-5
PS8’-1
99.2
0
W
P88’-6
4,4’- [OMeIz, 5,5 ‘-Mez, 7,7’-[OA~]z
169.0
89.6 153.0 138.4 121.8 150.4 112.3 161.9 111.2
C
224
R
crr.T-E \
0
0 0
0 ‘
Q33’-1 Q33’-2
K=H K = OCOEt
166.1 102.7 169.8 124.6 124.9 132.4 116.8 153.3 120.2 167.9 102.3 164.6 124.2 124.7 132.8 116.4 152.0 116.2 167.8
D C
28 29
Q33’-3
R=Me
168.8 106.8 164.1 124.0 124.6 132.4 116.4 152.1 117.0 167.3
C
29
433’-4 Q33’-5
R = CHzOMe K = CHzPh
168.2 104.0 164.7 124.0 124.7 132.6 116.4 152.1 116.6 168.8 105.4 164.7 124.0 124.6 132.4 116.3 152.1 117.0 167.5 164.2 123.9 151.9 116.1
C C
29 29
Q33’-6
R
169.1 105.8 164.8 124.1 124.6 132.4 116.4 152.3 117.1 167.3 105.5 164.1
C
29
433’-7
R = CH~CHZ-SMC
168.8 105.l 164.9 123.9 124.7 132.5 116.4 152.2 116.9 167.5 164.2
C
29
Q33’-8
R =PS
169.0 105.9 164.5 124.0 124.6 132.3 116.3 152.2 117.1 167.3 105.7 164.1 123.9
C
29
Q33’-9
R = Bu“
169.0 106.0 164.5 124.0 124.6 132.3 116.4 152.2 117.1 167.2 105.8 164.0 123.9
C
29
Q33’-10 Q33’-11
R = COOEt R = Ph
165.3 103.4 165.7 125.0 124.9 133.3 117.3 153.3 118.4 169.1 105.5 165.3 124.3 124.8 132.8 116.5 152.3 116.5 166.9 104.0 164.5
D C
28 29
Q33 ’-12
R = O-Cl-Ph
168.6 105.1 164.7 124.3 124.8 132.8 116.5 152.2 116.5 168.2 104.8
C
29
= Et
Q33’-13
R = p-Cl-Ph
169.1 105.2 165.9 124.4 124.9 132.7 116.6 152.3 116.6 166.7 103.7 164.5
C
29
Q33’-14
R = P-NOz-Ph
168.9 104.7 166.2 123.8 124.4 133.3 116.7 152.4 116.7 167.0 103.3 164.7
C
29
Q33’-15
163.2 126.9 142.4 124.3 115.0 155.2 135.7 148.7 113.9 161.1 121.9 138.1 124.2 114.9 155.7 135.6 148.4 114.2
P
225
433’-16
162.0 119.5 154.4 122.4 124.2 131.5 116.7 152.1 116.4 162.0 119.5 154.1 122.4 124.2 131.5 116.7 152.1 116.4
C
60
159.2 97.6 165.7 137.7 127.6 131.6 114.6 153.9 114.9 159.2 98.1 165.6 137.7 127.6 131.6 114.6 153.9 114.5 (no assignment of the data sets to the individual coumarin moieties)
C
24 1
OMe
HO
Q33‘-17
Me
Q37’-1 XI= H, X2 = OH
57.3 35.6 131.4 129.8 113.8 161.0 102.4 153.8 115.2 60.4 14.1 144.4 130.2 113.6 160.1 104.1 155.4 114.6 56.8 38.9 129.2 129.3 119.2 152.3 110.3 152.4 17.2 160.3 114.4 144.4 130.2 114.2 159.6 104.8 155.4 15.0
D
226
D
226
Q37’-3 X,= OMe, Xz= OH
159.8 135.7 131.0 109.4*145.8’ 150.4 102.8*147.5 10.2 160.1 113.8 144.1 129.9 113.6$157.1 104.0$155.1’ 14.4
D
227,228
437’-4 X I = OMe, Xz = OAc
160.3 140.0 129.3 105.7 149.0 155.4 112.0 142.1 16.6 158.7 114.2 142.8 126.7 109.0 156.5 115.2 145.7 15.3
C
228
Q37’-5 XI= OMe, X2 =7-OCO-C(Me)OH-C00-(6”-0-(3-~-glcp) 159.6 137.3 130.1 109.8 146.5 149.0 103.1 146.9 112.4 160.1 114.1 144.2 129.9 113.8 156.9 104.6 155.2 114.7
D
229
Q37’-6 X I = OMe, Xz=[O-~-D-~~~~-(~”-OCO-CH~-C(M~)OH-CH~-COOH)] 156.7 137.0 129.7 109.5 146.2 148.7 102.9 146.6 112.2 159.9 113.9 144.0 129.9 113.5 159.3 104.3 154.9 114.5
D
222
C
230
Q37’-2 XI H, X2 = OAC 7
MeO
Q38’-1
161.7 112.3 144.2 129.7 129.6 160.9 97.9 154.4 111.8 161.0 113.0 143.9 126.3 107.5 159.7 114.8 152.3 113.2
Q58’-1
160.0 108.9 139.1 148.0 104.1 155.6 112.1 152.7 101.8 160.0 111.1 145.0 128.0 112.8 159.3 110.1 152.8 111.3
D
23 1
Q68’-1
159.9 108.3 139.9 150.6 110.6 157.7 113.3 151.8 103.2 160.4 109.4 140.2 147.7 105.8 153.8 112.0 152.6 103.2
D
181
X
Q78’-1 Q78‘-2
X
55.3 114.0 53.0 112.8
A
232
161.4 112.3 144.1 128.2 108.0 157.3 114.0 53.8 112.4 160.0 113.4 143.8 126.3 113.8 160.0 117.8 153.8 113.8
C
232
161.1 112.8 144.1 128.5 118.4 155.7 117.4 152.8 112.7
C
233
161.4 113.5 145.6 128.2 114.6 158.7 114.8 161.0 112.6 145.9 129.9 115.3 163.5 114.3
X=OH = OMe
0
Q88’-1
0
M eO
Q88’-2
0
0
M AcO
90.1 161.7 107.2 154.0 103.3 90.6 162.0 107.4 154.6 103.7
C
106
161.8 113.1 143.7 127.6 108.2 153.0 116.6 160.5 113.0
C
234
160.6 112.7 143.8 126.7 106.3 153.4 115.0 160.3 112.5 160.4 112.4 143.7 127.5 106.6 153.3 115.0 160.2 112.4 (no assignment of the data sets to he individual coumann moieties)
C
234
161.3 110.3 138.5 155.1 161.5 110.6 138.7 155.6
e
OAc O8.
\ /
Q88’-3
X
0 /
0
I
Q88’-4
OAc
Me0
WO 8’
AcO
Q88’-5
Q3”3”’-1 Q3”3”’-2
OAc
X=OH X =Me
160.4 112.5 143.7 127.0 106.7 153.2 115.9 160.3 112.5
C
234
159.2 130.1 132.4 158.6 105.2 158.9 114.3 158.8 104.1 159.2 130.9 132.6 151.2 112.4 154.1 118.8 153.4 108.2
C C
190 235
01I
I10
R-1
011
158.8
R-2
R-3
R-4
0
0.6
07.3 110.6 110.6 148.0 139.2 136.6 107.3
1)
236
159.1 115.7 143.3 123.1 108.6 156.9 115.7 148.3 115.7 159.1 103.3 160.6 126.3 114.2 159.4 104.3 155.4 106.6
C
237
162.2 100.2 156.4 122.4 124.1 132.1 116.9 152.8 114.9
C
51
162.4 113.2 144.8 127.8 131.5 160.5 162.3 113.0 144.5 125.0 130.8 160.2
C
85
99.1 154.0 112.5 98.6 154.0 112.1
R-5
R=H
159.9 109.5 139.4 148.6 107.4 155.2 113.9 153.4 102.7 159.8 109.9 139.8 152.1 106.6 155.0 113.3 153.7 103.8
D
238
R-6
R = C(MeZ)CH=CH*
158.6 127.5 132.8 148.2 107.5 154.0 112.8 152.4 102.7 158.5 127.9 133.3 151.8 106.8 154.3 113.2 152.7 103.8
D
238
?
239
R-7
159.7 128.7 132.3 148.1
95.8 154.2 113.1 152.8 102.8
77.8
73.2
26.0
Tahle 13.
chemical shifts of tricournarins (S)
Substituents
c-2
c-3
c-4
c-5
C-6
c-7
C-8
c-9
c-10 Soh. Reference
S-37’8”-1
XI = X2 = OH
156.8 159.9 160.4
135.0 113.8 111.0
131.3 143.9 144.9
129.1 129.8 128.8
113.2 113.2 112.8
158.8 159.5 159.4
106.9 103.9 107.1
151.2 154.9 153.4
110.8 114.3 11 1.3
D
226
S-37’8”-2
XI = OH, X2 = 0-a-L-rhamp
156.5 159.9 160.1
135.0 113.8 113.0
131.1 143.9 144.6
129.1 129.9 129.4
1 1 1.4 113.0 113.3
158.6 159.4 157.0
106.6 104.2 109.6
150.9 154.8 152.6
110.6 114.3 113.5
D
226
S-37’8”-3
X ’ = OH.
156.6 159.9 160.1
135.1 113.8 112.8
130.9 144.0 144.6
128.9 129.8 129.2
11 1.3 112.8 113.1
158.4 159.4 157.4
106.6 104.2 109.5
151.0 154.9 152.6
110.7 114.3 113.1
D
222
S-37’8”-4
X’ = OAc, x2= apiof triacetate
157.0 160.8 161.1
140.3 115.5 115.1
127.2 143.5 143.7
128.5 129.8 130.1
120.3 114.8 111.6
151.2 159.4 156.9
I 14.0 106.6 109.8
150.5 156.0 153.5
117.2 115.8 115.0
C
222
S-37’8”-5
XI = OAc, X2 = O-a-L,-rhamp triacetate
156.3 159.9 160.3
138.5 113.9 1 14.5
129.4 144.4 144.8
129.8 130.4 130.8
120.2 114.2 1 10.9
150.6 159.3 155.6
113.3 104.5 108.3
150.1 155.4 152.5
117.3 115.1 114.2
D
226
1074 References to Section 6 (Tables 5-13) 1.
2. 3. 4. 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. 33. 34. 35.
