M E T H O D S IN E N Z Y M O L O G Y EDITORS-IN-CHIEF
John N. Abelson
Melvin I. Simon
DIVISION OF BIOLOGY CALIFORNIA...
13 downloads
959 Views
45MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
M E T H O D S IN E N Z Y M O L O G Y EDITORS-IN-CHIEF
John N. Abelson
Melvin I. Simon
DIVISION OF BIOLOGY CALIFORNIA INSTITUTE OF TECHNOLOGY PASADENA, CALIFORNIA
FOUNDING EDITORS
Sidney P. Colowick and Nathan O. Kaplan
Methods in Enzymology Volume 234
Oxygen Radicals in Biological Systems Part D E D I T E D BY
Lester Packer DEPARTMENT OF MOLECULAR AND CELL BIOLOGY UNIVERSITY OF CALIFORNIA, BERKELEY BERKELEY, CALIFORNIA
Editorial Advisory Board Bruce Ames Kelvin Davies Barry HaUiwell
Etsuo Niki William Pryor Helmut Sies
® ACADEMIC PRESS San Diego
A Division of Harcourt Brace & Company New York Boston London Sydney Tokyo
Toronto
Contributors to V o l u m e 2 3 4 Article numbers are in parentheses following the names o f contributors. Affiliations listed are current.
ELIAS N. ABOUJAOUDE (2), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720
MICHAEL BOCKSTETTE (13), Department of lmmunochemistry, Deutsches Krebsforschungszentrum, D-69120 Heidelberg, Germany WOLF BORS (41), Institutfiir Strahlenbiologie, GSF-Forschungszentrum fiir Umwelt und Gesundheit Neuherberg, D-85758 Oberschleissheim, Germany KARLIS BRIVIBA (37), Institut fiir Physiologische Chemie I, Heinrich-Heine-Universitgit, D-40001 Diisseldorf, Germany GARRY W. BUCHKO (8), Department of Radiobiology, Cross Cancer Institute, Edmonton, Alberta, Canada T6G 1Z2 MARK J. BURKITT (7), Division of Biochemical Sciences, Rowett Research Institute, Bucksburn, Aberdeen AB2 9SB, Scotland, United Kingdom LARRY G. BUTLER (42), Department of BiDchemistry, Purdue University, West Lafayette, Indiana 47907 JEAN CADET (8), D~partement de Recherche Fondamentale sur la Matidre Condens6e, SESAM/LAN, CEA - Centre d'Etudes Nacldaires de Grenoble, F-38054 Grenoble, France JOHN M. CARNEY (53), Department of BiDchemistry and Molecular Biology, Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma 73104 RAJAGOPAL CHATTOPADHYAYA (5), Bose Institute, 700054 Calcutta, India QIN CHEN (2), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 JOSIANE CILLARD (43), Laboratoire de Biologie Cellulaire et V~gdtale, Facult~ de Pharmacie, 35043 Rennes, France PIERRE CILLARD (43), Laboratoire de Biologie CeUulaire et V6gdtale, Facult~ de Pharmacie, 35043 Rennes, France
SHOSHY ALTUVIA (17), Cell Biology and Metabolism Branch, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892 BRUCE N. AMES (2, 8, 23), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 MARY E. ANDERSON (49, 50), Department of Biochemistry, Cornell University Medical College, New York, New York 10021 MIGUEL ASENSI (35), Departamento de Fisiolo~ia, Universidad de Valencia, 46010 Valencia, Spain LAURA AUGERI (10), Departments of Biochemistry and Radiation Oncology, Rollins Research Center, Emory University School of Medicine, Atlanta, Georgia 30322 PATRICK A. BAEUERLE (14), Institute of Biochemistry, University of Freiburg, D79104 Freiburg, Germany G. BARJA (31), Department of Animal Biology-H (Animal Physiology), Complutense University, 28040 Madrid, Spain SHARMILA BAsu-MODAK (18), Swiss Institute for Experimental Cancer Research, Physical Carcinogenesis Unit, CH-1066 Epalinges, Switzerland CHRISTA BAUMSTARK-KHAN (9), Radiologische Universitgitsklinik, Experimentelle Radiologie und Strahlenbiologie, Universiti~t Bonn, D-53105 Bonn, Germany JOHN S. BERTRAM (19), Cancer Research Center of Hawaii, University of Hawaii, Honolulu, Hawaii 96813 xi
xii
CONTRIBUTORS TO VOLUME 2 3 4
IAN A. COTGREAVE (48), Institute of Envi-
CHERYL A. EDBAUER-NECHAMEN (16), De-
ronmental Medicine, Division of Toxicology, Karolinska Institute, Stockholm 171 77, Sweden DANA R. CRAWFORD (16), Department of Biochemistry and Molecular Biology, The Albany Medical College, Albany, New York 12208 CARROLL E. CROSS (21), Department of Medicine and Physiology, PulmonaryCritical Care Medicine, University of California, Davis Medical Center, Sacramento, California 95817 TOM CURRAN (15), Department of Molecular Oncology and Virology, Roche Institute of Molecular Biology, Roche Research Center, Nutley, NJ 07110 DIPAK K. DAS (40), Cardiovascular Division, Department of Surgery, Surgical Research Center, University of Connecticut School of Medicine, Farmington, Connecticut 06030 JOANNA M. S. DAVIES (16), Department of Biochemistry and Molecular Biology, The Albany Medical College, Albany, New York 12208 KELVIN J. A. DAVIES (16), Department of Biochemistry and Molecular Biology, The Albany Medical College, Albany, New York 12208 MiRAL DIZDAROGLU (1), Chemical Science and Technology Laboratory, National Institute of Standards and Technology, Gaithersburg, Maryland 20899 PAUL W. DOETSCH (3, 10), Departments of Biochemistry and Radiation Oncology, Rollins Research Center, Emory University School of Medicine, Atlanta, Georgia 30322 WULF DROGE (13), Department oflmmunochemistry, Deutsehes Krebsforschungszentrum, D-69120 Heidelberg, Germany MARIA-THERESE DROY-LEFAIX (46), Department of Pharmacology, IPSEN Institute, 75016 Paris, France TIMOTHY R. DUVALL (22), California Regional Primate Research Center, University of California at Davis, Davis, California 95616
partment of Biochemistry and Molecular Biology, The Albany Medical College, Albany, New York 12208 BERND EPE (12), Institute of Pharmacology
and Toxicology, University of Wiirzburg, D-97078 Wiirzburg, Germany Jose M. ESTRELA (35), Departamento de Fi-
siologia, Universidad de Valencia, 46010 Valencia, Spain ROBERT A. FLOYD (6, 53), Department of
Biochemistry and Molecular Biology, Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma 73104 BALZ FREI (23), Whitaker Cardiovascular
Institute, Boston University Medical Center, Boston, Massachusetts 02118 HANS-JOACHIM FREISLEBEN (36), Depart-
ment of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 T. G. GANTCHEV (63), MRC Group in
the Radiation Sciences, Department of Nuclear Medicine and Radiobiology, University of Sherbrooke, Sherbrooke, Qudbec, Canada J1H 5N4 MONIQUE GARDgS-ALBERT (46), Rend Des-
cartes University, Physical Chemistry Laboratory, URA 400 CNRS , 75006 Paris, France ERNST GLINZ (26), Pharma Research, F.
Hoffmann-La Roche Ltd., CH-4002 Basel, Switzerland DAVID V. GOEDDEL (20), Department of
Immunology and Molecular Biology, Genentech, Inc., South San Francisco, California 94080 MATTHEW B. GRISHAM (57), Department of
Physiology and Biophysics, Louisiana State University Medical Cener, Shreveport, Louisiana 71130 ANN E. HAGERMAN (42), Department of Chemistry, Miami University, Oxford, Ohio 45056 EDWARD D. HALL (56), CNS Diseases Re-
search Unit, Upjohn Company, Kalamazoo, Michigan 49001
. o °
CONTRIBUTORS TO VOLUME 234
Xlll
BARRY HALLIWELL (21), Pharmacology
KAr~KI KOMIYAMA (29), Research Division,
Group, King's College, University of London, London SW3 6LX, United Kingdom KRISTA K. HAMILTON (3), Department of Biochemistry, Emory University School of Medicine, Atlanta, Georgia 30322 KRISTA K. HAMILTON (10), Departments of Biochemistry and Radiation Oncology, Rollins Research Center, Emory University School of Medicine, Atlanta, Georgia 30322 JUTTA HEGLER (12), Institute of Pharmacology and Toxicology, University of Wiirzburg, D-97078 Wiirzburg, Germany ERNST S. HENLE (5), Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, Massachusetts 02115 A. HERNANZ (31) Department of Biochemistry, lnsalud La Paz Hospital, 28046 Madrid, Spain LUBICA HOR~KOV.g. (58), Department of Neuropharmacology, Institute of Experimental Pharmacology, Slovak Academy of Sciences, 842 16 Bratislava, Slovakia J. R. S. HOULT (44), Department of Pharmacology, King's College London, London SW3 6LX, United Kingdom MASAYASU INOUE (32), Department of BiDchemistry, Osaka City University Medical School, Abeno, Osaka 545, Japan RICHARD L. JACKSON (51), Wyeth-Ayerst Laboratories, Philadelphia, Pennsylvania 19101 RUCHENG JIN (5), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 VALERIAN E. KAGAN (28, 33, 36, 63), Department of Environmental and Occupational Health, University of Pittsburgh, Pittsburgh, Pennsylvania 15238 S. KHWAIA (33), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 YONG K. KIM (16), Department of Biochemistry and Molecular Biology, The Albany Medical College, Albany, New York 12208
The Kitasato Institute, Minato-ku, Tokyo 108, Japan ALFRED W. KORMANN (26), Vitamins and Fine Chemicals Research, Research and Technology Development, F. HoffmannLa Roche Ltd., CH-4002 Basel, Switzerland KE1KO KOYAMA (32), Department of Biochemistry, Osaka City University Medical School, Abeno, Osaka 545, Japan MURALI C. KRISHNA (59), Radiation Oncology Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892 JEROLD A. LAST (22), California Regional Primate Research Center, University of California at Davis, Davis, California 95616 DIANA M. LEE (52), Free Radical Biology and Aging Research Program, Oklahoma Medical Resarch Foundation, Oklahoma City, Oklahoma 73104 KEUNMYOUNG LEE (3), Department of Botany and Plant Sciences, University of California at Riverside, Riverside, California 92521 GI~RARD LESCOAT (43), Laboratoire de Biologie Cellulaire et Vdg(tale, Facultd de Pharmacie, 35043 Rennes, France ELLEN J. LEVY (49, 50), Department of Anatomy and Cell Biology, State University of New York Health Science Center at Brooklyn, New York, New York 11203 DANIEL C. LIEBLER (27), Department of Pharmacology and Toxicology, College of Pharmacy, University of Arizona Health Sciences Center, Tucson, Arizona 85721 STUART LINN (5), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 MARIA A. LIVREA (39), lstituto de Chimica Biologica, Universit& di Palermo, Policlinico, 90127 Palermo, Italy CHARLES V. LOWRY (16), Department of Biochemistry and Molecular Biology, The Albany Medical College, Albany, New York 12208
xiv
CONTRIBUTORS TO VOLUME 2 3 4
YONGZHANG LUO (5), Department of Bio-
NICHOLAS J. MILLER (24), Free Radical Re-
logical Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, Massachusetts 02115 I. TONG MAK (62), Departments of Medicine and Physiology, George Washington University Medical Center, Washington, D.C. 20037 MASANOBU MANABE (36), Department of Anesthesiology and Resuscitology, Kochi Medical School, Nankoku-shi, Kochi 783, Japan SIMON J. T. MAO (51), Marion Merrell Dow Research Institute, Cincinnati, Ohio 45215 LUCIA MARCOCCI (46, 54), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 AMY M. MARTIN (10), Departments of Biochemistry and Radiation Oncology, Rollins Research Center, Emory University School of Medicine, Atlanta, Georgia 30322 JOHN M. MCCALL (56), Discovery Research, Upjohn Company, Kalamazoo, Michigan 49001 ALTON MEISTER (49, 50), Department of Biochemistry, Cornell University Medical College, New York, New York 10021
search Group, UMDS-dash Guy's Hospital, London SE1 9RT, United Kingdom PETER MOLDI~US (48), Institute of Environmental Medicine, Division of Toxicology, Karolinska Institute, Stockholm 171 77, Sweden OLIVIER H. MORAND (61), Pharma Division, Preclinical Research, F. HoffmanLaRoche Ltd., CH-4002 Basel, Switzerland ISABELLE MOREL (43), Laboratoire de Biologie Cellulaire et V~g~tale, Facult~ de Pharmacie, 35043 Rennes, France MICHELE A. MORONEY (44), Department of Pharmacology, King's College London, London SW3 6LX, United Kingdom PAUL A. MOTCHNIK (23), Xoma Corporation, Berkeley, California 94710 MAIK S. W. OBENDORF (4), Institutfiir Physiologische Chemie 1, Heinrich-HeineUniversitdt, D-40001 Diisseldorf, Germany
CARLOS FREDERICO MARTINS MENCK (11),
Department of Biology, Institute of Biosciences, University of Sao Paulo, CEP 05422-970 Sao Paulo, Brazil CI-IRISTA MICHEL (41), Institutfiir Strahlenbiologie, GSF Forschungszentrum fiir Umwelt und Gesundheit Neuherberg, D85758 Oberschleissheim, Germany SABINE MIHM (13), Department of Gastroenterology and Endocrinology, University of G6ttingen, D-37075 GiSttingen, Germany MASAYUKO MIKI (55), Department of Pediatrics, Osaka Medical College, Takatsuki, Osaka 569, Japan ALLEN M. MILES (57), Department of Physiology and Biophysics, Louisiana State University Medical Center, Shreveport, Louisiana 71130
LESTER PACKER (28, 33, 34, 36, 45, 46, 54),
Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 FEDERICO V. PALLARDO (35), Departamento de Fisiologia, Universidad de Valencia, 46010 Valencia, Spain MIGUEL PAYA (44), Departamento de Fisiologla, Universidad de Valencia, 46100 Valencia, Spain ORS B. RANALDER (26), Pharma Research, F. Hoffmann-La Roche Ltd., CH-4002 Basel, Switzerland JEAN-Luc RAVANAT (8), Ddpartement de Recherche Fondamentale sur la Matidre Condensde, SESAM/LAN, CEA - Centre d'Etudes Nucldaires de Grenoble, F38054 Grenoble, France ROSEMARIE RETTENMAIER (25), Vitamins and Fine Chemicals Research, Research and Technology Development, F. Hoffmann-La Roche Ltd., CH-4002 Basel, Switzerland CATHERINE RICE-EVANS (24), Free Radical Research Group, UMDS-Guy's Hospital, London SE1 9RT, United Kingdom
CONTRIBUTORS TO VOLUME 234
XV
GEORGES RISS (26), Vitamins and Fine
STEEN STEENKEN (58), Max-Planck-lnstitut
Chemicals Research, Research and Technology Development, F. Hoffmann-La Roche Ltd., CH-4002 Basel, Switzerland STEFFEN ROTH (13), Mannheimer Strasse 129, 68309 Mannheim, Germany SHARON L. SALMON (16), Department of Biochemistry and Molecular Biology, The Albany Medical College, Albany, New York 12208 AMRAM SAMUNI (59), Department of Molecular Biology, Hebrew University Medical School, 91010 Jerusalem, Israel MANFRED SARAN (41), Institutfiir Strahlenbiologie, GSF Forschungszentrum fiir Umwelt und Gesundheit Neuherberg, D85758 Oberschleissheirn, Germany JUAN SASTRE (35), Departamento de Fisiologia, Universidad de Valencia, 46010 Valencia, Spain RALF SCHRECK (14), Fred Hutchinson Cancer Center, Seattle, Washington 98104 WILLY SCHLrEP (25), Vitamins and Fine Chemicals Research, Research and Technology Development, F. Hoffmann-La Roche Ltd., CH-4002 Basel, Switzerland WOLFGANG A. SCHULZ (4), Institutfiir Physiologische Chemie 1, Heinrich-HeineUniversit?it, D-40001 Diisseldorf, Germany ABDELHAFID SEKAKI (46), Rend Descartes University, Physical Chemistry Laboratory, URA 400 CNRS , 75006 Paris, France ELENA A. SERBINOVA (33, 34), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 MARK K. SHIGENAGA (2), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 HELMUT SIES (4, 37, 38, 47, 58), Institutffw Physiologische Chemie I, HeinrichHeine-Universitiit, D-40001 Diisseldorf, Germany WILHELM STAHL (38), Institutfi2r Physiologische Chemie I, Heinrich-Heine-Universiti~t, D-40001 Diisseldorf, Germany
far Strahlenchemie, W-4330 Miilheim, Germany GISELA STORZ (17), Cell Biology and Metabolism Branch, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892 O. A. STOYANOVSKY (33, 63), Department of Environmental and Occupational Health, University of Pittsburgh, Pittsburgh, Pennsylvania 15238 ALFRED R. SUNDQUIST (37), Department of Chemistry, University of California at San Diego, San Diego, California 92093 YUICHIRO JUSTIN SUZUKI (45, 54), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 BRIAN K. TARKINGTON (22), California Regional Primate Research Center, University of California at Davis, Davis, California 95616 LUISA TESORIERE (39), Istituto de Chimica Biologica, Universitft di Palermo, Policlinico, 90127 Palermo, Italy MASAHIKO TSUCHIYA (36, 45, 54), Department of Anesthesiology and Resuscitology, Kochi Medical School, Nankokushi, Kochi 783, Japan REX M. TYRRELL (18), Swiss Institute for Experimental Cancer Research, Physical Carcinogenesis Unit, CH-I066 Epalinges, Switzerland J. E. VAN LIER (63), MRC Group in the Radiation Sciences, Department of Nuclear Medicine and Radiobiology, University of Sherbrooke, Sherbrooke, Qudbec, Canada J1H 5N4 GOVlND T. VATASSERY (30), Research Service, and Geriatric Research Education and Clinical Center, V.A. Medical Center, Minneapolis, Minnesota 55417 JOSE VIIZ4A (35), Departamento de Fisiologia, Universidad de Valencia, 46010 Valencia, Spain WILLI WALTHER (26), Pharma Research, F. Hoffmann-La Roche Ltd., CH-4002 Basel, Switzerland
xvi
CONTRIBUTORS TO VOLUME 2 3 4
WILLIAM B. WEGLICKI (62), Departments
of Medicine and Physiology, George Washington University Medical Center, Washington, D.C. 20037 HELEN WISEMAN (60), Departments of Pharmacology and Biochemistry, Royal Free Hopsital School of Medicine, London NW3 2PF, United Kingdom GRACE H. W. WONG (20), Cardiovascular Department, Genetech, Inc., South San Francisco, California 94080 PETER K. WONG (6), Free Radical Biology and Aging Research Program, Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma 73104 STEVEN XANTHOUDAKIS (15), Roche Institute of Molecular Biology, Roche Research Center, Nutley, New Jersey 07110 J. C. YALOWICH (63), Department of Pharmacology, University of Pittsburgh, and
Pittsburgh Cancer Institute, Pittsburgh, Pennsylvania 15238 MASAKAZU YAMAOKA (29), Applied Microbiology Division, National Institute of Bioscience and Human Technology, Tsukuba, Ibaraki 305, Japan MARK T. YATES (51), Marion Merrell Dow Research Institute, Cincinnati, Ohio 45215 HELEN C. YEO (8), Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, California 94720 PAULOS G. YOHANNES (10), Science Division, DeKalb College, Decatur, Georgia 30034 L1-XIN ZHANG (19), Cell Biology Section, Laboratory of Pulmonary Pathobiology, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709
Preface The importance of oxygen-derived radicals, reactive oxygen species, and antioxidants in health and disease is now recognized by every branch of medicine and biological science. Overwhelming evidence indicates that free radicals play a role in most major health problems of the industrialized world, including cardiovascular diseases, cancer, neurological disease, and aging, and that antioxidants play a critical role in wellness, health maintenance, and the prevention of chronic and degenerative diseases. Oxidants also play a role in some aspects of health, as in the oxidative burst of neutrophils and macrophages which allows them to kill foreign organisms. The discovery that endothelial relaxing factor is nitric oxide has provided further evidence of the role of reactive oxygen species in transcellular signaling pathways; the inducible nitric oxide synthetase in macrophages produces large amounts of nitric oxide which are cytotoxic. Transcellular signaling and cytotoxicity have generated enormous interest, not only in nitric oxide, but also in hydrogen peroxide, carbon monoxide, and other oxygen-containing compounds as modulators of cell proliferation and differentiation. Recently, a new branch of these studies has emerged. It is becoming increasingly evident that oxygen radicals and antioxidants have roles in modulating gene expression; e.g., reactive oxygen species affect transcription factors (NFK-B, AP-1) and early growth response genes (c-los, c-jun, etc.). These effects can be important both in normal growth as well as in pathological conditions. The discovery and continued exploration of such actions, as well as clarification of the subtle interactions between oxidants and antioxidants and between various antioxidants themselves, have been the result of new, more sensitive techniques for the detection and quantitation of oxygen radicals in biological systems and the merging of these techniques with the explosive and ever-changing fields of molecular biology and molecular genetics. The enormous array of technologies and new developments has required two new Methods in Enzymology volumes, Oxygen Radicals in Biological Systems (Part C, Volume 233, and Part D, Volume 234), to contain some of the best and most recent technical improvements in the field of oxidants in biological systems. The contributions to these volumes describe methods for the generation and determination of various radical species and antioxidant actions and for the study of the products of their attack on cellular components. xvii
. ° °
XVIII
PREFACE
We express great appreciation to the editorial advisory board--Bruce Ames, Kelvin J. A. Davies, Barry Halliwell, Etsuo Niki, William Pryor, and Helmut Sies--whose advice, suggestions, and contributions have helped these volumes represent the state of the art in new techniques and methods. LESTER PACKER
METHODS IN ENZYMOLOGY VOLUME I. Preparation and Assay of Enzymes
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME II. Preparation and Assay of Enzymes
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME III. Preparation and Assay of Substrates
Edited by SIDNEY P, COLOWICK AND NATHAN O. KAPLAN VOLUME IV. Special Techniques for the Enzymologist
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME V. Preparation and Assay of Enzymes
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VI. Preparation and Assay of Enzymes (Continued) Preparation and Assay of Substrates Special Techniques Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VII. Cumulative Subject Index
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VIII. Complex Carbohydrates
Edited by ELIZABETH F. NEUFELD AND VICTOR GINSBURG VOLUME IX. Carbohydrate Metabolism
Edited by WILLIS A. WOOD VOLUME X. Oxidation and Phosphorylation
Edited by RONALD W. ESTABROOK AND MAYNARD E. PULLMAN VOLUME XI. Enzyme Structure
Edited by C. H. W. HIas VOLUME XII. Nucleic Acids (Parts A and B)
Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XIII. Citric Acid Cycle
Edited by J. M. LOWENSTEIN VOLUME XIV. Lipids
Edited by J. M. LOWENSTEIN VOLUME XV. Steroids and Terpenoids
Edited by RAYMOND B. CLAYTON VOLUME XVI. Fast Reactions
Edited by KENNETH KUSTIN xix
XX
METHODS IN ENZYMOLOGY
VOLUME XVII. Metabolism of Amino Acids and Amines (Parts A and B)
Edited by HERBERT TABOR AND CELIA WHITE TABOR VOLUME XVIII. Vitamins and Coenzymes (Parts A, B, and C)
Edited by DONALD B. McCORMICK AND LEMUEL n . WRIGHT VOLUME X l X . Proteolytic Enzymes
Edited by GERTRUDE E. PERLMANN AND LASZLO LORAND
VOLUME XX. Nucleic Acids and Protein Synthesis (Part C)
Edited by KIVlE MOLDAVEAND LAWRENCEGROSSMAN VOLUME XXI. Nucleic Acids (Part D)
Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XXII. Enzyme Purification and Related Techniques
Edited by WILLIAM B. JAKOBY VOLUME XXIII. Photosynthesis (Part A)
Edited by ANTHONY SAN PIETRO VOLUME XXlV. Photosynthesis and Nitrogen Fixation (Part B)
Edited by ANTHONY SAN PIETRO VOLUME XXV. Enzyme Structure (Part B)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVI. Enzyme Structure (Part C)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVII. Enzyme Structure (Part D)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVIII. Complex Carbohydrates (Part B)
Edited by VICTORGINSBURG VOLUME XXIX. Nucleic Acids and Protein Synthesis (Part E)
Edited by LAWRENCEGROSSMANAND KIVIE MOLDAVE VOLUME XXX. Nucleic Acids and Protein Synthesis (Part F)
Edited by KIVIE MOLDAVEAND LAWRENCEGROSSMAN VOLUME XXXI. Biomembranes (Part A)
Edited by SIDNEY FLEISCHERAND LESTER PACKER VOLUME XXXII. Biomembranes (Part B)
Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME XXXIII. Cumulative Subject Index Volumes I-XXX
Edited by MARTHA G. DENNIS AND EDWARD A. DENNIS VOLUME XXXIV. Affinity Techniques (Enzyme Purification: Part B)
Edited by WILLIAM B. JAKOBY AND MEIR WILCHEK VOLUME XXXV. Lipids (Part B)
Edited by JOHN M. LOWENSTEIN
METHODS iN ENZYMOLOGY
xxi
VOLUME XXXVI. Hormone Action (Part A: Steroid Hormones)
Edited by BERT W. O'MALLEY AND JOEL G. HARDMAN VOLUME XXXVII. Hormone Action (Part B: Peptide Hormones)
Edited by BERT W. O'MALLEY AND JOEL G. HARDMAN VOLUME XXXVIII. Hormone Action (Part C: Cyclic Nucleotides)
Edited by JOEL G. HARDMAN AND BERT W. O'MALLEY VOLUME XXXIX. Hormone Action (Part D: Isolated Cells, Tissues, and Organ Systems) Edited by JOEL G. HARDMAN AND BERT W. O'MALLEY VOLUME XL. Hormone Action (Part E: Nuclear Structure and Function)
Edited by BERT W. O'MALLEY AND JOEL G. HARDMAN VOLUME XLI. Carbohydrate Metabolism (Part B)
Edited by W. A. WOOD VOLUME XLII. Carbohydrate Metabolism (Part C)
Edited by W. A. WOOD VOLUME XLIII. Antibiotics
Edited by JOHN H. HASH VOLUME XLIV. Immobilized Enzymes
Edited by KLAUS MOSBACH VOLUME XLV. Proteolytic Enzymes (Part B)
Edited by LASZLO LORAND VOLUME XLVI. Affinity Labeling
Edited by WILLIAM B. JAKOBY AND MEIR WiLCHEK VOLUME XLVII. Enzyme Structure (Part E)
Edited by C. H. W. HiRS AND SERGE N. TIMASHEEE VOLUME XLVIII. Enzyme Structure (Part F) Edited by C. H. W. HIRS AND SERGE N. TIMASHEEE VOLUME XLIX. Enzyme Structure (Part G)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEEE VOLUME L. Complex Carbohydrates (Part C)
Edited by VICTOR GINSBURG VOLUME LI. Purine and Pyrimidine Nucleotide Metabolism
Edited by PATRICIA A . HOFFEE AND MARY ELLEN JONES VOLUME LII. Biomembranes, (Part C" Biological Oxidations)
Edited by SIDNEY FLEISCHER~AND LESTER PACKER VOLUME LIII. Biomembranes (Part D: Biological Oxidations)
Edited by SIDNEY FLEISCHER' ~ND LESTER PACKER VOLUME LIV. Biomembranes (~?art E: Biological Oxidations)
Edited by SIDNEY FLEISCHER, AND LESTER PACKER
xxii
METHODS IN ENZYMOLOGY
VOLUME LV. Biomembranes (Part F: Bioenergetics)
Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LVI. Biomembranes (Part G: Bioenergetics)
Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LVII. Bioluminescence and Chemiluminescence
Edited by MARLENE A. DELuCA VOLUME LVIII. Cell Culture
Edited by WILLIAM B. JAKOBY AND IRA PASTAN VOLUME LIX. Nucleic Acids and Protein Synthesis (Part G)
Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME LX. Nucleic Acids and Protein Synthesis (Part H)
Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME 61. Enzyme Structure (Part H)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEEE VOLUME 62. Vitamins and Coenzymes (Part D)
Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME 63. Enzyme Kinetics and Mechanism (Part A: Initial Rate and Inhibitor Methods) Edited by DANIEL L. PURICH VOLUME 64. Enzyme Kinetics and Mechanism (Part B: Isotopic Probes and Complex Enzyme Systems) Edited by DANIEL L. PURICH VOLUME 65. Nucleic Acids (Part I)
Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME 66. Vitamins and Coenzymes (Part E)
Edited by DONALD B. McCORMICK AND LEMUEL D. WRIGHT VOLUME 67. Vitamins and Coenzymes (Part F)
Edited by DONALD B. McCoRMICK AND LEMUEL D. WRIGHT VOLUME 68. Recombinant DNA
Edited by RAY Wu VOLUME 69. Photosynthesis and Nitrogen Fixation (Part C)
Edited by ANTHONY SAN PIETRO VOLUME 70. Immunochemical Techniques (Part A)
Edited by HELEN VAN VUNAKIS AND JortN J. LANGONE VOLUME 71. Lipids (Part C)
Edited by JOHN M. LOWENSTEIN VOLUME 72. Lipids (Part D)
Edited by JOHN M. LOWENSTEIN
METHODS IN ENZYMOLOGY
XXlll
VOLUME 73. Immunochemical Techniques (Part B) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 74. Immunochemical Techniques (Part C) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 75. Cumulative Subject Index Volumes XXXI, XXXII, XXXIV-LX Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS VOLUME 76. Hemoglobins Edited by ERALDO ANTONINI, LUIGI ROSSI-BERNARDI, AND EMILIA CHIANCONE
VOLUME 77. Detoxication and Drug Metabolism Edited by WILLIAM B. JAKOBY VOLUME 78. Interferons (Part A) Edited by SIDNEY PESTKA VOLUME 79. Interferons (Part B) Edited by SIDNEY PESTKA VOLUME 80. Proteolytic Enzymes (Part C) Edited by LASZLO LORAND VOLUME 81. Biomembranes (Part H: Visual Pigments and Purple Membranes, I) Edited by LESTER PACKER VOLUME 82. Structural and Contractile Proteins (Part A: Extracellular Matrix) Edited by LEON W. CUNNINGHAM AND DIXIE W. FREDERIKSEN VOLUME 83. Complex Carbohydrates (Part D) Edited by VICTOR GINSBURG VOLUME 84. Immunochemical Techniques (Part D: Selected Immunoassays) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 85. Structural and Contractile Proteins (Part B: The Contractile Apparatus and the Cytoskeleton)
Edited by DIXIE W. FREDERIKSEN AND LEON W. CUNNINGHAM VOLUME 86. Prostaglandins and Arachidonate Metabolites Edited by WILLIAM E. M. LANDS AND WILLIAM L. SMITH VOLUME 87. Enzyme Kinetics and Mechanism (Part C: Intermediates, Stereochemistry, and Rate Studies)
Edited by DANIEL L. PURICH VOLUME 88. Biomembranes (Part I: Visual Pigments and Purple Membranes, II) Edited by LESTER PACKER VOLUME 89. Carbohydrate Metabolism (Part D) Edited by WILLIS A. WOOD
xxiv
METHODS IN ENZYMOLOGY
VOLUME 90. Carbohydrate Metabolism (Part E)
Edited by WILLIS A. WOOD VOLUME 91. Enzyme Structure (Part I)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEEE VOLUME 92. Immunochemical Techniques (Part E: Monoclonal Antibodies and General Immunoassay Methods) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 93. Immunochemical Techniques (Part F: Conventional Antibodies, Fc Receptors, and Cytotoxicity) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 94. Polyamines
Edited by HERBERT TABOR AND CELIA WHITE TABOR VOLUME 95. Cumulative Subject Index Volumes 61-74, 76-80
Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS VOLUME 96. Biomembranes [Part J: Membrane Biogenesis: Assembly and Targeting (General Methods; Eukaryotes)]
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 97. Biomembranes [Part K: Membrane Biogenesis: Assembly and Targeting (Prokaryotes, Mitochondria, and Chloroplasts)]
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 98. Biomembranes (Part L: Membrane Biogenesis: Processing and Recycling)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 99. Hormone Action (Part F: Protein Kinases)
Edited by JACKIE D. CORBIN AND JOEL G. HARDMAN VOLUME 100. Recombinant DNA (Part B)
Edited by RAY WU, LAWRENCE GROSSMAN, AND KIVIE MOLDAVE VOLUME 101. Recombinant DNA (Part C)
Edited by RAY Wu, LAWRENCEGROSSMAN,AND KIVlE MOLDAVE VOLUME 102. Hormone Action (Part G: Calmodulin and Calcium-Binding Proteins)
Edited by ANTHONY R. MEANS AND BERT W. O'MALLEY VOLUME 103. Hormone Action (Part H: Neuroendocrine Peptides)
Edited by P. MICHAELCONN VOLUME 104. Enzyme Purification and Related Techniques (Part C)
Edited by WILLIAMB. JAKOBY VOLUME 105. Oxygen Radicals in Biological Systems
Edited by LESTER PACKER VOLUME 106. Posttranslational Modifications (Part A)
Edited by FINN WOLD AND KIVIE MOLDAVE
METHODS IN ENZYMOLOGY
XXV
VOLUME 107. Posttranslational Modifications (Part B)
Edited by FINN WOLD AND KIVIE MOLDAVE VOLUME 108. Immunochemical Techniques (Part G: Separation and Characterization of Lymphoid Cells) Edited by GIOVANNI DI SABATO, JOHN J. LANGONE, AND HELEN VAN VUNAKIS
VOLUME 109. Hormone Action (Part I: Peptide Hormones)
Edited by LUTZ BIRNBAUMER AND BERT W. O'MALLEY VOLUME 110. Steroids and Isoprenoids (Part A)
Edited by JOHN H. LAW AND HANS C. RILLING VOLUME 111. Steroids and Isoprenoids (Part B)
Edited by JOHN H. LAW AND HANS C. RILLING VOLUME 112. Drug and Enzyme Targeting (Part A)
Edited by KENNETH J. WIDDERAND RALPH GREEN VOLUME 113. Glutamate, Glutamine, Glutathione, and Related Compounds
Edited by ALTON MEISTER VOLUME 114. Diffraction Methods for Biological Macromolecules (Part A)
Edited by HAROLDW. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF VOLUME 115. Diffraction Methods for Biological Macromolecules (Part B)
Edited by HAROLDW. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF VOLUME 116. Immunochemical Techniques (Part H: Effectors and Mediators of Lymphoid Cell Functions) Edited by GIOVANNIDI SABATO,JOHN J. LANGONE, AND HELEN VAN VUNAKIS
VOLUME 117. Enzyme Structure (Part J)
Edited by C. H. W. Hms AND SERGE N. TIMASHEFF VOLUME 118. Plant Molecular Biology
Edited by ARTHUR WEISSBACHAND HERBERT WEISSBACH VOLUME 119. Interferons (Part C)
Edited by SIDNEY PESTKA VOLUME 120. Cumulative Subject Index Volumes 81-94, 96-101 VOLUME 121. Immunochemical Techniques (Part I: Hybridoma Technology and Monoclonal Antibodies)
Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 122. Vitamins and Coenzymes (Part G)
Edited by FRANK CHYTIL AND DONALD B. McCoRMICK VOLUME 123. Vitamins and Coenzymes (Part H)
Edited by FRANK CHYTIL AND DONALD B. McCORMICK VOLUME 124. Hormone Action (Part J: Neuroendocrine Peptides)
Edited by P. MICHAELCONN
xxvi
M E T H O D S IN E N Z Y M O L O G Y
VOLUME 125. Biomembranes (Part M: Transport in Bacteria, Mitochondria, and Chloroplasts: General Approaches and Transport Systems) Edited by SIDNEY FLEISCHERAND BECCA FLEISCHER VOLUME 126. Biomembranes (Part N: Transport in Bacteria, Mitochondria, and Chloroplasts: Protonmotive Force) Edited by SIDNEY FLEISCHERAND BECCA FLEISCHER VOLUME 127. Biomembranes (Part O: Protons and Water: Structure and Translocation) Edited by LESTER PACKER VOLUME 128. Plasma Lipoproteins (Part A: Preparation, Structure, and Molecular Biology) Edited by JERE P. SEGRESTAND JOHN J. ALBERS VOLUME 129. Plasma Lipoproteins (Part B: Characterization, Cell Biology, and Metabolism) Edited by JOHN J. ALBERSAND JERE P. SEGREST VOLUME 130. Enzyme Structure (Part K)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 131. Enzyme Structure (Part L)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 132. Immunochemical Techniques (Part J: Phagocytosis and CellMediated Cytotoxicity) Edited by GIOVANNIDI SABATOAND JOHANNESEVERSE VOLUME 133. Bioluminescence and Chemiluminescence (Part B)
Edited by MARLENEDELUCA AND WILLIAMD. MCELROY VOLUME 134. Structural and Contractile Proteins (Part C: The Contractile Apparatus and the Cytoskeleton) Edited by RICHARDB. VALLEE VOLUME 135. Immobilized Enzymes and Cells (Part B)
Edited by KLAUS MOSBACH VOLUME 136. Immobilized Enzymes and Cells (Part C)
Edited by KLAUS MOSBACH VOLUME 137. Immobilized Enzymes and Cells (Part D)
Edited by KLAUS MOSBACH VOLUME 138. Complex Carbohydrates (Part E)
Edited by VICTOR GINSBURG VOLUME 139. Cellular Regulators (Part A: Calcium- and Calmodulin-Binding Proteins) Edited by ANTHONYR. MEANS AND P. MICHAELCONN VOLUME 140. Cumulative Subject Index Volumes 102-119, 121-134
METHODS IN ENZYMOLOGY
xxvii
VOLUME 141. Cellular Regulators (Part B: Calcium and Lipids)
Edited by P. MICHAEL CONN AND ANTHONY R. MEANS VOLUME 142. Metabolism of Aromatic Amino Acids and Amines Edited by SEYMOUR KAUFMAN VOLUME 143. Sulfur and Sulfur Amino Acids
Edited by WILLIAM B. JAKOBY AND OWEN GRIFEITH VOLUME 144. Structural and Contractile Proteins (Part D: Extracellular Matrix) Edited by LEON W. CUNNINGHAM VOLUME 145. Structural and Contractile Proteins (Part E: Extracellular Matrix)
Edited by LEON W. CUNNINGHAM VOLUME 146. Peptide Growth Factors (Part A)
Edited by DAVID BARNES AND DAVID A. SIRBASKU VOLUME 147. Peptide Growth Factors (Part B)
Edited by DAVID BARNES AND DAVID A. SIRBASKU VOLUME 148. Plant Cell Membranes
Edited by LESTER PACKER AND ROLAND DOUCE VOLUME 149. Drug and Enzyme Targeting (Part B)
Edited by RALPH GREEN AND KENNETH J. WIDDER VOLUME 150. Immunochemical Techniques (Part K: In Vitro Models of B and T Cell Functions and Lymphoid Cell Receptors) Edited by GIOVANNI DI SABATO VOLUME 151. Molecular Genetics of Mammalian Cells Edited by MICHAEL M. GOTTESMAN VOLUME 152. Guide to Molecular Cloning Techniques
Edited by SHELBY L. BERGER AND ALAN R. KIMMEL VOLUME 153. Recombinant DNA (Part D) Edited by RAY Wu AND LAWRENCE GROSSMAN VOLUME 154. Recombinant DNA (Part E)
Edited by RAY Wu AND LAWRENCE GROSSMAN VOLUME 155. Recombinant DNA (Part F) Edited by RAY Wu VOLUME 156. Biomembranes (Part P: ATP-Driven Pumps and Related Transport: The Na,K-Pump) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 157. Biomembranes (Part Q: ATP-Driven Pumps and Related Transport: Calcium, Proton, and Potassium Pumps) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 158. Metalloproteins (Part A) Edited by JAMES F. RIORDAN AND BERT L. VALLEE
.°. XXVIII
METHODS IN ENZYMOLOGY
VOLUME 159. Initiation and Termination of Cyclic Nucleotide Action
Edited by JACKIE D. CORBIN AND ROGER A. JOHNSON VOLUME 160. Biomass (Part A: Cellulose and HemiceUulose)
Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG VOLUME 161. Biomass (Part B: Lignin, Pectin, and Chitin)
Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG VOLUME 162. Immunochemical Techniques (Part L: Chemotaxis and Inflammation) Edited by GIOVANNIDI SABATO VOLUME 163. Immunochemical Techniques (Part M: Chemotaxis and Inflammation) Edited by GIOVANNIDI SABATO VOLUME 164. Ribosomes
Edited by HARRY F. NOLLER, JR., AND KIVlE MOLDAVE VOLUME 165. Microbial Toxins: Tools for Enzymology
Edited by SIDNEY HARSHMAN VOLUME 166. Branched-Chain Amino Acids
Edited by ROBERT HARRIS AND JOHN R. SOKATCH VOLUME 167. Cyanobacteria
Edited by LUSTER PACKER AND ALEXANDER N. GLAZER VOLUME 168. Hormone Action (Part K: Neuroendocrine Peptides)
Edited by P. MICHAELCONN VOLUME 169. Platelets: Receptors, Adhesion, Secretion (Part A)
Edited by JACEK HAWlGER VOLUME 170. Nucleosomes
Edited by PAUL M. WASSARMAN AND ROGER D. KORNBERG VOLUME 171. Biomembranes (Part R: Transport Theory: Cells and Model Membranes)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 172. Biomembranes (Part S: Transport: Membrane Isolation and Characterization)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 173. Biomembranes [Part T: Cellular and Subcellular Transport: Eukaryotic (Nonepithelial) Cells]
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 174. Biomembranes [Part U: Cellular and Subcellular Transport: Eukaryotic (Nonepithelial) Cells]
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 175. Cumulative Subject Index Volumes 135-139, 141-167
METHODS IN ENZYMOLOGY
xxix
VOLUME 176. Nuclear Magnetic Resonance (Part A: Spectral Techniques and Dynamics)
Edited by NORMAN J. OPPENHEIMER AND THOMAS L. JAMES VOLUME 177. Nuclear Magnetic Resonance (Part B: Structure and Mechanism)
Edited by NORMANJ. OPPENHEIMERAND THOMASL. JAMES VOLUME 178. Antibodies, Antigens, and Molecular Mimicry
Edited by JOHN J. LANGONE VOLUME 179. Complex Carbohydrates (Part F)
Edited by VICTOR GINSBURG VOLUME 180. RNA Processing (Part A: General Methods)
Edited by JAMES E. DAHLBERG AND JOHN N. ABELSON VOLUME 181. RNA Processing (Part B: Specific Methods)
Edited by JAMES E. DAHLBERG AND JOHN N. ABELSON VOLUME 182. Guide to Protein Purification
Edited by MURRAY P. DEUTSCHER VOLUME 183. Molecular Evolution: Computer Analysis of Protein and Nucleic Acid Sequences Edited by RUSSELL F. DOOLITTLE VOLUME 184. Avidin-Biotin Technology
Edited by MEIR WILCHEK AND EDWARD A. BAYER VOLUME 185. Gene Expression Technology
Edited by DAVID V. GOEDDEL VOLUME 186. Oxygen Radicals in Biological Systems (Part B: Oxygen Radicals and Antioxidants)
Edited by LESTER PACKER AND ALEXANDER N. GLAZER VOLUME 187. Arachidonate Related Lipid Mediators
Edited by ROBERT C. MURPHY AND FRANKA. FITZPATRICK VOLUME 188. Hydrocarbons and Methylotrophy
Edited by MARY E. LIDSTROM VOLUME 189. Retinoids (Part A: Molecular and Metabolic Aspects)
Edited by LESTERPACKER VOLUME 190. Retinoids (Part B: Cell Differentiation and Clinical Applications)
Edited by LESTER PACKER VOLUME 191. Biomembranes (Part V: Cellular and Subcellular Transport: Epithelial Cells)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 192. Biomembranes (Part W: Cellular and Subcellular Transport: Epithelial Cells) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
XXX
METHODS IN ENZYMOLOGY
VOLUME 193. Mass Spectrometry
Edited by JAMES A. McCLOSKEY VOLUME 194. Guide to Yeast Genetics and Molecular Biology
Edited by CHRISTINEGUTHRIE AND GERALD R. FINK VOLUME 195. Adenylyl Cyclase, G Proteins, and Guanylyl Cyclase
Edited by ROGER A. JOHNSON AND JACKIE D. CORBIN VOLUME 196. Molecular Motors and the Cytoskeleton
Edited by RICHARDB. VALLEE VOLUME 197. Phospholipases
Edited by EDWARD A. DENNIS VOLUME 198. Peptide Growth Factors (Part C)
Edited by DAVID BARNES, J. P. MATHER, AND GORDON H. SATO VOLUME 199. Cumulative Subject Index Volumes 168-174, 176-194 (in preparation) VOLUME 200. Protein Phosphorylation (Part A: Protein Kinases: Assays, Purification, Antibodies, Functional Analysis, Cloning, and Expression) Edited by TONY HUNTER AND BARTHOLOMEWM. SEFTON VOLUME 201. Protein Phosphorylation (Part B: Analysis of Protein Phosphorylation, Protein Kinase Inhibitors, and Protein Phosphatases) Edited by TONY HUNTER AND BARTHOLOMEWM. SEFTON VOLUME 202. Molecular Design and Modeling: Concepts and Applications (Part A: Proteins, Peptides, and Enzymes) Edited by JOHN J. LANGONE VOLUME 203. Molecular Design and Modeling: Concepts and Applications (Part B: Antibodies and Antigens, Nucleic Acids, Polysaccharides, and Drugs) Edited by JOHN J. LANGONE VOLUME 204. Bacterial Genetic Systems
Edited by JEFFREY H. MILLER VOLUME 205. Metallobiochemistry (Part B: Metallothionein and Related Molecules) Edited by JAMES F. RIORDANAND BERT L. VALLEE VOLUME 206. Cytochrome P450
Edited by MICHAELR. WATERMANAND ERIC F. JOHNSON VOLUME 207. Ion Channels Edited by BERNARDO RUDY AND LINDA E. IVERSON VOLUME 208. P r o t e i n - D N A
Interactions
Edited by ROBERT T. SAUER VOLUME 209. Phospholipid Biosynthesis Edited by EDWARD A. DENNIS AND DENNIS E. VANCE
METHODS IN ENZYMOLOGY
xxxi
VOLUME 210. Numerical Computer Methods Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 211. DNA Structures (Part A: Synthesis and Physical Analysis of DNA)
Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG VOLUME 212. DNA Structures (Part B: Chemical and Electrophoretic Analysis of DNA)
Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG VOLUME 213. Carotenoids (Part A: Chemistry, Separation, Quantitation, and Antioxidation) Edited by LESTER PACKER VOLUME 214. Carotenoids (Part B: Metabolism, Genetics, and Biosynthesis) Edited by LESTER PACKER VOLUME 215. Platelets: Receptors, Adhesion, Secretion (Part B) Edited by JACEK J. HAWIGER VOLUME 216. Recombinant DNA (Part G) Edited by RAY Wu VOLUME 217. Recombinant DNA (Part H) Edited by RAY Wu VOLUME 218. Recombinant DNA (Part I) Edited by RAY Wu VOLUME 219. Reconstitution of Intracellular Transport Edited by JAMES E. ROTHMAN VOLUME 220. Membrane Fusion Techniques (Part A) Edited by NEJAT DOZGUNE~ VOLUME 221. Membrane Fusion Techniques (Part B) Edited by NEJAT Dt2ZGONE~ VOLUME 222. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement Activation (Part A: Mammalian Blood Coagulation Factors and Inhibitots) Edited by LASZLO LORAND AND KENNETH G. MANN VOLUME 223. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement Activation (Part B: Complement Activation, Fibrinolysis, and Nonmammalian Blood Coagulation Factors) Edited by LASZLO LORAND AND KENNETH G. MANN VOLUME 224. Molecular Evolution: Producing the Biochemical Data Edited by ELIZABETH ANNE ZIMMER, THOMAS J. WHITE, REBECCA L. CANN, AND ALLAN C. WILSON
VOLUME 225. Guide to Techniques in Mouse Development Edited by PAUL M. WASSARMANAND MELVIN L. DEPAMPHILIS
xxxii
M E T H O D S IN E N Z Y M O L O G Y
VOLUME 226. Metallobiochemistry (Part C: Spectroscopic and Physical Methods for Probing Metal Ion Environments in Metalloenzymes and Metalloproteins) Edited by JAMES F. RIORDANAND BERT L. VALLEE VOLUME 227. Metallobiochemistry (Part D: Physical and Spectroscopic Methods for Probing Metal Ion Environments in Metalloproteins) Edited by JAMES F. RIORDANAND BERT L. VALLEE VOLUME 228. Aqueous Two-Phase Systems Edited by HARRY WALTERAND GOTE JOHANSSON VOLUME 229. Cumulative Subject Index Volumes 195-198, 200-227 (in preparation) VOLUME 230. Guide to Techniques in Glycobiology
Edited by WILLIAMJ. LENNARZAND GERALD W. HART VOLUME 231. Hemoglobins (Part B: Biochemical and Analytical Methods)
Edited by JOHANNES EVERSE, KIM D. VANDEGRIFFAND ROBERT M. WINSLOW VOLUME 232. Hemoglobins (Part C: Biophysical Methods)
Edited by JOHANNES EVERSE, KIM D. VANDEGRIFFAND ROBERT M. WINSLOW VOLUME 233. Oxygen Radicals in Biological Systems (Part C)
Edited by LESTER PACKER VOLUME 234. Oxygen Radicals in Biological Systems (Part D) Edited by LESTER PACKER VOLUME 235. Bacterial Pathogenesis (Part A: Identification and Regulation of Virulence Factors) Edited by VIRGINIAL. CLARKAND PATRIKM. BAVOIL VOLUME 236. Bacterial Pathogenesis (Part B: Integration of Pathogenic Bacteria with Host Cells) Edited by VIRGINIAL. CLARKAND PATRIKM. BAVOIL VOLUME 237. Heterotrimeric G Proteins
Edited by RAVI IYENGAR VOLUME 238. Heterotrimeric G-Protein Effectors (in preparation)
Edited by RAVI IYENGAR VOLUME 239. Nuclear Magnetic Resonance (Part C) (in preparation) Edited by THOMAS L. JAMES AND NORMANJ. OPPENHEIMER VOLUME 240. Numerical Computer Methods (Part B) (in preparation)
Edited by MICHAELL. JOHNSON AND LUDWIG BRAND VOLUME 241. Retroviral Proteases (in preparation)
Edited by LAWRENCEC. Kuo AND JULES A. SHAFER VOLUME 242. Neoglycoconjugates (Part A: Synthesis) (in preparation) Edited by Y. C. LEE AND REIKO T. LEE VOLUME 243. Inorganic Microbial Sulfur Metabolism (in preparation)
Edited by HARRY D. PECK, JR., AND JEAN LEGALL
METHODS
IN
ENZYMOLOGY
..° XXXIII
VOLUME 244. Proteolytic Enzymes: Serine and Cysteine Peptidases (in preparation) Edited by ALAN J. BARRETT VOLUME 245. Extracellular Matrix Components (in preparation) Edited by ERICKIRUOSLAHTI AND EVA ENGVALL VOLUME 246. Biochemical Spectroscopy (in preparation) Edited by KENNETH SAUER VOLUME 247. Neoglycoconjugates (Part B, Biomedical Applications) (in preparation) Edited by Y. C. LEE AND REIKO T. LEE VOLUME 248. Proteolytic Enzymes: Aspartic and Metallo Peptidases (in preparation) Edited by ALAN J. BARRETT VOLUME 249. Enzyme Kinetics and Mechanism (Part D) (in preparation) Edited by DANIEL L. PURICH
[1]
DETERMINING OXIDATIVEDNA DAMAGEBY GC/MS
3
[1] Chemical D e t e r m i n a t i o n o f Oxidative D N A D a m a g e by Gas C h r o m a t o g r a p h y - M a s s Spectrometry B y MIRAL DIZDAROGLU
Introduction Oxidative DNA damage produced by free radicals or other DNAdamaging agents has been implicated to play a role in mutagenesis, carcinogenesis, reproductive cell death, and aging.l Oxygen-derived species such as superoxide radical (O2-) and H202 are generated in all aerobic cells. 1,2 Excess generation of these species by endogenous sources or exogenous sources (e.g., redox-cyclic drugs, ionizing radiation) may cause damage to cellular DNA by a variety of mechanisms): The toxicity of these species in vivo, however, is thought to result from their metal ion-catalyzed conversion to the highly reactive hydroxyl radical (-OH). l Ionizing radiation can also produce .OH in cells and tissues by interacting with cellular water. 3 Hydroxyl radical causes formation of a number of modified bases and sugars in DNA, and DNA-protein cross-links in nucleoprotein) -7 A number of these lesions are also produced by nonradical mechanisms.8-15 For understanding of the role of oxidative DNA damage in carcinogenesis i B. Halliwell and J. M. C. Gutteridge, " F r e e Radicals in Biology and Medicine," 2nd ed. Oxford Univ. Press (Clarendon), Oxford, 1989. 2 I. Fridovich, Arch. Biochem. Biophys. 247, 1 (1986). 3 C. von Sonntag, "The Chemical Basis of Radiation Biology," pp. 116-166, 221-294. Taylor & Francis, London, 1987. 4 B. Halliwell and O. I. Aruoma, FEBS Lett. 281, 9 (1991). 5 N. L. Oleinick, S. Chiu, N. Ramakrishnan, and L. Xue, Br. J. Cancer 55, Suppl. 8, 135 (1987). 6 M. Dizdaroglu, Free Radical Biol. Med. 10, 225 (1991). 7 K. Frenkel, Pharmacol. Ther. 53, 127 (1992). 8 M. Dizdaroglu, E. Holwitt, M. P. Hagan, and W. F. Blakely, Biochem. J. 235, 531 (1986). 9 R. A. Floyd, M. S. West, K. L. Eneff, and J. E. Schneider, Arch. Biochem. Biophys. 273, 106 (1989). l0 S. Steenken, Chem. Rev. 89, 503 (1989). ii S. A. Akman, J. H. Doroshow, and M. Dizdaroglu, Arch. Biochem. Biophys. 282, 202 (1990). 12 D. J. Deeble, M. N. Schuchmann, S. Steenken, and C. von Sonntag, J. Phys. Chem. 94, 8186 (1990). J3 D. Angelov, M. Berger, J. Cadet, N. Getoff, E. Keskinova, and S. Solar, Radiat. Phys. Chem. 37, 717 (1991). 14 S. Boiteux, E. Gajewski, J. Laval, and M. Dizdaroglu, Biochemistry 31, 106 (1992). i5 H. Sies and C. F. M. Menck, Mutat. Res. 275, 367 (1992).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
4
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[l]
and other biological processes, it is essential to chemically characterize and quantify DNA lesions. In this chapter, we describe characterization and quantification of oxidative DNA damage by the technique of gas chromatography-mass spectrometry (GC/MS). Materials and Methods Reagents and enzymes used in this methodology are available commercially from a number of suppliers.16 Reference Compounds. Isobarbituric acid (5-hydroxyuracil), 5,6-dihydrothymine, 6-azathymine, 8-azaadenine, 5,6-dihydrouracil, alloxan, 5-(hydroxymethyl)uracil, 4,6-diamino-5-formamidopyrimidine, 8-bromoadenine, isoguanine (2-hydroxyadenine), 2,5,6-triamino-4-hydroxypyrimidine, and xanthine-l,3-15N2 are available from Sigma Chemical Company (St. Louis, MO), and 8-hydroxyguanine from Schweizerhall Inc. (formerly Chemical Dynamics Corporation) (Piscataway, N J). Thyminec~,ot,ot,6-2H4 is available from Merck & Co. Inc./Isotopes (St. Louis, MO). 8-Hydroxyadenine and 2,6-diamino-4-hydroxy-5-formamidopyrimidine are synthesized by treatment of 8-bromoadenine and 2,5,6-triamino4-hydroxypyrimidine with formic acid, respectively. 17,18cis-Thymine glycol is obtained by treatment of thymine with osmium tetroxide. 8 5-Hydroxy-5-methylhydantoin is obtained by the reaction between pyruvic acid and u r e a . 19 Isodialuric acid (5,6-dihydroxyuracil) is synthesized by oxidation of isobarbituric acid with bromine. E°5-Hydroxyhydantoin is obtained by treatment of alloxan with formic a c i d . E1 These and other reference compounds, and their stable isotope-labeled analogs, which are dealt with in this chapter, are available on a custom-synthesis basis from the Chemical Synthesis and Analysis Laboratory of Program Resources Inc./Dyncorp, National Cancer Institute-FCRD (Frederick, MD). The following stable isotope-containing analogs of modified DNA bases have also become available: 5,6-dihydrothymine-l,3JSNE-2J3C, 5,6-
dihydrouracil-l,3JSNE-EJ3C, 5-hydroxy-5-methylhydantoin-l,3JSNE-EJ3C, 16 Certain commercial equipment or materials are identified in this chapter in order to specify adequately the experimental procedure. Such identification does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose. 17 M. Dizdaroglu and D. S. Bergtold, Anal. Biochem. 156, 182 (1986). 18 L. F. Cavalieri and A. Bendich, J. Am. Chem. Soc. 72, 2587 (1950). ~9 S. Murahashi, H. Yuki, K. Kosai, and F. Doura, Bull. Chem. Soc. Jpn. 39, 1559 (1966). 2o R. Behrend and O. Roosen, Justus Liebig's Ann. Chem. 251, 235 (1889). 21 M. Dizdaroglu, FEBS Lett. 315, 1 (1993).
[1]
DETERMINING OXIDATIVE D N A DAMAGE BY G C / M S
5
alloxan-1,3-15N2-2,4-13C2, 5-hydroxyhydantoin-1,3A5N2-2,433C 2, 5-hydroxyuracil-l,3JSN2-2J3C , 5-(hydroxymethyl)uracil-2,433C2-a,a-2H 2 , 5-hydroxycytosine-l,3A5N2-2A3C, cis-thymine glycol-a,a,a,6-2H4, 5,6dihydroxyuracil-I,3J5N2-2J3C (isodialuric acid-l,3A5N2-233C), 4,6-diamino-5-formamidop yrimidine-l ,3-15N2-2-~3C-(5-formamidoA5 N,2 H ) , 8-hydroxyadenine-l,3,7-tSN3-2,8J3C2, 2,6-diamino-4-hydroxy-5-formamidopyrimidine-l,3J5N2-(5-amino-15N)-2A3C, and 8-hydroxyguanine-l,3JSN 2(2-aminoA5N)-233C. The synthesis of 5-(hydroxymethyl)uracil-2-13C-5 2H2-6-2H has also been reported elsewhere. 22 Most compounds listed above are soluble in water. 8-Hydroxyguanine and 8-hydroxyguanine-l,3-t5N2-(2-amino35N)-233C are not completely soluble. However, complete solubility can be obtained by increasing the pH of the solutions to 9.5 with dilute NaOH and then stirring the solutions for several hours at room temperature.
Hydrolysis The preparation of DNA or nucleoprotein samples for analysis by GC/MS involves hydrolysis followed by derivatization. DNA can be hydrolyzed by either acidic hydrolysis or enzymatic hydrolysis. Acidic Hydrolysis. Acidic hydrolysis cleaves the glycosidic bonds between bases and sugar moieties in DNA, releasing intact and modified bases. Formic acid is well suited for hydrolysis of DNA. z3-25 In the case of DNA-protein cross-links, nucleoprotein is hydrolyzed by the standard method of protein hydrolysis using 6 M HC1. By this procedure, peptide bonds in proteins as well as glycosidic bonds in DNA are cleaved to release DNA base-amino acid c r o s s - l i n k s . 26'27 The stability of numerous pyrimidine- and purine-derived modified DNA bases and their release from DNA have been studied under various conditions of formic acid hydrolysis, because this information is important for the assessment of the accuracy of the DNA damage measurement. 25,28 It was found that most of the modified bases are stable under all studied conditions of hydrolysis, and only a few undergo partial destruction, depending on the concentration of formic acid. 25Furthermore, the possibility that some of the modified bases may be formed by acidic treatment in 22 Z. Djuric, D. A. Luongo, and D. A. Harper, Chem. Res. Toxicol. 4, 687 (1991). 23 G. R. Wyatt and S. S. Cohen, Biochem. J. 55, 774 (1953). 24 M. Dizdaroglu, Anal. Biochem. 144, 593 (1985). z5 Z. Nackerdien, R. Olinski, and M. Dizdaroglu, Free Radical Res. Commun. 16, 259 (1992). 26 E. Gajewski, A. F. Fuciarelli, and M. Dizdaroglu, Int. J. Radiat. Biol. 54, 445 (1988). 27 M. Dizdaroglu, E. Gajewski, P. Reddy, and S. A. Margolis, Biochemistry 28, 3625 (1989). 28 A. F. Fuciarelli, B. J. Wegher, E. Gajewski, M. Dizdaroglu, and W. F. Blakely, Radiat. Res. 119, 219 (1989).
6
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[1]
DNA from corresponding intact bases has been investigated. The results indicated that formic acid caused no significant formation of modified bases under the conditions used. As a conclusion of these studies, formic acid at a concentration of 60% (v/v) has been found to be optimal for DNA hydrolysis. 25 The following procedure is used for DNA hydrolysis. An aliquot of DNA (10 to 100/zg) is treated with 0.5 ml of formic acid (60%) in evacuated and sealed tubes at 140° for 30 min. The sample is then transferred to a vial, frozen in liquid nitrogen, and lyophilized. If chromatin instead of pure DNA is to be hydrolyzed for detection of modified bases, the same hydrolysis procedure is followed. It should be pointed out that formic acid hydrolysis causes deamination and dehydration of cytosine-derived products as follows: cytosine glycol yields a mixture of 5-hydroxycytosine and 5-hydroxyuracil, the former by dehydration and the latter by dehydration and deamination. 8 5,6-Dihydrocytosine, 5-hydroxy-6-hydrocytosine and 5,6-dihydroxycytosine deaminate to give 5,6-dihydrouracil, 5-hydroxy-6-hydrouracil and 5,6-dihydroxyuracil, respectively. 5-Hydroxyhydantoin, which has been identified in the past as a product of cytosine, 29 results from acid-induced decarboxylation of alloxan. 21 For detection of DNA-protein cross-links, nucleoprotein containing 100 ~g DNA is hydrolyzed with 0.5 ml of 6 M HCI in evacuated and sealed tubes at 120° for 6 hr. Subsequently, the sample is transferred into a vial, frozen in liquid nitrogen and lyophilized. Enzymatic Hydrolysis. This type of hydrolysis has been discussed elsewhere in detail. 3° The following procedure can be used to hydrolyze DNA to nucleosides. An aliquot (0.1 mg) of DNA is incubated in 0.5 ml of 10 mM Tris-HCl buffer, pH 8.5 (containing 2 mM MgC12), with deoxyribonuclease I (100 units), spleen exonuclease (0.01 unit), snake venom exonuclease (0.5 units), and alkaline phosphatase (10 units) at 37 ° for 24 hr. The sample is then transferred to a vial, frozen in liquid nitrogen, and lyophilized. The drawback of enzymatic hydrolysis is that the products of the 2'-deoxycytidine moiety in DNA cannot be readily analyzed by GC/MS because of the poor gas chromatographic properties of cytidine derivatives. 31 On the other hand, generally less volatile trimethylsilyl [(CH3)3Si] derivatives of modified purine 2'-deoxynucleosides can be analyzed successfully by GC/MS.31'32 Deamination of 2'-deoxyadenosine products may 29 M. Polverelli and R. Troule, Z. Naturforsch., C: Biosci. 29C, 12 (1974). 3o p. F. Crain, this series, Vol. 193, p. 782. 3z M. Dizdaroglu, J. Chromatogr. 367, 357 0986). 32 M.-L. Dirksen, W. F. Blakely, E. Holwitt, and M. Dizdaroglu, Int. J. Radiat. Biol. 54, 195 (1988).
[1]
DETERMINING OXIDATIVEDNA DAMAGEBY GC/MS
7
occur during enzymatic hydrolysis owing to contaminating deaminase activity in the enzymes. Removal of excess salt from the hydrolyzates and removal of deaminases from the enzymes may prevent problems associated with analysis of 2'-deoxycytidine products and deamination of 2'-deoxyadenosine products. 3°'33
Derivatization DNA bases, nucleosides, and DNA base-amino acid cross-links are not sufficiently volatile for gas chromatography, and thus must be converted to volatile derivatives. For this purpose, trimethylsilylation is the mode of derivatization most frequently used. 34 (tert)-Butyldimethylsilylation is also u s e d . 33'35'36 The following procedure can be used for trimethylsilylation. Lyophilized hydrolyzates of DNA or nucleoprotein samples containing 0.01-0.1 mg of DNA are heated with 0.1 ml of a mixture of bis(trimethylsilyl)trifluoroacetamide (containing I% trimethylchlorosilane) and acetonitrile (4 : 1, v/v) at 140° for 30 rain in polytetrafluoroethylene-capped vials (sealed under dry nitrogen). The amounts of the reagents can be modified according to the amount of DNA. After derivatization, samples are cooled to room temperature. Without any further treatment, an aliquot (e.g., l tzl) of each derivatized sample is injected into the injection port of the gas chromatograph.
Instrumentation
A GC/MS instrument equipped with a capillary inlet system and a computer data system can be used. Data presented and reviewed here were obtained on commercial quadrupole mass spectrometers. Fusedsilica capillary columns are used for separation of derivatized hydrolyzates of DNA or nucleoprotein. These types of columns provide high inertness, excellent separation efficiency, and measurement of high sensitivity. Fused-silica capillary columns coated with cross-linked 5% phenyl methylsilicone gum phase appear to be best for the p u r p o s e . 24'37 Column length may vary depending on the type of analysis. A column 12.5 m in length (0.2 mm internal diameter, 0.33 tzm film thickness) is generally used for analysis of derivatized bases. A shorter column (e.g., 8 m, 0.2 mm internal 33 p. F. Crain, Mass Spectrom. Rev, 9, 505 (1990). 34 K. H. Schram, this series, Vol. 193, p. 791. 35 M. A. Quilliam and J. B. Westmore, Anal. Chem. 50, 59 (1978). 36 M. Dizdaroglu, BioTechniques 4, 536 (1986). 37 M. Dizdaroglu, J. Chromatogr. 295, 103 (1984).
8
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[1]
diameter, 0.11 /~m film thickness) serves best for analysis of derivatized DNA base-amino acid cross-links. Helium (ultrahigh purity) is used as the carrier gas. The split mode of injection is used to avoid overloading the column. The split ratio (i.e., ratio of the carrier gas flow through the splitter vent to carrier gas flow through the column) is adjusted according to the DNA amount in samples. An aliquot of DNA hydrolyzates corresponding to approximately 0.1 to 0.4 ~g of DNA on the GC column after splitting of the injected sample is generally sufficient. The injection port of the gas chromatograph and the GC/MS interface are kept at 250°. The temperature of the ion source of the mass spectrometer is usually kept at around 200° . Some instruments may permit variation of this temperature zone. The glass liner in the injection port of the gas chromatograph is filled with silanized glass wool, which allows homogeneous vaporization of injected samples. Electronionization (El) mode at 70 eV has been used to record mass spectra and to perform selected-ion monitoring in studies presented or reviewed here.
Gas Chromatography-Mass Spectrometry Free Bases. Gas chromatography on a fused-silica capillary column (usually 12.5 m long) permits separation of (CH3)3Si derivatives of a large number of modified bases from one another and from the four intact DNA bases in a single analysis. 24 Compounds eluting from the GC column are ionized in the ion source and then analyzed by the mass analyzer of the mass spectrometer) 8 Electron-ionization mass spectra of (CH3)3Si derivatives of modified DNA bases provide considerable structural detail that can be used for unequivocal identification. These mass spectra are characterized by an intense molecular ion (M "+ion), an intense (M - 15) ÷ ion, and other characteristic i o n s , 24'37 as are those of the intact bases, a9'4° The (M - 15) + ion results from the loss of a methyl radical from the M .+ ion. 39'4° In some instances, an intense (M - l) ÷ ion resulting from loss of an H atom from the M .+ ion also appears in the m a s s spectra. 24'37 Nucleosides. Trimethylsilyl derivatives of modified 2'-deoxynucleosides follow the same fragmentation patterns as those of other nucle-
38 j. T. Watson, "Introduction to Mass Spectrometry," Chapters 1 and 3. Raven, New York, 1985. 39 E. White, V. P. M. Krueger, and J. A. McCIoskey, J. Org. Chem. 37, 430 (1972). 40 j. A. McCloskey, in "Basic Principles in Nucleic Acid Chemistry" (P. O. P. Ts'o, ed.), Vol. l, p. 209. Academic Press, New York, 1974.
[1]
DETERMINING OXIDATIVE D N A DAMAGE BY G C / M S
9
osides/1,42 Prominent ions are the (base + H) + ion [(B + 1) + ion] and the (base + H - CH3) + ion, whereas the M .+ and the (M - 15) + ions are of low intensity. 31For example, in the mass spectra of (CH3)3Si derivatives of 8-hydroxy-2'-deoxyguanosine and 8-hydroxy-2'-deoxyadenosine, the (B + 1) + ion appears as the most prominent ion (100% relative intensity) owing to stabilization through an electron-donating substituent at C-8 of the purine ring. Trimethylsilyl derivatives of 8,5'-cyclopurine 2'-deoxynucleosides provide partly different fragmentation patterns from those of other nucleosides. 32'43These spectra are characterized by prominent ions containing the base plus portions of the sugar moiety and by an intense M "+ ion, most likely due to stabilization by the increased number of rings in the molecule. 44 DNA Base-Amino Acid Cross-links. Mass spectra of (CH3)3Si derivatives of DNA base-amino acid cross-links contain M .+ and (M - 15) + ions and other characteristic ions resulting from typical fragmentations of base and amino acid moieties, z6'~7'37 For example, the most prominent ion (m/z 448) in the mass spectrum of the (CH3)3Si derivatives of the thymine-tyrosine cross-link results from cleavage of the bond between the a and/3 carbons of the tyrosine moiety accompanied by an H atom transfer [(M - 218 + 1) + ion]. The high abundance of this ion is due to resonance stabilization through the aromatic ring. This cleavage, which is typical of (CH3)3Si derivatives of amino acids, also accounts for the intense m/z 218 ion when the charge is retained on the a carbon without an H atom transfer. In the case of DNA base-aliphatic amino acid crosslinks, these fragmentations also occur. However, an ion arising from loss of "CO~Si(CH3)3 from the M "+ ion generally appears as one of the most prominent ions in the mass spectra, whereas the abundance of the (M 218 + 1) + ions depends on the aliphatic amino acid residue. Figures 1 and 2 illustrate the structures of modified DNA bases and nucleosides and some DNA base-amino acid cross-links that can be measured by the use of the GC/MS technique. These compounds are formed in DNA or nucleoprotein by free radicals or other agents causing oxidative damage. 6,45 DNA base-amino acid cross-links involving thymine and the amino acids glycine, alanine, valine, leucine, isoleucine, and threonine, 41 H. Pang, K. H. Schram, D. L. Smith, S. P. Gupta, L. B. Towsend, and J. A. McCloskey, J. Org. Chem. 47, 3923 (1982). 42 j. A. McCloskey, this series, Vol. 193, p. 825. 43 M. Dizdaroglu, Biochem. J. 238, 247 (1986). 44 F. W. McLafferty, "Interpretation of Mass Spectra." Univ. Sci. Books, Mill Valley, California, 1980. 45 M. Dizdaroglu, Mutat. Res. 275, 331 (1992).
10
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
o
,~= •It"
z
o
~
-,-
o ~"
~
}
,~o
==
/ / ~
=
0
t.
Eta
== =
.=. ~;
"o
0
z: :ooz
"
:
= .d==
~ - o o ~-
.:
~
~.
_
~
~
~<~
-
~
Z
=._~o =
== o
-~'~
=
-~.
~!
==~o
0=~-=
=
"1"
z
-r
=
~
0
z0
o
::
.=
=... =:
[1]
=.
z
=
2
:~ ~ - - ~ / z
~._~ 6 =
0 0 0
~"
0
~
.~
•-
.j.
~= =%
"
o0
i
..Q ~ ".'"~ e,I ~-
.~
i. =
=
~==
~-~
[1]
DETERMINING OXIDATIVE D N A DAMAGE BY G C / M S
I1
which are also measurable by this technique, are not illustrated in Fig. 2. It should be noted that the exact structure of the thymine-tyrosine crosslink has been determined. Structures ofthymine-lysine and cytosine-tyrosine cross-links illustrated in Fig. 2 have been proposed on the basis of their mass spectral fragmentation patterns. 6 Measurement at Low Analyte Concentrations Identification and quantification of components of a complex mixture at low concentrations are generally carried out using GC/MS with selected-ion monitoring ( S I M ) . 46 When using this mode for identification, knowledge of the mass spectrum and the retention time of the analyte is required. A number of characteristic ions of an analyte are monitored by the mass spectrometer during the time period in which the analyte elutes from the GC column. If the analyte is present in the mixture, signals of the monitored ions will line up at its expected retention time. Subsequently, a partial spectrum is obtained on the basis of the monitored ions and their relative abundances. This spectrum is then compared with that of the authentic compound for unequivocal identification. For this purpose, the mass spectrum of the authentic compound should be recorded under the same tuning conditions of the mass spectrometer as are used to perform the SIM. This is because differences in the relative abundances of ions may occur depending on the tuning conditions of the mass spectrometer. Retention times of analytes also play an important role in reliable identification in addition to simultaneous measurement of masses because gas chromatography on fused-silica capillary columns permits measurement of retention times with great accuracy and precision. Quantification. The GC/MS-SIM technique also permits accurate quantification of components of a complex mixture.46 This is achieved by adding an aliquot of a suitable internal standard to aliquots of DNA samples at an early stage such as prior to hydrolysis. In mass spectrometry, a stable isotope-labeled analog of an analyte can be used as the internal standard. 46 This procedure is called isotope-dilution mass spectrometry. Because of the essentially same physical and chemical properties of the analyte and its analog, the procedure permits compensation for possible losses of the analyte during sample preparation and GC/MS analysis. If stable isotope-labeled analogs are not available, structurally similar compounds may be used as internal standards. 38 In the past, such compounds have been used for quantification of modified DNA bases. 6'z5'28 46 j. T. Watson, this series, Vol. 193, p. 86.
12
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[1]
•r
o
r~
"i 0
~
.,~ g g g=
=° ~
_
~._~,
,.,-
.,:_-i,
~
~
~
~o (J--(J--Z
co
< Z
-le
=-4
| •~ ! -
z
o -r -r
Y~ ,.
~.g
o= -r
,~
o
~
L T ~
L~ {D
[1]
DETERMINING OXIDATIVE D N A DAMAGE BY G C / M S
13
Recently, stable isotope-labeled analogs of modified DNA bases have become available (see Materials and Methods). 21,22 Mass spectral fragmentation patterns of stable isotope-labeled analogs are similar to those of corresponding unlabeled compounds. Masses of most ions in the mass spectra of labeled analogs are shifted to higher masses according to the isotope contents. 2]'22 Typical examples of mass spectra are illustrated in Fig. 3. Here are shown the mass spectra of (CH3)3Si derivatives of 4,6-diamino-5-formamidopyrimidine (Fig. 3A) and 4,6-diamino-5-formamidopyrimidine-tSN 3 ,13C,2H(Fig. 3B). In Fig. 3A, the M +-, the (M - 1) +, and the (M - 15) + ions appear at m/z 369, 368, and 354, respectively. The intense ion at m/z 280 results from the loss of •OSi(CH3)3 (89 Da) from the M .+ i o n . 24 In Fig. 3B, the masses of these ions are shifted by 5 Da to 374, 373,359, and 285 Da, respectively. The ion at m/z 372 may result from the M +. ion by loss of an 2H atom located at the formyl group of the molecule. Ions at m/z 73 and 147 in Fig. 3A,B are commonly observed with (CH3)3Si derivatives. 39 Trimethylsilyl derivatives of modified DNA bases coelute with those of their ]3C- and ]SN-labeled analogs, indicating no isotope effect on elution behaviors of corresponding compounds. 2] However, if a labeled analog contains several 2H atoms such as thymine glycol-2H4, a slight resolution from the unlabeled material is observed, 21indicating a well-known isotope
'ee]l~
A 38
40
147
18~
354
19e
15~
28~
m/z
258
88
388
35~
5
I~
15~
2~8
m/z
258
3~8
35~
FIG. 3. (A) Electron-ionization mass spectrum of the (CH3)3Si derivative of 4,6-diamino5-formamidopyrimidine; (B) EI mass spectrum of the (CH3)3Si derivative of 4,6-diamino-5formamidopyrimidineJSNs,uC,2H.(From Dizdaroglu. 21)
14
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[1]
effect on the elution behavior of 2H-containing compounds. Figure 4 illustrates superimposed selected ion-current profiles of three compounds and their labeled analogs. A discernible resolution of thymine glycol from its 2H-labeled analog is seen, whereas 2,6-diamino-4-hydroxy-5-formamidopyrimidine and 8-hydroxyguanine coelute with their respective ~3C- and lSN-labeled analogs. For quantification, calibration plots are obtained prior to GC/MS analysis for the response of the mass spectrometer to known quantities of both the analyte and its stable isotope-labeled analog as the internal standard. 46 For this purpose, mixtures containing known quantities of the analyte and analog are analyzed by GC/MS-SIM, and a number of prominent characteristic ions are monitored. The ratios of ion currents at selected masses are plotted as a function of the ratios of the molar amounts of an analyte and its analog. A linear relationship of the ratio of ion currents to the ratio of quantities should be obtained. Subsequently, quantities of analytes in a given mixture are calculated using the areas of the ion-current profiles of the monitored ions and the corresponding calibration plots. Examples of calibration plots are illustrated in Fig. 5. Although the data points were obtained by three independent measurements and contain standard deviations, 2~ the error bars are not discernable on the plots. This
A 10000"
C
B
s00o-
m/z 262
so00
m/z 446
90000~ 80000
m/z 444
70000
5000
8000"
60000-
7000"
4000
50000"
6000" 3000
5000" 4000
m/'z 259
Time
40000" 30000"
2000
20000.
30001 2000 6.40
m/z 442
lO00
(mln.)
6.70
10.60 Ttmo
m/z 440
10000' 10.B5 (mln.)
1
12.2
.9
Tlmo
(mln.)
FIG. 4. Superimposed selected ion-current profiles (A) at m/z 259 [(CH3)3Si derivative of thymine glycol] and 262 [(CH3)3Si derivative of thymine glycol-2H4]; (B) at m/z 442 [(CH3)3Si derivative of 2,6-diamino-4-hydroxy-5-formamidopyrimidine] and 446 [(CH3)3Si derivative of 2,6-diamino-4-hydroxy-5-formamidopyrimidineJSN3,13C]; (C) at rn/z 440 [(CH3)3Si derivative of 8-hydroxyguanine] and 444 [(CH3)3Si derivative of 8-hydroxyguanine15N3,13C]. Profiles were obtained during the GC/MS-SIM analysis of a trimethylsilylated hydrolyzate of y-irradiated chromatin. (From Dizdaroglu. 2~)
[1]
DETERMINING OXIDATIVEDNA DAMAGEBY GC/MS
15
2.0
g g "~ _o
1.o
0.0 0.0
0,5
1.0 ratio
1.5
2.0
2.5
of a m o u n t s
FI6.5. Calibration plots for ratios of ion currents at indicated masses versus ratios of molar amounts of (CH3)3Si derivatives of 8-hydroxyguanine and 8-hydroxyguanineJSN3,'C (Q, m/z 440/444), and (CH3)3Si derivatives of 4,6-diamino-5-formamidopyrimidine and 4,6diamino-5-formamidopyrimidine-tSN3,'C,2H (A, m/z 354/359). (From Dizdaroglu. 2~)
is because standard deviations were equal to or less than 1%, indicating the precision of such measurements by GC/MS-SIM. Selectivity and Sensitivity. The GC/MS-SIM technique not only permits high-sensitivity measurements but also provides high selectivity by virtue of monitoring a few selected ions that are characteristic of only one analyte in a complex mixture. This unique characteristic of mass spectrometry combined with the precise measurement of retention times by capillary gas chromatography makes unequivocal identification of analytes possible. Sensitivities in the range of approximately 1 fmol per compound applied to the GC column, or in the range of 1-3 modified residues in 10 6 DNA bases, can be achieved. It should be pointed out that the level of sensitivity may depend on the GC/MS instrument and the type of column and other factors. Mass spectrometers equipped with a highenergy dynode electron multiplier may provide approximately 10- to 20fold higher sensitivity than the sensitivity range mentioned above for modified DNA bases. Conclusions The GC/MS technique has been applied to a variety of in vitro and in vivo studies of oxidative DNA damage. Extensive reviews of the applications can be found elsewhere. 6'45 The technique offers the sensitivity,
16
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[2]
selectivity, speed, and versatility to solve a wide range of important measurement problems in terms of DNA base and sugar damage as well as DNA-protein cross-links. It also permits one to study enzymatic repair of DNA damage. 14,47 It appears that the GC/MS technique will continue to find a major role in studies of oxidative DNA damage and its repair in foreseeable future. Acknowledgment Work in the author's laboratoryis supported in part by the Officeof Health and Environmental Research, Officeof Energy Research, U.S. Department of Energy, Washington, D.C. 47M. Dizdaroglu, J. Laval, and S. Boiteux, Biochemistry 32, 12105(1993).
[2] A s s a y s o f O x i d a t i v e D N A D a m a g e B i o m a r k e r s 8 - O x o - 2 ' - d e o x y g u a n o s i n e a n d 8 - O x o g u a n i n e in N u c l e a r DNA and Biological Fluids by High-Performance Liquid Chromatography with Electrochemical Detection
By
MARK
K. S H I G E N A G A , E L I A S N. A B O U J A O U D E , QIN CHEN, and BRUCE N. AMES
Introduction Reactive oxygen species, which are generated as by-products of cellular metabolism, inflammation, ionizing radiation, and various xenobiotic treatments, 1 produce lesions in DNA that may result in mutations. The postulated importance of oxidative DNA damage in aging, cancer, and other age-related degenerative processes has prompted the development of methods that meet the following criteria: (1) ability to measure oxidative DNA damage with high sensitivity and selectivity in biological samples; (2) ability to survey this damage in a wide array of sample types such as DNA isolated from various organs and cultured cells as well as the excised repair products in biological fluids such as urine i B. Halliwell and J. M. Gutteridge, " F r e e Radicals in Biology and M e d i c i n e , " 2nd ed. Oxford Univ. Press (Clarendon), Oxford, 1989.
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994by AcademicPress, Inc. All rights of reproduction in any form reserved.
16
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[2]
selectivity, speed, and versatility to solve a wide range of important measurement problems in terms of DNA base and sugar damage as well as DNA-protein cross-links. It also permits one to study enzymatic repair of DNA damage. 14,47 It appears that the GC/MS technique will continue to find a major role in studies of oxidative DNA damage and its repair in foreseeable future. Acknowledgment Work in the author's laboratoryis supported in part by the Officeof Health and Environmental Research, Officeof Energy Research, U.S. Department of Energy, Washington, D.C. 47M. Dizdaroglu, J. Laval, and S. Boiteux, Biochemistry 32, 12105(1993).
[2] A s s a y s o f O x i d a t i v e D N A D a m a g e B i o m a r k e r s 8 - O x o - 2 ' - d e o x y g u a n o s i n e a n d 8 - O x o g u a n i n e in N u c l e a r DNA and Biological Fluids by High-Performance Liquid Chromatography with Electrochemical Detection
By
MARK
K. S H I G E N A G A , E L I A S N. A B O U J A O U D E , QIN CHEN, and BRUCE N. AMES
Introduction Reactive oxygen species, which are generated as by-products of cellular metabolism, inflammation, ionizing radiation, and various xenobiotic treatments, 1 produce lesions in DNA that may result in mutations. The postulated importance of oxidative DNA damage in aging, cancer, and other age-related degenerative processes has prompted the development of methods that meet the following criteria: (1) ability to measure oxidative DNA damage with high sensitivity and selectivity in biological samples; (2) ability to survey this damage in a wide array of sample types such as DNA isolated from various organs and cultured cells as well as the excised repair products in biological fluids such as urine i B. Halliwell and J. M. Gutteridge, " F r e e Radicals in Biology and M e d i c i n e , " 2nd ed. Oxford Univ. Press (Clarendon), Oxford, 1989.
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994by AcademicPress, Inc. All rights of reproduction in any form reserved.
[2]
ASSAYS FOR
oxo8dG IN D N A AND BIOLOGICAL FLUIDS
17
and spent cell culture or bacterial media; (3) capacity to measure such damage in organisms that have been exposed to endogenous and exogenous factors that are physiologically or pathologically relevant to aging and cancer; (4) sensitivity to explore the effects of nutritional status, an important determinant of cancer risk, and oxidative DNA damage. The identification of 7,8-dihydro-8-0xo-2'-deoxyguanosine (oxo8dG, 8OHdG) as a product of oxidative damage to DNA 2'3 together with the development of a high-performance liquid chromatography-electrochemical detection technique (HPLC-EC), which can detect this damage product with 2-3 orders of magnitude greater sensitivity than that of optical methods, 4 has fueled a growing effort aimed at understanding the relationship between oxidative DNA damage and cancer as well as other degenerative processes associated with aging. OxoSdG is a major product of oxidative DNA damage, 5 is implicated in G - T transversion mutagenesis, 6'7 and is elevated under various conditions associated with oxidant stress. 8 Application of H P L C - E C together with DNA isolation techniques and a monoclonal antibody-based immunoaffinity purification method for this biomarker of oxidative DNA damage 9 fulfills all the criteria listed above. In addition, quantitation of oxidative DNA damage in tissue DNA and biological fluids allows flexibility in designing experiments aimed at understanding the association between oxoSdG formation and its removal from DNA by glycosylase and excision repair under conditions of increased oxidant load. For a more complete overview of the many analytical techniques now available for the determination of oxo8dG readers are referred to reviews in this volume 1° and elsewhere. 1~
2 H. Kasai, H. Tanooka, and S. Nishimura, Gann 75, 1037 (1984). 3 M. Dizdaroglu, Biochemistry 24, 4476 (1985). 4 R. A. Floyd, J. J. Watson, J. Harris, M. West, and P. K. Wong, Biochem. Biophys. Res. Commun. 137, 841 (1986). 5 0 . I. Aruoma, B. Halliwell, E. Grajewski, and M. Dizdaroglu, Biochem. J. 273, 601 (1991). 6 M. L. Wood, M. Dizdaroglu, E. Gajewski, and J. M. Essigmann, Biochemistry 29, 7024 (1990). 7 K. C. Cheng, D. S. Cahill, H. Kasai, S. Nishimura, and L. A. Loeb, J. Biol. Chem. 267, 166 (1992). 8 R. A. Floyd, Carcinogenesis (London) 11, 1447 (1990). 9 E. M. Park, M. K. Shigenaga, P. Degan, T. S. Korn, J. W. Kitzler, C. M. Wehr, P. Kolachana, and B. N. Ames, Proc. Natl. Acad. Sci. U.S.A. 89, 3375 (1992). 10 M. Dizdaroglu, this volume [1]. 11 j. Cadet, F. Odin, J. F. Mouret, M. Polverelli, A. Audic, P. Giacomoni, A. Favier, and M. J. Richard, Mutat. Res. 275, 343 (1992).
18
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[2]
Analysis of Steady-State Levels of 8-Oxo-7,8-dihydro-2'-deoxyguanosine in DNA
Isolation and Preparation of DNA Samples for Chromatography DNA is isolated by modifications to a procedure described by Gupta. lz The following procedure is designed to isolate DNA from 50- to 200-mg tissue Samples or 3-6 x 106 cells from tissue culture, but may be scaled up as needed. Briefly, the procedure involves the preparation of cells by homogenization and/or lysis followed by protein and RNA digestion, phenol extractions, and enzymatic hydrolysis of DNA. Tissue Preparation. Samples (50-200 mg) are homogenized on ice or at 4 ° in 1 ml of homogenization buffer (0.1 M NaC1, 30 mM Tris, pH 8.0, I0 mM EDTA, 10 mM 2-mercaptoethanol, 0.5% (v/v) Triton X-100) with 6 passes of a Teflon-glass homogenizer at 200 rpm. The samples are centrifuged at 4° for 10 min at 1000 g to pellet nuclei. The supernatant is discarded and the crude nuclear pellet resuspended and rehomogenized in 1 ml of extraction buffer (0.1 M Tris, pH 8.0, 0.1 M NaC1, 20 mM EDTA) and recentrifuged as above for 2 min. The washed pellet is resuspended in 300 t~l of extraction buffer with a wide-orifice 200-tzl Pipetman tip. Cell Culture Preparation. Cells from adherent cultures are collected by trypsinization. At the end of trypsin treatment, cells are suspended in ice-cold calcium- and magnesium-free phosphate-buffered saline (PBS) containing 0.1 mM of the metal ion chelator deferoxamine mesylate (DFAM). Cells (3-6 x 106) are pelleted by centrifugation and resuspended in 300 pA of extraction buffer containing 2 mM butylated hydroxytoluene (BHT). For preparations containing greater than 107 cells, nuclei are pelleted by centrifugation and used for DNA isolation. Cell pellets are suspended in hypotonic buffer (10 mM Tris, 10 mM NaC1, 2 mM EDTA, 1 mM DFAM, and 0.2% Triton X-100, pH 8.0). The samples can be stored in this resuspension at - 2 0 ° before workup. The cell pellets are resuspended by vortexing, and the soluble fractions are removed after lowspeed centrifugation (500 g). DNA is isolated from this pellet by the phenol extraction method described below. Predigestion of Protein and RNA. RNA is digested on incubation of the samples with RNase (20/zl of the 20x stock containing 50 U/ml RNase A and 100 U/ml RNase T, Boehringer Mannheim) for 1 hr at 50 °. Protein is digested on incubation for 1 hr at 50° with a mixture of 40/xl of 10%
12 R. C. Gupta, Proc. Natl. Acad. Sci. U.S.A. 81, 6943 (1984).
[2]
ASSAYS FOR
oxo8dG
IN D N A AND BIOLOGICAL FLUIDS
19
sarkosyl and 40 /zl of a 5 mg/ml solution of proteinase K or pronase (Boehringer Mannheim) prepared in 10 mM Tris/1 mM EDTA, pH 7.4. Phenol Extractions. Add 400/zl of high-purity phenol (Clonetech, Palo Alto, CA) that has been saturated with the extraction buffer described above and vortex the sample vigorously for 1 min. For this purpose, a handheld multiple Eppendorf tube vortexer with capacities of 8 or 20 tubes is convenient and may be purchased from major scientific vendors. Centrifuge the sample at room temperature to separate the phases, transfer the upper aqueous phase to a new Eppendorf tube (often the DNA in the aqueous phase is highly viscous and may also be attached to cellular macromolecules residing at the interface; care should be taken to minimize carryover during transfer of the aqueous phase), and add 400/xl of extraction buffer-saturated phenol and 400 /zl of Sevag [chloroform:isoamyl alcohol (24 : 1, v/v)]. Vortex the sample vigorously for 1 min and centrifuge at 10,000 g for 10 min at room temperature. After transferring the upper aqueous phase to a new 1.5-ml Eppendorf tube, add 400/zl of Sevag to extract residual phenol from the aqueous phase and separate the phases according to the steps described above. Studies detailing the effects of phenol sensitization of DNA ~3warrant discussion. In our laboratory the effects of phenol have not revealed any measurable increases in oxo8dG caused by DNA extractions with phenol. Concentrating ethanol-treated enzymatic hydrolyzates of DNA with air can induce significant increases in oxo8dG levels; however, most DNA isolation protocols do not employ this procedure, but instead hydrolyze the sample after ethanol precipitation followed by decanting the ethanol and removing residual solvent by centrifugation under vacuum. This procedure is sufficient to prevent measurable increases in oxoSdG. Furthermore, studies on the effects of extraction of calf thymus DNA [three oxo8dG residues per 105 2'-deoxyguanosine (dG) residues] with the protocol described above that utilizes, as its phenol source, fresh high-purity phenol, old phenol (stored at room temperature for 1 year), or Sevag as a substitute, and comparing the levels to that of the same DNA that has not been subjected to this procedure, reveal essentially no difference in oxoSdG levels. In addition, rat liver DNA that has been subjected to up to four cycles of the phenol extraction protocol described above (where a single cycle is comprised of phenol, phenol-Sevag, and Sevag extractions) reveals no significant differences (2.0, 2.4, 1.9, and 2.0 oxoSdG residues per 105 dG residues for 1, 2, 3, and 4 cycles, respectively). Finally, it should be noted that phenol is an antioxidant, with oxidation potentials 13 H. G. Claycamp,
Carcinogenesis (London) 13, 1289 (1992).
20
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[2]
well below that of dG; thus, phenoxyl radicals or other contaminants of phenol (such as those present in old phenol preparations, see above) that could conceivably catalyze oxidation of dG to oxoSdG should be quenched by the vast molar excess of phenol that is used in the DNA extraction. Based on these observations, phenol extraction is judged to be an appropriate method of extracting DNA with minimal oxidation artifacts. Ethanol Precipitation. After transferring the aqueous phase from the last extraction to a fresh Eppendorf tube, precipitate the DNA by adding 40/.d of 3 M sodium acetate, pH 5.0, and 800 tzl of ice-cold ethanol to the sample. The precipitated DNA can immediately be centrifuged at 10,000 g for 15 min at 4 ° to pellet the DNA and washed with 1 ml of 70% ethanol in water (v/v) or allowed, particularly if the recovery of DNA is low, to sit overnight at - 2 0 ° to facilitate further precipitation. After washing the DNA pellet with 70% ethanol, the sample is centrifuged at 10,000 g for 15 min at 4° and the ethanol decanted from the sample. Although there is no measurable inhibition of P1 nuclease and alkaline phosphatase activities in incubation mixtures containing up to 10% ethanol by volume, poor chromatography of the subsequently analyzed enzymatic hydrolyzate is observed with as little as 2.5% ethanol (v/v). Consequently, to ensure good chromatographic separations, it is necessary to remove all traces of ethanol from the DNA pellet by centrifugation under vacuum for 10-20 min. Alternatively, ethanol can be carefully decanted from the sample tube, followed by air drying for approximately 30 min. Although resuspension of the DNA pellet following air drying is more convenient, possible oxidation artifacts could arise on exposure to air as described previously. ~3 Enzymatic Hydrolysis ofDNA. Dissolve 20-500/xg of DNA in 200/xl of 1 mM DFAM/20 mM sodium acetate, pH 5.0. Addition of DFAM, unlike EDTA, does not significantly inhibit the catalytic activity of either nuclease P1 or alkaline phosphatase but could limit metal-catalyzed oxidation reactions that contribute to artifacts and that may obscure the levels of oxo8dG that exist in vivo. DNA is hydrolyzed to nucleotides on incubation for 10 min at 65° with 4/zl of a 3.3 mg/ml suspension of nuclease P1 (Boehringer Mannheim, Indianapolis, IN) prepared in 20 mM sodium acetate, pH 5.0. Digestion of DNA near the temperature optimum for nuclease P1 activity, though rapid and more efficient than 1-hr incubations at 37°, requires that the incubation not exceed 15 min. Incubation periods in excess of 15 min at 65 ° result in an increase in oxidation artifacts that lead to a 2-fold increase in oxo8dG levels by 1 hr. This increase in oxoSdG levels is judged to be an artifact based on a parallel 2-fold increase in oxo8dG levels in a solution of dG that has been prepared identically. Under these incubation conditions, digestion of DNA to the corresponding
[2]
ASSAYS FOR
oxoSdG IN D N A AND BIOLOGICAL FLUIDS
21
nucleotides is found to be equal in efficiency to that of enzyme mixtures comprised of DNase, snake venom phosphodiesterase, and P I nuclease. ~4 The resulting mixture is pH adjusted to pH 8.5 by adding 20/zl of 1 M Tris-HC1 buffer, pH 8.5, and hydrolyzed to the corresponding nucleosides on incubation with 4/xl of 1 U//zl calf intestine alkaline phosphatase (Boehringer Mannheim) for 1 hr at 37°. If the pH values of incubation mixtures are not properly adjusted, rapid inactivation of alkaline phosphatase may occur, resulting in a DNA hydrolyzate containing a high percentage of undigested nucleotide monophosphates that elute with shorter retention times on reversed-phase HPLC. After alkaline phosphatase digestion, the pH of the hydrolyzed solution of DNA is once again lowered by adding 20 ~1 of 3 M sodium acetate buffer, pH 5.0, followed by the addition of 20/zl of 50 mM EDTA/10 mM DFAM solution prepared in water. The combination of this slightly acidic pH condition (which does not cause detectable depurination of oxoSdG to oxoSGua) and increased metal chelation minimizes posthydrolysis oxidation of the deoxynucleoside mixture. Transfer the contents to a 30,000-Da cutoff Millipore (Bedford, MA) UltraFree Eppendorf Filtration system and centrifuge at 10,000 g for 15-30 min at 4°. Analyze manually by injecting the filtrate directly or use a refrigerated (4°) automated system after transferring the filtrate to autosampier vials with low-volume polypropylene inserts (350/~I capacity). Samples that are placed in an autosampler should be analyzed within 48 hr of enzymatic hydrolysis to avoid oxidation artifacts.
Chromatography of Enzyme-Hydrolyzates of DNA For the chromatographic separation of electrochemically active and UV-absorbing material in the enzymatic hydrolyzates of DNA, any solvent delivery system in combination with a flow-through electrochemical and UV detection systems of high quality may be used. For our studies a Waters Associates (Milford, MA) Model 625 solvent delivery system and a refrigerated WISP Model 712 autoinjector are used. Other solvent delivery systems that minimize pump pulsation may be substituted; for example, the Hewlett Packard (Palo Alto, CA) Model 1090 systems provide baseline noise values that are adequate though slightly higher than that of the Waters system described above. Separations are performed by utilizing a linear gradient of 2.5-6.25% methanol in 50 mM KH2PO4 buffer, pH 5.5, over 20 min, although isocratic elution at 5% methanol in the same buffer is sufficient for samples containing over 20/xg DNA. The separa14 R. A. Floyd, M. S. West, K. L. Eneff, J. E. Schneider, P. K. Wong, D. T. Tingey, and W. E. Hogsett, Anal. Biochem. 188, 155 (1990).
22
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[2]
tions are achieved on a Supelco (Bellefonte, PA) LC-18 DB Supelcosil 3 /xm column (4.6 mm × 15 cm) equipped with a LC-18 DB precolumn cartridge assembly at a flow rate of 1.0 ml/min with a back pressure of approximately 2400 psi. Samples are analyzed by electrochemical and UV detection systems linked in series via low dead-volume polytetrafluoroethylene (PTFE) tubing. Although it is customary to detect UV-absorbing material before the possibly destructive electrochemical detector, we choose to place the electrochemical cell in line before the UV detector. This arrangement increases sensitivity slightly by improving peak shape, and at the oxidation potentials employed chemical alteration of the normal deoxynucleosides does not occur. Electrochemical detection of oxoSdG utilizes an ESA (Bedford, MA) Model 5100 Coulochem detector equipped with a 5011 analytical cell with potentials of electrodes 1 and 2 adjusted to 0.1 and 0.4 V, respectively. Detection at 260 or 290 nm utilizes a Kratos (Westwood, NJ) Model 773 UV detector. Because the average absorbance of nucleosides at 290 nm is substantially lower, detection at this wavelength increases the amount of DNA that can be quantitated. As shown in Fig. 1A, analysis of 47.1 /xg of an enzyme hydrolyzate of rat liver DNA following HPLC separation reveals a readily detectable peak at 19.23 min that corresponds to synthetic oxo8dG (Wako Chemicals USA, Inc., Richmond, VA). Shown in Fig. 1B is a chromatogram of the same sample monitored by UV detection; the three most prominent peaks correspond to the normal deoxynucleosides 2'-deoxycytidine (retention time 5.94 min), dG (14.37 min), and thymidine (16.76 min) (2'-deoxyadenosine elutes during the column washout at - 2 6 min). At the sensitivity setting used, the UV absorbance signal for oxoSdG, unlike that of the electrochemical signal for this compound (see above), is not observed. Depending on the source of DNA, quantitation of 5-20 fmol of oxo8dG per microgram of DNA can be achieved readily on as little as 10/xg of DNA. For this sample, the oxo8dG level is equivalent to 1.8 residues of oxo8dG per 105 residues of dG or 11.3 fmol of oxo8dG per/zg DNA, Although signals for oxo8dG can be detected with amounts of DNA that are significantly less than 10/xg, the levels of oxo8dG in such samples sometimes appear to be inappropriately high compared to the levels observed of DNA hydrolyzates obtained from the same sample but where the recovery allows one to analyze at least 10 tzg of DNA.15 The reason(s) for this discrepancy is not known. Levels of oxoSdG in various DNA samples measured by electrochemical detection are expressed as values relative to the amount of dG detected by UV absorbance at 260 or 290 nm. Alternatively, the amount of oxo8dG ~5 K. B. Beckman and B. N. Ames, unpublished observation (1993).
[2]
ASSAYS FOR
oxoSdG IN
D N A AND BIOLOGICAL FLUIDS
23
|1 -2 nA
A to ~J
re
uJ
dC
B
0.05 AU
E c-
dG
O
dT e,,
8¢.j >
0
5
I
I
I
]
10
15
20
25
Time (rain)
FIG. 1. Ultraviolet and electrochemical chromatograms of an enzyme hydrolyzate of rat liver DNA following gradient reversed-phase HPLC. An enzyme hydrolyzate, equivalent to 47.1 /~g of DNA, is separated by reversed-phase HPLC and analyzed by two detection systems linked in series. (A) HPLC chromatogram with electrochemical detection at +0.4 V. (B) HPLC chromatogram with UV detection at 290 nm. dC, Deoxycytidine; dG, deoxyguanosine; dT, thymidine. The steady-state level of oxoSdG in this sample is 1.8 molecules per 105 dG or 11.3 fmol of oxoSdG//~g DNA.
may be normalized to the amount of DNA analyzed; this value can be readily calculated by using the conversion factor 0.648 nmol dG per microgram of DNA. This conversion factor applies only to DNA composed of 0.21 parts dG, a value typical for nuclear DNA isolated from eukaryotic cells. The amount of oxoadG may therefore be expressed in terms of the molar ratio of oxoSdG and dG or femtomoles oxoSdG per microgram of DNA. Typical values obtained from the analysis of freshly isolated DNA
24
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[2]
from a variety of untreated tissues are 1-2 residues of oxoSdG per l0 s residues of dG, which is equivalent to approximately 6.5-13.0 fmol oxoSdG per microgram of DNA. Enzymatic hydrolysis is judged to be comparable to formic acid hydrolysis in its ability to liberate residues containing 8-oxoguanine (oxoSGua) from D N A since ratio values of I-2 residues oxoSgua/10 5 Gua are observed with acid-catalyzed hydrolysis of D N A as well. 16 The combination of HPLC with electrochemical and UV detection has been utilized in studies aimed at quantitating oxoSdG in D N A isolated from various sources including animal tissues, 17-19human blood, 2° human sperm, 21 rat liver mitochondriafl 2 tissue culture cells, 23 and bacteria, z4 Analysis of 7,8-Dihydro-8-oxo-2'-deoxyguanosine, 8-Oxoguanine, and 8-Oxoguanosine in Biological Fluids Analysis of oxoSGua and oxoSdG in biological fluids 9,a5-28 provides an estimate of D N A repair of this oxidative D N A damage product, although other cellular processes such as mitochondrial D N A turnover, cell turnover induced by cell necrosis or apoptosis, and oxidation of cytosolic dGTP pools can all contribute to the levels of these adducts, particularly oxo8dG. Under normal conditions, oxoSGua excretion rates are approximately 10-fold higher than that of oxo8dG, a ratio that appears to increase I6 H. C. Yeo and B. N. Ames, unpublished observation (1993). i7 H. Kasai, P. F. Crain, Y. Kuchino, S. Nishimura, A. Ootsuyama, and H. Tanooka, Carcinogenesis (London) 7, 1849 (1986). i8 C. G. Fraga, M. K. Shigenaga, J. W. Park, P. Degan, and B. N. Ames, Proc. Natl. Acad. Sci. U.S.A. 87, 4533 (1990). i9 H. Yoshiji, D. Nakae, Y. Mizumoto, K. Horiguchi, K. Tamura, A. Denda, T. Tsujii, and Y. Konishi, Carcinogenesis (London) 13, 1227 (1992). 20 H. Kiyosawa, M. Suko, H. Okudaira, K. Murata, T. Miyamoto, M. H. Chung, H. Kasai, and S. Nishimura, Free Radical Res. Commun. 11, 23 (1990). 2~ C. G. Fraga, P. A. Motchnik, M. K. Shigenaga, H. J. Helbock, R. A. Jacob, and B. N. Ames, Proc. Natl. Acad. Sci. U.S.A. 88, 11003 (1991). 22 C. Richter, J. W. Park, and B. N. Ames, Proc. Natl. Acad. Sci. U.S.A. 85, 6465 (1988). 23 B. C. Beehler, J. Przybyszewski, H. B. Box, and M. M. Kulesz, Carcinogenesis (London) 13, 2003 (1992). 24 T. Bessho, K. Tano, H. Kasai, and S. Nishimura, Biochem. Biophys. Res. Commun. 188, 372 (1992). 25 M. K. Shigenaga, C. J. Gimeno, and B. N. Ames, Proc. Natl. Acad. Sci. U.S.A. 86, 9697 (1989). 26 p. Degan, M. K. Shigenaga, E. M. Park, P. E. Alperin, and B. N. Ames, Carcinogenesis (London) 12, 865 (1991). z7 D. S. Bergtold, M. G. Simic, H. Alessio, and R. G. Cutler, Basic Life Sci. 49, 483 (1988). 28 S. Loft, K. Vistisen, M. Ewertz, A. Tjonneland, K. Overvad, and H. E. Poulsen, Carcinogenesis (London) 13, 2241 (1992).
[2]
ASSAYS FOR
oxoSdG IN D N A AND BIOLOGICAL FLUIDS
25
during cell proliferation (>10) and decrease following cell death (<10). Thus, it is proposed that oxo8dG is derived from the numerous pathways described above, whereas the factors that determine oxo8Gua excretion are dominated by active repair of oxo8dG lesions in DNA, presumably by an 8-0xoguanine glycosylase activity that exists in mammalian cells. 29 So, despite the complexities associated with interpreting oxoSdG levels in biological fluids, the ratio of the excretion rates for oxoaGua and oxo8dG can signify relative rates of cell proliferation or cell turnover when compared to the ratio estimated in unperturbed animal or cell culture systems. Absolute excretion rates of oxoSGua, in turn, are likely to signify oxidative hit rates in viable, actively repairing cells.
Synthesis of Radiolabeled Standards The addition of radioisotopically labeled internal standards of oxo8dG, oxo8Gua, and 8-oxoguanosine (oxoSG) to biological fluid samples is required to calculate the recovery of the respective compounds following solid-phase extraction (SPE) and/or immunoaffinity purification. Synthesis of radiolabeled oxo8dG is performed as described previously. 25 [U-14C]Oxo8G is synthesized by the same protocol. Synthesis of 8-Oxo[UJ4C]guanine. To [UJ4C]guanosine (Moravek Biochemicals, Brea, CA) [50/~Ci, 0.160/xmol (specific activity 473/xCi/ /xmol)] is added 4/.d of a 1 : 200 (v/v) dilution of Br2 in water (-2/zmol). Bromination at the C-8 position occurs rapidly (<1 min) to afford 8bromo[UJ4C]guanosine in a 65% yield. After HPLC purification of 8bromo[UlaC]guanosine, 8-oxo[U-laC]guanine) ([u-laC]oxo8Gua) is prepared by concentrating an aliquot of the bromo derivative to dryness and reacting it with 100-200/xl of undiluted reagent-grade formic acid at 130° overnight. The crude reaction mixture is concentrated to dryness in a centrifuge under vacuum, and the formic acid hydrolysis step is repeated. After this second hydrolysis the reaction mixture is concentrated to dryness, resuspended in water, and purified by reversed-phase HPLC using a 5/zm Supelcosil LC-18 column (4.6 x 25 cm) (Supelco, Bellefonte, PA) and a 50 mM KH2PO4 buffer, pH 4.0, mobile phase at a flow rate of 1 ml/ min; [14C]oxo8Gua elutes with a retention time of 11.4 min. The conversion of 8-bromo[U-14C]guanosine to [UJ4C]oxo8Gua is essentially quantitative with loss of one-half of the radioactivity through glycosidic bond cleavage of the 14C-containing ribose substituent. HPLC-purified [uJaC]oxo8Gua is stored at 4° in a polypropylene tube as a solution in 50 mM Tris-HC1, pH 7.0. Radiochemical purity of U-lnC-containing standards are checked routinely if stored for a period of greater than 1 month. 29 F. Yamamoto, S. Nishimura, and H. Kasai, Biochem. Biophys. Res. Commun. 187, 809 (1992).
26
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[2]
Immunoaffinity Column Preparation Preparation of immunoaffinity columns according to Degan et al. 3° are described briefly. Lipid is removed from mouse ascites fluid by adding 2 parts of the delipidating agent Sero-clear (Calbiochem Corp., La Jolla, CA) to 3 parts of ascites fluid. The sample is mixed thoroughly by turning gently end to end for 1 rain at room temperature, followed by centrifugation at 3000 g for 2 min. Transfer the monoclonal antibody (MAb)-containing top aqueous layer to a polypropylene culture tube containing wet CNBractivated Sepharose [pretreated according to instructions of the manufacturer (Pharmacia, Piscataway, N J] and add an equal volume of 0.1 M Na2CO3/0.5 M NaC1 buffer, pH 8.3. Place the tube in a multipurpose rotator set at a gentle mixing speed and allow it to react for 2 hr at room temperature (or overnight at 4°). Centrifuge the tube at I000 rpm for 2 rain and discard the supernatant. Add enough 0.1 M Tris-HC1 buffer, pH 8.5, to the tube to suspend the remaining pellet, place the tube on the rotator, and again allow the sample to mix for 2 hr at room temperature (or overnight at 4°). Transfer the contents of the tube to a glass filter and wash the gel with 5 volumes of 0.1 M Na2CO3/0.5 M NaC1 buffer, pH 8.3, followed by 5 volumes of water. The prepared immunoaffinity matrix can be diluted according to the sample type that will be purified. For urine samples, the immunoaffinity matrix is diluted with an equal volume of Sepharose 4B. Pack this mixture into 0.7 cm × 10 cm Econo-Columns (Bio-Rad, Hercules, CA) for a final bed volume of 2 ml and store wet at 4°. Approximately 5 columns can be prepared from a 0.5-ml aliquot of ascites fluid. Prior to using the column for the first time, pretreat it by washing with 10 ml each of the following: water, 1 M NaC1, water, acetonitrile, methanol, and water.
Sample Preparation and Isolation of S-Oxoguanine, 7,8-Dihydro-8-oxo2'-deoxyguanosine, and 8-Oxoguanosine from Biological Fluids Urine. Urine samples (1-5 ml), which are stored at - 2 0 ° prior to workup, are diluted with an equal volume of ! M NaCI, spiked with 3000-10,000 cpm of the appropriate radiolabeled tracer(s) ([3H]oxoSdG, [14C]oxoaGua, or [14C]oxo8G), and applied to a preconditioned Perkin E1mer/Analytichem (Harbor City, CA) C18OH solid-phase extraction (SPE) column. 25 The SPE column is then washed with 5 ml of 50 mM KH2PO4 buffer, pH 7.5, and retained compounds are eluted with 3 ml of 15% methanol in the same buffer. The eluate is applied to the immunoaffinity column at 4 °. The column is washed in sequential order with 5 ml each of the following: water, 1 M NaCI, water, and acetonitrile. Less than 10% 3o p. Degan, R. Montesano, and C. P. Wild, Cancer Res. 48, 5065 (1988).
[2]
ASSAYS FOR
oxoSdG IN
D N A AND BIOLOGICAL FLUIDS
27
of the applied radioactivity, corresponding to the radiolabeled internal standards for oxoSdG and oxoaG, are lost during the series of washes. Depending on the quality and age of the immunoaffinity column, 10-50% of the radioactivity corresponding to oxoSGua can be lost at this stage. Because the binding affinity of the MAb is roughly 20-fold less for oxo8Gua (2-amino-6,8-dihydroxypurine, Aldrich Chemical Co., Milwaukee, WI) than that for oxoSdG, such an increase in loss of radiolabeled oxo8Gua is diagnostic of degrading column performance. Immediately following the acetonitrile wash, the antibody-binding compounds are eluted with methanol (5 ml) into a polypropylene culture tube. The resulting methanol eluate is concentrated to dryness under a stream of nitrogen in a 40°-45 ° water bath for 1-2 hr. The sample is resuspended in 200/xl of water, and a 20- to 50-/zl aliquot is analyzed by H P L C - E C as described below. Recoveries of oxo8dG, oxoSGua, and oxo8G from the immunoaffinity columns are typically 60-90%. The urinary excretion rate values are expressed as picomoles of oxoSdG excreted in a 24-hr urine void per kilogram of body weight (pmol kg -~ day-l). Alternatively, the urinary excretion rates can also be normalized to creatinine. It is important when normalizing to creatinine to consider how creatinine excretion could be affected in a given sample population. For example, creatinine, a by-product of muscle protein, is turned over at a faster rate after exercise and decreases significantly in old age as muscle mass and metabolic rate decrease. Therefore, normalization to creatinine can be misleading, if uncorrected, in exercise or aging studies. Because of the large contribution of oxoSGua from dietary sources, 9 presumably from peptic acid hydrolysis of DNA containing oxo8dG, urine must be obtained from animals or individuals placed on a nucleic acidfree diet if quantitative estimates of its excretion rate are to be determined. Unfortunately, this problem severely limits the use of the urinary assay in human epidemiological surveys since it is impractical to place individuals on such a restrictive diet. Although urinary oxo8dG is not so affected it is derived from various sources other than nonspecific nuclease repair of this lesion in DNA, thus complicating the interpretation of results obtained by this assay. Maintenance oflmmunoaffinity Columns. The column is regenerated immediately following the methanol elution step described above by applying an additional aliquot of methanol (10 ml). The bed volume is reequilibrated and resuspended in water (I0 ml, 2 times) with a Pasteur pipette to remove air bubbles, and columns are stored at 4° in high-purity water between uses. If columns are stored for long periods of time (> 1 month), an aqueous solution of 0.02% sodium azide can be used to wet the bed
28
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[2]
volume. The immunoaffinity columns are quite durable and can be used for at least 10 cycles without apparent loss of binding capacity. However, with prolonged use that is accompanied by loss of binding capacity for oxoaGua, the immunoaffinity column also exhibits decreased binding specificity. Of particular concern is the retention of an electrochemically active artifact with HPLC chromatographic characteristics similar to that of oxoSdG. Although the compound typically elutes 1-2 min prior to oxoSdG, the electrochemical signal produced by the artifact can obscure the oxoSdG peak; at this point the immunoaffinity columns are discarded. Tissue Culture. Measurement of oxoSGua, oxoSdG, and oxoSG produced by cells in culture is complicated by the presence of high levels of these compounds in serum, a component of many cell culture media. Commercially available dialyzed fetal bovine serum (FBS), for example, contains less oxo8Gua, oxo8dG, and oxo8G than normal serum; however, the modifed residues are still present at levels that are typically 5 times higher than that produced by cells in cultures. Therefore, to obtain a satisfactory signal-to-noise ratio, it is necessary to remove oxoSGua, oxoSG, and oxoadG from dialyzed FBS by the immunoaffinity matrix described above. The anti-oxoSdG antibody-conjugated Sepharose 4B matrix (10 ml) is added to dialyzed FBS (25 ml) contained in a 50-ml polypropylene culture tube with a screw top (this procedure can be scaled accordingly). The mixture is swirled and the tube turned end to end to ensure thorough mixing. The mixture is centrifuged at 1000 g for 4 min at 4 °, and the serum (supernatant) is transferred to a fresh culture tube; at this point the process is repeated. After two consecutive "scrubbing" cycles, the serum is sterile filtered through a 0.22/zm Nalgene (Rochester, NY) filtration apparatus. Typically, greater than 95% of serum oxo8Gua, oxoadG, and oxoaG are removed. Residual background levels of the adducts are accounted for in each experiment by subtracting the amounts found in medium controls. The anti-oxoSdG antibody-conjugated Sepharose 4B matrix (beads) can be reused. After removing the serum, the remaining beads ( - 1 0 ml) are washed two times with 40 ml of ice-cold sterilized water in a 50-ml polypropylene tissue culture tube. After centrifugation (1000 g for 4 min at 4°), the supernatant wash volume is removed. The immunoaffinity beads are washed further with two cycles of methanol (40 ml), which removes oxo8Gua, oxoSdG, and oxoSG from the antibodies. After two cycles, the beads are rinsed further with two cycles of distilled water. The beads are stored at 4 ° in water between uses. The levels of oxoSGua, oxoSdG, and oxoSG produced by most types of cells in culture are low. Large amounts of cells in culture are usually necessary for estimating the levels of these adducts in tissue culture media.
[2]
ASSAYS FOR
oxoSdG IN
D N A AND BIOLOGICAL FLUIDS
29
For adherent cultures such as normal fibroblast cells (human dermal fibroblast cells F65 (Naval Biosciences Laboratory, Oakland, CA), CRL 1489 (ATCC, Rockville, MD), human fetal lung fibroblast cells IMR-90 and WI-38 (Coriell Institute for Medical Research, Camden, N J), mouse fibroblast cells NIH 3T3 and C3H10T1/2 (ATCC, Rockville, MD), cultures of three to six 100-mm dishes are used for a single determination. For suspension cultures such as HeLa cells, a volume of 250 ml containing 3-8 × 105 cell/ml is used for a single determination. To measure the levels of oxoSGua, oxo8dG, and oxoSG produced by endogenous oxidative DNA damage, cells are seeded in 100-mm dishes containing 10 ml of 10% FBS that has been processed by treatment with the immunoaffinity beads. Medium is replaced in order to remove unattached cells, and the cell number (Do) is determined on the next day. After several days (TI), the number of which is dependent on both the cell type and seeding density, cultures reach confluency. The conditioned medium collected over the period To to TI is recovered and the number of cells (D) is determined. Alternatively, the conditioned medium may be collected and D determined after the cells are postconfluent for a number of days (TII). Levels of oxoSGua, oxoSdG, and oxoSG in conditioned medium are determined as for urine by HPLC-EC after isolation using the anti-oxo8dG monoclonal antibody immunoaffinity column. The conditioned (spent) medium from six 100-mm dishes of human fibroblast cells or from 250 ml of HeLa cell cultures is filtered through a 0.22/xm Nalgene filtration apparatus and pooled prior to processing by the immunoaffinity column. A trace amount of [3H]oxoSdG (10,000 cpm) added to each sample before applying the conditioned medium to the column, is used to correct for losses during sample workup. It is also possible to spike, within the same sample, with [14C]oxoSGua (3,000-5,000 cpm) to determine the recovery efficiency of oxoSGua. After the conditioned medium is allowed to pass through the affinity column by gravity, the columns are washed as described for the urinary assay (see above) and the adducts are eluted with methanol. Samples are concentrated to dryness and resuspended in buffer containing 100 mM Tris, 20 mM EDTA, and I mM DFAM. Excretion R a t e Calculations. The number of residues per cell per day (X) is calculated based on the assumption that each cell excretes adducts daily at a constant rate and that cell numbers do not change significantly after cultures reach confluence. The rate of oxoSGua, oxoSdG, and oxoSG excretion by each cell per day (X) is calculated by the following equation:
N A Y = (Do + D1
+
D2 +
"'"
+ D , + "" + D ) X
30
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[2]
For exponentially growing cells, D, = 2 nr × Do. Therefore, NA Y = X f-JlD dDn = X fT, 2nr Do dn DO
(D - Do)TIx
For growth arrested cells, NAY = DTIIX Each variable is defined as follows: Y is the total moles of adducts from a single culture dish; N A is Avogadro's number (6.03 × 1023per mole); X is the number of residues produced by a single cell per day; Do is the number of cells before conditioning the medium; D is the cell number when the medium is collected; r is the cell growth rate; TI is the number of days of exponential cell growth; Tu is the number of days that the cells are confluent. Blood Plasma. Human blood plasma (5-10 ml) is spiked with the appropriate radiolabeled tracer(s) and the proteins precipitated by adding an equal volume of acetonitrile. The precipitated proteins are separated by centrifugation at 3000 g for 15 min at 4 °, and the supernatant is transferred to a new culture tube and mixed with 8 volumes of water (40 ml). The resulting sample is applied directly to an immunoaffinity column and processed according to the protocol described for the isolation of these adducts from urine. Alternatively, following centrifugation, the sample can be stored at - 2 0 ° for 2 hr to allow separation of the organic and aqueous phases. Following removal of the organic phase with a Pasteur pipette, the remaining aqueous layer is transferred directly to the immunoaffinity column and processed according to the protocol described above. Approximately 5 ml of blood plasma is required for sample workup; the equivalent of 1 ml is analyzed. Overall recovery of oxoSdG is approximately 40%. The average steady-state concentration of oxoSdG in blood plasma is approximately 70-120 pM, or roughly 100-fold less than the average concentration of 10 nM detected in human urine. Bacterial Media. Measurement of oxoSGua and its nucleoside derivatives in spent media of bacterial cultures can be used to investigate the role of various known mutants (i.e., mutM, mutT, mutY, oxyR, and soxR) and the relationship between oxidative damage and mutation rates as well as to understand the adaptive response of a cell to oxidant stress. Wild-type Escherichia coli K-12 is cultured in Vogel Bonner citrate medium with 0.4% glucose (w/v). Cultures (100 ml) are inoculated with 3 ml of an overnight culture and incubated at 37° for 8 hr (A600 = 1.0).
[2]
ASSAYS FOR
oxoSdG IN
D N A AND BIOLOGICAL FLUIDS
31
Cells are removed from the medium by centrifugation at 3000 g for 10 min at 4° and the medium filtered through a 0.22 txm membrane filter unit (Nalgene). The medium (100 ml) is spiked with the appropriate radiolabeled tracer, applied directly to the antibody column, and processed as described above. A second method for growing E. coli and other bacteria, and which is likely to support better growth of bacterial mutants compared to that of the minimal media described above, employs an amino acidsupplemented medium. This media is prepared by adding, as supplements to 0.4% glucose Vogel Bonner citrate, the following: 50x of GIBCO (Grand Island, NY) MEM essential amino acids (10 ml), 50× of GIBCO MEM nonessential amino acids (5 ml), 100× of GIBCO vitamin mixture (5 ml), and 100x thiamin (5 ml) in a total volume of 500 ml. Bacterial media prepared with these amino acid and vitamin supplements are found to support more vigorous bacterial growth without the high background levels of oxoSGua derivatives that obscure endogenous oxoSGua production and which are observed in complete media (i.e., Luria-Bertani medium (LB), casamino acids, etc.).
Chromatography of Immunoaffinity-Purified Biological Fluids For the chromatographic separation of electrochemically active compounds isolated from immunoaffinity-purifiedbiological fluids, the solvent delivery systems, HPLC columns, and mobile phase conditions are used as described for the determination of oxoSdG in DNA. Data are digitized by a Nelson (Cupertino, CA) Model 760 analytical interface and processed by PE/Nelson TurboChrom 3 data acquisition software on an IBM PS/2 Model 70 computer. Figure 2 shows a representative chromatogram of human urine processed by a combination of SPE/MAb-based immunoaffinity columns and analyzed by HPLC-EC. In addition to oxo8dG (peak 6), which elutes at 20.9 min, other 6,8-dioxopurines are observed including uric acid (peak 1, retention time 3.2 min), oxo8Gua (peak 2, 5.1 min), 7-methyl-8-oxoguanine (peak 3, 12.2 min), Nz-methyl-oxoSGua (peak 4, 13.2 rain), and oxoSG (peak 5, 14.5 min). Similar chromatographic profiles are observed for other biological fluids. Detection is normally performed at 200 nanoamperes (nA) full scale (gain 500) and reprocessed according to the signal obtained for oxo8dG. In Fig. 2, reprocessed data are shown at 50 nA full scale. Detection of approximately 250-1500 fmol of oxoSdG is achieved readily by analyzing the equivalent of 25-100/xl of human urine from a workup volume of 1-5 ml. The value obtained is normalized to the volume of urine excreted in a 24-hr period and the body weight of the individual. Results obtained by this method indicate that humans excrete approximately 130-300 pmol of oxoSdG per kilogram of body weight per 24 hr. Similar results have been obtained by other methods. 27,28
32
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
Peak ~t
[2]
Compound O
H Uric Acid
O/~,N ,/~-HN O HN
T5 nA
14 N
oxo~Gua
H2. c~, 3
.N
O
4
2
rnToxoSOua
H~
5
O u.I
O H HN~ % , _ ^
1
"-.~.--~./-"
N2methyl -
oxo~ua
CHa O 5
HN
H N
OXOeG i
I
5
ml~s* 0
6
ItN
14 N
0
H2N~..N~I N~=
I
I
10 15 Time (min)
I
I
20
25
OXO~IG
I dRl~se
Fro. 2. HPLC-EC chromatogram of SPE/immunoaffinity-purifiedhuman urine. The sample, whichis equivalentto 0.42 ml urinebased on the recoveryof [3H]oxoSdG,contained 3.98 pmoloxoSdG(9.48nM). Becausethe SPE stepusedto process the samplewas optimized for oxoSdG,recoveries of earlier elutingcompoundsare in manycases signifcantlylower. (Adapted from Park et al. 9)
Summary High-performance liquid chromatography with electrochemical detection is a highly sensitive and selective method for detecting oxoSdG and oxoSGua, biomarkers of oxidative DNA damage. When employed together with the DNA isolation and monoclonal antibody-based immunoaffinity purification methods described, oxoSdG and oxoSGua in DNA and urine can be readily detected and quantitated, offering a powerful approach for assessing oxidative DNA damage in vivo. Application of the technique to the detection of oxoSdG from DNA permits quantitation of the steadystate levels of this oxidatively modified deoxynucleoside and overcomes the detection problems associated with the extremely low levels present in DNA. In addition, the selectivity gained by this detection method eliminates the problem of separating the signal for oxoSdG from those of normal deoxynucleosides. The quantitation of oxoSdG and oxoSGua in
[3]
DETECTING REPAIR ENZYMES FOR OXIDATIVE DAMAGE
33
biological fluids is noninvasive and complements the measurement of oxoSdG in DNA by estimating the rate of oxidative DNA damage occurring within the body or in a population of cells. This analytical approach may allow one to estimate oxidative DNA damage in an animal or individual exposed to prooxidant conditions associated with lifestyle, genetic predisposition, degenerative diseases, or environmental toxins. Furthermore, these assays may allow one to assess the potentially beneficial effects of intervention strategies that protect DNA from such damage. Acknowledgments This work was supported by National Institutes of Health Grant CA39910to B.N.A., by National Institute of Environmental Health Sciences Center Grant ES01896 and the Colgate PalmolivePostdoctoralFellowshipGrant M1597from the Societyof Toxicologyto Q.C.
[3] D e t e c t i o n a n d C h a r a c t e r i z a t i o n o f E u k a r y o t i c E n z y m e s That Recognize Oxidative DNA Damage
By
KRISTA
K.
HAMILTON, KEUNMYOUNG LEE,
and
PAUL W. DOETSCH
Introduction Reactive oxygen species generated by ionizing radiation, near-UV light, normal cellular metabolism, and certain chemical agents have been shown to cause significant damage to DNA, including single-strand breaks, damage to deoxyribose moieties, and modifications of the DNA bases.~'2 The types of DNA base damage produced by oxidizing agents include ring saturation, fragmentation, contraction, and exocyclic modifications of pyrimidines. The purine products 2,6-diamino-4-hydroxy-5-formamidopyrimidine, hypoxanthine, and 8-oxoguanine, as well as abasic sites, also form as a result of oxidative insult, z The importance of the removal of these DNA lesions is underscored by the number of enzymes identified in both prokaryotes and eukaryotes which recognize and act on such DNA lesions. These repair enzymes act via N-glycosylase and apurinic/
I R. Teoule, Int. J. Radiat. Biol. 51, 573 (1987). 2 F. Hutchinson, Prog. Nucleic Acid Res. Mol. Biol. 32, 115 (1985).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproduction in any form reserved.
[3]
DETECTING REPAIR ENZYMES FOR OXIDATIVE DAMAGE
33
biological fluids is noninvasive and complements the measurement of oxoSdG in DNA by estimating the rate of oxidative DNA damage occurring within the body or in a population of cells. This analytical approach may allow one to estimate oxidative DNA damage in an animal or individual exposed to prooxidant conditions associated with lifestyle, genetic predisposition, degenerative diseases, or environmental toxins. Furthermore, these assays may allow one to assess the potentially beneficial effects of intervention strategies that protect DNA from such damage. Acknowledgments This work was supported by National Institutes of Health Grant CA39910to B.N.A., by National Institute of Environmental Health Sciences Center Grant ES01896 and the Colgate PalmolivePostdoctoralFellowshipGrant M1597from the Societyof Toxicologyto Q.C.
[3] D e t e c t i o n a n d C h a r a c t e r i z a t i o n o f E u k a r y o t i c E n z y m e s That Recognize Oxidative DNA Damage
By
KRISTA
K.
HAMILTON, KEUNMYOUNG LEE,
and
PAUL W. DOETSCH
Introduction Reactive oxygen species generated by ionizing radiation, near-UV light, normal cellular metabolism, and certain chemical agents have been shown to cause significant damage to DNA, including single-strand breaks, damage to deoxyribose moieties, and modifications of the DNA bases.~'2 The types of DNA base damage produced by oxidizing agents include ring saturation, fragmentation, contraction, and exocyclic modifications of pyrimidines. The purine products 2,6-diamino-4-hydroxy-5-formamidopyrimidine, hypoxanthine, and 8-oxoguanine, as well as abasic sites, also form as a result of oxidative insult, z The importance of the removal of these DNA lesions is underscored by the number of enzymes identified in both prokaryotes and eukaryotes which recognize and act on such DNA lesions. These repair enzymes act via N-glycosylase and apurinic/
I R. Teoule, Int. J. Radiat. Biol. 51, 573 (1987). 2 F. Hutchinson, Prog. Nucleic Acid Res. Mol. Biol. 32, 115 (1985).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproduction in any form reserved.
34
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[3]
apyrimidinic (AP) endonuclease or lyase activities at sites of DNA lesions to initiate repair through the base excision repair pathway. 3-5 Currently, techniques to examine DNA repair enzymes involve either the detection of released, damaged DNA bases as indicators of N-glycosylase activity or assays measuring nicks in duplex DNA substrates. N-Glycosylase activity can be demonstrated by thin-layer6'7or high-performance liquid chromatographic7'8 analysis of modified bases released from a radiolabeled, oxidatively damaged DNA substrate. The abasic site generated by a DNA N-glycosylase activity can be monitored colorimetrically through the use of an aldehyde-reactive probe, namely, a biotin-tagged derivative of O-(carboxymethyl)hydroxylamine.9 The methods are especially useful in the confirmation of the chemical nature of the base damage and quantitation of the amount of product released. Nicks in duplex DNA are commonly detected by following the conversion of supercoiled plasmids (form I DNA) containing specific DNA lesions to nicked plasmids (form II DNA) by agarose gel electrophoresis 1° or by nitrocellulose filter binding of nicked, radiolabeled plasmids. 1~ These assay methods, while rapid and quantitative, do not identify the nucleotide site of endonuclease incision, nor do they provide any indication of the mechanism by which strand cleavage occurs. They are also often unsuitable for use when examining crude cellular extracts which may contain multiple enzyme activities that possess the ability to cleave the DNA phosphodiester backbone. The method described here utilizes end-labeled DNA fragments of defined sequence containing various base damages in conjunction with base-specific chemical cleavage reactions for the detection and characterization of eukaryotic enzymes that recognize oxidative DNA damage. This method was originally employed for the characterization of prokaryotic enzymes that recognize and cleave DNA at sites of UV light-induced cyclobutane pyrimidine dimers and to define the site of damage and DNA 3 S. S. Wallace, Enoiron. Mol. Mutagen. 12, 431 (1988). 4 p. W. Doetsch, K. K. Hamilton, L. Rapkin, S. A. Okenquist, and J. Lenz, in "Ionizing Radiation Damage to DNA: Molecular Aspects" (S. Wallace and R. Painter, eds.), p. 109. Wiley-Liss, New York, 1990. 5 p. W. Doetsch and R. P. Cunningham, Murat. Res. 236, 173 (1990). 6 E. H. Radany and E. C. Friedberg, J. Virol. 41, 88 (1982). 7 L. H. Breimer and T. Lindahl, J. Biol. Chem. 259, 5543 (1984). 8 p. W. Doetsch, D. E. Helland, and W. A. Haseltine, Biochemistry 25, 2212 (1986). 9 K. Kubo, H. Ide, S. S. Wallace, and Y. W. Kow, Biochemistry 31, 3703 (1992). l0 p. C. Seawell and A. K. Ganesan, in " D N A Repair: A Laboratory Manual of Research Procedures" (E. C. Friedberg and P. C. Hanawalt, eds.), Vol. 1B, p. 425. Dekker, New York and Basel, 1981. II A. G. Braun, in " D N A Repair: A Laboratory Manual of Research Procedures" (E. C. Friedberg and P. C. Hanawalt, eds.), Vol. 1B, p. 447. Dekker, New York and Basel, 1981.
[3]
DETECTING REPAIR ENZYMES FOR OXIDATIVE DAMAGE
35
cleavage by chemical agents such as neocarzinostatin and bleomycin. 12 End-labeled DNA substrates are incubated with the proteins of interest under conditions which eliminate the activities of interfering, nonspecific endonucleases. The DNA scission products are analyzed on DNA sequencing gels to determine both the nucleotide location of phosphodiester backbone breakage and details relevant to the mode of DNA strand cleavage. In our hands, this method can detect a significant number of DNA cleavage events following the incubation of approximately 0.01 ng of a highly purified DNA repair enzyme such as Escherichia coli endonuclease III with 4 ng of a DNA substrate containing thymine glycol.13 There are multiple advantages to the use of this method for characterizing DNA repair enzymes. The use of a substrate which contains several chemically different DNA lesions distributed at various locations within the same DNA fragment population allows for the identification of multiple enzyme activities contained within a single cellular extract, provided the proteins are separable using conventional chromatographic methods. In other situations, DNA substrates containing only a single, defined type of base lesion are useful in the comparison of functionally similar enzymes isolated from different species and cell types. Finally, this method also has the advantage that one can examine the chemical nature of the termini produced during enzyme-mediated DNA strand scission in order to gain insight into the mechanism of DNA cleavage by DNA repair enzymes.
General Approach The DNA substrate is a plasmid restriction fragment labeled on either the 5' or the 3' end of one DNA strand. The labeled DNA substrate of 50-250 base pairs (bp) in length is purified by nondenaturing polyacrylamide gel electrophoresis. TM As depicted in Fig. l, end-labeled DNA is then modified to contain the lesion(s) of choice by subjecting the DNA to various physical or chemical damaging agents under conditions which generate approximately one DNA base lesion per DNA molecule. For example, UV-irradiation,15 3~_irradiation,2 and treatment with copper plus H20216 generate a broad spectrum of different oxidative lesions. Other 12 W. A. Haseltine, C. P. Lindan, A. D. D'Andrea, and L. Johnsrud, this series, Vol. 65, p. 325. t3 L. Augeri and P. W. Doetsch, unpublished results (1992). 14 p. W. Doetsch, G. L. Chart, and W. A. Haseltine, Nucleic Acids Res. 13, 3285 (1985). ~5 W. Harm, "Biological Effects of Ultraviolet Radiation." Cambridge Univ. Press, London, 1980. 16 J.-L. Sagripanti and K. Kraemer, J. Biol. Chem. 264, 1729 (1984).
36
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR 3'- end-labeled DNA
5'
T
C
T T
o
c c
G O
[3]
. a,<_.u p . ;~
Oxidative DNA Damage
DNA containing modifiol bas~
5'
~) T T
C ~
O
* 3'
o
C
~)
.~
Enzymatic DNA Cleavage
Oeaved DNA 5 ' m fr~ts
C
T r
C
G O
~ 3' .
Denaturing PAGE
DRm 5 t
A T T G C A A G C 3'
G
GA CT C DNA DNA
•m
m
m
m
mm
m
m
m
<--- uncleared, full-length DNA ~pair enzyme. cleavage at sites of oxidative DNA dAmAge
m m m
m
Gel Autoradiogram
FIG. 1. Scheme depicting the use of end-labeled DNA fragments of defined sequence containing various oxidative base damages in conjunction with base-specific chemical cleavage reactions for the detection and characterization of eukaryotic enzymes that recognize oxidative DNA damage.
agents such as OsO417 produce almost exclusively a single type of base damage (thymine glycol). The substrate DNA is then incubated with a cellular extract or enzyme preparation under conditions which inhibit nonspecific nucleases (10 mM EDTA). The reaction products are analyzed by denaturing polyacrylamide gel electrophoresis alongside of the Maxam and Gilbert base-specific DNA sequencing reaction products is of the same fragment in order to determine the exact nucleotide site of cleavage. 17 K. Burton and W. T. Riley, Biochem. J. 98, 70 (1966). m A. M. Maxam and W. Gilbert, this series, Vol. 65, p. 499.
[3]
DETECTING REPAIRENZYMESFOR OXIDATIVEDAMAGE
37
Identification and Characterization of Multiple Proteins in Saccharomyces cerevisiae That Recognize Oxidative DNA Damage Ultraviolet radiation exposure to DNA induces a variety of DNA base damage products including cyclobutane pyrimidine dimers, (6-4)pyrimidine-pyrimidone photoproducts, monobasic pyrimidine hydrates (e.g., cytosine hydrate),15 and 8-oxoguanine. 19A DNA substrate which has been irradiated with high doses of UV light can therefore be exploited to identify multiple enzymatic activities within the same cellular extract. Individual enzymatic activities are distinguished from one another by the specific pattern of DNA cleavage products which they produce on a DNA sequencing gel following incubation with a UV-irradiated DNA substrate. A comparison of the enzyme-mediated DNA cleavage products with the basespecific chemical cleavage reaction products will reveal the exact nucleotide location of incision by the enzymes of interest. In addition, the DNA scission products produced by various well-characterized DNA repair enzymes and chemical treatments of oxidatively or radiation-damaged DNA serve as important tools for comparing the behavior of the eukaryotic DNA repair enzymes under investigation. UV light-induced cyclobutane pyrimidine dimers are specifically recognized and cleaved by T4 endonuclease V (T4 endo V) 2° and cytosine photohydrates and thymine glycol are cleaved by E. coli endonuclease III (endo III).17 Preparation of UV-Damaged DNA Substrates. A 3'-end-labeled 92-bp EcoRI-PvuII restriction fragment 14'21generated from the plasmid pGEM-2 (Promega, Madison, WI) is irradiated on ice with 10 kJ/m 2 UV light (Model UVGL-25 mineralight lamp, 254 nm) in 10 mM Tris (pH 8.0), 1 mM EDTA (TE buffer). Radiant incidence is measured using a UVX Radiometer (UVP, Inc., San Gabriel, CA). Incubation of DNA Substrates with Eukaryotic Enzyme Preparations. A cellular extract from Saccharomyces cerevisiae, from which the genomic DNA has been removed, is loaded onto a phosphocellulose cationexchange column under low-salt conditions and eluted with a 0.05-1.0 M NaC1 gradient as described previously, z2 Aliquots (25 /zl) of the 12-ml column fractions collected are incubated with UV-irradiated DNA in a total reaction volume of 100/zl under the following buffer conditions: 15 19B. C. Beehler, J. Przybyszewski,H. B. Box, and M. F. Kulesz-Martin,Carcinogenesis (London) 13, 2003 (1992). 2oE. C. Friedberg, Photochem. Photobiol. 21, 277 (1975). 21S. Tabor and K. Struhl, in "Current Protocols in MolecularBiology"(F. M. Ausubel, R. Brent, R. E. Kingston,D. D. Moore,J. G. Seidman,J. A. Smith,and K. Struhl, eds.), p. 3.5.7. Greene PublishingAssociatesand Wiley(Interscience), New York. 22j. Gossett, R. P. Cunningham,and P. W. Doetsch,Biochemistry27, 2629 (1988).
38
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[3]
mM KH2PO 4 (pH 6.8), 10 mM EDTA, and 10 mM 2-mercaptoethanol (buffer B). The final NaCI concentration varies from 14 to 250 mM in this activity assay owing to the variable NaC1 concentration of the column fractions. Incubations with purified T4 endo V (7.5 ng) and endo III (80 ng) are performed in an identical manner, except that they contain a final KC1 concentration of 10 and 40 mM, respectively. Enzyme incubations are for 60 min at 37°. Reactions are terminated by extracting the mixture with phenol-chloroform-isoamyl alcohol (25:24: 1, v/v/v) three times. The DNA samples are recovered by ethanol precipitation, then resuspended in TE buffer. Analysis of DNA Cleavage Products. The DNA samples from above are lyophilized, resuspended in formamide loading dye (80% formamide, 0.2% bromphenol blue, 0.2% xylene cyanol), and loaded onto a 15% polyacrylamide DNA sequencing gel containing 7 M urea and run in TBE buffer [89 mM Tris (pH 8.0), 89 mM boric acid, 2 mM EDTA] at 1500 V for approximately 6 hr. The samples are heated to 90 ° for 30 sec prior to loading. The Maxam and Gilbert base-specific chemical cleavage reaction products are generated as described ~8 and treated in the same manner as the enzyme-generated DNA cleavage products. The DNA sequencing gels are visualized by autoradiography. Interpretation of Results. Figure 2 depicts the DNA cleavage products resulting from the incubation of aliquots of column fractions with the UVirradiated end-labeled DNA substrate. Multiple endonuclease activities are detectable in these fractions, many of which are not completely resolved from one another by this chromatographic step. One activity (arrow 1, Fig. 2) is present in the highest amount in fraction 88, with the greatest extent of strand cleavage occurring at nucleotide C50. This is also a predominant site of cleavage by endo III and corresponds to cleavage at the site of a cytosine hydrate. The yeast enzyme responsible for this activity has been previously identified as being analogous to endo III and is termed yeast redoxyendonuclease. 22 A second endonuclease activity (arrow 2, Fig. 2) is detected in fractions 64 and 72. The greatest extent of DNA cleavage occurs at nucleotide T45, also a site of cleavage by T4 endo V, and occurs at the location of a cyclobutane pyrimidine dimer located at T44-T45 in the DNA substrate. The yeast enzyme responsible for this cleavage is termed yeast pyrimidine dimer endonuclease, and it has been identified as an analog of T4 endo V in S. cerevisiae. 23 A third yeast endonuclease activity can be seen in a relatively wide range of fractions (72-120), with a peak of activity (arrows 3, Fig. 2) in fractions 104 and 112. This enzyme cleaves UV-damaged 23 K. K. Hamilton, P. M. H. Kim, and P. W. Doetsch, Nature (London) 356, 725 (1992).
[3]
39
DETECTING REPAIR ENZYMES FOR OXIDATIVE DAMAGE
-~lE
Z
;~,
~
i::: -'"
~
t" =
00(3 ~61. t;,gl. 9LI. 891. 091. 1~1. 9£1. 8~1. 0~1. ~1.1. "1~01. 96
~
~o
. Z ~
.-= ~ ~ c:
O8 ~L t;,9 9~
~'N
,,,.=
,~..=,.,
==,-- ~ * - , =
~ o
"~ _ ~
~-~- o ~ ~ . . ' -
01~ 0
;~ ._~ ~
E ~ ~ ~.=--.~-
,-,i~g
"~"
(.'3 (Si,~l:(~(JOO~C,~Jl-
I-- ¢.O(D¢.~Jl- ¢.90~:(.90 i,O
• U3
t
ttt
¢~ ~ ' . ,~~
o
°.~"~
"'~
'
40
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[3]
DNA at sites of guanine and probably recognizes 8-oxoguanine which occurs in DNA under certain UV irradiation conditions.19 This putative 8-oxoguanine endonuclease has not yet been completely characterized in yeast. Hence, in a single experiment this method has identified three endonucleases which recognize different classes of UV-damaged DNA bases. Two of the three activities are directed toward oxidative DNA damage. Further use of such UV-irradiated DNA substrates also serves as a highly specific activity assay to monitor these activities during subsequent purification steps. Comparison of Redoxyendonucleases Isolated from Various Sources Both prokaryotic and eukaryotic cells possess a set of enzymes which recognize and remove a variety of oxidatively modified DNA bases. The substrate specificity of these enzymes, termed redoxyendonucleases, includes products of pyrimidine ring saturation (e.g., thymine glycol), ring contraction (e.g., 5-hydroxy-5-methylhydantoin), and ring fragmentation (e.g., urea), as well as other types of modified DNA bases. 3-5,24Redoxyendonucleases have been detected and characterized from a variety of sources including human HeLa, lymphoblast, and fibroblast cells, mouse plasmacytoma cells, S. cerevisiae, Micrococcus luteus, and E. coli. The best characterized member of the family is E. coli endonuclease III. 24'25 Redoxyendonucleases from all sources examined thus far appear to act in a similar manner, via a combined N-glycosylase/AP lyase activity in the initial steps of the base excision repair of oxidatively damaged DNA.3-5 The use in the cleavage assay of a DNA substrate which contains only thymine glycol (TG), a major oxidative DNA base damage product, 2 allows for the determination of the base specificity of cleavage by redoxyendonucleases isolated from various species and cell types. Preparation of Osmium Tetroxide-Damaged DNA Substrates. A 3'end-labeled 211-bp SalI-PvuII restriction fragment 21 generated from the plasmid pUC1826 is damaged with osmium tetroxide (OsO4) under conditions which introduce primarily thymine glycol into the DNA substrate. 17 The DNA is incubated with 400/zg/ml of OsO 4 in a total reaction volume of 100/A for 20 min at 70°. The reaction is terminated by extracting three times with 200 kd of diethyl ether to remove the OsO4. DNA samples are ethanol precipitated and resuspended in TE buffer. The distribution of TG in the DNA substrate is confirmed by treating OsO4-damaged DNA 24 L. H. Breimer and T. Lindahi, J. Biol. Chem. 259, 5543 (1984). 25 H. L. Katcher and S. S. Wallace, Biochemistry 22, 4071 (1983). 26 C. Yanisch-Perron, J. Vieira, and J. Messing, Gene 33, 103 (1985).
[3]
DETECTING REPAIR ENZYMES FOR OXIDATIVE DAMAGE
HeLa LB GA CT ~TG TG
CT TG
41
HA E III TG TG
5' C T G A C A T G 40- A T T A C G A A T T C G A
30-G C T
C G 3'
FIG. 3. Redoxyendonuclease cleavage of DNA at sites of thymine glycol. 3'-End-labeled OsO4-damaged DNA (lanes TG) and undamaged DNA (blank lanes) were incubated with redoxyendonuclease isolated from HeLa cells (HeLa), human lymphoblasts (LB), calf thymus (CT), and endo III (E III) or subjected to hot alkali treatment (HA). The purine-specific (GA) and the pyrimidine-specific (CT) chemical cleavage DNA sequencing reactions were loaded adjacent to the enzyme reaction lanes. Arrows indicate redoxyendonuclease incision at sites of TG. Base numbering starts from the 3'-endqabeled terminus of the DNA fragment. (Autoradiogram reproduced with permission from Doetsch et al. 28)
with 1 M piperidine in a volume of 100 ~1 at 90 ° for 30 min (hot alkali treatment). Under these conditions DNA is quantitatively cleaved at sites of TG. 27
Redoxyendonuclease Digestion of Thymine Glycol-Containing DNA. Thymine glycol-containing DNA (lanes TG, Fig. 3) and undamaged DNA (blank lanes, Fig. 3) are digested with partially purified redoxyendonuclease preparations 28 from HeLa cells (lanes HeLa), human lymphoblasts (lanes LB), calf thymus (lanes CT), and highly purified E. coli endo III (lanes E III). The incubations are conducted in a total volume of 40/zl in 10 mM Tris (pH 8.0), 10 mM EDTA, 10 mM 2-mercaptoethanol, and 40 mM KCI for 30 rain at 37 °. Reactions are terminated and the DNA recovered for analysis in DNA sequencing gels and autoradiography as described above. 27 T. Friedman and D. M. Brown, Nucleic Acids Res. 5, 615 (1978), 2s p. W. Doetsch, W. D. Henner, R. P. Cunningham, J. H. Toney, and D. E. Helland, Mol. Cell. Biol. 7, 26 (1987).