E. Wenkert, B. L. Buckwalter, I. R. Burfitt, M. J. Gasic, H. E. Gottlieb, E. W. Hagaman, F. M. Schell, P. M. Wovkulich and A. Zheleva, in: G. C. Levy (Ed), Topics in Carbon-13 NMR Spectroscopy, Vol. 2, Wiley, New York, 1976, pp. 81-121. H. Gunther, J. Prestien and P. Joseph-Nathan, Org. Magn. Reson., 1 (1975) 339-344. S. A. Sojka, 7. Org. Chem., 40 (1975) 1175-1178. N. J. Cussans and T. N. Huckerby, Tetrahedron, 31 (1975) 2587-2590. L. Ernst, J. Magn. Reson., 21 (1976) 241-246. C.-J. Chang, H. G. Floss and W. Steck, J. Org. Chem., 42 (1977) 1337-1340. K. K. Chan, D. D. Giannini, A. H. Cain, J. D. Roberts, W. Porter and W. F. Trager, Tetrahedron, 33 (1977) 899-906. M. H. A. Elgamal, N. H. Elewa, E. A. M. Elkhrisy and H. Duddeck, Phytochemistry, 18 (1979) 139-143. H. E. Gottlieb, R. A. de Lima and F. delle Monache, J. Chem. Soc, Perkin Trans. 2, (1979) 435-437. H. Duddeck, M. H. A. Elgamal, F. K. Abd Elhady and N. M. M. Shalaby, Org. Magn. Reson., 14 (1980) 256-257. O. A. Subbotin, P. L Zakharov, V. A. Zagorevskii and D. A. Zykov, Khim. Prir. Soedin., (1975)476-479. L W. J. Still, N. Plavac, D. M. McKinnon and M. S. Chauhan, Can. J. Chem., 54 (1976) 280289. W. V. Turner and W. H. Pirkle, J. Org. Chem., 39 (1974) 1935-1937. R. D. Lapper, Tetrahedron Lett, (1974) 4293-4296. L. F. Johnson and W. C. Jankowski, Carbon-13 NMR Spectra, No. 333. Wiley, New York, 1972. A. Patra, A. K. Mukhopadhyay, A. Ghosh and A. K. Mitra, Indian J. Chem., 17B (1979) 385387. A. Chatterjee, S. Sarkar and J. N. Shoolery, Phytochemistry, 19 (1980) 2219-2220. K. Jewers and K. A. Zirvi, Planta Med., 33 (1978) 403-406. M. Baiwir and G. Llabres, Spectrochim. Acta, 39A (1983) 693-698. Sadtler Standard Carbon-13 NMR Spectra, No. 2157, Sadtler Research Laboratories, Philadelphia (1977). N. J. Cussans and T. N. Huckerby, Tetrahedron, 31 (1975) 2719-2726. A. K. Mitra, A. K. Mukhopadhyay, S. K. Misra and A. Patra, Indian J. Chem., 21B (1982) 834-837. L. M. Ryzhenko, V. L Labunskaya, B. F. Ryzhenko and A. D. Shebaldowa, Khim. Geterosikl. Soedin., 12 (1988) 1611-1614 [Chem. Heterocyclic Comp. (Consultants Bureau) 12 (1988) 1330-1333]. T.Liptaj, P. Hmciar, A. Gaplovsky and J. Donovalowd, Chem. Papers, 44 (1990) 71-76. H. Duddeck and M. Kaiser, Org. Magn. Reson., 20 (1982) 55-72. N. J. Cussans and T. N. Huckerby, Tetrahedron, 31 (1975) 2591-2594. A. Rabaron, J.-R. Didry, B. S. Kirkiacharian and M. M. Plat, Org. Magn. Reson., 12 (1979) 284-288. B. S. Kirkiacharian, A. Rabaron and M. Plat, C. R. Acad. ScL, Ser. C, 284 (1977) 697-700. O. Convert, C. Deville and J.-J. Godfroid, Org. Magn. Reson., 10 (1977) 220-223. E. G. Frandsen and J. P. Jacobsen, Org. Magn. Reson., 10 (1977) 43-46. P. Joseph-Nathan, J. Hidalgo and D. Abramo-Bruno, Phytochemistry, 17 (1978) 583-584. M. Nagesam and M. S. Raju, Indian J. Chem., 32B (1993) 705-708. A. Patra, A. K. Mukhopadhyay and A. K. Mitra, Indian J. Chem., 17B (1979) 1-2. K. Jain, M. S. Alam, M. KamU, M. Eyas, M. AH, Phytochemistry, 30 (1991) 3826-3827 (questionable data). A. Patra, A. K. Mukhopadhyay and A. K. Mitra, Indian J. Chem., 25B (1986) 1167-1170.
1075 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. 63. 64. 65. 66. 67. 68. 69. 70.
L.-J. Lin, L.-Z. Lin, N. Ruangrungsi and G. Cordell, Phytochemistry, 34 (1993) 825-830. A. Patra and S. K. Misra, Magn. Reson. Chem., 29 (1991) 749-752. A. G. Osborne, Magn. Reson. Chem,, 27 (1989) 348-354. A. G. Osborne, Tetrahedron, 37 (1981) 2021-2025. F. A. Macias, G. M. Massanet, F. Rodriguez-Luis and J. Salva, Magn. Reson. Chem., 27 (1989) 892-894. F. A. Macias, R. Hemandez-Galan, G. M. Massanet, F. Rodriguez-Luis, M. Vasquez and J. Salva, Magn. Reson. Chenu, 28 (1990) 732-735. A. L Gray, Phytochemistry, 20 (1981) 1711-1713. B. T. Ngadjui, S. M. Mouncherou, J. F. Ayafor, B. L. Sondengam and F. TiUequin, Phytochemistry, 30 (1991) 2809-2811. A. Patra and A. K. Mitra, Org. Magn. Reson., 17 (1981) 222-224. D. Bergenthal, Z. R6sza, L Mester and J. Reisch, Arc/i. Pharm., 311 (1978) 1026-1029. B. T. Ngadjui, J. F. Ayafor, B. L. Sondengam and J. D. Connolly, J. Nat. Prod., 52 (1989) 243-247. M. P. DiFazio and A. T. Sneden, J. Nat. Prod., 53 (1990) 1357-1361. M. A. Rashid, A. L Gray and P. G. Waterman, J. Nat. Prod., 55 (1992) 851-858. G. Appendino, H. Q. Ozen, S. Tagliapietra and M. Cisero, Phytochemistry, 31 (1992) 32113213. G. Appendino, S. Tagliapietra, G. M. Nano and J. Jakupovic, Phytochemistry, 35 (1994) 183186. G. Appendino, H. C- Ozen, G. M. Nano and M. Cisero, Phytochemistry, 31 (1992) 42234226. H. R. W. Dharmaratne, S. Sootheeswaran, S. Balasubramaniam and E. S. Waight, Phytochemistry, 24 (19S5) 1553-1556. M. Rahmani, T. Y. Y. Hin, H. B. M. Ismail, M. A. Sukari and A. R. Manas, Planta Med., 59 (1993)93-94. J. A. Marco, J. F. Sanz, A. Yuste and A. Rustaiyan, Liebigs Ann. Chem., (1991) 929-931. A. Patra, S. K. Panda, K. C. Majumdar, A. T. Khan and S. Saha, Magn. Reson. Chem., 29 (1991)631-634. D. D. Giannini, K. K. Chan and J. D. Roberts, Proc. Nat. Acad. Sci. USA, 71 (1974) 42214223. G. Appendino, G. Cravotto, L. Toma, R. Annunziata and G. Palmisano, J. Org. Chem. 59 (1994)5556-5564. M. G. Valle, G. Appendino, G. M. Nano and V. Picci, Phytochemistry, 26 (1987) 253-256. G. Appendino, S. Taghapietra, P. Gariboldi, G. M. Nano and V. Picci, Phytochemistry, 27 (1988)3619-3624. D. Lamnaouer, O. Fraigui, M. T. Martin, and B. Bodo, Phytochemistry, 30 (1991) 2383-2386. A. Pelter, R. S. Ward and T. L Gray, J. Chem. Soc, Perkin Trans. 1, (1976) 2475-2483. S. D. Sarker, A. I. Gray and P. G. Waterman, J. Nat. Prod., 57 (1994) 324-327. G. Appendino, S. Tagliapietra, P. Gariboldi, G. M. Nano and V. Picci, Phytochemistry, 27 (1988)944-946. M. A. Kirpichenok, I. I. Granberg, L. K. Denisov and L. M. Melnikova, Izy. Timiryazevsk. S-kh.Akad., 3 (19^5) \12-m. P. G. Lihong, L. Xian and Z. Tingru, Shenyang Yaoxueyuan Xuebao, 6 (1989) 6-11. J. Redondo, F. Sanchez-Ferrando, M. Vails and A. Virgih, Magn. Reson. Chem., 26 (1988) 511-517. D. S. Yufit, M. A. Kirpichenok, Y. T. Struchkov, L. A. Karandasova and 1.1. Grandberg, Bull. Acad. Sci. USSR, Chem. Sci., (1991) 702-710. V. S. Parmar, S. Singh and P. M. Boll, Magn. Reson. Chem., 26 (1988) 430-433. Sadtler Standard Carbon-13 NMR Spectra, No. 2025, Sadtler Research Laboratories, Philadelphia (1977). R. E. Pastor, J. Fabron and A. Cambon, Can. J. Chem., 65 (1987) 1356-1360.
1076 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. 97. 98. 99. 100. 101. 102. 103. 104.
Sadder Standard Carbon-13 NMR Spectra, No. 5217, Sadtler Research Laboratories, Philadelphia (1977). T. Harayama, K. Katsuno, Y. Nishita and M. Fujii, Chem. Pharm. Bull, 42 (1994) 550-1552. A. Z. Abyshev, V. P. Zmeikov and I. P. Sidorova, Khim, Prir. Soedin., (1982) 294-301. M. Fujita, T. Inoue and M. Nagai, Yakugaku Zasshi, 105 (1985) 240-248. P. Forgacs, J.-F. Desconclois, J.-L. Pousset and A. Rabaron, Tetrahedron Lett., (1978) 47834786. S. Sibanda, B. Ndengu, G. Multari, V. Pompi and C. Galeffi, Phytochemistry, 28 (1989) 15501552. G. E. Jackson, W. E. CampbeU and B. Davidowitz, Spectrosc. Lett., 23 (1990) 359-367. A. D. Vdovin, E. K. Batirov, A. D. Matkarimov, M. R. Yagudaew and V. M. Malikov, Khim. Prir.Soedin. (1987)796-799. W. Vilegas, N. Boralle, A. Cabrera, A. C. Bemardi, G. L. Pozetti and S. F. Arantes, Phytochemistry, 38 (1995) 1017-1019. M. Kuroyanagi, M. Shiotsu, T. Ebihara, H. Kawai, A. Ueno and S. Fukushima, Chem. Pharm. Bw//., 34 (1986) 4012-4017. M. Mizuno, H. Kojima, M. linuma, T. Tanaka and K. Goto, Phytochemistry, 31 (1992) 717719. S. P. Sati, D. C. Chaukiyal and O. P. Sati, J. Nat. Prod, 52 (1989) 376-379. N. Matsuda and M. Kikuchi, Phytochemistry, 38 (1995) 803-804. M. Nicoletti, F. Delle Monache and G. B. Marini-Bettollo, Planta Med, 45 (1982) 250-251. J. Reisch, H. M. T. B. Herath, D. Bergenthal and N. S. Kumar, Liehigs Ann. Chem., (1991) 1233-1235. F. A. Macias, G. M. Massanet, F. Rodnguez-Luis and J. Salva, Magn. Reson. Chem., 27, (1989) 705-707. T. Masuda, Y. Muroya and N. Nakatani, Phytochemistry, 31 (1992) 1363-1366. P. C. Joshi, S. Mandal, P. C. Das and A. Chatterjee, Phytochemistry, 30 (1991) 2094-2096. A. Patra, J. Indian. Chem. Soc, 63 (1986) 417-419. Y. Ikeshiro, I. Mase and Y. Tomita, Phytochemistry, 35 (1994) 1339-1341. M. A. Rashid, J. A. Armstrong, A. I. Gray and P. G. Waterman, Z. Naturforsch., 47b (1992) 284-287. P. Tantivatana, N. Ruangningsi, V. Vaisiriroj, D. C. Lankin, N. S. Bhacca, R. P. Bonis, G. A. Cordell and L. Johnson, J. Org. Chem., 48 (1983) 268-270. K. Baba, Y. Maisuyama and M. Kozawa, Chem. Pharm. Bull., 30 (1982) 2025-2035. M. Kozawa, Y. Matsuyama and M. Fukumoto and K. Baba, Chem. Pharm. Bull., 31 (1983) 64-69. Z. Rozsa, I. Mester, J. Reisch and K. Szendrej, Planta Med, 55 (1989) 68-69. M. Ju-ichi, M. Inoue, I. Kajiura, M. Omura, C. Ito and H. Furukawa, Chem. Pharm. Bull., 36 (1988)3202-3205. H. Furukawa, C. Ito, T. Mizuno, M. Ju-ichi, M. Inoue, I. Kajiura and M. Omura., J. Chem. Soc. Perkin Trans. 1, (1990) 1593-1599. Y. Takemura, M. Inoue, H. Kawaguchi, M. Ju-ichi, C. Ito, H. Furukawa and M. Omura, Heterocycles, 34 (1992) 2363-2372. Y. Takemura, M. Ju-ichi, M. Omura, M. Haruna, C. Ito and H. Furukawa, Heterocycles, 38 (1994) 1937-1941. J. Reisch and A. A. W. Voerste, J. Chem. Soc, Perkin Trans. 1, (1994) 3251-3256. J. Tianyi, H. Meifang, P. Jingxian and Y. Xianbin, Bopuxue Zazhi, 9 (1992) 413-417. D. Lamnaouer, O. Fraigui, M.-T. Martin, J.-F. Gallard and B. Bodo, J. Nat. Prod., 54 (1991) 576-578. M. H. A. Elgamal, N. M. M. Shalaby, H. Duddeck and M. Hiegemann, Phytochemistry, 34 (1993) 819-823. P. Ceccherelli, M. Curini, M. C. Marcotulio and G. Madruzza, J. Nat. Prod, 53 (1990) 536538.