42
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[3]
Interpretation of Results. All of the redoxyendonuclease preparations produced a similar pattern of cleavage at sites of thymine glycol (arrows, Fig. 3). None of the enzymes had activity against undamaged DNA. Variations in the extent of cleavage by the enzymes from different sources are due to the variable extent of redoxyendonuclease purification among the different preparations. The electrophoretic mobilities of enzyme-generated DNA scission products were the same for all the redoxyendonucleases and were identical to those produced by hot alkali treatment of TG-containing DNA (lane HA). This result suggests that this collection of enzymes from different cell sources recognizes and cleaves oxidatively damaged DNA in a similar manner. Therefore, the use of a DNA substrate which contains only thymine glycol in this assay system reveals that redoxyendonucleases isolated from various sources have similar base specificity of cleavage and incise the DNA backbone in a similar manner. Determining Mode of DNA Strand Cleavage The sequencing method has the additional advantage of revealing information about the nature of the termini produced following enzyme-mediated DNA strand scission and provides direct information concerning the mechanism of DNA strand cleavage by DNA repair enzymes. 8 This is accomplished by posttreatment of DNA scission products with enzymes which alter the 3' and 5' termini. For example, the nonspecific phosphatase activity of calf alkaline phosphatase 29or the specific 3'-phosphatase activity of T4 polynucleotide kinase 3° can be utilized to determine if the DNA strand scission products contain phosphoryl groups at either the 5' or 3' termini, respectively. DNA fragments containing phosphoryl groups at their termini migrate at faster rates through DNA sequencing gels than do DNA fragments of equivalent length from which the phosphoryl group has been removed. 31'32 Preparation of Y- and Y-End-Labeled Osmium Tetroxide-Damaged DNA Substrates. A 201-by SaII-PvuII restriction fragment is generated from plasmid pUC1926 and either 3'-end-labeled (Fig. 4A) with a-32Plabeled deoxynucleoside triphosphates (dNTPs) and Klenow enzyme zl or 29 S. Tabor, in "Current Protocols in Molecular Biology" (F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl, eds.), p. 3.10.1. Greene Publishing Associates and Wiley (Interscience), New York. 3o C. C. Richardson, in "The Enzymes" (P. D. Boyer, ed.), 3rd ed., Vol. 14, p. 299. Academic Press, New York, 1981. 31 L. K. Gordon and W. A. Haseltine, J. Biol. Chem. 256, 6608 (1981). 32 D. P. Tapper and D. A. Clayton, Nucleic Acids Res. 9, 6787 (1981).
[3]
DETECTING REPAIR ENZYMES FOR OXIDATIVE DAMAGE
43
A 5.GACT
1
2
3
4
1
2
3
4
5
6
7
8
A
GA CT 3' A "~ C G
5
6
7
8
'~
15-T
A 5" FIG. 4. 5'- and 3'-terminal analysis of human redoxyendonuclease-generated DNA scission products. 3'-End-labeled (A) and 5'-end-labeled (B) TG-DNA (lanes 1-3 and 5-7) and undamaged DNA (lanes 4 and 8) were incubated with HRE (lanes 1, 3, and 4) or treated with hot alkali (lanes 5, 7, and 8). A portion of the enzyme- and hot alkali-generated products were treated separately with the 5'-phosphatase activity of calf alkaline phosphatase (A) or with the 3'-phosphatase activity associated with T4 polynucleotide kinase (B) to remove the 5'- and 3'-terminal phosphoryl groups from the DNA fragments. Lanes 2 and 6 represent mixtures of DNA products analyzed in lanes 1 and 3 and lanes 5 and 7, respectively. The purine-specific (GA) and pyrimidine-specific (CT) chemical cleavage DNA sequencing reactions are loaded in the far left-hand lanes. Base numbering starts from the end-labeled terminus, and the arrows indicate differences in the mobilities of DNA fragments with (arrows b) and fragments without (arrows a) terminal phosphoryl groups. For simplicity, only a portion of the autoradiogram is shown in (A) and (B) to demonstrate the 5'- and 3'terminal analysis, respectively, of only one DNA strand scission product for each assay.
5'-end-labeled (Fig. 4B) with [y-32p]ATP and polynucleotide kinase. 33 The end-labeled DNA substrates are treated with OsO4 as described above to produce TG-containing DNA substrates.
Redoxyendonuclease Digestion of Thymine Glycol-ContainingDNA. Damaged DNA (lanes 1-3 and 5-7, Fig. 4) and undamaged DNA (lanes 4 and 8, Fig. 4) are incubated with 9/.~g of a partially purified human redoxyendonuclease preparation (HRE) isolated from human lymphoblasts (as described previously z8) or subjected to hot alkali treatment. Incubations are carried out in buffer B (see above) plus 40 mM KC1 at 37° in a total volume of 50/.d. Incubations are terminated after 30 min 33 S. Tabor, in "Current Protocols in Molecular Biology" (F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl, eds.), p. 3.10.2. Greene Publishing Associates and Wiley (Interscience), New York, 1989.
44
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[3]
and the DNA recovered by phenol-chloroform-isoamyl alcohol extraction and ethanol precipitation.
Posttreatment of Redoxyendonuclease-Generated DNA Cleavage Products. A portion of the HRE or hot alkali-generated 3'-end-labeled DNA scission products are further digested with 25 U of calf alkaline phosphatase (5'-phosphatase) in 20 tzl TE buffer at 37° for 4 hr. This treatment removes the phosphoryl group from the 5' terminus, leaving a hydroxyl group in its place, z9 A portion of the HRE or hot alkali-generated 5'-end-labeled DNA scission products is digested with 20 U of T4 polynucleotide kinase (3'-phosphatase) in 20 /xl of the reaction buffer [25 mM 2-(N-morpholino)ethanesulfonic acid (pH 6.0), 8 mM MgCI2, 1.5 mM ammonium acetate, and 5 mM 2-mercaptoethanol] at 37° for 4 hr. This treatment removes phosphoryl groups from 3' termini, producing DNA fragments containing 3'-terminal hydroxyl groups. 3° The DNA is recovered and analyzed on DNA sequencing gels as described above. Interpretation of Results. The 5'-terminal analysis (Fig. 4A) reveals that the HRE cleaved TG-DNA between the damaged base and its adjacent 3'-phosphoryl group of the phosphodiester backbone to produce a strand scission product containing a 5'-terminal phosphoryl group (lane 1, Fig. 4). The HRE-generated cleavage product has an identical electrophoretic mobility as that produced by hot alkali (lane 5, Fig. 4), a treatment known to produce fragments with both 5' and 3' phosphoryl groups (arrows b, Fig. 4). 27 The presence of a Y-terminal phosphoryl group was further confirmed on digestion with alkaline phosphatase to remove the 5'-phosphoryl group (lanes 3 and 7, Fig. 4), producing a DNA fragment with a slower electrophoretic mobility (arrows a, Fig. 4). The 3'-terminal analysis (Fig. 4B) reveals that the HRE also cleaves TG-DNA between the damaged base and its adjacent 5'-phosphoryl group to produce a strand scission product containing a 3'-terminal phosphoryl group (lane 1, Fig. 4). The comigration of the HRE- and hot alkali-generated products (lane 5, Fig. 4) and the conversion to a slower migrating fragment containing a 3'hydroxyl group following 3'-phosphatase treatment (lanes 3 and 7, Fig. 4) confirm this notion. It should be noted that phosphatase removal in both the 3' and 5' analyses was not complete, leaving a fraction of the DNA cleavage products containing terminal phosphoryl groups and producing a doublet band pattern (arrows b, lanes 3 and 7, Fig. 4A,B).
Acknowledgments We thank R. Cunningham and R. S. Lloyd for gifts of purified endo III and T4 endo V, respectively. This work was supported by Grants CA42607, CA01441, and Training Grant T32 GM08367 from the National Institutes of Health and Grant NP-806 from the American Cancer Society.
[4]
LOCALIZING STRAND BREAKS IN PLASMID D N A
45
[4] Localization of Strand Breaks in Plasmid D N A Treated
with Reactive Oxygen Species By
WOLFGANG A . SCHULZ, M A I K S. W . OBENDORF, a n d H E L M U T SIES
Introduction A procedure for determining the localization of single-strand breaks in plasmid DNA is describedJ The damaged strand of DNA is selectively end-labeled, and its size is defined on a denaturing polyacrylamide gel. Damage to DNA other than strand breaks, such as base modifications that may interfere with other methods, 2 do not usually affect the result. The intensity of the signals obtained allows an estimation of the proportion of strands broken at each site. This method has been used to show that strand breaks induced by singlet oxygen in a metallothionein promoter contained in a plasmid occur selectively at guanosine residues, with similar frequency at different positions in the sequenceJ The method is also suitable for examining other agents that damage DNA, and it is applicable to sequences other than the metallothionein promoter. Principle of Method The plasmid vector pBluescriptII is treated with the activated oxygen species of choice; the damaging agent is then removed, and the amount and type of damage inflicted are examined by agarose gel electrophoresis. Appropriate samples and controls, including untreated plasmid, are digested to completion with a restriction enzyme leaving a Y-protruding end, usually KpnI (Fig. 1). Then the plasmid is cut a few bases away with a restriction enzyme that leaves Y-protruding ends (e.g., XhoI). The 5' overhang is filled in with Sequenase 2.0, a DNA polymerase devoid of exonuclease activity, using appropriate deoxynucleotide triphosphates with the last nucleotide to be filled in labeled at the a-phosphate group (e.g., dTTP and [a-32p]dCTP may be used for the XhoI site). Following denaturation of the plasmid DNA, two pieces of DNA remain labeled. A 13-bp fragment of DNA between the KpnI and the XhoI site will result from one strand ( + ) of every plasmid molecule. A second fragment will 1 T. P. A. Devasagayam, S. Steenken, M. S. W. Obendorf, W. A. Schulz, and H. Sies, Biochemistry 30, 6283 (1991). 2 D. T. Ribeiro, F. Bourre, A. Sarasin, P. Di Mascio, and C. F. M. Menck, Nucleic Acids Res. 20, 2465 (1992).
METHODSIN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
46
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[4]
(.)
Ks Primer
cgoggt cgocgg totcg
--------q ceGtAc
cc,6ccccccc~ tCGAC.~tCGAC~tArCe
CATGGCC~ G G A G C T
Kpnl
. ~ C ~ C TGCCATAGC Xhol
J __2 A>
pBluescr i pt
C
V
s> ,
FIG. 1. Detection of single-strand breaks by end-labeling. DNA strands are labeled ( + ) and ( - ) ; A, B, and C denote different strand breaks in individual plasmid molecules. The nucleotides added by Sequenase are underlined, and the one carrying the radioactive label is printed in boldface type.
originate from the opposite strand ( - ) of every damaged molecule. This latter fragment will start at a single-strand break and contain the labeled cytidine nucleotide at its 3' end. Its size will indicate the location of the strand break. The denatured end-labeled fragments are separated on a denaturing polyacrylamide sequencing gel alongside marker lanes. DNA fragments for the marker lanes are synthesized by standard dideoxynucleotide termination sequencing reactions using an appropriate primer (e.g., KS primer, Fig. 1) with the plasmid DNA. The marker lanes therefore display the DNA sequence of the ( + ) strand. The locations of strand breaks in the ( - ) strand are determined by reference to the marker lanes as shown in Fig. 2.
C T
~
I I1|.|
C-marker
FIG. 2. Localization of DNA strand breaks. Autoradiograph of a denaturing polyacrylamide gel, showing lanes resulting from (C) a control plasmid and (T) a plasmid treated with UV laser light (W. A. Schulz, S. Steenken, U. Adamek, and H. Sies, unpublished work) by end-labeling as described in the text. The cytosine marker (C-marker) lane was obtained from pBluescript using KS primer (see Fig. 1). Some bases in the C-marker lane are connected by lines to the corresponding bases in the treated plasmid lane.
[4]
LOCALIZING STRAND BREAKS IN PLASMID D N A
47
In this experiment, strand breaks are observed at every base in plasmid DNA treated with UV light. Reference to the cytosine marker lane shows, however, that strand breaks are most frequent at certain G residues. Untreated plasmid on end-labeling yields a weak unspecific smear at high molecular weight. In this fashion, several hundred base pairs of DNA downstream from the KpnI and XhoI site can be examined. Whereas analysis of pBluescript DNA will suffice for many purposes, DNA sequences of interest (e.g., containing a particular frequency of specific bases) may be inserted into the polylinker of pBluescript II. Reagents and Solutions pBluescriptII: 1 ~g//zl, >95% supercoiled form; the plasmid, available from Stratagene (Heidelberg, Germany), is purified from Escherichia coli strains DHFa or XLl-blu by CsCI double banding 3 or by chromatography on affinity columns (Diagen, Dtisseldorf, Germany) TE buffer: 10 mM Tris, 1 mM EDTA, pH 8.0
Agarose Gel Electrophoresis Loading buffer (agarose gel): 20 mM EDTA, 40% glycerol, pH 7.4, 0.4 mg/ml bromphenol blue TBE buffer: 45 mM Tris, 45 mM boric acid, 10 mM EDTA, pH 8.3 Agarose gel: 1% in TBE, boil for 2 min, let cool to 65°, pour Molecular weight marker: h phage DNA digested with BstEII Ethidium bromide: 10 mg/ml stock solution
End-Labeling of Damaged DNA Buffers and nucleotides may have to be changed when enzymes other than KpnI and XhoI are to be used. 10 × Restriction buffer: 100 mM Tris-HC1, pH 7.5, 100 mM MgCI2, I0 mM dithiothreitol (DTT) Restriction enzymes: KpnI and XhoI at - 10 U//.d; restriction enzymes must be devoid of endonuclease activity, which is assayed by incubating overnight 1/xg of supercoiled plasma DNA that does not contain recognition sites for the enzymes used (e.g., pBR322) with 10 U enzyme each in restriction buffer and checking for loss of supercoiled form as described below 3 H. Miller, this series, Vol. 152, p. 145.
48
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[4]
Adaptation buffer: 10 mM Tris-HCl, I0 mM MgCIz, 200 mM NaC1, 1 mM DTT, pH 7.5 Dilution buffer: 30 mM Tris, 14 mM MgCl 2 , 14 mM DTT, pH 7.5, 25 p.M dTTP, 0.5/.tCi//zl [a-3Zp]dCTP (3000 Ci/mmol) Sequenase 2.0: Available from United States Biochemical (Bad Homburg, Germany) Loading buffer (polyacrylamide gel): Included in Sequenase kit, see below
Preparation of Sequence Marker Lanes Sequenase 2.0 kit: Available from United States Biochemical KS primer: Available from Stratagene
Sequencing Gel A 5% polyacrylamide denaturing gel is prepared and run under standard conditions. 4 Procedure
Treatment of Plasmid DNA with Activated Oxygen It is essential to start with a plasmid preparation containing as high a fraction of the intact supercoiled form as possible (usually >95%). Two micrograms DNA is incubated with activated oxygen, the reaction is terminated, and the source of activated oxygen is removed. The DNA is dialyzed and/or purified by ethanol precipitation 5 to remove contaminants from the reaction. Enzymes such as glucose oxidase or xanthine oxidase used to generate activated oxygen species may be removed by proteinase K digestion followed by extraction with high-quality phenol. 6 Finally, the DNA is redissolved in 10 mM Tris, 1 mM EDTA, pH 8.0, at 0.1 mg/ml.
Determining Amount of Damage Before proceeding to localization of the strand breaks, the amount and type of DNA damage are determined by agarose gel electrophoresis. This is useful for several reasons. The method described below will not distinguish between double- and single-strand breaks. Also, among multiple 4 j. Sambrook, E. F. Fritsch, and T. Maniatis, "Molecular Cloning," 2nd Ed., Chapter 13. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1989. 5 D. M. Wallace, this series, Vol. 152, p. 41. 6 D. M. Wallace, this series, Vol. 152, p. 33.
[4]
LOCALIZING STRAND BREAKS IN PLASMID D N A
C
1
2
3
4
5
6
7
8
9
49
L
FIc. 3. Estimation of DNA damage by agarose gel electrophoresis. One microgram of pBluescript plasmid DNA was treated for 10 min with different concentrations of DNase I in the presence of Mg2÷ ; the reaction was terminated by addition of EDTA and heating at 65° for 10 min. The products were analyzed on an agarose gel as described in the text. C, Untreated plasmid; L, h phage DNA digested with BstElI; lanes 1-9, plasmid DNA treated with 10-is-10 -7 U DNase I.
single-strand breaks in the same molecule only the one closest to the restriction sites will be seen. Spontaneous strand breaks will also be detected, and their frequency limits the sensitivity of the assay. Therefore, a level of damage with a maximum of single-strand breaks and few doublestrand breaks is optimal. Moreover, with some reagents the level of damage may be difficult to reproduce. In that case it is advisable to carry several samples differing in exposure to the damaging agent through the procedure in addition to appropriate controls. Aliquots containing approximately 1 p.g plasmid DNA treated with activated oxygen species are mixed with agarose gel loading buffer and electrophoresed on a 1% agarose gel alongside 1 k~g untreated pBluescript DNA. 7 A molecular weight marker, for example, ?~-phage DNA digested with BstEII, is included on the gel. Figure 3 shows a model experiment with DNase I using conditions where the enzyme introduces mostly singlestrand breaks. The control plasmid still contains greater than 90% supercoiled form after incubation in buffer alone, but also a fraction of relaxed circular plasmid form with single-strand breaks that runs more slowly. With increasing concentration of DNase I the proportion of plasmid in the relaxed form increases, and eventually double-strand breaks appear 7 R. C. Ogden and D. A. Adams, this series, Vol. 152, p. 61.
50
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[4]
as indicated by a fragment at 3.0 kb, the size expected for linearized plasmid. Further increases in DNase I concentration cause degradation of DNA detectable as a smear of fragments smaller than 3 kb. In this experiment, samples 4-7 in addition to the control plasmid would have been chosen for further investigation.
End-Labeling Analysis For end-labeling analysis, 7.2 ~1 of each sample in a small reaction tube is mixed with 0.8/zl of 10 × restriction buffer. Eight units KpnI in a maximum volume of 0.8/.d is added (or another chosen enzyme yielding 3'-protruding ends). Following thorough mixing by hand, the reaction mixture is incubated at 37° for 1 hr. Then 4/zl is removed and checked on an agarose minigel for complete digestion. Meanwhile, 4/zl of adaptation buffer and 5 U of XhoI are added to the remainder of the sample, and incubation is continued for at least I hr. After the minigel has been analyzed and digestion by KpnI has been found to be complete, 8/xl of dilution buffer containing the 32p-labeled deoxynucleotide triphosphate and 2.5 U of Sequenase 2.0 are added. Following incubation for another 30 min, 4 ttl of loading buffer is added; the samples are denatured for 2 min at 95 ° and put rapidly on ice, where they may be stored until loading. Approximately 4-5 /zl of each sample is loaded on a standard 5% polyacrylamide sequencing gel 4 alongside the 35S-labeled marker DNA. The four marker lanes should be next to one another, but it is useful to alternate empty ones with the sample lanes, since/3 radiation from 32p may obscure the signal from the markers and from weaker samples and the control. Marker DNA is prepared by running standard sequencing reactions 4 with pBluescript plasmid DNA using an appropriate primer and a-35Slabeled deoxynucleotide triphosphates. With the above enzymes, KS primer is most convenient to use. The marker DNA can be prepared in advance and may be stored for several weeks at - 20°. Following fixation and drying, the gel is exposed to X-ray film. 8 An amplifying screen is usually not required.
Interpretation of Data The location of single-strand breaks is readily determined from the autoradiograph (Fig. 2). Note, however, that the strand investigated for strand breaks is opposite to the one synthesized in the marker DNA. Thus, a band in the C-marker lane corresponds to a strand break at G in 8 W. M. Bonner, this series, Vol. 152, p. 55.
[S]
Fe/H202
DNA
DAMAGE PRODUCTS
51
the sample, and so on. Also, marker DNA and sample fragments do not start at the same position. Depending on which enzyme is used, which nucleotides are filled in, and which primer is employed, marker and samples are off by several bases. For instance, when cutting with XhoI, filling in with dTTP and labeled dCTP, and using KS primer for marker synthesis, the fragments differ in length by only a single base (Fig. 1). To facilitate interpretation of complex damage patterns, it has proved useful to load marker DNAs on both sides of the samples on the sequencing gel. To compare the frequency of single-strand breaks at individual sites, suitable autoradiographs are analyzed using one of several commercially available densitometer systems (e.g., GSXL, LKB, Piscataway, N J).
[5] D e t e c t i n g D N A D a m a g e C a u s e d b y I r o n a n d Hydrogen Peroxide B y YONGZHANG LUO, ERNST S. HENLE, RAJAGOPAL CHATOPADHYAYA, RUCHENG JIN, and STUART LINN
Introduction
This chapter describes methods to assess DNA damage mediated by Fenton reactions via hydroxyl radicals (.OH): Fe 2+ + H202 + H+---~ Fe 3+ + H20 + "OH
Depending on the state of chelation of the iron, including DNA chelation, other oxygen radicals such as ferryl radicals 1 are possible, and these may also be reactive toward DNA. Additionally superoxide radical (02 ~) is formed by the following scavenging reaction: •OH
+
H202 ---> H20 + O2: + H ÷
Identification of the stable products derived from DNA after exposure to Fenton reactions is difficult owing to the multitude of products. Therefore, we have devised methods for observing the damage spectra of the four deoxynucleosides, nucleotides, oligonucleotides, homopolymers, and finally duplex DNA. After enzymatic conversion to nucleosides, products are identified by high-performance liquid chromatography (HPLC) retention times, radiolabeling, mass spectrometry, nuclear magnetic resonance, i j. D. Rush and W. H. Koppenol, J. Biol. Chem. 261, 6730 (1986).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[S]
Fe/H202
DNA
DAMAGE PRODUCTS
51
the sample, and so on. Also, marker DNA and sample fragments do not start at the same position. Depending on which enzyme is used, which nucleotides are filled in, and which primer is employed, marker and samples are off by several bases. For instance, when cutting with XhoI, filling in with dTTP and labeled dCTP, and using KS primer for marker synthesis, the fragments differ in length by only a single base (Fig. 1). To facilitate interpretation of complex damage patterns, it has proved useful to load marker DNAs on both sides of the samples on the sequencing gel. To compare the frequency of single-strand breaks at individual sites, suitable autoradiographs are analyzed using one of several commercially available densitometer systems (e.g., GSXL, LKB, Piscataway, N J).
[5] D e t e c t i n g D N A D a m a g e C a u s e d b y I r o n a n d Hydrogen Peroxide B y YONGZHANG LUO, ERNST S. HENLE, RAJAGOPAL CHATOPADHYAYA, RUCHENG JIN, and STUART LINN
Introduction
This chapter describes methods to assess DNA damage mediated by Fenton reactions via hydroxyl radicals (.OH): Fe 2+ + H202 + H+---~ Fe 3+ + H20 + "OH
Depending on the state of chelation of the iron, including DNA chelation, other oxygen radicals such as ferryl radicals 1 are possible, and these may also be reactive toward DNA. Additionally superoxide radical (02 ~) is formed by the following scavenging reaction: •OH
+
H202 ---> H20 + O2: + H ÷
Identification of the stable products derived from DNA after exposure to Fenton reactions is difficult owing to the multitude of products. Therefore, we have devised methods for observing the damage spectra of the four deoxynucleosides, nucleotides, oligonucleotides, homopolymers, and finally duplex DNA. After enzymatic conversion to nucleosides, products are identified by high-performance liquid chromatography (HPLC) retention times, radiolabeling, mass spectrometry, nuclear magnetic resonance, i j. D. Rush and W. H. Koppenol, J. Biol. Chem. 261, 6730 (1986).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
52
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[5]
and UV absorption spectra. Quantitation is achieved either by UV absorption or radiolabeling. Materials Chromatography Columns. Analytical silica-based C18 reversed-phase (RP) columns (250 × 4.6 mm, 5/zm) and semipreparative normal-phase and RP columns are from Alltech (Deerfield, IL). Polystyrene reversedphase (PRP) columns are from Hamilton (Reno, NV). Saturator columns are constructed from empty analytical columns which are gravity-filled with chromatographic grade 230-400 mesh silica gel from Sigma (St. Louis, MO). Enzymes. DNase I is from Sigma, and units are defined by Kunitz. 2 P1 nuclease is from Boehringer Mannheim (Indianapolis, IN) with units as defined by Fujimoto et al. 3 Snake venom phosphodiesterase (SVP) is from Worthington (Freehold, N J) with units defined by Razell and Khorana. 4 Bacterial alkaline phosphatase (BAP) is from Worthington with units defined by Garen and Levinthal. 5Enzymes are dialyzed extensively versus 25 mM triethylammonium bicarbonate (TEAB), pH 7.5, to remove contaminants which may be mistaken for products and to remove nonvolatile salts which interfere with subsequent analyses by mass spectrometry (MS) or nuclear magnetic resonance (NMR). Substrates. Substrates are repurified by HPLC where possible. When unavailable, radiolabeled deoxynucleoside monophosphates (dNMPs) can be produced from commercially available deoxynucleoside triphosphates (dNTPs) by treatment with S V P 6 and then purified by ion-pairing C~8 HPLC (mobile phase TEAB, pH 7, 0-30% methanol), followed by C18 RP-HPLC (mobile phase water) with a 100-fold molar excess of NaHCO3 injected with the sample. Treatment of nucleotides with BAP will yield the corresponding nucleosides, which are purified by RP-HPLC (mobile phase water). Suitable dNpNs and dNpNpNs are purified by ion-pairing HPLC (mobile phase TEAB, pH 9, 0-30% methanol) on a PRP column. The sample is then further purified on a C18 RP-HPLC column (mobile phase water for 10 min followed by a 0-50% methanol gradient), with a 100-fold molar excess of NaHCO3 injected with the sample. The dNpN and dNpNpN structures are confirmed by SVP digestion, to dNp and dN, and quantitation after HPLC. 2 M. Kunitz, J. Gen. Physiol. 33, 349 (1950). s M. Fujimoto, A. Kuninaka, and H. Yoshino, Agric. Biol. Chem. 38, 777 (1974). 4 W. E. RazeU and H. G. Khorana, J. Biol. Chem. 234, 2105 (1959). 5 A. Garen and C. Levinthal, Biochim. Biophys. Acta 38, 470 (1960). 6 S. E. Pollack and D. S. Auld, Anal. Biochem. 127, 81 (1982).
[5]
Fe/HzO2 DNA DAMAGEPRODUCTS
53
DNA is best obtained from purified virus particles. Covalently closed and supercoiled DNA is preferred, since its integrity is easily tested. Most eukaryotic DNAs contain 5-methylcytosine which adds to the product spectrum, so we use supercoiled phage PM2 DNA which lacks modified bases and which can be radiolabeled in vivo 7 or by nick-translation. The DNA is extensively dialyzed against EDTA then against 25 mM NaCI. It can also be used to monitor nicking by the Fenton reagents. 8 Concentrations of H202 are measured by UV absorbance, assuming an extinction coefficient (e240) of 39.4 M - 1 cm- 1.9 Ferrous concentrations are determined either with Ferrozine, assuming that Fe 2+(Ferrozine)3 has an 8562 value of 28,000 M - 1 cm- 1,10 or with o-phenanthroline, assuming that Fe2+(o-phenanthroline)3 has an e510 value of l 1,100 M -~ cm -1. Methods Fenton Reaction Conditions. Substrates can be subjected to H202induced damage under a variety of conditions using Fe z+ or Fe 3+. NADH
and/or ethanol can be present, and the reaction can be made anaerobic by flushing with N 2 . Two millimolar H202 is used because this concentration gives maximal killing of Escherichia coli. 8 In most cases, damage is proportional to Fe z+ concentrations up to 1.2 mM. For 1 mM substrate, ! mM iron is used to get 5-50% damage, and 2 mM NADH is used where desired. To avoid secondary attack on the highly reactive compounds dC and dCMP, 0.4 mM iron and 0.8 mM NADH are used. Damage caused by freely diffusible .OH can be eliminated in the presence of 100 mM ethanol, an .OH scavenger. When purging O2 with N2, for a 2-ml volume the Oz concentration drops from 8 to less than or equal to 0.1 ppm by 5 min. Prior to H 2 0 2 addition, the pH of reaction mixtures is adjusted with NaOH to between pH 6.5 and 6.8. The order of reactant addition is substrate-iron-(NADH)-(ethanol)-H202 so that the iron can associate with the substrate before it reacts with HzO z or is chelated by NADH. The reaction is started by adding a 10 mM H2Oz stock solution while vortexing and then is maintained at 25° for 30 min. Vortexing during rapid addition of HzO2 is necessary to transient differences in concentrations. Postreaction Digestion to Nucleosides. For most products, separation and purification are best attained by enzymatic digestion to nucleosides followed by C18 RP-HPLC. DNase I digests DNA to short oligomers which 7 U. Kuhnlein, E. E. Penhoet, and S. Linn, Proc. Natl. Acad. Sci. U.S.A. 73, 1169 (1976). 8 j. A. Imlay and S. Linn, Science 240, 1302 (1988). 9 D. P. Nelson and L. A. Kiesow, Anal. Biochem. 49, 474 (1972). I0 T. J. Giovanniello and T. Peters, Jr., in "Standard Methods of Clinical Chemistry" (D. Seligson, ed.), Vol. 4, p. 139. Academic Press, New York, 1963.
54
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[5]
are substrates for P1 nuclease and SVP. Both P1 nuclease and SVP should be used to ensure complete digestion of oligonucleotides to mononucleotides.l~'12 Because deaminases may contaminate SVP preparations, excessive use of the enzyme should be avoided. Mononucleotides are dephosphorylated to nucleosides by BAP. To our knowledge, no damage products are refractory to digestion or altered by these procedures with the exception that we did not observe dephosphorylation of 2-deoxy-D-ribono-8lactone-5-phosphate by BAP. For DNA, reactions contain 1 mM substrate (nucleotide residues), 20 mM Tris-HCl, pH 7.5, 2.25 mM MgC12 , 50/xM CaC12, and 20-100 units/ ml DNase I and are incubated for 30 min at 25 °. Twenty microliters of 1 M sodium acetate, pH 5.2, and 30 units of P1 nuclease are then added per milliliter, and incubation is continued for 30 min. Twenty-five microliters of 1 M Tris-HC1, pH 8.0, 20 units SVP, and 20 units BAP are then added per milliliter, and incubation is continued for 30 min. For homopolymers, the substrate is suspended in 20 mM sodium acetate, pH 5.5, 3 mM MgC12, and digestion with P1 nuclease, SVP, and BAP is carried out as above. For mononucleotides, the substrate is suspended in 20 mM TrisHCI, pH 7.5, and digested with BAP as above. Digests are centrifuged for 1 min at 5000 g, then the supernatants are filtered through 0.22-/xm Millipore (Bedford, MA) filters and finally brought to 10 mM potassium phosphate, pH 6.5. For large-scale reactions of dNs or dNMPs for NMR or MS analysis 1 M N H 4 H C O 3 o r TEAB is added to a 500-ml Fenton reaction mixture to bring the pH to 6-7. After centrifugation for 1 min at 5000 g, the supernatant is passed through a 0.22-/.~m Millipore filter and then dried under vacuum. TEAB is generally preferred as a buffer for desiccation since it is not so volatile as N H 4 H C O 3 . It does not interfere with fast atom bombardment MS in the negative ion mode [FAB-MS(- )], but will interfere in the positive ion mode [FAB-MS(+)]. N H 4 H C O 3 interferes with both modes of FAB, but since it is so volatile, it can be adequately removed from sample preparations destined for FAB-MS(+). The dried sample is redissolved in the mobile phase before HPLC. High-Performance Liquid Chromatography. We use a binary Perkin Elmer (Norwalk, CT) LC-250 pump and a saturator column upstream of the injector to prolong the life of silica-based columns. Mobile phases are purged with helium to minimize bubble formation and UV absorption by oxygen below 210 nm. The pH of the mobile phase is important for obtaining symmetric peaks. When the pH approaches the PKa of an eluate, its II M. Liuzzi, M. Weinfeld, and M. C. Paterson, J. Biol. Chem. 264, 6355 (1989). 12 M. Weinfeld, M. Liuzzi, and M. C. Paterson, Nucleic Acids Res. 17, 3735 (1989).
[5]
Fe/H202 D N A DAMAGE PRODUCTS 4000
-
55
4 dG
3000
o. 0
2000
2
1000
1
67
3
5 8
0
~ 0
5
9
----'-"~ 10 15 Time
1
11
"~"
~ 20
r. 25
-'--I 30
(min)
FIG. 1. Radiochromatogram of products from [I',2'-3H]dG (dashed line) and [8-14C]dG (solid line) after exposure to FeZ÷/H202 under aerobic conditions after RP-HPLC. The reaction mixture contained 1 mM" dG, 1 mM FeSO4, and 2 mM H202 and was processed as described in the text. Approximately 105 cpm of each isotope was injected. Peaks: (1) unknown; (2) 2-deoxy-o-ribono-~-lactone and unknown 247-Da compound; (3) unknown 179-Da compound; (4) unknown 273-Da compound; (5) unknown; (6) 5',8-cyclodeoxyguanosine; (7) guanine; (8) 9-(2'-deoxy-fl-D-erythro-pentopyranosyl)guanine; (9) 9-(2'-deoxy-fl-oerythro-pento-l,5-dialdo-l,4-furanosyl)guanine; (10) 9-(2'-deoxy-a-D-erythro-pentofuranosyl)guanine; (dG) 2'-deoxyguanosine; (11) 5',8-cyclo-2',5'-dideoxyguanosine.
peak may be distorted. We have found that pH 6.5 is well-suited for the separations. Potassium phosphate buffer is used because it can chelate residual iron that would otherwise damage the column. For nucleoside mixtures, a C]8 RP-HPLC column is eluted with 2% methanol v/v in 10 mM potassium phosphate, pH 6.5, at 1 ml/min for 10 min followed by a gradient to 30% methanol v/v 70% 10 mM potassium phosphate, pH 6.5, for 40 min. For dC products, all products elute before the gradient is applied. For resolution of nucleotides by C]8 RP-HPLC for FAB-MS( - ), elution is with 10 mM TEAB, pH 7, for 60 min followed by 100% methanol. Alternatively, for resolution by PRP columns, the TEAB is adjusted to pH 9. Samples for FAB-MS(+) or NMR spectroscopy are purified by RP-HPLC with water at 1 ml/min for 60 min followed by 100% methanol. The dC products are not well-resolved by RP-HPLC because of their greater polarity. Accordingly they are separated first by normal-phase silica HPLC 13 followed by RP-HPLC with 2% methanol in 10 mM potassium phosphate, pH 6.5, before RP-HPLC with water. The HPLC eluates are monitored by a diode array detector. We use a Perkin Elmer LC-480 detector which continuously records UV spectra t3 j. R. Wagner, J. E. van Lier, C. Decarroz, M. Berger, and J. Cadet, this series, Vol. 186, p. 502.
56
OXIDATIVE DAMAGETO DNA AND DNA REPAIR 400 -
10-15 1 -.__55 8 -
a.
dC
16-18
/
300'
0
[5]
200" 21-22 100"
19-20 ~
23
A_
0 0
2
4
6 Time
;
'0
1 '2
1'4
(min)
FIG. 2. Radiochromatogram of products from [UJ4C]dCMP after exposure to Fe2+/H202 under aerobic conditions resolved on RP-HPLC prior to final resolution on normal-phase HPLC. The reaction mixture contained 1 mM dCMP, 0.4 mM FeSO4, and 2 mM H202 and was processed as described in the text. Peaks: (1-5) 2-deoxy-D-ribono-8-1actone-5phosphate, 2'-deoxyribosyl-NLformyl-N2-glyoxylurea,4-amino-l-formyl-5-hydroxy-2-oxoA3-imidazoline, parabanic acid, alloxan; (6-7) 2'-deoxyribosylurea, 5,6-dihydro-5,6-dihydroxydeoxycytidine; (8-9) 2'-deoxyribosylformamide, trans-l-carbamoylimidazolidone-4, 5-diol; (10-15) cis/trans-dihydroxyuracil, cis-dihydroxyuracil, cis/trans-dihydroxydeoxyuridine, cis/trans-dihydroxyuracil, 2,4,5-trihydroxydeoxypyrimidine, 2'-deoxyribosyl-aminol-formyl-5-hydroxy-2-oxo-A3-imidazoline;(16-18) cytosine, 5-hydroxyhydantoin, cis/transdihydroxyuracil; (19-20) 2'-deoxyribosyl-trans-1-carbamoylimidazolidone-4,5-diol,oxaluric acid; (21-22) 2'-deoxyribosyl-5-hydroxyhydantoin, 2'-deoxyribosylbiouret; (23) 1-carbamoyl-l-carboxy-4-(2-deoxy-fl-D-erythro-pentofuranosyl) glycinamide; (dC) 2'-deoxycytidine. for storage on a computer. A diode array detector allows immediate UV spectra for identification of products, quantitation by integration of UV profiles at the maximum absorbance for each product, a useful UV trace without test runs, and determination of peak purity by assessing absorption ratios across a peak. Radiochromatograms have the advantages of reliable quantitation, reduction of the number of observable products to avoid overcrowding, and confirmation of structure by which labeled atoms remain. Tritiated water from 3H abstraction b y . O H or 14CO2 from decarboxylation can be r e m o v e d by desiccation. Substrate radiolabeling is essential for identifying degradation products from iron/NADH/H202 systems, since UV-absorbing N A D H and its degradation products coelute with products of interest. Characterization o f Eluates. Eluates may be characterized by their H P L C retention times, UV absorbance spectra, and/or the use of specifically radiolabeled substrates. If standard compounds are unavailable, MS
[5]
Fe/H202 DNA DAMAGEPRODUCTS
dA
dT
L
0.050 ~D
57
5
A
0.0
-
i
0
i
i
lo
i
I
11
I
I
i
i
5
i
I
i
l
10 Time
i
i
15
i
i
i
i
20
(rain)
FIG. 3. Ultraviolet absorbance profile of products from poly(dA) after exposure of poly (dA) : oligo(dT) to Fe 2+/H202 under aerobic conditions resolved by RP-HPLC. The reaction mixture contained 0.52 mM (dA)2000 and 0.48 mM (dT)z0 (nucleotide residues), 50 mM NaCl, 1 mM FeSO4, and 2 mM H202. After reaction, 8 M was added to disrupt the duplex structure, and the mixture was centrifuged in a Centricon concentrator 30 (Amicon, Danvers, MA) at 4000 rpm. Small molecules (e.g., released bases) and thymine oligomers not associated with the poly(dA) passed through the filter. The filter was washed 7 times with water to remove urea, and the retentate was removed from the filter by reverse centrifugation and then digested to nucleosides and processed as described in the text. Peaks: (1) unknown; (2) 2'-deoxyribosyl-4,6-diamino-5-formamidopyrimidine; (3) unknown from thymine family; (4) unknown; (5) unknown from thymine family; (6, 7) unknown; (8) 2'-deoxyribosyl-5formyluracil; (9) 2'-deoxyribosylisoguanosine; (10) 5',8-cyclo-dA (R and S); (dT) 2'-deoxythymidine; (11) adenine; (12) unknown; (dA) 2'-deoxyadenosine; (13) 8-oxo-7,8dihydrodeoxyadenosine.
and NMR analysis can be used to determine a structure. Samples for NMR are desiccated twice in D20 and redisolved in D20 or deuterated dimethyl sulfoxide (DMSO). We have chosen FAB-MS over other forms of MS since the molecular ion is less likely to be fragmented, and polar samlbles are more easily volatilized. 14 FAB-MS(+) is used for bases and nucleosides, since these generally bear no charge but can associate with a proton under these conditions. F A B - M S ( - ) is sensitive for anionic dNMP and dNpN residues. A glycerol matrix is used as the internal marker. Figures 1-4 are representative chromatograms, with preliminary peak identifications, from a member of the guanine, cytosine, adenine, and thymine families, respectively. 14 A. G. Harrison and R. J. Cotter, this series, Vol. 193, p. 3.
58
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[5]
dT
2501
2001
3
15° I
8
O lOOoi O
J
0
.
.
.
.
i
5
.
.
.
.
i
10
.
.
.
.
i
•
5
20 25 Time (rain)
-
i
30
•
•
,~,
i
35
.
.
.
.
i
40
FIG. 4. Radiochromatogram of products from [methyl, l',2'-3H]dT-labeled PM2 DNA under aerobic conditions resolved by RP-HPLC. The reaction mixture contained 1 mM DNA (nucleotide residues), 10 mM NaC1, 1 mM FeSO4, and 2 mM H202. After 30 min at 25°, the reaction mixture was subjected to enzymatic digestion as described in the text. Peaks: (1) 2'-deoxyribosyl-5-hydroxylmethyluracil; (2, 3, 4, 5) 2'-deoxyribosyithymine glycol (four isomers); (X) contaminant from undamaged DNA substrate; (6) 2'-deoxyribosylpyruvylurea; (7) 2'-deoxyribosyl-5-hydroxy-5-methylhydantoin; (8) thymine; (Y) contaminant from undamaged DNA substrate; (dT) 2'-deoxythymidine.
Discussion The above procedures have been applied to characterize DNA damage at the nucleoside, nucleotide, and polynucleotide levels. Enzymatic conversion to nucleosides avoids harsh chemical hydrolysis conditions which can destroy (or form) products. To our knowledge no products are lost owing to their resistance to enzymatic conversion. However, these procedures are not necessarily recommended as a replacement for chemical hydrolysis before analysis, but rather as a complement to those techniques. An isotope effect is noted with dG (Fig. 1). Singly or multiply labeled [14C]dG coelutes with its nonradioactive counterpart, but [ I ' , 2 ' - 3 H ] d G elutes 2.2% faster under isocratic conditions. The [I',2'-3H]dGTP, from which the nucleoside was prepared, was tested for its authenticity by incorporation into DNA with DNA polymerase I and digesting the purified DNA to nucleosides. The recovered 3H-labeled material behaves in the manner described above. Similar 3H isotope effects have been previously reported. 15,16 15 K. Frenkel, M. S. Goldstein, and G. W. Teebor, Biochemistry 20, 7566 (1981). 16 p. D. Klein, in "Advances in Chromatography" (J. C. Giddings and R. A. Keller, eds.), pp. 3-65. Dekker, New York, 1966.
[6]
PHOTOCHEMICAL
SYNTHESIS
OF 8-HYDROXYGUANINE
59
Acknowledgments This research was supported by U.S. Public Health Service Grants R37 GM19020, T32 ES07075, and P30 ES01896 and a grant from the Chevron Corporation Risk Assessment Program. We are indebted to Drs. Richard Wagner and Clinton Ballou for helpful advice, to Sherry Ogden for MS analyses, and to Richard Mazzarisi and Graham Ball for NMR analyses.
[6] P h o t o c h e m i c a l 8-Hydroxyguanine
Synthesis
of
Nucleosides
B y PETER K. WONG and ROBERT A. FLOYD
Introduction O x y g e n free radicals have been implicated in the etiology of processes associated with m a n y diseases and pathological states. The molecular basis of the action of oxygen free radicals is probably manifold but most likely involves oxidative damage to macromolecules including proteins and nucleic acids. Oxygen free radical-mediated damage to nucleic acids to form base adducts represents an attractive notion to explain their effect in cancer development. Kasai and Nishimura I first noted that reducing agents, involving the action of hydroxyl free radicals, mediated formation of 8-hydroxy-2'-deoxyguanosine (8-oxo-dG) when these agents acted on either isolated D N A or the free nucleoside. Since these first observations, n u m e r o u s biological models have been investigated to determine if 8-oxo-dG can be implicated in carcinogenesis. The status of research in this area has been summarized, and it is clear that in models where oxidative damage has been implicated, there is a close association between the increased presence of 8-oxo-dG and the d e v e l o p m e n t of cancer, z In addition, there have been extensive studies devoted to the use of 8-oxodG excretion in the urine as an index of oxidative stress and as a p a r a m e t e r to assess the role of oxygen free radicals in aging.3 Singlet oxygen produced by the action of light on methylene blue or by chemical systems has been shown to form 8-oxo-dG in D N A . 4,5 The use of high-performance liquid l H. Kasai and S. Nishimura, Nucleic Acids Res. 12, 2137 (1984). z R. A. Floyd, Carcinogenesis (London) 11, 1447 (1990). 3 j. McCann, V. Simmon, D. Streitwieser, and B. N. Ames, Proe. Natl. Acad. Sci. U.S.A. 72, 3190 (1975). 4 R. A. Floyd, M. S. West, K. L. Eneff, and J. E. Schneider, Arch. Biochem. Biophys. 273, 106 (1989). 5 T. P. A. Devasagayam, S. Steenken, M. S. W. Obendorf, W. A. Schulz, and H. Sies, Biochemistry 30, 6283 (1991).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[6]
PHOTOCHEMICAL
SYNTHESIS
OF 8-HYDROXYGUANINE
59
Acknowledgments This research was supported by U.S. Public Health Service Grants R37 GM19020, T32 ES07075, and P30 ES01896 and a grant from the Chevron Corporation Risk Assessment Program. We are indebted to Drs. Richard Wagner and Clinton Ballou for helpful advice, to Sherry Ogden for MS analyses, and to Richard Mazzarisi and Graham Ball for NMR analyses.
[6] P h o t o c h e m i c a l 8-Hydroxyguanine
Synthesis
of
Nucleosides
B y PETER K. WONG and ROBERT A. FLOYD
Introduction O x y g e n free radicals have been implicated in the etiology of processes associated with m a n y diseases and pathological states. The molecular basis of the action of oxygen free radicals is probably manifold but most likely involves oxidative damage to macromolecules including proteins and nucleic acids. Oxygen free radical-mediated damage to nucleic acids to form base adducts represents an attractive notion to explain their effect in cancer development. Kasai and Nishimura I first noted that reducing agents, involving the action of hydroxyl free radicals, mediated formation of 8-hydroxy-2'-deoxyguanosine (8-oxo-dG) when these agents acted on either isolated D N A or the free nucleoside. Since these first observations, n u m e r o u s biological models have been investigated to determine if 8-oxo-dG can be implicated in carcinogenesis. The status of research in this area has been summarized, and it is clear that in models where oxidative damage has been implicated, there is a close association between the increased presence of 8-oxo-dG and the d e v e l o p m e n t of cancer, z In addition, there have been extensive studies devoted to the use of 8-oxodG excretion in the urine as an index of oxidative stress and as a p a r a m e t e r to assess the role of oxygen free radicals in aging.3 Singlet oxygen produced by the action of light on methylene blue or by chemical systems has been shown to form 8-oxo-dG in D N A . 4,5 The use of high-performance liquid l H. Kasai and S. Nishimura, Nucleic Acids Res. 12, 2137 (1984). z R. A. Floyd, Carcinogenesis (London) 11, 1447 (1990). 3 j. McCann, V. Simmon, D. Streitwieser, and B. N. Ames, Proe. Natl. Acad. Sci. U.S.A. 72, 3190 (1975). 4 R. A. Floyd, M. S. West, K. L. Eneff, and J. E. Schneider, Arch. Biochem. Biophys. 273, 106 (1989). 5 T. P. A. Devasagayam, S. Steenken, M. S. W. Obendorf, W. A. Schulz, and H. Sies, Biochemistry 30, 6283 (1991).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
60
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[6]
2ram ID
FIG. 1. Glass coil used to carry out the irradiation reaction in the synthesis of 8-OH derivatives of guanosine or deoxyguanosine by the action of methylene blue and light.
chromatography with electrochemical detection (HPLC/ED) as an extremely sensitive method to measure 8-oxoguanine in DNA and RNA 6 has prompted us to develop newer and easier methods to prepare the 8-oxo-dG and 8-hydroxyguanosine (8-oxo-G). Here we present methods based on the use of methylene blue plus light. The ease of synthesis by these newer methods is valuable for researchers in the field.
Materials and Methods
Reagents Guanosine, 2'-deoxyguanosine, and 8-hydroxyguanine; Sigma Chemical Company (St. Louis, MO) Methylene blue; Aldrich Chemical Company (Milwaukee, WI) AG 50W-X2 cation-exchange resin (200-400 mesh, H + form); BioRad Laboratories (Richmond, CA) Sephadex G- 15; Pharmacia Fine Chemicals (Piscataway, N J)
Synthesis of 8-Hydroxyguanosine by Methylene Blue and Light Prepare a reaction mixture consisting of 450 ml of 2.2 mM guanosine in water and 50 ml of 1.0 mM methylene blue in water. The above mixture is stored in a reservoir and allowed to slowly pass through a glass coil having dimensions as shown in Fig. 1. The flow rate of the mixture is adjusted so that a portion of the mixture is exposed to a 100-W flood light at a distance of 15 cm for approximately 30 min. The liquid, which contains unreacted guanosine, 8-oxo-G, and a small amount of by-products as well 6 R. A. Floyd, J. J. Watson, P. K. Wong, D. H. Altmiller, and R. C. Rickard, Free Radical Res. C o m m u n . 1, 163 (1986),
[6]
PHOTOCHEMICAL SYNTHESIS OF 8-HYDROXYGUANINE
61
as methylene blue, is collected as it exits from the coil and is stored at 4° in the dark. The reaction mixture is then first passed through an AG 50W-X2 cation-exchange column, 38 x 50 mm, at a flow rate of 1.5 ml/ min. Methylene blue and unreacted guanosine are strongly retained on the column while 8-oxo-G and some by-products are eluted more quickly. After all the liquid containing the reaction products has passed through the column, 190 ml of deionized water is added to the column to wash out the remaining 8-oxo-G. The first 160 ml of eluate does not contain 8-oxo-G and is discarded, and the remaining clear eluate is collected and evaporated under vacuum to about 150 ml. The solution containing 8-oxo-G is then loaded on to a second AG 50W-X2 column, 38 x 120 mm, and allowed to run through at a rate of 1.5 ml/min. Deionized water is added to continue the flow. The eluate is collected in 20-ml fractions. Fractions 14 through 25 are rich in 8-oxo-G and are pooled together for a total of 220 ml. Because the elution profile changes as the chromatographic conditions vary, it is necessary to determine if 8-oxo-G is present by injecting each fraction into a HPLC/ED system. The pooled fractions are then evaporated under vacuum to a volume of about 2 ml and stored at 4° for crystallization. The yield of product is about 3%.
Synthesis of 8-Hydroxyguanosine by Udenfriend System The method of 8-oxo-G synthesis is similar to that described by Kasai and Nishimura ~ for 8-oxo-dG synthesis with some modifications. The reaction mixture consists of the following: 0.25 g guanosine in 390 ml of 0.1 M phosphate buffer (pH 6.8), 32 ml of 0.1 M EDTA, 8 ml of 0.1 M FeSO4, and 70 ml of 0.1 M ascorbic acid. The reaction is allowed to proceed at 37° with constant bubbling of oxygen for 3 hr in the dark. The removal of iron, ascorbate, and buffer is accomplished by adding about 100 ml of the reaction mixture onto a Sephadex G-15 gel filtration column, 38 x 200 mm. This is followed by addition of deionized water. The eluate is collected in 20-ml fractions. The first 300 ml of liquid is discarded. At this point, all the yellow bands should have passed out of the column. The remaining fractions, about 180 ml, are collected and pooled. The elution profile is monitored for products using HPLC/ED. The maximum load of the column is about 100 ml, so the remaining reaction mixture has to be run in 100-ml batches following the same procedure. The eluates containing 8-oxo-G collected from each run are pooled together and vacuum evaporated to a final volume of about 150 ml. Guanosine and other by-products are removed by the procedure described previously for the purification of 8-oxo-G formed by the action of methylene blue and light. Thus, 150 ml of the pooled eluate, which
62
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[6]
consists mainly of unreacted guanosine and 8-oxo-G, are loaded onto an AG 50W-X2 cation-exchange column, 38 × 120 mm, and allowed to run through at a rate of 1.5 ml/min. Deionized water is added to continue the elution. The eluate is collected in 20-ml fractions. Fractions rich in products (14 through 25) are pooled and then evaporated under vacuum to a volume of about 2 ml, followed by storage at about 4° for crystallization. The yield is about 4%.
Synthesis of 8-Hydroxy-2'-deoxyguanosine by Methylene Blue and Light The reaction mixture consists of 450 ml of 3.8 mM 2'-deoxyguanosine and 50 ml of 1.0 mM methylene blue in water. The above mixture is stored in a reservoir and slowly passed through a glass coil as described before. Experimental conditions are similar to those used for the 8-oxo-G synthesis. After exposure to light the reaction mixture, consisting of dG, 8-oxo-dG, methylene blue, and a small amount of by-products, is fractionated, collected, and stored in the dark at about 4° for further treatment as described before. About 500 ml of the reaction mixture is first passed through an AG 50W-X2 cation-exchange column, 38 × 50 mm, at a flow rate of 1.5 ml/min. Methylene blue and unreacted deoxyguanosine are strongly absorbed on the column, while the product, 8-oxo-dG, and small amount of by-products are eluted faster. After all the liquid containing the reaction product has passed through the column, about 250 ml of deionized water is added to wash out the remaining product. After discarding the first 180 ml of liquid, the rest of the eluate is collected and evaporated under vacuum to about 150 ml. This is then loaded onto a second AG 50W-X2 cation-exchange column, 38 × 120 mm, as described before, and allowed to run through at a rate of 1.5 ml/min. The first 420 ml is discarded, and fractions 22 through 36, a total of 280 ml rich in product, are pooled, evaporated under vacuum to about 2 ml, and stored at 4 ° for crystallization. The yield is about 1%.
Characterization of 8-Hydroxy-2'-deoxyguanosine and 8-Hydroxyguanosine The UV spectra of 8-oxo-dG and 8-oxo-G were obtained, and the spectral parameters agree with previously obtained v a l u e s , 6 a s listed below. Ultraviolet Spectra. Spectral parameters for 8-oxo-dG and 8-oxo-G are as follows: 8-oxo-dG, max (e) 248 nm (12,034), 295 nm (10,440), 206 nm (26,170); 8-oxo-G, max (e) 248 nm (12,880), 295 nm (10,940), 206 nm (25,640).
[6]
PHOTOCHEMICAL
SYNTHESIS OF 8-HYDROXYGUANINE
63
High-Performance Liquid Chromatography with Electrochemical Detection Figure 2 s h o w s the H P L C / E D traces w h e n 8 - o x o - G and 8 - o x o - d G p r e p a r e d by the K a s a i and N i s h i m u r a p r o c e d u r e o r the m e t h y l e n e blue plus light p r o c e d u r e w e r e injected and run. It is clear that the p r o d u c t s p r e p a r e d b y either p r o c e d u r e are identical. Figure 3 s h o w s the h y d r o d y n a m i c v o l t a m m o g r a m s o f 8 - o x o - G p r e p a r e d by the K a s a i and N i s h i m u r a ~ p r o c e d u r e as c o m p a r e d to the m e t h y l e n e blue plus light p r o c e d u r e . T h e plots are identical. T h e same is true f o r
T C
® t'O
B
O~ nr'~ ¢9 LU
1
A
I
I
I
I
0
10
20
30
(Min)
Fro. 2. HPLC chromatograms of 8-oxo-O (peak 1) and 8-oxo-dO (peak 2). (A) Ten microliters of a solution containing 10 mM 8-oxo-G and 10 mM 8-oxo-dG prepared by the method described by Kasai and Nishimura 1was analyzed. (B) Ten microliters of a solution consisting of 10 mM 8-oxo-G and 10 mM of 8-oxo-dG prepared by the action of methylene blue plus light was analyzed. In (C) 10 t~l each of the solutions used in (A) and (B), for 20 ~l total, was injected onto the HPLC/ED system. Chromatographic conditions: mobile phase; acetate/citrate buffer, pH 5.1, with 4% methanol; flow rate, 0.8 ml/min; column, Hibar Cl8 column, 10 ram, 4 x 250 ram; electrochemical detector, set at +0.750 V versus Ag/AgC1; UV detector, 0.05 AUFS, 254 nm.
64
OXIDATIVEDAMAGETODNA ANDDNA REPAIR
[6]
1.0
0.8
u) C o CL
0.6
n" C]
o
0.4
0.2 ~ 0
0.~
|
0'.6
i
!
0.8
|
110
E vs Ag/AgCI (volt)
FIG. 3. Hydrodynamic voltammograms of (e) 8-oxo-G prepared by the method described by Kasai and NishimuraI and ((3) 8-oxo-G prepared by the action of methylene blue plus light.
1.00.8ta
c
O O nO I.U
0.6
02/
S
0.4
0
|
o.,
0'.6
o18
11o
E vs Ag/AgCI (volt)
FIG. 4. Hydrodynamic voltammograms of(e) 8-oxo-dG prepared by the method described by Kasai and NishimuraI and (©) 8-oxo-dG prepared by the action of methylene blue plus light.
[6]
PHOTOCHEMICAL SYNTHESIS OF 8-HYDROXYGUANINE
65
the hydrodynamic voltammograms of 8-oxo-dG prepared by the Kasai and Nishimura I procedure as compared to the methylene blue plus light procedure (Fig. 4). Discussion The Udenfriend system for synthesis of 8-oxo-dG or 8-oxo-G requires that reaction components and buffer present in the incubation mixture be removed before the evaporation step. Incomplete removal of these components could result in the formation of large amount of an unknown white precipitate which interferes with the isolation of the product. In the Kasai and Nishimura I method, active charcoal was used to accomplish this goal. Because properties of charcoal vary, some strongly absorbing the product and thus causing problems in the desorption step, it was found that the use of the gel filtration material Sephadex G-15 can better achieve separation of product from the starting materials and buffer. By using the methylene blue system for 8-hydroxyguanine nucleoside synthesis, the isolation step is simplified because only a few components are present in the system, and the use of the AG 50W-X2 cation-exchange resin can then effectively separate the product from the mother compounds. Guanosine or deoxyguanosine were strongly retained on the AG 50W-X2 cation exchanger. The use of stronger acid may then be required to elute the compound. Alternatively, the resin can be regenerated by stirring in ascorbic acid solution. Ascorbate reduces methylene blue, a cation, to a colorless neutral leuco form of methylene blue which can then be washed from the resin. It was found that deoxyguanosine or guanosine dissolved in water reacts with methylene blue to produce the 8-OH derivative in the highest yield. Phosphate buffer ranging from pH 5.2 to 7.4 was tested, and all systems gave less product than the reaction mixture without any buffer present. If the reaction mixture was placed inside an open beaker and irradiated with light, the product generated was susceptible to further changes; therefore, the timing of irradiation is important. Thus, when a glass coil was used as described above, the product was stable and fewer by-products were formed.
66
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[7] C o p p e r - D N A
[7]
Adducts
By MARK J. BURKITT Introduction Redox active metal ions such as iron and copper are believed to play a central role in the formation of reactive oxygen species in biological systemsJ ,2 Although iron has received the most attention in this capacity (perhaps because of its greater cellular abundance), the role of copper may be particularly crucial because copper occurs in the mammalian cell nucleus where it may be involved in the condensation of DNA-histone fibers into higher order chromatin structures. 3-5 Consequently, to those concerned with the role of oxygen radicals in biomolecular damage, interest in copper-DNA adducts stems from the possibility that endogenous, DNA-associated copper may be able to promote oxidative damage to DNA. 4-7 As with iron, copper serves to convert superoxide (02"-3 and hydrogen peroxide, formed during the metabolism of oxygen 8,9 (particularly in the presence of redox-cycling xenobiotics l°'H), to the highly reactive hydroxyl radical (.OH), which is generally considered to be the oxidizing species l B. HaUiwell and J. M. C. Gutteridge, this series, Vol. 186, p. 1. 2 S. D. Aust, L. A. Morehouse, and C. Thomas, J. Free Radicals Biol. Med. 1, 3 (1985). 3 C. D. Lewis and U. K. Laemmli, Cell (Cambridge, Mass.) 29, 171 0982). 4 A. M. George, S. A. Sabovljev, L. E. Hart, W. A. Cramp, G. Harris, and S. H o r n s e y , Br. J. Cancer 55, Suppl. 8, 141 (1987). 5 W. A. Cramp, A. M. George, H. Khan, and M. B. Yatvin, in "Free Radicals, Metal Ions and Biopolymers" (P. C. Beaumont, D. J. Deeble, B. J. Parsons, and C. Rice-Evans, eds.), p. 127. Richelieu Press, London, 1989. 6 W. A. Cramp, A. M. George, J. C. Edwards, S. A. Sabovljev, G. Harris, L. E. Hart, H. Lambert, and M. B. Yatvin, in "Prostaglandin and Lipid Metabolism in Radiation Injury" (T. L. Walden, Jr. and H. N. Hughes, eds.), p. 59. Plenum, New York, 1987. 7 R. Stoewe and W. Prtitz, Free Radical Biol. Med. 3, 97 (1987). s A. Boveris and E. Cadenas, in "Superoxide Dismutase" (L. Oberley, ed.), Vol. 2, p. 15. CRC Press, Boca Raton, Florida, 1982. 9 N. Oshino, B. Chance, H. Sies, and T. Biicher, Arch. Biochem. Biophys. 154, 117 (1973). i0 R. P. Mason, Environ. Health Perspect. 87, 237 (1990). 11 G. V. Rumyantseva and L. M. Weiner, FEBS Lett. 234, 459 (1988).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All fights of reproductionin any form reserved.
[7]
COPPER-DNA ADDUCTS
67
responsible for the induction of biomolecular damage [reactions ( 1 ) - ( 3 ) ] . 1'2'7A2 The ability of DNA-bound copper ions to participate in O f + Cu 2 + ~ 02 + Cu + 202-7+ 2H + ~ H202 + 02 Cu ÷ + H202--~ Cu 2+ + .OH + O H -
(1) (2) (3)
the catalysis of hydroxyl radical formation has been demonstrated by numerous investigators in well-defined chemical s y s t e m s . 7,13-16 Evidence for the participation of endogenous copper in such reactions in intact cells is, however, less extensive) 7 In this chapter, methods are described for the investigation of DNA oxidation by bound copper ions in well-defined systems, and approaches are suggested for demonstrating the (possible) participation of endogenous copper ions in the mediation of damage to DNA in intact cells exposed to oxidants. The oxidative lesions induced in DNA following exposure to copper ions and reduced oxygen species, which include strand breaks and hydroxylated base products, are expected to be similar in nature (though not distribution) to those induced by oxygen radicals generated via other means (e.g., the Fe2+-H202 system). ~5Therefore, any of the conventional methods of determining oxidative damage to DNA may be applied to systems employing copper. These methods, which are described in detail elsewhere, include alkaline elution, DNA unwinding, and plasmid nicking techniques for the determination of strand breaks and gas chromatography-mass spectrometry (GC-MS) techniques for the determination of oxidized base products (e.g., 8-hydroxyguanine). ~3-~5'18-2~In this chapter, full experimental details are given only for techniques that are particularly applicable to copper systems.
12 M. J. Burkitt and B. C. Gilbert, Free Radical Res. Commun. 14, 107 (1991). 13 J.-L. Sagripanti and K. H. Kraemer, J. Biol. Chem. 264, 1729 (1989). 14 C. J. Reed and K. T. Douglas, Biochem. J. 275, 601 (1991). 15 O. I. Aruoma, B. Halliwell, E. Gajewski, and M. Dizdaroglu, Biochem. J. 273, 601 (1991). 16 L. Milne, P. Nicotera, S. Orrenius, and M. J. Burkitt, Arch. Biochem. Biophys. 304, 102 (1993). ~7 H. C. Birnboim, Arch. Biochem. Biophys. 294, 17 (1992). 18 G. Ahnstr6m and K. Erixon, in " D N A Repair: A Laboratory Manual of Research Procedures" (E. C. Friedberg and P. C. Hanawalt, eds.), Vol. 1, Part B, p. 403. Dekker, New York, 1981. ~9C. von Sonntag and H.-P. Schuchmann, this series, Vol. 186, p. 511. 2o M. K. Shigenaga, J.-W. Park, K. C. Cundy, C. J. Gimeno, and B. N. Ames, this series, Vol. 186, p. 521. 21 M. Dizdaroglu and E. Gajewski, this series, Vol. 186, p. 530.
68
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[7]
Physical Properties of Copper-DNA Complexes
Copper(II) In 1968, Eichhorn and Shin 22 described the effects of various divalent metal ions, including those of the first transition series, on DNA melting and annealing. Metal ions believed to bind to DNA primarily at the phosphate moiety [e.g., Mg(II)] were found to stabilize DNA, whereas those which bind primarily at the base residues [such as Cd(II) and Cu(II)] were found to destabilize the double helix. Copper(II) was shown to bind to DNA with a higher affinity than any of the other divalent cations studied, which is believed to be one reason why copper is particularly effective in promoting the oxidation of DNA: oxidation by H 2 0 2 with copper as a catalyst has been reported to be 50 times faster than that with iron. 7 Kagawa et al.23 have resolved the crystal structure (to 1.2 A resolution) of CuCIE-soaked duplex DNA and suggested that Cu(II) forms a regular octahedral complex, with four water ligands in the equatorial plane and a fifth water along with the N-7 of a guanine residue in the axial positions. This appears to be the prototypical structure, of which all other complexes with guanine are a variation. 23 Although studies employing Cu(II)-GMP complexes have indicated that the Y-phosphate oxygen atoms from neighboring GMP molecules can also act as equatorial ligands, 24 this has not been observed in duplex DNA. 23 Regular octahedral binding is not observed at adenine residues, but an axially distorted trigonal bipyrimidal complex may occur when both adenine and guanine residues are able to interact and share a single copper(II) center. 25
Copper(l) Whereas Cu(II) binding causes destabilization of the double helix, 22 Cu(I) binding causes stabilization, 26 and it may involve the conversion of certain DNA regions from the B conformation to the Z conformation, z7 Indeed, Minchenkova and Ivanov suggested that the oxidation state of copper may play a role in the control of DNA synthesis by determining 22 G. L. Eichhorn and Y. Ae Shin, J. Am. Chem. Soc. 90, 7323 (1968). 23 Z. F. Kagawa, B. H. Geierstanger, A. H.-J. Wang, and P. Shing Ho, J. Biol. Chem. 266, 20175 (1991). 24 K. Aoki, G. R. Clark, and J. D. Orbell, Biochim. Biophys. Acta 425, 369 (1976). 25 B. H. Geierstanger, T. F. Kagawa, S.-L. Chen, G. J. Quigley, and P. Shing Ho, J. Biol. Chem. 266, 20185 (1991). 26 S. J. Atherton, in " F r e e Radicals, Metal Ions and Biopolymers" (P. C. Beaumont, D. J. Deeble, B. J. Parsons, and C. Rice-Evans, eds.), p. 93. Richelieu Press, London, 1989. 27 W. A. Prfitz, J. Butler, and E. J. Land, Int. J. Radiat. Biol. 58, 215 (1990).
[7]
COPPER-DNA ADDUCTS
69
the conformation of the double helix. 28 Compared with Cu(II), there have been considerably fewer studies on the binding of copper(I) to DNA. This is presumably a reflection of the instability of the copper(I) oxidation state. Copper(I) ions in aqueous solution undergo rapid disproportionation, z9 Although stable stock solutions of copper(l) can be prepared using nonprotic solvents, such as acetonitrile, 3° the interaction of Cu(I) with DNA has been studied most successfully following its generation in situ via the addition of a reducing agent (e.g., ascorbate) to DNA in the presence of Cu(II). 7'28 Copper(I) may also be formed via the radiolytic reduction of Cu(II). 27 This involves the y-irradiation (0.33 Gy sec -1) of air-saturated formate solutions. The radiolysis products from water are rapidly converted to the reducing superoxide radical [reactions (4)-(8)]. Binding of the Cu(I) to DNA competes with its loss via disproportionation. H20 _1~ -OH(0.28), eaq(0.28), H'(0.06) •OH + HCO2- ~ H20 + CO27 C O 2 "7"q- 0 2 ~
CO 2 + 027
e~-q(H-) + 02--~ (H ÷) + O Z 027 + Cu 2+ ~ 02 + Cu +
(4) (5) (6) (7) (8)
Using this method of Cu(I) generation, Prtitz et al. 27 have obtained UV absorbance difference spectra of the DNA-Cu(I) complex. Spectra are characterized by a minimun at 250-255 nm, a maximum at 295 nm, and an isobestic point around 270 nm. These features are believed to be due primarily to the fixation of Cu(I) to the guanine and cytosine bases, accompanied by enolization, proton transfer from the N-l-guanine to the N-3-cytosine, and conformational changes, z7'28 Similar spectra have been obtained using ascorbate as the reductant of Cu(II). 28 Detection of Copper-Dependent Hydroxyl Radical Formation Copper-dependent hydroxyl radical formation has been demonstrated in many model systems, generally via the application of standard methods of .OH determination. For example, van Steveninck et al. 31 have used salicylate hydroxylation to detect the hydroxyl radical in systems containing HzO2, Cu(II), and a reducing agent. However, the most direct method of hydroxyl radical detection is electron spin resonance (ESR) spectros2s L. E. Minchenkova and V. I. Ivanov, Biopolymers 5, 615 (1967). 29 F. A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry," 5th Ed. Wiley (Interscience), New York, 1988. 3o p. M. Hanna and R. P. Mason, Arch. Biochem. Biophys. 295, 205 (1992). 31 j. van Steveninck, J. van der Zee, and T. M. A. R. Dubbelman, Biochem. J. 232, 309 (1985).
70
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[7]
copy in conjunction with a spin trapping reagent. 32'33 ESR spin-trapping evidence for .OH formation via reaction (3) has been obtained by Hanna and Mason, 3° who used Cu(I) prepared in deoxygenated acetonitrile. The •OH radical was detected as its adduct to the spin trap 5,5-dimethyll-pyrroline N-oxide (DMPO). The DMPO/.OH adduct can also arise, however, via a nonradical mechanism involving the Cu(II)-stimulated nucleophilic addition of water to DMPO. 3°'34 Although this can lead to problems in the detection of copper-dependent. OH formation, the contribution of the nucleophilic route to DMPO/.OH formation can be minimized by the careful use of copper chelating agents (e.g., by using a citrate-based buffer. 34,35 A further problem is that the DMPO/.OH adduct is prone to reduction to an ESR-silent hydroxylamine in the presence of Cu(I)) 5 Since the radical adducts resulting from the trapping of carbon-centered radicals are generally more stable than those derived from oxygen-centered radicals, 36 a more reliable indication of .OH formation is obtained if the radical is first reacted with a suitable scavenger molecule to form a carbon-centered radical before trapping (see below and Refs. 16, 35, and 36). Electron Spin Resonance Detection of Copper-DNA Adduct-Dependent Hydroxyl Radical Formation The pattern of hydroxylated base products detected following the incubation of DNA with copper(II), H202, and a reducing agent indicates that DNA-bound copper ions can support hydroxyl radical formation.15 Indeed, Yamamoto and Kawanishi 37 have described possible mechanisms by which DNA-bound copper ions may interact with hydrogen peroxide to form oxidizing species. These workers suggested that a peroxide bridge is formed between two copper centers prior to electron transfer. In support of this hypothesis, Kagawa and co-workers have described models which indicate that DNA can accommodate peroxide-bridged copper(II) centers at adjacent guanine residues. 23 ESR spin-trapping techniques have been applied to the detection of hydroxyl radical formation following the interaction of DNA-bound copper(II) ions with ascorbate and H202 [reactions (9) and (10), in which 32 G. R. Buettner and R. P. Mason, this series, Vol. 186, p. 127. 33 R. P. Mason, P. M. Hanna, M. J. Burkitt, and M. B. Kadiiska, Environ. Health Perspect. (in press). 34 p. M. Hanna, W. Chamulitrat, and R. P. Mason, Arch. Biochem. Biophys. 296, 640 (1992). 35 M. J. Burkitt, L. Milne, S. Y. Tsang, and S. C. Tam, Arch. Biochem. Biophys., (in press). 36 M. J. Burkitt, Free Radical Res. Commun. 18, 43 (1993). 37 K. Yamamoto and S. Kawanishi, J. Biol. Chem. 264, 15435 (1989).
[7]
COPPER-DNA ADDUCTS
71
Asc 2- and Asc: represent the dianion of ascorbic acid and the ascorbyl radical, respectively]. 16 Because oxygen-radical adducts of DMPO (the D N A - C u 2+ + ASC 2- ~ D N A - C u + + Asc-:D N A - C u + + H202----~ D N A - C u z+ + OH- + .OH
(9) (10)
spin trap used most commonly to detect .OH) can undergo conversion to ESR-silent hydroxylamines in the presence of reductants (e.g., superoxide), 32 and because of the instability of the DMPO/-OH adduct in the presence of metal ions, 36 a secondary trapping technique has been employed in which the .OH radical is first converted to the methyl radical ('CH3) via scavenging with dimethyl sulfoxide (DMSO). The methyl radical is then trapped and detected as its relatively stable adduct to the spin trap N-tert-butyl-et-phenylnitrone (PB N), PBN/. CH 3 [reactions (11) and (12)].38 (CH3)280 + "OH---> CHaSO2H + "CH3 PBN + -CH 3~ PBN/'CH 3
(11) (12)
Method The standard 2-ml reaction mixture contains 25 mM potassium phosphate buffer (pH 7), 2 M DMSO, 100 mM PBN, 2 mg/ml DNA, 1 mM CuCI2, 1 mM H202, and 1 mM ascorbate. The phosphate buffer is added from a 100 mM stock solution treated with chelating resin (from Sigma, St. Louis, MO) to remove contaminating metal i o n s . 39 The DNA (sodium salt, from Sigma, St. Louis, MO) is added from a concentrated stock solution prepared daily (8 mg/ml, in water). The CuCI: and H202 (prepared daily) are added from concentrated stock solutions prepared in water (100 and 200 mM, respectively), and ascorbic acid (added last) is taken from a 25 mM stock solution prepared daily in the 100 mM phosphate buffer and adjusted to pH 7. Following the initiation of reactions with ascorbate, mixtures are transferred to a quartz fiat cell, either manually or using a rapid sampling device. 4° Spectra are then recorded using a Bruker E 109 time constant, 41 ms; spectrometer (Bruker Spectrospin Ltd., Coventry, England) with the following instrument settings: sweep width, 60 G (1 G = 10 -4 T ) , modulation frequency, 100 kHz; modulation amplitude, 1 G; sweep time, 84 sec; power, 20 mW. Hydroxyl radical formation is indicated by the observation of the six-line signal (a N = 16.3 G, a~ = 3.6 G) from the 38 M. J. Burkitt and R. P. Mason, Proc. Natl. Acad. Sci. U.S.A. 88, 8440 (1991). 39 G. R. Buettner, J. Biochem. Biophys. Methods 16, 27 (1988). 4o R. P. Mason, this series, Vol. 105, p. 416.
72
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[7]
PBN/'CH3 adduct (Fig. IA). The weaker six-line signal (a TM= 14.8 G, aft = 3.5 G), also present in the spectrum in Fig. 1A, is believed to arise from the PBN-methoxyl adduct, PBN/.OCH3, formed following the reaction of methyl radicals with oxygen prior to trapping with PBN. 38 The dependence of hydroxyl radical formation on copper, hydrogen peroxide, and ascorbate is demonstrated in Fig. 1B-D, respectively. The weak doublet signal in Fig. 1C is from the ascorbyl radical, the formation of
IO.OG X/ Cu 2+ DNA
/~
sscoF'os|e
minus Cu 2+
B
minus H202
C
minus ascorbate
D ,....--,..
FIG. 1. Demonstration of copper-DNA adduct-dependent hydroxyl radical formation. Reaction mixtures contained 2 mg/ml DNA, 1 mM CuC12, 1 mM H202, 0.1 M PBN, 2 M DMSO, and 1 mM ascorbate in 25 mM KHPO4 buffer, pH 7. (A) Complete system. (B) As in (A), but without CuC12. (C) As in (A), but without H20:. (D) As in (A), but without ascorbate. Hydroxyl radical formation is indicated by the detection of the six-line ESR signal from the PBN-methyl radical adduct.
[7]
COPPER-DNA ADDUCTS
73
which is not H202-dependent [see reaction (9)]. Hydroxyl radical formation can also be demonstrated to ocur when DNA is omitted from the above reaction mixture. Indeed, the signal from the PBN/.CH3 adduct detected in the absence of DNA is more intense than that detected in the presence of DNA (data not shown). This may reflect differences in the reactivities of the free and bound metal ions: the rate constant for the reaction of the aquacopper(I) ion with H20 2 is 4 x 103 M -1 sec -1, whereas the corresponding value for the slower reaction of the DNA-bound Cu(I) ion is less than 1.3 M -1 sec-l. ~6 The intensity of the PBN/.CH 3 signal from reactions containing DNA increases with time, whereas that from incubations not containing DNA is stable (not shown). The steady increase in the production of hydroxyl radicals in the presence of DNA may reflect the redox cycling of copper ions following the generation of reducing radicals on the DNA. 41 The spin-trapping technique described above can be used to examine the effects of a variety of agents on hydroxyl radical formation catalyzed by DNA-bound copper ions. For example, glutathione, which occurs in cell nuclei at a relatively high concentration (-19.2 mM),4z has been shown to suppress -OH formation when included in reaction mixtures containing DNA, copper(II), hydrogen peroxide, and ascorbateJ 6 Glutathione is believed to suppress -OH formation by stabilizing copper in the + 1 oxidation state, thereby preventing its participation in reaction (3)J 6 Measurement of Copper-Dependent DNA Oxidation Many of the standard, well-documented techniques for measuring DNA oxidation have been applied to copper systems. 13-15,17In the author's laboratory, use has been made of the ethidium-binding assay,16 developed initially by Prtitz for studies on radiation damage to DNA. 43 The assay, which has been applied subsequently to copper systems,7 provides a quantitative determination of copper-dependent oxidative damage to DNA under well-defined reaction conditions. The assay is based on the fact that a highly fluorescent complex is formed between native DNA and the intercalating agent ethidium bromide. When DNA is modified following exposure to hydroxyl radicals, intercalation by ethidium bromide is disrupted and the fluorescence from the ethidium bromide-DNA complex 41M. J. Burkitt, M. Fitchett, and B. C. Gilbert, in "Medical, Biochemicaland Chemical Aspects of Free Radicals" (O. Hayaishi,E. Niki, M. Kondo, and T. Yoshikawa,eds.), p. 63. Elsevier, Amsterdam, 1989. 42G. Bellomo, M. Vairetti, L. Stivala, F. Mirabelli, P. Richelmi, and S. Orrenius, Proc. Natl. Acad. Sci. U.S.A. 89, 4412 (1992). 43W. A. Prfitz, Radiat. Environ. Biophys. 23, 1 (1984).
74
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[7]
is compromised. Several forms of DNA lesions, including strand scission, base oxidation, and base liberation, are believed to contribute to the loss of fluorescence. Hence, the assay is not specific for any single lesion. The basic assay, essentially as used by Stoewe and Priitz, 7 is described below, along with modifications which permit its application to the investigation of the effects of chelating agents on damage. 16 Method The 2-ml standard reaction mixture contains (in order of addition): 20 mM phosphate buffer (chelating resin-treated, pH 7), 100 /zg/ml DNA (sodium salt, Sigma), 50/zM CuClz, 2 mM H202, and 2 mM ascorbate. DNA is added from a concentrated stock solution prepared freshly (400 /~g/ml in water). The CuCI2 and H202 (prepared daily) are added from concentrated aqueous stock solutions (10 and 80 mM, respectively). Reactions are initiated via the addition ofascorbate from a 40 mM stock solution prepared daily in phosphate buffer and adjusted to pH 7. Reactions are carried out at 25° in open tubes. In the original method described by Stoewe and Prtitz, 7 reactions are terminated via the addition of 10 mM EDTA, to form a redox-inactive complex with the metal ion. Although EDTA can be used successfully to terminate reactions in the simple incubation described above, the reagent may not be able to stop reactions in more complex incubations containing other copper chelating agents. For example, in the presence of 1,10phenanthroline (OP), which promotes the degradation of DNA by copper (involving the formation of a reactive ternary complex), 16 EDTA would not be expected to remove and inactivate the metal ion. To broaden the application of the ethidium bromide binding assay to a greater variety of reaction conditions, catalase can be used to terminate reactions~6: following the addition of 400 units catalase (in 5/zl water), 50 /zM ethidium bromide is added to tubes and the fluorescence intensity measured with excitation at 510 nm and emission at 590 nm. Because the enhancement in the fluorescence of ethidium bromide measured following intercalation with DNA is a measure of the integrity of the DNA, the 100% value on the instrument is set using a solution containing the same reagents as the reaction under study with the exception of H202 , and catalase is added before the CuCI 2 and ascorbate. The zero reading on the machine is set using a solution prepared the same as the 100% reference solution except that DNA is omitted. Using these reference solutions, 100% fluorescence intensity refers to a 100% enhancement in the fluorescence ofethidium bromide following intercalation with (undamaged) DNA. Damage is then indicated as a loss in the enhancement of fluorescence.
[7]
cOPPER-DNA ADDUCTS
75
Figure 2 shows a time course of copper-dependent DNA oxidation in which reactions were terminated using either catalase or EDTA. Clearly, catalase is as effective as EDTA in the termination of DNA oxidation. The findings from ESR spin-trapping measurements of .OH formation carried out as described above indicate, however, that although both EDTA and catalase prevent the formation of .OH in the absence of 1,10phenanthroline, only catalase is able to inhibit .OH formation in the presence of the copper chelator (data not shown). A time course of DNA oxidation by copper in the presence of OP, terminated using catalase, is shown in Fig. 3; at the low concentration of copper used (5 ~M), no oxidation is detected in the absence of OP. Intact Cells: Catalysis by Endogenous Copper The oxidation of DNA following incubation with exogenous copper has been demonstrated in a variety of in vitro model systems. ~3-16 However, considerably fewer studies have addressed the possibility that the endogenous, DNA-bound metal ion may serve to promote the oxidation of DNA under conditions of oxidative stress. DNA single-strand breaks are often
100
...
80 60 40
0
20
0
i
i
i
i
5
10
15
20
i
25
i
30
Time ( m i n ) FIG. 2. Time course of copper-dependent DNA oxidation. DNA (100/xg/ml) was incubated with 50 ~M CuCI~, 2 mM H202, and 2 mM ascorbate in 20 mM KHPO4 buffer, pH 7. Reactions were terminated at the times indicated via the addition of either 10 mM EDTA (O) or 400 units catalase (©). Ethidium bromide (50 ~M) was then added and damage indicated as the failure of the DNA to cause a 100% enhancement in the fluorescence of the dye (compared with undamaged DNA). Values represent means with standard deviation values no greater than 2.6 (n = 2). (From Milne e t al., 16 with permission.)
76
OXIDATIVEDAMAGETO DNA AND DNA REPAIR
[7]
100- ,~:
80
60'
40-
20
. 0
5
.
. 10
. 15
20
J
i
25
30
Time (min)
Fie. 3. Effect of 1,10-phenanthroline (OP) on copper-dependent DNA oxidation. DNA (100/xg/ml) was incubated with 5 p~MCuCIz, 200 p.M H202, and 200 p.M ascorbate, in either the presence (O) or absence (0) of 15 ~M OP, in 20 mM KHPO4 buffer, pH 7. Reactions were terminated at the times indicated via the addition of 400 units catalase. Ethidium bromide (50 p.M) was then added and damage indicated as the failure of the DNA to cause a 100% enhancement in the fluorescence of the dye (compared with undamaged DNA). Values represent means with S.D. values no greater than 0.9 (n = 3). (From Milne e t a l . , 16 with permission.)
detected following the exposure of cells to either hydrogen peroxide or c o m p o u n d s that promote the cellular formation of superoxide and hydrogen peroxide.44 Although a significant proportion of the D N A strand breaks detected in cells following exposure to reactive oxygen species may be brought about by the action o f calcium-dependent nucleases, 44,45 the fact that oxidized purine and pyrimidine bases have been detected in cells following such treatment indicates that direct radical damage to D N A must also occur.44 Lipophilic metal ion-chelating agents have been used to probe for the participation o f metals of the induction of strand breaks in cells exposed to reduced o x y g e n species, such as superoxide and hydrogen peroxide. A key issue under investigation by several groups is the elucidation o f the identity o f the metal ion(s) responsible for the promotion of hydroxyl radical formation, and hence biomolecular damage, in cells exposed to 44B. Halliwell and O. I. Aruoma, F E B S Lett. 281, 9 (1991). 45S. Orrenius, M. J. Burkitt, G. E. N. Kass, J. M. Dypbukt, and P. Nicotera, Ann. Neurol. 32, $33 (1992).
[7]
COPPER-DNA ADDUCTS
77
o x i d a n t s . 17,35,46-49 Although
low molecular weight iron chelates may support •OH formation, the possibility that DNA-associated copper may promote formation in the nucleus is particularly interesting and highly relevant to the mechanism(s) of DNA damage.
Use of 1,10-Phenanthroline and Neocuproine as Probes for the Involvement of Copper and Iron in DNA Oxidation Although OP and 2,9-dimethyl- 1,10-phenanthroline (neocuproine, NC) will each chelate both copper and iron, their effects on the redox chemistry of the two metal ions are distinct: Cu(I) chelated to OP reacts readily with HzOz to form .OH [see reaction (3)]: 0 whereas reaction of the Fe(II) complex is inhibited46; NC inhibits the reaction of Cu(I) with HzO25~ but appears to have little effect o n F e ( I I ) . 46 Therefore, by determining the effects of OP and NC on DNA oxidation in cells exposed to oxidants, it is possible, in principle, to determine whether iron or copper is the catalytic metal responsible for -OH formation and damage. 35 From the data shown in Fig. 4, 5z demonstrating that H202-induced DNA fragmentation in HeLa cells is inhibited by OP and not NC, it appears that iron rather than copper is responsible for the catalysis of .OH formation: had copper been responsible, NC would have been expected to inhibit fragmentation and OP would have had, if anything, a stimulatory effect (see Fig. 3). OP does not remove copper from DNA, but it participates in the formation of a reactive ternary complex with both copper and DNA. 53-55 It is considered unlikely, therefore, that the protection afforded by OP could be due to its removal of copper, and hence -OH formation, to a site distant to the target molecule. Lipophilic chelating agents can be used as probes for the participation of copper and iron in DNA oxidation in experiments employing most of the standard techniques for the determination of end points of damage. 46 A. C. Mello-Fihlo and R. Meneghini, Mutat. Res. 251, 109 (1991). 47 A. C. Mello-Filho and R. Meneghini, Biochim. Biophys. Acta 781, 56 (1984). 48 I. Schraufstatter, P. A. Hyslop, J. H. Jackson, and C. G. Cochrane, J. Clin. Invest. 82, 1040 (1988). 49 O. Cantoni, P. Sestili, F. Cattabeni, G. Bellomo, S. Pou, M. Cohen, and P. Cerutti, Eur. J. Biochem. 182, 209 (1989). 50 S. Goldstein and G. Czapski, J. Am. Chem. Soc. 108, 2244 (1986). 51 G. Czapski and S. Goldstein, Free Radical Res. Commun. 1, 157 (1986). 52 C. M. Gedik and A. R. Collins, Nucleic Acids Res. 18, 1007 (1990). 53 S. Goldstein and G. Czapski, J. Free Radicals Biol. Med. 2, 3 (1986). 54 F. Liu, K. A. Meadows, and D. R. McMillin, J. Am. Chem. Soc. 115, 6699 (1993). 55 D. S. Sigman, T. W. Bruice, A. Mazumder, and C. L. Sutton, Acc. Chem. Res. 26, 98 (1993).
78
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[7]
Control
h
100 .aM H202
100 gM H202 + 1 mM OP
100 p.M H202 + 1 mM NC i
0
i
20
40
h
i
60
% ss breaks
FIG. 4. Effects of 1,10-phenanthroline (OP) and neocuproine (NC) on the level of DNA single-strand breaks detected in HeLa cells following incubation with H20 2. Cells were grown and DNA labeled with [3H]thymidine according to Gedik and Collins.52After a wash in phosphate-buffered saline (PBS), and following preincubation (4°) for 5 min with either 1 mM OP or 1 mM NC (each added from 15 mM aqueous stock solutions), cells were incubated with 100 p.M H20 2, and additions as indicated, for 15 min (4°). Cells not exposed to OP or NC were also preincubated for 5 min. No additions were made to the control incubations. Strand breaks were then determined by alkaline denaturation and hydroxyapatite chromatography/8 Values represent means -+1 S.D. (n = 3). (Data from Burkett et al., 35 with permission). H o w e v e r , w i t h t h e e x c e p t i o n o f t h e findings f r o m s o m e r a d i a t i o n s t u d ies,4-6,56 t h e r e is little e v i d e n c e f o r t h e p a r t i c i p a t i o n o f e n d o g e n o u s c o p p e r in D N A o x i d a t i o n d u r i n g o x i d a t i v e s t r e s s . F o r e x a m p l e , B i r n b o i m 17 h a s s h o w n t h a t O P e n h a n c e s D N A s t r a n d b r e a k a g e in p h o r b o l e s t e r - t r e a t e d l e u k o c y t e s . It r e m a i n s to d e m o n s t r a t e w h e t h e r t h e c e l l u l a r p o o l o f c o p p e r r e s p o n s i b l e f o r this is i n d e e d t h a t w h i c h e x i s t s in t h e n u c l e u s b o u n d to D N A . Concluding Remarks It h a s b e c o m e i n c r e a s i n g l y a p p a r e n t t h a t c o p p e r o c c u r s n a t u r a l l y in c h r o m o s o m e s , w h e r e it is b e l i e v e d to p l a y a r o l e in t h e a t t a c h m e n t o f D N A to s c a f f o r d p r o t e i n s v i a t h e f o r m a t i o n o f m e t a l l o p r o t e i n b r i d g e s ) I n 56 S.-M. Chiu, L.-Y. Xue, L. R. Friedman, and N. L. Oleinick, Biochemistry 32, 6214 (1993).
[8]
HPLC AND MS ANALYSISOF DNA DAMAGEPRODUCTS
79
view of the increasing awareness of the role played by redox active metal ions and reactive oxygen species in the inducement of biomolecular damage and disease, it seems particularly remarkable that the interconversion of bound Cu(I) and Cu(II) ions should play a role in the regulation of the conformation of such a critical cellular target molecule as D N A . 4-6 It is anticipated that this interesting dilemma will continue to stimulate research from which an appreciation of the true significance of D N A oxidation by the bound metal ion will emerge. Acknowledgments The author thanks Mrs. L. Milne for technical assistance and Scottish Office Agriculture and Fisheries Department for support.
[8] S i n g l e t O x y g e n D N A D a m a g e : C h r o m a t o g r a p h i c and Mass Spectrometric Analysis of Damage Products B y JEAN CADET, JEAN-LUC RAVANAT, GARRY W. BUCHKO,
HELEN C. YEO, and BRUCE N. AMES Introduction Singlet oxygen (~O2), the lowest excited state of molecular oxygen (lAg, 94.2 kJ/mol), m a y be generated through energy transfer involving a type II photosensitized reaction L2 and by thermal decomposition of endoperoxides 3 and dioxetanes.4 In addition, the formation of 10 2 in biological systems m a y be mediated by several enzymatic reactions 5 and by chemiexcitation during lipid peroxidation. 6 Like the highly reactive hydroxyl radical (-OH), but in a more specific manner, IO 2 is capable of inducing genotoxic, carcinogenic, and mutagenic effects 7-9 and is likely involved in aging. ~° 1c. s. Foote, in "Free Radicals in Biology" (W. A. Pryor, ed.), Vol. 2, p. 85. Academic Press, New York, 1976. 2 E. Sage, T. Le Doan, V. Boyer, D. E. Helland, L. Kittler, C. Hfl~ne, and E. Moustacchi, J. Mol. Biol. 209, 297 (1989). 3 p. di Mascio and H. Sies, J. Am. Chem. Soc. 111, 2909 (1989). 4 W. Adam and C. Cilento, Angew. Chem., Int. Ed. Engl. 22, 529 (1983). 5 E. Cadenas and H. Sies, this series, Vol. 105, p. 221. 6 E. Cadenas and H. Sies, Eur. J. Biochem. 124, 349 (1982). 7 j. Piette, J. Photochem. Photobiol., B 4, 335 (1990). 8 R. A. Floyd, Carcinogenesis (London) 11, 1447 (1990). 9 B. Epe, Chem.-Biol. Interact. 83, 239 (1991). 10B. N. Ames, Science 221, 1258 (1983).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[8]
HPLC AND MS ANALYSISOF DNA DAMAGEPRODUCTS
79
view of the increasing awareness of the role played by redox active metal ions and reactive oxygen species in the inducement of biomolecular damage and disease, it seems particularly remarkable that the interconversion of bound Cu(I) and Cu(II) ions should play a role in the regulation of the conformation of such a critical cellular target molecule as D N A . 4-6 It is anticipated that this interesting dilemma will continue to stimulate research from which an appreciation of the true significance of D N A oxidation by the bound metal ion will emerge. Acknowledgments The author thanks Mrs. L. Milne for technical assistance and Scottish Office Agriculture and Fisheries Department for support.
[8] S i n g l e t O x y g e n D N A D a m a g e : C h r o m a t o g r a p h i c and Mass Spectrometric Analysis of Damage Products B y JEAN CADET, JEAN-LUC RAVANAT, GARRY W. BUCHKO,
HELEN C. YEO, and BRUCE N. AMES Introduction Singlet oxygen (~O2), the lowest excited state of molecular oxygen (lAg, 94.2 kJ/mol), m a y be generated through energy transfer involving a type II photosensitized reaction L2 and by thermal decomposition of endoperoxides 3 and dioxetanes.4 In addition, the formation of 10 2 in biological systems m a y be mediated by several enzymatic reactions 5 and by chemiexcitation during lipid peroxidation. 6 Like the highly reactive hydroxyl radical (-OH), but in a more specific manner, IO 2 is capable of inducing genotoxic, carcinogenic, and mutagenic effects 7-9 and is likely involved in aging. ~° 1c. s. Foote, in "Free Radicals in Biology" (W. A. Pryor, ed.), Vol. 2, p. 85. Academic Press, New York, 1976. 2 E. Sage, T. Le Doan, V. Boyer, D. E. Helland, L. Kittler, C. Hfl~ne, and E. Moustacchi, J. Mol. Biol. 209, 297 (1989). 3 p. di Mascio and H. Sies, J. Am. Chem. Soc. 111, 2909 (1989). 4 W. Adam and C. Cilento, Angew. Chem., Int. Ed. Engl. 22, 529 (1983). 5 E. Cadenas and H. Sies, this series, Vol. 105, p. 221. 6 E. Cadenas and H. Sies, Eur. J. Biochem. 124, 349 (1982). 7 j. Piette, J. Photochem. Photobiol., B 4, 335 (1990). 8 R. A. Floyd, Carcinogenesis (London) 11, 1447 (1990). 9 B. Epe, Chem.-Biol. Interact. 83, 239 (1991). 10B. N. Ames, Science 221, 1258 (1983).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
80
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[8]
One major biological target of 10 2 is DNA. It has been shown that D N A synthesis blocking lesions induced by a chemical source of singlet oxygen are targeted to guanine residues in single-stranded DNA. 11 This may be accounted for by the specific oxidation of the guanine moiety of DNA, which is the most reactive nucleobase toward electrophilic addition of IOz .12 One of the main resulting oxidation products, at least in DNA, has been identified as 7,8-dihydro-8-oxo-2'-deoxyguanine (8-oxo-dG). 13-15 Various studies involving site-specific incorporation of 8-oxo-dG into oligonucleotides have shown that this oxidized nucleoside is potentially mutagenic 16and is a substrate for the formamidopyrimidine glycosylase (FPG) D N A repair protein. 17It should be noted that a sensitive high-performance liquid chromatography (HPLC)-electrochemical detection assay 18 and postlabeling techniques 19 are now available to monitor the formation of 8-oxo-dG in both isolated and cellular DNA. In addition, a monoclonal antibody column has been developed for the prepurification of 8-oxo-dG in urine. 2° However, the formation of 8-oxo-dG cannot be used as a probe for 10 z reactions in cellular D N A since this modified nucleoside is also produced through the reaction of .OH (and related reactive species of the Fenton reaction) with the guanine moiety. 21 In addition, it has been observed that the hydration reaction of the guanine radical cation within D N A leads to the predominant formation of 8 - o x o - d G . z2 On the other hand, the 4R* and 4S* diastereoisomers of 4,8-dihydro-4-hydroxy-8-oxoH D. T. Ribeiro, F. Bourre, A. Sarasin, P. Di Mascio, and C. F. M. Menck, Nucleic Acids Res. 211, 2465 (1992). 12j. Cadet and P. Vigny, in "Bioorganic Photochemistry" (H. Morrison, ed.), Vol. 1, p. 1. Wiley, New York, 1990. ~3R. A. Floyd, M. S. West, K. L. Eneff, and J. E. Schneider, Free Radical Biol. Med. 8, 327 (1990). 14 j. E. Schneider, S. Price, L. Maidt, J. M. C. Gutteridge, and R. A. Floyd, Nucleic Acids Res. 18, 631 (1990). 15 T. P. A. Devasagayam, S. Steenken, M. S. W. Obendorf, W. A. Schultz, and H. Sies, Biochemistry 311, 6283 (1991). 16 S. Shibutani, M. Takeshida, and A. P. Grollman, Nature (London) 349, 431 (1991). 17 j. Tchou, H. Kasai, S. Shibutani, M.-H. Chung, J. Laval, A. P. Grollman, and S. Nishimura, Proc. Natl. Acad. Sci. U.S.A. 88, 4690 (1991). ~8 R. A. Floyd, J. J. Watson, P. T. Wong, D. H. Altmiller, and R. C. Rickard, Free Radical Res. Commun. 1, 163 (1986). 19 j. Cadet, F. Odin, J.-F. Mouret, M. Polverelli, A. Audic, P. Giacomoni, A. Favier, and M.-J. Richard, Mutat. Res. 275, 343 (1992). 20 E.-M. Park, M. K. Shigenaga, P. Degan, T. S. Korn, J. W. Kitzler, C. M. Wehr, P. Kolachana, and B. N. Ames, Proc. Natl. Acad. Sci. U.S.A. 89, 3375 (1992). 2~ K. Kasai and S. Nishimura, in "Oxidative Stress, Oxidants and Antioxidants" (H. Sies, ed.), p. 99. Academic Press, San Diego, 1991. 22 H. Kasai, Z. Yamaizumi, M. Berger, and J. Cadet, J. Am. Chem. Soc. 114, 9692 (1992).
[8]
HPLC AND MS ANALYSISOF DNA DAMAGEPRODUCTS
81
2'-deoxyguanosine (4-hydroxy-8-oxo-dG), 23-25 whose formation arises from initial [2 ÷ 4] Diels-Alder cycloaddition of 10 2 across the 4,8-purine carbons, are specific for 102-mediated DNA oxidation. 12Emphasis in this chapter is placed on the chromatographic behavior and mass spectroscopic (MS) features of the diastereoisomers of 4-hydroxy-8-oxo-dG and the corresponding base. Materials and Methods 2'-Deoxyguanosine is obtained from Sigma Chemical Co. (St. Louis, MO) and used without further purification. Acetonitrile (HPLC grade) and ammonium formate are purchased from Carlo Erba (Farmitalia, Carlo Erba, Milano, Italy) and Kodak (Eastman Kodak Co., Rochester, NY), respectively. Phthalocyanine complexed with zinc (ZnPcS2), prepared as described by Langlois et a1.,26 is a gift from Prof. J. E. van Lier (University of Sherbrooke, Qu6bec, Canada). The two diastereoisomers of 4,8-dihydro-4-hydroxy-8-oxo-2'-deoxyguanosine are prepared by dye photosensitization of 2'-deoxyguanosine. Typically, aerated Milli-Q deionized aqueous solutions of 1.0 mM 2'deoxyguanosine containing ZnPcS2 (OD6732.0) are exposed to the visible light generated from a 100-W halogen lamp equipped with a Kodak Model 23A (590 nm) cutoff filter. The cooling of the irradiation system is achieved by interfacing a heat filter (10 mm, circulating water) between the lamp and the photolyzed solution. The 4R* and 4S* diastereoisomers of 4,8-dihydro-4-hydroxy-8-oxo-2'deoxyguanosine are separated by HPLC on an analytical (250 x 4.6 mm i.d.) amino-substituted silica gel (mean particle size 5 /zm) Lichrocart column (Merck, Darmstadt, Germany) under isocratic conditions using an 8 : 2 (v/v) mixture of acetonitrile and 50 mM ammonium formate at a flow rate of 1.0 ml/min. 24 The HPLC system consists of two Model 302 Gilson dual pumps (Middleton, WI) equipped with a sil-9A Shimadzu automatic injector (Touzart & Matignon, Paris, France) and a L-4000 UV spectrophotometer (Hitachi, Tokyo, Japan) set at 230 nm. The pump is interfaced to an Apple II microcomputer that controls the eluent and to a Model 621 Gilson Data Master for quantitative analysis. 23 J.-L. Ravanat, M. Berger, F. Benard, R. Langlois, R. Ouellet, J. E. van Lier, and J. Cadet, Photochem. Photobiol. 55, 809 (1992). 24 J.-L. Ravanat, Ph.D. Thesis, University of Grenoble (1992). 25 G. W. Buchko, J. Cadet, M. Berger, and J.-L. Ravanat, Nucleic Acids Res. 20, 4847 (1992). 26 R. Langlois, H. Ali, N. Brasseur, J. R. Wagner, and J. E. van Lier, Photochem. Photobiol. 44, 117 (1986).
82
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[8]
Fast atom bombardment (FAB) mass spectra (glycerol matrix, 35-keV cesium atoms) in the positive and negative ion modes are obtained on a VG ZAB 2-EQ spectrometer (Fisons-V.G., Manchester, UK). A Hewlett Packard (HP) 5890 Series II gas chromatograph interfaced with a HP 5971 mass-selective detector is used for gas chromatography-mass spectrometry (GC-MS) analysis of the 4R* and 4S* diastereoisomers of 4-hydroxy-8-oxo-dG and the corresponding purine base. The GC apparatus is equipped with a fused-silica capillary column (12 m, 0.2 mm i.d., 0.33/zm film thickness) coated with cross-linked 5% phenylmethylsilicone (w/w). The GC oven is held at 150° for 1 min and programmed at 10°/min to a final temperature of 280 °. The temperature of both the injector and the detector is 280 °. Prior to GC-MS analysis, the 4R* and 4S* diastereoisomers of 4hydroxy-8-oxo-dG are subjected to acid hydrolysis in evacuated tubes with either 88% formic acid (v/v) for 30 min at 140°27 or 48% aqueous hydrofluoric acid (v/v) for 30 min at 0°28 in evacuated tubes. The samples are then lyophilized to dryness and subsequently derivatized with a mixture of bis(trimethylsilyl)trifluoroacetamide and acetonitrile (2 : 1, v/v) at 130° for 30 min. Photosensitized Reactions of 2'-Deoxyguanosine One of the problems associated with using phthalocyanine or other dyes, such as methylene blue, to generate type II photooxidation products is that they usually also give rise to a significant amount of type I photoproducts. Consequently it is necessary to be able to distinguish between the two main classes of photooxidation damage. One approach is to use a dye, such as benzophenone, which produces predominantly type I photooxidation lesions. The main type I photosensitization products of 2'-deoxyguanosine are 2,2-diamino-4-[2-deoxy-fl-D-erythro-pentofuranosyl)amino]-5(2H)-oxazolone, its unstable precursor 2-amino-5-[(2-deoxy~-D-erythro-pentofuranosyl)amino]-4H-imidazol-4-one, 29 and (2S)-2,5'anhydro- 1-(2-deoxy-fl-D-erythro-pentofuranosyl)-5-guanidinylidene-2hydroxy-4-oxoimidazolidine.30 The formation of these photoproducts may be rationalized in terms of quantitative initial hydrogen or electron abstraction from the guanine ring by photoexcited benzophenone. It is interesting 27 S. Boiteux, E. Gajewski, J. Laval, and M. Dizdaroglu, Biochemistry 167, 347 (1992). 28 j. Catania, B. C. Keenan, G. P. Margison, and D. S. Fairweather, Anal. Biochem. 167, 347 (1987). 29 j. Cadet, M. Berger, C. Decarroz, J.-F. Mouret, J. E. van Lier, and J. R. Wagner, J. Chim. Phys. 88, 1021 (1991). 30 G. W. Buchko, J. Cadet, J.-L. Ravanat, and P. Labataille, Int. J. Radiat. Biol. 63, 669 (1993).
[8]
HPLC AND MS ANALYSISOF DNA DAMAGEPRODUCTS
83
also to note that these compounds have also been shown to be produced from initial .OH addition at the C-4 position of the guanine ring followed by a fast dehydration reaction, 29'3~giving rise to a strong oxidizing neutral radical. Another method of distinguishing between type I and type II photooxidation products is to identify singlet oxygen-mediated lesions on the basis of the significant D20 enhancement effect or by using the thermal decomposition of the endoperoxide of 3,3'-(1,4-naphthylidene) dipropionate as a clean source of singlet oxygen effect. 23 Using these methods the two main 2'-deoxyguanosine type II photooxidation products are characterized as the 4R* and 4S* diastereoisomers of 4,8-dihydro-4-hydroxy-8-oxo-2'-deoxyguanosine on the basis of extensive nuclear magnetic resonance (NMR) measurements and mass spectrometry analysisfl 4'25 The formation of the two modified nucleosides is likely to be explained by the initial [2 + 4] cycloaddition of 10 2 to the 4,8-carbons leading to unstable endoperoxides (Fig. l) which, subsequently, undergo thermal decomposition. In addition, 7,8-dihydro-8-oxo-2'-deoxyguanosine has also been characterized as a relatively minor ~O2 oxidation product of 2'-deoxyguanosine. 23Note that 8-oxo-dG becomes a competitive substrate with respect to 2'-deoxyguanosine as soon as its yield of formation reaches a value of 0.75%. The two 4-hydroxy-8-oxo-dG diastereoisomers are the main singlet oxygen secondary oxidation products of 8 - o x o - d G . 23 It is likely that the formation of 8-oxo-dG involves the reduction of the transient 2'-deoxyguanosine 4,8-endoperoxides as inferred from the observation that the presence of Fe 2+ in the photooxidation reaction leads to a significant decrease in the yield of diastereoisomeric 4-hydroxy-8oxo-dG with a concomitant increase in the formation of 8-oxo-dG. This effect, which was also shown to decrease the yield of the oxazolone compound, is likely to explain the decrease in the ratio of 4-hydroxy8-oxo-dG to 8-oxo-dG in singlet oxygen-mediated oxidation of doublestranded DNA. Chromatographic Separation of the 4R* and 4S* Diastereoisomers of 4,8-Dihydro-4-hydroxy-8-oxo-2'-deoxyguanosine Separation of the ZnPcS2-mediated photooxidation products of 2'deoxyguanosine on the amino-substituted column is illustrated in Fig. 2. Note that the two diastereoisomers of 4-hydroxy-8-oxo-dG are well separated from 2'-deoxyguanosine and 2,2-diamino-4-[(2-deoxy-fl-Derythro-pentofuranosyl)amino]-5(2H)-oxazolone (the main stable type I photooxidation product). A baseline separation of the 4R* and 4S* diaster31 S. Steenken, Chem. Rev. 89, 503 (1989).
0
"0
Z
Z
"0
=
Z
TZ
Z
i\
Z
0
0
0
0
"0
[
1 Z
_= "0
_= cJ
0
-~
o 0 -r L
I (M
0 Q
"lZ
0
o~-,
=p 0"0
"~r"N
o
No
0 1-
._~ "0
[8]
H P L C AND M S ANALYSIS OF D N A DAMAGE PRODUCTS
85
dGuo
2 -
O E C
1 -
i
0.00
i
i
i
=
5.00
i
i
I
i
i
10.00
i
i
I
i
i
15.00
|
i
I
i
!
20.00
Time (min)
Fl•. 2. HPLC elution profile of 2'-deoxyguanosine (dGuo), 2,2-diamino-4-[(2-deoxy-fl-
o-erythro-pentofuranosyl)amino]-5(2H)-oxazolone (type I), and the S* and R* diastereoisomers of 7,8-dihydro-4-hydroxy-8-oxo-2'-deoxyguanosine(type II) on a Lichrocart aminosubstituted analytical column (250 x 4.6 mm i.d.). The eluent consisted of an 8 : 2 mixture of acetonitrile and 50 mM ammonium formate at a flow rate of 1 ml/min. Detection of the compounds was achieved by a variable wavelength spectrophotometer set at 230 nm.
eoisomers (the early eluting c o m p o u n d has been assigned as the 4S diastere o i s o m e r 25) is obtained. The purity of each of the diastereoisomers, as determined by 1H N M R spectroscopy, is greater than 95%. Hence, the amino-substituted silica gel column constitutes an appropriate analytical s y s t e m to separate these two specific products of singlet-mediated oxidation of 2'-deoxyguanosine. Mass Spectrometry of the 4R* and 4S* Diastereoisomers of 4,8-Dihydro-4-hydroxy-8-oxo-2'-deoxyguanosine A quasi-molecular ion at m/z 322 corresponding to (M + Na) ÷ was o b s e r v e d in the positive mode F A B - m a s s spectrum of both diastereoisomers of 4-hydroxy-8-oxo-dG. In addition, the negative m o d e F A B - m a s s spectrum of both oxidized nucleosides exhibited a parent ion at m/z 298 corresponding to the pseudomolecular ion (M - H ) - . 25
86
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[8]
73
6,000,000:
43o
5,000,000 ~
4,000,000
.1~ <
188
a,ooo,ooo 171
1,000,000 , 50
100
150
200
250
300
:J.,l...t..,.k 350
400
.
.
.
.
.
.
450
.
.
.
.
.
500
550
M/Z
FIG. 3. Mass spectrum of the pentatrimethylsilyl derivative of 4,8-dihydro-4-hydroxy-8oxoguanine from GC-MS analysis. Although the use of a high-resolution GC-MS method has proved to be useful for assaying certain modified nucleosides, 32 the trimethylsilyl (TMS) derivative of both diastereoisomers of 4-hydroxy-8-oxo-dG did not produce a prominent molecular ion at m/z 659 as confirmed using an extended mass range spectrometer. It is interesting to note that the mass spectra of the two diastereoisomers differed, depending on the stereochemistry of the hydroxyl group at the C-4 position. In particular, we observed a major ion at m/z 500 and 384 for the 4R* and 4S* diastereoisomers, respectively. On the other hand, the mass spectrum of the pentatrimethylsilyl derivative of 4-hydroxy-8-oxo-Gua, the corresponding base, obtained by GC-MS analysis exhibited a molecular peak at m/z 543 (Fig. 3). In addition, the loss of a methyl group from the molecular ion (m/z 528) is characteristic of the fragmentation of TMS derivatives. It should be noted that the identity of the molecular peak was confirmed by highresolution mass spectrometry measurement. Chromatographic and Mass Spectrometric Analysis of 4,8-Dihydro-4hydroxy-8-oxo-2'-deoxyguanosine and Related Base Component The 4R* and 4S* diastereoisomers of 4-hydroxy-8-oxo-dG may be obtained from ]O2-oxidized DNA by the combined use of both spleen 32 M. Dizdaroglu, J. Chromatogr. 367, 357 (1986).
[8]
HPLC AND MS ANALYSISOF DNA DAMAGEPRODUCTS
87
TABLE I HYDROLYSIS STABILITY OF 4,8-DIHYDRO-4-HYDROXY-8-OXOGUANINEa Conditions
Relative peak area (n = 3)
Control HF/H20 HCOOH
5.107 -+ 0.113 4.938 +- 0.114 4.646 -+ 0.075
a The quantitative importance of 4-hydroxy-8oxo-Gua was determined by HPLC analysis.
and snake venom phosphodiesterases. Although the 4-hydroxy-8-oxo-dG compounds are resistant to digestion by snake venom phosphodiesterase, they will be hydrolyzed by spleen phosphodiesterase, although at a rate slower than for unmodified nucleosides. 25 Such a digestion of the oxidized DNA should release 4-hydroxy-8-oxo-dG nucleosides which may then be detected by an on-line HPLC FAB-mass spectrometry assay. Alternatively, the inability of snake venom phosphodiesterase to hydrolyze the 5'-internucleotide phosphodiester bond adjacent to 4-hydroxy-8-oxo-dG may be used in a postlabeling assay to detect these lesions as dinucleoside monophosphates.33 The GC-MS method appears to be a suitable approach for the measurement of TMS derivatives of 4-hydroxy-8-oxo-dG, whose separation is well resolved on the capillary columns. Attempts have been made to extend the application of this GC-MS assay to the quantitative analysis of 4-hydroxy-8-oxo-Gua. The advantage of monitoring the specific ~O2 oxidation product of guanine is the increased sensitivity due to one peak formation, compared to two peaks for the related GC-MS analysis of the 4-hydroxy-8-oxo-dG nucleosides. Modified bases can be released from DNA by using various acids including formic acid (HCOOH), aqueous hydrofluoric acid (HF/H20), and hydrofluoric acid-pyridine (HF/Pyr). 34 The 4R* and 4S* diastereoisomers of 4-hydroxy-8-oxo-dG were found to be specifically converted to the corresponding enantiomers by acid treatment. The stability of 4-hydroxy-8-oxo-Gua toward acid hydrolysis conditions have been tested with HCOOH and HF/H20 (Table I). The use of HCOOH presents some drawbacks since the GC chromatogram of the resulting acidic hydrolyzate exhibited extra peaks which are indicative 33 G. W. Buchko, M. Weinfeld, F. Karimi-Booshehri, J. Cadet, and R. A. Floyd, 41st Radiat. Res. Conf., Dallas, Texas, 1993. 34 M. Polverelli, M. Berger, J.-F. Mouret, F. Odin, and J. Cadet, Nucleosides Nucleotides 9, 451 (1990).
88
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[9]
of decomposition processes. As an interesting alternative, HF/Pyr was found to be efficient in converting the diastereoisomeric 4-hydroxy-8-oxodG to the corresponding enantiomeric base derivatives. 24 The mild HF/ Pyr hydrolysis approach should be applicable in forthcoming GC-MS assays capable of monitoring the formation of a large array of oxidized DNA bases including 4-hydroxy-8-oxo-Gua. Acknowledgments This work was supported by National Cancer Institute Outstanding Investigator Grant CA 39910 to B.N.A. and by National Institute of Environmental Health Sciences Grant ES 01896 and Tobacco-RelatedDisease Research Program Fellowship Award 3FT-0401 to H.C.Y.
[9] A l k a l i n e E l u t i o n v e r s u s F l u o r e s c e n c e A n a l y s i s of DNA Unwinding
By CHRISTA BAUMSTARK-KHAN Introduction Induction and rejoining of DNA strand breaks in X-irradiated cells may be measured by two different methods: alkaline filter elution and fluorometric analysis of DNA unwinding (FADU). Both methods are compared in this chapter, and the FADU method has proved to be as sensitive as the alkaline filter elution approach in detecting X-ray-induced DNA damage. Even though the generation of free radicals is involved in many metabolic reactions in vivo, 1 free radicals are potentially dangerous for living c e l l s . 2'3 The issue of risk assessment from exposure to chemicals as well as from exposure to radiation, both of which produce free radicals (the reactive oxygen species in particular: hydrogen peroxide, superoxide anion, and hydroxyl radicals), centers around oncogenic, mutagenic, and teratogenic effects. Because DNA strand breaks may play an important role in mutagenesis and oncogenesis, the dose-effect relationships and rejoining kinetics are essential indicators reflecting cellular response pro-
l R. J. Y o u n g m a n , Trends Biochem. Sci. 9, 280 (1984). 2 B. N. A m e s , Science 221, 1256 (1983). 3 p. A. Cerutti, Science 227, 375 (1985).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994by AcademicPress, Inc. All fights of reproduction in any form reserved.
88
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[9]
of decomposition processes. As an interesting alternative, HF/Pyr was found to be efficient in converting the diastereoisomeric 4-hydroxy-8-oxodG to the corresponding enantiomeric base derivatives. 24 The mild HF/ Pyr hydrolysis approach should be applicable in forthcoming GC-MS assays capable of monitoring the formation of a large array of oxidized DNA bases including 4-hydroxy-8-oxo-Gua. Acknowledgments This work was supported by National Cancer Institute Outstanding Investigator Grant CA 39910 to B.N.A. and by National Institute of Environmental Health Sciences Grant ES 01896 and Tobacco-RelatedDisease Research Program Fellowship Award 3FT-0401 to H.C.Y.
[9] A l k a l i n e E l u t i o n v e r s u s F l u o r e s c e n c e A n a l y s i s of DNA Unwinding
By CHRISTA BAUMSTARK-KHAN Introduction Induction and rejoining of DNA strand breaks in X-irradiated cells may be measured by two different methods: alkaline filter elution and fluorometric analysis of DNA unwinding (FADU). Both methods are compared in this chapter, and the FADU method has proved to be as sensitive as the alkaline filter elution approach in detecting X-ray-induced DNA damage. Even though the generation of free radicals is involved in many metabolic reactions in vivo, 1 free radicals are potentially dangerous for living c e l l s . 2'3 The issue of risk assessment from exposure to chemicals as well as from exposure to radiation, both of which produce free radicals (the reactive oxygen species in particular: hydrogen peroxide, superoxide anion, and hydroxyl radicals), centers around oncogenic, mutagenic, and teratogenic effects. Because DNA strand breaks may play an important role in mutagenesis and oncogenesis, the dose-effect relationships and rejoining kinetics are essential indicators reflecting cellular response pro-
l R. J. Y o u n g m a n , Trends Biochem. Sci. 9, 280 (1984). 2 B. N. A m e s , Science 221, 1256 (1983). 3 p. A. Cerutti, Science 227, 375 (1985).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994by AcademicPress, Inc. All fights of reproduction in any form reserved.
[9]
FLUORESCENCE ANALYSIS OF D N A STRAND BREAKS
89
c e s s e s . 4-6 In addition to the primary induced breaks many other DNA lesions may be transformed to strand breaks and can thus easily be measured and quantified. It is therefore necessary to have methods at hand which measure DNA strand breaks rapidly and reproducibly. A number of methods for quantitative measurement of DNA strand breaks have been developed. 7-~° Because they are based on different principles and performed under different conditions, there are a number of questions to be discussed: (1) what does a specific method really measure, (2) what are the limitations of the method, and (3) what interfering factors may affect the results? In this chapter two methods which measure DNA strand breaks are introduced and compared: the alkaline elution method 8 and the fluorescence analysis of DNA unwinding.~°-13 Both methods are based on DNA denaturation (unwinding) in alkaline solutions. Unwinding begins from all free DNA ends (DNA double-strand breaks and DNA single-strand breaks), which are multiplied in cells that have been exposed to radiation or DNA-reactive agents. 12-~5 Under stringent alkaline conditions DNA unwinding should be complete. Under moderate alkaline conditions DNA unwinding is a time-dependent process which has to be stopped after a certain period of time by neutralization.
Experimental Procedures Cell Culture
Chinese hamster ovary cells (CHO-9) are grown in monolayers on plastic in Eagle's minimal essential medium (MEM) supplemented with 15% fetal bovine serum, glutamine (0.292 /xg/ml), and pyruvate (0.11 /zg/ml) at 37° in a humidified incubator with 95% air and 5% CO2. Stock cultures are propagated by subculturing every third day at an initial density of 1 × 10 4 cells/cm 2 (5 ml MEM) in 25-cm 2 flasks. After 3 days stock CHO 4 I. R. Radford, Int. J. Radiat. Biol. 49, 611 (1986). 5 E. Dikomey, Int. J. Radiat. Biol. 57, 1169 (1990). 6 C. Baumstark-Khan, E. Aufderheide, and H. Rink, Ophthalmic Res. 24, 220 (1992). 7 R. A. McGrath and R. Williams, Nature (London) 212, 534 (1966). 8 K. W. Kohn and R. A. Grimek-Ewig, Cancer Res. 33, 1849 (1973). 9 G. Ahnstr6m and K. A. Edvardsson, Int. J. Radiat. Biol. 26, 493 (1974). i0 H. C. Birnboim and J. J. Jevcak, Cancer Res. 41, 1889 (1981). 11 p. M. Kanter and H. S. Schwartz, Mol. Pharmacol. 22, 145 (1982). ~2 C. Baumstark-Khan, U. Griesenbach, and H. Rink, Free Radical Res. Commun. 16, 381 (1992). 13 H. C. Birnboim, this series, Vol. 186, p. 550. 14 G. Ahnstr6m and K. Erixon, Int. J. Radiat. Biol. 23, 285 (1973). I~ B. Rydberg, Radiat. Res. 61, 274 (1975).
90
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[9]
cells are used to prepare the cells for experiments. Plateau-phase cultures are obtained by growing 5 × 10 3 cells/cm z in 30-mm tissue culture dishes (1.5 ml) for 4 days without changing the medium. Flow cytometric measurements indicate that such resting phase cultures contain 87.6% cells in the G1/G0 phase, 4.2% cells in the S phase, and 8.2% cells in the G2/M phase.
Induction and Rejoining of DNA Strand Breaks Irradiation is an invaluable source of primary free radicals, which in turn give rise to secondary radicals. To expose the cells to free radicals the cells are irradiated with 200-keV X-rays (0.5-mm Cu filter, dose rate 2 Gy/min). Cells are held on ice 30 min before and during irradiation in order to prevent any enzymatic cellular response during the treatment. Dose-response curves for the induction of strand breaks are established using either the alkaline filter elution technique 8 or fluorometric analysis of DNA unwinding (FADU). 1°-~3 To allow rejoining of strand breaks after irradiation, cells are further incubated at 37 ° and analyzed thereafter.
Alkaline Elution The number of DNA strand breaks is measured by a modified alkaline elution method. 8 Exponentially growing cells are labeled with 0.751.85 x 10 4 Bq/ml [methyl-3H]thymidine (Amersham Radiochemical Center, Amersham, UK, TRK 120, 185 GBq/mmol). The cells enter the plateau phase after at least two generation times, after which the radioactive medium is discarded and the cells are washed twice with phosphatebuffered saline (PBS). Cells are then incubated for at least 12 hr in nonradioactive medium before further treatment. The elution technique is adopted from the original protocol with minor modifications. After irradiation the radioactive labeled cells are scraped from the culture dish, layered onto 25-mm (0.22/xm pore size) polyvinyl chloride filters (Millipore, Bedford, MA, GVLP 25) placed in disposable filter holders (Fig. 1), and washed twice with ice-cold PBS. The filter holders are connected to 20-ml syringes, and the cells are lysed at room temperature for 30-60 min on the filters using 2.5 ml of lysis buffer (40 mM H4EDTA, 2 M NaCI, 7 mM N-lauroylsarcosine, pH 10.0), followed by washing with 5 ml of 20 mM/liter Na2EDTA, pH 10.0. Elution buffer (20 mM H4EDTA, adjusted to pH 12.4 with tetrapropylammonium hydroxide) is filled into the syringes, and alkaline elution is performed with continuous pumping at a flow rate of 0.05 ml/min, with fractions collected at 30-min intervals. Elution is finished after collecting 10 fractions per
[9]
FLUORESCENCE ANALYSIS OF D N A STRAND BREAKS
91
FIG. 1. Alkaline elution of D N A from polyvinyl chloride filters. The filters (0.22/~m pore size) were placed in disposable filter holders and connected to syringes serving as buffer reservoirs. The outlets of the filter holders were connected over a multichannel p u m p to a fraction collector equipped with a timer.
column. After neutralization, the radioactivities of all samples and filters are determined by liquid scintillation counting. The quantity of DNA retained on the filter is calculated for each fraction from the radioactivity of the 10 samples and the corresponding filter. Using these values elution profile curves are established for all columns. For determining dose-effect curves, the amounts of nonelutable DNA (F) which correspond to DNA retained on the filters for fraction 5 (after 7.5 ml elution) are used.
Fluorometric Analysis of DNA Unwinding To quantify DNA strand breaks by the fluorometric procedure, ~°-13 the inoculated cells are grown to confluence. After irradiation, trypsinized cells (trypsin at 0.2 g/liter in 135 mM NaC1, 3 mM KC1, 5 mM NazHPO 4 , 1 mM KH2PO 4, 5 mM trishydroxyaminomethane, 0.45 mM Na2EDTA) are transferred to ice-cold plastic tubes (1 ml).
92
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[9]
The FADU method uses the fluorescence data from test samples (P) as well as from unirradiated reference samples (T and B). The background contribution to fluorescence by components other than double-stranded DNA (including unbound dye) is estimated from unirradiated samples (B fluorescence). In these B samples the cell extract is sonicated under alkaline conditions in order to break all DNA and to bring about complete unwinding of the low molecular weight DNA fragments. B samples show the lowest level of fluorescence. Procedure. Suspended cells (1 ml) of the B samples are lysed by addition of 1 ml of 0.1 M NaOH, sonicated (Braun Labsonic 1510, Melsungen, Germany, 100 W, 15 sec), incubated at 20 ° for the whole unwinding period (30 min, if not indicated otherwise), and subsequently neutralized by addition of 1 ml of 1 M HC1 and resonicated. The second reference set is used for estimating total fluorescence (T fluorescence). In these T samples the measured contribution to fluorescence comes from double-stranded DNA and contaminations which include the unbound dye. The cell extract is sonicated under neutral conditions to prevent unwinding of the DNA. The difference T minus B (T - B) provides an estimate of the amount of double-stranded DNA in unirradiated cell extracts. Procedure. Suspended cells (1 ml) of the T samples are lysed by the addition of 1 ml of 0.1 M NaOH immediately followed by neutralization (addition of 1 ml of 1 M HCI), incubated at 20 ° for the whole unwinding period, and sonicated (Braun Labsonic 1510, 100 W, 15 sec) afterward. The third samples set is used to estimate the unwinding rate of radiation-damaged DNA (Fig. 2). The cells are exposed to alkaline conditions
II II
'
treated cells
FLUOROMETRIC ANALYSIS l + blsbenzemlde 1.25p mol ,,~ ~ ,.~
U
LYSIS: ÷ NeOH0.1tool
/ 1
NEUTRALIZATION', * HOI0.1reel SONICATION'.. 100 W, 15s
FIG. 2. Schematic drawing of the alkaline unwinding procedure for P samples and fluorescence analysis of double-stranded DNA.
[9]
FLUORESCENCE ANALYSIS OF
DNA STRAND
BREAKS
93
which bring about partial unwinding of DNA. The degree of unwinding in these P samples, which are subjected to X-irradiation, is related to the number of DNA breaks (P fluorescence). The difference P minus B (P B) provides an estimate of the amount of DNA that remains doublestranded, thus indicating the amount of the induced DNA damage. Procedure. Suspended cells (1 ml) of the test samples are lysed by gentle addition of 1 ml of 0.1 M NaOH without mixing, followed by incubation at 20° (30 rain, if not indicated otherwise). During this unwinding period the alkali diffuses into the viscous lysate to give a final pH of 12.4. The tubes are shielded from light and vibrations to avoid artificially induced strand breaks. After DNA unwinding, the P samples are neutralized by addition of 1 ml of 1 M HC1 (pH 7,1-7.4) and sonicated to render the samples homogeneous, to minimize renaturation of DNA, and to reduce the fraction of DNA molecules containing both single- and doublestranded regions. All samples (B, T, and P) then receive 1 ml of fluorochrome-containing buffer (1.25/zM bisbenzamide, if not indicated otherwise, in 0.15 M phosphate buffer, pH 7.6). After mixing, relative fluorescence intensities are read from a spectrofluorimeter operating at 355 nm (excitation) and 450 nm (emission). The fraction of double-stranded DNA (F) is calculated according to F = (P - B)/(T - B), where T, P, and B are fluorescence intensities of the T, P, and B samples, respectively. Dose-Effect Curves
Data from the fractions of nonelutable DNA (alkaline elution) or fractions of double-stranded DNA (FADU) are standardized according to Fo/Fo~o (where F Dand Fo=oare the DNA fractions of treated and untreated ceils, respectively). Resulting data are expressed as strand scission factors (SSFs) which are defined as -ln(Fo/Fo=o). The dose-effect curves are fitted by least-squares linear regression analysis of SSFs versus X-ray dose. Rejoining Curves
The FADU method is used to quantify the time-dependent repair of DNA damage. In these measurements the irradiated cells are kept under nongrowth conditions for different periods of time and analyzed thereafter. The quantity of double-stranded DNA (SSFs) is plotted over the repair time. The resulting hyperbolic curve shows the decay of DNA damage with time. It can be expressed as the sum of three terms {SSFs are defined in this case a s --ln[FD(o/FD=o(t=o)]} 16 according to Eq. (I), reflecting three 16 E. Dikomey and J. Franzke, Int. J. Radiat. Biol. 50, 893 (1986).
94
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[9]
different components of DNA repair. The halftimes for the different repair components z(i) , Z(n) , and r(m ) are calculated for each component using Eq. (2). The appropriate ordinate intercepts are a(~), a(n), and a(iii). SSF = (slope(i) t + act)) + (slope(ii)t + a(ii)) + (slope(m)t + a(nl)) z = ( - 1/slope) log 2
(1) (2)
Results and Discussion Methods which are used to measure DNA strand breaks (single-strand breaks and double-strand breaks) make use of alkaline solutions for cell lysis. T M Such alkaline treatment inactivates DNA-degradative enzymes, removes proteins from the DNA, and separates the two strands of the double helix. The number of strand breaks may then be calculated either from the size distribution of damaged DNA 7 or from the amount of singleor double-stranded forms. 14 The method of alkaline elution as well as DNA-unwinding protocols are used frequently, probably because they are easy to apply and have simple fundamental principles.
Alkaline Elution The alkaline elution technique requires simple and inexpensive filters, a multichannel pump, a fraction collector (Fig. 1), and a liquid scintillation counter. Up to l0 samples can be handled simultaneously. The method is based on the rate at which single-stranded DNA elutes through an inert filter membrane under stringent denaturing conditions. The filter does not absorb DNA under the given conditions but rather mechanically impedes the passage of DNA. Because elution volumes are usually high (15-30 ml), cells have to be radioactively labeled (e.g., with [methyl-3H]thymidine) for accurate DNA measurements. Radioactive labeling is expensive (label, scintillation fluids and vials, disposal of radioactivity) and troublesome to perform (need for safety). Figure 3A shows the fraction of eluted radioactivity of irradiated CHO cells to be dependent on elution time (fraction number). The damaged DNA passed through the filter, thus increasing the level of radioactivity, in the first two fractions (30-60 min). This in turn resulted in reduced amounts of radioactive label remaining on the filter that is available for radioactivity measurements at the end of the elution (5 hi'). The fraction of filter-retained DNA for corresponding times (fraction number) can be calculated from the eluted radioactivity and the radioactivity on the filter. The elution rate (Fig. 3B) decreases exponentially with elution time (fraction number).
[9]
95
FLUORESCENCE ANALYSIS OF D N A STRAND BREAKS
• u °9t
B k°
t
kl!li~.-.
°8t
~t-o:7
/ I%,.
....
k O8 k °.6
o
o 7 J ,': " J:'i
/
,,lli',~ " , lii'~ -"i...
-~Po.s /
~-
(2£
o.sd:,;I It'i
~\ '~A11 ii x\, \.
!"11 -'-'los
S
,k
0
/
I
,
',
",~
/
0 ~ !I l '? I 10""~I~1 I II~'~|0"1 0 1 2 5 4 5 6 7 8 910fi 0 1 5 4 5 6 7 8 910 FRACTION NUMBER FRACTIONNUMBER
'<
FIG. 3. Alkaline elution of DNA from polyvinyl chloride filters. CHO cells prelabeled with [3H]thymidine were lysed on the filters and eluted immediately after X-irradiation with varying doses. Fractions (1.5 ml) were collected at 30-rain intervals, and radioactivities of the samples and the filters (fi) were determined (A). The fractions of DNA remaining on
the filter at the time of fraction collection(B) were calculatedfrom the data shown in (A). (A) 0 Gy; (zx) 2 Gy; (u) 5 Gy; ([]) 7 Gy; (o) 10 Gy; (o) 15 Gy.
The long periods of time needed for the alkaline elution process may cause a problem inasmuch as it is difficult to avoid a partial hydrolysis of abasic sites in the DNA. 17 Elution buffers (pH 12.1-12.4) accelerate the elution rate, if the sample contains alkali-labile sites which get transformed to DNA strand breaks during elution. In addition, cell DNA suffers a spontaneous loss of purines, which increases with increasing pH. 18'19 The problem could be minimized by performing the elution at low temperatures, but even at 4° two-thirds of depurinated DNA gets hydrolyzed in 0.1 M NaOH within 10 m i ni 7
Fluorometric Analysis of DNA Unwinding The DNA-unwinding method needs fewer instruments than the alkaline elution. It can be performed in any laboratory and requires only an ice bath, a sonicator, and a fluorimeter. A large number of samples can thus be handled and analyzed within a few hours. iv M. V. M. Lafleur, J. Wolohuis, and H. Loman, Int. J. Radiat. Biol. 39, 113 (1981). 18 T. Lindahl and B. Nyberg, Biochemistry 11, 3610 (1972). 19 T. Lindahl and O. Karlstr6m, Biochemistry 12, 5151 (1973).
96
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[9]
150->-
i--
_z
5
125 10075 ~
50-
g 25 o 1
2
3
4
5
6
7
8
9
10
CELL NUMBERS ( x l O s)
FIG. 4. Fluorescence enhancement of bisbenzamide by cell DNA. Varying numbers of CHO cells were treated with unwinding solution (T and B samples) and mixed afterward with bisbenzamide-containing buffer. (o) B, Single-stranded DNA; (e) T, double-stranded DNA.
The basis for DNA unwinding is a time-dependent transformation of duplex DNA to single-stranded DNA under moderate alkaline conditions. 9-15 Quantification of the remaining double-stranded DNA can be performed either by scintillation counting of [3H]thymidine in fractions of double- and single-stranded DNA after hydroxyapatite chromatography 9 or by fluorescence measurements, thus omitting radioactive labeling. The fluorochrome bisbenzamide reacts specifically with DNA but not with RNA and protein. This results in an enhanced fluorescence, the extent of which depends on the adenine-thymine content of the DNA. 2° The degree of fluorescence of DNA-bound bisbenzamide is a function of the amount and state of the DNA (double- or single-stranded form). Data in Fig. 4 show the fluorescence of different numbers of CHO cells which were subjected to unwinding conditions (T and B samples). There are linear relationships between cell numbers and fluorescence, showing that the degree of fluorescence is dependent on the amount (cell numbers) and the state of DNA. The fluorescence of double-stranded DNA is about twice as high as that of single-stranded DNA for cell numbers up to 1 × 106. For higher cell numbers, the intensity of fluorescence of doublestranded DNA becomes saturated (data not shown), but not that for singlestranded DNA because binding affinities of bisbenzamide to duplex and 20 C. F. Ceasrone, C. Bolognesias, and L. Santi, Anal. Biochem. 100, 188 (1979).
[9]
FLUORESCENCE ANALYSIS OF D N A STRAND BREAKS
97
single-stranded DNA are different. The difference in fluorescence (T minus B) for high cell numbers (> 106) is smaller than for low cell numbers (< 106) owing to unsaturated binding between the dye and duplex DNA. Figure 5 shows the dependence of fluorescence intensity on dye concentration up to 2.5/xM for T and B samples prepared from 2 x 10 6 CHO cells. For higher dye concentrations the fluorescence intensity becomes saturated. It appears that the optimal values for cell numbers and dye concentrations in the FADU assay are 5 x 105 to 1 x 106 cells and 1.25 /xM bisbenzamide, respectively. The FADU technique is based on time-dependent alkaline denaturation of DNA under moderate denaturing conditions (Fig. 6). Aliquots of cells (5 x 105) were subjected to the alkaline unwinding treatment for different periods of time. Unwinding starts at the strand ends or other unwinding units. The fluorescence intensities for B samples are low and stay constant during the unwinding period. Because of extensive DNA fragmentation, conversion of the native DNA to single-stranded DNA comes to a completion during the first seconds of the treatment (F = 0). The DNA in T samples remain in duplex form and show no decrease in fluorescence intensity (F = I). The P samples contain DNA in a partially denatured form; accordingly, the amount of double-stranded DNA is a function of unwinding time and dose (0 < F < 1). As shown in Fig. 6 in a log-log plot, DNA unwinding proceeds linearly with time. The initial unwinding 300 275
~> 2sod.
•
•
225tm
~-
z
L~
200-
175-
¢
150-
"'""
125' 100-
O J L
-~-~"
75 50-
2s~ O2 0.0
0.5
1.0
1.5
2,0
2.5
3.0
5.5
4.0
DYE CONCENTRATION (/~M)
FIG. 5. Fluorescence enhancement of bisbenzamide by cell DNA. CHO cells (2 x 106 cells) were treated with unwinding solution (T and B samples) and mixed afterward with different amounts of bisbenzamide-containing buffer. (o) T; (e) B; (o) T - B.
98
[9]
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
100 U3 Z
(iI Z
q]i,::-÷"-.
8O i'0
LJ G) Z
60
.. ~'-m , ~-.....,, " "~
L~ C) 6~ Ld
50
o
t
4O L
50
,
0.1
r
+++
• i
,rll~]
r
~o
i
i
iilll
I
r
T i
i
illl~j
~o.o
l~ i
I
i
i
~oo.o
UNWINDING TIME (MIN)
FIG. 6. Kinetics of DNA unwinding in an alkaline solution. Cells were lysed immediately after X-irradiation. Fluorescence intensities of bisbenzamide-DNA complexes were measured after different unwinding times for B, T, and P samples. (.) T; (o) B; (.) 0 Gy; (I) 2 Gy; ([]) 5 Gy; (e) 10 Gy.
rate (fast unwinding) is thus related to the amount of strand breaks. With a growing length of single strands the degree of free movement decreases, and unwinding proceeds at a reduced rate. Davison 21 showed that the time necessary for complete strand separation in alkali for phage DNA of various sizes increases with the square of its molecular mass. For a DNA molecule of 1.2 x .108 Da it takes about 25 sec for complete strand separation. For DNA of molecular masses as high as 10 9, 101°, and 10u Da total unwinding would require 30 min, 50 hr, and 200 days, respectively, z~ For DNA of CHO cells irradiated with 2, 5, and 10 Gy the times necessary for complete unwinding are 20 days, 20 hr, and 160 min, respectively (Fig. 6). Assuming that X-irradiation induces randomly distributed strand breaks along the DNA molecule and that the slow denaturation process proceeds continuously, severely damaged DNA with only short duplex portions would separate completely under alkaline conditions. This would result in a wrong estimate of strand break induction at high doses, particularly when long unwinding times are used. Therefore, a relatively short period of alkali treatment was chosen. Under these conditions (20°, 30 min) even severely damaged DNA will not unwind completely. Thus, even relatively closely spaced DNA strand breaks may be detected as 2i p. F. Davison, J. Mol. Biol. 22, 97 (1966).
[9]
FLUORESCENCE ANALYSIS OF D N A STRAND BREAKS
99
2.0-U3 0 i--
1.5-
<~ L Z 0
1.0
u~ El) r..p 0.5 z <~ [z" kco
0.0 Ej ' ' ~
.....
0.0
I"l'~r'"l 2.5
....
5.0
"~"1
.........
7.5
I'"
10.0
" J' " 1 ~q~ ' ' ' r ~ r r n 12.5 15.0
X - R A Y DOSE (GY)
FI6.7. Comparison of results obtained by alkaline elution (©) and fluorometric analysis of DNA unwinding (e) after X-irradiation with varying doses. Strand scission factors [-ln(Fo/Fo=o)]were compared using Student's t-test (significance level: 2p -< 0.05). Values for SSF/Gy were 0.1231 and 0.1207 for FADU and the alkaline elution, respectively. Data obtained by the two methods were nearly identical (2p -> 0.9).
single events. Similar kinetics of unwinding have been determined by hydroxyapatite chromatography.J5,22
DNA Damage Induction The elution profiles (Fig. 3B) or the unwinding kinetics (Fig. 6) can be used to determine the d o s e - r e s p o n s e relationships for the X-irradiated C H O cells. Results for D N A breaks in X-irradiated cells determined by F A D U and by alkaline elution are c o m p a r e d in Fig. 7. Results are expressed as strand scission factors [SSF = -ln(Fo/F~=o)]. There are no differences b e t w e e n the two methods (2p -> 0.9). The F A D U technique detects D N A damage in the same dose range as does alkaline elution (0-15 Gy). Reliable m e a s u r e m e n t s are possible in the range between 10 and 100% content of double-stranded D N A in a sample (corresponding to3>SSF>0). As shown in Fig. 7, no significant difference was found in the d o s e response curves for strand b r e a k induction by X-irradiation b e t w e e n alkaline elution and fluorometric analysis of D N A unwinding (FADU). Other groups also found no differences between the results of different methods 22 H. Rink and R. Zblewski, Ophthalmic Res. 15, 45 (1983).
100
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[9]
for measuring the extent of all DNA strand breaks (double-strand breaks, single-strand breaks, alkali-labile sites): Ahnstr6m and co-workers 9'14 showed comparable results for DNA unwinding of X-irradiated V79 cells as determined by hydroxyapatite chromatography and alkaline velocity sedimentation, Kanter and Schwartz 11 found no differences for L1210 cells as determined by hydroxyapatite chromatography and FADU, and Olive et al. 23 found no differences for alkaline sucrose gradient centrifugation and alkaline elution with V79 cells. When FADU data are compared with data from methods for measuring damage other than the sum of strand breaks, the differences are striking. With mutagens, fluorescence analysis made it possible to detect DNA repair for doses that were 3-4 times lower than those used with UDS (unscheduled DNA synthesis, i.e., excision repair of thymidine dimers). 24The nondenaturing elution method, which only determines the DNA double-strand breaks, requires doses for damage assessment that are 10-20 times higher than those needed for FADU.12'25 DNA Damage Repair Cultured cells have the ability to rapidly rejoin the DNA strand breaks induced by irradiation. 23-29 Figure 8 shows that the FADU method is suitable for following this process in X-irradiated (12 Gy) CHO cells. During the first minutes after irradiation the DNA damage (SSF) decreases very rapidly followed by a slower decrease. For incubation intervals exceeding 40 min the repair kinetics can be described by an exponential decrease [SSF -ln(Fou)/FD=o)]. This exponential component represents the kinetics of strand breaks with the longest repair halftime. The halftime '/'(lid and the initial fraction a~tit) for this repair component were determined by a regression analysis for the time range between 40 and 420 min to be 760 rain and 6.5%, respectively. When the amount of remaining damage (SSF) calculated for component 1II is subtracted from the experimental data, component II can be calculated by regression analysis for the time range between 15 and 40 min (r~ii)of 15.9 min, a(ii) corresponding to 45.3%). After subtracting component II from the remaining values the parameters for component I are obtained (r~i)of 5.4 min, a~i)corresponding to 48.2%). =
23 p. L. Olive, J. Hilton, and R. E. Durand, Radiat. Res. 107, 115 (1986). 24 L. Celotti, P. Ferraro, and M. R. Biasin, Mutat. Res. 281, 17 (1992). 25 C. Baumstark-Khan, Int. J. Radiat. Biol. 61, 191 (1992). 26 p. E. Bryant and D. Bl6cher, Int. J. Radiat. Biol. 38, 335 (1980). 27 B. Rydberg and K. J. Johanson, Radiat. Res. 64, 281 (1975). 28 E. Dikomey and J. Franzke, Radiat. Environ. Biophys. 27, 29 (1988). 29 K. F. Weibezahn and T. Coquerelle, Nucleic Acids Res. 9, 3139 (1981).
[9]
FLUORESCENCE ANALYSIS OF D N A STRAND BREAKS
101
2.0-
O0 CO
1.8component
1,6--
I
0OL
1.4-
£D <~
1.2-
1I III
I---
L~
"r
fraction
6.4 rain 15.9 rain 760 rain
48.2% 45.5% 6.5%
1.0--
z 0.8
-
9 9 0
60
120
180
240
,300
360
420
INCUBATION TIME (MIN)
FIG. 8. Rejoining kinetics (SSF versus incubation time) of X-ray-induced DNA strand breaks. For incubation times longer than 40 min the curve was fitted by linear regression (component III). After subtraction of component III from the experimental data, the residual repair kinetics of component II was calculated for incubation times between 15 and 40 min. After subtraction of component II, the residual repair kinetics of component I was determined for incubation times shorter than 15 min.
These calculated halftimes are in good agreement with published values for CHO cells. 5'28 It was assumed z6 that the repair components I and II represent the repair of single-strand breaks, whereas component III represents the repair of double-strand breaks. This assumption is based on the fact that the initial fraction (a(iii) corresponding to 6.5%) and the half-time of component III are in agreement with measurements using the neutral sedimentation method, which detects only double-strand breaks. 26 The repair components I and II were discussed 16,28as representing two different classes of single-strand breaks, or were attributed to primary single-strand breaks induced by X-irradiation (I) and secondary single-strand breaks caused by incision enzymes during excision repair (II). When interpreting strand break data from DNA repair experiments, small differences between control cells and repairing cells have to be considered for long repair times. The sensitivity of the method in resolving such differences depends on the background level of normal strand breaks (which are required for the metabolic function of DNA, e.g., D N A synthesis) rather than the absolute detection capacity. The background level of single-strand breaks was calculated to be of the order of 3 per 10 l° Da,
102
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[10]
but in extreme cases may be much lower. 3° Rydberg 3° calculated that the alkaline unwinding method permits strand break detection for doses below 0.1 Gy. As shown in Fig. 8 no measurable strand break damage remained after long repair incubations. In contrast, strand break frequencies were lower for irradiated cells than for unirradiated controls. This does not mean that the repaired DNA has a molecular mass larger than the native (nondamaged) DNA; rather, it implies that the DNA-damaging treatment could change the pattern of cell metabolism (e.g., shutdown of DNA synthesis) which in turn affects the background level of strand breaks. 31'32 3o B. 31 G. 32 G. E.
Rydberg, Radiat. Res. 81, 492 (1980). Ahnstr6m, Int. J. Radiat. Biol. 54, 695 (1988). Ahnstr6m and K. A. Edvardson, in "DNA Repair Mechanisms" (P. C. Hanawalt, C. Friedberg, and C. F. Fox, eds.), p. 469. Academic Press, New York, 1978.
[10] P u r i f i c a t i o n a n d P r o p e r t i e s o f Yeast Redoxyendonuclease B y L A U R A A U G E R I , KRISTA K. HAMILTON, A M Y M . MARTIN, PAULOS YOHANNES,
and P A U L W . DOETSCH
Introduction
Saccharomyces cerevisiae is an attractive system for enzymological studies of eukaryotic DNA repair enzymes as it offers an inexpensive and abundant supply of cells for purification of proteins as well as an amenable system for subsequent cloning and genetic experiments. In addition, several DNA repair enzymes from yeast are functionally similar to ones found in other eukaryotic organisms, including mammalian cells. Yeast redoxyendonuclease is a DNA repair enzyme from S. cerevisiae that recognizes and removes a variety of oxidatively damaged bases and is functionally similar to a well-characterized prokaryotic enzyme from Escherichia coli, endonuclease III (endo III). 1Yeast redoxyendonuclease removes ring-saturated and ring-cleaved base damage products via a combined N-glycosylase-apurinic/apyrimidinic (AP) lyase mechanism resulting in DNA strand scission at the site of damage. I j. Gossett, K. Lee, R. P. Cunningham, and P. W. Doetsch, Biochemistry 27, 2629 (1988).
METHODSIN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
102
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[10]
but in extreme cases may be much lower. 3° Rydberg 3° calculated that the alkaline unwinding method permits strand break detection for doses below 0.1 Gy. As shown in Fig. 8 no measurable strand break damage remained after long repair incubations. In contrast, strand break frequencies were lower for irradiated cells than for unirradiated controls. This does not mean that the repaired DNA has a molecular mass larger than the native (nondamaged) DNA; rather, it implies that the DNA-damaging treatment could change the pattern of cell metabolism (e.g., shutdown of DNA synthesis) which in turn affects the background level of strand breaks. 31'32 3o B. 31 G. 32 G. E.