1077 105. J. Abaul, E. Philogene, P. Bourgeois, C. Poupal, A. Ahond and P. Potier, J. Nat. Prod., 57 (1994)946-848. 106. M. I. M. Wazeer, V. Kumar and D. B. M. Wickramaratna, Can. J. Spectrosc, 35 (1990) 5759. 107. A. K. Dey, B. R. Bank and T. Bhaumik, J. Indian. Chem. Soc, 67 (1990) 440. 108. T. Fukai, L. Zeng, J. Nishizawa, Y.-H. Wang and T. Nomura, Phytochemistry, 36 (1994) 233236. 109. B. T. Ngadjui, J. F. Ayafor, B. L. Sondengam and J. D. Connolly, Phytochemistry, 28 (1989) 585-589. 110. D. Bergenthal, K. Szendrei and J. Reisch, Arch. Pharm., 310 (1977) 390. 111. A. Ulubelen, R. R. Kerr and T. J. Mabry, Phytochemistry, 21 (1982) 1145-1147. 112. G. Delle Monache, I. Messana, B. Botta and E. Gacs-Baitz, Magn. Reson. Chem., 27 (1989) 1181-1183. 113. G. Reher, Lj. Kraus, V. SinnweU and W. A. Konig, Phytochemistry, 22 (1983) 1524-1525. 114. R. Aquino, M. D'Agostino, F. de Simone and C. Pizza, Phytochemistry, 27 (1988) 1827-1830. 115. R. Mata, F. Calzada, M. R. Garcia and M. T. Reguero, J. Nat. Prod., 50 (1987) 866-871. 116. K. Hori, T. Satake, Y. Saiki, T. Murakami and C.-M. Chen, Yakugaku Zasshi, 107 (1987) 491-494. 117. Sadtler Standard Carbon-13 NMR Spectra No. 5238, Sadtler Research Laboratories Philadelphia (1977). 118. A. D. Vdovin, D. Batsuren, E. Kh. Baiirov, M. R. Yagudaev and V. M. Malikov, Khim. Prir. 5oe^m., (1983) 441-445. 119. K. Hirakura, I. Saida, T. Fukai and T. Nomura, Heterocycles, 23 (1985) 2239-2242. 120. A. R. Bilia, C. Cecchini, A. Marsili and I. MoreUi, J. Nat. Prod., 56 (1993) 2142-2148. 121. H. Ishii, T. Ishikawa, H. Wada, H. Miyazaki, Y. Kaneko and T. Harayama, Chem. Pharm. Bull, 40 (1992) 2614-2619. 122. T. Konishi, S. Wada and S. Kiyosawa, Yakugaku Zasshi, 113 (1993) 670-675. 123. H. Greger, O. Hofer and W. Robien, J. Nat. Prod., 46 (1983) 512-516. 124. S. D. Sarker, J. A. Armstrong, A. I. Gray and P. G. Waterman, Phytochemistry, 37 (1994) 1287-1294. 125. A. B. Ray and S. K. Chattopadhyay, Tetrahedon Lett., 21 (1980) 4477-4480. 126. H. Greger, E. Haslinger and O. Hofer, Monatsh. Chem., 113 (1982) 375-379. 127. A. San Feliciano, M. Medarde, J. L. Lopez, J. M. Miguel Del Corral and A. F. Barrero, An. Quim., Ser C, 82 (1986) 170-172. 128. E. Kh. Batirov, A. D. Matkarimov, V. M. Malikov and E. Seitmuratov, Khim. Prir. Soedin., (1982)780-781. 129. R. D. Waigh, B. M. Zerihun and D. J. Maitland, Phytochemistry, 30 (1991) 333-335. 130. E. G. Crichton and P. G. Waterman, Phytochemistry, 17 (1978) 1783-1785. 131. W. Bremser, L. Ernst, B. Franke, R. Gerhards and A. Hardt, Carbon-13 NMR Spectra Data, No. 10772, Verlag Chemie, Weinheim (1981). 132. W. Bremser, L. Ernst, B. Franke, R. Gerhards and A. Hardt, Carbon-13 NMR Spectra Data, No. 10773. Verlag Chemie, Weinheim (1981). 133. G. Delle Monache, B. Botta, F. Menichini and R. M. Pinheiro, Bull. Chem. Soc. Ethiop., 1 (1987) 65-70. 134. G. Delle Monache, B. Botta, V. Vinciguerra and R. M. Pinheiro, Phytochemistry, 29 (1990) 3984-3986. 135. M. D'Agostino, V. De Feo, F. De Simone and C. Pizza, Phytochemistry, 28 (1989) 17731774. 136. M. D'Agostino, F. De Simone, A. Dini and C. Pizza, J. Nat Prod., 53 (1990) 161-162. 137. Sadtler Standard Carbon-13 NMR Spectra No.4315, Sadtler Research Laboratories, Philadelphia (1977). 138. W. Bremser, L. Ernst, B. Franke, R. Gerhards and A. Hardt, Carbon-13 NMR Spectra Data, No. 10770, Verlag Chemie, Weinheim (1981).
1078 139. W. Bremser, L. Ernst, B. Franke, R. Gerhards and A. Hardt, Carbon-13 NMR Spectra Data, No. 10771, Verlag Chemie,Weinheim (1981). 140. A. Jakobs, G. Llabres and M. Baiwir, Magn, Reson. Chem., 31 (1993) 1S6-1S1. 141. A. K. Mitra, A. Patra and A. Ghosh, Indian J. Chem., 17B (1979) 385-387. 142. A. Z. Abyshev, V. P. Zmeikov, and I. P. Sidorova, Khim. Prir. Soed., 3 (1982) 301-307. 143. F. A. Macias, G. M. Massanet, F. Rodriguez-Luis and J. Salva, Magn, Reson. Chem., 28 (1990) 219-222. 144. G. Llabres, M. Baiwir, W. Vilegas, G. L. Pozetti, J. H. Y. Vilegas, Spectrochim. Acta, 48A (1992) 1347-1353. 145. H. Vuorela, C. A. J. Erdelmeier, Sz. Nyiredy, K. Dallenbach-Tolke, C. Anklin, R. Hiltunen and O. Sticher, Planta Med., 54 (1988) 538-542. 146. P. C. Joshi, S. Mandal, P. C. Das and A. Chatterjee, Phytochemistry, 32 (1993) 481-483. 147. V. Lakshmi, D. Prakash, K. Raj, R. S. Kapil and S. Popli, Phytochemistry, 23 (1984) 26292631. 148. N. U. Khan, S. W. L Naqvi and K. Ishratullah, Phytochemistry, 11 (1983) 2624-2625. 149. O. Thastrup and J. Lemmich, Phytochemistry, 11 (1983) 2035-2037. 150. M. H. A. Elgamal, N. M. Elewa, E. A. M. Elkhrisy and H. Duddeck, Symp.-Pap., lUPAC Int. Symp. Chem. Nat. Prod., Ilth (2) 111 (1978). 151. H. Sun, J. Ding, Z. Lin, Y. Yi and J. Fu, Yaoxue Tongbao, 17 (1982) 121. 152. A. Agrawal, L R. Siddiqui and J. Singh, Phytochemistry, 28 (1989) 1229-1231. 153. H. Kohda and M. Satake, Shoyakugaku Zasshi, 36 (1982) 88-97. 154. T. Okuyama, M. Takata and S. Shibata, Planta Med., 55 (1989) 64-67. 155. L Sakakibara, T. Okuyama and S. Shibata, Planta Med., 44 (1982) 199-203. 156. T. Asahara, L Sakakibara, T. Okuyama and S. Shibata, Planta Med., 50 (1984) 488-492. 157. J. M. Amaro-Luis, G. M. Massanet, E. Pando, F. Rodriguez-Luis and E. Zubia, Planta Med., 56 (1990) 304-306. 158. J. Lemmich and M. Shabana, Phytochemistry, 23 (1984) 863-865. 159. W. Wilegas and G. L. Pozetti, J. Nat. Prod., 56 (1993) 416-417. 160. J. Lemmich, S. Havelund and O. Thastrup, Phytochemistry, 11 (1983) 553-555. 161. A. K. Bose, H. Fujiwara, V. S. Kamat, G. K. Trivedi and S.C. Bhattacharyya, Tetrahedron, 35 (1979) 13-16. 162. P. Rodighiero, P. Manzini, G. Pastorini and A. Guiotto, J. Heterocyclic Chem., 21 (1984) 235240. 163. N. M. D. Brown, A. De, J. S. A. Brunskill and H. Jeffrey, Org. Magn. Reson., 18 (1982) 211213. 164. K. R. Gustafson, H. R. Bokesch, R. W. Fuller, J. H. CardeUina II, M. R. Kadushin, D. D. Soejarto and M. R. Boyd, Tetrahedron Lett., 35 (1994) 5821-5824. 165. T.-T. Jong, H.-C. Hwang, M.-Y. Jean, J. Nat. Prod., 55 (1992) 1396-2401. 166. A. Ulubelen, A. H. Mericli, F. F. Mericli and N. Tan, J. Nat. Prod., 56 (1993) 1184-1186. 167. A. Mizuno, M. Takata, Y. Okada, T. Okuyama, H. Nishino, A. Nishino, J. Takayasu and A. Iwashima, Planta Med., 60 (1994) 333-336. 168. P. Harmala, S. Kaltia, H. Vuorela and R. Hiltunen, Planta Med., 58 (1992) 287-288. 169. M. Rahmani, Y. H. Taufiq-Yap, H. B. M. Ismail, A. Sukari and P.G. Waterman, Phytochemistry, 37 (1994) 561-564. 170. P. S. Steyn, R. Vleggaar, P. L. Wessels and D. B. Scott, J. Chem. Soc, Chem. Commun., (1975) 193-195. 171. K. G. R. Pachler, P. S. Steyn, R. Vleggaar, P. L. Wessels and D. B. Scott, J. Chem. Soc, Perkin Trans. 1, (1976) 1182-1189. 172. A. Osborne, Monatsh. Chem., 115 (1984) 749-756. 173. G. Haas, J. L. Stanton and T. Winkler, J. Heterocyclic Chem., 18 (1981) 619-622. 174. A. E. Nkengfack, J. Kouam, T. W. Vouffo, M. Meyer, M. S. Tempesta and Z. T. Fomum, Phytochemistry, 35 (1994) 521-526. 175. W. Herz and M. Bruno, Phytochemistry, 25 (1986) 1913-1916.