Rydberg, Radiat. Res. 81, 492 (1980). Ahnstr6m, Int. J. Radiat. Biol. 54, 695 (1988). Ahnstr6m and K. A. Edvardson, in "DNA Repair Mechanisms" (P. C. Hanawalt, C. Friedberg, and C. F. Fox, eds.), p. 469. Academic Press, New York, 1978.
[10] P u r i f i c a t i o n a n d P r o p e r t i e s o f Yeast Redoxyendonuclease B y L A U R A A U G E R I , KRISTA K. HAMILTON, A M Y M . MARTIN, PAULOS YOHANNES,
and P A U L W . DOETSCH
Introduction
Saccharomyces cerevisiae is an attractive system for enzymological studies of eukaryotic DNA repair enzymes as it offers an inexpensive and abundant supply of cells for purification of proteins as well as an amenable system for subsequent cloning and genetic experiments. In addition, several DNA repair enzymes from yeast are functionally similar to ones found in other eukaryotic organisms, including mammalian cells. Yeast redoxyendonuclease is a DNA repair enzyme from S. cerevisiae that recognizes and removes a variety of oxidatively damaged bases and is functionally similar to a well-characterized prokaryotic enzyme from Escherichia coli, endonuclease III (endo III). 1Yeast redoxyendonuclease removes ring-saturated and ring-cleaved base damage products via a combined N-glycosylase-apurinic/apyrimidinic (AP) lyase mechanism resulting in DNA strand scission at the site of damage. I j. Gossett, K. Lee, R. P. Cunningham, and P. W. Doetsch, Biochemistry 27, 2629 (1988).
METHODSIN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
[10]
YEASTREDOXYENDONUCLEASE
103
Assay Method
Principle. Yeast redoxyendonuclease activity can be directly determined by monitoring base-specific DNA strand breakage in oxidatively damaged end-labeled DNA fragments of defined sequence. Comparison of redoxyendonuclease-generatedDNA strand scission products with basespecific chemical cleavage reaction products 2 on DNA sequencing gels reveals the nucleotide location of cleavage and provides information on redoxyendonuclease substrate specificity, base specificity of cleavage, and mode of phosphodiester bond cleavage. We have found this method, which is described in detail elsewhere in this volume, 3 to be particularly useful for the detection of several prokaryotic and eukaryotic DNA repair enzymes. Although the method can be utilized for specific activity measurements at various stages of enzyme purification, we have not reported such values for the present purification of yeast redoxyendonuclease. Our past experiences with the purification of this enzyme indicate that such measurements are misleading indicators of the actual purity and stability of the enzyme because of the presence of suspected inhibitory proteins purified along with yeast redoxyendonuclease that are removed during the later purification steps. Generation of Redoxyendonuclease Substrates. Several different 3'and 5'-32p end-labeled DNA fragments can be utilized as redoxyendonuclease substrates. Although restriction fragments from the plasmids pUC19 and pGEM-2 (Promega, Madison, WI) are used in these studies, virtually any end-labeled plasmid DNA fragment in a size range of 50-250 bp can be utilized as redoxyendonuclease substrates. Generation of substrates involves plasmid cutting with a first restriction enzyme, 3'- or 5'-32P end-labeling, cutting with a second restriction enzyme to generate a suitable length DNA fragment (50-250 bp) labeled at only one terminus, and purification on a preparative, nondenaturing 8% (w/v) polyacrylamide gel as previously described. 4 The 3' end-labeling reactions are carded out with [o~-32p]dATP(specific activity 3000 Ci/mmol) and DNA polymerase I (Klenow); 5' end-labeling reactions are carried out with [y-32p]ATP (specific activity 5000 Ci/mmol) and polynucleotide kinase as previously described. 4 The yeast redoxyendonuclease substrates are as follows: fragment l, 92-bp 3'-end-labeled EcoRI-PvuII restriction fragment generated from pGEM-2; fragment 2, 91-bp 3'-end-labeled EcoRI-PouII restriction fragment generated from pUCl9; fragment 3, 179-bp 5'-end-labeled HindIII-PvuII restriction fragment generated from pUC19. 2 A. M. Maxam and W. Gilbert, this series, Vol. 65, p. 499. 3 K. K. Hamilton, K, Lee, and P. W, Doetsch, this volume [3]. 4 p. W. Doetsch, G. L. Chan, and W. A. Haseltine, Nucleic Acids Res. 13, 3285 (1985).
104
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[10]
If phenol extractions are employed to remove protein at any stage of plasmid DNA preparation, labeling, or processing, it is important to only use fresh phenol preparations (less than 2 months old) that contains 0.5% (w/v) 8-hydroxyquinoline (radical scavenger). Use of phenol without a radical scavenger can introduce oxidative damage into the DNA substrates, resulting in a high background of DNA cleavage observed in the analytical DNA sequencing gels. Two different types of DNA damage substrates are used in the purification of yeast redoxyendonuclease. Heavily UV-irradiated (254 nm, 10,000 J/m 2) DNA fragment 1 contains cytosine and thymine photohydrates which are substrates for yeast redoxyendonuclease. 1 DNA fragment 1 is also damaged with OsO4 to generate substrates containing primarily cis-thymine glycol, another redoxyendonuclease substrate. 5 The generation of UV- and OsOa-damaged substrates is described in detail elsewhere in this volume) Although the generation of UV-damaged substrates is rapid and less labor-intensive than the preparation of OsO4-damaged substrates, it should be borne in mind that such UV-damaged fragments contain a large percentage of base damage products such as cyclobutane pyrimidine dimers and (6-4)pyrimidine-pyrimidone photoproducts which are not redoxyendonuclease substrates, and therefore the resulting assay is less specific. 6 End-labeled DNA substrates containing AP sites (AP-DNA) for use in AP lyase experiments are generated by chemical depurination of 3'-endlabeled DNA fragment 2 and 5' end-labeled DNA fragment 3. To achieve depurination primarily at sites of guanine, 10-20 ng of DNA substrate ( - 2 × 10 6 cpm in 25/xl) is alkylated with 1/.d dimethyl sulfate (DMS) in 200/zl DMS buffer (50 mM sodium cacodylate, pH 8.0, 1 mM EDTA) for 30 sec at room temperature. 7 The DMS reaction is terminated by addition of 50 /xl of DMS stop solution (1.5 M sodium acetate, pH 7.0, 1 M 2-mercaptoethanol), and the DNA is recovered by ethanol precipitation. The alkylated DNA samples are resuspended in 10 mM Tris-HC1, pH 4.5, 1 mM EDTA and incubated at 50 ° for 20 min followed by ethanol precipitation. Procedure. The final reaction volume for yeast redoxyendonuclease assays is variable (20-100/zl) depending on the yeast redoxyendonuclease purification step. Early steps in the purification procedure require larger enzyme sample volumes to provide detectable levels of activity. End5 K. Burton and W. T. Riley, Biochem. J. 98, 70 (1966). 6 W. Harm, "Biological Effects of Ultraviolet Radiation." Cambridge Univ. Press, London, 1980. 7 j. Lenz, S. A. Okenquist, J. E. LoSardo, K. K. Hamilton, and P. W. Doetsch, Proc. Natl. Acad. Sci. U.S.A. 87, 3396 (1990).
[10]
YEAST REDOXYENDONUCLEASE
105
labeled DNA fragments I-3 (3-10 × 10 4 cpm, 4-8 ng DNA) containing either thymine glycol or UV-induced pyrimidine photohydrates are incubated with 2-30/zl of a particular column fraction or enzyme preparation. A 1/10th volume of 10 × reaction buffer (150 mM KH2PO4, pH 6.8,400 mM KC1, 100 mM EDTA, 100 mM 2-mercaptoethanol) and a variable volume of deionized water are added prior to enzyme addition to bring the reaction mixture to the appropriate final volume. Reactions are carried out at 37° for 60 min followed by extraction twice with phenol-chloroform-isoamyl alcohol (25 : 24 : 1, v/v/v). DNA samples are ethanol precipitated, dried, and resuspended in 10/zl of deionized water, and the radioactivity is determined by Cerenkov counting. The appropriate volumes of resuspended DNA sample corresponding to the desired amount of radioactivity are transferred to fresh, siliconized bullet tubes, dried, resuspended in 2/zl of loading dye (80% (v/v) deionized formamide, 0.2% (w/v) bromphenol blue, 0.2% (w/v) xylene cyanol), loaded onto denaturing, 15% polyacrylamide DNA sequencing gels, and subjected to electrophoresis (750-1500 V, 5-12 hr) followed by autoradiography. Purification Procedure Purification steps employing fast protein liquid chromatography (FPLC) are carried out at room temperature. All other purification steps are carried out at 00-4 °. Preparation of Yeast Cell Crude Extract. Yeast cells (1.7 kg) from dried, active, commercial bakers' yeast (Universal Foods Corp., Milwaukee, WI) are subdivided into 170-g batches and soaked for 30-45 min in 150 ml of buffer Y (50 mM KHzPO 4 , pH 8.7, 0.1 mM phenylmethylsulfonyl fluoride, 0.7/~g/ml pepstatin A, 0.1/~g/ml aprotinin, 0.7/~g/ml leupeptin, 1% dimethyl sulfoxide) plus 0.2 M NaCI. The resulting cell paste is further divided into three equal portions, and each portion is mixed with 210 g of glass beads (0.5 mm diameter) and 50 ml buffer Y plus 0.12 M NaCI and loaded into a Bead Beater (Biospec Products, Bartlesville, OK). The cell paste is subjected to four spin bursts of 1 min with a 4-rain holding on an ice bath between each burst. The glass beads are removed from the homogenates, the subdivided preparations are pooled, and 5 M NaC1 is added to bring the final NaCI concentration to 0.7 M. The samples are then held on ice for 30 min to disrupt the chromatin.l Buffer Y is added to reduce the NaCI concentration to 0.3 M, and the resulting crude homogenate is centrifuged at 100,000 g for 30 min at 4°. The high-speed supernatants are pooled and constitute fraction I (16 liters). Ammonium Sulfate Precipitation and DEAE-Cellulose Chromatography. Fraction I is gradually brought to 65% saturation with ammonium
106
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[10]
sulfate, then gently stirred on an ice bath for 15 hr. Small volumes of 2 M NHaOH are added periodically during the ammonium sulfate precipitation step to maintain a pH of 8.0. The material is centrifuged (11,000 g), and the resulting pellets are resuspended in buffer A (10 mM Tris-HCl, pH 8.0, 10 mM 2-mercaptoethanol, 10 mM EDTA) plus 0.3 M NaCI in a total volume of 2 liters and extensively dialyzed against buffer A plus 0.3 M NaCI. The resulting material (fraction II, 2 liters) is applied to a large DEAE-cellulose (Whatman, Clifton, N J, DE-52) column (906 ml) equilibrated in buffer A plus 0.3 M NaCI for the removal of nucleic acids and some proteins. Yeast redoxyendonuclease does not bind to the column and is eluted as a broad peak of activity with the same buffer in the flow-through fractions. The redoxyendonuclease-containing fractions are pooled (3 liters, fraction III). Phosphocellulose Chromatography. Fraction III is dialyzed extensively against buffer A plus 50 mM NaCI and applied to a phosphocellulose (Whatman P11) column (553 ml) equilibrated with the same buffer. The column is washed with buffer A plus 50 mM NaCI until the absorbance at 280 nm (A280) of the column wash is near zero. The column is eluted with a 0.05-1.0 M NaCI linear gradient (1000 ml/1000 ml) in buffer A (20-ml fractions), and redoxyendonuclease activity is monitored on UVdamaged DNA substrates using the activity assay described above. The enzyme elutes in fractions 41-55 (-0.2-0.4 M NaC1), which are pooled, dialyzed against buffer B (15 mM KH2PO4, pH 6.8, I0 mM EDTA, I0 mM 2-mercaptoethanol) plus 50 mM NaC1, and concentrated to a volume of 186 ml in an Amicon (Danvers, MA) concentrator (fraction IV). Preparative Mono S Fast Protein Liquid Chromatography. Fraction IV is divided into six 28.6-ml portions (300 mg each), and each portion is separately fractionated on a preparative Mono S FPLC column (Pharmacia, Piscataway, NJ, HR10/10) equilibrated in buffer B plus 50 mM NaCI. The column is washed with buffer B until the A280 values of the wash fractions are near zero, and then a 0-1 M NaCI linear gradient (40 ml/40 ml) in buffer B is applied (0.5-ml fractions collected, Fig. 1). The redoxyendonuclease peak (fractions 55-60, 0.3-0.4 M NaC1) obtained from these steps is pooled together with Mono S peak fractions from a similar purification (Table I) and concentrated (15 ml, fraction V).
Sephacryl S-200 Chromatography and Analytical Mono S Fast Protein Liquid Chromatography. Fraction V is brought to 0.5 M NaCI and applied to a 2.5 x 120 cm Sephacryl S-200 column (Pharmacia) equilibrated in buffer B plus 0.5 M NaCI and eluted with the same buffer (2-ml fractions). Redoxyendonuclease elutes in fractions 134-155, which are pooled, concentrated, and dialyzed against buffer B plus 0.15 M NaCI (fraction VI). Fraction VI is applied to an analytical Mono S FPLC column (Pharmacia,
A
m1.0
1.00~
.90 -"0.8
.80 .70 ,60
,,~ ~
-, . f
4. " ~
---0.6
:5 --0.4
.30 .20 t
I'
~_
--0.2
.10 0
B
I li ll~lll]lll
GC GATC
I I II I i|1 I II I II IIII
I I
--0
/ .ac i 15 50 52 54 55 56 58 60 62 64 65 68 70
3 FIG. 1, Preparative Mono S FPLC purification step of yeast redoxyendonuclease. An aliquot of fraction IV (28.6 ml, 300 rag) was fractionated on a preparative Mono S FPLC column as described in the text. (A) FPLC elution profile [solid line, absorbance units (A.U.) at 280 nm] versus fraction number following elution with the NaCI gradient (dashed line). (B) Gel autoradiograrn of redoxyendonuclease activity present in corresponding fractions (2-/~1 aliquots) on 3'-end-labeled, UV-damaged substrates (DNA fragment 1) occurring at sites of C (cytosine photohydrates) and G. Base-specific DNA sequencing reactions were run in the left-hand lanes.
108
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[10]
TABLE I PURIFICATION OF Saccharomyces cerevisiae REDOXYENDONUCLEASE
Fraction
Step
I II
Crude extract
III IV V VI VII VIII
DEAE-cellulose Phosphocellulose Mono S FPLC d Sephacryl S-200 Mono S FPLC e Superose 6/12 FPLC
A m m o n i u m sulfate
Protein ~
Fraction
Step
N.D. b 45.1 g
Ic II c
32.5 g 1.95 g 100 mg 17 mg 3.5 mg 0.1 mg
III c IV c
Crude extract A m m o n i u m sulfate precipitate DEAE-cellulose Phosphocellulose
precipitate
Protein a 512 g 370 g 275 g 6.74 g
a Protein determination by Bio-Rad (Richmond, CA) dye binding assay with bovine
plasma y-globulin standard. b Not determined. c Separate purification starting from 4.76 kg of dried yeast cells. Peak redoxyendonuclease fractions from the phosphocellulose chromatography steps (IV and IV0 were run over the same preparative Mono S FPLC column, and peak fractions were pooled together and subjected to the final three purification steps. d Preparative Mono S FPLC. e Analytical Mono S FPLC.
HR 5/5) equilibrated in buffer B plus 0.15 M NaC1 and washed with the same buffer until the A2so values of the wash fractions are near zero. The column is eluted at room temperature with a 0.15-1 M NaC1 linear gradient (15 ml/15 ml) in buffer B (0.2-ml fractions), and the peak of redoxyendonuclease activity (fractions 106-111, 0.3-0.4 M NaCI) is pooled, concentrated, and brought to 0.5 M NaCI (3 ml, fraction VII). Superose 6-Superose 12 Fast Protein Liquid Chromatography. Fraction VII is divided into three portions, and each portion is applied and chromatographed in separate runs on Superose 6 (Pharmacia, HR 10/30) and Superose 12 (Pharmacia, HR 10/30) columns connected in tandem and eluted with buffer B plus 0.5 M NaCI (0.2-ml fractions). Aliquots of fractions 148-172 are electrophoresed in sodium dodecyl sulfate-15% polyacrylamide gels (SDS-PAGE) and silver stained for protein analysis as described 8'9 as well as assayed directly for redoxyendonuclease activity (Fig. 2). The presence of a band corresponding to a well-resolved protein of about 40 kDa coincided exactly with the presence of redoxyendonuclease activity. The redoxyendonuclease peak fractions (fractions 153-158) are 8 U. K. Laemmli, Nature (London) 227, 680 (1970). 9 C. R. Merril, D. Goldman, S. A. Sedman, and M. H. Ebert, Science 211, 1437 (1987).
[10]
YEAST REDOXYENDONUCLEASE
109
o
A
G c ,GA
TC
~__~--
~
'P" , r - ,¢,- ,i-- 'P- ,t-"
T
A
c
::?N
r c
B
M
150
152
153
154
155
156
157
158
159
160
161
162
163
164
80~ 49~
32~ 27~
FIG. 2. Superose 6-Superose 12 FPLC purification step of yeast redoxyendonuclease. Fraction VII was further purified on Superose 6 and Superose 12 FPLC columns connected in tandem as described in the text. (A) Gel autoradiogram of redoxyendonuclease cleavage of 3'-end-labeled, OsO4-damaged substrates (DNA fragment 1). Aliquots (5/zl) of the indicated fractions (149-171) were assayed for cleavage at sites of thymine glycol and compared to endo II1 cleavage (Endo III lanes) or no treatment (NT lanes) of the OsO4-damaged (TG lanes) or undamaged ( - lanes) substrates. Base-specific DNA sequencing reactions were run in the left-hand lanes. Arrows indicate redoxyendonuclease cleavage at sites of thymine glycol. (B) Aliquots (5/,1) of the indicated column fractions (150-164) were analyzed for protein content in a sodium dodecyl sulfate-15% polyacrylamide gel followed by silver staining as described in the text. Small arrow indicates band corresponding to redoxyendonuclease. Lane M contains prestained (Bio-Rad) molecular size markers: phosphorylase b, 106 kDa; bovine serum albumin, 80 kDa; ovalbumin, 49.5 kDa; carbonate dehydratase, 32.5 kDa; soybean trypsin inhibitor, 27.5 kDa (sizes given correspond to protein-dye complex size).
110
OXIDATIVE DAMAGETO DNA AND DNA REPAIR
[10]
pooled and concentrated (0.2 ml, fraction VIII). The entire purification procedure is summarized in Table I. General Properties
Enzyme Storage and Dilution. Any of the redoxyendonuclease preparations subsequent to the phosphocellulose chromatography step can be stored at - 2 0 ° in 50% (v/v) glycerol in buffer B. The preparations are active for at least 1 year. Redoxyendonuclease is diluted before use with buffer B. Purity and Physical Properties. Fraction VIII is not a homogeneous preparation of yeast redoxyendonuclease and contains at least six other proteins (Fig. 2B). The preparation has no detectable DNA cleavage activity on undamaged, duplex DNA. Yeast redoxyendonuclease appears to be a monomeric protein of about 40 kDa (SDS-PAGE), in good agreement with the size estimate of 38-42 kDa based on Sephadex G-100 (Superfine) chromatography) Yeast redoxyendonuclease has no divalent cation or other known cofactor requirements and is fully active in 10 mM EDTA. The enzyme is active throughout a broad salt concentration range (10-200 mM NaCI). Substrate Specificity and Mode of Action
Base Damages Recognized. Experiments with end-labeled restriction fragments damaged with OsO4 or UV light show that yeast redoxyendonuclease recognizes and cleaves duplex DNA at sites of thymine glycol (5,6-dihydroxy-5,6-dihydrothymine) and cytosine photohydrates (6-hydroxy-5,6-dihydrocytosine).1In addition, we have prepared a number of synthetic oligonucleotides containing specific, individual types of oxidative DNA base damages and have found that yeast redoxyendonuclease recognizes and cleaves substrates containing 5,6-dihydrothymine, 5,6dihydrouridine, and 8-oxoguanine.l° N-Glycosylase Activity. To assay N-glycosylase activity, a duplex DNA substrate is generated by random primer labeling 11 of linearized pUC19 DNA with [methyl-3H]TTP (specific activity 40-60 Ci/mmol) and subsequent treatment with 5% OsO4 at 70° for 20 min, then processed as described for OsO4-damaged DNA substrates elsewhere in this volume) Approximately 2 × 105 cpm (3.2 ng) of the resulting product containing l0 L. Augeri and P. W. Doetsch, unpublishedresults (1993). 11S. Tabor and K. Struhl, in "Current Protocols in Molecular Biology"(F. M. Ausubel, R. Brent, R. E. Kinston, D. D. Moore, J. G. Seidman,J. A. Smith, and K. Struhl, eds.), p. 3.5.9. Greene PublishingAssociates and Wiley(Interscience), New York, 1989.
[10]
YEAST REDOXYENDONUCLEASE
111
3000 -
TG 2500,
1
2000
E
1500
o 1o0o 5O0
0 0
5
10
Retention
15
Time
20
25
30
(min)
FIG. 3. HPLC analysis of thymine glycol released from OsO4-damaged [methyl-3H]thymine-labeled DNA. Ethanol-soluble products released from thymine glycol-containing DNA substrates were incubated with redoxyendonuclease (0) or buffer B alone (x) and undamaged DNA substrates incubated with redoxyendonuclease (A) or buffer B alone (D). Released products were analyzed on a Cl8 reversed-phase HPLC column (Adsorbosphere HS, Alltech Associates, Deerfield, IL) eluted with water (1 ml/min flow rate). The radioactivity (cpm) in each fraction (1 ml) was determined by liquid scintillation counting. Arrow TG indicates the retention time of authentic, unlabeled thymine glycol marker.
[methyl-3H]thymine glycol (or an equivalent amount of undamaged DNA) is incubated with yeast redoxyendonuclease (purified through the Sephacryl S-200 step) in 15 mM KH2PO4, pH 6.8, 40 mM KCI, 10 mM EDTA, 10 mM 2-mercaptoethanol for 30 min at 37° (standard conditions), and the ethanol-soluble supernatants are analyzed by reversed-phase highperformance liquid chromatography (HPLC) together with authentic, unlabeled (marker) thymine glycol (Fig. 3). Yeast redoxyendonuclease causes the release of thymine glycol from the OsOn-damaged DNA substrates, indicating that the enzyme removes oxidative DNA damage via an N-glycosylase activity and in this regard is analogous to endo III (Fig. 5). Apurinic/Apyrimidinic Lyase Activity. Yeast redoxyendonuclease also cleaves DNA substrates containing AP sites in a manner similar to endo III and other AP lyases via fl-elimination.12 Depurinated 3'-end-labeled DNA fragment 2 or depurinated 5'-end-labeled fragment 3 is incubated with yeast redoxyendonuclease (fraction V) under standard conditions, and the enzyme-generated DNA strand scission products are analyzed on DNA sequencing gels (Fig. 4). The DNA cleavage products are compared 12 V. Bailly and W. G. Verly, Biochem. J. 242, 565 (1987).
112
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
,,,
~
r~
A CG -IC T AGAp-
AP
LU
-
AP
o
~
-
AP
-
AP
[10]
"O c -
AP
60 ¸
,J
FIG. 4. AP lyase activity of yeast redoxyendonuclease. Base numbering is from 3'- or 5'-end-labeled termini. Base-specific DNA sequencing reactions were run in the left-hand lanes. (A) Gel autoradiogram of 3'-end-labeled AP-DNA substrate (AP lanes) or undamaged substrate ( - lanes) treated with hot alkali to cleave DNA quantitatively at sites of depurinated guanines (HA lanes), yeast pyrimidine dimer endonuclease (YPDE lanes), T4 endonuclease V (T4 Endo V lanes), endonuclease IV (Endo IV lanes), yeast redoxyendonuclease (YRE lanes), and endonuclease III (Endo III lanes). (B) Gel autoradiogram of 5'-end-labeled APDNA substrate treated with hot alkali or with the same enzymes as in (A). Large arrow indicates DNA cleavage event at AP site G34 depicted by the following small arrows: a, hot alkali-generated DNA cleavage product containing 3'-terminal phosphoryl group; b, AP lyase-generated DNA cleavage products containing 3'-terminal modified deoxyribose; c, AP endonuclease-generated DNA cleavage product containing 3'-terminal hydroxyl group. [Autoradiogram in (B) reprinted with permission from Hamilton et al. Js]
[10]
YEAST REDOXYENDONUCLEASE O
B
,,,
GaACT
C
,.
113
u n
= -
O
,,,
FIG. 4. (continued)
to the DNA cleavage products generated by incubation of the same substrate with several well-characterized prokaryotic enzymes that are either hydrolytic 5' AP endonucleases (Escherichia coli endonuclease IV) or 3' AP lyases (T4 endonuclease V, endo III, and yeast pyrimidine dimer endonuclease). 12-15 All of the AP lyases cleave 3'-end-labeled A P - D N A substrates in an identical manner, producing DNA scission products containing 3' terminal phosphoryl groups (Fig. 4A). When 5'-end-labeled AP DNA substrates are digested with the same battery of enzymes, all of the AP lyases (yeast pyrimidine dimer endonuclease, T4 endonuclease V, yeast redoxyendonuclease, and endo III) produce DNA scission products with identical electrophoretic mobilities (b, Fig. 4B), indicating the presence of a 13 V. Bailly and W. G. Verly, Biochem. J. 259, 761 (1989). 14 M. A. Manoharan, S. C. Mazumder, J. A. Ransom, J. A. Gerlt, and P. H. Bolton, J. Am. Chem. Soc. U0, 2690 (1988). J5 K. K. Hamilton, P. M. H. Kim, and P. W. Doetsch, Nature (London) 356, 725 (1992).
114
O X I D A T I V E D A M A G E TO D N A
AND DNA
REPAIR
[10]
o
z
0
o
;>.
E..~o
=.~ z _
o
=at. - o - - % - " " =
-,-o
o
-,-
o.
II
o - - - o - - - o.='%~'=
,0
~..=
o--~-o--
,
~.=-"~
.,.
o,.,~
~ o~
~~ " 0 c~ -+"~ ;:> o ~ ~ .~,
'
~)
.o==° -~~
,..
N=E_.~
0
~
=a
"~
;-s<~
.= o i~ O~
''z
0
*" > " " ~ "I"
[n "
:z:
o ii
il <\ ~-I ~:~L..n__^
" -- o--
q: "~''z
0
'~"
0
-
0
o
.,.j II \ z~L.o~~ o~'-"O--m--O--q.~-'=
,,
T-"---
!
~
,
~
~--. 0 |
.t-
,0
~
~
-
0
~"~
~ ~ 0"~ ' ' z
=, II
~.~
~,~==
0
o
° ~ °
i~
0
f~,=
~
=
~
0
~ o'," ~ u = "~._~ c~
~ o
'
z
e-L
o
/
~>~ ~ "=" ~
o
=.= ='r
o
.,.
II
0--=--- O - ,
~=~
o" _~I z
0,,.~
~=-=
~=
-
-~'~
.
[1 1]
SINGLET OXYGEN AND SHUTTLE VECTORS
115
3'-terminal a,fl-unsaturated aldehyde (4-hydroxy-trans-2-pentenal). Such DNA scission products have electrophoretic mobilities different from those produced by hot alkali cleavage (a, 3'-phosphoryl, Fig. 4B) or endonuclease IV cleavage (c, 3'-hydroxyl, Fig. 4B) of the same A P - D N A substrates. Yeast redoxyendonuclease generates DNA scission products identical to those produced by other AP lyases, placing this enzyme into the same category of DNA base excision repair enzymes as endo III with respect to the N-glycosylase/AP lyase mode of action (Fig. 5). Acknowledgments We thank the followingindividuals for gifts of enzymes: R. Cunningham (endonuclease III), R. S. Lloyd (T4 endonuclease V), and B. Demple (endonuclease IV). This work was supported by Grants CA42607,CA01441,and Training Grant T32GM08367from the National Institutes of Health and Grant NP-806from the American Cancer Society.
[1 1] S h u t t l e V e c t o r b e t w e e n P r o k a r y o t e s a n d E u k a r y o t e s for A s s a y i n g S i n g l e t O x y g e n - I n d u c e d D N A D a m a g e and Mutagenicity
By
CARLOS FREDERICO MARTINS M E N C K
Introduction Investigations on the in oioo consequences of the interactions of singlet oxygen (102) with genetic material face some difficulties owing to the high reactivity and short lifetime of singlet oxygen. To overcome these difficulties methods are needed in which DNA is treated in vitro with 102 and then introduced into living cells in order to analyze how the cells handle this modified genetic molecule and the subsequent outcomes. For mammalian cells, special plasmids, the shuttle vectors, can be used as exogenous probes to obtain data on the mutagenicity of lO2-induced DNA damage. Shuttle vectors are plasmid molecules which can replicate in both bacteria and mammalian cells.l The basic idea is to submit the vector to IO2 treatment, to analyze the damaging effects, and to introduce it into mammalian cells. The DNA lesions are processed, that is, replicated and repaired in the eukaryotic environment, causing mutations. Finally, the 1 A. Sarasin, J. Photochem. Photobiol., B 3, 143 (1989).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproductionin any form reserved.
[1 1]
SINGLET OXYGEN AND SHUTTLE VECTORS
115
3'-terminal a,fl-unsaturated aldehyde (4-hydroxy-trans-2-pentenal). Such DNA scission products have electrophoretic mobilities different from those produced by hot alkali cleavage (a, 3'-phosphoryl, Fig. 4B) or endonuclease IV cleavage (c, 3'-hydroxyl, Fig. 4B) of the same A P - D N A substrates. Yeast redoxyendonuclease generates DNA scission products identical to those produced by other AP lyases, placing this enzyme into the same category of DNA base excision repair enzymes as endo III with respect to the N-glycosylase/AP lyase mode of action (Fig. 5). Acknowledgments We thank the followingindividuals for gifts of enzymes: R. Cunningham (endonuclease III), R. S. Lloyd (T4 endonuclease V), and B. Demple (endonuclease IV). This work was supported by Grants CA42607,CA01441,and Training Grant T32GM08367from the National Institutes of Health and Grant NP-806from the American Cancer Society.
[1 1] S h u t t l e V e c t o r b e t w e e n P r o k a r y o t e s a n d E u k a r y o t e s for A s s a y i n g S i n g l e t O x y g e n - I n d u c e d D N A D a m a g e and Mutagenicity
By
CARLOS FREDERICO MARTINS M E N C K
Introduction Investigations on the in oioo consequences of the interactions of singlet oxygen (102) with genetic material face some difficulties owing to the high reactivity and short lifetime of singlet oxygen. To overcome these difficulties methods are needed in which DNA is treated in vitro with 102 and then introduced into living cells in order to analyze how the cells handle this modified genetic molecule and the subsequent outcomes. For mammalian cells, special plasmids, the shuttle vectors, can be used as exogenous probes to obtain data on the mutagenicity of lO2-induced DNA damage. Shuttle vectors are plasmid molecules which can replicate in both bacteria and mammalian cells.l The basic idea is to submit the vector to IO2 treatment, to analyze the damaging effects, and to introduce it into mammalian cells. The DNA lesions are processed, that is, replicated and repaired in the eukaryotic environment, causing mutations. Finally, the 1 A. Sarasin, J. Photochem. Photobiol., B 3, 143 (1989).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproductionin any form reserved.
116
OXIDATIVE DAMAGETO DNA AND DNA REPAIR
[11]
vectors are rescued back into E s c h e r i c h i a coli, where mutations in a target gene can be screened and examined at the sequence level. Specific shuttle vectors have been used for the investigation of 10 2 effects on genetic material. 2'3 These vectors contain the origin of replication of pBR322 and the c a t (chloramphenicol acetyltransferase) gene for amplification and selection (chloramphenicol resistance) in E. coli. They carry the SV40 origin of replication, so when introduced in monkey COS7 cells, which supply the T-antigen in trans, 4 the plasmid replicates episomally. They also have the SV40 late genes, responsible for production of capsid proteins, and thus the plasmids are packaged and amplified as pseudoviruses. As the target gene for mutation studies, the vectors have the tRNA s u p F gene of E. coli. One of the vectors used has a particular and interesting feature. It also carries the replication origin from the bacteriophage fl, which allows the plasmid to enter the phage replication mode in permissive bacteria after infection with a helper phage, generating single-stranded DNA (ssDNA). Therefore, the use of these shuttle vectors also enables the comparison of IO2 action on ssDNA and double-stranded DNA (dsDNA) structures. The shuttle vector approach is straightforward (Fig. 1). Plasmid Treatment and DNA Damage Assays Shuttle vectors are exposed to IO2generated by thermal decomposition of the water-soluble endoperoxide of disodium 3,3'-(1,4-naphthylidene) dipropionate (NDPO2). This chemical method is ideally suited for these studies because it is simple and produces IO2 as the only reactive molecule in solution. 5 Other methods, such as the use of photosensitizers, may generate other reactive products under some conditions, making interpretation of results uncertain. DNA samples (2/zg/200/A) are incubated with NDPO2, at a concentration up to 100 mM, in 50 mM sodium phosphate buffer in D20, pD 7.4. The reaction proceeds for 90 min at 37°, yielding 3,3'-(1,4-naphthylidene) dipropionate (NDP) and molecular oxygen, half in the ground state and half in the excited singlet state, which reacts with DNA. During the first 30 min of incubation, the samples are vortexed every 2 min in order to homogenize the IO2 produced in the solution. After treatment, DNA may 2 p. Di Mascio, C. F. M. Menck, R. G. Nigro, A. Sarasin, and H. Sies, Photochem. Photobiol. 51, 293 (1990). 3D. T. Ribeiro, C. Madzak, A. Sarasin, P. Di Mascio, H. Sies, and C. F. M. Menck, Photochem. Photobiol. 55, 39 (1992). 4 y. Gluzman, Cell (Cambridge, Mass.) 23, 175 (1981). s p. Di Mascio and H. Sies, J. Am. Chem. Soc. 111, 2909 (1989).
[11]
SINGLET OXYGEN AND SHUTTLE VECTORS
117
Virus
Sitesthat b l o c k DNAbreaks DNAsynthesis DNAdaiageassays ssDNA ~.~ dsDNA
NDPO2 treatment
Mutant sequencing ~
~nfection
q~:~o:~3 extraction E. co/i
~ transformation ~ i "~ ~,°~(~ supfmutants screening
FIG. 1. Schematicrepresentation of the general use o£ shuttle vectors for assayingthe DNA-damaging and mutageni¢effects of IO2 in mammaliancells. be purified by the addition of 0.6 volume of a solution of 20% polyethylene glycol (PEG) and 2.5 M NaCI to the samples, which are kept for 1 hr in ice and then centrifuged (13,000 rpm, 20 min). The precipitated DNA is resuspended for further analysis. The DNA can be checked for detectable lesions using any of the procedures described in the literature, such as chromatographic6and enzymatic assays.7 For 102-treated shuttle vectors, single- and double-stranded breaks can be detected by conventional electrophoresis in 0.7% agarose gels, in Tris-borate buffer, pH 7.5. 8 After migration the DNA is stained by ethidium bromide (0.5 /~g/ml) and visualized by fluorescence in an UV (330 nm) transilluminator. This methodology allows discrimination between unbroken plasmid molecules (supercoiled structure) and those having single-stranded (open relaxed circles) or double-stranded (linear) breaks, and the relative number of any of the three forms can be quantified by scanning the gel pictures with a densitometer. For ssDNA, agarose 6 M. K. Shigenaga, J. W. Park, K. C. Cundy, C. J. Gimeno, and B. N. Ames, this series, Vol. 186, p. 521. 7 S. Boiteux, E. Gajewski, J. Laval, and M. Dizdaroglu, Biochemistry 31, 106 (1992). 8 j. Sambrook, E. F. Fritsch, and T. Maniatis, 2nd ed. "Molecular Cloning: A Laboratory Manual." Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1989.
118
OXIDATIVE DAMAGETO DNA AND DNA REPAIR
[11]
gel electrophoresis also separates covalently closed and whole linear molecules (resulting from a single break in the DNA). In this case, degradation of the ssDNA can also be observed, as a result of at least two nonspecific breaks in the same molecule. This generates single-stranded linear fragments with different DNA sizes which migrate to various positions and, consequently, are dispersed in the gel. The relative number of covalently closed and whole linear molecules can also be obtained as for dsDNA. In both cases, the number of breaks can be calculated based on the Poisson distribution. The results obtained after treatment of shuttle vectors with NDPO2 indicate that 102 induces single- and double-strand breaks in dsDNA. 2 It also reacts with ssDNA, yielding breaks in this molecule. Quantification of such lesions revealed that more breaks are induced in ssDNA than in dsDNA. 3 A second approach was employed in order to examine DNA damage in 102-treated shuttle vectors, namely, the analysis of sites that block in vitro DNA synthesis by DNA polymerases. This method was first described by Moore and Strauss, 9 who found that DNA lesions induced by UV-irradiation block DNA synthesis by the DNA polymerase I from E. coli. The method is simple and uses conventional sequencing techniques. 102-treated and denatured DNA (1/zg) is annealed with a specific primer (30 ng), which is complementary to the vector in a position next to the sequence to be investigated. Annealing is performed in 10/zl of 40 mM Tris-HCl, pH 7.5, 20 mM MgCI2,and 50 mM NaC1, at 65° for 5 min, followed by gentle cooling to room temperature. To 5/~1 of annealed DNA are added standard buffer [26.5 mM Tris-HCl, pH 7.5, 13.3 mM MgCI2, 16.6 mM NaCI, and 6.7 mM dithiothreitol (DTT)] and 0.3 /zM of each [35S]dATP (1000 Ci/mmol), dGTP, dCTP, and dTTPs and DNA polymerase (final volume of 7.5/zl). A mixture (1/~1) of 83/zM of the four deoxynucleoside triphosphates (dNTPs) is added for DNA elongation. Both time and temperature vary for labeling and elongation reactions, based on the DNA polymerase used. The reactions are terminated by the addition of 0.6 volume of stop solution (95% formamide, 20 mM EDTA, 0.05% bromphenol blue, and 0.05% xylene cyanol). The products of the polymerization reactions are heat denatured and then loaded and electrophoresed on standard high-resolution denaturing sequencing gels (8% polyacrylamide, 7 M urea). After migration the gels are dried and subjected to autoradiography. Sites on the template DNA containing lesions that block DNA synthesis appear in the autoradiogram as bands, with the specific location being identified by comparison with the regular DNA sequence in the same gel. 9 p. D. Moore and B. S. Strauss, Nature (London) 278, 664 (1979).
[11]
SINGLET OXYGEN AND SHUTTLE VECTORS
119
Although the nature of the blocking lesions cannot be determined by this kind of experiment, at least two important results are obtained: (i) the identified lesions correspond to biologically significant damages that interrupt DNA synthesis, and (ii) the final result is a DNA damage spectrum induced by 102 in a specific DNA sequence. In the case of shuttle vectors, the DNA sequence investigated corresponds to the supF gene, which is also used to obtain the mutation spectrum. Thus, one can correlate the mutagenicity of a specific region of the supF gene with the presence of lesions in the same position. The data obtained 1° show that, depending on the enzyme used, DNA synthesis is interrupted either opposite or one nucleotide 3' to the deoxyguanosine positions on the template. This suggests that the blocking lesions induced by IO2 are specifically located at deoxyguanosine residues, confirming the high and specific reactivity of 102 with guanine. Moreover, there are marked variations among the different deoxyguanosines of the supF gene in the efficiency of blocking DNA synthesis, which may reflect the distribution of blocking lesions in the sequence. In general, no particular sequence context showed specific susceptibility to the attack of 1 0 2 .
Shuttling DNA into Mammalian Cells The IO2-damaged vectors can be introduced in monkey COS7 cells by any of a number of established transfection procedures. The DEAE-dextran technique n works well for this cell line. Basically, a sterile solution is prepared in Dulbecco's modified Eagle's medium (without serum or sodium bicarbonate) containing 2/.~g/ml of DNA, 50 mM Tris-HCl, pH 7.4, and 0.75 mg/ml of DEAE-dextran. A 90-mm cell culture dish with approximately 10 6 cells is rinsed twice with phosphate-buffered saline (PBS), and 0.5 ml of the DNA solution is carefully spread over the cells. The cells are incubated for 30 min at room temperature and then rinsed twice with PBS. Culture proceeds for 3 to 7 days in order to amplify the vector. The cells are harvested, and low molecular weight DNA is extracted by the small-scale alkaline lysis procedure. An alkaline sodium dodecyl sulfate (SDS) solution (450/zl of 2.7% sucrose, 1.7% Triton X-100, 16 mM EDTA, 16 mM Tris-HCl, 0.67 N NaOH, and 0.7% SDS) is applied over the cell monolayer. The culture dish is rocked gently, and the cell lysate is scraped into an Eppendorf tube. Ammonium acetate (7.5 M, pH 7.8, ~0 D. T. Ribeiro, F. Bourre, A. Sarasin, P. Di Mascio, and C. F. M. Menck, Nucleic Acids Res. 20, 2465 (1992). u j. H. Wilson, Virology 91, 380 (1978).
120
OXIDATIVE DAMAGETO DNA AND DNA REPAIR
[11]
225/~1) is mixed into the lysate; the tube is then cooled on ice for 10 min and centrifuged (13,000 rpm, 30 min). The clear supernatant is transferred to a fresh tube and incubated with RNase I (50/zg/ml) for 15 min, at 45 °. DNA is purified by two phenol-chloroform extractions and sedimented after addition of 0.6 volume of 2-propanol by centrifugation (13,000 rpm, 15 min). The pellet (almost invisible) is rinsed with 70% ethanol and resuspended in 100 /zl of 10 mM Tris-HCl pH 8.0, 1 mM EDTA. The plasmids are ready to transform bacteria in order to check for relative survival and to screen for mutants in the supF gene induced by ~O2 treatment.
Shuttling Back to Escherichia coli and Screening for Mutations Plasmid DNA is rescued in E. coli made competent by the method of Hanahan.12 The transformants are plated on LB medium containing chloramphenicol (34/~g/ml), X-Gal (5-bromo-4-chloroindolyl-fl-D-galactoside, 0.08 mg/ml), and IPTG (isopropyl-fl-D-thiogalactoside, 0.1 raM). The number of colonies recovered reflects the amount of DNA molecules replicated in the monkey cells, and thus plasmid survival can be defined as the ratio of colonies transformed with DNA from cells transfected with ~O2-treated vector to those obtained with untreated vector. Such analysis indicated that lesions induced by 102 in ssDNA strongly inhibit its conversion to the double-stranded replicative form after transfection in mammalian cells. This is in contrast with the high recovery of plasmids from cells transfected with damaged dsDNA. 3 The E. coil strain MBM7070 is used for screening of plasmids containing mutations that inactivate the tRNA supF gene. The product of the supF gene suppresses an amber mutation in the lacZ gene of the bacterial chromosome, leading to the production of an active fl-galactosidase and, thus, metabolization of the indicator dye X-Gal. Consequently, colonies bearing plasmids with a functional supF gene have a bright blue phenotype. Mutations that inactivate the suppressor tRNA lead to the formation of white or light blue colonies. These colonies are screened in a background of blue colonies and restreaked three times for confirmation of the mutant phenotype. The mutation frequency is defined as the ratio of white or light blue colonies to total colonies examined. The experiments with '02treated shuttle vectors have shown that ~O2-induced lesions are highly mutagenic in mammalian cells. The mutagenicity of ~O2 is higher for ssDNA than for dsDNA. However, when comparison of the data obtained 12 D. Hanahan, J. Mol. Biol. 166, 557 (1983).
[11]
SINGLET OXYGEN AND SHUTTLE VECTORS
121
for both DNA structures is based on the number of lesions induced on each molecule, it seems that damages on the dsDNA have a higher mutability, t3 Mutation Analysis The mutated plasmids are amplifed in bacteria and extracted by conventional procedure. 8 The supF locus is then sequenced by the Sanger chain-termination method, using specific primers complementary to the plasmid DNA in positions adjacent to the supF gene. This locus is particularly interesting as a mutation target, since extensive studies have demonstrated that base substitutions at almost any site in the 85 structural base pairs of the tRNA inactivate the supF function, with few silent mutations. 14 For double-stranded shuttle vectors, mostly single and multiple base substitutions were found among the ~O2-induced mutants. The great majority of these point mutations involve G : C base pairs, resulting mainly in G : C to T : A followed by G : C to C : G transversions) s Consistent with these data, the sequence of mutants obtained in experiments employing singlestranded vectors indicated that transversions involving G (G to T and G to C) are the most frequent mutations induced by ~O2 (unpublished results). Similar results were also obtained for the M13 lacZ phage system replicated in E. co[i. 16 Therefore, mutagenesis, mediated by 102-induced DNA damage, is targeted selectively at guanine residues in both prokaryotic and eukaryotic cells. Concluding Remarks Shuttle vector systems, in which DNA is exposed to damaging agents outside of the cell, are being used in order to understand the biological consequences of the interaction of ~O2with DNA. These and other studies provided information showing that lO2-induced DNA damages may interfere with DNA replication, are repaired, and are mutagenic in both prokaryotic and eukaryotic cells. The mechanisms and the enzymology of these processes are just beginning to be understood, especially in mammalian cells. The use of shuttle vectors may continue to contribute to solving these questions by shuttling ~O2-damaged plasmids into cells impaired in the metabolism of damaged DNA, such as those derived from persons 13 H. Sies and C. F. M. Menck, Mutat. 14 K. H. Kraemer and M. M. Seidman, ~5R. Costa de Oliveira, D. T. Ribeiro, Nucleic Acids Res. 20, 4319 (1992). 16 D. Decuyper-Debergh, J. Piette, and
Res. 275, 367 (1992). Mutat. Res. 220, 61 (1989). R. G. Nigro, P. Di Mascio, and C. F. M. Menck,
A. Van de Vorst, EMBO J. 10, 3155 (1987).
122
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[12]
suffering from syndromes like xeroderma pigmentosum, ataxia telangiectasia, Fanconi anemia, Cockayne, and Bloom syndromes. If any of the genes related to these syndromes is involved in the processing of DNA damages, the replication or the mutagenicity of the ~O2-treated vectors may be affected. Not only will such information help in understanding the syndrome itself, but it may also give clues regarding the deleterious role of ~O2 over the genetic material in oivo and how the cells deal with DNA damage induced by this excited molecule. Acknowledgments I a m grateful to Dr. H e l m u t Sies and Dr. Paolo Di Mascio for continuing e n c o u r a g e m e n t during the course of this research. This work was supported by grants from Fundaqfm de A m p a r o ~ P e s q u i s a do E s t a d o de S~o Paulo (FAPESP) and C o n s e l h o Nacional de Desenvolvim e n t o Cientffico e Tecnol6gico (CNPQ), Brazil.
[12] O x i d a t i v e D N A D a m a g e : E n d o n u c l e a s e F i n g e r p r i n t i n g By BERND EPE and JUTTA HEGLER
Introduction Most cells contain a number of repair endonucleases which specifically recognize types of DNA modifications that are induced by reactive oxygen species (hydroxyl radicals, singlet oxygen) and some other agents, for example, UV-irradiation and methylating agents, that apparently constitute a natural hazard for the DNA. These specific repair endonucleases work independently from and in addition to the nonspecific nucleotide excision repair of the cells, which is represented in Escherichia coli by the uvrABC endonucleases. Several specific repair endonucleases have been cloned and well characterized; the substrate specificities of some of them, according to the available data, are summarized in Table I. Several types of both base modifications and AP sites (apurinic/apyrimidinic sites, sites of base loss) are recognized. The spectrum of substrate modifications comprises most of the modifications known to be generated by reactive oxygen species. All the enzymes shown in Table I incise the DNA at the site of a modification, that is, they generate a DNA single-strand break. The enzymes recognizing base modifications have combined glycosylase and AP endonuclease activity: they first remove the modified base from the deoxyribose moiety and then incise the DNA backbone at the AP site generated. Therefore, regular AP sites (i.e., those not oxidized in the sugar moiety) are substrates for all the enzymes. METHODS 1N ENZYMOLOGY, VOL. 234
Copyright © 1994by AcademicPress, Inc. All rights of reproduction in any form reserved.
122
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[12]
suffering from syndromes like xeroderma pigmentosum, ataxia telangiectasia, Fanconi anemia, Cockayne, and Bloom syndromes. If any of the genes related to these syndromes is involved in the processing of DNA damages, the replication or the mutagenicity of the ~O2-treated vectors may be affected. Not only will such information help in understanding the syndrome itself, but it may also give clues regarding the deleterious role of ~O2 over the genetic material in oivo and how the cells deal with DNA damage induced by this excited molecule. Acknowledgments I a m grateful to Dr. H e l m u t Sies and Dr. Paolo Di Mascio for continuing e n c o u r a g e m e n t during the course of this research. This work was supported by grants from Fundaqfm de A m p a r o ~ P e s q u i s a do E s t a d o de S~o Paulo (FAPESP) and C o n s e l h o Nacional de Desenvolvim e n t o Cientffico e Tecnol6gico (CNPQ), Brazil.
[12] O x i d a t i v e D N A D a m a g e : E n d o n u c l e a s e F i n g e r p r i n t i n g By BERND EPE and JUTTA HEGLER
Introduction Most cells contain a number of repair endonucleases which specifically recognize types of DNA modifications that are induced by reactive oxygen species (hydroxyl radicals, singlet oxygen) and some other agents, for example, UV-irradiation and methylating agents, that apparently constitute a natural hazard for the DNA. These specific repair endonucleases work independently from and in addition to the nonspecific nucleotide excision repair of the cells, which is represented in Escherichia coli by the uvrABC endonucleases. Several specific repair endonucleases have been cloned and well characterized; the substrate specificities of some of them, according to the available data, are summarized in Table I. Several types of both base modifications and AP sites (apurinic/apyrimidinic sites, sites of base loss) are recognized. The spectrum of substrate modifications comprises most of the modifications known to be generated by reactive oxygen species. All the enzymes shown in Table I incise the DNA at the site of a modification, that is, they generate a DNA single-strand break. The enzymes recognizing base modifications have combined glycosylase and AP endonuclease activity: they first remove the modified base from the deoxyribose moiety and then incise the DNA backbone at the AP site generated. Therefore, regular AP sites (i.e., those not oxidized in the sugar moiety) are substrates for all the enzymes. METHODS 1N ENZYMOLOGY, VOL. 234
Copyright © 1994by AcademicPress, Inc. All rights of reproduction in any form reserved.
[12]
ENDONUCLEASE FINGERPRINTING
+ .~,.~
'q.
123
,,.:,
~4
,-,ca .-a te~ o x ul <
~g
,<
< Z
O
.> % < N
©
'D + + + + + + , ~
O
u5 i;4
,..4
g
o',
°
N O
O
>, I
.1.<
I
I
I
+
+
~ e,I
O
.4
Z
e,i
[.-. ~u ul M < I.-,
+
+
+
+
+
+ EL
¢}
,< al .1
<
Z ~
Z O
.
.,.--
dd
Z
~2
N ~'~
~ ~
,.I'~
e.~ e-
¢.a 0 e-,
g
<
~
2
,-- v
O
I ~
~.
'~
,
¢a.1 a= ~a~
tM M
•.-
o '1
~
g ~ ,.
Eo
..- . : .<
o ~a., m
t-- [-
124
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[12]
DNA strand breaks can be detected with high sensitivity and quantified by a variety of techniques (see below), and the same methods can be used to quantify the breaks created by repair endonucleases at endonucleasesensitive sites. If several repair endonucleases are used to quantify DNA modifications, DNA damage profiles are obtained, which indicate the relative frequencies of various types of modifications. The number of single-strand breaks generated directly by the damaging agents is obtained in all cases from experiments without any repair endonuclease. As different damaging agents (e.g., hydroxyl radicals and singlet oxygen) give rise to different DNA modifications or at least to different ratios of common modifications, a DNA damage profile can serve as a fingerprint of the ultimate DNA-damaging agent. Fingerprints obtained from DNA that has been modified under cell-free conditions, for which the ultimate DNA-damaging species is known, can be compared with those obtained from cellular DNA and thus allow identification of the species and mechanism ultimately responsible for the cellular DNA damage. Here we describe methods of obtaining cell-free and cellular DNA damage profiles. Two types of assays are used to quantify strand breaks and the incisions of repair endonucleases: a relaxation assay with supercoiled DNA which is suitable for cell-free DNA, bacterial plasmid DNA, and mitochondrial DNA (mtDNA), and a modified alkaline elution assay which allows analysis of damage in chromosomal DNA of cultured mammalian cells. Endonuclease Preparations Only a few repair endonucleases, for example, exonuclease III and T4 UV endonuclease, are commercially available. The purification of several repair endonucleases both from wild-type bacteria and from overproducing strains (containing plasmids with the endonuclease genes) has been described (see Table I for references). Before a new lot of endonuclease is used for damage analysis, the protein concentration required to saturate the incision reaction at a suitable reference damage site (a known substrate of the enzyme) should be determined. Some types of reference damage are listed in Table I, and a typical saturation curve is shown in Fig. 1. Similarly, it should be verified that the endonuclease preparation has no or negligible activity against DNA containing modifications that are not substrates. Even in relatively crude cell extracts from bacteria, the activity of several repair endonucleases is high enough for damage analysis, and both DNA and major EDTA-resistant contaminating nucleases can be readily removed [e.g., by diethylaminoethyl (DEAE)-cellulose chromatography].
[12]
ENDONUCLEASE FINGERPRINTING ~. e~ o
"O
125
2,0"
1.5'
1,0'
.N c
0,5"
o
9 to
~m v......I ........ I ........ I ........ I
0.0
0
1
10
100
1000
FPG protein concentration (ng/ml)
FiG. 1. Recognition by FPG protein of DNA modifications induced by methylene blue (10/zg/ml) plus visible light (4 sec, 1000-W halogen lamp at 93 cm distance) in phosphate buffer. For details, see E. Mfiller, S. Boiteux, R. P. Cunningham, and B. Epe, Nucleic Acids Res. 18, 5969 (1990). Therefore, overproducing or even wild-type strains deficient in some of the repair endonucleases may provide valuable endonuclease preparations which do not require purification to homogeneity. Endonuclease preparations may be kept for months at - 2 5 ° in 50% glycerol or frozen in aliquots in BE1 buffer (20 mM Tris-HCl, p H 7.5, 100 m M NaCI, I mM EDTA) at - 7 0 °. Damage Profiles from Cell-Free DNA To quantify strand breaks and endonuclease-sensitive sites induced under cell-flee conditions a relaxation assay is used which requires supercoiled D N A as target. We routinely use bacteriophage PM2 DNA (104 bp) prepared by the method described by Salditt, 1 but other types of D N A (e.g., plasmids) are equally suitable when prepared under conditions that avoid damage. For the modification, the D N A (10/zg/ml) is exposed to the damaging agent in, for example, phosphate buffer. Acidic conditions (pH < 6.5) have to be avoided because they generate AP sites. The D N A is precipitated by ethanol/sodium acetate and redissolved in BE I buffer at a concentration of 10/zg/ml for damage analysis by the relaxation assay (see below). Damage Profiles from Mitochondrial DNA To obtain damage profiles from mtDNA, 2 mitochondria are prepared from porcine or rat liver or kidney (10 g) as described by Johnson and 1M. Salditt, S. N. Brannstein, R. D. Camerini-Otero, and R. M. Franklin, Virology 48, 259 (1972). 2 j. Hegler, D. Binner, S. Boiteux, and B. Epe, Carcinogenesis (London) 14, 2309 (1993).
126
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[12]
L a r d y ) Exposure of the mitochondria (8 ml; 2 mg mitochondrial protein/ ml) to various agents is carried out in a buffer containing 250 mM sucrose, 20 mM Tris-HC1 (pH 7.6), and 0.1% bovine serum albumin. After centrifugation (5 min, 4900 g), the mitochondria are resuspended in 500 /zl TEG buffer (50 mM glucose, 10 mM EDTA, 25 mM TrisHCI, pH 8.0) containing RNase A (100/zg/ml). The mtDNA is isolated by alkali lysis, strictly avoiding acidic conditions (pH < 6.5): 500 /zl of 0.2 N NaOH containing 1% sodium dodecyl sulfate (SDS) is added at 0°, and directly afterward 500 /xl of a neutralizing solution (3 M potassium acetate adjusted to pH 6.6 with acetic acid). After centrifugation (10 min, 12,000 g, 0°) the supernatant is extracted twice with phenol and subsequently with chloroform to remove proteins. The aqueous solution is applied to an anion-exchange minicolumn (Quiagen Inc., Chatsworth, CA), and the mtDNA is eluted according to the manufacturer's protocol. After ethanol precipitation, the mtDNA is dissolved in BE1 buffer for damage analysis by means of the relaxation assay (see below). Bacteriophage PM2 DNA may be added as an "internal control" to the mitochondria prior to lysis, because both its supercoiled and relaxed form run separately from the mtDNA during gel electrophoresis. This control ensures that no DNA damage is induced during the workup procedure. Alternatively, the addition of PM2 DNA with a reference damage (Table I) can be useful to demonstrate that the mtDNA preparation does not contain contaminants (membrane particles?) that impair the activity of the repair endonucleases.
Damage Profiles from Bacterial DNA Plasmid-containing bacteria (2 x 101°/40 ml) are exposed to the damaging agents in minimal medium (10 g/liter K2HPO4, 200 mg/liter MgSO4, 2 g/liter citric acid, 50 g/liter glucose, pH 7.0). Isolation and purification of the plasmids are carried out essentially as described above for mitochondria; the incubation with alkali, however, is extended to approximately 5 min. To avoid contact of the plasmids with alkali, neutral ]ysis has been used for Salmonella typhimurium strains which have a cell wall deficiency (rfa). 4 Subsequent damage analysis by the relaxation assay is described below. 3 D. Johnson and H. Lardy, this series, Vol. 10, p. 41. 4 B. Epe, J. Hegler, and D. Wild, Carcinogenesis (London) 10, 2019 (1989).
[12]
ENDONUCLEASE FINGERPRINTING
127
Relaxation Assay: Quantification of Strand Breaks and Endonuclease-Sensitive Modifications in Supercoiled DNA Principle. The relaxation assay makes use of the fact that a single strand break or an incision by a repair endonuclease converts a supercoiled DNA molecule to a relaxed (nickel) form that migrates separately from the supercoiled form in agarose gel electrophoresis. 5The relative amounts of supercoiled and relaxed DNA are determined by fluorescence scanning after staining with ethidium bromide and are used to calculate the number of single-strand breaks and endonuclease-sensitive modifications per DNA molecule. A Poisson formula is used to account for the fact that only the first single-strand break (or endonuclease incision) in each molecule induces the relaxation [Eq. (1)]. ssb + ess = - l n [ 1 . 4 S C / ( 1 . 4 S C + OC)]
(1)
In Eq. (1), ssb represents the number of single-strand breaks per DNA molecule induced directly by the damaging agent, and ess is the number of endonuclease-sensitive sites per molecule that are incised by a repair endonuclease (if an incubation with the enzyme precedes the gel electrophoresis). S C and OC are the fluorescence intensities of the supercoiled and relaxed circular forms of the DNA, respectively. An empirical factor of 1.4 is used to account for the relatively lower fluorescence of ethidium bromide in supercoiled compared to the relaxed form of DNA. 6 Reliable quantification is possible in the range between 0.1 and 3 modifications (ssb + ess) per DNA molecule. The sensitivity of the assay is therefore proportional to the size of the DNA; with PM2 DNA (10,000 bp), approximately 1 modification per 105 base pairs can be measured. The number of DNA double-strand breaks (dsb) can be calculated from the relative amount of linear DNA (L), which migrates as a third band during gel electrophoresis [Eq. (2)]. In this case no correction for dsb -- L/(1.4SC + OC + L)
(2)
multiple breaks can be made; therefore, the number of double-strand breaks per DNA molecule should be low. Procedure. Aliquots (0.2 p~g) of the modified DNA in 20/zl BE1 buffer (see above) are incubated for 30 min at 37° with l0/zl of a repair endonuclease preparation in BEt buffer for most enzymes or TC buffer (20 mM Tris-HC1, pH 8.0, 100 mM NaCl, 15 mM CaCI2) for exonuclease III. The 5p. s. Seawelland A. K. Ganesan, in DNA Repair: A LaboratoryHandbookof Research Procedures" (E. C. Friedbergand P. C. Hanawalt,eds.), Vol. l, Part B, p. 425. Dekker, New York, 1981. 6R. S. Lloyd, C. W. Haidle, and D. L. Robberson, Biochemistry 17, 1890(1978).
128
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
[12]
reactions are stopped by the addition of 3/A of 10% SDS, and the DNA is applied to an agarose slab electrophoresis gel (0.9% agarose in 40 mM Tris-HCl, 5 mM sodium acetate, 1 mM EDTA, pH 7.9). After staining the gel with ethidium bromide, the relative amounts of the supercoiled and the relaxed form of PM2 DNA are determined by means of a fluorescence scanner (FTR20, Eppendorf Instruments, Berlin), and the number of modifications is calculated using E q. (1). Value s obtained with untreated control DNA are subtracted; typical background values observed with PM2 DNA are 0.05-0.15 sites per l04 bp depending on the enzyme used.
Modified Alkaline Elution Assay: Damage Profiles from Cultured Mammalian Cells The alkaline elution assay follows essentially the protocol of Kohn e t Modifications involving the use of repair endonucleases have been described. 8 The apparatus used should allow analysis of several samples in parallel and immersion of the filter holders in a water bath during the incubation with enzymes. Pure repair endonucleases are required to keep incisions by nonspecific contaminating nucleases low. Cultured cells are exposed to the damaging agents in PBSG (140 mM NaCl, 3 mM KC1, 8 mM NazHPO 4, I mM KH2PO 4, 1 mM CaCl 2 , 0.5 mM MgC12 , 0.1% glucose, pH 7.4). Subsequently, 1 × 106 cells are applied to a polycarbonate filter (25 mm diameter, 2/.tm pore size). After washing with 5 ml PBSG without C a 2+ and M g 2+ , a lysing solution (0.1 M glycine, 20 mM Na2EDTA, 2% SDS, pH 10.0) is pumped through the filter for 1 hr at 25 °. After washing with 25 ml BE~ buffer, 2 ml repair endonuclease solution (e.g., 2/~g/ml FPG protein in BE1 buffer or 100 U/ml exonuclease III in a buffer containing 36 mM Tris-HCl, 18 mM CaC12 , 0.5 mg/ml bovine albumin, pH 8.0) is applied and pumped through the filter for 40 min at 37° . To quantify direct strand breaks, this step is carried out without endonucleases. After washing with BE1 buffer, 5 ml proteinase K solution (500/~g proteinase K in lysing solution) is passed through the filter at 25 ° for 30 min. After another washing with BE1, the DNA is eluted at 2.3 ml/ hr with a solution of 20 mM EDTA (acid form) adjusted to pH 12.1 with tetraethylammonium hydroxide and collected for l0 hr in 4.6-ml fractions. The DNA eluted and retained on the filter is quantified after neutralization by Hoechst 33258 fluorescence measurement (excitation at 360 nm; emission at 450 nm; final dye concentration 0.5/~M). The number of modificaal. 7
7 K. W. Kohn, L. C. Erickson, R. A. G. Ewig, and C. A. Friedman, Biochemistry 15, 4629 (1976). 8 A. J. Fornace, Jr., Mutat. Res. 94, 263 (1982).
[12]
ENDONUCLEASEFINGERPRINTING
129
tions per 106 bp is calculated from the slopes of the elution curves after subtraction of the slopes obtained with unmodified control cells. For calibration, X-rays are used as reference damage, assuming that a dose of 6 Gy generates 1 single-strand break per 106 b p . 7
Examples
Figure 2A depicts several DNA damage profiles induced in PM2 DNA under cell-free conditions (phosphate buffer). The damage profile induced by singlet oxygen [generated by thermal decomposition of an aromatic endoperoxide of 3,3 '-( 1,4-naphthylidene) dipropionate (NDPO2)9] is dominated by base modifications sensitive to FPG protein, whereas both strand breaks and AP sites are rare. In contrast, the damage induced by hydroxyl radicals [generated by ionizing radiation or by xanthine plus xanthine oxidase in the presence of Fe(III)-EDTA] consists of approximately equal levels of base modifications, AP sites, and single strand breaks (Fig. 2A). The influence on the damage profile of various quenchers and scavengers and of D20 as solvent has been studied to demonstrate that all types of modifications in the damage profiles are indeed generated by the same ultimate species (singlet oxygen or hydroxyl radicals).l° The damage profiles induced under cell-free conditions by acridine orange (Fig. 2A) and several other photosensitizers absorbing visible light are very similar to the damage profile induced by singlet oxygen, even in cases in which the damaging mechanism seems to involve direct electron transfer (type I reaction) rather than singlet oxygen (type II reaction). 1~ The damage profile induced by Cu(II)-phenanthroline (Fig. 2A) is interesting, as it indicates the presence of AP sites that are sensitive to endonuclease IV and exonuclease III, but not to the other endonucleases. These AP sites are most probably oxidized in the 1' position.12 In Fig. 2B, DNA damage profiles from L1210 mouse leukemia cells treated with acridine orange plus light or with H202 are depicted. A comparison of these fingerprints with those of the cell-free profiles indicates that the cellular DNA damage induced by the excited photosensitizer is similar to the damage induced under cell-free conditions. Therefore, cellular generation of hydroxyl radicals (via the Fenton reaction) or activation of cellular nucleases is not involved in the production of this particular DNA damage. On the other hand, the damage profile observed after treat9 p. i0 B. 11 B. 12 L.
Di Mascio and H. Sies, J. Am. Chem. Soc. U l , 2909 (1989). Epe, P. Miatzel, and W. Adam, Chem.-Biol. Interact. 67, 149 (1988). Epe, M. Pflaum, and S. Boiteux, Mutat. Res. 299, 135 (1993). F. Povirk and R. Steighner, Mutat. Res. 214, 13 (1989).
130
OXIDATIVE DAMAGE TO D N A AND D N A REPAIR
A
[12]
cell-free (phosphate buffer) 2.0
~~
°~1 " 5
a.1.0
:~0 0.0 NDI~02 xant;llne X-rays Cu-I)hen .cr;dlne + H202 or. +light
B D.
L1210 cells endonucleasesens. modifications:
o,.o] H o,5t
•
[]
FPG protein
[] [] [] n
endon. III UV endon. endon. IV exon. III
single strand breaks
I or.
+light
H202
FiG. 2. Cell-free and cellular damage profiles. Column heights indicate the number of various endonuclease-sensitive modifications and of single-strand breaks induced in PM2 DNA (A) by treatment with (a) NDPO2 (3.5 raM; 2 hr in D20 buffer), (b) superoxide, generated from xanthine (8/zM) and xanthine oxidase in the presence of Fe(III)-EDTA, (c) ionizing radiation (20 Gy), (d) Cu(II)-phenanthroline (4/zM) in the presence of H202 , and (e) acridine orange (2/xg/ml) plus light (1 min) and induced in L1210 mouse leukemia cells (B) by treatment with (a) acridine orange (2/zg/ml) plus light (10 sec) and (b) H~O2 (100/zM, 20 min, 0°). For details, see B. Epe, M. Pflaum, M. H/iring, J. Hegler, and H. Riidiger, Toxicol. Lett. 67, 57 (1993), and B. Epe, M. Pflaum, and S. Boiteux, Murat. Res. 299, 135 (1993). m e n t o f L1210 c e l l s w i t h H 2 0 x is c o n s i s t e n t w i t h t h e a s s u m p t i o n t h a t hydroxyl radicals are the ultimate damaging species. Further Applications and Related Methods Additional information can be obtained by incubation of the modified D N A at 60 ° (e.g., f o r 2 hr) p r i o r to e n z y m e t r e a t m e n t . A l k y l a t e d D N A
[12]
ENDONUCLEASE FINGERPRINTING
131
bases such as 7-alkylguanine are converted to AP sites and thus become sensitive to exonuclease III and other repair endonucleases. On the other hand, pyrimidine hydrates, which are sensitive to endonuclease III, may eliminate water and become insensitive. 13 DNA damage profiles of oxidative base modifications have been obtained by GC/MS/SIM techniques (gas chromatography coupled with mass spectrometry in the selected ion-monitoring mode). 14The two methods may complement one another as GC/MS/SIM allows differentiation between base modifications that are sensitive to the same repair endonuclease whereas oxidized and regular AP sites and strand breaks cannot be quantified. In a combination of the two methods, DNA base modifications excised by repair endonucleases can be analyzed by GC/MS/SIM. 13 T. Ganguly and N. J. Duker, Nucleic Acids Res. 197 3319 (1991). 14 M. Dizdaroglu and D. S. Bergtold, Anal. Biochem. 1567 182 (1986).
[13]
EFFECT OF THIOLS ON T LYMPHOCYTES
135
[13] E f f e c t o f R e a c t i v e O x y g e n Intermediates and Antioxidants on Proliferation and Function o f T L y m p h o c y t e s By W U L F DROGE, SABINE M I H M , MICHAEL BOCKSTETTE, a n d STEFFEN ROTH
Introduction The importance of thiols and especially of cysteine and glutathione to lymphocyte function has been known for many years. 1-5 The extracellular concentrations of thiols have a profound effect on B and T cell responses in vitro. There are also strong indications that cyst(e)ine may play a limiting role for the immune system in rio0. 4'6 Certain lymphocyte functions, however, are potentiated by hydrogen peroxide or other reactive oxygen intermediates. 7'8 The intact immune system thus appears to require a delicate balance between prooxidant and antioxidant conditions. This is obviously maintained by a limited and well-regulated supply of cysteine. 8 Expectedly, the pathological changes in cellular cysteine supply that are seen in persons infected with the human immunodeficiency virus (HIV) are associated with an immunological disorder. 6'8 Here we describe some of the experimental approaches that are being used to determine the intracellular concentrations, redox states, and functions of cysteine and glutathione and the role of functionally related metabolites.
Determination of Intracellular Glutathione and Amino Acid Levels The procedure for determining the total intracellular glutathione level has been described elsewhere in this series. 9 To determine the amount of oxidized glutathione (GSSG), the acid-soluble supernatants are treated M. W. Fanger, D. A. Hart, J. V. Wells, and A. Nisonoff, J. lmmunol. 105, 1043 (1970). 2 J.-C. Cerottini, H. D. Engers, H. R. MacDonald, and K. T. Brunner, J. Exp. Med. 140, 703 (1974). 3 D. L. Hamilos and H. J. Wedner, J. Immunol. 135, 2740 (1985). 4 W. Dr6ge, C. Pottmeyer-Gerber, H. Schmidt, and S. Nick, Irnmunobiology 172, 151 (1986). 5 H. Gmfinder, H.-P. Eck, B. Benninghoff, S. Roth, and W. Dr6ge, Cell. Immunol. 129, 32 (1990). 6 H.-P. Eck, T. Mertens, H. Rasokat, G. F/itkenheuer, C. Pohl, M. Schrappe, V. Daniel, H. N~iher, D. Petzoldt, P. Drings, and W. Dr6ge, Int. Immunol. 4, 7 (1992). 7 S. Roth and W. Dr6ge, Cell. lmmunol. 108, 417 (1987). 8 W. Dr6ge, H.-P. Eck, and S. Mihm, lmmunol. Today 13, 211 (1992). 9 W. Dr6ge, H.-P. Eck, and S. Mihm, this series, Vol. 233 [59].
METHODS IN ENZYMOLOGY.VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any form reserved.
136
ASSAY OF STRESS GENES/PROTEINS
[13]
TABLE I INTRACELLULAR LEVELS OF REDUCED AND OXIDIZED GLUTATHIONEa Cell line
GSH
GSSG
GSH/GSSG
Total thiol
Thiol/GSSG
U937 Molt-4
34.8 -+ 3.7 17.4 -+ 2.5
2.04 -+ 0.31 3.58 - 0.50
17.1 4.9
48.9 -+ 6.5 26.7 -+ 2.5
24.0 7.5
Data in nmol/mg protein.
with 2-vinylpyridine (2/zl/100/zl) for 2 min and incubated with 6/xl triethanolamine for 30 min at 20 °. 5,5'-Dithiobis-2-nitrobenzoic acid (DTNB) is then added, and glutathione is determined photometrically as described by Dr6ge e t al. 9 The amount of reduced glutathione (GSH) can be determined as the difference between total glutathione and GSSG. For the determination ofintracellular cysteine, cystine, and GSH levels with the amino acid analyzer, ~° cells (about 2 × 107) are washed twice with phosphate-buffered saline (PBS) and resuspended in 0.08 ml PBS. At 4 °, the cell suspension is mixed with 40 mM NEM, and the protein is subsequently precipitated with 2.5% sulfosalicylic acid. The samples are centrifuged at 1700 g for 10 min, and the supernatants are studied with an amino acid analyzer. With this method, the total glutathione content of freshly prepared peripheral blood mononuclear cells (PBMC) of healthy human individuals was found to be 24 nmol/mg protein and that of peripheral blood monocytes 34 nmol/mg protein, n For comparison, murine lymphocytes from spleen or lymph nodes and murine peritoneal macrophages were found to contain approximately 8 and 15 nmol of total glutathione/mg protein, respectively (H.-P. Eck and W. Dr6ge, unpublished observation), suggesting that mice may have generally a lower intracellular level than humans. The intracellular glutathione levels of the human promonocytic line U937 and the human T cell line Molt-4 in standard RPMI 1640 cultures without 2-mercaptoethanol were found to be similar to those of freshly prepared human blood monocytes and PBMC, respectively (Table I). The T lineage cells appear to have not only a markedly lower intracellular glutathione level but also a markedly lower GSH/GSSG ratio. The GSH/GSSG ratio is believed to have a strong influence on several enzyme activities. 12 Jo j. A. Schneider, K. Bradley, and J. S. Seegmiller, Science 157, 1321 (1967). H H.-P. Eck, H. Gmiinder, M. Hartmann, D. Petzoldt, V. Daniel, and W. Dr6ge, Biol. Chem. Hoppe-Seyler 3711, 101 (1989). 12 D. M. Ziegler, Annu. Rev. Bioehem. 54, 305 (1985).
[13]
EFFECT OF THIOLS ON T LYMPHOCYTES
137
A similar redox state of T lineage cells and monocyte/macrophage lines is seen when the cysteine/cystine redox couple is analyzed in a culture medium with approximately physiological amino acid levels (Table II). The analysis is performed with cell lines that are routinely propagated in standard RPMI 1640 medium but incubated before the analysis for 30 hr in NCTC 135 culture medium with 10% fetal calf serum (FCS) at a density of 1 × 106/ml. Cells of the murine T cell lymphoma L5178Y ESb have been prepared either e x v i v o from a tumor-bearing mouse 3 days prior to the experiment or propagated as separate sublines in 2-mercaptoethanol-free cultures for several months. ~7 The data in Table II indicate also that the intracellular GSH levels are lower under these conditions than in RPMI medium, especially in the U937 cells. Moreover, the ratio glutathione/cysteine is markedly lower in cells of the monocyte/macrophage line as compared to T lineage cells, suggesting that T lineage cells have a higher 7-glutamylcysteine synthetase (glutamate-cysteine ligase) or glutathione synthase activity. This conclusion is in agreement with the observation that the intracellular glutathione levels of U937 cells are profoundly augmented by the exogenous addition of GSH but not by cysteine or N-acetylcysteine, whereas intracellular glutathione levels of Molt-4 cells are strongly elevated by all three thiols. ~3 The high rate of glutathione biosynthesis in the T lymphocytes may be necessary to compensate for their weak cystine transport activity (see below). Membrane Transport Systems for Cysteine and Cystine: Influence of Extracellular Glutamate Concentration on Lymphocyte Functions Procedures for the determination of amino acid transport activities have been reported elsewhere 14-17and are not described here. However, the following key facts are particularly important for a detailed understanding of the redox regulation of lymphoid cells and for the interpretation of experiments with lymphocyte cultures. The special role of cysteine among all protein-forming amino acids is determined by the extracellular concentrations of cystine and cysteine and by the membrane transport activities of lymphoid cells for the two amino acids (see Table III). Cysteine is transported mainly by the transport system ASC which is strongly expressed in lymphocytes. However, the plasma concentration of cysteine i~ S. Mihm, J. Ennen, U. Pessara, R. Kurth, and W. Dr6ge, AIDS (London) 5, 497 (1991). 14 H. Watanabe and S. Bannai, J. Exp. Med. 165, 628 (1987). 15 T. Ishii, Y. Sugita, and S. Bannai, J. Cell. Physiol. 133, 330 (1987). t6 H. Gm0nder, H.-P. Eck, and W. Dr6ge, Eur. J. Biochem. 201, 113 (1991). 17 J.-S. Lim, H.-P. Eck, H. Gm0nder, and W. Dr6ge, Cell. lmmunol. 1411, 345 (1992).
138
ASSAY
OF
STRESS
GENES/PROTEINS
[13]
O
O
L~
% .1 LO > LO
I-
+1 +1
+1 ÷1 +1
+1 +1
+1 +1 ÷1
+1+1
+l+l+l
+l +1
+1 +1 +1
t Z < Z 0
<
r,..) Z
r,.) < .1 .1 M
E e~
z
E O e., O
..= e~ O
[13]
EFFECT OF THIOLS ON T LYMPHOCYTES
139
TABLE III MEMBRANE TRANSPORT SYSTEMS FOR CYSTEINE AND CYSTEINE DERIVATIVE Compound Cysteine Cystine
Disulfide of cysteine and 2-mercaptoethanol
Comments Transport system ASC shared by glycine, alanine, methionine, etc. Concentration in blood plasma - 1 × 10-5 M Transport system Xc- shared by glutamate and homocysteate Concentration in blood plasma - 0 . 5 - 1 × 10-4 M Transport system shared by phenylalanine and leucine Most relevant cysteine derivative in standard lymphocyte culture medium
(-1 × 10-5 M) is extremely small in comparison with that of all other protein-forming amino acids. The plasma concentration of the disulfide cystine (0.5-1 x 10-4 M) is about 10 times higher in terms of cysteine equivalents. This large amino acid uses another membrane transport system (X~-), however, which is shared by glutamate and which is extremely weak in T lymphocytes. 14-17In spite of the weak X~- transport activity, the intracellular glutathione concentrations of human PBMC are significantly correlated with the plasma cystine but not with the (relatively low) concentrations of cysteine (R. Kinscherf, S. Mihm, and W. Dr6ge, unpublished observation). Mercaptoethanol (2-ME) is widely used by immunologists to ensure an adequate supply of cysteine for lymphocytes in cell cultures. For the interpretation of experiments with 2-ME, it is important to know (i) that 2-ME and cysteine are oxidized under standard conditions with a half-life of less than ! hr and (ii) that 2-ME generates with cyst(e)ine a mixed disulfide which, in contrast to cystine, is quite effectively transported into lymphocytes.15 Experimental Procedure to Determine Functional Consequences of Elevated Extracellular Glutamate Concentrations
Glutamate inhibits competitively the membrane transport of cystine. 14 Pathologically elevated extracellular glutamate concentrations occur in several etiologically unrelated diseases including cancer and the acquired immunodeficiency syndrome (AIDS) and are correlated with immunological dysfunctions. 6'9 The following procedure is designed to determine the functional consequences of pathologically relevant changes of glutamate concentrations in vitro. Lymphocytes are incubated in a culture medium with approximately physiological amino acid concentrations (NCTC 135 medium plus 10% FCS without 2-ME). Some cultures receive additional
140
ASSAY OF STRESS GENES/PROTEINS
[13]
÷
50# M glutamate 200p M glutamate
5O/t M glu + 100# M cysteine 200# M glu + 100# M cysteine 1
2
3
4
5
3H - thymidine incorporation ( cpmxl0 -4 )
FIG. 1. Inhibition of the mitogenic response of concanavalin A-stimulated spleen cells by glutamate (glu) in the absence or presence of additional cysteine (modified NCTC 135 medium). For details, see text.
amounts of glutamate and/or cysteine, and the DNA synthetic activity is determined by [3H]thymidine incorporation after 2-3 days. Figure 1 shows the inhibitory effect of glutamate in cultures with murine spleen cells that have been stimulated with concanavalin A (5/zg/ml). Influence of Extracellular Cysteine Concentration on Intracellular Glutathione Levels and T Cell Functions In view of the weak cystine transport activity and the relatively strong transport activity for cysteine, it is not unexpected that T cells respond with extreme sensitivity to moderate changes of extracellular cysteine concentration even in the presence of relatively high and approximately physiological concentrations of cystine. ~1This effect may play a regulatory role during T cell activation because macrophages are capable of releasing considerable amounts of cysteine and come into close contact with T cells in the process of antigen stimulation. 5,m The degree of cysteine dependency can be determined by incubating T cell preparations or T cell clones (see Table IV) at a density of 2-4 x 106/20 ml with human recombinant interleukin 2 (rlL-2; 30 units/ml) in modified NCTC 135 medium with 10% FCS. Some of the cultures may receive additional amounts of 5 x 10-5 mol/liter culture volume of 2-mercaptoethanol (+2-ME) or 3 x 10-3 mol/liter ofglutathione (+GSH) at the start or 3 × 10-5 mol/liter ofcysteine (+Cys) 4 times per day. The experiments in Table IV illustrate that intracellular glutathione levels and the viability of the T cell clone 29 and clone 18 H.-P. Eck and W. DrOge, Biol. Chem. Hoppe-Seyler 370, 109 (1989).
[13]
EFFECT OF THIOLS ON T LYMPHOCYTES
141
TABLE IV EFFECT OF CYSTEINE STARVATION ON VIABILITY AND INTRACELLULARGLUTATHIONE CONTENT OF MURINE T CELL CLONESa Total
Clone D10.G4.1 with TPA/Ca2+ ionophore
D10.G4.1 without TPA/Ca 2+ ionophore 29 without TPA/Ca 2+ ionophore W-2 without TPA/Ca 2+ ionophore
Conditions
Viable cell number (% of cell input)
glutathione (nmol/mg protein)
-Cys +Cys +2-ME +GSH -Cys +Cys -Cys +Cys -Cys +Cys
185 200 190 260 109 109 42 151 34 73
3.9 20.2 19.0 36.2 11.8 21.4 0.6 11.0 5.9 25.3
a Glutathione and protein were determined on aliquots of the cultured cells on day 2 except for
the stimulated D10.G4.1 cells, which were harvested after only 24 hr. These cells had been stimulated with 10 ng/ml tetradecanoylphorbol acetate (TPA) plus 0.5/zg/ml calcium ionophore at the start of the culture.
W-2 decrease rapidly in the absence of cysteine. 17The clone D10.G4.1,19,20 in contrast, maintains substantial intracellular glutathione levels in the absence of cysteine unless it receives an additional stimulus.
Procedure to Evaluate Cysteine Requirement During T Cell Activation Interaction of the antigen receptor with its specific antigen transforms the T cell within a few hours from a small resting lymphocyte with little cytoplasm into a relatively large lymphoblast with abundant cytoplasm. During the process, there is an extreme demand for various metabolites up to the limits of membrane transport capabilities. This is the time in the life of a lymphocyte when the cysteine supply is expected to be most limiting. Certain T cell clones which grow continuously but are still capable of responding to antigenic stimulation with either increased proliferation or lymphokine production provide a convenient model for studies on the metabolic requirements during activation. The T cell clone D10.G4.1 is a useful example. 19'2° It was shown to express the phenotype CD4 + and to produce interleukin 4 (IL-4) after appropriate stimulation. It is routinely 19 j. Kaye, S. B. Mizel, E. M. Shevach, T. R. Malek, C. A. Dinarello, L. B. Lachman, and C. A. Janeway, Jr., J. Immunol. 133, 1339 (1984). 2o E. A. Kurt-Jones, S. Hamberg, J. Ohara, W. E. Paul, and A. K. Abbas, J. Exp. Med. 166, 1774 (1987).
142
ASSAY OF STRESS GENES/PROTEINS
[13]
propagated in IL-2-containing cell culture medium and stimulated every 3 weeks with irradiated (2000 rad) C57BL/6 spleen cells. Alternatively, this T cell clone can be stimulated effectively with tetradecanoylphorbol acetate (TPA) in combination with a calcium ionophore or mitogenic lectin. The biochemical basis of this method of T cell stimulation has been described elsewhere. 21 The results of a typical experiment are also shown in Table IV. The stimulated DI0.G4.I cells have relatively low intracellular glutathione levels unless the cultures are provided with additional amounts of cysteine, 2-mercaptoethanol, or glutathione. The unstimulated D10.G4. I cells, in contrast, maintain relatively high intracellular glutathione levels even in the absence of cysteine. Because the T cell clones 29 and W-2 always show a strong requirement for cysteine, they appear to resemble highly activated T cells. Indeed, their routine propagation requires 2-mercaptoethanol-containing culture medium. 8,~7 Procedures to Evaluate Consequences of Cysteine Deficiency and Glutathione Depletion on Different T Cell Functions and T Cell Subsets At face value, one might have expected that the rate of protein synthesis may be most sensitive against cysteine starvation. Unexpectedly, however, the rate of DNA synthesis appeared to be markedly more sensitive than that of protein synthesis. In studies with several T cell clones or mitogenically stimulated T cells, it was consistently observed that the inhibition of DNA synthesis ([3H]thymidine incorporation) under conditions of cysteine deficiency precedes the decrease of cell viability by several days, whereas the inhibition of protein synthesis ([3H]leucine incorporation) usually parallels the loss of viability and therefore may be a direct consequence of cell death (H. Gmtinder, H.-P. Eck, and W. Dr6ge, unpublished observation). The cysteine dependency of DNA synthesis reflects the dependency on the intracellular glutathione level: when the clone 29 or W-2 cells are treated with 10-100/zM buthionine sulfoximine (BSO), a specific inhibitor of glutathione biosynthesis, DNA synthesis is also strongly inhibited even in the presence of relatively high extracellular cysteine concentrations. 5'22'23The unstimulated D10.G4.1 clone maintains a high level of DNA synthesis even in cultures without cysteine or other 2t W. Dr6ge, lmmunol. Today 7, 340 (1986). 22 H. Gmtinder and W. DrOge, Cell. lmmunol. 138, 229 (1991). 23 H. Gmiinder, S. Roth, H.-P. Eck, H. Gallas, S. Mihm, and W. Drfge, Cell. lmmunol. 130, 520 (1990).
[13]
EFFECT OF THIOLS ON T LYMPHOCYTES
143
thiols. ~v,22Expectedly, however, the rate of DNA synthesis in stimulated D10.G4. I cells shows again a strong cysteine dependency (Fig. 2). To characterize the cysteine requirement at different time intervals after stimulation (Fig. 2), D10.G4.1 cells are incubated in NCTC 135 culture medium with 10% FCS and 5 U/ml IL-2 but without 2-mercaptoethanol for 24 hr at a density of 1 × 10 6 cells/ml. The cells are then transferred into microcultures with fresh NCTC 135 culture medium without 2-ME (1.2 × 10 4 cells/0.2 ml) and are stimulated with a combination of TPA (10 ng/ml) and concanavalin A (5 ~g/ml). [3H]Thymidine (0.5/xCi) is added 24 hr later, and the incorporation of [3H]thymidine is tested after another 7 hr with a cell harvester. Some of the cultures receive a single dose of cysteine corresponding to a final concentration of 30/zM at different times after stimulation, or they receive cysteine (30/xM) repeatedly every 2 hr, starting at the indicated time intervals after culture initiation. The results of the experiment (Fig. 2) show that the increased demand for cysteine does not occur immediately after stimulation but only 6-8 hr later. Cultures which received the first dose of cysteine 6 or 8 hr after initiation and subsequent doses of cysteine every 2 hr showed essentially the same rate of thymidine incorporation as cultures that received the first dose at the start. The strong decline of thymidine incorporation in cultures which received the first dose of cysteine only 12-20 hr after incubation indicated that this is the period with the greatest demand for cysteine. The decline is not the consequence of the decreasing cumulative 20r
18i
4~
0
i
0
i
i
i
i
i
~
i
i
i
i
_
2 4 6 8 10 12 14 16 18 20 22 24 hours
FIG. 2. Cysteine requirement at different time intervals after stimulation of the DI0.G4.1 T cell clone. (©) Single dose of cysteine at indicated time, ([]) cysteine every 2 hr, starting at indicated time, and (A) control (no cysteine). For details, see text.
144
ASSAY OF STRESS GENES/PROTEINS
[13]
amounts of cysteine added, since the same curve was obtained in cultures which received the same doses ofcysteine every 4 hr (not shown). Control cultures received no cysteine but rather corresponding amounts of solvent (culture medium) (Fig. 2). Another highly cysteine-dependent lymphocyte function is the activation of cytotoxic T cell activity. The activation of lymphokine (IL-2) production, on the other hand, is not at all affected by cysteine starvation or glutathione depletion and is even inhibited by the addition of exogenous glutathione.4.5.16,23,24 The differential sensitivity of different T cell subsets to cysteine starvation and glutathione depletion can be determined by cytofluorographic analysis. Spleen cells from C3H mice (2 × 107) are incubated with 5 x 10 6 irradiated (1500 rad) C57BL/6 stimulator cells in RPMI 1640 culture medium with 10% FCS. On day 3, cultures are treated with graded concentrations of BSO, and on day 5 1 × 10 6 cells are resuspended in 50/xl PBS that contains 5% FCS. The cells are mixed with 4/zl fluorescein-conjugated anti-Lyt 2 (Becton Dickinson, Lincoln Park, NJ, No. 1353) and 4 /~1 phycoerythrin-conjugated anti-L3T4 antibody (Becton Dickinson, No. 1447). After incubation for 30 min at 4° in the dark, the cells are washed twice and resuspended in 250/zl PBS with 5% FCS. Dead cells are stained by the addition of 250/zl propidium iodide (Sigma, St. Louis, MO, P4170) (I /xg/ml in PBS). The resulting cell suspension is subjected to cytofluorographic analysis [i.e., with a FACscan (Becton Dickinson)] with a life gate to exclude the dead cells. The analysis revealed that CD8 + T cells are markedly more sensitive against glutathione depletion on day 3 than CD4 + T cells and that the large CD8 + T cell blasts are in fact the most sensitive cells. The activation of cytotoxic T cell activity, a typical function of CD8 + T cell blasts, is also strongly decreased after glutathione depletion on day 3 and is rescued by the addition of 2-6 mM glutathione to the culture medium. 22 These data do not exclude the possibility that CD4 + T cells are similarly sensitive to glutathione depletion during the first 24 hr after activation.
Capacity o f Various Cysteine Derivatives to Substitute for Cysteine Cultures of cysteine-dependent T cell clones can also be used to determine whether physiologically or pharmacologically relevant cysteine derivatives such as cystine, N-acetylcysteine (NAC), 8 or thiazolidine deriva-
24 S. Roth and W. Dr6ge, Eur. J. Immunol. 21, 1933 (1991).
[13]
EFFECT OF THIOLS ON T LYMPHOCYTES
145
Thr
Ser
Glu Ala OTC
Cystine Cysteine ~/////////////////////////////////~ NAC
GSH~ 0
A 10
~
~///////////////////////A ~///////////////////////'///////A ¢////////////////////////////////////////~ ~/////////////////////A
20
intracellular glutathione ( nmol / mg proteinZ
30
lb
2b
ab
3H - thymidine incorporation
( cpmx 10 "3 )
FIG. 3. Capacity of different thiol derivatives to support intracellular glutathione levels and DNA synthesis. For details, see text.
tives z5 can substitute for cysteine. In the experiment of Fig. 3, intracellular glutathione concentrations and the rate of [3H]thymidine incorporation of the T cell clone 29 have been determined after 24 hr of incubation. The results show that cysteine (3 × 10 -5 M) can be substituted by 3 x 10 -5 M NAC, 3 × 10 -5 M 2-ME, or 5 × 10 -3 M GSH. In contrast, the addition of 1.5 x 10 -5 M cystine or 3 × 10 -5 M 2-oxo-4-thiazolidine carboxylate (OTC) had no effect. Similar results have been obtained with other T cell clones (not shown). The T cells either lack the enzyme to convert OTC to cysteine or fail to transport OTC through the plasma membrane.
Methods to Evaluate Role of Reactive Oxygen Intermediates in Lymphocyte Activation In view of the many potentiating effects of cysteine and glutathione, it is a surprising finding that certain T cell functions are augmented by reactive oxygen intermediates (ROI). 7 These processes are of course antagonized by antioxidants such as cysteine and GSH, 24 and this may explain the physiological relevance of the weak cystine transport activity of lymphoid cells. A profound immunopotentiating effect of hydrogen peroxide and superoxide anion (02.-) can be demonstrated with the following experimental system. Accessory cell-depleted splenic T cell preparations from mice are prepared by incubating the spleen cells in two consecutive cycles in a z5 H. J. Debey, J. B. Mackenzie, and C. G. Mackenzie, J. Nutr. 66, 607 (1958).
146
ASSAY OF STRESS GENES/PROTEINS
[13]
nylon wool column as described by Julius e t al. 26 The T cell-enriched nonadherent spleen cell fraction is subsequently incubated for 2 min at a density of 1 spleen equivalent per milliliter in a mixture of 9 parts of a 0.83% NH4C1 solution in water and 1 part 2% Tris buffer adjusted with HCI to pH 7.5. The accessory cell-depleted splenic T cells are washed 3 times with culture medium and incubated in RPMI 1640 culture medium without mercaptoethanol but with 10% FCS, 3 x 10-2 M lactate, and 5/~g/ml concanavalin A at a density of 5 x 10 6 cells/ml. (Concentrations of 10-50 mM lactate are expected to occur in the vicinity of glycolytically active macrophages. The role of lactate is discussed below.) After an incubation of 48 hr, the culture supernatants are harvested and assayed for IL-2 activity with the IL-2-responsive T cell clone W - 2 . 7 In this system it can be shown that the production of IL-2 is strongly augmented by hydrogen peroxide at a physiologically relevant concentration of 1 x 10-5 M. Higher concentrations are toxic. A similar augmentation of IL-2 production is observed with an O2T-generating system, containing 1 x 10-5 M xanthine and graded concentrations of xanthine oxidase. 7 Hydrogen peroxide (1 x 10-5 M) was also found to augment the DNA synthetic activity in cultures of allogeneically stimulated murine spleen cells. 7 Moreover, studies on the T cell clone D10.G4.1 reveal that 1 x 10-5 M hydrogen peroxide causes a marked augmentation of the rate of DNA synthesis in the absence but not in the presence of concanavalin A (Fig. 4), suggesting that a strong mitogenic stimulus may either bypass the requirement for ROI or may induce by itself the generation of ROI within the cell. The available evidence supports the latter interpretation. 8 The D10.G4.1 cells, which are routinely propagated in RPMI 1640 medium plus 10% FCS, are transferred into NCTC 135 culture medium plus 10% FCS and 6 U/ml IL-2 without 2-ME 16 hr prior to the start of the experiment. Approximately 2.5 x 104 D10.G4.1 cells are then incubated in 0.2 ml of fresh NCTC 135 culture medium with FCS for 32 hr. Graded amounts of hydrogen peroxide are added at the start of the culture, and 3 x 10-5 M cysteine is added 8 hr later. [3H]Thymidine (0.5 ~Ci) is added 20 hr after initiation of the culture, and thymidine incorporation is finally determined 12 hr later with a cell harvester. The results of the experiment (Fig. 4) are interpreted to mean that exogenous hydrogen peroxide may facilitate T cell activation in response to a suboptimal stimulus but may be inhibitory if the cells are exposed to a stronger activating signal. It is important to point out in this context that freshly prepared media usually generate hydrogen peroxide on incubation at 37 °. Using RPMI 26 M. H. Julius, E. Simpson, and L. A. Herzenberg, Eur. J. Immunol. 3, 645 (1973).
[13]
EFFECT
OF THIOLS
ON T LYMPHOCYTES
147
6
0 0 c
E tT',
____~
/]
0
__L~
I
I
3.3
10
30
90
UM H202
FIG. 4. Effect of hydrogen peroxide on DNA synthesis in the T cell clone DI0.G4.1. (D) No concanavalin A, (e) 2/zg/ml concanavalin A. For details, see text.
1640 powder medium (GIBCO, Grand Island, NY) we typically found an increasing concentration of hydrogen peroxide, reaching 11-12/zM H202 after 20 min of incubation (S. Roth, 1992, unpublished observation). This concentration decreased gradually over 2 days. RPMI medium without phenol red (Sigma) was found to generate up to 80/zM hydrogen peroxide on incubation at 37°. The reaction that generates hydrogen peroxide in the culture medium is not yet understood, but the initial presence of 10-5 M hydrogen peroxide may be an important and generally positive factor in standard mixed lymphocyte cultures. Effect of Lactate and Pyruvate on Cysteine Requirement of T Lineage Cells and on Cysteine Assay The reactive a-keto group of pyruvate can undergo spontaneous reactions with cysteine 27'28and with hydrogen peroxide. 29The latter effect can potentially interfere with the GSH peroxidase reaction as demonstrated in the following experiment. Hydrogen peroxide (50/xM) is incubated for l0 min together with graded concentrations of pyruvate and then mixed 27 M. P. Schubert, J. Biol. Chem. 114, 341 (1936). 28 p. C. Jocelyn, "Biochemistry of the SH group," p. 71. Academic Press, London, 1972. 29 j. O'Donnell-Tormey, C. F. Nathan, K. Lanks, C. J. DeBoer, and J. de la Harpe, J. Exp. Med. 165, 500 (1987).
148
ASSAY OF STRESS GENES/PROTEINS
[13]
50
401 w
30 CO
20
10
0
i
z
i
i
31
63
125
250
500 1000 2000 pyruvate [IJM]
FIG. 5. Effect of p y r u v a t e on the G S H peroxidase reaction. For details, see text.
with 1 mM GSH plus 1 U GSH peroxidase. GSSG is determined 30 min later. The results of this experiment (Fig. 5) show that the generation of GSSG is markedly inhibited by millimolar concentrations of pyruvate. For glycolytically active T cells and in the vicinity of lactate-producing macrophages, these are certainly physiologically relevant concentrations. The chemical interaction between pyruvate and cysteine (Fig. 6) can be detected by the DTNB assay for thiols. If graded concentrations of cysteine are incubated with 10-fold higher concentrations of pyruvate, the apparent cysteine concentration in the ninhydrin assay 3°,31 is not decreased, whereas the apparent cysteine concentration in the thiol assay 9 is markedly decreased (Fig. 7). This indicates that the covalent complexes (mainly thiazolidine derivatives) are dissociated under the conditions of the ninhydrin assay but not in the thiol assay. The former detects the total cysteine, the latter detects free cysteine only. Thiazolidine derivatives of cysteine and pyruvate are, in principle, also demonstrable by nuclear magnetic resonance (NMR) analysis (W. E. Hull and W. Dr6ge, unpublished observation). However, the sensitivity of this method has not been sufficient to detect these derivatives in intact cells. It is believed that the immunoregulatory effects of lactate are largely based on its conversion to pyruvate by the lactate dehydrogenase reaction and the subsequent interaction of pyruvate with cysteine (Fig. 6). This 3o M. K. Gaitonde, Biochem. J. 1114, 627 (1967). 31 S. Bannai and T. Ishii, J. Cell. Physiol. 104, 215 (1980).
[13]
EFFECT OF THIOLS ON T LYMPHOCYTES
149
COOH
L
CH - - NH 2
COOH [ C=O
C=O+
CH 3
I CH 3
I
COOH
COH I H2N--CH HS--CH
I
COOH COOH I [ .~. OH H 2 N - - C H
¢ 2
~ C~. I CH 3
COOH
COOH
I
[ ~.
NH-
CH
)'~" CH a
S _
C]H2
I 'q S
-
CH 2
I CH 3
pyruvate
pyruvate +
lactate
cysteine
~
~.- h e m i m e r c a p t a l
derivative
•
IP thiazolidine
derivative
FIG. 6. Chemical interaction between cysteine and pyruvate/lactate. LDH, Lactate dehydrogenase. For details, see text.
interpretation is supported by the fact that lactate causes a decrease in intracellular glutathione levels and a profound inhibition of typically glutathione-dependent lymphocyte functions. 32 In contrast, the production of IL-2, which is inhibited by high glutathione levels, is markedly augmented by lactate. 7'24'33
Conclusion and Perspectives The available evidence clearly shows that the magnitude of T cell responses is profoundly influenced even by relatively moderate changes of the extracellular cysteine concentration and/or intracellular glutathione level. The limiting role of cysteine is mainly determined by the weak cystine transport activity of lymphocytes and by the relatively low extracellular concentrations of reduced cysteine in blood plasma and tissue culture media. The physiological relevance of this limited and well-regulated supply of cysteine must be seen in the fact that the intact immune system requires a delicate balance between prooxidant and antioxidant conditions. As a rule, prooxidant conditions appear to be beneficial (i) for T cell responses against a weak antigenic stimulus and (ii) for lymphokine (IL-2) production. A particularly strong supply of cysteine and glutathione seems to favor (i) responses against a strong antigenic stimulus and (ii) proliferative T cell responses (i.e., IL-2-dependent DNA synthesis) and the activation of cytotoxic T cells. High GSH concentrations are required especially in the relatively late phase of the immune response, that is, 6
3z S. Roth, H. Gmiinder, and W. Dr6ge, Cell. lmmunol. 136, 95 (1991). 33 W. Dr6ge, S. Roth, A. Altmann, and S. Mihm, Cell. lmmunol. 108, 405 (1987).
150
ASSAY OF STRESS GENES/PROTEINS
[13]
[~M] 4501 400 >, 350 300 "~
250
,.~
200 150
oo
100
~o
50 .9 0
,
0
~
c
_~
J
p
~
,
50 100 150 2 0 0 2 5 0 3 0 0 3 5 0 4 0 0 4 5 0
cysteine added [IJMI
FIG. 7. Determination of cysteine in the presence of pyruvate. For details, see text.
or 8 hr after antigenic stimulation when the activated T cells start to enter the DNA synthetic phase. We have obtained some information about the intracellular concentrations of the key players cysteine and glutathione in lymphoid cells and about the redox states of the corresponding redox couples. We still have little information about the quantitative regulation of redox states in the course of an immune response. Apart from the antigen receptor, there are several cell surface determinants on T cells whose costimulatory effects have already been characterized in much detail. One may predict that at least some of these signals may modulate membrane transport activities and/or enzyme activities such as glutathione reductase, the hexose monophosphate shunt, or glutathione biosynthetic enzymes which ultimately determine the intracellular GSH and GSSG concentrations. We also need to know which of the relevant transport activities and enzyme activities are genetically fixed in all T lineage cells, or which of these activities are subject to regulation by activating signals, and which of these activities simply adapt to microenvironmental conditions. Yodoi and colleagues have provided a substantial body of evidence that the thioredoxin redox couple may also play an important role in the regulation of lymphocyte functions. 34 However, there is still little information about the regulation of the thioredoxin reductase activity which ultimately determines the intracellular concentrations of oxidized 34 j. Yodoi and T. Uchiyama, lmmunol. Today 13, 405 (1992).
[14]
INDUCIBLETRANSCRIPTIONFACTORNF-KB
151
and reduced thioredoxin. Also, the analytical procedures for the quantitative determination of reduced and oxidized thioredoxin still have to be established. Much work in the near future is needed in these areas. Acknowledgment The assistance of I. Fryson in preparation of the manuscript is gratefully acknowledged.
[14] A s s e s s i n g O x y g e n R a d i c a l s as M e d i a t o r s in A c t i v a t i o n o f Inducible Eukaryotic Transcription Factor NF-KB B y RALF SCHRECK a n d PATRICK A. BAEUERLE
Introduction Eukaryotic organisms can undergo dramatic changes in their intracellular concentrations of dioxygen and reactive oxygen intermediates (ROIs). Injury, ischemia, or certain drugs can cause a hypoxic or anoxic condition in cells or tissues. The opposite condition, oxidative stress, can result, for instance, from reperfusion of ischemic tissue, the oxidative burst reaction of neutrophils during inflammatory processes, or the stimulation of cells with inflammatory cytokines and certain drugs. In extreme cases, cells are killed as a consequence of oxidative damage to DNA, proteins, and lipids. During evolution, eukaryotic cells have developed efficient mechanisms to counteract the detrimental effects of dioxygen and ROIs. These include antioxidants such as glutathione and vitamins C and E, as well as a multitude of enzymes specialized in interconverting or eliminating ROI species and repairing damaged protein, DNA, and lipids. This antioxidative apparatus is used to establish a "normoxic" condition in the cell. ' Any perturbation of ROI homeostasis is apparently sensed by regulatory molecules in the cell and triggers reactions reestablishing physiological levels of ROIs. In bacteria, the molecular mechanisms controlling ROI homeostasis are fairly well understood. Two transcription factor systems, called oxyW and soxRS, 2 are specifically activated on exposure of cells to hydrogen i G. Storz, L. A. Tartaglia, S. B. Farr, and B. N. Ames, Trends Genet. 6, 363 (1990). 2 B. Demple, Annu. Rev. Genet. 25, 315 (1991).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[14]
INDUCIBLETRANSCRIPTIONFACTORNF-KB
151
and reduced thioredoxin. Also, the analytical procedures for the quantitative determination of reduced and oxidized thioredoxin still have to be established. Much work in the near future is needed in these areas. Acknowledgment The assistance of I. Fryson in preparation of the manuscript is gratefully acknowledged.
[14] A s s e s s i n g O x y g e n R a d i c a l s as M e d i a t o r s in A c t i v a t i o n o f Inducible Eukaryotic Transcription Factor NF-KB B y RALF SCHRECK a n d PATRICK A. BAEUERLE
Introduction Eukaryotic organisms can undergo dramatic changes in their intracellular concentrations of dioxygen and reactive oxygen intermediates (ROIs). Injury, ischemia, or certain drugs can cause a hypoxic or anoxic condition in cells or tissues. The opposite condition, oxidative stress, can result, for instance, from reperfusion of ischemic tissue, the oxidative burst reaction of neutrophils during inflammatory processes, or the stimulation of cells with inflammatory cytokines and certain drugs. In extreme cases, cells are killed as a consequence of oxidative damage to DNA, proteins, and lipids. During evolution, eukaryotic cells have developed efficient mechanisms to counteract the detrimental effects of dioxygen and ROIs. These include antioxidants such as glutathione and vitamins C and E, as well as a multitude of enzymes specialized in interconverting or eliminating ROI species and repairing damaged protein, DNA, and lipids. This antioxidative apparatus is used to establish a "normoxic" condition in the cell. ' Any perturbation of ROI homeostasis is apparently sensed by regulatory molecules in the cell and triggers reactions reestablishing physiological levels of ROIs. In bacteria, the molecular mechanisms controlling ROI homeostasis are fairly well understood. Two transcription factor systems, called oxyW and soxRS, 2 are specifically activated on exposure of cells to hydrogen i G. Storz, L. A. Tartaglia, S. B. Farr, and B. N. Ames, Trends Genet. 6, 363 (1990). 2 B. Demple, Annu. Rev. Genet. 25, 315 (1991).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
152
ASSAY OF STRESS GENES/PROTEINS
[14]
peroxide and superoxide, respectively. The ROIs apparently modify and thereby induce the activity of the transcriptional activator proteins. The oxyR and soxRS systems activate in response to oxidant stimuli numerous genes encoding antioxidative enzymes protecting cells from oxidative damage. Eukaryotic cells also possess such regulatory mechanisms involving gene transcription. In higher vertebrate cells, expression of genes encoding the antioxidative enzymes thioredoxin (ADF), Mn2+-dependent superoxide dismutase (Mn-SOD), glutathione S-transferase (Ya subunit), and NAD(P)H : quinone reductase is induced by either oxidant or antioxidant stimuli. However, the responsible transcription factor systems have not yet been identified. One eukaryotic transcription factor system specifically activated by peroxides is NF-KB. 3'4 Micromolar concentrations of H202 can mobilize the sequestered cytoplasmic form of NF-KB in cultured cells. This involves release of the regulatory subunit IKB from a heterodimer of DNA-binding p50 and p65 (also called Rel-A) subunits and nuclear translocation of p50-p65. Presumably, oxidants activate protein kinases that triggers dissociation of the NF-KB-IKB by a yet unknown mechanism. Additional evidence that NF-KB is an oxidative stress-responsive transcription factor comes from the inhibitory effects of various structurally unrelated antioxidants on the activation of NF-KB in response to many diverse stimuli. 4,5 In this chapter, we describe in detail how the activation of NF-KB by oxidants and the inhibitory effects of antioxidants on activation of NF-KB are investigated using intact cultured cells.
General Considerations The effects of oxidant/antioxidant treatments on NF-KB can be investigated either in cultured cells or under cell-free conditions. The results are quite contradictory: whereas DNA binding and transactivation by NF-KB is induced by H202 and blocked by antioxidants in intact cells, thiol reagents increase and oxidizing agents abrogate DNA binding of NF-KB when added to nuclear extracts. 6 The in vitro effects are explained by reversible oxidation of a conserved cysteine residue in the DNA-binding
3 R. Schreck, P. Rieber, and P. A. Baeuerle, EMBO J. 10, 2247 (1991). 4 R. Schreck, K. Albermann, and P. A. Baeuerle, Free RadicalRes. Commun. 17, 221 (1992). 5 R. Schreck, B. Meier, D. M/innel, W. Drfge, and P. A. Baeuerle, J. Exp. Med. 175, ll81 (1992). 6 M. B. Toledano and W. J. Leonard, Proc. Natl. Acad. Sci. U.S.A. 88, 4328 (1992).
[14]
INDUCIBLETRANSCRIPTIONFACTORNF-KB
153
domains of p50 and p65. 7 Redox-dependent activity was also observed with many cytosolic enzymes and might simply reflect the circumstance that cells maintain a strongly reducing environment in the cytosol in order to maintain cysteine residues in a biologically active state. The observation that NF-KB is not inactivated but induced on oxidant treatment of intact cells indicates that H202 treatment does not lead to an appreciable oxidation of cysteine residues in cytosolic or nuclear NF-KB under physiological conditions. Therefore, redox regulation of NF-KB activity in uiuo must be achieved by a mechanism fundamentally distinct from the one observed under cell-free conditions. Because so far we have been unsuccessful in activating NF-KB in cell-free systems by treatments with oxidants, current experiments are restricted to intact cells. The activation of NF-KB is most conveniently monitored by the electrophoretic mobility shift assay (EMSA). 8 This highly sensitive method detects the specific DNA-binding activity of NF-KB in extracts from either nuclei or whole cells. It is also advisable to assay nuclear or whole cell extracts with one or more DNA probes detecting DNA-binding activities unrelated to NF-KB. That way, it is possible to test the integrity of the extract and the specificity of an inducing or inhibiting effect. Alternatively or in addition, NF-KB activation is assayed by transfection of reporter genes controlled by cis-acting NF-KB binding sites. The amount of reporter gene product accumulating in a treated cell might better quantitate an effect on transcription than the amount of protein-DNA complex detected in EMSAs. Useful reporter genes are those controlled by the human immunodeficiency virus type 1 (HIV-1) long terminal repeat (LTR) which contains two perfect KB sites. 9 The use of reporter constructs controlled by transcription factors unrelated to NFKB indicate whether a given treatment has a specific inhibitory effect on NF-KB or is simply toxic. Because transcriptional activation and replication of HIV-1 are tightly coupled and dependent on NF-KB, an alternative readout system is the production of virus in latently infected T cells, as determined by p24 enzyme-linked immunosorbent assay (ELISA) or syncytium induction assays. 3 It might also be feasible to test the effects of antioxidants and oxidants by measuring the induction of endogenous NF-KB-controlled genes using Northern blot or reverse polymerase chain reaction (PCR) analysis of mRNA levels or immunodetection assays for 7 j. R. Matthews, N. Wakasugi, J.-L. Virelizier, Y. Yodoi, and R. T. Hay, Nucleic Acids Res. 20, 3821 (1992). 8 j. Carey, this series, Vol. 208, p. 103. 9 G. Nabel and D. Baltimore, Nature (London) 326, 711 (1987).
154
ASSAY OF STRESS GENES/PROTEINS
[14]
the gene products. An obvious disadvantage of the latter method is that one or more of the multiple steps between the activation of a gene and the final appearance of its protein product might also be affected by a given treatment.
Cell Lines and Culture Conditions Most of the data on the redox regulation of NF-KB have been obtained with the human cell lines Jurkat and HeLa. Jurkat is a CD4 ÷ T lymphoma line which has been extensively used for T cell activation studies. We found that Jurkat subclones from various sources are differentially responsive to H202 treatment. 4 The subclone JR (obtained from Professor T. Hiinig, University of Wiirzburg, Germany) is a high responder, whereas a commercially available subclone (ECACC 8805240 I) is much less responsive. However, costimulation with phorbol 12-myristoyl 13-acetate (PMA) showed a synergistic effect with H202 in both subclones and yielded similar amounts of activated NF-KB. HeLa is a epithelioid carcinoma line and is perhaps the most extensively studied cell line available. The HeLa subclone analyzed was strongly responsive to H202, as tested by EMSAs and transactivation assays. 10 In both cell types, NF-KB is almost exclusively present in an inducible form in the cytoplasm and is activated by a variety of stimuli including treatments with PMA, tumor necrosis factor-a (TNF-a), and double-stranded RNA. Jurkat cells are cultured in RPMI 1640 (GIBCO/BRL, Gaithersburg, MD) at densities between 2 × 105 and 1 × 10 6 cells/ml. RPMI 1640 medium was chosen because it is essentially free of iron salts that could catalyze decomposition of added H202 by the Fenton reaction or interfere with the antioxidant effect of metal chelators. The adherent HeLa cells are cultured in iron-flee minimal essential medium (MEM; GIBCO/BRL). The media are supplemented with heat-inactivated 10% fetal calf serum (FCS) and antibiotics (penicillin/streptomycin). We noted that the extent of NF-KB activation in response to H202 showed variations with the serum batch. This could be related to the content of catalases or free metal ions in a given batch. Cells are cultured at 37°C and 5% CO2 in a humidified atmosphere. The evening before an experiment, cells are supplied with flesh medium and kept in log phase during the experiment (i.e., 6-8 × 105 cells/ml or 60-75% confluency). To obtain maximal effects of oxidants and antioxidants, only cells with a low passage number should be used.
10 M. Meyer, R. Schreck, and P. A. Baeuerle, E M B O J. 12, 2205 (1993).
[14]
INDUCIBLETRANSCRIPTIONFACTORNF-KB
155
Treatment of Cells with Hydrogen Peroxide and Other Oxidants The activation of NF-rB by' H202 may show a cell type-specific dose-response. NF-KB DNA binding was optimally induced in the Jurkat subclone JR between 100 and 150 tzM H202,3 whereas transactivation of a HIV-1 LTR-controlled CAT (chloramphenicol acetyltransferase) reporter gene responded optimally between 30 and 50/zM H202.1° The difference might be due to distinct incubation periods. Although DNA binding of NF-KB can be detected by EMSAs within 30 min following addition of 150 IzM H202 to the culture medium, transactivation assays usually require much longer incubation with the oxidant to allow for accumulation of the reporter gene product. Therefore, a toxic effect of higher H20: concentrations might become more evident in the transactivation assay than in the EMSA. HeLa cells require slightly higher H202 concentrations for maximal stimulation of NF-KB than Jurkat T cells. Approximately 200-250/zM H202 gave optimal induction of DNA binding as well as expression of a CAT reporter gene controlled by two KB sites) ° In the monocytic cell line Monomac-6, NF-KB was not detectably induced even at high H202 concentrations. H It is possible that cells capable of a respiratory burst reaction are generally less susceptible to oxidants because of a more efficient self-protection from ROIs. In view of the cell type-specific differences in H202 susceptibility, it is advisable to test the effect of H202 for a given cell line in a concentration range between 10 and 500 IzM. As a source of H202 we use dilutions of a 30% solution of H202 (Perhydrol; Merck, Darmstadt, Germany) in deionized, freshly tapped Millipore (Bedford, MA) water. Alternatively, H202 can be generated enzymatically in the medium by addition of a glucose/glucose oxidase system. The latter might provide a more continuous but less defined supply of the oxidant rather than a single "pulse" with a known concentration of H202. We do not know the half-life of H:O: in our cell cultures but expect it to vary with serum batch, cell type, and degree of cell lysis which could liberate catalases. Whereas all this can be determined, it is much more difficult to assess the effective intracellular concentration of H202 activating NF-KB. Because some H202 solutions contain stabilizing agents, we recommend the use of catalase to control an H202 effect. We observe that 200 U of catalase per milliliter of medium prevents the induction of NF-KB by H202. It is also advisable to include a positive control for NF-KB activation, for example, a treatment with 200 U/ml TNF-a or 50 ng/ml phorbol ester (PMA). " Dr. L6ms Ziegler-Heitbrock, University of Munich, personal communication.
156
ASSAY OF STRESS GENES/PROTEINS
[14]
To test whether NF-rB is specifically activated by H202, we tested various agents that have been reported to lead to production of other ROI species, including superoxide and singlet oxygen. 4 With the exception of 300/zM butyl peroxide, none of the agents could activate NF-rB at various incubation periods and concentrations tested, suggesting that NF-KB is a peroxide-inducible factor, as has been shown for the bacterial oxyR protein. Treatments of Cells with Antioxidants Before testing the inhibiting effect of a novel antioxidant on NF-KB activation or the effect of established antioxidants on NF-KB activation by a novel inducer, a cell line should be tested with known activators or antioxidants, respectively. Induction of NF-KB by all inducers tested so far was suppressed by antioxidants. 4 Therefore, there is a choice among many inducers. In most cell types, TNF-a very strongly activates NF-KB. In appropriate cell lines, interleukin 1 (IL-1), bacterial lipopolysaccharide (LPS), active phorbol esters, or a combination of phorbol esters with lectins or calcium ionophores also potently induce the factor. Antioxidative chemicals might have pleiotropic effects. Any information concerning their chemical and biological properties, toxicity, solubility, membrane permeability, and turnover should be considered. Antioxidants might act by (i) directly scavenging radicals, (ii) chelating metal ions required for the Fenton reaction, (iii) sustaining the activity of antioxidative enzymes, or (iv) inhibiting the activity of oxidizing enzymes. A good indication that the effect of an agent relies on its antioxidative activity is if a second chemically and structurally unrelated antioxidative compound shows a similar effect. We therefore recommend testing more than one agent in cell cultures when antioxidants are used to investigate the role of ROIs in novel NF-rB-inducing conditions. NF-KB activation in response to phorbol ester or TNF-a treatment can be suppressed with different classes of compounds4: dithiocarbamates, the metal chelators desferroxamine and o-phenanthroline, cysteine and derivatives, vitamin E, quinones, and a-lipoic acid. 12 If an inhibiting effect is observed with at least two agents belonging to different classes, it is very likely that a novel inducer acts via a ROI-dependent pathway. In addition to these types of experiments, it should be demonstrated directly that an inducer elicits the production of H202 or superoxide in stimulated cells. The kinetics of H202 and superoxide production can be determined by chemical or enzymatic analysis of cell culture supernatants. 5 J2 y. 1. Suzuki, B. B. Aggarwar, and L. Packer, Biochem. Biophys. Res. Commun. 189, 1709 (1993).
[14]
INDUCIBLETRANSCRIPTIONFACTORNF-KB
157
In our cell culture experiments, pyrrolidine dithiocarbamate (PDTC) (and other dithiocarbamates) turned out to be the most potent inhibitor of NF-KB activation. 5 This might come from the triple activity of the compounds: first, they chelate free metal ions catalyzing decomposition of H202 into hydroxyl radicals; second, they react directly with and neutralize radicals; and third, they inhibit cytosolic superoxide dismutase, thereby preventing the generation of H2Oz from superoxide. In Jurkat cells, 100-150/zM PDTC almost completely suppresses NF-KB activation by PMA as detected by EMSAs and CAT assays. ]° In HeLa cells, 60 ~M PDTC is required. 3 For use in cell cultures, the ammonium salt of PDTC (Sigma) is freshly dissolved in degassed phosphate-buffered saline (PBS) as a 1 M stock solution. We observed that iron decreases the inhibitory effect of PDTC and therefore recommend avoiding the use of a metal spatula. Cell cultures are pretreated for 30-60 min with 10-250 ~M PDTC. The compound seems to enter the cell very quickly because it shows some effect even when applied at the same time as the inducer. 5 Incubations with 100-150 /zM PDTC exceeding 10 hr may lead to desensitization of cells for PDTC. No significant cell lysis or decrease in cell number is observed when Jurkat cells are treated with 250 ~M PDTC for up to 24 hr. Other dithiocarbamates tested were more toxic. Because activation of NF-KB is a rapid posttranslational process, long-term incubations with PDTC are not required. PDTC shows a biphasic dose-response curve as observed for other antioxidants; compared to micromolar concentrations, millimolar concentrations of PDTC are no more inhibitory on NF-KB activation by TNF-a. 5 Another frequently used antioxidant, N-acetylcysteine (NAC), is dissolved in water and adjusted to pH 7.4 by the addition of 1 N NaOH. Up to 30 mM NAC is required to see strong inhibitory effects on NF-KB activation, o-Phenanthroline and desferroxamine are also potent in the micromolar range but require much longer preincubation periods than PDTC or NAC (several hours). Overexpression of Enzymes A relatively specific way to interfere with the metabolism of ROIs is via the overexpression of enzymes converting or eliminating ROIs. Overexpression can be either transient or stable. Low transfection efficiencies in transient expression experiments may allow monitoring NFKB activation only by transactivation assays, not by EMSAs. In these experiments, an expression vector coding for an enzyme is cotransfected with a KB-controlled reporter construct or with control vectors. Stable expression requires selection for cell lines containing stably integrated copies of the transgene under constitutive or inducible transcriptional
158
ASSAY OF STRESS GENES/PROTEINS
[14]
control, and subsequent biochemical analysis to verify the presence in a cell line of enhanced levels of a particular enzymatic activity. An advantage of stable lines is that every cell in the culture overexpresses the enzyme and will show an effect on stimulation by an inducer in both EMSAs and transactivation assays. We have investigated the effect of Mn-SOD overexpression on activation of NF-KB in a stably transfected human epithelial breast carcinoma cell line, MCF-7, and a nontransfected parental line. 5 Activation of NFKB in response to TNF-a, IL-1, and PMA was not inhibited, supporting the idea that superoxide is not directly involved. Several other enzymes should be tested in this way, including catalases, various peroxidases, thioredoxin, and combinations thereof. Preparation of Cell Extracts To exploit fully the analytical potential of EMSAs, great care and attention should be paid to the preparation of protein extracts from cells treated with oxidants and antioxidants. Most important is a quick preparation providing an efficient inhibition of proteases. We use two types of protein extracts: whole cell extracts and nuclear extracts. The advantage of whole extracts is that far fewer cells are required for an analysis and the preparation is very simple and rapid, thus allowing one to screen for many parameters within short time. The advantage of nuclear extracts is that they usually show less background DNA-binding activities in EMSAs and that additional information about the subcellular distribution of NFKB is provided. Preparation of nuclear extracts requires some experience in the subcellular fractionation of cells, is more time-consuming, and consumes many more cells than the preparation of whole cell extracts. For whole cell extracts, cells are lysed with nonionic detergent and proteins extracted by salt from chromatin and other cellular structures. Approximately 2 × 105 cells are sufficient to prepare an extract, but an experienced experimenter might use less. Cells are collected from the medium by centrifugation for 10 min at 500 g and 4 °. The cell pellet is resuspended in 1 ml of ice-cold PBS and cells sedimented again by centrifugation. The PBS is carefully removed as completely as possible from the tube with a drawn-out glass pipette under mild vacuum. The cell pellet is resuspended in 10 /A lysis buffer [20 mM H E P E S - K O H , pH 7.5, 350 mM KCI, 1 mM MgC12, 0.5 mM EDTA, 0.1 mM EGTA, 5 mM dithiothreitol (DTT), 10/zg/ml leupeptin, 10/~g/ml aprotinin, 0.5% (v/v) of a saturated phenylmethylsulfonyl fluoride (PMSF) solution in ethanol, 20% (v/v) glycerol, and 1% (v/v) Nonidet P-40 (NP-40)] per 105 cells.
[14]
INDUCIBLETRANSCRIPTIONFACTORNF-KB
159
After 15 min on ice, the lysate is centrifuged in a cooled microcentrifuge for 20 min at 15,800 g. The resulting supernatant is diluted with l volume of buffer H [20 mM H E P E S - K O H , pH 7.9, 0.2 mM EDTA, and 20% (v/v) glycerol] in order to reduce the salt concentration. The average protein concentration of whole cell extracts is between 8 and 15/xg/ml, as determined by a Bradford microassay procedure (Bio-Rad, Richmond, CA). Equal amounts of protein (10-30/zg) are immediately used in DNAbinding reactions for EMSAs. Remaining extract may be quick-frozen in liquid nitrogen and stored at - 70°. Because some DNA-binding activities are reduced on repeated freeze-thawing, extracts should be stored and used in aliquots. Nuclear extracts are prepared according to Dignam et al. ~3with modifications as follows. Cells (not less than 10 7) a r e collected by centrifugation (10 min, 500 g, 4 °) and washed once in ice-cold PBS followed by a second centrifugation. The PBS is carefully removed from the cell pellet and cells resuspended in 160 ml buffer A (I0 mM H E P E S - K O H , pH 7.9, 10 mM KCI, 1.5 mM MgC12 , 2 mM DTT, and 4/xg/ml aprotinin). One microliter of a saturated solution of PMSF in ethanol is added and then added a second time after a 10-min incubation on ice. Cells are transferred into a cooled small-scale glass homogenizer (Dounce type B pestle, Kontes, Vineland, NJ) and lysed by 10-15 strokes. Lysis conditions have to be established for a given cell type. Warming of the lysate and air bubbles are avoided during homogenization. The degree of cell lysis, which is monitored by phase-contrast microscopy, should not exceed 80% because from this point on nuclei also start to disintegrate. The homogenate receives 1/~l of PMSF stock solution and is centrifuged for 5 sec in a microcentrifuge to pellet nuclei. The supernatant is recentrifuged at 4 ° for 20 min at 15,800 g. The resulting supernatant (cytoplasm) receives 10% (v/v) glycerol and may be analyzed by EMSA for sequestered NF-KB following a treatment with deoxycholate. ~4 The nuclear pellet obtained after the 5-sec spin is thoroughly resuspended using a cut microtip in 60/zl of the high-salt extraction buffer B [20 mM H E P E S - K O H , pH 7.9,400 mM KCI, 0.1 mM EDTA, 25% (v/v) glycerol]. A wash of the nuclear pellet in buffer A is optional. The suspension is incubated on ice for 20 min, preceded and followed by addition of I/zl of PMSF solution. The extract is cleared by centrifugation (20 min at 4 ° and 15,800 g). The resulting supernatant is diluted with l volume buffer D containing 0.25% (v/v) NP-40. Nuclear extracts are immediately used t3 D. Dignam, R. M. Lebovitz, and R. G. Roeder, Nucleic Acids. Res. 11, 1475 (1983). 14 p. A. Baeuerle and D. Baltimore, Cell (Cambridge, Mass.) 53, 211 (1988).
160
ASSAY OF STRESS GENES/PROTEINS
[14]
for EMSAs or are frozen in liquid nitrogen and stored at - 7 0 °. Average protein concentrations of nuclear extracts are between 1 and 3/~g/ml and for cytoplasmic fractions between 4 and 12 ~g//zl. Monitoring NF-rB Activity by Electrophoretic Mobility Shift Assays The EMSA method is based on the specific binding of protein to a 32p-labeled DNA fragment, followed by separation of the protein-DNA complex from unbound excess DNA on a low percentage native polyacrylamide gel. 8 The detection limit of the method is in the femtomole (fmol) range and relies on the high specific activity of 32p-labeled ligands which, in the case of nucleic acid molecules, can be varied by the labeling procedure. For the detection of the DNA-binding activity of NF-KB in whole cell extracts and nuclear extracts, we use double-stranded oligonucleotides containing the high-affinity binding motif 5'-GGGACTTTCC-3'. 15 For the detection of unrelated DNA-binding activities serving as internal controls, we recommend probes detecting octamer-binding proteins (OTFs), Spl, or the cAMP-response factor (CREB). Oligonucleotides are labeled either by the Klenow enzyme filling in 5'-overhangs with a-32p-labeled deoxynucleoside triphosphates (dNTPs) or by T4 polynucleotide kinase phosphorylating free 5'-hydroxy groups of ribose residues using [y-32p]ATP. Complexes of NF-KB with DNA have a characteristic low mobility in 4% polyacrylamide gels. Nonspecific DNA-binding activities in extracts can be efficiently blocked by inclusion of microgram amounts of the competitor poly(dI-dC). This nucleic acid barely interferes with NF-KB binding to the specific radioactive probe.15 The identity of a complex with an NF-KB-DNA complex can be verified by the following criteria. (1) The complex should appear de n o v o following treatments known to activate NF-KB and should comigrate with a defined NF-KB complex in native gels. (Some cell types might contain constitutive NF-KB, such as B lymphoma cells.) (2) It should exhibit KB-specific DNA binding which can be demonstrated by various techniques. Most convenient are competition experiments using excess unlabeled specific and nonspecific DNA fragments. They are mixed with the labeled DNA probe before the addition of extract. More elaborate are methylation interference assays. (3) The complex should immunoreact with at least one antibody specific for a NFKB subunit. Monoclonal antibodies are commercially available (Santa Cruz Biotechnology, Santa Cruz, CA), but we do not know whether they alter the mobility of NF-KB complexes in EMSAs. 15 U. Zabel, R. Schrcck, and P. A. Baeuerle, J. Biol. Chem. 266, 252 (1991).
[14]
INDUCIBLETRANSCRIPTIONFACTORNF-KB
161
There are now three more proteins known to contain the DNA-binding and dimerization domain shared by p50 and p65 (Rel-A): c-Rel, p52, and Rel-B. 16 In vitro, heterotypic dimerization is observed among the five protein subunits, but it is not known to what extent this occurs in intact cells and what the DNA-binding specificity of dimers are. Future studies might require that N F - r B - D N A complexes be investigated in more detail with regard to subunit composition. However, using the classic rB motif 5'-GGGACTTTCC-3', only p50-p65 complexes were purified from placenta and lung tissue. It is thus possible that other subunit combinations have a distinct binding specificity and are not readily detected under the EMSA conditions described here. The EMSA conditions for human NF-KB were described in detail. ~5 Binding reactions are carried out in a final volume of 20/xl using 1-5/zl of whole cell extract, nuclear extract, or cytoplasm. The reaction is started by adding protein extracts to a DNA-binding mix containing 1-2.5/zg of poly(dI-dC), 20/xg bovine serum albumin (BSA) as a carrier, 2/zl of a l0 × binding buffer [100 mM Tris-HC1, pH 7.5, 500 mM NaCl, l0 mM EDTA, 50% (v/v) glycerol, 2 mM DTT, and 2/zl of 1% (v/v) NP-40], and approximately 10,000 cpm (Cerenkov counting) of a 32p-labeled DNA probe with a specific activity not less than 3000 Ci/mmol (approximately 0.1 ng of a 30-mer double-stranded oligonucleotide). The DNA binding of NF-rB is optimal in the presence of 100 mM NaC1 or KCI. Water or NaCI/KC1 solutions are added to the DNA binding mix such that the final volume of a single reaction including the extract volume is 20 /xl and has a final concentration of approximately 100 mM Na + or K +. Binding reactions proceed for 30 rain on ice. To test for the DNA binding specificity, known molar excesses of competitor oligonucleotides are added to the DNA binding mix prior to addition of protein extracts. Binding reactions are loaded on a prerun (30 min at 150 V) vertical polyacrylamide gel, and electrophoretic separation of protein-DNA complexes from unbound DNA is started immediately after loading of gels with a Hamilton syringe. Gels (220 x 160 x 15 ram) are polymerized from 4% (w/v) acrylamide and 0.1% (w/v) bisacrylamide in low ionic strength electrophoresis buffer (6.8 mM Tris-HCl, pH 7.5, 3.4 mM sodium acetate, and 1 mM E D T A - N a O H , pH 7.5). During electrophoresis, the electrophoresis buffer is constantly recirculated using a peristaltic pump. Electrophoresis is performed at 180-220 V for approximately 2 hr. It is stopped when a bromphenol blue dye marker has migrated 10 cm. Glass plates are then placed horizontally and the top plate is slowly lifted with the help of a spatula. A double layer of Whatman (Clifton, N J) 3MM 16 V. Blank, P. Kourilsky, and A. Israel, Trends Biochem. Sci. 17, 135 (1992).
162
ASSAY OF STRESS GENES/PROTEINS
[14]
paper is carefully laid on top of the gel on the bottom plate. The sandwich is inverted and the glass plate lifted off. The gel, which is stuck to the filter paper, is covered with Saran wrap and immediately dried at 80 ° under vacuum. The dried gel is exposed to Kodak (Rochester, NY) XAR film at - 7 0 ° for 6-24 hr. Monitoring NF-KB Activity by Transactivation Assays Transactivation assays measure the activation of the NF-KB transcription factor by the expression of reporter genes controlled by well-characterized, usually reiterated cis-acting KB elements. Frequently used reporter proteins are chloramphenicol acetyltransferase (CAT) and luciferase. We use either a construct with two copies of the 5'-GGGACTTTCC-3' KB motif upstream of a minimal thymidine kinase (tk) promoter 1°or the HIV-1 LTR controlling a CAT gene. 3,9 As a specificity control, constructs with the tk promoter alone or an HIV-1 LTR with mutated NF-KB sites are used in parallel experiments. To test for nonspecific effects of oxidant/antioxidant treatments on transactivation, reporter constructs controlled by strong constitutive enhancers and promoters are used, for instance, the elongation factor-la promoter or the Rous sarcoma virus LTR. Both elements presumably act independently of NF-KB. Cells can be transfected with reporter constructs by various techniques. ~7Often used are DEAE-dextran, calcium phosphate precipitation, liposomes, or electroporation. It should be verified that a transfection method by itself does not induce NF-KB. Depending on the method and cell type, transfection efficiencies can be quite variable. Transfected cells are harvested after 36-48 hr. Treatments with antioxidants, oxidants, and other inducers are performed 5-10 hr prior to harvesting of cells. Longer treatments might affect the viability of transfected cells. Cell extracts are prepared and the amount of accumulated reporter gene product determined according to published procedures.17 Perspective The procedures described in this chapter are particularly useful in achieving two goals. The first is to find novel inhibitors of NF-KB that may be more potent antioxidative drugs interfering with the release of IKB from the cytoplasmic complex of NF-rB. The experimental approaches allow screening for other types of inhibitors affecting, for in17W. A. Keown, C. R. Campbell, and R. S. Kucherlapati, this series, Vol. 185, p. 527.
[15]
c-Fos AND c-Jun REDOX-DEPENDENTDNA BINDING
163
stance, DNA binding or dimerization properties of NF-KB. In view of the important biological roles of NF-KB, 18'19 inhibitors of the transcription factor might find broad clinical application as anti-inflammatory and immunosuppressive drugs and, perhaps, as inhibitors of acute-phase response, allograft rejection, and malignant growth. The second goal is to understand molecular mechanisms underlying the oxidative stress response in higher eukaryotes, that is, to determine what molecules sense a disturbance of the intracellular levels of ROIs and how they transmit their signals to the nucleus where genes are newly transcribed. The methods presented are not restricted to analyzing NF-KB, but have also been applied to investigate the unrelated inducible transcription factor AP-1.~° Treatments of cell cultures with oxidants and antioxidants and sensitive methods to detect transcription factor activities provide important tools for analyzing ROI-controlled gene regulation in general. Acknowledgments We are indebted to Dr. Heike Pahl for helpful comments on the manuscript. This work was supported by grants from the Bundesministerium for Forschung und Technologie and the Deutsche Forschungsgemeinschaft (SEB 217) awarded to P.A.B. 18 p. A. Baeuerle and D. Baltimore, in "Molecular Aspects of Cellular Regulation, Hormonal Control of Gene Transcription" (P. Cohen and J. G. Foulkes, eds.), Vol. 6, p. 409. Elsevier, Amsterdam. 19 S. Grimm and P. A. Baeuerle, Biochem. J. 290, 297 (1993).
[15] A n a l y s i s o f c-Fos a n d c - J u n R e d o x - D e p e n d e n t D N A Binding Activity
By STEVEN XANTHOUDAKISand TOM CURRAN Introduction The control of gene expression by sequence-specific DNA binding proteins is mediated, in part, by regulatory mechanisms that involve posttranslational modification. Insights into these modes of regulation have been facilitated in recent years because of the increased availability of purified recombinant transcription factors. Studies of eukaryotic transcription factors overproduced and isolated from bacterial sources have helped to identify posttranslational modifications that influence DNA binding activity. METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All fights of reproduction in any form reserved.
[15]
c-Fos AND c-Jun REDOX-DEPENDENTDNA BINDING
163
stance, DNA binding or dimerization properties of NF-KB. In view of the important biological roles of NF-KB, 18'19 inhibitors of the transcription factor might find broad clinical application as anti-inflammatory and immunosuppressive drugs and, perhaps, as inhibitors of acute-phase response, allograft rejection, and malignant growth. The second goal is to understand molecular mechanisms underlying the oxidative stress response in higher eukaryotes, that is, to determine what molecules sense a disturbance of the intracellular levels of ROIs and how they transmit their signals to the nucleus where genes are newly transcribed. The methods presented are not restricted to analyzing NF-KB, but have also been applied to investigate the unrelated inducible transcription factor AP-1.~° Treatments of cell cultures with oxidants and antioxidants and sensitive methods to detect transcription factor activities provide important tools for analyzing ROI-controlled gene regulation in general. Acknowledgments We are indebted to Dr. Heike Pahl for helpful comments on the manuscript. This work was supported by grants from the Bundesministerium for Forschung und Technologie and the Deutsche Forschungsgemeinschaft (SEB 217) awarded to P.A.B. 18 p. A. Baeuerle and D. Baltimore, in "Molecular Aspects of Cellular Regulation, Hormonal Control of Gene Transcription" (P. Cohen and J. G. Foulkes, eds.), Vol. 6, p. 409. Elsevier, Amsterdam. 19 S. Grimm and P. A. Baeuerle, Biochem. J. 290, 297 (1993).
[15] A n a l y s i s o f c-Fos a n d c - J u n R e d o x - D e p e n d e n t D N A Binding Activity
By STEVEN XANTHOUDAKISand TOM CURRAN Introduction The control of gene expression by sequence-specific DNA binding proteins is mediated, in part, by regulatory mechanisms that involve posttranslational modification. Insights into these modes of regulation have been facilitated in recent years because of the increased availability of purified recombinant transcription factors. Studies of eukaryotic transcription factors overproduced and isolated from bacterial sources have helped to identify posttranslational modifications that influence DNA binding activity. METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All fights of reproduction in any form reserved.
164
ASSAY OF STRESS GENES/PROTEINS
Fos FosB Frol FroZ
RRERNKMAA~!KCR~RRRELT RRERNKLAA~IKCR~RRRELT RRERNKLAAAKCRINRRKELT RRERNKLAAAKCRINRRRELT
3un 3unB 3unD
KRMRNRIAASKCRKRKLERI KRLRNRLAATKCRKRKLERI KRLRNRIAA: KCRKRKLERI
[15]
FIG. 1. Sequence alignment of the basic region of different Fos- and Jun-related proteins. The amino acid sequences corresponding to the basic region (DNA binding domain) of several Fos and Jun family members are aligned. The position of the conserved cysteine residue and flanking amino acids is indicated by the box.
In studying the properties of recombinant Fos and Jun, purified from bacteria, we discovered that their DNA binding activity was dependent on their redox state.l'2 These proteins form dimeric complexes that bind to activator protein-I (AP-1) binding sites. 3,4 A bipartite DNA binding domain is formed as a direct consequence of dimerization through the leucine zipper motif. 4-6 Redox regulation of Fos and Jun maps to a single conserved cysteine residue flanked by basic amino acids in the DNA binding domain of the proteins (Fig. 1). 7 Reduction of the critical cysteine residue by chemical reducing agents (e.g., dithiothreitol and 2-mercaptoethanol) or by a nuclear protein, Ref-1, present in nuclear extracts from mammalian cells, can convert the inactive (presumably oxidized) form of recombinant Fos and Jun to an active state that is permissive for DNA binding. 7-9 Biochemical analysis indicates the reduced state of the cysteine residue is a sulfhydryl based on its sensitivity to various modifying agents. 7 However, the oxidation state of the cysteine has not yet been identified, but it does not involve the formation of a disulfide bond. 7 i C. Abate, D. Luk, and T. Curran, Cell Growth Differ. 1, 455 (1990). 2 C. Abate, D. Luk, R. Gentz, F. J. Rauscher, III, and T. Curran, Proc. Natl. Acad. Sci. U.S.A. 87, 1032 (1990). 3 T. Curran and B. R. Franza, Jr., Cell (Cambridge, Mass.) 55, 395 (1988). 4 F. J. Rauscher, III, P. J. Voulalas, B. R. Franza, Jr., and T. Curran, Genes Dev. 2, 1687 (1988). 5 T. K. Kerppola and T. Curran, Curr. Opin. Struct. Biol. 1, 71 (1991). 6 W. H. Landschulz, P. F. Johnson, and S. L. McKnight, Science 240, 1759 (1988). 7 C. Abate, L. Patel, F. J. Rauscher, III, and T. Curran, Science 249, 1157 (1990). 8 S. Xanthoudakis and T. Curran, EMBO J. 8, 653 (1991). 9 S. Xanthoudakis, G. Miao, F. Wang, Y.-C. E. Pan, and T. Curran, EMBO J. 1!, 3323 (1991).
[15]
c-Fos AND c-Jun REDOX-DEPENDENTDNA BINDING
165
In the following chapter we describe the method used to assay the redox-dependent DNA binding activity of Fos and Jun in vitro. The general strategy involves established procedures that have been modified for our assay. The procedures used to prepare and manipulate recombinant Fos and Jun for redox analysis are summarized in detail, as are the DNA binding and gel retardation assays. ~°,N Overview of Procedure To study redox regulation of Fos and Jun, the proteins are prepared under conditions in which they can be reversibly shuttled between the oxidized and reduced forms. Oxidation of the proteins (prepared initially under reducing conditions) is readily achieved by lowering the concentration of reducing agent in the sample either through dialysis or by dilution of the protein to working concentrations in a buffer that lacks or contains low levels (<0.2 raM) of reducing agents. An aliquot of the oxidized proteins is then incubated under reducing or nonreducing conditions together with a radiolabeled oligonucleotide containing an AP-1 site. Protein-DNA complexes are analyzed on nondenaturing polyacrylamide gels using standard gel retardation assays. 10m Under these conditions efficient DNA binding to the AP-1 probe is obtained only if an appropriate reducing agent is incubated together with Fos and Jun prior to the addition of the AP-1 oligonucleotide. In the absence of a reducing agent, Fos and Jun remain in an oxidized state and consequently fail to bind to DNA with high affinity. Reduction of Fos and Jun by Cellular Proteins In the absence of chemical reducing compounds such as dithiothreitol or 2-mercaptoethanol, reduction can be catalyzed by activities present in cellular extracts. ~'2 Analysis of different cell lines has shown that the nuclear fraction of a cell extract is a major source of AP-1 reducing activity.~ Nuclear extracts are prepared according to standard protocols, and AP-1 reducing activities are generally detectable even in crude preparations.~.tz The major Fos-Jun reducing protein in HeLa nuclear extracts, designated redox factor-1 (Ref-1), has been isolated and cloned from human cells. 8'9 The redox assay described here was used to follow the purification of Ref-1 throughout the fractionation procedure. 8 Recombinant 10 M. M. Garner and A. Revzin, Nucleic Acids Res. 9, 3047 (1981). H j. Carey, this series, Vol. 208, p. 103. 12 M. R. Briggs, J. T. Kadonaga, S. D. Bell, and R. Tjian, Science 234, 47 (1988).
166
ASSAY OF STRESS GENES/PROTEINS
[15]
Ref-1, expressed and isolated from bacteria, was also found to stimulate the DNA binding activity of Fos and Jun as well as that of other redoxdependent transcription factors. 9 In addition, analysis of the cloned ref-1 gene product has revealed that Ref-1 is a bifunctional protein with an additional activity acting as an apurinic/apyrimidinic endonuclease in DNA repair. 9 Interestingly, the AP-I reducing potential of Ref-1 is influenced by its redox state. Oxidation of Ref- 1 selectively inhibits its redox activity without altering its DNA repair activity. 9 Following oxidation, the AP-I reducing activity of Ref-1 can be regenerated, at least in vitro, using a thioredoxin reducing system. 9 However, thioredoxin alone fails to stimulate binding of Fos and Jun to DNA.7'9 Given the susceptibility of Ref-I to inactivation by oxidation it is necessary to exercise caution when manipulating the protein or analogous activities. It is therefore recommended that any protein preparations used for the experiments described here be handled under reducing conditions. Nevertheless, when assaying the reducing activity of a protein preparation, the concentration of reducing agents contributed by the buffer in which the protein is stored should not be so high as to give a false-positive result in the binding assay. As a rule, the final concentration of chemical reducing agents contributed by any source should not exceed 0.2 mM in the DNA binding reaction.
Bacterial Expression and Purification of Recombinant Fos and Jun A number of bacterial expression systems have been developed that allow efficient overproduction of recombinant proteins. We routinely use an inducible pDS56-6xHis/RBSII vector for overexpression of Fos and Jun (and derivatives thereof) in Escherichia coli. 2'13'14 This vector fuses six contiguous histidine residues to the N terminus of the recombinant protein. Bacterial cells containing a plasmid encoding the Lac repressor are transformed with the expression vector and treated with isopropyl-/3-Dthiogalactoside during mid-logarithmic growth to induce expression from a bacteriophage T5 promoter under the control of a lac operator. 13.14Following induction, cultures are harvested and lysed in a guanidine hydrochloride buffer. Cellular debris is removed from the lysate by centrifugation, and the solubilized fraction containing the recombinant protein is fractionated by nickel-chelate chromatography using a multistep guanidine pH I3 R. Gentz, C.-H. Chen, and C. A. Rosen, Proc. Natl. Acad. Sci. U.S.A. 86, 821 (1989). ~4H. Bujard, R. Gentz, M. Lanzer, D. Stueber, M. Mueller, I. Ibrahimi, M. T. Haeuptle, and B. Dobberstein, this series, Vol. 155, p. 416.
[15]
c-Fos AND c-Jun REDOX-DEPENDENTDNA BINDING
167
gradient. 13,15Guanidine is slowly removed from the preparation by stepwise dialysis against a series of buffers containing decreasing concentrations of guanidine. The renatured protein is generally greater than 95% pure as determined by sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) analysis. A complete expression-purification kit (QIAexpress) is commercially available through QIAGEN Inc. (Chatsworth, CA). QIAGEN will also provide, on request, a detailed instruction manual that outlines the theoretical and practical applications of this system. The main advantage of the expression-purification strategy is that it combines high-level expression with the convenience of a rapid single-step purification. Fos and Jun proteins purified under denaturing conditions maintain their characteristic DNA binding properties as determined by gel retardation, DNA footprinting, and UV cross-linking analysis and are functionally active in in vitro transcription a s s a y s . 2,8'16A7 Most importantly, the proteins maintain their redox properties. However, a potential drawback to this procedure is the problem of protein renaturation during dialysis. In the case of Fos, and in particular Jun, there is occasionally a tendency for the proteins to precipitate from solution, although this is less of a factor when working with various truncated forms of the proteins. The problem of precipitation during dialysis can be partly circumvented by stepping down the concentration of guanidine in the dialysis buffer in a more gradual fashion. This is achieved by increasing the number of steps and dialyzing against a shallower guanidine gradient. In addition, we routinely add detergent and reducing agents to the dialysis buffer to help stabilize the proteins throughout the process. Recovery can vary between 10 and 80% depending on the protein, although the total protein yield is generally in the range of 1-5 mg of protein per liter of bacterial culture. At least for Fos and Jun, these yields are considerably higher than those obtained through purification under nondenaturing conditions, in which the proteins often fractionate with the insoluble component of the crude bacterial lysate. Materials and Reagents
Bacterial strains: E. coli strain M15 ~8 (transformed with plasmid PDMI1 encoding the Lac repressor, Kan R ~3 Parental expression vector: pDS56-x6His/RBSII; IPTG-inducible, AmpR 13,14 15 E. Hochuli, H. Doebeli, and A. Schacher, J. Chromatogr. 411, 177 (1987). 16 C. Abate, D. Luk, E. Gagne, R. G. Roeder, and T. Curran, Mol. Cell. Biol. 10, 5532 (1990). i7 C. Abate, D. Luk, and T. Curran, Mol. Cell. Biol. 11, 3624 (1991). Is M. R. Villarejio and I. Zabin, J. Bacteriol. 110, 171 (1974).