1079 176. N. M. D. Brown, A. De, J. S. A. Brunskill and H. Jeffrey, 7. Heterocyclic Chem., 22 (1985) 619-656. 177. E: Maldonado, E. Hernandez and A. Ortega, Phytochemistry, 31 (1992) 1413-1414. 178. S. L. Debenedetti, P. S. Palacios, E. I. Nadinic, J. D. Coussio, N. De Kimpe, M. Boeykens, J. Feneau-Dupont and J.-P. Declercq, J. Nat Prod., 57 (1994) 1539-1542. 179. G. S. R. Subba Rao, K. Raj and V. P. Sashi Kumar, Indian J. Chem., 20B (1981) 88-89. 180. I. Mester, K. Szendrei and J. Reisch, Planta Med., 32 (1977) 81-85. 181. M. Ju-ichi, Y. Takemura, M. Azuma, K. Tanaka, M. Okano, N. Fukamiya, C. Ito and H. Furukawa, Chem. Pharm. Bull, 39 (1991) 2252-2255. 182. K. V. Subba Raju, G. Srimannarayana, B. Temai, R. Stanley and K. R. Markham, T^rra/z^^rofz, 37(1981)957-962. 183. M. Takata, T. Okuyama and S. Shibata, Planta Med, 54 (1988) 323-327. 184. C. A. J. Erdelmeier and O. Sticher, Planta Med, 51(1985) 407-409. 185. A. I. Gray, M. A. Rashid and P. G. Waterman, J. Nat. Prod, 55 (1992) 681-684. 186. W. Herz, S. V. Govindan and N. Kumar, Phytochemistry, 20 (1981) 1343-1347. 187. I. Sakakibara, T. Okuyama and S. Shibata, Planta Med, 50 (1984) 117-120. 188. M. A. Rashid, J. A. Armstrong, A. I. Gray and P. G. Waterman, Phytochemistry, 31 (1992) 3583-3588. 189. Y. Takemura, T. Kurozumi, M. Ju-ichi, M. Okano, N. Fukamiya, C. Ito, T. Ono and H. Furukawa, Chem. Pharm. Bull., 41 (1993) 1757-1759. 190. Y. Takemura, T. Nakata, M. Ju-ichi, M. Okano, N. Fukamiya, C. Ito and H. Furukawa, Chem. Pharm. Bull, 42 (1994) 1213-1215. 191. Y. Takemura, S. Maki, M. Ju-ichi, M. Omura, C. Ito and H. Furukawa, Heterocycles, 36 (1993) 675-680. 192. M. Ahsan, A. I. Gray, G. Leach and P. G. Waterman, Phytochemistry, 36 (1994) 771-lSO. 193. T. Shamsuddin, W. Rahman, S. A. Khan, K. M. Shamsuddin und J. P. Kintzinger, Phytochemistry, 27 (1988) 1908-1909. 194. M. De Abreu e Silva, T. J. Nagem and M. Yoshida, Revista Latinoam. Quim., 18 (1987) 134135. 195. F. A. Macias, G. M. Massanet, F. Rodriguez-Luis, J. Salva and F. R. Fronczek, Magn. Reson. C/i^m., 27(1989)653-658. 196. Y. Ikeshiro, I. Mase and Y. Tomita, Phytochemistry, 31 (1992) 4303-4306. 197. A. Shoeb, M. D. Manandhar, R. S. Kapil and S. P. Poph, J. Chem. Soc, Chem. Commun., (1978)281-282. 198. T. M. Swager and J. H. Cardellina H, Phytochemistry, 24 (1985) 805-813. 199. Y. Ikeshiro, I. Maze and I. Tomita, Phytochemistry, 33 (1993) 1543-1545. 200. T. Okuyama and S. Shibata, Planta Med, 42 (1981) 89-96. 201. C.-Y. Duh, S.-K. Wang and Y.-C. Wu, Phytochemistry, 31 (1992) 1829-1830. 202. C.-Y. Duh, S.-K. Wang and Y.-C. Wu, Phytochemistry, 30 (1991) 2812-2814. 203. B. W. Seong, C. S. Yook, H. S. Chung and W. S. Woo, Planta Med, 57 (1991) 496-497. 204. G. Delgado and J. Garduno, Phytochemistry, 26 (1987) 1139-1141. 205. S. L. Debenedetti, E. L. Nadinic, M. A. Gomez, J. D. Coussio, N. De Kimpe and M. Boeykens, Phytochemistry, 31 (1992) 3284-3285. 206. J. B. Rampal, K. D. Berlin, N. S. Pantaleo, A. McGuffy and D. van der Hehn, J. Am. Chem. Soc, 103(1981)2032-2036. 207. B. I. Fozdar, S. A. Khan, T. Shamsudin, K. M. Shamsuddin and J. P. Kintzinger, Phytochemistry, 28 (1989) 2459-2461. 208. M. Boeykens, N. De Kimpe, S. L. Debenedetti, E. L. Nadinic, M. A. Gomez, J. D. Coussio, A. Z. Abyshev and V. A. Gindin, Phytochemistry, 36 (1994) 1559-1560. 209. M. A. Rashid, J. A. Armstrong, A. I. Gray and P. G. Waterman, Nat. Prod. Lett., 1 (1992) 7984. 210. S. L. Debenedetti, E. L. Nadinic, J. D. Coussio, N. De Kimpe, J. Feneau-Dupont and J. P. Declerq, Phytochemistry, 30 (1991) 2757-2758.
1080 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240.
241.
A. B. Ray and S. K. Chattopadhyay, Heterocycles, 19 (1982) 19-22. Z. Lin-gen, O. Seligmann and H. Wagner, Phytochemistry, 22 (1983) 617-619. P. Bhandari, P. Pant and R. P. Rastogi, Phytochemistry, 21 (1982) 2147-2149. A. Amoldi, A. Amone and L. Merlini, Heterocycles, 22 (1984) 1537-1544. A. B. Ray, S.K. Chattopadhyay and S. Kumar, Tetrahedron, 41 (1985) 209-214. M. Nagesam and M. S. Raju, Indian J. Chem., 32B (1993) 308-310. K. Nozawa, S. Nakajima, K.-I. Kawai, S.-I. Udagawa and M. Miyaji, Phytochemistry, 35 (1994) 1049-1051. M. R. TePaske, J. B. Gloer, J. Nat. Prod., 55 (1992) 1080-1086. J. A. Laakso, E. D. Narske, J. B. Gloer, D. T. Wicklow and P. F. Dowd, J. Nat. Prod, 57 (1994) 128-133. S. C. Basa, D. P. Das and R. N. Tripathy, V. Elango and M. Shamma, Heterocycles, 22 (1984) 333-337. K. Nozawa, H. Seyea, S. Nakajima, S. Udagawa and K. Kawai, J. Chem. Soc, Perkin Trans. 1, (1987) 1735-1738. K. Baba, M. Taniguti, Y. Yoneda and M. Kozawa, Phytochemistry, 29 (1990) 247-249. P. C. Joshi, S. Mandal and P. C. Das, Phytochemistry, 28 (1989) 1281-1283. H. G. Gutler, F. G. Crumley, R. H. Cox, O. Hernandez, R. J. Cole and J. W. Domer, J. Agric. Food Chem., 27 (1979) 592-595. C. M. Hasan, D. Kong. A. I. Gray, P. G. Waterman, J. Nat. Prod., 56 (1993) 1839-1842. K. Baba, Y. Tabata, M. Taniguti and M. Kozawa, Phytochemistry, 28 (1989) 221-225. G. A. Cordell, J. Nat. Prod., 47 (1984) 84-88. R. Chakrabarti, B. Das and J. Banerji, Phytochemistry, 25 (1986) 557-558. H. Fischer, A. Romer, B. Ulbrich and H. Arens, Planta Med., 54 (1988) 398-400. C. Ito, M. Nakagawa, M. Inoue, Y. Takemura, M. Ju-ichi, M. Omura and H. Furukawa, Chem. Pharm. Bull, 41 (1993) 1657-1658. Y. Takemura, M. Ju-ichi, T. Kurozumi, M. Azuma, C. Ito, K. Nakagawa, M. Omura and H. Furukawa, Chem. Pharm. Bull, 41 (1993) 73-76. C. Ito, T. Ono, Y. Takemura, Y. Nakata, H. Ten, M. Ju-ichi, M. Okano, N. Fukamiya and H. Furukawa, Chem. Pharm. Bull, 41 (1993) 1302-1304. D. J. Jung, A. Porzel and S. Huneck, Phytochemistry, 30 (1991) 710-712. C. Ito, T. Mizuno, S. Tanahashi, H. Furukawa, M. Ju-ichi, M. Inoue, M. Muraguchi, M. Omura, D. R. McPhail and A. T. McPhail, Chem. Pharm. Bull, 38 (1990) 2102-2107. M. Ju-ichi, Y. Takemura, M. Okano, N. Fukamiya, C. Ito and H. Furukawa, Heterocycles, 32 (1991)1189-1194. M. A. M. Nawwar and M. A. Souleman, Phytochemistry, 23 (1984) 2966-2967. P. Bandhari and R. P. Rastogi, Phytochemistry, 20 (1981) 2044-2047. Y. Takemura, M. Ju-ichi, K. Hatano, C. Ito and H. Furukawa, Chem. Pharm. Bull, 42 (1994) 2436-2440. Y. Takemura, T. Nakata, H. Uchida, M. Ju-ichi, K. Hatano, C. Ito and H. Furukawa, Chem. Pharm. Bull, 41 (1993) 2061-2062. (a) B. R. Barik, A. K. Dey, P. C. Das, A. Chatterjee and J. N. Shoolery, Phytochemistry, 22 (1983) 792-794; (b) Errata: B. R. Barik, A. K. Dey, P. C. Das, A. Chatterjee and J. N. Shoolery, Phytochemistry, 22 (1983) 2889. P. G. Waterman, S. M. Zhong, J. A. Jeffreys and B. Zakaria, J. Chem. Res. (S), (1985) 2-3; J. Chem. Res. (M), (1985) 0101-0144.