168
ASSAY OF STRESS GENES/PROTEINS
[15]
Expression vector pDS56Junl-334: Full-length rat c-jun coding sequences (amino acids 1-334) cloned into the SphI-BamHI site of pDS56-x6His/RBSI116 Expression vector pDS56Fosl-380: Full-length rat c-fos coding sequences (amino acids 1-380) cloned into the SphI-BamHI site of pDS56-x6His/RBSII16; the cloned c-fos cDNA has been reconstructed using a strategy involving the polymerase chain reaction (PCR), without altering the primary amino acid structure, to optimize for codon usage in E. coli 2 Isopropyl-fl-D-thiogalactoside (IPTG): Prepared as a 1 M stock in distilled water and sterilized by filtration Buffer A: 6 M guanidine hydrochloride, 10 mM 2-mercaptoethanol, 25 mM Na2HPO4/NaH2PO4, pH 8.0 Buffer B: 6 M guanidine hydrochloride, 10 mM 2-mercaptoethanol, 25 mM Na2HPO4/NaH2PO4, pH 6.5 Buffer C: 6 M guanidine hydrochloride, 10 mM 2-mercaptoethanol, 25 mM Na2HPO4/NaH2PO4, pH 5.0 Buffer D: 25 mM NazHPO4/NaH2PO4, pH 7.0, 1 mM dithiothreitol (DTT), 1 mM EDTA, 0.01% Nonidet P-40 (NP-40), 5% glycerol, guanidine hydrochloride (variable concentration of 0, 0.1, 1.0, 3.0 M) Nickel-chelate resin: Ni2+-NTA (nitrilotriacetic acid)-agarose (QIAGEN, Cat. No. 30230)
Note: The pH of buffers A - C is always verified immediately prior to use. Procedure 1. Streak out a glycerol culture of M15 E. coli cells transformed with the appropriate Fos/Jun expression plasmid on an LB agar plate containing 50/~g/ml kanamycin and 50/xg/ml ampicillin (+ Kan/Amp). 2. Pick a single bacterial colony from the freshly streaked plate to inoculate 2 ml of LB media (+ Kan/Amp) and grow an overnight starter culture at 37°. The starter culture is used as the inoculum for a second 200-ml overnight culture. 3. The 200-ml overnight culture is diluted 5-fold with LB media ( + Kan/Amp) and grown at 37° for approximately 20 min to a final optical density at 600 nm (OD600) of 0.50. 4. Isopropyl-fl-o-thiogalactoside is added to the culture at a final concentration of 1 mM to induce protein expression. The culture is then incubated for an additional 2 hr at 37° before harvesting. (Note: Because
[15]
c-Fos AND c-Jun REDOX-DEPENDENTDNA BINDING
169
the time course of induction may vary for different proteins, it is recommended that the optimal incubation time be determined on an individual basis.) 5. Cells are harvested by centrifugation at 5000 g for 10 min at 4 °. The cell pellet is resuspended by pipetting in 15 ml of buffer A and solubilized by gently mixing the sample on a rotator for 2 hr at room temperature (or overnight at 4 ° for convenience). 6. The cell lysate (15 ml) is cleared by centrifugation at 80,000 g for 30 min at 4°. The supernatant is removed and loaded onto a disposable 10-ml chromatography column (Bio-Rad, Richmond, CA, Cat. No. 7311550) packed with a 1-ml bed volume of nickel-chelate resin (QIAGEN, Cat. No. 30230). Before loading, the resin is washed with 10 ml of distilled water and equilibrated with 10 ml of buffer A (these steps are normally performed while the lysate is being centrifuged). 7. The flow-through from the column is collected by gravity and passed over the resin a second time (optional). After the second flowthrough fraction is collected the column is washed with 20 ml of buffer A, followed by 10 ml of buffer B. The two wash steps are necessary in order to elute proteins that are nonspecifically bound to the resin. 8. The recombinant protein is eluted from the column with 5 ml of buffer C. The sample is stable in this form and can be stored at - 8 0 ° indefinitely. 9. The purified protein is dialyzed overnight at 4° against several changes of buffer D to remove guanidine. A series of guanidine-containing buffers (1 liter each) are set up as a gradient to lower the concentration of guanidine in the sample in a stepwise manner. Buffers 1-3 contain 3.0, 1.0, and 0.1 M guanidine, respectively. The buffer used in the last four changes does not contain any guanidine. If protein precipitation is a problem, a shallower gradient using additional guanidine steps is recommended. 10. After dialysis the sample is transferred to 15-ml Corex tube, and the precipitate is cleared by centrifugation at 15,000 g in a swinging-bucket rotor for 15 min at 4 °. 11. The supernatant is removed carefully, and an aliquot of the purified protein is analyzed by SDS-PAGE. Protein concentration is estimated using the Bio-Rad protein assay (Bio-Rad, Cat. No. 500-0006). The protein is divided into aliquots and stored frozen at - 8 0 °. [Note: The protein precipitate may be resolubilized in a small volume of guanidine hydrochloride, pH 8.0 (0.1-1.0 M), or used directly to inject animals for antibody generation. However, a fraction of the precipitate should first be analyzed for purity by SDS-PAGE.]
170
ASSAY OF STRESS GENES/PROTEINS
[15]
Preparation of AP-10ligonucleotide Probe Fos-Jun DNA binding activity is assayed using a synthetic 25-bp double-stranded oligonucleotide (a-32p-labeled) containing the human metallothionein IIA promoter AP-1 site and synthetic XhoI and SalI restriction sites) 9'2° The oligonucleotide is radiolabeled by filling in 5' (XhoI) and 3' (SalI) overhanging ends with all four (AGCT) a-32p-labeled deoxynucleoside triphosphates (Amersham, Arlington Heights, IL). The sensitivity of the DNA binding assay is markedly enhanced because of the high specific activity of the AP-1 probe (1 × l09 cpm//~g). However, for the same reason the probe is more susceptible to radiolysis and is only stable at - 8 0 ° for approximately l0 days.
Materials and Reagents Double-stranded AP-1 oligonucleotide: 5'-TCGAGCGTGACTCAGCGCGCG-3' (top strand) 5'-TCGACGCGCGCTGAGTCACGC-3' (bottom strand) 10 x Reverse transcriptase buffer: 500 mM Tris-HC1, pH 8.0, 400 mM KCI, 100 mM MgC12, l0 mM DTT (added fresh) Reverse transcriptase: Avian myeloblastosis virus (AMV) reverse transcriptase (GIBCO-BRL, Gaithersburg, MD, Cat. No. 8020S). Radionucleotides: a-32P-Labeled deoxynucleoside triphosphates (dNTPs: dATP, dCTP, dGTP, TTP; >3000 Ci/mmol, Amcrsham) Sephadex G-25 spin column: STE SELECT-D G-25 (5 Prime-3 Prime, Inc., Cat. No. 5301-233431). Procedure 1. The AP-1 oligonucleotides are synthesized on an Applied Biosysterns (Foster City, CA) Model 380B DNA synthesizer and purified by 15-20% denaturing polyacrylamide gel electrophoresis.21 Equimolar quantities of the purified complementary DNA strands are annealed in STE buffer (10 mM Tris-HCl, 100 mM NaCl, pH 8.0, 1 mM EDTA) by heating the oligonucleotide mixture at 80° for 15 min and then slowly cooling to room temperature. The concentration of the single- and double-stranded oligonuclcotidcs is estimated by spectrophotometry (A260). Stock preparations of the AP-I oligonucleotide (usually 0.2 mg/ml) are stored frozen at - 20°. 19 p. Angel, M. Imagawa, R. Cbiu, B. Stein, R. J. Imbra, H. J. Rahmsdorf, C. Jonat, P. Herdich, and M. Karin, Cell (Cambridge, Mass.) 49, 729 (1987). 20 F. J. Rauscher III, L. C. Sambucetti, T. Curran, R. J. Distel, and B. M. Spiegelman, Cell (Cambridge, Mass.) 52, 471 (1988). 2i j. Sambrook, E. F. Fritsch, and T. Maniatis, "Molecular Cloning: A Laboratory Manual," 2nd ed., Appendix E.19. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1989.
[15]
c-Fos AND c-Jun REDOX-DEPENDENTDNA BINDING
171
The oligonucleotide labeling reaction is performed in a final volume of 30/zl containing the following: .
Double-stranded AP-1 oligonucleotide (0.2 mg/ml) Sterile water 10 × Reverse transcriptase buffer [a-32p]dNTPs (5 Izl of each labeled nucleotide A, G, C, T; >3000 Ci/mmol) AMV reverse transcriptase (1 unit//zl)
1.0/zl 4.0/zl 3.0/A 20.0/xl 2.0/zl
3. The reaction mixture is incubated at 37° for 30 min. 4. The labeling reaction is terminated by the addition of 20/xl of 50 mM EDTA, pH 8.0, and the sample is chilled on ice for 2 min. A 1-txl aliquot is set aside to determine the specific activity of the probe [incorporation of the isotope is measured by monitoring trichloroacetic acid (TCA)precipitable counts]. 21 5. Unincorporated a-32p-labeled nucleotides are removed from the reaction mixture by centrifugation through a 1.0-ml Sephadex G-25 spin column (5 Prime-3 Prime, Inc.) at 500 g for 5 min as per the manufacturer's recommendations. 6. The radioactivity in the sample is measured as described in Step 4, and the probe is stored at frozen at - 8 0 °.
DNA Binding/Gel Retardation Assay The redox state of Fos and Jun is determined indirectly by assaying DNA binding activity. Protein-DNA complexes formed in solution are analyzed on nondenaturing polyacrylamide gels. 10,11In this assay purified protein stocks of Fos and Jun are freshly diluted to working concentrations (1.0 tzM), and then an aliquot of each diluted protein is coincubated at 37° to allow the proteins to form dimers. The final concentration of Fos/ Jun dimers in the binding reaction is approximately 50 nM. Coincubation is performed in the presence of a carrier protein (e.g., bovine serum albumin) and a nonionic detergent (e.g., NP-40) which help stabilize nascent dimers. It is important to maintain a total protein concentration of 0. I-0.2 mg/ml in the reaction. Therefore, the addition of a carrier protein is not necessary if proteins from other sources are to be added to the sample (e.g., nuclear extract, see below). Oxidation of Fos and Jun occurs rapidly when the concentration of the reducing agent in the buffer falls below 0.2 mM. This is achieved by diluting the protein to a concentration of 1.0 ~M in a buffer that lacks any reducing agents. During the initial coincubation step in which the oxidized proteins are further diluted to 50 nM, they maintain their ability
172
ASSAY OF STRESS GENES/PROTEINS
[15]
to dimerize and will act as substrates for suitable reducing agents, either chemical (e.g., DTT, 2-mercaptoethanol) or biological (e.g., Ref-l). The addition of reducing agents to the reaction during the dimerization process will convert the proteins to a form that is active for AP-1 DNA binding. Following the dimerization/reduction step a suitable nonspecific copolymer DNA [e.g., poly(dI-dC):poly(dI-dC)] is added to the reaction. When present in excess over the radiolabeled oligonucleotide, poly (dI-dC):poly(dI-dC) acts to reduce the background of nonspecific protein-DNA complexes. 22'23 This is particularly important when assaying the AP-1 reducing activity of crude cell extracts in which nonspecific DNA binding proteins (e.g., histones) may impede binding of Fos/Jun dimers to the AP-1 probe. Labeled AP-1 oligonucleotide (25 fmol) is added to the reaction mixture in the final step of the DNA binding assay. After a short period of incubation with the probe, Fos/Jun-DNA complexes are resolved on nondenaturing polyacrylamide gels. As an alternative to using recombinant Fos and Jun proteins in this assay it is possible to use crude or partially purified AP-1 activity from mammalian cells to demonstrate a redox effect on DNA binding activity. HeLa cell nuclear extracts are an abundant source of AP-1 proteins, and routine procedures for purification of this activity are described in the literature. 24,25 Oxidation of AP-1 proteins isolated from nuclear extracts is achieved by removing reducing agents (commonly DTT) from the preparation through dialysis. The dialyzed AP-I preparation can then be directly substituted in the assay described here.
Materials and Reagents Radiolabeled AP-1 oligonucleotide (see above for preparation) Stock preparations of purified recombinant Fos and Jun (see above for preparation) Poly(dI-dC):poly(dI-dC): The lyophilized powder (Pharmacia LKB Biotechnologies, Piscataway, NJ, Cat. No. 27-7880) is resuspended in TNE buffer (I0 mM Tris-HC1, pH 7.5, 1 mM EDTA, 100 mM NaC1) to a final concentration of 1.0 mg/ml, denatured by heating at 100° for 5 min, and reannealed by slowly cooling to room temperature over a period of 3-4 hr; the preparation can be used directly and is stored frozen in aliquots at - 2 0 ° 22 H. Singh, R. Sen, D. Baltimore, and P. A. Sharp, Nature (London) 319, 154 (1986). 23 L. Hennighausen and H. Lubon, this series, Vol. 152, p. 721. 24 W, Lee, P. Mitchell, and R. Tjian, Cell (Cambridge, Mass.) 49, 741 (1987). 25 j. T. Kadonaga, "Protein Purification and Characterization," in press. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1994.
[15]
c-Fos AND c-Jun REDOX-DEPENDENTDNA BINDING
173
Binding buffer: 10 mM Tris-HCl, pH 7.5, 50 mM NaC1, 5 mM MgCI 2 , 1 mM EDTA, 5% glycerol, 5% sucrose, 0.01% NP-40, 0.1 mg/ml bovine serum albumin (BSA fraction V, ICN, Costa Mesa, CA, Cat. No. 810033) Protein dilution buffer: Same as binding buffer without BSA Gel loading buffer: Same as binding buffer with 0.05% bromophenol (w/v) blue and without BSA 10 × Tris-glycine buffer: 1.95 M glycine, 0.25 M Tris base (filter sterilized to remove particulate matter); it is not necessary to adjust the pH of the solution
Procedure 1. The DNA binding assay is performed in a 20-/~1 reaction as indicated below. The final reaction contains 50 nM of Fos-Jun dimer, 1.25 nM AP-1 probe (400,000 cpm), and 50/zg/ml poly(dI-dC):poly(dI-dC). Reduction of Fos and Jun is stimulated by the addition of chemical reducing agents to a final concentration of 10 mM. Alternatively, nuclear extract (1.0-2.0 /~g) or a purified recombinant reducing protein (e.g., Ref-1, 5.0-100 ng) is added to the reaction to stimulate DNA binding. However, if any protein preparations are to be assayed for redox activity and contain reducing agents, it is important to avoid carrying over an excess of the reducing agents into the binding reaction. The final concentration of reducing agent contributed by any protein preparation should not exceed 0.2 mM in the DNA binding reaction. Because our nuclear extracts or Ref-1 stock preparations are normally stored in buffers containing 1.0 mM DTT, the carryover of DTT into the binding reaction, following dilution, is roughly 50/zM as indicated below. Binding buffer (nonreducing) Purified Fos (freshly diluted to 1.0/~M with dilution buffer) Purified Jun (freshly diluted to 1.0/zM with dilution buffer) Reducing agent (nuclear extract, Ref-1, 10 mM DTT, 10 mM 2-mercaptoethanol)
15 ~1 1.0 ~1 1.0/~1 1.0/zl
(Note: Stock preparations of nuclear extract or Ref-1 are stored in a buffer containing 1 mM DTT.) 2. Incubate the reaction at 37° for 10 min to allow the proteins to dimerize and to catalyze the reduction of Fos and Jun. 3. Add 1.0/zl poly(dI-dC):poly(dI-dC) (1.0 mg/ml) and incubate for 5 min at room temperature. 4. Add 1.0/zl [o~-32p]AP-1 oligonucleotide (25 fmol; 400,000 cpm) and incubate for 15 min at room temperature.
174
ASSAY OF STRESS GENES/PROTEINS
[15]
5. Add 1.0/zl of gel loading buffer and quickly load the sample on a 4.5% nondenaturing polyacrylamide gel. The gel (1.5 mm thick) and running buffer are prepared in 25 mM Tris-195 mM glycine. Gels are run at 200-300 V (constant) for 2.5 hr at 4°. 6. The wet gel is blotted to Whatman (Clifton, N J) 3MM paper (Cat. No. 3030917) and dried under vacuum for 35 min at 80°. It is not necessary to fix the gel before drying. Protein-DNA complexes are visualized by autoradiography of the dried gel at - 80°. (Note: Native Tris-glycine gels have a tendency to turn yellow on drying.) Conclusion The procedures outlined in this chapter provide a rapid and sensitive method to study the regulation of AP-1 DNA binding activity in vitro. The assay has been used to demonstrate redox-dependent DNA binding activity of other classes of DNA binding proteins (e.g., Myb, NF-rB, ATF/CREB) and should therefore be applicable to other transcription factor systems with only minor modification.9 However, it is important to note that the DNA binding assay can only be used as an indirect measure of the redox state of a transcription factor. Although in the case of Fos and Jun we know that the reduced state of the reactive cysteine residue is a sulfhydryl based on sensitivity to certain cysteine modifying agents, the assay on its own does not provide information about the chemical nature of the oxidized and reduced states of the cysteine residue. 7 Alternative strategies are needed to make these determinations. Acknowledgments We thank C. Abate for contributions in helping to develop the AP-1 redox assay. S. Xanthoudakisis the recipientof a MedicalResearchCouncilof CanadaPostdoctoralFellowship.
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
175
[16] A s s e s s i n g G e n e E x p r e s s i o n d u r i n g O x i d a t i v e S t r e s s By DANA R. CRAWFORD, CHERYL A. EDBAUER-NECHAMEN,
CHARLES V. LOWRY, SHARON L. SALMON, YONG K. KIM, JOANNA M. S. DAVIES, and KELVIN J. A. DAVIES Introduction One of the exciting developments in the field of biological oxidation is the discovery of differentially regulated gene expression in response to oxidative stress. Both prokaryotic organisms and eukaryotic cells have now been shown to modulate gene expression as a function of oxygen tension, exposure to reactive oxygen species (such as superoxide and hydroperoxides), or exposure to various oxidation-linked cofactors or catalysts (such as quinones and transition metals). Such discoveries have led to the realization that biological defenses against oxidative insults are not static; rather, they respond to oxidative stimuli and can allow the organism to adapt to stress. Throughout this chapter the term gene expression is used to include all processes beginning with the initiation of gene transcription and ending with a functional protein product. For convenience the chapter has been subdivided into methods for assessing messenger RNA (mRNA) products through transcriptional regulation and methods for assessing altered translation from de novo protein synthesis. More advanced methods used to study the regulation of transcriptional processes [such as the role(s) of various transcription factors] are somewhat beyond the scope of this chapter. The concept of differentially regulated gene expression in response to oxygen or oxidative stress can be traced both to studies of altered enzyme activities under varied exposure conditions and to demonstrations of organismal adaptation or enhanced survival capacity. Increases in cellular antioxidant activities (e.g., superoxide dismutase, glutathione peroxidase) have been reported by many investigators to result from ox,idative stress. Similarly, adaptational improvements in the ability to survive an oxidative stress have been shown, by several laboratories, to result from preexposure to relatively low stress levels. 2-7 Some of the most striking demonstral K. J. A. Davies, ed., "Oxidative Damage and Repair: Chemical, Biological,and Medical Aspects." Pergamon, Oxford, 1991. 2 B. Demple and J. H. Halbrook, Nature (London) 304, 466 0983). METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
176
ASSAY OF STRESS GENES/PROTEINS
[16]
tions of both enzymatic and adaptational plasticity come from studies of aerobic versus anaerobic gene expression in facultative aerobes such as bacteria and yeast. 8'9 To cope with oxygen and various activated oxygen species prokaryotes and eukaryotes utilize numerous dietary antioxidant compounds and express a wide series of antioxidant enzymes and damage removal/repair enzymes. The antioxidant enzymes and compounds minimize oxidative damage but are not 100% effective. Damage removal/repair systems for oxidized proteins, lipids, and DNA act to prevent the accumulation of damage products and to repair membrane and DNA lesions. A current view of the interplay between oxidant, antioxidant, and damage removal/ repair systems is shown in Fig. 1. Many of the enzymes and some of the compounds shown in Fig. 1 exhibit increased levels of expression in response to various forms of oxidative stress. In writing this chapter we anticipate that many readers will wish to utilize various gene expression techniques in order to increase our understanding of antioxidants and damage removal/repair systems. The techniques described below are also suited to the discovery of potential new enzymes. From our current limited knowledge it is entirely possible that enzymes vital to surviving an oxidative stress await discovery. In approaching studies of the regulation of gene expression one must consider an inviting array of exciting research opportunities. Useful models range from acute to chronic regulation and from simple prokaryotes to studies in human beings. Systems that have already proved valuable include acute stress, acute and chronic adaptation, development and differentiation, aging, and clonal selection. Direct intervention methods include the use of expression vectors, deletion or mutation procedures, antisense RNA methods, and the engineering of transgenic cells or animals.
3 M. F. Christman, R. W. Morgan, F. S. Jacobson, and B. N. Ames, Cell (Cambridge, Mass.) 41, 753 0985). 4 K. J. A. Davies, A. G. Wiese, A. Sevanian, and E. H. Kim, in "Molecular Biology of Aging" (C. E. Finch and T. E. Johnson, eds.), p. 123. Alan R. Liss, New York, 1990. 5 S. W. Ryter, R. E. Pacifici, and K. J. A. Davies, in "Biological Oxidation Systems" (C. C. Reddy, G. A. Hamilton, and K. M. Madyastha, eds.), Vol. 2, p. 929. Academic Press, San Diego, 1990. 6 R. E. Pacifici and K. J. A. Davies, Gerontology 37, 166 (1991). 7 K. J. A. Davies, A. G. Wiese, R. E. Pacifici, and J. M. S. Davies, in "Free Radicals: From Basic Science to Medicine" (G. Poll, M. U. Dianzani, and E. Albano, eds.), p. 45. Birkhaeuser, Basel, 1993. s S. E. Chuang, D. L. Daniels, and F. R. Blattner, J. Bacteriol. 175, 2026 (1993). 9 R. S. Zitomer and C. V. Lowry, Microbiol. Rev. 56, 1 (1992).
[16]
G E N E EXPRESSION D U R I N G O X I D A T I V E STRESS
177
The methods presented in this chapter are designed to provide the basic tools necessary to begin useful studies in the regulation of gene expression. Readers should approach the chapter from the perspective that only the methods of interest need be selected. We have also attempted to provide original references, or references to more detailed review papers, whenever possible. For those who have limited experience in gene expression techniques, any of the more basic methods presented can be utilized. All of the techniques we discuss are applicable to even the most basic laboratory and require a minimum of specialized equipment.
Experimental Strategies Researchers who want a comprehensive survey of genes that may be differentially expressed may use subtractive hybridization, differential hybridization, or differential display techniques to isolate cDNA clones. In cases where a protein is well-characterized and the gene encoding it has been sequenced, a cDNA probe may be commercially available. Even genes coding for unknown proteins can be isolated by sequencing the purified protein. From the protein sequence, an oligonucleotide is designed and used to probe a cDNA library. Once an investigator has a cDNA probe, RNA may be analyzed by Northern blots or by transcriptional run-on assays. Certain methods described in this chapter are more appropriate for prokaryotes than eukaryotes (and vice versa), as summarized in Table I. For instance, a genetic approach (i.e., the isolation of mutants), is particularly well-suited for the simple genome of prokaryotes and lower eukaryores, such as yeast. The complex genome of higher eukaryotes, however, makes the isolation of mutants less practical. Techniques for isolating genes that are differentially transcribed, such as subtractive hybridization, differential hybridization, and differential display, are primarily reserved for eukaryotic systems since they take advantage of polyadenylated mRNA sequence. (Note: These techniques can still be used for some ATrich bacteria such as Bacillus subtilis.) Northern blot analysis is performed in eukaryotic systems to determine the size and levels of RNA transcripts. It is less commonly done in prokaryotic systems owing to rapid mRNA turnover and transcription-translation coupling. The rate of synthesis of primary RNA transcripts is determined in higher eukaryotic systems by transcriptional run-on assays. Lower eukaryotes and prokaryotes can be easily labeled with [3Ep]UTP and quantified directly, obviating the need for transcriptional run-on assays. Methods for analyzing proteins, such as two-dimensional gel electrophoresis,
178
ASSAY OF STRESS GENES/PROTEINS
~++o
-
~
ill
~
r-
""m-~ ~ • Om
i#ie-
.o_m
m
~ _ ~
I
_>~
~.,
/
"-"
~.~.1
w
\
c
/
:.=+
m
/ ~'~ ~ !
~_~.~|
+>>:=>:>
0)~
o
/
o++°m I i
6=,-
i~
~ -
. ~. . . .- .- .- - ..,i. . . .
•- -
[16]
m
I+p) J Im
"~,, ~
m-
/ ~
iI
,-
~1 ~ o . < . ~
!
\ i-
l<.. ©~
i l ) ~.e
m
\~,
o
~.lt.~
l
.-o
"8 o~ i
]
.~'~
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
179
TABLE I METHODS FOR ASSESSINGGENE EXPRESSION Method
Prokaryotes
Lower eukaryotes
Higher eukaryotes
Isolation of mutants Subtractive hybridization Differential hybridization Differential display Northern blot Transcriptional run-on RNA pulse-chase 2D gel electrophoresis Western blot Immunoprecipitation
+ + + + + +
+ + + + + + + + +
+ + + + + + + +
Western blotting, and immunoprecipitation, are useful in studying gene expression in both prokaryotes and eukaryotes. Protein Techniques T h e r e h a v e b e e n m a n y r e p o r t s o v e r the y e a r s d e s c r i b i n g m o d u l a t i o n of gene e x p r e s s i o n at the p r o t e i n level. M o s t studies h a v e u s e d e l e c t r o p h o resis as the t e c h n i q u e of c h o i c e to identify these c h a n g e s . O n e - d i m e n s i o n a l a n d e s p e c i a l l y t w o - d i m e n s i o n a l (2D) p o l y a c r y l a m i d e gel e l e c t r o p h o r e s i s ( P A G E ) t e c h n i q u e s h a v e b e e n v a l u a b l e m e t h o d s for i d e n t i f y i n g u n k n o w n m o d u l a t e d p r o t e i n s . W e s t e r n b l o t t i n g a n d i m m u n o p r e c i p i t a t i o n h a v e also b e e n u s e d to a s s e s s c h a n g e s in k n o w n p r o t e i n s to w h i c h a n t i b o d i e s are available. I n this s e c t i o n , we p r e s e n t p r o t o c o l s for these c o m m o n techn i q u e s that h a v e b e e n u s e d s u c e s s f u l l y in a s s e s s i n g o x i d a n t s t r e s s - i n d u c e d c h a n g e s in p r o t e i n levels.
Polyacrylamide Gel Electrophoresis There are two primary types of PAGE: sodium dodecyl sulfate (SDS) and native. The most popular form of SDS-PAGE is the discontinuous
FIG. 1. Pathways of oxidative damage and repair. Radicals that escape detoxification by antioxidant defenses contribute to the production of oxidatively damaged cell components. The steady-state level of damaged components is a function of the rate of production and the rate of removal and/or repair. Oxidatively damaged biomolecules may undergo direct enzymatic repair or degradation by catabolic enzymes that selectively recognize oxidatively damaged products, Reusable building blocks (amino acids, fatty acids, nucleic acids) are conserved for biosynthetic processes, and irreversibly damaged molcules are excreted.
180
ASSAY OF STRESS GENES/PROTEINS
[16]
Laemmli system. 10Denatured proteins (subunits if multimers) run on slab gels migrate strictly according to differences in molecular weight. In native gels migration is based on a combination of molecular weight, charge, and conformation. These gels consist of a lower separating gel and an upper stacking gel. The stacking gel concentrates large sample volumes, resulting in better band resolution. Resolving gels are poured first and typically range between 7 and 15% (w/v) polyacrylamide. High percentage gels allow better resolution of small proteins, and low percentage gels resolve high molecular weight proteins more clearly. A gradient gel, for instance, 6-12%, gives good resolution of large and small proteins. Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis. The resolving gel (e.g., 200 ml of gel with 12.5% (w/v) polyacrylamide) contains 0.375 M Tris (pH 8.8), 0.1% (w/v) SDS, and 12.5% (w/v) of a stock solution of acrylamide/bisacrylamide in a 30 : 0.8 ratio. After degassing the gel for 15 min, polymerization is catalyzed by the addition of 25/zl TEMED (N,N,N',N'-tetramethylethylenediamine) and 50 /zl of 10% (w/v) freshly prepared ammonium persulfate. Pour the solution immediately to a level 1-2 cm below the top of the apparatus. Overlay with water or butanol. We prefer butanol because of the smoother interface it leaves after polymerization, but it must be rinsed off with water within 1 hr or the polymerized gel will start to dehydrate. We usually pour the separating gel the day before an experiment and wrap it with Saran wrap to prevent water evaporation. Following polymerization, a 4.8% (w/v) stacking gel solution is then poured on top. (Note: The top of the separating gel is first dried down to the gel surface by blotting with Kimwipe tissues or paper towels). The stacking gel consists of 0.125 M Tris (pH 6.8), 0. I% (w/v) SDS, and 4.8% (w/v) acrylamide/bisacrylamide (30:0.8). A sample well comb is then inserted for one-dimensional gel electrophoresis. The gel is polymerized through the addition of 40 /zl TEMED and 400 /zl of 10% ammonium persulfate, Please note that higher concentrations of TEMED and ammonium persulfate are required for cross-linking of low percentage gels. When polymerization is complete, again rinse well with water. Protein samples are then added (see below). This dimension is run in buffer consisting of 25 mM Tris, 191 mM glycine, and 1% (w/v) SDS at a pH of 8.3. Native Polyacrylamide Gel Electrophoresis. For native gels, the same general procedure, reagents, and concentrations are used except that the SDS is omitted. Running Conditions. The SDS gels should be run at constant current. Because the proteins that run through the gels are denatured, they have low resistance and run quickly. A typical SDS gel is run at a constant l0 U. K. Laemmli, Nature (London) 277, 680 (1970).
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
181
current of 35-40 mA. The voltage will rise during this time from about 150 to 250-300 V. The gel is usually run until the bromphenol blue marker dye (see "Sample Preparation" below) reaches the bottom. For native gels, the resistance is much greater, and increasing voltage can become a heat problem. This is of concern since many native gels are run with the intent of maintaining protein conformation and hence activity. Therefore, it is preferable to run native gels at a constant 150 V. The amperage will drop during the run, however, so the overall time of electrophoresis is significantly greater than for SDS gels. We usually run our native gels overnight, also until the bromphenol blue dye reaches the bottom. Sample Preparation. For SDS gels, protein samples are mixed with an equal volume of 2 x SDS buffer, which consists of 250 mM Tris (pH 6.8), 2% (w/v) SDS, 100 mM dithiothreitol (DTT), 10% glycerol, and 0.025% (w/v) bromphenol blue. Boil samples for 3 min and apply to the gel wells. For native gels, samples can be applied directly with glycerol and bromphenol blue. In this case, no SDS, DTT, nor boiling is included. The 125 mM Tris buffer is usually enough to overcome the pH effects of other less concentrated buffers.
Two-Dimensional Protein Electrophoresis Two-dimensional PAGE li refers to the separation of proteins in the first dimension according to differences in charge followed by denaturation and further separation in the second dimension according to differences in molecular weight. Separation based on charge is referred to as isoelectric focusing (IEF). Both IEF and SDS-PAGE (second dimension) are standard procedures, although the number of commercial vendors and separation systems varies. The 2D PAGE technique has been used to identify oxidantmodulated sequences in both bacterial 3'12'13 and mammalian 14-16 cells (Fig. 2). Isoelectric Focusing. There are a number of types of apparatus commercially available for IEF. Each includes step-by-step protocols. The following is a general protocol for equipment of various shapes and sizes. I1 p. H. O'FarreU, J. Biol. Chem. 250, 4007 (1975). 12 S. B, Farr and T. Kogoma, Microbiol. Rev. 55, 561 (1991). t3 B. Demple, Y. Daikh, J. Greenberg, and A. Johnson, in "Anticarcinogenesis and Radiation Protection" (P. A. Cerutti, O. F. Nygaard, and M. G. Simi~, eds.), p. 151. Plenum, New York, 1987. t4 A. G. Wiese, Ph.D. Dissertation, Univ. of Southern California, Los Angeles (1992). i~ D. R. Crawford and J. S. Greenberger, Blood 76, 229a (1990). 16 K. Nose, M. Shibanuma, K. Kikuchi, H. Kageyama, S. Sakiyama, and T. Kuroki, Eur. J. Biochem. 201, 99 (1991).
A
A
77--
.=.m~
~=~
7~9 W
45-
rf /
0 w / 0 12-
I 5.5
I 7.4
I 8.4
ISOELECTRIC POINT
B
A
77-
4"
7-
~uJ
45 -
< --J
"7 0 W ...1 0
1
I
I
5.5
7.4
8.4
ISOELECTRIC POINT
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
183
Sample preparation. A stock solubilization solution consists of 9 M urea, 4% (v/v) Nonidet P-40 (NP-40), 2% (v/v) 2-mercaptoethanol, and 2% ampholytes of choice. Urea is first slowly dissolved in water with warming before addition of the other reagents. Dissolve the protein sample(s) in this solution. Usually 1 hr of incubation at room temperature is sufficient for dissolution. Ampholytes. The main operating principle on which IEF is based is the establishment of a pH gradient using charged carrier molecules known generically as ampholytes. These various polyamino-polycarboxylic acids are mainly supplied by Pharmacia-LKB (Piscataway, N J) and Bio-Rad (Richmond, CA). Each company has its own trademark names. Ampholytes from Pharmacia are known as Ampholines, Pharmalytes, and Immobilines; those from Bio-Rad are called Bio-lytes. Ampholines, Pharmolytes, and Bio-lytes use the conventional methodology of establishing a given pH gradient after an electric field has been applied. Immobilines are used for preformed pH gradients cast before application of an electric field. The pH range selected for IEF can be easily varied by mixing different range ampholytes together. The range is usually chosen based on what information is known about the protein(s) of interest and can be narrow (e.g., pH 7-9) to maximize resolution. For most situations, however, a broad pH range is used. This would be most appropriate for assessing proteins modulated by oxidant stress, where all possibilities could be analyzed. First-dimension isolectric focusing. Mark IEF tubes to the desired height, then wrap the bottom of the tubes with Parafilm. Set the tubes in a level casting stand and prepare the polymerization solution. A standard solution includes 9 M urea (final concentration), 1% (v/v) NP-40, 4%/0.2% (w/v) acrylamide/bisacrylamide, and 4% (v/v) ampholytes. The urea is dissolved by gradual warming in water, NP-40, and acrylamide. Once the urea is dissolved, the ampholytes are added followed by a 15-min degassing. Polymerization is initiated by the addition of stock TEMED (add 7 /xl per 10 ml solution) and freshly prepared 10% (w/v) ammonium persulfate with gentle swirling. The solution is then injected into the bottom of the tube and the syringe (wide internal diameter tube) or fine-gauge needle
FIG. 2. Two-dimensional gel analyses of 32D mouse promyelocytes treated with tertbutyl hydroperoxide (tBHP). The cells were treated with 0 (A) or 700 tzM (B) tBHP for 3 hr and harvested as described in the text. Samples were then run on a pH 4-9 isoelectric focusing gradient in the first dimension using Immobilines (Pharmacia), followed by separation on an 11% SDS-polyacrylamide gel in the second dimension. The gels were then silver stained. Induction of a peptide called transformin at 700 txM tBHP is indicated.
184
ASSAY OF STRESS GENES/PROTEINS
[16]
(small internal diameter tube) gradually withdrawn until the mark is reached. Top off with water. Polymerization should be completed within an hour. After polymerization, rinse the tops of the gels with water and remove the Parafilm and water. Insert the tubes in the IEF apparatus after covering the bottoms with wetted dialysis tubing held in place with rubber bands. Sample is added and tubes are topped off with the top chamber solution (freshly degassed 20 mM NaOH). The lower chamber solution is 10 mM H 3 P O 4 . Alternatively, a protective overlay solution can first be added to the IEF tubes and the samples loaded underneath this. The overlay solution usually consists of detergent such as 2% (v/v) NP-40 or Triton X-100 and 5% (v/v) ampholytes. After coolant is circulating, focus for 6-15 hr at a maximum of 1000 V. The amperage should start around 30 mA and drop near zero after a few hours.
Application of lsoelectric focusing Gel to Sodium Dodecyl Sulfate Gel. After the run, the first-dimension IEF gel is removed from the glass tube by cooling on ice for a few minutes. The dialysis tubing is removed, and the gel is gently pushed from the tube using a syringe filled with water and connected to the glass tube with Tygon tubing. It may be necessary to loosen the gel first by "rimming" it with a water-filled syringe and small-gauge needle. We prefer the use of a beveled short glass plate when pouring Laemmli gels for 2D electrophoresis to facilitate placement of the IEF gel. Soak the first-dimension IEF gel for at least 30 min in 0.125 M Tris (pH 6.8), 2% (w/v) SDS, and 0.1% (v/v) 2-mercaptoethanol. The gel is then applied to the second dimension in a makeshift Parafilm boat using a spatula to carefully move it onto the Laemmli gel (prepared as above). The best results are obtained when the IEF gel is lined up over the second dimension and the gel allowed to slowly slide off as the Parafilm boat is moved across the surface. In this case, the spatula acts more as a guide. We find that too much tugging and pushing of the IEF gel with the spatula eventually shows up as elongated, distorted sections following protein staining. The IEF gel can also be held in place by an overlay of 1% agarose containing 62.5 mM Tris buffer (pH 6.8), 2.3% (w/v) SDS, and enough bromphenol blue to give a strong color. A stacking gel usually is necessary for larger diameter IEF gels for improved resolution (3 mm). For IEF gels of 1.5 mm and below, only a separating gel need be poured. The second dimension is then electrophoresed at constant amperage (2 cm/A) until the bromphenol blue running dye is almost to the bottom. At this stage, remove the gel and stain with silver or Coomassie blue. In the case of [35S]methionine labeling, dry the gel and expose to X-ray film. Gradient Gels. Superior 2D gels are often obtained using gradient (e.g., 5-15% (w/v) polyacrylamide) rather than the above linear slab gels. Gradi-
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
185
ent SDS gels not only improve the resolution of a given spot but also the range of peptides that can be resolved on a given gel. Gradient gels are available commercially from several vendors. We have had success with prepoured gels from Integrated Separation Systems (ISS, Natick MA), and Isolabs (Akron, OH). To prepare them manually, prepare light and dense solutions. For a 5-15% gradient gel, prepare 100 ml each of 0.375 M Tris buffer (pH 8.8), 0.1% (w/v) SDS, 5% (w/v) (light) and 15% (dense) acrylamide (from a 30% acrylamide/0.8% bisacrylamide stock). TEMED (0.0005% v/v) and ammonium persulfate (0.0001% w/v) are added just before pouring. We usually pour gels from the bottom, in which case the gradient is poured from light to dense. Put light solution into a gel gradient mixing chamber, let some across the connector, then add dense solution to the other chamber. Start a stir bar in the mixing chamber, open the connector, and run the solutions into the gel(s) with a peristaltic pump. Fill to within 1-2 cm of the top for the stacking gel. Top off with water or butanol and allow the gel to polymerize for 1 to 2 hr. Add stacking gel and equilibrated IEF gels as above. Staining. Silver stain is much more sensitive than Coomassie blue and is usually the detection method of choice. However, when microsequencing is to be done on a spot of interest, silver staining should not be used (except for analytical purposes) owing to amino acid modifications. In this case, Coomassie blue staining in the absence of acetic acid is recommended. A general rule of thumb is that if you can detect a spot this way, there is sufficient quantity for microsequencing. For silver staining, we prefer to buy stains commercially and have used kits from ISS and BioRad successfully. The method of Blum et al., 17 while laborious, gives excellent detection with little background. There are a number of steps to the silver staining protocol, however, and background can be a major problem. We recognize that companies who offer kits have carefully minimized problems in their own reagents. In addition, it is very important to use top quality distilled, deionized water (e.g., 18 Mf~ Millipore, Bedford, MA, MilliQ purified), clean glassware, and gloves. For Coomassie blue staining, dissolve Coomassie Brilliant Blue R250 in 50% (v/v) methanol/ 12% (v/v) glacial acetic acid. For best solubilization, dissolve the Coomassie blue in the methanol/acetic acid only, then add water. Stain gels for 30-60 min, then destain first in 50% (v/v) methanol/12% (v/v) acetic acid, then 20% (v/v) methanol/12% (v/v) acetic acid. For microsequencing, eliminate the acetic acid from the protocol. Radioactive Pulse Labeling. Pulse labeling is an effective technique for assessing the effects of oxidant stress for two reasons. First, it increases i7 H. Blum, H. Beier, and H. J. Gross,
Electrophoresis 8, 93 (1987).
186
ASSAY OF STRESS GENES/PROTEINS
[16]
the sensitivity for identifying modulated proteins. Second, it allows for a detailed analysis of modulated proteins at specific time points following oxidant exposure. The latter analysis is greatly facilitated by the pulse technique, which keeps background signal low. The following technique has been used successfully to identify proteins modulated following exposure to H202.14 At the end of the stress period, wash the cells free of oxidant. At various times from this point, pulselabel for 1 hr (shorter time with prokaryotes) with 100/zCi [35S]methionine/ cysteine per dish in methionine-cysteine-free medium. Remove the medium and wash the cells two times with phosphate-buffered saline (PBS), then chase with unlabeled medium to the desired end point (for these studies, 18 hr). As an example, a 4-hr time point pulse-chase experiment with mammalian cells would consist of removing the oxidant after cell exposure, replacing with complete medium, removing the medium at 4 hr and replacing with [35S]methionine/cysteine in nonlabeled methionine-cysteine-free medium, pulsing for 1 hr, removing the radioactive medium, and chasing with unlabeled medium. At the end of the chase, wash the cells twice with PBS, harvest, and lyse cells in the lysis buffer of choice (e.g., 0.3% (w/v) SDS, 5% (v/v) 2-mercaptoethanol, and 60 mM Tris). Incubate the lysate with 1/zg/ml DNase and RNase, boil for 5 min, then bring to 8 M urea. Equal amounts of radioactivity (determined by scintillation counting) are then applied to the first-dimension IEF'gel. Amino-N-Terminal Microsequencing. An advantage to 2D electrophoresis is the separation of peptides into spots from which they can be sequenced. Because we are dealing with only a small amount of protein in a given spot, limited by the amount of protein that can be loaded on a gel, conventional protein sequencing is not possible. However, it is possible to do microsequencing if sufficient peptide is present. This technique, if successful, will give the sequence of the first 20-30 amino acids of the peptide. This information can then be used to scan a protein sequence database to determine whether the peptide is known or not. Such protein sequences can also be used to generate homologous DNA oligomers which can in turn be used to screen a cDNA library to pull out a corresponding clone. It is beyond the scope of this chapter to describe protein microsequencing in detail. Practically speaking, however, spots of interest are usually passed on to protein core laboratories or companies who specialize in the technique. There are a number of important points to keep in mind. First, the best way to present a peptide spot for microsequencing is to have it bound to a polyvinylidene fluoride (PVDF) filter. To do this, the peptides are electroblotted to the membrane following the second dimension run. The membranes can be stained with Coomassie blue (minus the acetic acid)
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
187
tub ~tr '¢
<
CI 30
C I 41
,¢z O~
o
~-
,-- ,--
o
~
2
~:
0
N
~
0
¢~
~
~o
m
F1G. 3. Effect of phorbol myristate acetate (PMA) on catalase protein levels in mouse epidermal JB6 cells. JB6 subclones 30 (C130) and 41 (CI 41) were treated in monolayer cultures with 10 ng/ml PMA for the indicated lengths of time. Cell sonicates were electrophoresed on a 12.5% SDS-polyacrylamide gel and electrotransferred to membranes, as described in the text. The membranes were treated with rabbit anti-human catalase antibody and then incubated with tzsI-labeled goat anti-rabbit antibody. Protein extracted from non-PMAtreated cells (Cont) was included as a control. [Reproduced (in part) from Crawford et al., 19 with permission.]
to locate the spot, which is then cut out for sequencing. The electroblot is a convenient way to present a concentrated form of the peptide, free from interferring glycine and SDS. One pitfall with the entire protocol is something called N-terminal blockage, where the N-terminal amino acid is modified such that sequencing cannot even start. The combination of in vivo blockage and/or subsequent handling effects, especially N-terminal carbamylation by urea, makes this a common problem. Usually the sequencer can detect this within a few cycles and stop the sequencing to keep costs down. If N-terminal blockage is a problem, the peptide can be cleaved internally by trypsin, proteinase K, or cyanogen bromide to generate unmodified ends.
Western Blotting Western blotting has been a popular technique for assessing changes in protein levels for many years. It has also been used with both bacterial and mammalian cells to study oxidant stress response and the modulation of antioxidant enzymes 18'~9(Fig. 3). Most commonly, one-dimensional slab gels are used for Western blotting, although 2D gels can also be utilized.
Sample Preparation and Sodium Dodecyl Sulfate and Native Polyacrylamide Gel Electrophoresis. Samples for the gels and gel preparation and running conditions are as described above. 18 G. Storz, L. A. Tartaglia, and B. N. Ames, Science 248, 189 (1990). 19 D. R. Crawford, P. A. Amstad, D. D. Y. Yin Foo, and P. A. Cerutti, Mol. Carcinog. 2, 136 (1989).
188
ASSAY OF STRESS GENES/PROTEINS
[16]
Electroblotting. Proteins separated on SDS and native gels are commonly electroblotted onto Millipore Immobilon PVDF membranes. Soak the gel and membrane for l0 min in transfer buffer [25 mM Tris and 192 mM glycine (pH 8.3)]. Methanol is included to 20% (v/v) for SDS gels. Prepare a gel-filter sandwich under transfer buffer to avoid bubble formation with two additional Whatman (Clifton, N J) papers on both sides. Correctly orient the sandwich in a Trans-Blot (Bio-Rad, Richmond, CA) apparatus (filter to positive side, gel to negative side). Conduct the transfer for 2 hr at high amperage - 9 5 0 mA, 60 V. Antibody Incubation. Incubate the filters for 2 hr (in a slowly shaking bath) at room temperature with saturation buffer consisting of 10 mM Tris (pH 7.4), 100 mM MgC12,0.1% (v/v) Triton X-100, 0.5% (w/v) bovine serum albumin (BSA), and 1 x Denhardts' solution (see "Northern Blot Electrophoresis" section below) filtered through a 0.45-/xm filter. Remove and add fresh saturation buffer containing antibody (5/xl per I0 ml buffer as an initial approximation). Incubate overnight at 4 °. Pour out the antibody solution and add 10 mM Tris (pH 7.4) plus 150 mM NaCl (TBS) to the bag. Mix, pour out, remove the filter, and wash the filter four times for 5 min each at room temperature with TBS. Then rinse for 10 min with saturation buffer and put the filter in a plastic bag with fresh saturation buffer containing 100,000 cpm 125I-labeled second antibody per milliliter. The second antibody binds to the filter-bound primary antibody from the previous step. For example, if the first antibody is of rabbit origin, the second labeled one might be goat anti-rabbit immunoglobulin G (IgG). Incubate for 2 hr at room temperature with gentle mixing. Remove the labeled solution, rinse the filter with TBS, then remove the filter from the bag and wash 4 times with TBS. Wrap the damp filter with Saran wrap and expose to X-ray film with an intensifying screen at - 7 0 °. Alternatively, the antibody reaction can be quantitated nonisotopically by adding the second antibody conjugated to alkaline phosphatase or horseradish peroxidase. After incubation, the second antibody is reacted with a suitable colorimetric or chemiluminescent substrate. Kits are available from numerous commercial distributors. Immunoprecipitation Where antiserum is available, immunoprecipitation is a useful technique for analyzing proteins in their native form. Protein-protein interactions may be detected by coprecipitation of immunologically unrelated proteins. In addition, the function of a protein can be confirmed by removing the protein from a cell extract by immunoprecipitation and verifying
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
189
loss of activity in the system. In other words, one can analyze a protein in a condition approaching its physiological state in the cell. There are two factors to keep in mind when using immunoprecipitation. The antibody must recognize native protein. Antibody raised against denatured protein or synthetic peptides may not recognize conformational epitopes in the native protein. Such an antibody will, however, probably recognize denatured protein on a Western blot. Immunoprecipitation is most useful when a protein can be radioactively labeled in situ. The labeled protein can then be studied directly after electrophoresis on an SDS-polyacrylamide gel and exposure to X-ray film. Proteins that cannot be labeled in situ (e.g,, proteins in erythrocytes) can be detected indirectly by Western blotting of immunoprecipitates. Labeling of CellExtracts. Proteins are radioactively labeled by incubating ceils (e.g., 60-mm dish) with 30-200/zCi of [35S]methionine/cysteine in 1 ml of methionine- and cysteine-free medium for 2 hr. 2° Cells are washed 3 times with PBS and lysed in 0.5 ml of lysis buffer [20 mM Tris (pH 7.5), 150 mM NaCI, 1% sodium deoxycholate (w/v), 0.5% NP-40 (v/v), 1 mM phenylmethylsulfonyl fluoride]. Cell debris is removed by centrifuging at 14,000 rpm for 15 min in a microcentrifuge at 4°. Immunoprecipitation of Protein. Antiserum (e.g., 15/zl) is added to 0.25 ml of the clarified cell extract and incubated at room temperature for 30 rain.21 Antiserum should be titrated to determine the optimal concentration for immunoprecipitation. Protein A linked to Sepharose or agarose (e.g., 15/zl) is incubated with antigen-antibody complexes for 15 min at room temperature and precipitated by centrifuging in a microcentrifuge at 14,000 rpm for 1 rain at room temperature. Immunoprecipitates are washed once in sucrose buffer [20 mM Tris (pH 7.5), 5 mM EDTA, 20% (w/v) sucrose, 0.5 M NaCI, 0.05% (v/v) NP-40] and 3 times in low salt buffer [10 mM Tris (pH 8.0), 5 mM EDTA, 150 mM NaC1]. Antigen is released from antigen-antibody complexes by suspending precipitates in 1× sample buffer [1.0% (w/v) SDS, 1.0% (v/v) 2-mercaptoethanol, I0 mM Tris (pH 8.5), 1 mM EDTA, 20% (v/v) glycerol, 20 mM bromophenol blue], boiling for 2-5 min, and centrifuging for 5 min at room temperature in a microcentrifuge. The supernatant is analyzed by SDS-polyacrylamide gel electrophoresis as described above. Radiofluorography. Gels are fixed in 10% (v/v) acetic acid, 45% (v/v) methanol for 1.5-18 hr. Radiofluorography is performed by treating gels 2o y. Yin, M. Tainsky, F. Bischoff, L. Strong, and G. Wahl, Cell (Cambridge, Mass.) 70, 937 (1992). 2i S. Ross, A. Levine, R. Galos, J. Williams, and T. Shenk, Virology 103, 475 (1980).
A
myc --~
I
I
I
I
~:~
o
0
:
:
:
8~
:
:
8 ~× C130
.
-
0
C141
28S ~)"
Fos --~ 18S --~
ACTIN --)-
JB30
(NON-PROMOTABLE)
I
B
JB41
(PROMOTABLE)
II
Con 51zg10~g
2hr
4hr
8hr
I
Con
2hr
4hr
8hr
CAT GPX JB30
JB41 II
I
Con 5vg 10pg
SOD
2hr
4hr
8hr
Con
2hr
4hr
8hr
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
191
with EN3HANCE (NEN-Du Pont, Boston, MA), twice for 30 min each. The gel is incubated in water for 15 min and in 3% glycerol for 1 hr before drying. Gels are exposed to Kodak (Rochester, NY) X-OMAT X-ray film at - 7 0 ° that has been preflashed to get a linear range of exposure. RNA Techniques RNA techniques have become increasingly popular in analyzing gene expression changes as more and more clones have been made available for use as probes; however, the basic techniques have remained almost unchanged. Both transcriptional run-on and Northern blot hybridization techniques have been used to assess oxidant stress-induced modulation of RNA, including protooncogene and antioxidant enzyme m R N A s 16'22-24 (Figs. 4 and 5) Transcriptional run-on is a technique for studying the expression of specific genes in eukaryotic cells. It is used to quantify the amount of primary RNA transcripts transcribed from isolated cell nuclei. Northern blot hybridization is used to assess the modulation of mRNA levels in the cell, regardless of whether the changes are due to altered transcription, stability, or transport. Basic Nucleic Acid Techniques
Although a detailed description for handling nucleic acids is beyond the scope of this chapter, there are some basic procedures that should be mentioned. First of all, there, are several toxic componds that should be used in a chemical fume hood. These include formaldehyde, diethyl pyrocarbonate (DEPC), ethidium bromide, phenol, and dimethyldichlorosilane. Water should always be deionized distilled and, preferably, of highest purity (18 Mfl Millipore and Barnstead units). For RNA work, 22 D. Crawford, I. Zbinden, P. Amstad, and P. Cerutti, Oncogene 3, 27 (1988). 23 M. Shibanuma, T. Kuroki, and K. Nose, Oncogene 5, 1025 (1990). 24 S. ShuU, N. H. Heintz, M. Periasamy, M. Manohar, Y. M, W. Janssen, J. P. Marsh, and B. T. Mossman, J. Biol. Chem. 266, 24398 (1991).
FIG. 4. Northern blot analyses of modulated mRNAs in response to active oxygen (A) and PMA (B). Mouse epidermal JB6 cells (subclones 30 or 41) were treated with 20/zg/ml xanthine and 2/~g/ml xanthine oxidase (A) or 10 ng/ml PMA (B) for the indicated lengths of time and total RNA extracted. The RNAs were run on agarose gels, transferred, and hybridized as described in the text. CO, Nontreated control cells; X/XO, xanthine plus xanthine oxidase; CAT, catalase; GPX, glutathione peroxidase; SOD, copper, zinc-superoxide dismutase. [Reproduced (in part) from Crawford e t al., 19'22 with permission.]
192
ASSAY OF STRESS GENES/PROTEINS
[16]
£= 0 X
0 X
0 X
X O
fos myc GAPDH ACTIN SP65
fos
8.40
0.01
-
0.62
-
myc
0.04
0.01
-
0.06
0.04
GAPDH
1
1
1
1
1
ACTIN
7.20
0.09
0.05
1.82
0.69
SP65
.
.
.
.
.
D e n s i t o m e t e r readinqs normalized relative to G A P D H
FIG. 5. Transcriptional run-on analysis of c-fos, c-myc, and fl-actin genes in response to active oxygen. The experiment was conducted as described for Fig. 4A, and cell nuclei were extracted and D N A transcribed as detailed in the text. COX, Cells exposed to xanthine only; GAPDH, glyceraldehyde phosphate dehydrogenase control used for normalization; SP65, control, nonmammalian DNA. [Reproduced (in part) from Crawford e t a1.,22 with permision.]
the water is treated with DEPC (Sigma, St. Louis, MO) to inactivate RNase. Add 1/zl DEPC per milliliter of water, mix and shake well, and leave overnight at room temperature. The next day, autoclave to break down the toxic DEPC. Glassware for RNA work can also be treated with DEPC. The D N A mutagen ethidium bromide is prepared as a 10 mg/ml stock and stored at 4 °. Stain agarose gels for RNA or D N A by soaking in a 10,000-fold dilution of the stock, then destaining in water. Phenol is prepared by heating high-quality frozen stock to 65 °, dissolving 8-hydroxyquinoline to 0.1% once the solution is thawed, then immediately adding an equal volume of DEPC-treated water. (Note: Phenol freezes at room temperature.) Mix well, then let separate. Repeat this at least several times more. Siliconization of tubes and pipettes that are used to manipulate nucleic acid is an advantage because it prevents loss of valuable sample to the wall of these items, especially at low concentration. We siliconize Eppendoff tubes, Corex tubes, Pasteur pipettes, etc., by soaking in a 5% (v/v) solution of dimethyldichlorosilane (Sigma) in chloroform for 30 min followed by washing with DEPC-treated water. For RNA work, any glass-
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
193
ware should be baked for 8 hr at 200 ° to inactivate RNase. (Note: Autoclaving does not destroy all of this very stable enzyme.) This includes siliconized glassware. For all RNA work, gloves should be worn. Basic nucleic acid precipitation consists of adding 0. I volume of 3 M sodium acetate (pH 5.3) or NaCI to 0. I M, mixing, then adding 2 volumes of absolute ethanol. Precipitate overnight at - 2 0 °, for 1 hr at - 7 0 °, or for 15-30 min in dry ice/alcohol slush. Spin the samples (e.g., 15 min at high speed in a microcentrifuge at 4°), remove the supernatant, wash the pellet with ice-cold 75% ethanol, and respin. The supernatant is again removed and the pellet semidried in the open on a bench top (-15 min; no longer for RNA) or in a Speed Vac (Farmingdale, N.Y.) concentrator. The pellet is then resuspended, preferably in DEPC-treated water or low ionic strength buffer. Readers are referred to Sambrook et al., 25 and Wallace 26 for excellent detailed procedures on working with RNA and DNA.
RNA Extraction Extraction of quality DNA is one of the more difficult molecular biology techniques despite the availability of relatively simple protocols. This is due to the presence of RNase, an extremely active and heat-stable enzyme that degrades RNA. RNase is present in cells, on fingers, and on contaminated glassware. Of immediate concern is the release of the enzyme during cell extraction. Most techniques therefore employ either guanidinium isothiocyanate or SDS/proteinase K to inactivate RNase during extraction. We use the strong protein denaturant guanidinium thiocyanate for preparation of total RNA. [Note: Usually less than 5% of total RNA is poly(A) + mRNA, the bulk consisting of ribosomal RNAs.] The necessary reagents for this procedure are also available commercially from Biotecx Labs., Inc. (Houston, Texas). For poly(A) + mRNA preparation (i.e., most of ribosomal RNA removed), we use a relatively new technique that combines total RNA preparation with poly(A) + selection. Protocols for both applications are given below. Total R N A Preparation (Guanidinium Thiocyanate). The following procedure 27gives typical yields of 25-50/xg total RNA per 2 × 10 6 Chinese hamster ovary (CHO)-derived fibroblasts. The guanidinium solution is prepared by dissolving 250 g commercial guanidinium thiocyanate powder directly in the vendor's bottle with 293 ml water, 17.6 ml of 0.75 M sodium 25 j. Sambrook, E. F. Fritsch, and T. Maniatis, eds., "Molecular Cloning: A Laboratory Manual." Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1989. 26 D. M, Wallace, in "Guide to Molecular Cloning Techniques" (S. L. Berger and A. R. Kimmel, eds.), p. 33. Academic Press, San Diego, 1987. 27 p. Chomczynski and N. Sacchi, Anal. Biochem. 162, 156 (1987).
194
ASSAY OF STRESS GENES/PROTEINS
[16]
citrate (pH 7.0), and 26.4 ml of 10% (w/v) sarkosyl at 65° followed by 0.36 ml of 2-mercaptoethanol per 50 ml (this denaturing solution is stable for 3 months at room temperature without mercaptoethanol and 1 month with). Additional solutions that need be prepared include sodium acetate (pH 4.0), water-saturated phenol (warning: toxic), and chloroformisoamyl alcohol (49 : 1 (v/v)). Cells in suspension should first be spun out and washed with PBS before guanidinium addition. Monolayer cultures should also be washed with PBS. Add 2 ml of the denaturing solution per 100-mm monolayer plates (1.5 ml per 60-mm plate), let sit for 2 min, triturate, and remove the lysate to a sterile Falcon tube on ice. Add 0.1 volume of 2 M sodium acetate (pH 4.0), 1 volume of water-saturated phenol, and 0.2 volumes of chloroform-isoamyl alcohol. Shake the tubes vigorously for 15 sec, then put back on ice for 15 min longer. Transfer the samples to baked, siliconized Corex tubes and spin at 10,000 g for 20 min at 4 °. (Note: Be sure to use the inside, paper-facing side of the Parafilm facing into the Corex tube.) At this pH, DNA goes into the phenol phase. Remove the top aqueous phase to a fresh Corex tube and precipitate the RNA with an equal volume of 2-propanol. Mix well and store at -20 ° for at least 1 hr to precipitate the RNA. Pellet the RNA by spinning again at I0,000 g for 20 min, dissolve the pellet in 0.3 ml of the denaturing solution, transfer to a sterile siliconized Eppendorf tube (Krackler Scientific, Albany, N.Y.), and again precipitate with an equal volume of 2propanol. Spin for I0 rain in a microcentrifuge, then wash the pellet with 75% ethanol and respin. Remove the supernatant again and let the pellet semidry open on the laboratory bench or in a Speed-Vac. (Note: If the pellet is dried to completion, it is very difficult to redissolve.) Dissolve the RNA on ice with a small volume of DEPC-treated water or with brief heating at 65° if necessary to approximately 2/xg//xl. Dilute a small aliquot in water and immediately quantify by optical density at 260 and 280 nm. An OD260 value of 1.0 corresponds to 40 /~g/ml RNA assuming the 0026o/280ratio is close to 2.0. A significantly lower value means the sample is probably contaminated with protein or phenol. Poly(A) ÷ m R N A Preparation. The following procedure, described by Badley et al., z8 gives relatively rapid preparations of quality poly(A) + mRNA selected on oligo(dT)-cellulose and does not require phenol. Spin suspension cells or scraped (plastic cell scraper, Falcon) monolayer cells, resuspend the cell pellet in PBS, and respin and rewash with PBS. Remove the supernatant and vortex or tap the pellet so it coats the sides of the Falcon tube. Add l0 ml lysis buffer [0.2 M NaCl, 0.2 M Tris (pH 7.5), 28 j, E. Badley, G. A. Bishop, T. St. John, and J. A. Frelinger, BioTechniques 6, 114 (1988).
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
195
1.5 mM MgCI2, 2% (w/v) SDS, and 0.2 mg/ml proteinase K] per 108 cells. Draw up and expel the lysate several times with a syringe equipped with an 18-gauge needle, then repeat with a 22-gauge needle. This will disperse the lysate and shear the viscous genomic DNA. Incubate for 60-90 min on a rocker platform at 45 °. The proteinase K will digest the RNases liberated by the SDS. During the incubation, prepare the oligo(dT)cellulose resin (Type 3, Collaborative Research, Bedford, MA). First, hydrate it with a 20-fold excess volume of 10 mM Tris (pH 7.5), subject it to a quick spin at low speed, then suspend the pelleted resin in a 20fold excess of binding buffer [10 mM Tris (pH 7.5) with 0.5 M NaC1]. [Note: RNA will bind to oligo(dT)-cellulose in the presence of the high salt and will elute at low ionic strength.] Pellet and resuspend a second time in binding buffer, then respin again and resuspend in about 1 ml of the same. Keep samples at room temperature until further use. Following the proteinase K/SDS incubation, the lysate (already 0.2 M NaCI) is brought to 0.5 M with addition of more NaC1. Mix the sample thoroughly, then add to the preequilibrated oligo(dT)-ceilulose. Incubate for 20 min at room temperature with constant mixing. Pellet at room temperature, remove the supernatant, and resuspend the resin pellet in fresh binding buffer. Repeat this until the cloudiness is gone from the supernatant. Prerinse a small disposable column [Quik-Snap columns (Isolab)] and fill part way with binding buffer. Add the resuspended oligo(dT)cellulose/RNA to the column and wash with binding buffer until the OD260 value is less than 0.05. Then allow the column to run dry. Elute the bound mRNA with DEPC-treated water (5 successive volumes of 0.5 ml) into 15-ml siliconized, baked Corex tubes. Precipitate the RNA with 0.1 volume of 3 M sodium acetate (pH 5.3) and 2 volumes of cold absolute ethanol (high quality such as Aldrich, Milwaukee, WI, HPLC grade). If working with relatively small amounts of RNA, add 1 /zl (20/xg) of commercial molecular biology-grade glycogen (Boehringer Mannheim, Indianapolis, IN) before the alcohol to aid precipitation and subsequent detection of the RNA pellet. Spin out the RNA at maximum allowable speed for the Corex tubes (30 min at 4°), remove the supernatant, wash the pellet in 75% ethanol, and respin. Semidry, resuspend, and quantitate the pellet as for the guanidinium procedure. Notes. For both procedures, it is a good idea to save supernatants in case the RNA does not pellet at any one step. Note that the poly(A) ÷ preparation will still have appreciable amounts of ribosomal RNA present (approximately half the RNA). To improve purity, we often repeat the poly(A)* selection step by heating the combined fractions to 65° for 3 min, cooling rapidly on ice, and adjusting to 0.5 M NaCI for a second
196
ASSAY OF STRESS GENES/PROTEINS
[16]
round of oligo(dT)-cellulose chromatography on the same column. The drawback of the additional handling and heating is that more poly(A) + degradation can occur.
Northern Blot Electrophoresis Northern blotting and hybridization constitute a methodology for quantifying and sizing mRNA. z9Often times, as in the case of oxidant treatment, there are multiple samples that can be easily compared for modulation. The basic technique involves running RNA out on a denaturing agarose gel in an electric field. The RNA is then transferred from the gel to a more permanent matrix such as a nitrocellulose or nylon membrane. After hybridizing with radiolabeled DNA, one can then locate the RNA of interest by autoradiography. Denaturing Gel. RNA will not effectively bind to filters unless completely denatured. This can be accomplished by a variety of reagents including formaldehyde, our method of choice. Prepare a 1.0-1.5% (w/v) agarose gel by first weighing out high-quality, RNase-free agarose (e.g., Sigma "Molecular Biology" grade) in water and microwaving into solution. Let cool for 5 min, then add 20× MOPS buffer [400 mM 3-(Nmorpholino)propanesulfonic acid, 100 mM sodium acetate, and I0 mM EDTA (pH 7.0)] to 1× final concentration and formaldehyde to a final concentration of 6% (v/v). Pour the gel into an electrophoresis apparatus with a loading comb and allow to set. The RNA sample is then denatured as follows: per 25-/xl final volume (excluding loading dye), add 12.5 ~1 of highest quality, deionized formamide; 4/xl of formaldehyde; 1.25/.d of 20x MOPS buffer; and 6/~1 of RNA sample plus water. Heat for 15 min at 60 °, place on ice, and add 2.5/~1 of sterile-filtered 10 × loading buffer [0.25% (w/v) bromphenol blue, 0.25% xylene cyanol, 50% (v/v) glycerol, and 1 mM EDTA (pH 8.0)]. It is also helpful to run a separate lane with RNA size standards available from G I B C O - B R L (Gaithersburg, MD). Fill the electrophoresis chamber with l x MOPS buffer. For best results, electrophorese overnight at 1-1.5 V/cm until the bromphenol blue dye nears the bottom. Alternatively, the gel can be run during the day at about 4-5 V/cm. When the run is complete, rinse the gel several times in DEPC-treated water to remove the formaldehyde. Cut the gel to the desired size for blotting (usually 1-2 cm below the dye) and mark by cutting the upper left corner. The RNA is then transferred from the gel to a nitrocellulose or nylon filter as described below. z9 p. S. Thomas, Proc. Natl. Acad. Sci., U.S.A. 77, 5201 (1980).
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
197
Transfer to Filter. Transfer of denatured RNA to a nitrocellulose or nylon filter can be accomplished by several methods. We prefer transfer by capillary elution. This method is well described in Sambrook et al. 25 and by Wahl et al. 3° The blot is set up in the following manner. A large shallow tray is lined with sponges or a blotting stone (Altec Labs, Boston, MA) saturated with 10-20× SSC [20× SSC is 3 M NaCI and 0.3 M sodium citrate (pH 7.0)]. Wetted Whatman 3MM paper (larger than the gel) is placed on the sponges. Care is taken to avoid air bubbles in all steps. The gel is placed on the Whatman 3MM paper, and a nylon or nitrocellulose membrane, cut to the size of the gel, is placed on the gel. Thin strips of Parafilm should then be placed around the gel to prevent the paper towel from short-circuiting with the buffer. Whatman 3MM paper and a stack of paper towels, cut to the size of the gel, are then placed on top. The stack is kept in place with a small weight on a plastic tray. RNA transfer is carried out overnight in 10-20× SSC, added to just below the top of the sponges or blotting stone. Liquid flows from the dish through the sponges or blotting stone, paper, gel, and filter, depositing the nucleic acids on the filter. The next morning, the wells of the gel are marked on the filter with a ballpoint pen or by piercing with a needle. The filter is briefly rinsed in 6× SSC to remove paper and gel bits, air dried, and baked at 80° for 2 hr in a vacuum oven. (Note: Only a vacuum oven is used for nitrocellulose, which is explosive.) Alternatively, the filter is blotted to dampness, then immediately UV-irradiated (Stratagene, La Jolla, CA, Stratalinker) for 2 min. The filter is now ready for hybridization. Hybridization. There are two primary methods for setting up hybridization. The conventional method is to place the filter in a sealable plastic bag (e.g., Seal-a-meal bags, available from supermarkets) and seal securely all the way around. The amount of prehybridization solution is dependent on the size of the filter. One needs enough to cover the filter, but not so much that the radioactive probe is diluted significantly. A good guide is 0.2 ml of prehybridization solution per square centimeter of filter. The second method involves the use of a hybridization oven (Hybaid oven, Laboratory Product Sales, Rochester, N.Y.), which we strongly prefer owing to the ease of use and, especially, the reduced exposure to radioactivity. For this procedure, an appropriately sized piece of support mesh is chosen and prewet along with the filter in 2× SSC. Orient the filter to fit exactly over the mesh, then tightly roll up. Insert the roll into a hybridization bottle and add 30 ml of 2x SSC. Allow the filter and mesh to 3o G. M. Wahl, J. L. Meinkoth, and A. R. Kimmel, in "Guide to Molecular Cloning Techniques" (S. L. Berger and A. R. Kimmel, eds.), p. 572. Academic Press, San Diego, 1987.
198
ASSAY OF STRESS GENES/PROTEINS
[16]
slowly unwind around the inside of the bottle, being careful to avoid air bubbles. Once the filter is in place, pour out the 2x SSC and replace with prehybridization fluid. Replace the cap and put the bottle on rotisserie. For either method, prehybridize the filters for at least 2 hr in 10% (w/v) dextran sulfate, 50 mM sodium phosphate (pH 6.8), 1 x SET [150 mM NaCI, 1 mM EDTA, and 20 mM Tris (pH 7.8)], 5x Denhardt's [50x stock equals 1% (w/v) each of ficoll (Type 400, Pharmacia), polyvinylpyrrolidone, and bovine serum albumin (Fraction V, Sigma)], 0.5% (w/v) SDS, and 200/zg/ml denatured, fragmented salmon sperm DNA at 65 ° in a gently shaking water bath. Add at least 10 7 cpm radiolabeled probe to the bag. (Note: It is important to denature both the salmon sperm DNA and probes before hybridization. Heat for 5 min at 100°, then cool immediately on ice before addition.) Hybridize overnight, also in a 65 ° gently shaking water bath. We use the Oligolabeling kit (Pharmacia) for random primer labeling. Typically, we obtain 10 9 cpm per 50 ng of DNA clone using this very simple kit procedure. For nylon membranes, the prehybridization solution should be removed completely and fresh solution plus radiolabeled probe added. The next day, wash the blots 2 times with gentle agitation at room temperature with 2x SSC plus 0.2% (w/v) SDS, then wash twice for 20 min at 65° at 0.2x SSC plus 0.2% (w/v) SDS. Drain and blot the filters with Kimwipes, then wrap the still damp membranes with Saran wrap and expose to Xray film. Notes. It is important to reemphasize here that, when working with RNA, always wear gloves. Also, work quickly and store the RNA in aliquots at - 7 0 ° so the entire sample is not constantly freeze-thawed. All glassware should be either baked (8 hr, 200°) or, at least, extensively autoclaved. Use sterile, disposable plastic when possible. Solutions should be made with DEPC-treated deionized, distilled water. After washing and exposing, filters can be rehybridized with a new probe without significant loss of the RNA, although nitrocellulose filters will withstand only two to three hybridizations. A main advantage of nylon is its ability to be reprobed many times. Nylon is also less brittle than nitrocellulose and hence easier to handle. After washing, do not allow the membrane to dry out. Moist membranes exposed to X-ray film can be rewashed at l o w e r " stringency" if there is too much background radioactivity. This is not possible once the membrane has been dried. Stringency is an important concept in RNA and DNA hybridization procedures. The binding of one nucleic acid polymer to another is dependent on both the salt concentration and temperature. High salt concentration and low temperature, referred to as low stringency, promote the hybridization of two nucleic acid polymers that are not completely homologous. As the salt concentration is decreased and/or the temperature in-
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
199
creased (high stringency), a closer match in sequence is required to maintain hybridization. For exact hybridization (i.e., DNA probes from the same species as the RNA being probed), high stringency conditions (0.1 x SSC, 0.2% (w/v) SDS, washing at 65°) can be used to minimize nonspecific binding. For cross-species hybridizations where there is inexact homology, decrease the final wash stringency (e.g., 0.4x SSC, 60°).