1081 SUBJECT INDEX Absolute configuration 607-646 of liverwort sesquiterpenoids 607-646 Acanthacerebroside A 481 Acanthaster planci 481 Acetogenins 193 stereoselective synthesis of 193-227 (+)-Acetoxycrenulide 22-28 5-Acetoxyarctigenin monocetate 601 2a-Acetoxycholestanone 888 2P-Acetoxycholestanone 888 5-Acetoxydiniethylmatairesinol 602 5-Acetoxymethyltrachelogenin 602 5-Acetoxymethyltrachelogenin monoacetate 602 5-Acetoxytrachelogenin diacetate 601 N-Acetyl-muramyl-L-alanyl-D-iso-glutaminyl -S -tenbutyl-cysteamine 927 Acetylcholine receptor 863 Acetylene-zipper reaction 469 Acrylamide reagents 320 Aerylonitrile reagents 318 ACRL toxins 178-185 synthesis of 178-185 N-Acylnorreticuline 74 A^-Acylsphingosines 714 p-Adrenergic receptors 720 Anatoxin B, 711 Agelas mauritianus 460,467 agelasphins from 460,467 Agelasphin-9b 467 synthesis of 467-469 Agelasphins 460,467 from Agelas mauritianus 460,467 9-£/7/-Alatol 746 Alexandrium ostenfeldii 703 Alexandrium tamarebnsuis 703 (±)-Ajmalicine 332 Alatol 745 Alcyonidium gelatinosum 695 Aldimine 680 Alexandriumfundyense 703 (-)-Alloyohimban 384 (+)-3-£pi-Alloyohimban 384 .4Isophila pometaria 681 Amadori rearrangement 680 Amathia convoluta 715 Amathia wilsoni 693 Amathamides C,D,E and F 715
American celastraceae 757-764 triterpenes from 757-764 Aminocoumarins 978 O-Aminobenzylalcohol 164 2-Aminobenzylteu-ahydroisoquinoline 73 a-Amino ketones 885-887 steroid-pyrazine dimers via 885-887 Anatoxin 697,698 Androstanolone 885 Anhydrodeacyltautomycin 271,284 Annona bullata 111 isodeacetyl uvaricin from 221 Annona muricata 213 muricatacin from 213 corrossolon from 220 Annotinine 341 Anthocidaris crassispina 486 ganglioside GM5 from 486 Anthricin 555 from Anthricus sylvestris 555 Anthricus sylvestins 555 anthricin from 555 Anthypoxic activity 373 Antiallergic activity 674 Antibacterial activity 777 Anticaries activity 673 Antifeedant assays 771-774 Antifungal activity 229 Antiinflammatory activity 775 Antimicrobial activity 776-778 of triterpenes 776-778 of dimer uiterpenes 776-778 Antihyperglycemic activity 672 Antipyretic activity 775 Antisweet activity 671,672 of gymnemic acid 671,672 Antitumoral activity 739 Antiviral activity 674,776 Apergillus oryzae 807 monohexosylceramides from 807 Apium graveolens 507,515 Aplidiuni californicum 716 Aplysia brasiliana 625 Aplysiatoxins 294 (±)-Apovincamine 331 synthesis of 331 APT spectrum 973 Arabidopsis thaliana 721 Arbuzov reaction 236 Arctigenin monoacetate 601 Arenaria kansuensis 721
1082 Arenes 430 microbial oxidation of 430-432 Arigons enzymatic method 171 Aristlane 607 Aromadendrane 607 4a-Aryldecahydroisoquinolines 81 a-AryI enamine substrates 327 p-Aryl enamine substrates 333 4-Aryltetralin-type lignan 586-588 synthesis of 586-588 Asclepiadaceae 649 Aspergillusflavus 111 Aspergillus fumigatus 469,807,809 monohexosylceramides from 807 Aspergillus versicolor 807,809 monohexosylceramides from 807 Aspidospermine alkaloid 338 Astropecten polyacanthus 725 Asymmetric aza-annulation reaction 378,379 Asymmetric epoxidation 205,207 Asymmetric induction 373-386,484 in aza-annulation reaction 373-386 Asymmetric synthesis 202 ofsolamin 202-206 Atelopus chiriquiensis llA Aulacohorafemoralis chinensis 111 (-)-Austalide B 32-37 from Aspergillus ustus 32 Austalides A-F 32,37 Autoimmunodeficiency syndrome 908 Avidin 919 Aza-annulation 315-386 of enamine related substrate 315-386 Aza-annulation reaction 373 asymmetric induction in 373-386 3-Azetidinol 677,678 from Chara globularis 611 Azidodeoxy-myo-inositol 411 3-Azido-3-deoxy-myc>-inositol 2,4,5-uisphosphate 411 4-Azoniaspiro [3,3] heptane-2,6 diol 677 from Chara globularis 677 Babylonia japonica 724 Bacillus subtilis 692 Bacteriorhodopsin 720,823 Baeyer-Villiger oxidation 172,231,243,257,258,337, 347,509,515,520,533 Baeyer-Villiger reaction 521 Barton radical decarboxylation 75
Basidiomycetes 806,813 glycosphingolipids of 813,814 Bazzania tridens 639 tridensone from 639 Benzochromones 978 N-Benzoyl C,8-phytosphingosine 461 A^-Benzoyl phytosphingosine 462 Benzyltetrahydroisoquinoline 72,76 Berberis cell cultures 53 Bifurcaria bifurcata 713 Bioactive conformations 819-866 of hormones 819-866 Bioactive polyketides 193-227 Biological activity 971 Biological activities 196,459,670-674,771-778 of sphingolipids 459,460 Biological properties 857-863 oflipo-gastrin 857-863 oflipo-CCK 857-863 Biosynthesis 51 of morphine 51-55 Biotinylated peptides 919 Biotinylated gasuim 920 Birch reduction 72,616,638 Boariol 743,752 0-//-Boc-amino benzaldehyde 163 Botanical juvenile hormones 498 Brasilane sesquiterpene 633-638 from Laurencia implicata 633 Brasilenol 625 from Laurencia obtusa 625 Brassica napus 495,498 brassinolide from 495 Brassinolide 495 from Brassica napus 495 Brassinone 500 Brassinosteroid biosynthesis 520 Brassinosteroid metabolites 534-546 Brassinoteroids 520-533 metabolism of 520-533 8-Bramodesoxypicropodophyllin 593,596 1-a-Bromo-desoxypodophylIotoxin 598 l-(3-Bromo-desoxypodophyllotoxin 598 8-Bromodesoxypodophyllotoxin 593,595 6-Bromo-N,-methyl-Nj,-formyltryptamine 691,726 2'-Bromopodophyllotoxin 576 (-)-Bromothebaine 87 Brown's crotylboration 280 Bryostatin A and B 716 Bryostatins 696,697 B. subtilis 11%
1083 Bufo regularis 804 Bugula neritina 715,716 Bungarotoxins 698 Bursehemin 554,556 Burseraceae 558 Bursera schlechtendalii 556 S-O-r-Butyldimethylsilyldehydrononactate 242 tert-B utyldimethylsilyldehydropodophyllotoxin 565 (±)-8-0-/err-ButyIdimethylsilylnonactate 242 rerr-Butyldimethylsilylpicropodophyllin 564 rerr-Butyldimethylsilylpodophyllotoxin 563
Ca'* binding 851 ofgasUn 851-857 ofCCK 851-857 CalyculinA 269 Campesterol 520 l(S)-(-)-Camphanic chloride 607,613,627 (±)-Cainptotiiecin 333,349,355 Candida cyclindracea 429 Cangorinin 760 Cangorosin B 665 Carbonyl reduction 189 3-Carboxycoumarin 978 P-Carotene 708 (-)-Carvone 623 Caryophyllane 607 Cassine 741 Castasteron 495,503,522 from Omithopus sativus 503 Catalytic asymmetric aldol reaction 485 Catalytic receptors 694 Catharanthus roseus 520,521 Caulerpalean algae 688 Caulerpa racemosa 689,714 (±)-Cavinton 331 CCK-10,-12,-13 838 CCK-analog [Thr,NIe]-CCK-9 836 CCK-A antagonist 863 CCK-A receptor 819,825,827,836,857,862-865 CCK-B antagonist 863 CCK-B receptor 824,827,836,839,840,857,858,864 CCK hormone 834 CCK-peptides 825,827,836,852,854,857,859,860 CCK-radioUgand 861 CCK-4,-5,-8,-18,-25,-33,-39, and -58 825 (+)-Ceroplastol I 20-22 Celangulin 771,772 from Celastrus angidatus 111 Celapanol 744
Celastraceae 739-778 Celastrol 757,776 Celastrus 741,753 Celastrus angulatus 111 celangulin from 771 Celorbicol 743 Cephalodiscus gilchristi 875,876,901,902 Cephalostatin 876 (±)-Cephalotaxine 319 Cephalothrix linearis 725 (+)-Ceroplastol I 20-22 Channel-linked receptors 694 Chara globularis 677,698 charaminfrom 677 3-azelidinol from 677 4-azoniaspiro [3,3] heptane-2,6-dioI 677 Charamin 677 from Chara globularis 677 Charatoxins 697 Charonia sauliae 724 Chartella papyracea 692 Chemotaxonomy 701 (+)-Chiloscypholone 624 Chiloscyphone 609-614 from Chiloscyphus polyanthos 609 Chiloscyphus polyanthos 609 chiloscyphone from 609 Chiral auxiliary 164,411,422,607 Chiral chromatography 411 Chiral induction 480 Chiro-\no^\io\ 415,421-439,444 Cholecystokinin (CCK) 824 Chondrillasp. 718 Chondrillin 718 Chondrosia collectrix 719 Cicer arietinum 719 (R)-Citronellol 26 Cinnanioniuni camphola 558 (-)-dimethylmatairesinol 558 Citreoviridine 176 Citropten 979 Citrus unshiu 682 Claisen condensation 243 Claisen rearrangement 23-25,30,259 (+)-Clemeolide 28-32 Cleome viscosa 28 Cleome icoSandra 28 Clitocybe geotropa 813,814 Clitocybe nebularis 813,814 Codeine 47,74,87,91 Codeinone 49,55,57,74,91
1084 Colaphellus loweringi 111 Collision-induced-dissociation (CID) 195 Conocephalenol 625-633,636 from Conocephalum conicum 625,632 OD-Conotoxin GVIA 722 Coprotoxins 698 Corals 716 Corchorus acutangulus 650 23-hydroxylongispinogenin from 650 3p,16p,23,28-tetrahydroxyoIean-12-ene from 650 Corepoxylone 195 Corey's oxazaborolidine catalysts 182 fp/'-Corrossolin 200 synthesis of 200,201 Corrossolone 219 hemi-synthesis of 219,221 from Annona muricata 219 Corticium caeruleum 111 (±)-Corynanaieal 332 Corytuberine 58 (±)-Costaclavine 336 Costaticella hasta 726 barman from 726 pavettine from 726 Cotton effect 623 Coumarins 971-1080 C,g-phytosphingosine 485 entioselective synthesis of 485,486 Crinum asiaticum 687 Crithidia guilhermei 791,792 Crithidia luciliae 791,792 Crithidia oncopelti 791 (-)-(E)-Crotyldiisopinocamphenyl borane 280,281 Cryomixture dimethylsulfoxide 836 Cupressaceae 558 c/5-Cyclogonionenin 222 /rfl/i5-Cyclogonionenin 222 Cystoseira elegans 712,713 Cytidine diphosphate diacylglycerols 433 Cytochromes 914 Cytosolic inositol (tris) phosphate 857 Cytostatic activity 115-11% of P-dihydro-agarofuran skeleton sesquiterpenes 775 of triterpenes 115-11% of dimer triterpenes 116-11% Cyctotoxic T-lymphocytes 921 Deacylgymnemic acid 655,656,661 Debromoaplysiatoxin 295,297 (E)-8-e/7/-2,3-Dehydrononactate 236
Dehydrodesoxypodophyllotoxin 553,555 1,2,3,4-Dehydrodesoxypodophyllotoxin 557,561 3,4-Dehydrodesoxypodophyllotoxin 588 (6S,8R)-(E)-2,3-Debydrononactate 237 Dehydropodophyllotoxin 554 D-n^o-Dehydrophytosphingosines 465 1,2-Dehydroreticuline reductase 53 1,2-Dehydroreticuline 53 1,2-Dehydroreticulinium ion 53 3-Dehydroteasteron 500,512,520 from Distylium racemosum 500 from Triticum aestivum 500 synthesis of 512,513 Demethylenepodophyllotoxin dimethyl ether 597 6-Deoxo-24-epicastasterone 503,514 6-Deoxocastasteron 503 6-Deoxo-28-norcastasterone 507 3-Deoxydebromoaplysiatoxin 295 2-Deoxy-24-epibrassinolide 507 3-Deoxy-3-fluoro-D-m>'o-inositol 439 3-Deoxy-magellanol 745 4-Deoxy-magellanol 745 6-Deoxy-magellanol 745 (±)-Deoxymannojirimycin 347 3-Deoxy-maytol 745 (±)-Deplancheine 333 Desoxypicropodophyllin 552,553,555,596 from Hernandia ovigera 552 (±)-Desoxypodophyllotoxin 586,588,590 D-and L-2,4-Di-(9-benzyI-myo-inositol 427 Dianilinophosphoric chloride 397 Dianilinophosphoric esters 397 1,2-Dibutyroyl inositol 409 l,3-Dichloro-l,l,3,3-teu-aisopropyldisiloxane 409 Dictyodendrilla cavernosa 718 Dicyclohexylidene-myo-inosital 403 1,2:4,5 -Dicyclohexylidene-myo-inositol 401 2,3-Dideoxy-maytol 745 3,4-Dideoxy-maytol 745 3,13-Dideoxy-evoninol 746 Dieckmann cyclization 17 Diels-Alder reaction 8,65,91,258,580,586 3,24-Diepibrassinolide 530 3,24-Diepibrassinolide-3P-laurate 533 3,24-Diepibrassinolide-3P-myrisiate 533 3,24-Diepibrassinolide-3p-palmitate 533 3,24-Diepicastasteron 496,530 Diepomuricanin A 212,219 rflc-l,2-Di-fattyacyl-3-mercaptoglycerol 841 P-Dihydro-agarofuran skeleton sesquiterpenes 775 cytostatic activity of 775
1085 Dihydrocephalostatin 902-904 Dihydrocodeinone 96,98 (±)-Dihydrocorynantheol 331 (±)-P-A'-Dibydrodesoxycodeine methyl ether 65 Dihydronustramine C 691,725 Dihydroflustramine C-N-oxide 691 (±)-Dihydroprotoemetine 330 Dihydropyranocoumarins 984 Dihydrosphingosine 786 Dihydrothebainone 59 p-Dihydrothebainone 66 (±)-Dihydrothebainone 70,72,73,77,78 2p,4P-Dihydroxy-6-deoxy-celorbicol 744 2a,3P-Dihydroxy-B-homo-6a-oxa-5a-pregnane-6,20dione 530 2a,4p-Dihydroxy-8-epi-celapanol 746 2a,3P-Dihydroxy-5a-pregnane-6,20-dione 530,534 (22R,23R)-22,23-DihydroxysUgmasterol 515 (22R,23R,24R)-22,23-Dihydroxy-2a,3a-epoxy-24inelhyl-5a-cholestan-6-one 512 (22R,23R,24S)-22,23-Dihydroxy-28-homoergosterol 520 Dihydroudoteal 688 N,N-Diisopropylphosphoramidites 398 Dimeric triterpenes 764-709 (-)-rrfln5-2-(3,4-Dimethoxybenzyl)-3-(3,4-methylenedioxybenzyl)-Y-butyrolactone 556 from Bursera schlechtendalii 556 (-)-Dimethylmatairesinol 558 (-)-Dimethylmaiairesinol 558,566 Dimethyl phosphorochloridite 399-402 (-)-Dimethylmatairesinol 558 from Cinnamonium camphola 558 Di-myristoylphosphatidylcholine (DMPC) 839 Dinophysistoxin-1 269 3,4-Di-(9-acetyl-l,2:5,6-di-0-cyclohexylideiie-m}'c>inositol 426 l,4-Di-0-benzoyl-m>'0-inositol 421 l,2:3,4-Di-0-cyclohexylidene-m>'o-inositol 428 Di-palmitoylphosphatidylcholine (DPPC) 839 Distylium racenwsum 498,500,512,520 3-dehydroteasterone from 500 Diterpenes 756 DM-CCK sonication 844 DL-epiisopodophyllotoxin 600 D-mannitol 443 D-myo-inosito] 391 DL-myo-inositol 1,3,4,5-tetrakisphosphate 413 Dopamine 53 Drimane 607 Drosophila melanogaster 698
E. coli 921 4,6-0-Ethylidene-A^-benzoyl-D-gIucosamine 462 4-Eicosphingenin 786 Elaodendrum 741 Electrophorus electricus 121 (±)-Emetine 330 Enamine related substrates 315-386 aza-annulation of 315-386 p-Enamino imine substrate 343,366 Enantioselective reduction 288 Enantioselective synthesis of 485 Enzymatic reaction 396 Enzyme-aided enantioselective acylation 428 Enzyme-aided enantioselective hydrolysis 426-428 /-Ephedrine 601 Epiaschantin 552 from Hernandia ovigera 552 24-Epibrassinolide 509,511,529,530,533,534 24-Epicastasteron 503,511,514,529,530,533 from Ornithopus sativus 503 (±)-Epilamprolobine 367 Epimagnolin 552 from Hernandia ovigera 552 (22S,23S)-Epimeric 24-epicastasterone 509 Epimeric 2,3-epoxy brassinosteroids 512 synthesis of 512 8-Epinonactic acid 235 Epipodophyllotoxin 597-601 synthesis of 561,597-601 (±)-5-Epipumiliotoxin 340 synthesis of 340 (-)-6-Epislaframine 386 (±)-5-Epitashiromine 345,353 22R-Epitautomycin 294 Epomuricenin A 212,219 2,3-Epoxyeleganolone 713 Equisetum arvense 495 Eremophilane-type sesquiterpene 639 Ergosterol 509 Ergosterol mesylate 509 Erythronolide B 181 Escherichia coli 709,722,727 24-Ethylbrassinone 500 Euoniminol 747 Euonomynus 741 Euonymus 753 Evan's reduction 253 Evolution 677-728 of secondary metabolites 677-728
1086 Evoninol 747 Exciton-chirality method 748 Ferrier reaction 433-437 Fetizon reagent 28 (±)-Festuclavine 386 Fluorescence indicator fura-2 856 Flustrabromine 691 Flustrafoliacea 689,692,693,708,725 Hustramide A and B 691 Rustramine A,B,C,D and E 690,691 Flustramine D-A^-oxide 691 Flustraminol A and B 691 FlustrarineB 690 Rustriidae 690-693 Friedel-Crafts acylation 234 Friedel-Crafts alkylation 231 Friedel-Crafts annulation 70 Friedelin 770 Fritsch-Buttenberg-Wiechell rearrangement 171 Frullania dilatata 607 Frullania tamarisci 607,614,623 (-)-frullanolide from 607 (-)-tamariscol from 614,623 (-)-Frullanolide 607 from Frullania tamarisci 607 Fungi and protozoa 785-814 glycolipids from 785-814 Fuchs synthesis 892-895 of tetrahydrocephalostatin 892-895 Furanocoumarins 978,979,985 Fusarium lini 807 monohexosylceramides from 807 Fusarium solani 807 monohexosylceramides from 807 Fusicoccum amygdali 807 monohexosylceramides from 807 N-2-hydroxy-3-trans-octadecenoyl-l-0-D-glucosyl9-methyl-cis-4,8-sphingadienine 807 Galactinol 396 a-Galactosylceramides 460 GaINAcpi-^[NeuAca2-^3)Gaipi-4Glcpi->'lCer488 Ganglioside GM5 486 from Anthocidaris crassispina 486 Ganglioside 486,786,788 synthesis of 486-488 Gastrin and CCK 824-840
Gastrin family 819-866 hormones of 819-866 Germacrane 607 Gigantecin 221,222 from Goniothalamus giganteus 111 hemi-synthesis of 221,222 Gigantetronenin 221 Gleosporiumfructigenum 715 Glucosaminyl-chiro-inositol phosphate 434 23-0-P-D-Glucophyranosyl-2-epi-25-methyldolichosterone 495 from Phaseolus vulgaris 495 23-0-p-D-Glucophyranosyl-25-methyldolichosterone 495 2-0-P-D-Glucophyranosyl-3,24-diepicastasterone 529,532 3-0-p-D-Glucophyranosyl-3,24-diepicastasterone 529,541 23-0-P-D-Glucophyranosyl-brassinolide 522 from Vigna radiata 522 26-P-D-Glucopyranosyloxy-24-epi-brassinolide 541 25-P-D-Glucopyranosyloxy-24-epi-brassinolide 539 D-Glucurono-6,3-lactone 396,442 D-Glucosamine 461 P-Glucuronidase 649,650 from Helix proniatia 655 Glycolipids 785-814 from fungi and protozoa 785-814 N-Glycolylneuraminic acid 786 a-Glycosyl-c/»rc>-inositol 434 Glycosphingolipids 457,459,785-814 isolation and purification of 786-797 of Trypanosoma cruzi 796-802 of fungi 806 from zygomycetes 806 from denteromycetes 806 of basidiomycetes 806 Glycosylphosphatidylinositol anchor 840 Gobius criniger 48 Goniocin 193 Goniothalamus giganteus 221,222 gigantecin from 221 Gorgonane 607 G-protein-coupled receptors (GPCR) 822 G-protein-linked receptors 694 G-proteins 861 Grignard reaction 473,630 Guaiane 607 Gyminda 741 Gymnemagenin 649,650 from Gymnema sylvestre 650
1087 Gymnemanol 650,665 Gymnemasides 650 Gymnema sylvestre 649,650,653,661,671,673 gymnemic acids from 649-676 Gymnemic acids 649-676 from Gymnema sylvestre 649-676 biological activity of 670-674 Gymnestrogenin 649,650,656 from Gymnema sylvestre 649,650,656 Gynostemma pentaphyllum 662 Gypenosides 650
Halicerebroside A 459 Halichondria japonica 475 Halichondria panicea 719 Haliclona sp. 459 Halimeda incrassata 689 Halimeda tuna 689 Hannoa klaineana 726 Hapalochlaena maculosa 724 Harman 726 from Costaticella hastata 726 Heathcock's asymmetric aldol reaction 283 Heck reaction 96 Helix promatia 655 Hemi-synthesis 219-222 of acetogenius 219-222 ofsolamin 219,220 of reUculatacin 219,220 of corrossolone 220,221 of isodeacetyl uvariein 221 ofgigantccin 221,222 Heroin 48 Hemandiaceae 558 Hemandia cordigera 561 5'-methoxypodorhizol from 561 Hemandia ovigera L. 552,600 hemandion from 600 isohemandion from 600 isohemandion from 600 desoxypicropodophyllin from 600 epiaschantin from 600 epimagnolin from 600 Hemandin 553,571 synthesis of 579-594 Hemandion 552,555 from Hemandia ovigera 552 Hemiarin 974,976 Hemolactone 554,565,569,566 from Hemandia ovigera 569
Hetero-Diels-Alder reaction 187 Hexadecyltrimethylammonium hydroxide (CTAH) 832 3p,16p,2ip,22a,23,28-Hexahydroxyolean-12-ene 649 Hexepi-uvBiicin 207 synthesis of 207-211 Hinge-peptide 930-958 Hinokinin 566 Hippocratea 741,754 Hippocrateaceae 740 Historical perspective 43-107 of morphine synthesis 43-107 Histrionicotoxins 698 Hoffmann degradation 51 Hoffmann'scycloaddition 257 (±)-Homononactate 253 Homo sapiens 705 Homoteasterone 507 from Raphanus sativus 507 Hormones 819-866 of gastrin family 819-866 bioactive conformations of 819-866 Homer-Emmons condensation 288 Homer-Emmons olefmation 288,633 Horner-Emmons reaction 481,586 Human IgG, Hinge-fragment 907-958 Human leukemia cells 269 Hijckel MO method 981 (±)-Huperzine A 321,324 Hydractinia echinafa 701 Hydrazone 67 Hydrodictyon reticulatuni 495,507 4P-Hydroxy-alatol 747 25-Hydroxy-celapanol 744 4p-Hydroxy-celapanol 744 4p-Hydroxy-celorbicol 743 (20R)-Hydroxy-3,24-diepicastasteron 531,539,542 (20R)-Hydroxy-3,24-diepibrassinolide 531 4p-Hydroxy-6-deoxy-6-deoxy-celorbicol 743 25-Hydroxy-3,24-diepibrassinolide 532 25-Hydroxy-24-epicastasterone 529 26-Hydroxy-24-epicastasterone 529 (S)-l-(l'-Hydroxyethyl)-P-carboline 726 (S)-p-Hydroxyisobutyric acid 172 2'-Hydroxyhexadecanoyl-l-0-p-D-glucopyranosyl-9methyl-4,8-D-erytho-sphingadienine 814 23-Hydroxylongispinogenin 650,656 from Corchorus acutangulus 650 5-Hydroxymatairesinol 601 5-fl//o-Hydroxymatairesinol 601 (4E,8E)-N-D-2'-HydroxypaImitoyl-l-0-P-Dglucopyranosyl-9-methyl-4,8-sphingadienine 813
1088 6p-Hydroxy-pentahydroxy-agarofurano 746 N-2-Hydroxy-3-trans-octadecenoyl-1 -O-D-glucosyl-9methyl-cis-4,8-sphingadienine 807 from Fusicoccum amygdali 807 25-Hydroperoxy-2a,3p,12p,20S-tetrahydroxydammar23-ene 650 Hypoglycemic activity 672 Ichthotoxicity 607,716 Iguesterin 151,116 (+)-Ikanigamycin 10-17 p-Imino sulfoxide substrate 382 Immunoadjuvant monophosphoryl lipid A in 918 Immunogens 909-920 Immunosuppressive activity 739,921 INDO MO method 981,985 D-c/z/r(?-Inositol 439 L-c/if>-Inositol 2,3,5-trisphosphate 416 Inositol phosphates 393-451 synthesis of 393-451 Insecticidal activity 229,704 Intramolecular aldol condensation 269,297 Intramolecular aldol cyclization 633 Intramolecular Claisen condensation 299 lonophoretic activity 857 lophotoxins 697,698 Ireland-Claisen rearrangement 230,231,259 Isoalatol 745 Isoboldine 58 2-/50caproyl-3R-hydroxymethyl-'y-butyro lactone 700 Isocelorbicol 743 Isochiloscyphone 614 Iso-1-cytochrome 915,916 Iso-1-cytochromeC 832,838 Isodeacetyl uvaricin 221 from Uvaria narum 111 from Annona bullata 111 hemi-synthesis of 221,222 Isodesoxypicropodophyllin 588 Isoergosterol 509 Isoeuoniminol 747 Isoflustramine D 691 Isohemandin 586 Isohemandion 552 from Hemandia ovigera 552 Isoiguesterim 776 Isolation and purification 786-788 of glycosphingolipids 786-788 Isomagellanol 746 Isomolvizarin 193
Iso-P-peltatin B methyl ether 59 (R)-2,3-Isopropylidene glyceraldehyde 181 L-Isopodophyllotoxone 600 Isopristimerin III 760 Isosalutaridine 58 (±)-Isosophoramine 323 Isostigamasterol 515 Isotingenone 760 Isotingenone III 760
Jones oxidation 235,514,627,641 Katsuki-Sharpless epoxidation 481 Ketamine 680 Krapcho decarboxylation 246
5-Lactams 315 synthesis of 315-386 Lactosylceramide 796 (-)-Laminitol 434 (±)-Lamprolobine 367 LammcuUa apialis 716 Laudanine 49,73 Laurencia implicata 633 brasilane sesquiterpene from 633 Laurencia obtusa 625 brasilenol from 625 Leishnmnia 791,793,794 Leishmania aniazonensis 794 Lemieuix oxidation 171 Lemieuix-Jounson oxidation 81 Lepidoptera 772 Light scanning detection (LSD) 195 Lignaiis 551 from Hernandia ovigera 551 synthesis of 551 Lilium longifolium 495,507,522 leasterone myristate from 507 teasterone 3-myristate from 495,522 Linensfuscoviridis 725 Lipo-CCK 844 physical properties of 844-848 biological properties of 857-863 Lipo-gastrin 844 physical properties of 844-848 biological properties of 844-848
1089 Lipophilic derivatives 840-851 of gastrin 840-851 ofCCK 840-851 Lissodendoryx isodactyalis 715,716 Liverwort sesquiterpenoids 607-646 absolute configuration of 607-646 Locusta migratoria manilensis 111 Locusta migratoria migratorioides 111 L-quebrachitol 439 L-Selectride reduction 235 Luffariella variabilius 111 Luffariellin A and B 717 (±)-Lupinine 345,356,370 Lycopersicon esculentum 523,529,532,533 (±)-Lycopodine 321 Lycopodium alkaloid 3-7 Lycopodium magellanicum 3 Lycopodium paniculatum 3 Lycopodium alkaloids 341 Lymphocytic leukemia system 696 Lyngbya majuscula 294
Macrocyclic sesquiterpene alklaoids 753-755 Magellanin 665 Magellaninone 3-7 Magellanine 3-7 Magellanol 746 Magnoliophyta 740 Magnolitae 740 m-Maleimidobenzoyl-N-hydroxysuccinimide ester 911 N"-Maleoyl-P-alanyl-gastrin-[2-17] derivatives 913 Malkanguniol 744 Z)-Mannitol 157,159 D-Mannitol diacetonide 157 (±)-Mannonolactam 347 2-a-Mannopyranosyl-1 -phosphatidlyl-D-myo-inositol 449 Mannosylated phosphoinositides 393 Manoalide 717 Maytansinoids 739 Maytenus genus 739,740 Maytenus rigida 111 wilfordin from 771 Maytol 747 M. boaria 152,115 M. buxifolia 757 MDLISIS-Base 971 Medorinone 338 N-Mesitylenesulfonyltriazole 401 Mesyloxy oxazoline 461
Metabolism 523-533 of brassinosteroids 523-533 of 24-epibrassinolide 523-533 of 24-epicastasterome 523-533 Metachromin D 643 5-Methoxydesoxypodophyllotoxin 576 5'-Methoxypodorhizol 559,561,566 from Hernandia cordigera 561 7-Methoxycoumarin 976 4-Methylumbelliferone 985 (±)-Na-Methyl-Np-acetylphlegmarine 341 0-Methylvulgarolide 20 30-Methoxyscillatoxin D 294 sythesis of 294-309 Metridium senile 813 Michael acceptor 315,329,334,339,369 Michael addition 10,63,81,333,343,355,374,378,369 Michael type addition 65,269,302,303,305,307,309 Microbial oxidation 430-432 ofarenes 430-432 Microcystins 269 M. illicifolia 665 Mitsunobu condensation 88 Mitsunobu inversion 7,237,246,259,281,470,893,903 Mitsunobu reaction 174,217,235,471 M. loesner 757 Molecular recognition 819 Monohexosylceramide 799,807,814 from Sporothrix schenckii 807 from Fusicoccum amygdali 807 from Fusarium lini 807 from Fusarium solani 807 from Aspergillus oryzae 807 ivom A. fumigatus 807 from A. versicolor 807 Monovalent peptide 914-918 Morphine 43-109 history of 45-55 biosynthesis of 51-55 synthesis of 56-98 Morphine synthesis 43-107 historical perspective of 43-107 Morphinan 76 Mortonia greggi 751 Morton ins A,B,C and D 751 Morionol 751 [wm Mortonia greggi 751 Mugil cephalus 486 Mukaiyama aldol reaction 292 Multiple antigen 920 M. umbelata 665
1090 Muricatacin 195,213-218 from Annona muricata 213 (+)-Muricatacin 198 Musca domestica 698 Mycobacteria 393 Mycobacterium tuberculosis 351 A/yo-inositol 391 M>'0-inositol inflates 439 Mytilitol 434
(±)-Na-benzyl-20-desethylaspidospermidine 323 synthesis of 323 N-acetyl-D-mannosamine 469 N-acetylgalactosamine 786,792 N-acetylglucosamine 786,789 N-acetylmuramyl-dipeptide 928 N-acetyl-muramyl-L-alanyl-D-isoglutamine 923 N-acetyl-muramyl-L-alanyl-D-isoglutaminyl-N^ palmitoyl-L-lysyl-S-r^rr-butyl-thio-cysteamine 923 NADP oxidoreductase 55 NCI assay 878 Nectria haematococca 709 Neopinone 55 Nereistoxin 697 Netzahualcoyene 777 Netzahualcoyondiol 777 Netzahualcoyone 764,765,777,778 Netzahualcoyonol 777 Nicotinic acetylcholine receptor 695 Nicotinic toxins 697-699 Nitzschia palea 698 NMR spectroscopy 534-546 of brassinosteroid metabolites 534-546 Noctuidae 111 Nodularin 269 Nonacticacid 229-268 synthesis of 229-268 Nonactin-potassium thiocyanate complex 229 Non-covalent peptide 918-920 Non-peptide antagonists 863 S-Norcoclaurine 53 Noroxymorphone 47,48 (-)-Norreticuline 86 Nucleophilic substitution 438-445 Octahydro-lH-[l]-benzopyrano-[4,3,2,-e,f]isoquinolines 81 Octahydro-lH-benzofuro-[3,2,-e]-isoquinolines 81
(-)-1,2,3,4,6,7,12,12b-Octahydroindolo[2,3-a] quinolizine 383 Octant rule 623,628 Octyl-P-D-glucopyranoside 832 Octyl-p-D-glucopyranoside micelles 838 Ophidiaster opidiamus 460 Opium alkaloid 45 Ornithopus sativus 503,514,520,523,529-533 castasterone from 503 24-epicastasteron from 503 Orthosphenia 754 Orthosphenia mexicana 756,758 Orthosphenin 755 Oryzasativa 522 Oscillatoria nigroviridis 294 Oscillatoxin D 269-309 sythesis of 269-309 Otivarin 193 Oxazolidinone auxiliaries 161 £:MM-Oxo-2,33-dihydrosolamin 197 synthesis of 197-199 4-Oxo-2,33-dihydrosolamin 199 19-Oxo-3p,20S-dihydroxydammar-24-ene 650 (±)-8-Oxocephalotaxine 319 Oxy-Cope reaimgement 17 A*-7-0xygenated brassinosteroids 515 synthesis of 515-520
Pachysiima 741 Pacifigorgia admsii 614,615 pacifigorgiol from 614 Pacifigorgiol 624,625 from Pacifigorgia admsii 624 N-Palmitoyl-S-[2,3-bis (palmitoyloxy) propyl]-L-CysSer 926 (+)-Pallescensin A 17 Pancreatic acinar cells 857 Paniculatine 4 Papaver somniferum 45,54 Papilio xuthus 682 Papyraceabromine-A 692 (+)-Pallescensin A 17 (±)-P-Peltatin A methyl ether 586,589,592,596 Penares sp. 460 Penaresidines 460 from Penares sp. 460 Penicillus dumetosus 688 Pentahydroxy-agarofuran 744 2oc3p,12p,20S,25-Pentahydroxydammar-23-ene 650 3p,16p,21p,23,28-Pentahydroxyolean-12-ene 650
1091 Pentalenolactone antibiotic 7-10 Pentaporafasciata 715 Peptide antagonists 863-865 Peritassa 754 Peritassin A and B 755 Pesticidal activity 196 Phaseolus vulgaris 495,522 23-0-P-D-glucopyranosyl-2-epi-25-methyldolichosteron from 495,522 8-Phenylmenthylglyoxalates 190 Af-(R)-1-Phenylethylcarbamates 427 Phoenix dactylifera 507 Phosphatidyl-myo-inositol 4,5-bisphosphate 392 Phosphatidylcholine vesicles 848,856 Phosphatidylinositol 3,4,5-trisphosphate 410,427 Phosphatidylinositol 3-kinase 394 Phosphatidylinositols 445-450 synthesis of 445-450 Phospholipid bilayer 848 Phosphoramidites 398 Phosphorylation 395,397-402 Phycomyces blakeslearus 806 Physical properties 844-848 ofUpo-gastrin 844-848 oflipo-CCK 844-848 D-Jcy/o-C,^-Phytosphingosine 479 Phytosphingolipids 457-490 synthesis of 457-490 biological activities of 459,460 Phytosphingosines 786,806 D-n770-C„-Phytosphingosine 471,478,485 D-n770-Phytosphingosines 465 Phytosterols 520 Piceaabies 498 Picropodophyllone 600 Pieris rapae 111 Pinacol coupling 443 Pinacol rearrangement 174 (±).Pinitol 431 Pinus thunbergii 498 Piperaceae 558 Pisum sativum 708,710 Plakortis angulospiculatus 719 Plakortis halichondroides 719 Plakortis lita 718,719 Plocamium cartilagineaum 714 Podophyllotoxin 554,555,561,597-601 Podophyllotoxin 597-601 synthesis of 597-601 Podophyliotoxone 599
Podophyllum emodi 556 podorhizol from 556 Podophyllum peltatum 557 podorhizol from 556,557 Podorhizol 554,556,565 from Podophyllum emodi 556 from Podophyllum peltatum 557 (+)-Podorhizon 588 Polydimehtylacryl-amide resin 921 Polyester sesquiterpenes 743 Polyketides 155 synthesis of 155-192 Polymastia sp. 716 Polypropionate 155 Prins cyclization 174 Prins reaction 882,891 Pristimerin 757,760,765,776,778 (±)-Prosopinine 347 Protection of inositols 403-421 PSbioassay 878 flf-Pseudoephedrine 602 Pseudomonas puiida 430 Pseudomonas sp. 429,430 Pseudophrynamines 726 Pseudophrynaminol 725 Pseudophryne coriacea 725 Pseudophryne guentheri 726 Pseudophryne occidentalis 126 Pseudosuberites hyalinus 696 Psoralen 978 (R)-(+)-Pulegone 319 (-)-Pumiliotoxin C 319 Pyridine substrates 369 Pyridones 315 selective synthesis of 315-386
D-and L-Quebrachitol 396 Radioimmunoassay 500 Ramberg-Backlund olefmation 205,206 Kaphanus sativus 507 homoteasterone from 507 Regioseleciive acylaiion 404 Regio-pyranocoumarins 993 Regioseleciive aza-annulation 319,327 Regioseleciive Michael addition 319 Regioseleciive nucleophilic substitution 406 Regioseleciive phosphorylation 399 Reserpine 64
1092 Reticulatacin 205 hemi-synthesis of 219,220 Reticulatamol 212 R-Reticuline 53,54 S-Reticuline 53 Reticulatamone 195,212 Retro-Michael reaction 18 Retro-aldol reaction 284,285 Retro-Wittig reaction 78,176 Rhipocephalus phoenix 689 Rhodopsin 823 Ritterazines 881-884 synthesis of 900-902 Ritterella tokioka 881,882 Robinson annulation 33 Rollinia mucosa 208 roUiniastatin from 208 £"/ir-Rolliniastatin-l 210 synthesis of 210,211 Rolliniastatin-2 208,209 from Rollinia mucosa 208 synthesis of 208,209 Rzedowskia tolantonguensis 756,764
S. bikiniensis 700 S. griseus 700 Saccharomyces cerevisiae 721,722,727 Salacia 741 Salutaridine 54,58,71,73 Salutaridinol 54,55 Sapogenins 649,650 Saponins 649,650 Saxitoxins 697,698 Schaefferia cuneifolia 761 Schizophyllum commune 460,813,814 Schizosaccharomycespombe 111 Schizothrix calcicola 19A Sclerotinia sclerotiolum 269 SDS micelles 832 Secale cereale 500,502,512 Secomanoalide 717 Secondary metabolites 611-12% evolution of 611-11% Securiflustra securifrons 691 Selective synthesis 315 ofS-lactams 315-386 ofpyridones 315-386 Sesquiterpene alkaloids 753 Sesquiterpenes 743-752 SFORD techniques 972
Sharpless asymmeU'ic dihydroxylation 197 Sharpless asymmetric epoxidation 217 Sharpless epoxidation 197,204,205,244 Shizandra chinensis B 589 Sigmatropic 1,3-H-migration 166 Sitophilus granarius 698 (-)-Slaframine 386 Sodoptera littorais 111 Solamin 193,197,199,202,205 asymmetric synthesis of 202-206 hemi-synthesis of 219-220 Spennatophyta 740 D-(+)-erv///ro-Sphingosine 461 Sphingofungins 460,469 Sphingosines 786 Spiroethers 269-309 synthesis of 269-309 Spiroketal reduction 276,277 with DIBAH 276,277 with Silane-Lewis acid 277,278 Spodoptera littoraUs 111 Sporothrix schenckii 807 monohexosylceramides 807 Staphylococcus aureus 111,11% Stereochemistry 303 of spiroethers 303-305 Stereoselective synthesis of 193-227 ofacetogenins 193-227 Stereospecific synthesis 197 from chiral pool 197-202 Stem-Volmer iodide-quenching constant 847 Steni-Volmer quenching constant 832,844 Steroid-pyrazine 875-905 via a-amino ketones 885-887 Stigmasterol 515 Streptavidin 919 Streptococcus mutans 673 Streptomyces griseochromogenes 269 Streptomyces species 229 Streptomyces phaeochromogenes 10 Streptomyces spiroverticHiatus 269 tautomycin from 269 Suchilactone 590 Surface-immunoglobulins 915,916 Swern oxidation 91,176-178,198,201,210,260,282,283, 297,475,623,633,641 SwinholideA 186-190 synthesis of 186-190 Syntliesis of ACRL toxin 178-185 of agelasphin-ab 467-469
1093 of (±)-apovincamine 331 of 4-aryltetralin-type lignan 586-588 of (±)-N,-benzyl-20-(lesethylaspidospermidine 323 ofbrassinosteroids 507-520 of cephalostatins 900-902 of e/7i-corrossoline 200 of 3-dehydro-24-epiteasterone 512,513 of 3-dehydroteasterone 512,513 of 2-deoxy-3,24-diepibrassinolide 515 of 6-deoxo-24-ep/-castasterone 512,513 of 2-deoxy-24-epibrassinolide 515 of epimeric 2,3-epoxy brassinosteroids 512 of epipodophyllotoxin 597-601 of (±)-5-epipumiliotoxin 340 of 24-epiteasterone 515 of 24-epityphasterol 515 of gangliosides 486-488 ofhemandin 579-584 of hexepi'Uwmcin 207-216 oflignans 551-606 of30-methyloscillatoxinD 294-309 of inositol phosphates 391 -451 of morphine 56-98 ofnonacticacid 229-268 ofnonactin 260-265 ofoscillatoxinD 269-309 of A*-7-oxygenated brassinosteroid 515-520 of enM-oxo-2,33-dihydrosolamin 197 of phosphatidylinositols 445-450 of phytosphingolipids 457-490 of podophyllotoxin 597-601 ofritterazineK 900-902 of ent-rolliniastatin-1 210,211 of enr-rolliniastatin-2 208,209 of secasterone 512 of spiroethers 269-309 ofswinholideA 178-185 of (±)-tashiromine 353 oftautomycin 269-309 of A^'^-unsaturated brassinosteroids 515-520 Synthetic peptide 920-926 Syringaldehyde 569 T and B-cell epitope chimeras 929,930 T. dionisii 801-804 T verpertillionis 801,804 (-)-Tamariscol 614,623,625 from Frullania tamarisci 614-624 Taricha torosa 724
(±)-Tashiromine 353 synthesis of 353 Tautomycin 269-309 synthesis of 269-309 from Streptomyces spiroverticHiatus 269 0-TBDPS lactaldehyde 164 Teasteron 500,512,520 Teasteron 3-myristale 495,500 from Lilium longiflorum 495 D-1,2,5,6-Tetra-O-benzyl-myo-inositol 425 D-1,4,5,6-Tetra-0-benzyl-m}'o-inositol 444 Tertiary metabolites 680,681 2,3,13,15-Tetra-deoxy-evoninol 744 Tetrabenzyl-myo-inositol 444 2,3,13,15-Tetradeoxy-isoeuoniminol 745 Tetrahydrocephalostatin 12, 892-895 fuchs synthesis of 892-895 (22R,23R,24S)-2a,3a,22,23-Tetrahydroxy-24methyI-B-homo-6a-oxa-5a-cholestan-6-one 495 5,6,7,8-Tetrahydroisoquinoline 58,59 3P,16p,23,28-Tetrahydroxyolean-12-ene 650 Tetrasubsiituted enamino esters 358 Teirodotoxin 697 Thapsia villosa 685 Thebaine 49,55,57 P-Thebainone 65 Thorectandra excavatiis 111 Thorectolide monoacetate 717 Tingenol 778 Tingenone 151,160,116 TMAO-urea complex 678 Torpedo califronica 721 l,3,5-Tri-(9-benzoyl-myo-inositol 422 2,3,6-Tri-(9-buiyroyl-/7iyo-inositol 1,4-5-trisphosphate 409 Tridensone 639-643 from Bozzania tridens 639 D-2,3,6-Tribenzyl-m.yo-inositol 397 2,3,13-Trideoxy-evoninol 745 3,4,13-Trideoxy-evoninol 745 2,3,13-Trideoxy-isoeuoniminol 746 2a,3p,6P-Trihydroxy-5a-pregnane-20-one 532,532 3,4,5-Trimethoxycinnamate 580 Trishomononactate 253 Triteipenes 757-769 from American celastraceae 757-769 Trificum aestivum 500,503,512,520 3-dehydroteasterom from 500 Trypamosofiia 791,793 Trypanosimatids 793 Trypanosoma brucei 448,449
1094 Trypanosoma cruzi 796-802 glycosphingolipids 796-802 Trypanosoma mega 804 Trypterigium 753 Trypterigium wilfordii 111 tryptofordin from 771 Tryptofordin 771,772 from Trypterigium wilfordii 111 d-Tubocurarine 698 Tubulanus punctatus 725 (±)-Tubulosine 330 Tutufa lissostoma 724 Typhasterol 500,512,520 Umbelliferae 8 A^'-Unsaturated brassinosteroids 515 synthesis of 515-520 Unsymmetrical pyrazines 887-892 Uvaria acuminata 193 uvaricin from 193 Uvaria narum 111 isodeacetyl uvaricin from 221 Uvaricin 193 itom Uvaria acuminata 193
y-Valerolactone 198 (±)-Vallesamidine 318 Verticillane 607 Vigna radiata 522 23-O-p-D-glucopyranosyl brassinolide from 522 Vilsmeier formylation 233 Vinylogous amide susbtrates 376,386 Vinylogous carbamate substrates 344,377 Vinylogous urea substrates 363 Virola sebifera 726 Von Braun degradation 51 (-)-Vulgarolide 19,20 Wacker oxidation 633,273,283 Wadsworth-Enmions reaction 897 1,2-Wagner-Meerwein shift 882 Weiss reaction 286 Wieland-Miescher ketone 29 Wilfordin 771,772 from Maytenus rigida 111 Wilkinson's catalyst 207 Wittig olefmation 8 Wittig homologation 196,216
Wiitig methylenation 177 Wittig reaction 178,233,234,255,256,288,297,618 Wittig-Homer homologation 333 Wittig-Homer reaction 247,623 Wolff-Kishner reduction 35
Xenochemicals 680 Xenochemicals 680,681 XestinAB 718,719 Xestospongia sp. 718
(-)-Yatein 557,566 (-)-Yohimban 384
Zeuxis siquizorensis 124 Zeylasterone 2,3-dimeihyl ehter 765 Ziziphin 671 Zizyphus jujuba 671 Zygomycetes 806 glycosphingolipids from 806 Zygomycetes 806,814 glycosphinolipids from 806