Transcriptional Run-on The basic technique 31'32 described below involves isolating cell nuclei from control and oxidant-treated cells and labeling the nascent RNA transcripts produced with [a-3zP]UTP. The labeled RNA is then hybridized to DNA clones of specific genes bound to a nitrocellulose membrane and quantified by densitometric scanning of X-ray film bands. The following protocol is based on a starting cell number of 3 x 107 mammalian cells. Nuclei Isolation. Following oxidant treatment, collect cultured cells by centrifugation (4 rain, 1100 g, 4 °) and wash with cold phosphate-buffered saline. Resuspend in 9 ml standard RSB buffer [10 mM Tris buffer (pH 7.4), 10 mM NaCI, and 3 mM MgC12] further supplemented with 0.1 mM CaCI2. After resuspension, add 1 ml of 5% (v/v) NP-40 and let sit 30 sec on ice. Centrifuge at 1100 g for 3 min at 4 °. Discard the supernatant and resuspend the pellet in 9 ml of RSB buffer containing 0.1 mM CaCI 2 . Add 1 ml of 5% (v/v) NP-40 to this mixture and centrifuge again as above. Discard the supernatant and resuspend the pellet in approximately 50/zl of nuclei storage buffer [40% (v/v) glycerol, 50 mM Tris buffer (pH 8.3), 5 mM MgCI 2, and 0.1 mM EDTA]. The nuclei suspension can be stored at - 7 0 ° or used immediately for the next step. Transcription. Add the following mixture to the nuclei suspension: 30 ~1 of glycerol; 5 /xl of 100 mM DTT; 2 ~1 of 100 mM MgCI2; 14/xl of 1 M KCI; 2 ~1 each of 25 mM GTP, CTP, and ATP; and 250 /zCi of [c~-32p]UTP (760 Ci/mmol, 10 mCi/ml). Bring the volume to 100/xl with nuclease-free water. Incubate at 37° for 15 min, then add 100 volumes of 20 mM RSB buffer containing 0.1 M CaCI~. Spin at 1100 g for 3 min at 4°. Primary Transcript Extraction. Resuspend nuclei in 3 ml of 0.5 M NaC1, 10 mM Tris buffer (pH 7.4), 50 mM MgCI2, and 2 mM CaC12 . Add 6/xl of RNase free-DNase 1 (10 mg/ml), put the mixture into a 65 ° water bath, and triturate with a siliconized Pipetman tip until fluid ( - 1 5 sec). Be careful not to lose the pellet by its sticking to the tip at this step. Add 60/xl of 20% (w/v) SDS, 60/xl of 0.5 M EDTA, and 66/zl of 3 M sodium acetate (pH 5.4) followed by 5 ml of 60 mM sodium acetate (pH 5.4) containing 10 mM EDTA. Add 400/~g of RNase-free yeast tRNA and 31 A. Udardy and K. Seifart, Eur. J. Biochem 62, 353 (1976). 32 E. Hofer and J. E. Darnell, Sr., Cell (Cambridge, Mass.) 23, 585 (1981).
200
ASSAY OF STRESS GENES/PROTEINS
[16]
extract twice with phenol-chloroform-isoamyl alcohol (25 : 24 : l, v/v/v) preequilibrated with 60 mM sodium acetate (pH 5.4) and 10 mM EDTA, then extract once more with chloroform. Precipitate the RNA with 0.3 M sodium acetate (pH 5.4) and 2 volumes of 95% ethanol at - 2 0 ° overnight. Centrifuge at 200,000 g for 40 min at 4 ° and resuspend in 0.5 M of 10 mM TES [N-tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid] buffer (pH 7.4). Hybridization. The DNA clones of interest (10/xg) are bound to nitrocellulose using a slot-blot apparatus (e.g., BRL Convertible) and baked for 2 hr in a vacuum oven. Prehybridize for 2-4 hr in 10 mM TES buffer (pH 7.4) containing 0.2% (w/v) SDS, 10 mM EDTA, and 300 mM NaCI at 65°. Remove the solution and replace with fresh prehybridization buffer containing 5 × 106 to 107 cpm of labeled RNA transcripts for 36 hr at 65 °. Then wash filter several times with 2× SSC at 65 ° followed by addition of 10/zg/ml RNase A in 2× SSC for 30 min at 37°. Finally, wash the filter for 1 hr at 37° in 2× SSC, air dry, and expose to X-ray film. Notes. For the above example, we used two 15-cm cell culture dishes that were 90% confluent, a yield of about 3 × 107 cells. From this, we typically get 1.2 × 107 cpm of 32p incorporation. The efficiency of release of the small spheroid nuclei from cells can be assessed by microscopy. The calcium and magnesium salts should be autoclaved separately and later added to the RSB solution. In the transcription reaction mixture, the solution should be made up in DEPC-treated water.
DNA Techniques Several different DNA-based techniques are available for identifying mRNAs and assessing modulated expression in either different cell types or in cells treated with modifying agents. These techniques include differential hybridization, subtractive hybridization, and the more recently described differential display. Both differential and subtractive hydridization have been successfully used to isolate sequences modulated by oxidant stress. 33,34Differential hybridization is the most commonly used of these techniques. It is a significantly easier technique than subtractive hybridization. Subtractive hybridization, however, is much more sensitive and is able to detect modulated sequences up to 50 times less abundant than necessary for differential hybridization. Differential display is a new technique whose sensitivity approaches that of subtractive hybridization. Its 33 S. M. Keyse and E. A. Emslie, Nature (London) 359, 644 (1992). 34 A. J. Fornace, Jr., D. W. Nebert, M. C. Hollander, J. D. Luethy, M. Papathanasiou, J. Fargnoli, and N. J. Holbrook, Mol. Cell. 9, 4196 (1989).
[16]
G E N E EXPRESSION D U R I N G O X I D A T I V E STRESS
201
value in identifying oxidant-modulated sequences is unproved. This section includes methodology for each of these major techniques.
Differential Hybridization Differential hybridization, like subtractive hybridization and differential display, is a technique that compares two preparations of mRNAs and identifies those species that are present at different levels. An example would be mRNA extracted from control cells and from cells that have been exposed to heat shock or oxidative stress. Invariably, the levels of the majority of mRNAs do not change after treatment. However, a small proportion of the mRNAs will increase or decrease in the stressed cells. DNA reverse transcribed from mRNA is known as complementary DNA (cDNA). Synthesis of cDNA can be done in the presence of a labeled deoxynucleotide, such that the product (first-strand 32p-labeled cDNA) can be used as a probe. The probe is used to detect any RNA or DNA that shares sequence similarity. A library is a collection of DNA sequences that represent all the mRNA ( " c D N A library") or genes ("genomic library") present in a cell. A labeled cDNA probe can thus be used to determine whether any of the cDNAs present in a library are homologous. This is known as screening a cDNA library. For example, if the labeled probe is a known sequence such as catalase, screening a library will identify those clones that are catalase and nothing else. (Note. A clone is a cDNA that has been inserted into another type of DNA known as a vector that is amplified many times over, usually in bacteria or virus. The vectors thus have the essential quality of being able to replicate in the bacteria or virus.) In the case of differential hybridization, the probe consists of cDNA derived from an entire population of mRNA. First, a cDNA library is made from mRNA derived from oxidant-treated cells. The library is then probed with labeled cDNA made from mRNA extracted from control cells, followed by reprobing with cDNA made from mRNA extracted from t r e a t e d cells. 33 Clones that hybridize preferentially to one of the cDNA probes are selected for further characterization. Alternatively, a genomic DNA library may be probed with control and with treated cDNA. In a genomic library, however, each clone contains the coding region of several genes, only one of which is likely to be induced. A reverse Northern blot (see below) will help delineate which gene is induced. There are two potential drawbacks to differential hybridization. First, rare mRNA sequences will have low specific probe concentrations and may not hybridize to the DNA from the library in a reasonable amount
202
ASSAY OF STRESS GENES/PROTEINS
[16]
of time. 35 Second, genes that show a low level of induction (e.g., 2- to 3fold) will not give a substantially stronger hybridization signal and may be difficult to detect. Labeling of cDNA Probes. To label rare cDNA molecules, cDNA probes must be representative of the mRNA population and labeled to high specific activity (e.g., 0.5-2.0 x 108 cpm). 36 Reverse transcriptase reactions are performed with poly(A)-containing RNA as a template and random hexanucleotides as primers. Reactions contain 0.5/xg of poly(A)containing RNA, 1.25/~g random hexanucleotides, 50 mM Tris (pH 8.3), 6 mM MgC12 , 40 mM KCI, 10 mM DTT, 100/xg/ml BSA, 0.5 units RNase Block II (Stratagene), 0.5 mM dATP, 0.5 mM dGTP, 0.5 mM dTTP, and 50/xCi [a-32p]dCTP (3000 Ci/mmol) in a final volume of 20 ~l. Reactions are incubated for 60-90 min at 37° after the addition of l0 units of Moloney murine leukemia virus (M-MuLV) reverse transcriptase (Boehringer Mannheim). (Note: it is important that relatively fresh, highly active transcriptase be used for the reaction.) The reaction is stopped by adding l /xl of 0.5 M EDTA and 0.5/xl of 20% (w/v) SDS. The RNA is hydrolyzed by the addition of 3/zl of 3 N NaOH and incubated at 68 ° for 30 min. The reaction is allowed to cool to room temperature, and then l0/zl of 1 M Tris (pH 7.4) and 3/A of 2 N HC1 are added. Probes are precipitated with trichloroacetic acid (TCA) to determine specific activity. 37 Bacterial Colony Replicas. To isolate genes responding to oxidative stress, a plasmid cDNA or genomic DNA library is constructed 25 and plated out onto filters. Membranes made of nylon or a blend of nylon and nitrocellulose are less brittle than nitrocellulose and stand up well to reprobing. Membranes (137 ram) are placed on top of an LB agar 150mm plate, and 40,000 bacterial colonies are plated onto the membranes at a density of 1000-2000 colonies per membrane. We have found that plasmid libraries give a stronger hybridization signal than do bacteriophage plaques. Replicas of each plate are made by placing the master membrane colony-side up on a sheet of Whatman 3MM paper, overlaying it with 35 j. H. Weiss, in "Current Protocols in Molecular Biology" (F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl, eds.), p. 5.8.14. Greene Publishing Associates and Wiley (Interscience), New York, 1989. 36 D. D. Chaplin and B. H. Brownstein, in "Current Protocols in Molecular Biology" (F. Ausubel, R, Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl, eds.), p. 6.3.1. Greene Publishing Associates and Wiley (Interscience), New York, 1989. 37 H. Perry-O'Keefe and C. M. Kissinger, in "Current Protocols in Molecular Biology" (F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl, eds.), p. 3.4.7. Greene Publishing Associates and Wiley (Interscience), New York, 1989.
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
203
another membrane, covering it with a 20 × 20 cm glass plate, and pressing down. 19 The master membrane is then placed colony-side-up back onto the LB agar plate and left at room temperature overnight. The newly made replica is also overlaid onto an LB agar plate and incubated at 37° overnight. Up to two replicas may be made from each master filter. After overnight growth, the master filters are stored at 4° while screening the replicas. Lysis of Bacterial Colonies. Filters are removed from the LB agar plates and placed colony-side-up on a sheet of Whatman 3MM paper, soaked with 10% (w/v) SDS for 3 min. 25 Filters are denatured on a sheet of Whatman 3MM paper, soaked with 1.5 M NaCl and 0.5 M NaOH, and incubated for 5 min. The third transfer is to a Whatman 3MM sheet, soaked with 1 M Tris (pH 7.4) and 1.5 M NaC1 for 5 min to allow neutralization to occur. Filters are then transferred to a sheet soaked with 2x SSC for 5 min. The filters are air dried for 30 min and baked at 80° for 90 min. Hybridization of Colony Lifts. A set of filters may be probed with control cDNA and reprobed with treated cDNA, or, alternatively, duplicate sets of filters may be probed with control or treated cDNA. Hybridizations are performed in heat-sealable plastic bags (Dazey Industrial Airport, KS). Hybridization to filters is conducted as for Northern blots on a shaking platform. Each bag has two filters in 10 ml of hybridization solution and is prehybridized at 65° for 2-3 hr. Approximately 5 ~g of control RNA is labeled, and 2 x l 0 7 cpm of cDNA is added to each bag for 18 hr. It is important not to use too much probe in the hybridization. If all potential binding sites are saturated during the first hybridization (e.g., with control cDNA), there will be none available for binding during the second hybridization (e.g., with treated cDNA). Consequently, induced genes would not be detected. The filters are washed two times with 20 ml of 0.4× SET plus 0.2% (w/v) SDS at 65° for 90 min each. After washing, all filters are covered with plastic wrap and exposed at -70 ° under Kodak X-OMAT X-ray film with an intensifying screen. Filters should not be allowed to dry out between hybridizations. The whole process is repeated when probing the same filters with the treated cDNA probe. A comparison of the same filters probed with both control and treated cDNA should yield colonies that differ in hybridization signal. These are the clones that are of interest. Candidate clones should be confirmed by using them as probes for Northern blots. This is important because differential hybridization is a screening technique and is not absolute proof that a gene is differentially expressed. Reverse Northern Blotting. When a genomic DNA library is constructed, genomic DNA is digested with a restriction enzyme and cloned
204
ASSAY OF STRESS GENES/PROTEINS
[16]
into a vector. Because restriction sites occur randomly, a given clone may contain all or part of the coding region of several genes. A genomic DNA clone that appears to be positive by differential hybridization, therefore, contains several genes, only one of which is likely to be differentially expressed. Reverse Northern blotting is a useful technique for determining which of the genes is induced. In this method, candidate clones are digested with several restriction enzymes to generate fragments containing 300-400 base pairs. Fragments of this size are unlikely to contain the coding region of more than one gene. After electrophoresis and transfer to a membrane, clones are probed with control cDNA and with treated cDNA. Typically one or two fragments per clone will be differentially expressed, corresponding to one gene. Digests are run on agarose gels [typically 1% (w/v) agarose in 1× TEAC buffer (2 mM EDTA, 18 mM NaCI, 12 mM sodium acetate, and 40 mM Tris, pH 8.05)] and prepared for blotting. DNA is fragmented by exposing the gel to UV light for 30 sec followed by soaking in 0.5 N NaOH plus 1.5 M NaC1 for 1 hr. Neutralization occurs during a 1-hr incubation in 0.5 M Tris (pH 7.5) plus 1.5 M NaCI. The gel is then blotted to filter as previously described above. The membrane is probed with control cDNA and then reprobed with treated cDNA. Labeling of cDNA probes and hybridization of membranes are as described earlier in this section, except that filters are washed twice in 100 ml per filter of 0.4× SET plus 0.2% (w/v) SDS at 65 ° for 90 min each. Examination of the membrane after probing with control and treated cDNA will reveal one or two fragments per clone that are differentially expressed, presumably corresponding to one gene. The fragments can be isolated and used as probes for Northern blots. This technique can also be used to reaffirm putative positive cDNA clones derived from a differential or subtractive hybridization screening. In this case, cDNA (instead of genomic DNA) from a number of potential positive clones are run out on an agarose gel, transferred, and probed as above.
Subtractive Hybridization A number of different subtractive hybridization techniques are now described in the literature, including ones based on the polymerase chain reaction (PCR) and phagemid vectors. We prefer a more conventional version of the procedure, described by Fargnoli et a1.,38because it has been successfully used to detect and clone oxidant stress-modulated sequences, 38 j. Fargnoli, N. J. Holbrook, and A. J. Fornace, Jr., Anal. Biochem. 187, 364 (1990).
[16]
G E N E EXPRESSION D U R I N G O X I D A T I V E STRESS
205
and because it best mimics in vitro the ratios of mRNA that actually exist in control versus treated cells. In the procedure, the critical hybridization step is carried out between the mRNA of control cells and first-strand cDNA of treated cells, a minimal change from actual cell concentrations. Although the PCR and phagemid procedures are more powerful and flexible, a number of steps must be performed before the actual hybridization takes place. Each step, including second-strand synthesis, cloning, singlestrand phage rescue, phage recovery, and extractions, can and does vary between samples and hence may become less and less faithful representations. The steps additively increase the chances that hybridization will not be reflective of in situ ratios. We therefore recommend the following procedure previously described 34 with some modification. Cells and Treatments. A critical part of any subtractive hybridization protocol is the choice of cells from which mRNA is to be extracted and their state at the time of extraction. The key is to match the cells as closely as possible or, in the case of some cell treatment including oxidant stress, to use as specific a modifying agent as possible. When properly executed, subtractive hybridization is a very sensitive technique, especially when conditions are used that maximize the detection of small inductions and reductions of low-abundance mRNA species. If widely disparate cell lines are used (such as certain transformed versus nontransformed or oxidantresistant versus oxidant-sensitive lines), many hundreds of subtracted clones will be generated, most of which will be nonspecific. Solutions. The usual extra precautions for RNase-free solutions and glassware are of absolute necessity. In addition, the use of molecular' biology grade reagents, such as those offered by Sigma, is strongly recommended. These reagents have no detectable RNase or DNase activities and are low in metal content, all definite advantages for the procedure. First-Strand Synthesis. Procedures for the initial cDNA synthesis, and selected other steps, are taken from Fargnoli et al. 38 For first-strand synthesis, 30-100/~g (preferably 100/zg) of poly(A) + mRNA from oxidantstressed cells is heated at 75 ° for 3 min. First-strand reagents are then added to the following final concentrations: 60 mM KCI; 50 mM Tris buffer (pH 7.6); 7 mM MgC12; 300/zg/ml RNase inhibitor; 2 mM DTT, 60 ~g/ml actinomycin D; 7 /~g/ml oligo(dT)12_18 and 200/xg/ml random hexamers (both from Pharmacia); 1 mM each of dATP, dCTP, dGTP, and dTTP; 50/xCi/ml [32p]dATP or [32P]dCTP; and 40/zg/ml poly(A) ÷ RNA. The mix is incubated for 10 min at 37°, then 16,000 units/ml M-MuLV reverse transcriptase is added, and the reaction is allowed to continue for 75 rain at 37°. Add EDTA to 20 mM to stop the reaction, then hydrolyze the RNA in 150 mM NaOH for 30 min at 65°. Neutralize with 2 M Tris buffer (pH 7.5) added to a final concentration of 150 mM. Remove lower
206
ASSAY OF STRESS GENES/PROTEINS
[16]
molecular weight molecules (<30,000) and concentrate the first-strand preparation sample using prewashed Microcon 30 concentrators (Amicon, Danvers, MA). The prewash should include I0 mM Tris buffer (pH 7.5), 1 mM EDTA, and 0.05% (w/v) SDS. Save 1 /xg of the concentrate for later use as the nonsubstracted probe. The rest should be organically extracted with an equal volume of phenol-chloroform (see "Basic Nucleic Acid Techniques" section above). Vortex well, then spin for 2 rain in a microcentrifuge. Remove the top aqeous layer to a fresh tube, then repeat the extraction with chloroform-isoamyl alcohol (24 : 1 (v/v)) only. Recentrifuge, remove the top phase, and wash extensively with 50 mM HEPES buffer (pH 7.6), 1 mM EDTA, and 0.1% (w/v) SDS, again using a prewashed Microcon 30. Biotinylation. Dilute control cell mRNA to 1/zg//zl if the concentration is greater than this. Mix an equal volume of the mRNA with Photoprobe biotin (Vector Labs, Burlingame, CA). Irradiate in an ice bath 10 cm below the sunlamp for 15 min. Unreacted biotin is removed by bringing the reaction to 100 mM Tris (pH 9.0) and 100/zl final volume, then extracting with an equal volume of 2-butanol saturated with 10 mM Tris (pH 8.0) plus 1 mM EDTA (TE). Vortex, centrifuge, and discard the upper layer. Extract the remaining RNA with chloroform-isoamyl alcohol, then precipitate the upper, RNA-containing phase with 0.1 volume of 3 M sodium acetate (pH 5.3) and 2.5 volumes of HPLC-grade ethanol. Two cycles of biotinylation yield a biotin label for every 50-100 RNA bases. The biotinylation step has been well described for subtractive hybridization by Sive and St. J o h n 39 and Nottenburg et al. 4° Hybridization. The biotinylated RNA is then hybridized to singlestranded cDNA from the oxidant-treated cells. The final mix should contain 50 m M HEPES buffer (pH 7.6), 0.1% (w/v) SDS, I mM EDTA, and 500 mM NaCI (added last). An equal mass ratio of cDNA to RNA is used. Overlay the reaction mix with paraffin oil, heat to 95 ° for 3 min, then hybridize at 65 ° for 24-48 hr. Dilute to 100/.d with same buffer minus SDS. Add streptavidin (Vector Labs), mix for 1 min at room temperature, and extract with TE-saturated phenol-chloroform (1 : 1, v/v). Centrifuge the sample and transfer the unhybridized cDNA-containing upper phase to a fresh tube. Back-extract the remaining lower organic phase by adding 25/.d buffer, vortexing, and respinning. Combine this upper phase with the previously collected upper phase. Repeat the streptavidin step on the combined upper phases, then extract the pooled aqueous phases with chloroform. Finally, nucleic acid is precipitated with sodium acetate (pH 5.3) and ethanol. After centrifuging, washing the pellet with 75% ethanol, 39 H. L. Sive and T. St. John, Nucleic Acids Res. 16, 10937 (1988). 40 C. Nottenburg, W. M. Gallatin, and T. St. John, Gene 95, 279 (1990).
[16]
GENE EXPRESSION DURING OXIDATIVE STRESS
207
and recentrifuging, resuspend the pellet in 50 mM HEPES buffer (pH 7.6), 0.1% SDS (w/v), 1 mM EDTA, and 500 mM NaC1. Positively select the resuspended pellet by repeating the hybridization step, this time using RNA from oxidant-treated cells. This positive selection removes nonhybridizable species. The solution is then adjusted to 0.1 M NaOH, heated to 65 °, and neutralized with 2 M Tris buffer (pH 7.5). Remove hydrolyzed RNA with the Microcon 30 by washing with TES buffer, then precipitate with ethanol and resuspend in water. Cloning. Clone into h ZAPII according to the manufacturer (Stratagene). This vector combines a high-efficiency, high-yield h vector with an option to subclone automatically into a phagemid vector. This usually means high-titer libraries with an option to screen the library as bacterial colonies. Bacterial colonies contain much greater amounts of DNA than phage plaques, maximizing the chances of detecting low-abundance sequences. Alternatively, h Shlox (Novagen, Madison, WI), a similarly designed cloning vector, can be used. The library is then screened by standard technique as described above (see "Differential Hybridization") and in further detail in Sambrook et al. 25 An example of colony blots hybridized with probes before and after subtractive enrichment is shown in Fig. 6. Notes. The above protocol uses low-ratio mRNA : cDNA hybridization conditions. This protocol has been championed by several investigators over conventional high-ratio subtractive hybridization and is our method of choice. The conventional high-ratio procedure has somehow evolved to include a 10- to 20-fold excess of control mRNA to treated cDNA. Although this procedure is simpler than low-ratio hybridization, it is also biased toward, and will only pick up, strongly induced mRNAs. We prefer low-ratio hybridization because it can detect mRNAs modulated only severalfold, some of which may be as critical to stress response as strongly modulated mRNAs. The downside to this approach is a higher number of positives clones, both true and false, that require screening.
Differential Display The differential display technique was described by Liang and Pardee 41 and used to isolate sequences that differ between normal and transformed cells. The basic protocol involves extraction of mRNA followed by firststrand cDNA synthesis for use as a template for subsequent PCR amplification. The key to the technique is the use of a number of combinations 41 p. Liang and A. B. Pardee, Science 257, 967 (1992).
208
ASSAY OF STRESS GENES/PROTEINS
[16]
A
B
FIG. 6. Subtractive hybridization. Replicate filters of bacterial colonies containing part of a cDNA library were probed with cDNA made to mRNA prior to enrichment (A) or cDNA made to mRNA after subtractive enrichment (B), as described in the text. Colonies preferentially hybridizing to the enriched probe are numbered. Colony A represents a reduced, originally abundant sequence. (Reproduced from Fargnoli et al. ?8 with permission.)
of 5' and 3' primers that, in total, represent all the mRNA species present in an average cell. The PCR products are then run on a standard sequencing gel. Modulated sequences are determined by simple comparison of lanes. The comparative aspect of the procedure is similar to 2D protein gel electrophoresis. Although this technique has not been widely used, its sensitivity approaches that of subtractive hybridization. Furthermore, reduced as well as induced sequences are easily detectable as opposed to subtractive and differential hybridization, where usually only induced species are identifiable.
[16]
G E N E EXPRESSION D U R I N G O X I D A T I V E STRESS
209
Cells and Treatments, Solutions, and mRNA Preparation. The preparation of cells, solutions, and mRNA is performed as for subtractive hybridization above. Primers. Combinations of 5' arbitrary primers and 3' dinucleotideextended oligo(dT) primers are prepared as described by the manufacturer (e.g., Applied Biosystems, Foster City, CA, DNA synthesizer) or are obtained commercially. The arbitrary primers should be 10-mers and contain 50-60% GC residues. These arbitrary primers will bind to the cDNA template in a degenerate fashion, acting as 6- to 7-mers when annealed at 40 ° and will thus anneal fairly frequently to the DNA. The 3' primers will be 5' TTTTTTTTTTTTMA 3' (where M is any base), TI2MC, T12MG, and T~zMT. Theoretical calculations indicate that screening 25 arbitrary primers with the above four T~2MN oligomers will analyze for roughly 50% of the mRNA species present in a given cell, assuming there are approximately 15,000 different mRNA species per cell. Screening 75 arbitrary primers plus all four T~2MN oligomers covers about 90% of the cellular mRNA species. Polymerase Chain Reaction. Although poly(A) ÷ mRNA preparations are sufficient for the PCR, cleaner results are actually obtained by using total RNA. We either prepare the reagents ourselves or buy a commercial kit (RNAzol, Biotecx Labs, Inc.). The RNA is first treated with RNasefree DNase (as for subtractive hybridization), checked for integrity by ethidium bromide staining of formaldehyde gels (see "Northern Blot Electrophoresis"), and spectrophotometrically quantified. Dilute these preparations (usually RNA from control and oxidant-treated cells) to 0.1 /~g//.d and then pipette the following reagents to produce a first-strand cDNA template: 9.4 ~1 DEPC-treated water, 4/xl of 5 × reverse transcriptase buffer, 1.6 ~1 of a 250/zM mix of each deoxynucleoside triphosphate (dNTP), 2/~1 of the 0.1 txg//xl DNase-treated RNA, and 2 tzl of the T12MN of choice (10/xM). The solutions are heated for 5 min at 65° and then put at 37° for 10 min; 1 /.d of stock M-MuLV reverse transcriptase is added to each tube, and the tubes are further incubated at 37° for 50 min longer. The tubes are then heated for 5 min at 95 °, centrifuged briefly (3 sec), and put on ice for the subsequent PCR amplification. Alternatively, the reaction tubes can be stored at - 2 0 °. For the PCR, add the following: 9.2/xl water, 2/zl of 10× PCR buffer [100 mM Tris buffer (pH 8.3), 15 mM MgC12, 500 mM KCI], 1.6 tzl of a 25/xM solution containing a mix of each dNTP, 2 ~1 of an arbitrary primer (2/~M) of choice, 2/xl of a T12 oligomer (10/zM) of choice, 2/.d from the above reverse transcription reaction, 1/~1 [35S]dATP (1200 Ci/mmol), and 0.2/zl Taq polymerase. Usually a core mix of all of these reagents, except primer(s) and reverse transcription reaction mix, is made first and dis-
210
ASSAY OF STRESS GENES/PROTEINS
[16]
pensed into each tube for more uniform results. After mixing, reactions are overlayered with 25/xl mineral oil, and the PCR is run for 40 cycles, each cycle consisting of the following three segments: 94 ° for 30 sec (denaturation), then 40 ° for 2 min (annealing), then 72° for 30 sec (elongation) with a 5-min extension at 72 ° at the last (fortieth cycle) elongation. The tubes are then placed on ice. Sequencing. Mix 3.5/~1 of each reaction with 2/xl of sequencing loading dye, heat for 2 min at 80°, place on ice, and apply to a prerun 6% (w/v) urea sequencing gel. We use Gel-Mix 6 (GIBCO-BRL) as our sequencing solution source. Gels are run at 55 W until the xylene cyanol dye is 10 cm from the bottom. Remove the gel and dry directly without acetic acid/ methanol treatment. Mark around the gel with radioactive ink (ink-well pen that has [35S]methionine mixed into the ink) for later alignment. Expose to X-ray film for 2 days at room temperature. Analysis and Processing. Simply compare control with treated lanes and look for modulation. Carefully line up the radioactive ink signals over the dried gel and cut through the film with a razor blade to excise the modulated band of interest. (Note: We expose the gel to X-ray film a second time before band excision to retain a permanent record of the modulation.) Place each band in 100/A water (in a siliconized Eppendorf tube) for 10 min. Boil for 15 min in a heating block, then centrifuge for 2 min in a microcentrifuge and collect the acrylamide/Whatman paper-free supernatant. Precipitate with 3 M sodium acetate (pH 5.2), glycogen, and ethanol (to 85%). Spin for 10 min at 4 °, wash the pellet with 85% cold ethanol, respin, and thoroughly remove the ethanol. Let the pellet semidry on the laboratory bench, then resuspend in 10/A of distilled water. Amplify by the PCR as before except use 250/zM stock dNTPs and twice the volume (40/zl). Also, no radioactivity is required at this stage. After 40 cycles, check the product on a gel. If a good signal is observed, move on to cloning the probe. If not, conduct the PCR again using 4/zl of the first PCR mix as template. Cloning. TA cloning kit (Invitrogen, San Diego, CA) is used for the cloning step. The specifics can be obtained from the manufacturer. In general, the above PCR products are ligated overnight into the specially designed cloning vector pCRII. The unique feature of this vector is that it has a single A overhang on each end. This allows it to readily hybridize to Taq polymerase PCR fragments, which characteristically have a onebase T overhang. The ligation products are then cloned in bacteria and plated. Colonies are grown in liquid culture followed by minipreparation DNA extraction. 25 Inserts are excised, run on agarose gels, and the bands excised either by Qiaex (Qiagen Inc., Chatsworth, CA) or in low melt agarose. The purified bands are then labeled for use as hybridization
[16]
G E N E EXPRESSION D U R I N G O X I D A T I V E STRESS
211
probes of Northern blots (see below). Any modulated mRNAs that show up on the blots, and that correspond with those observed on the original differential display, represent bona fide oxidant-modulated sequences. Notes. The differential display technique has several advantages over subtractive hybridization. For one, it is able to pick up both reduced and induced species. Second, it allows one to monitor the effectiveness of the technique throughout the procedure (unlike the subtractive method, where the result is not known until the end). Third, it requires considerably less RNA. However, we have been somewhat disappointed by the modest modulations we see on the sequencing gels as well as by the percentage of false positives as assessed by Northern blot hybridization. It is likely that the technique works significantly better when comparing two different cell lines (e.g., oxidant-adapted versus nonadapted) rather than the same cell line with oxidant treatment versus no treatment. In fact, the original procedure was worked out using two different cell lines (transformed versus nontransformed). In that case, a number of modulated sequences were identified based on their complete absence in one cell line. Otherwise, modulation of control mRNAs already present is susceptible to differences in PCR amplification between samples, especially at the plateau phase where the sample with lesser signal can "catch up." Moreover, although the protocol does not sound overwhelming in terms of numbers of PCR setups and sequencing gels on paper, in practice we have had to run many repeat reactions and gels (including duplicate reverse transcription template reactions) to obtain satisfactory results. An example of a potential modulated sequence is shown in Fig. 7. It is important to repeat the differential display result using cDNA template derived from a separate reverse transcription reaction. If the results are not confirmed, the sequence is used as a probe to assess mRNA modulation directly on a Northern blot. Genetic Strategies The existence of cellular factors responsible for resistance to oxidative damage can be demonstrated using genetic techniques. Mutations may cause either loss of normal resistance (i.e., a hypersensitive phenotype) or increased resistance (a resistant phenotype). General methods for obtaining and exploiting mutations of each type are outlined below. Mutations Causing Decreased Resistance A principle of genetics is that functions are discovered when they are lost by mutation. Hence, if we can isolate mutants which are hypersensitive to oxidative damage we can identify genes responsible for resistance
212
ASSAY OF STRESS GENES/PROTEINS
[16]
C P
IP-
FIG. 7. Differential display of PCR-amplified DNA derived from mouse epithelial cell mRNAs. Total RNA was extracted from the monolayer cells, and first-strand cDNA synthesis and PCR amplifcation with specific primers were executed according to the text. C, Control untreated cells; P, cells incubated with 25 p~M H202 for 75 min. Arrow indicates location of potential down-regulated sequence. (Note. It is important to repeat any modulation as explained in the text.)
in one way or another. Such mutants can be obtained in microbes by replica plating. Because the number of colonies which can be screened is limited in practical terms, it is desirable to increase the mutation frequency above the spontaneous rate, usually with a chemical mutagen such as EMS (ethylmethanesulfonic acid). After mutagenesis and a brief recovery period in rich medium (e.g., about 1 hr for bacteria and 3 hr for yeast), cells are seeded onto agar plates containing rich medium at a density sufficient to produce about 800 colonies per 100-mm plate. The colonies are then replica-plated onto plates containing sublethal concentrations of oxidant (e.g., hydrogen peroxide or paraquat). The sublethal concentration (that which permits about 80% survival) is determined ahead of time on plates containing various concentrations of the oxidant; the replica plates are then made up exactly the same way. Colonies which show reduced
[16]
G E N E EXPRESSION D U R I N G O X I D A T I V E STRESS
213
growth when replica-plated are picked from the master plate for further testing. To confirm the mutation, the mutant strain should be streaked onto rich medium plates for isolation of single colonies. A few of these are then streaked on an oxidant-containing plate, alongside the parent strain, to confirm hypersensitivity. When a collection of confirmed mutants is obtained, each isolate is tested for stability by plating about 107 cells on I0 oxidant test plates to be sure that the reversion is less than 10-8, indicated by absence of colonies on the plates. It is better to avoid high cell densities for this test because of the possibility of local destruction of the oxidant by large cell concentrations on the agar; to be sure of the validity of the reversion test it is well to mix a small number of wild-type cells into the aliquot of mutant cells added to one plate, to be sure that revertants would form colonies if they were present. Mutants are then stored at - 8 0 ° in medium containing 15% glycerol. Mutations can affect a single gene, or they can be pleiotropic, affecting several genes. One quick way of assessing this is by running 2D protein gels of the mutant strains. In other experiments, one would have established whether the organism being studied produces any proteins in response to oxidant stress which are detectable on a 2D gel. If so the mutant can be tested under the same conditions, provided that the inducing level of oxidant is sublethal for the mutant. If cells are too sensitive to induce with oxidant it may still be possible to see protein differences from the wild type in untreated cells. A pleiotropic mutant would show the absence of more than one protein. This type of mutant would be particularly interesting because the gene affected by the mutation would be likely to be involved in regulating genes involved in mounting resistance to the oxidant. Mutations affecting single loci (e.g., glutathione peroxidase) would show up as a loss of a single protein. Genes identified by mutagenesis can usually be readily cloned by complementation provided the species involved can be transformed by plasmids. The simplest approach is to introduce a plasmid library constructed with wild-type chromosomal DNA 25 into the mutant to produce a pool of transformants, among which is one carrying the wild-type gene. If the species involved is bacterial the vector used in construction of the library should if possible provide an inducible promoter in case the gene of interest has been cloned into the library without inclusion of its own promoter. The pool is plated onto the oxidant plate, and, barring complications, cells carrying the wild-type gene would regain resistance and be able to grow. The resulting colony constitutes a clone of the gene of interest, which is recovered by purification, Before sequencing the complementing region should be narrowed down by subcloning, using standard recombinant
214
ASSAY OF STRESS GENES/PROTEINS
[16]
techniques25; various segments are deleted, and the resulting set of plasmids are retested for ability to confer resistance. Once the gene has been localized by this procedure it may be sequenced, providing some indication, it is hoped, of its function, by comparison with sequences of other genes in the database. For complementation cloning as described the species must be competent for transformation. Advances in electroporation techniques (see, e.g., promotional literature from Bio-Rad) have greatly expanded the range of conveniently transformable species. Some can be transformed with linear DNA by homologous recombination (e.g., Bacillus subtilis or Neurospora crassa). In such cases one can construct a genomic library by cloning DNA digested with a six-base cutter restriction endonuclease into an E. coli vector (such as pUC19) cut with the same enzyme. DNA fragments excised from the library are then used to transform the mutant to wildtype genotype and phenotype. In this case the plasmid clones carried in E. coli have to be correlated with transformed mutant colonies, since the integrated gene cannot be recovered from the transformed mutant. To identify the plasmid carrying the gene, fractions of the library (i.e., if 10,000 clones are sufficient to include the entire genome, it is partitioned into pools of 1000 clones) are used to transform mutant cells. After two generations of growth for recovery, transformants are selected for complementation of the mutation, scored as resistance to the oxidant; if the library has been appropriately partitioned only one or two of the ten transformations will produce resistant colonies. The library subpool containing the complementing DNA is partitioned again (into pools of 100 clones at this point), and the process is repeated 3-4 times to obtain a homogeneous pool of plasmid-bearing E. coli, constituting in effect a clone of a fragment carrying the gene of interest. Subcloning by deletion, as described above, and retesting for complementation would lead to isolation of the gene in a form which can be sequenced. Mutations Causing Increased Resistance
In addition to mutations which cause increased sensitivity it is also possible to isolate mutants with increased resistance, most likely achieved by virtue of overexpression of protective factors. Mutants of this type have been used to identify two regulons in E. coli involved in resistance to oxidants, namely, o x y R 42 and soxR. 2 Such mutations are easier to obtain at the outset, because there is a positive selection for them, but the genes are generally more difficult to clone. Resistant mutants are selected simply 42 M. F. Christian, G. Storz, and B. N. Ames, Proc. Natl. Acad. Sci. U.S.A. 86, 3484 (1989).
[16]
G E N E EXPRESSION D U R I N G O X I D A T I V E STRESS
215
by plating cells on barely lethal concentrations of oxidant, the concentration being predetermined as that which kills about 99.8% of the cells in the same sensitivity test described above. Resistant colonies are picked and retested (the majority may actually prove to be still sensitive because of the nonstringent selection, so retesting is essential; however, a reversion test is unnecessary). The 2D gel test is used again, but now to look for proteins that are overexpressed in the mutant under inducing and/or noninducing conditions. If the mutation is pleiotropic, causing overexpression of a number of proteins, there are two likely causes: the mutation has either inactivated a repressor or altered an activator in such a way that it is constitutively active or overactive. Usually the former type of mutation would be recessive, the latter dominant. As there is no simple way of knowing or testing this a priori, it may be necessary to attempt different cloning strategies. On the assumption that the mutant gene encodes a repressor, the mutant can be transformed with a wild-type plasmid library and transformants then replica-plated on oxidant-containing plates to identify colonies that have been complemented, that is, restored to wild-type sensitivity. The plasmid carrying the complementing gene can then be recovered from the colony on the master plate. On the other assumption, that the mutant gene encodes a constitutive activator, a library can be constructed from DNA prepared from the mutant and used to transform wild-type cells, which are then plated on oxidant-containing plates to select for cells carrying a plasmid encoding the mutant activator. Detection o f Resistance Mutations in Higher Eukaryotes The foregoing describes methods available for identifying resistance genes in haploid organisms, where recessive mutations (loss-of-function) are phenotypically evident. In diploid organisms, recessive mutations are silent, and mutants are generally much more difficult to detect. By obtaining mutations in both chromosomal copies of a gene it is possible to determine the null phenotype, but this is only practical with current technology when the gene has already been identified and cloned. Hence it is not generally practicable to use mutagenesis to search for previously unidentified functions. The one possible exception to this is the case of a regulon which is controlled by an activator, in which it would be possible to obtain pleiotropic dominant constitutive mutations as described above. The success of such an approach would depend first on the nature of the regulatory pathways of the resistance system and second on whether a method was available for screening large numbers of cells for resistance to oxidation, given that the hypothetical dominant mutation would be
216
ASSAY OF STRESS GENES/PROTEINS
[16]
rather rare. In general we believe methods based on differential mRNA expression or on isolation and sequencing of oxidant-induced proteins are a better approach to hunting for resistance factors in higher eukaryotes. Concluding Remarks In general, use of the above methods to assess oxidant-modulated gene expression has been a relatively recent event and has significantly lagged behind the availability of the techniques themselves. This is probably due to the fact that the oxidant field did not enjoy widespread popularity until recently, as well as to the lack of recognition that modulation was such a prevalent event in response to oxidative stress. We now know that a number of different sequences are significantly modulated by oxidative stress. These sequences are included in the general groups of antioxidant enzymes, protooncogenes, DNA repair genes, growth arrest genes, cytoplasmic signal transduction factors, and heme oxygenase. With increased research in the field and the availability of the above techniques, it is likely that additional modulated sequences within these classes will be identified as well as those from other classes such as differentiation and apoptosis. Identification of protective response genes will be valuable in understanding the specifics as to how different cells respond to oxidative stress. Detrimental response genes may ultimately be valuable in the detection and therapy of oxidant-related pathologies such as aging and cancer. Within the next several years, we are likely to see the development of some newer and more powerful techniques, especially those involving the PCR. We have described one such method, differential display, in this chapter. There are also several PCR-based subtractive hybridization procedures in the literature that may ultimately make this a much simpler procedure while maintaining sensitivity. New approaches for identifying modulated gene expression in bacteria should also appear, including one procedure that uses a reverse Northern blotting approach to identify different stress-response genes. 8 Improvements in the protein field are also likely, especially in the development of preparative gels that will allow easier identification of modulated proteins. In this chapter, we have included methods that assess oxidant-modulated products of gene expression. Owing to space limitations, we have not described how oxidants can modulate expression at the gene promoter level. It is worth noting, however, that promoter studies have been used successfully to monitor gene promoter responses to oxidant stress. These include study of the bacterial OxyR promoter with an OxyR/fl-galactosidase fusion constructl8; study of the mammalian antioxidant response
[17]
OxyR REGULON
217
element (ARE) with an ARE/chloramphenicol acetyltransferase fusion construct43; and study of human immunodeficiency virus type 1 (HIV-1) activation by oxidant-induced NF-KB by gel shift analysis. 44 For an overview of gene regulation and oxidative stress, especially in bacteria where a greater gene response to oxidative stress has been observed, the reader is referred to Farr and Kogoma. ~2 43 T. H. Rushmore, M. R. Morton, and C. B. Pickett, J. Biol. Chem. 266, 11632 (1991). 44 R. Schreck, P. Rieber, and P. A. Baeuerle, EMBO J. 10, 2247 (1991).
[ 17] O x y R R e g u l o n
By GISELA STORZ and SHOSHY ALTUVIA Introduction
Escherichia coli and Salmonella typhimurium possess several different inducible responses to defend against the deleterious effects of reactive oxygen species) ,2 The oxyR-controlled regulon of hydrogen peroxideinducible genes in S. typhimurium and E. coli has provided a useful model for studying how bacterial cells defend against oxidative stress. Treatment of bacterial cells with low doses of hydrogen peroxide results in the induction of at least 30 proteins and resistance to killing by higher doses of hydrogen peroxide. The expression of nine of the hydrogen peroxideinducible proteins, including catalase, an alkyl hydroperoxide reductase, and glutathione reductase, is controlled by the positive regulator oxyR. Strains carrying deletions of the oxyR gene are hypersensitive to hydrogen peroxide and are uninducible for the nine hydrogen peroxide-inducible proteins. The oxyR deletion strains also have significantly higher rates of spontaneous mutagenesis during aerobic growth that are probably due to oxidative damage. OxyR is homologous to the LysR family of bacterial regulatory proteins. As with other members of the LysR family, OxyR is an activator of a regulon of genes and a repressor of its own expression. The mechanism by which the OxyR protein regulates gene expression has been characterized through in vitro assays. Oxidized but not reduced OxyR is able to activate transcription, and although both forms of OxyR bind to DNA, the two forms interact differently with the promoter sequences. These I B. Demple, Annu. Rev. Genet. 25, 315 (1991). 2 S. B. Farr and T. Kogoma, Microbiol. Rev. 55, 561 (1991).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproduction in any formreserved.
[17]
OxyR REGULON
217
element (ARE) with an ARE/chloramphenicol acetyltransferase fusion construct43; and study of human immunodeficiency virus type 1 (HIV-1) activation by oxidant-induced NF-KB by gel shift analysis. 44 For an overview of gene regulation and oxidative stress, especially in bacteria where a greater gene response to oxidative stress has been observed, the reader is referred to Farr and Kogoma. ~2 43 T. H. Rushmore, M. R. Morton, and C. B. Pickett, J. Biol. Chem. 266, 11632 (1991). 44 R. Schreck, P. Rieber, and P. A. Baeuerle, EMBO J. 10, 2247 (1991).
[ 17] O x y R R e g u l o n
By GISELA STORZ and SHOSHY ALTUVIA Introduction
Escherichia coli and Salmonella typhimurium possess several different inducible responses to defend against the deleterious effects of reactive oxygen species) ,2 The oxyR-controlled regulon of hydrogen peroxideinducible genes in S. typhimurium and E. coli has provided a useful model for studying how bacterial cells defend against oxidative stress. Treatment of bacterial cells with low doses of hydrogen peroxide results in the induction of at least 30 proteins and resistance to killing by higher doses of hydrogen peroxide. The expression of nine of the hydrogen peroxideinducible proteins, including catalase, an alkyl hydroperoxide reductase, and glutathione reductase, is controlled by the positive regulator oxyR. Strains carrying deletions of the oxyR gene are hypersensitive to hydrogen peroxide and are uninducible for the nine hydrogen peroxide-inducible proteins. The oxyR deletion strains also have significantly higher rates of spontaneous mutagenesis during aerobic growth that are probably due to oxidative damage. OxyR is homologous to the LysR family of bacterial regulatory proteins. As with other members of the LysR family, OxyR is an activator of a regulon of genes and a repressor of its own expression. The mechanism by which the OxyR protein regulates gene expression has been characterized through in vitro assays. Oxidized but not reduced OxyR is able to activate transcription, and although both forms of OxyR bind to DNA, the two forms interact differently with the promoter sequences. These I B. Demple, Annu. Rev. Genet. 25, 315 (1991). 2 S. B. Farr and T. Kogoma, Microbiol. Rev. 55, 561 (1991).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproduction in any formreserved.
218
ASSAY OF STRESS GENES/PROTEINS
[17]
properties suggest that oxidation of the OxyR protein brings about a conformational change which leads to RNA polymerase activation and that OxyR is the direct sensor and transducer of an oxidative stress signal. In this chapter, we outline several of the assays that have been used to characterize the OxyR protein. These approaches should be applicable to OxyR homologs found in other organisms as well as to the characterization of additional regulators of oxidative stress responses. The assays described here should complement a chapter in this series focusing on the genetic approaches used to characterize the bacterial responses to oxidative stress. 3
Methods R N A Isolation
The analysis of RNA isolated from cells treated with different conditions of oxidative stress or from mutant cells can be extremely informative. We find that extraction of cells with acid phenol is the simplest method for obtaining RNA with minimal degradation of the notoriously shortlived bacterial RNA. 4 Cells (I0 ml) are collected by centrifugation and then resuspended in 387/xl of TKM (I0 mM Tris-Cl, pH 7.5, I0 mM KCI, and 5 mM MgC12). After the addition of 27/xl of 9 mg/ml lysozyme in TKM, the cells are frozen at - 6 5 °. (The samples can be stored at this point.) Immediately on thawing, 57 /A of 10% sodium dodecyl sulfate (SDS), 30/xl of 3 M sodium acetate, and 887/.d of acid phenol (prepared by dissolving 100 g solid acid phenol in I00 ml of water with 0.5 ml of 5 M NaC1 at 65 °) are added to the sample. The mixture is vortexed, then heated at 65 ° for 4 min and vortexed again. The layers are separated by centrifugation, and the top layer is reextracted with 887 ~1 of phenol with a second round of heating and vortexing. After chloroform extraction of the upper aqueous layer, the RNA is precipitated by the addition of sodium acetate (to a final concentration of 0.3 M) and ethanol. The resulting pellet can be washed with 70% ethanol and reprecipitated. To prevent the introduction of RNases, disposable tubes and pipettes are used, and all solutions are treated with diethyl pyrocarbonate (0.5 ml per liter for 1 hr in a hood) and then autoclaved. 3 G. Storz and M. B. Toledano, this series, Vol. 236, p. 196 (1994). 4 W. Salser, R. F. Gesteland, and A. Bolle, Nature (London) 215, 588 (1967).
[17]
OxyR REGULON
219
Primer Extension
A primer extension assay is one of the simplest methods for determining the 5' start of a transcript and for comparing the amounts of RNA present under different conditions. The following protocol is derived from McKnight and Kingsbury.5 First, a 17- to 30-base oligonucleotide is 5'-endlabeled with polynucleotide kinase and [32p]ATP and then extracted with phenol and chloroform and precipitated with ethanol. The labeled primer [250-500 counts/sec (cps)] and 5-50 /zg of total RNA (the amount of RNA assayed depends on the anticipated abundance) are reprecipitated together, and the pellet is washed with 70% ethanol. The pellet is then resuspended in 16/zl of water and 4/~1 of 5 × buffer (1.25 M KC1, 10 mM Tris-Cl, pH 7.9, 1 mM EDTA, pH 8). Again, the use of diethyl pyrocarbonate-treated solutions prevents RNA degradation. The sample is heated at 60° for 5 min, cooled slowly to 45 ° (over the course of I hr), and then cooled to room temperature (for an additional 1 hr) to anneal the primer to the RNA template. After annealing, a 50-/A reaction mixture containing 10 mM MgCI2,5 mM dithiothreitol (DTT), 100/zg/ml actinomycin D (a lyophilized aliquot of a 2 mg/ml stock in 100% ethanol), 0.33 mM deoxynucleoside triphosphates (dNTPs), 20 mM Tris-Cl, pH 8.7, and 10 units of reverse transcriptase is added to each sample. The reactions are incubated at 37° for 60 min, and then the extension products are precipitated by the addition of 600/zl of ethanol. The resulting pellets are washed with 70% (v/v) ethanol, resuspended in 6/zl of formamide dye and 3/zl of 0.1 M NaOH, heated at 85° to 95 ° for 2 to 3 min, and loaded on a sequencing gel. The addition of NaOH (to hydrolyze the RNA) is important if large amounts of RNA are assayed. Sequencing reactions primed with the same labeled primer can be run alongside the primer extension reactions to determine the exact start of the RNA species. We observe substantial induction of OxyR-regulated genes when RNA samples are obtained from logarithmically growing cells treated with 60 p~M hydrogen peroxide. We also employ the following simplified protocol, especially for abundant RNA species. 6 The RNA (3-15 ~g) is mixed with the labeled primer (250-500 cps) in water in a total volume of 9/zl. The sample is heated at 65 ° for 5 min and then cooled on ice for 10 min. After the annealing, 3/zl of 5 × buffer (0.25 M Tris-C1, pH 8.3, 0.375 M KCI, 15 mM MgC12), 1.5 /zl of 0.1 M DTT, 0.5/zl of 25 mM dNTPs, and 0.5/zl of water are added and then warmed at 37 ° for 5 min. Reverse transcriptase is added (0.5/zl 5 S. L. McKnight and R. Kingsbury, Science 217, 316 (1982). 6 S. Altuvia, D. Kornitzer, D. Teff, and A. B. Oppenheim, J. Mol. Biol. 210, 265 (1989).
220
ASSAY OF STRESS GENES/PROTEINS
[171
of 200 U//zl), and the sample is incubated at 37° for 30 min. The reaction is stopped by the addition of 7.5/xl formamide dye and then heated as above.
OxyR Purification Three different procedures have been used to purify the OxyR protein. 7-9 In all cases, the protein was first overproduced on cloning into an overexpression vector. We use the following method to purify OxyR from the soluble fraction of E. coli cells. 7 The cells (10 g) are resuspended in 30 ml of buffer Z (50 mM HEPES, pH 8.0, 0.5 mM EDTA, pH 8.0, 10 mM MgCI2) containing l0 mM KCI. Cells are lysed by three passages through a French press. Unlysed cells are then removed by a low-speed spin. To remove excess DNA, the cell lysate can then be treated with 200/xg DNase I at 4 ° for 30 rain in the presence of 0.1 to 1.0 mM phenylmethylsulfonyl fluoride (PMSF is a protein inhibitor and can be stored as a 0.1 M stock in ethanol). The soluble fraction is cleared of the membrane fraction twice by centrifugation at 9000 rpm. The volume of the supernatant is then adjusted to 50 ml with buffer Z and 5 ml of 100% glycerol. The lysate can be stored frozen at - 7 0 °. The first step in the purification of OxyR involves a heparin Sepharose column (150 ml). The supernatant is loaded at a flow rate of 1-3 ml/min. The column is then washed with 500 ml of buffer Z with 10% (v/v) glycerol and 0.1 M KCI followed by 500 ml of buffer Z with 10% glycerol and 0.2 M KCI. The OxyR protein is eluted with a 700-ml linear gradient of buffer Z with 10% glycerol and 0.2 to 0.5 M KC1. Fractions (10 ml) are collected throughout the gradient. The individual heparin fractions carrying OxyR can further be purified on a Mono S column (1 ml, Pharmacia, Piscataway, NJ). First, the KCI concentration of the heparin fraction is adjusted to 0. I M by dilution with 30 ml of buffer Z with 10% glycerol. The sample is then loaded at a flow rate of 0.2-0.5 ml/min. The column is washed with greater than 10 ml of buffer Z with 10% glycerol and 0.1 M KCI, and OxyR is eluted in 1-ml fractions with a 20-ml linear gradient of buffer Z with 10% glycerol and 0.1 to 1.0 M KCI. We find that the peak fractions of OxyR are greater than 95% pure at this point. OxyR can also be purified from the membrane pellet separated from the above supernatant. The pellet is resuspended in I00 ml of buffer Z
7 G. Storz, L. A. Tartaglia, and B. N. Ames, Science 248, 189 (1990). 8 M. B61ker and R. Kahmann, EMBO J. 8, 2403 (1989). 9 K. Tao, K. Makino, S. Yonei, A. Kakata, and H. Shinagawa, J. Biochem. (Tokyo) 109, 262 (1991).
[17]
OxyR REaULON
221
with 10% glycerol and 1 M KC1 and stirred for 30-45 rain at 4 °. The supernatant is cleared by centrifugation at 18,000 g for 30 to 45 min and then diluted with 900 ml of buffer Z with 10% glycerol. The solubilized protein can be further purified as above by loading on a heparin Sepharose column possibly followed by a Mono S column. The OxyR protein retains activity if stored at - 7 0 ° in buffer Z with 0.3 to 0.4 M KCI and 10% glycerol. OxyR is not very soluble in the absence of salt and is more soluble at pH 8.0 than at pH 7.6 or pH 6.8. The protein also appears to be stabilized by the glycerol. We have found that OxyR purified as described above is in the oxidized form and is capable of activating transcription. The protein can be converted to the reduced form by the addition of ->100 mM DTT.
DNA Mobility Shift The ability of a protein to bind to DNA can most simply be examined by gel retardation assays. Numerous gel conditions for the retardation assays have been described; however, we have found that low ionic strength gel conditions give significantly better shifts with the OxyR protein. The protein sample, either purified OxyR protein (0.2 to 5 ng) or a soluble extract from cells lysed by sonication ( - 1 ~g), is diluted in 1 x TM [50 mM Tris-C1, pH 7.9, 12.5 mM MgC12,20% glycerol, 1 mM EDTA, pH 8, 0.1% Nonidet P-40 (NP-40), and 100 mM KCI]. The labeled DNA sample (5000-10,000 cpm, 1-5 fmol), either plasmid restriction fragments or complementary 5'-end-labeled oligonucleotides annealed in 0.1 MNaC1 at 65 °, is diluted in water. Then the protein and the binding sites are mixed to give a final concentration of 0.5 × TM and a final volume of 25/zl. The binding reaction is incubated at room temperature for 10 to 15 min and subsequently loaded directly onto a low ionic strength gel prepared exactly as described) ° Nonspecific poly(dI-dC) (0.1 to 0.5 ~g) can be added if less pure samples of OxyR are assayed. To assay the ability of OxyR to bind a DNA sequence under reducing conditions we add DTT (---100 mM final concentration) to the binding reactions. The state of OxyR binding may change once the binding reaction has been loaded on the polyacrylamide gel; however, DTT can also be added to the running buffer. (DTT cannot be added to the gel since it prevents polymerization.)
~0F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl, eds., "Current Protocols in Molecular Biology." Wiley (Interscience), New York, 1987.
222
ASSAY OF STRESS GENES/PROTEINS
[17]
DNase I Footprinting DNase I footprinting assays can be used to map the site of OxyR binding precisely. For the footprinting reactions, a DNA fragment that is labeled only on one end is required. All of the in vivo OxyR binding sites described to date map at or near the - 35 promoter consensus sequence, and we generally label a restriction fragment of 300 to 400 base pairs (bp) carrying the promoter as follows. DNA (20/zg) is digested with a restriction enzyme giving a 5' overhang (the restriction enzymes EcoRI, BamHI, and HindlII are ideal) between 100 and 200 bp from the putative binding site. The restriction enzymes are removed by phenol and chloroform extractions and ethanol precipitation. Phosphatase is added to remove the 5'-phosphate groups and then removed by thorough phenol and chloroform extractions. The 5' ends are then labeled by treating 5/~g of the cut and phosphatase-treated DNA (in 20/zl) with kinase in the presence of 200-300 /zCi [32p]ATP. The volume of the reaction is then adjusted to 50/xl, and the DNA is digested with a restriction enzyme that cuts 100 to 200 bp from the other side of the putative binding site. To purify the fragment, we load the entire labeled and digested sample on a 6%, 0.8-mm-thick acrylamide gel and cut out the fragment of interest after a 1- to 2-rain exposure to film. The fragment is eluted from the gel slice in 200 to 400 tzl TE at 37° for greater than 8 hr. After elution, the fragment is extracted with phenol and chloroform and precipitated. The DNase I required for the footprinting reactions is obtained as a lyophilized powder and then dissolved in cold water at a concentration of 2.5 mg/ml. Aliquots (10/xl) of the stock solution can be frozen in liquid nitrogen and stored at - 70°. The binding reactions are carried out almost exactly as for the mobility shift assay. Less than 30 ng of OxyR protein in I × TM is mixed with 2000 to 6000 cpm (3 fmol) 5'-end-labeled DNA in water to give a total volume of 25 /zl at 0.5 z TM and then incubated for I0 min at room temperature. Again DTT (-> 100 mM) can be added to examine the binding under reducing conditions. During the 10-min incubation, the DNase I stock is diluted to 1 to 3/xg/ml in cold water (use lower concentrations for larger fragments to obtain a more complete set of partially digested fragments). After the 10-min incubation, 25 ~1 of a Mg 2+,Ca2+-containing solution (10 mM MgCI2 and 5 mM CaCI2) is added to all tubes. Immediately thereafter, 2/xl diluted DNase I is added to each tube at room temperature at 10- to 15-sec intervals. Then each DNase I digestion is stopped after exactly 1 min by adding 200/~1 of stop solution (20 mM EDTA, pH 8.0, 1% SDS, 0.2 M NaCI, and 250/~g/ml total yeast RNA which is purified by abundant phenol and chloroform extractions). The digested fragments are subsequently extracted with phenol and chloroform and precipitated
[17]
OxyR REGULON
223
with 600 txl of ethanol (no additional salt is needed for the precipitation), The pellet can be washed with 70% ethanol, dried, resuspended in 2-4 tzl sequencing formamide dye mix, heated at 90° to I00 ° for 2 to 3 min, and then loaded on an 8% sequencing gel. A sample corresponding to the G and A sequence can be generated as follows. Five micrograms of yeast total RNA and 80,000 cpm of the labeled DNA fragment are dried, resuspended in 6/~1 of fresh 2% (v/v) formic acid (J. T. Baker Chemical Co., Phillipsburg, NJ), and incubated at 37° for 5 min (use shorter times if too much cleavage is observed). The formic acid is removed by evaporation for 2 hr in a Speed Vac concentrator. The dried pellet is resuspended in 100/xl of 1 M piperidine (Fisher Scientific, Fair Lawn, N J) and heated at 90 ° for 40 min. The piperidine is removed by evaporation overnight. Then the dried DNA is resuspended in 50 /xl of water and the water removed by evaporation. Finally the cleaved DNA is resuspended in 100 /xl of water, extracted twice with chloroform, and then precipitated on the addition of salt and ethanol. An amount equivalent to only one-third of the radioactivity in the footprinting lanes should be loaded. In Vitro Transcription
The ability of OxyR to activate transcription can be assayed in vitro as follows. Purified OxyR (-0.5 mg/ml) with 15 /zg of bovine serum albumin in a total volume of 50 tzl is exchanged into transcription buffer (40 mM Tris-Cl, pH 7.9, 0.1 M KCI, 10 mM MgCI2) containing 1 mM DTT, 5% glycerol, and 0.1% NP-40 by centrifugation through 800 ~1 of Sephadex. An aliquot (5/zl) together with 2.5 /xl water is the incubated with a supercoiled template DNA (0.2/zg in 36.5/zl of transcription buffer) for 10 min at 37°. RNA polymerase holoenzyme (0.5/zg in 5/zl of transcription buffer) is then added to the bound template, and the reaction is incubated for an additional 10 min at 37 °. After the addition of 1 /~1 of a 25 mM mixture of dNTPs, the reactions are incubated another 5 min at 37°. The transcription reactions are terminated by the addition of phenol, and the mRNA is extracted with phenol and chloroform several times and then assayed by primer extension.
224
ASSAY OF STRESS G E N E S / P R O T E I N S
[18]
[18] T r a n s i e n t E n h a n c e m e n t o f H e m e O x y g e n a s e 1 m R N A A c c u m u l a t i o n : A M a r k e r o f O x i d a t i v e S t r e s s to E u k a r y o t i c Cells By REX M. TYRRELL and SHARMILA BASU-MODAK Introduction
Bacteria respond to oxidative stress by the rapid and transient expression of a large number of genes. Two major regulatory pathways have been recognized to date, namely, the oxyR and soxR systems. The oxyR system regulates nine of the genes induced by hydrogen peroxide in both Escherichia coli and Salmonella typhimurium. 1 Several of the oxyR-controlled genes (e.g., catalase, glutathione reductase, alkyl hydroperoxide reductase) clearly play a role in defense against oxidative stress. OxyR codes for a regulatory protein which appears to act as an oxygen sensor by being oxidized directly. 2 Another 40 or so genes are turned on by redox cycling agents in E. coli, and 9 of these have been shown to have a common regulatory pathway and have been grouped together as the soxR regulon. 3 The detection of expression of the oxyR regulon and/or the soxR regulon should provide an early marker of oxidative stress in prokaryotes. The binding of several eukaryotic transcription factors such as NFKB and possibly the fos-jun heterodimer (AP-1) may be redox regulated. 4,5 Expression of the c-jun oncogene is stimulated by either UVC radiation (254 nm) or hydrogen peroxide. 6 On the other hand, expression of the human heme oxygenase 1 (HO-1) gene is not induced by UVC radiation, whereas it is strongly induced by oxidizing agents such as UVA (320-380 nm) radiation or hydrogen peroxide. 7 In addition to oxidative stress, the gene is induced by other agents such as phorbol esters, heavy metals, and sodium arsenite. 7 Although this phenomenon was originally observed in fibroblasts cultured from human skin, induction occurs in most human l L. A. Tartaglia, G. Storz, S. B. Farr, and B. N. Ames, in "Oxidative Stress: Oxidants and Antioxidants" (H. Sies, ed.), p. 155. Academic Press, London, 1991. 2 G. Storz, L. A. Tartaglia, and B. N. Ames, Science 248, 189 (1990). 3 B. Demple and J. D. Levin, in "Oxidative Stress: Oxidants and Antioxidants" (H. Sies, ed.), p. ll9. Academic Press, London, 1991. 4 C. Abate, L. Patel, F. J. Rauscher III, and T. Curran, Science 249, 1157 (1990). R. Schreck, P. Rieber, and P. A. Baeuede EMBO J. 10, 2247 (1991). 6 y . Devary, R. A. Gottlieb, L. F. Lau, and M. Karin, Mol. Cell. Biol. 11, 2804 (1991). 7 S. M. Keyse and R. M. Tyrrell, Proc. Natl. Acad. Sci. U.S.A. 86, 99 (1989).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[18]
INDUCTION OF HO-1 AS OXIDATIVE STRESS MARKER
225
cell types and all mammalian cell types so far tested. 8 The induction is clearly related to the cellular redox state since lowering cellular glutathione levels strongly enhances HO-1 mRNA accumulation) For these reasons, altered expression of the HO-1 gene appears to be a fairly sensitive marker of oxidative stress. Indeed, enhanced expression of the HO-1 gene is now being used in several laboratories as a positive control when testing other genes suspected of being oxidant-inducible. Choice of Assay
Because oxidants induce expression of the HO-1 gene by enhancing the transcription rate, 1° altered expression can be monitored at a variety of levels. Direct measurement of altered transcription rates is probably the most sensitive method, but the run-off transcription assays usually employed for this measurement are more labor-intensive and subject to greater interexperimental variation than assays that measure a later step. One-dimensional sodium dodecyl sulfate (SDS)-poylacrylamide gels are normally sensitive enough to detect induction of de novo synthesis of the 32-kDa protein corresponding to HO-l,n but background levels are high due to the large number of constitutive proteins in this molecular size range. Induction of HO-1 enzyme activity several hours after the initial treatment is also fairly simple to measure by a spectrophotometric assay,~Z although the biliverdin reductase required for the coupled assay is not available commercially. However, a resolution problem now arises because HO-1 cannot be distinguished enzymatically from the constitutive HO-2 form which is present in various amounts according to cell t y p e ) TM The two forms can be distinguished by Western blot analysis, but antibodies to the proteins are not yet commercially available. With these considerations in mind, the current method of choice is measurement of the accumulation of HO-1 mRNA using a specific cDNA probe. The most commonly used techniques for measuring such accumulation are the dot-blot and Northern blot procedures, the latter being preferred given the high level of nonspecific background that can be associated with the dot-blot procedure. RNA extraction and Northern blot methods 8 L. A. Applegate, P. Luscher, and R. M. Tyrrell, Cancer Res. 51, 974 (1991). 9 D. Lautier, P. Luscher, and R. M. Tyrrell, Carcinogenesis (London) 13, 227 (1992). l0 S. M. Keyse, L. A. Applegate, Y. Tromvoukis, and R. M. Tyrrell, Mol. Cell. Biol. 10, 4967 (1990). ii S. M. Keyse and R. M. Tyrrell, J. Biol. Chem. 262, 14821 (1987). 12 S. Shibahara, T. Yoshida, and G. Kikuchi, Arch. Biochem. Biophys. 188, 243 (1978). 13 M. D. Maines, G. M. Trakshel, and R. K. Kutty, J. Biol. Chem. 261, 411 (1986). 14 G. M. Trakshel, R. K. Kutty, and M. D. Maines, J. Biol. Chem. 261, 11131 (1986).
226
ASSAY OF STRESS GENES/PROTEINS
[18]
are commonly described in laboratory manuals for molecular biology, 15,16 but we describe here the sequence of procedures that works best for us for our work with HO-1 mRNA. The precise conditions of cell culture and preparation before and after treatment will depend on the cell type employed. We describe the procedure for human fibroblasts cultured as attached monolayers and treated with either UVA radiation or hydrogen peroxide.
Methods Treatment Procedure Irradiation with UVA. A broad-spectrum UVA lamp with a large beam area (e.g., we use the UVASUN 3000 lamp supplied by Mutzhas, Munich Germany) is a convenient source of UVA radiation for the processing of many culture dishes simultaneously.
1. Human fibroblast cells (FEK 417) are seeded in 10-cm dishes at a density of 5 × 105 cells in 10 ml Earle's minimal essential medium [supplemented with 50 U/ml penicillin, 50/zg/ml streptomycin, 0.2% (w/v) sodium bicarbonate, and 15% fetal calf serum (FCS, v/v) (heat-inactivated at 56°)] per dish and cultured for 3 days at 37° in 5% CO2, at which time they reach 60-80% confluency. 2. Remove the conditioned medium and keep aside at 37°. Rinse the cell monolayers with 10-15 ml phosphate-buffered saline (PBS) at room temperature. 3. Add 5 ml of PBS to each 10-cm dish. The PBS should be supplemented with CaZ'-/Mg2+ salts (each 0.01% final concentration)just prior to use. 4. Irradiate the culture dishes for the time required at a distance from the UVA source that avoids a temperature rise during irradiation. A fluence range of 0-1 MJ m -2 is generally used for the HO-1 gene, and we find that the HO-1 mRNA levels peak at fluences between 400 and 500 kJ m -2. The fluence rate at 30 cm from the UVASUN 3000 lamp source is approximately 300 W m-2 as measured by an IL 1700 radiometer (International Light Inc., Newburyport, MA). 15 j. Sambrook, E. F. Fritsch, and T. Maniatis, "Molecular Cloning: A Laboratory Manual," 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1989. ~6 L. G. Davis, M. D. Dibner, and J. F. Battey, "Basic Methods in Molecular Biology." Elsevier, New York, 1986. 17 R. M. Tyrrell and M. Pidoux, Cancer Res. 46, 2665 (1986).
[18]
INDUCTION OF HO-1 AS OXIDATIVE STRESS MARKER
227
5. After irradiation, aspirate the buffer and rinse the monolayer with 10-15 ml of PBS. Add back the original medium and incubate the cells for 3 hr before extracting total RNA, Addition of fresh medium instead of the conditioned medium induces the gene and gives incorrect estimates of the basal levels of the HO-1 mRNA.
Treatment with Oxidant 1. Cells are grown to 60-80% confluency and prepared for chemical treatment as described above for irradiation with UVA. 2. Add 5 ml of Ca2+/Mg2+-containing PBS to each culture dish and then hydrogen peroxide to a final concentration of 100 /xM. A dilute solution of hydrogen peroxide is prepared in sterile water just prior to use. At 240 nm the molar extinction coefficient of hydrogen peroxide is 43.6 M 1cm-l. 18The appropriate (i.e., nontoxic but effective) concentration of hydrogen perioxide is highly dependent on the cell number and needs to be determined for each given set of experimental conditions. 3. Incubate the cell monolayers with the oxidant for 30 min at 37° in the CO2 incubator. After chemical treatment, aspirate the buffer containing the oxidant, rinse the monolayers with PBS, and add back the original medium. Incubate cultures for 3 hr at 37° before extracting total RNA.
Isolation of Total Cellular RNA We use the acid-guanidinum thiocyanate-phenol-chloroform (AGPC) extraction method ~9 for isolation of total cellular RNA as it is rapid and a large number of samples can be processed simultaneously. All solutions (except Tris and EDTA) are prepared in water treated with DEPC (diethyl pyrocarbonate). 15 RNase-free glassware and plasticware is used for all manipulations. Use of gloves during all procedures is obligatory for RNA work.
Solutions Guanidinium thiocyanate stock: 4 M guanidinium thiocyanate (250 g plus 293 ml of water), 25 mM sodium citrate (pH 7.0) (17.6 ml of 0.75 M stock), 0.5% sarkosyl (26.4 ml of 10% stock); guanidinium thiocyanate is dissolved at 65 ° directly in the manufacturer's bottle; the stock solution is stable at room temperature for 3 months ~s A. Claiborne, in "CRC Handbook of Methods for Oxygen Radical Research" (R. A. Greenwald, ed.), p. 283. CRC Press, Boca Raton, Florida, 1985. 19 p. Chomczynski and N. Sacchi, Anal, Biochem. 162, 156 (1987).
228
ASSAY OF STRESS GENES/PROTEINS
[18]
Solution D: 0.36 ml of 2-mercaptoethanol in 50 ml of guanidinium thiocyanate stock; the solution is stable at room temperature for 1 month Phenol equilibrated with Tris-Cl (pH 8.0) 15 Chloroform-isoamyl alcohol mixture (49:1, v/v) 75% Ethanol (v/v) 2 M Sodium acetate (pH 4.0)
Procedure 1. Rinse the cell monolayers with 10-15 ml PBS and add 2 ml solution D per 10-cm dish. Cells are lysed directly in the culture dishes for 5 min at room temperature, after which the dishes are left in an inclined position for 2 min to allow the viscous lysate to accumulate on one side of the dish. This allows the collection of 95% of the lysate and is especially useful when a large number of dishes are processed simultaneously. 2. Transfer the lysate to polypropylene tubes with caps and add 0.2 ml of 2 M sodium acetate (pH 4.0), 2.5 ml phenol, and 0.4 ml chloroform-isoamyl mixture sequentially. Mix well by rapidly inverting the tube for at least 30 sec after each addition. The mixing steps should be thorough but gentle. 3. Cool suspensions on ice for 15 min and centrifuge at 10,000 g for 20 min at 4 ° to separate the phases. 4. Collect the aqueous phase in a fresh tube and precipitate the RNA with an equal volume of 2-propanol at - 2 0 ° for at least 1 hr. When a large number of samples are being processed, it may be more convenient to leave the samples overnight at - 2 0 ° at this step. 5. Pellet the RNA by centrifuging at 10,000 g for 30 min and redissolve (room temperature) the pellet in 0.3 ml of solution D. At this step, samples are transferred to 1.5-ml Eppendorf tubes and reprecipitated with an equal volume of 2-propanol (0.3 ml) at - 2 0 ° for I hr. 6. Pellet the RNA by centrifuging in a microcentrifuge at 4 °. Wash the RNA pellet twice with 75% ethanol as follows. Add 500/xl of 75% ethanol to each tube and release the RNA pellet by tapping the tube gently against the laboratory bench. Microcentrifuge for I0 min at 4 ° and then aspirate the supernatant. 7. Vacuum-dry the pellet for 5 min. If RNA is dried for too long, then it does not go into solution easily. 8. Dissolve the RNA in 25 /~1 of DEPC-treated water by heating to 65 ° for 10 min followed by quick cooling on ice. Determine the RNA concentration by measuring the absorbance of an aliquot at 260 and 280 nm. The A26o/A2so ratios obtained should be between 1.95 and 2.0. Store aliquots containing 12-15/~g of total RNA at - 2 0 ° until further use.
[18]
INDUCTION OF H O - 1 AS OXIDATIVE STRESS MARKER
229
Northern Analysis of HO-1 mRNA Electrophoresis of RNA. The gel casting trays, combs, and buffer tanks should be used routinely only for RNA work and should be soaked in 1% SDS (w/v) and rinsed well with ultrapure deionized water prior to each use. If RNase is also used in the laboratory, then extra attention should be given to cleaning the gel electrophoresis accessories. 15 We use formaldehyde/agarose gels for separation of RNA. Solutions 10× MOPS [3-(N-morpholino)propanesulfonic acid] buffer (final pH between 5.5 and 7.0)*: 0.2 M MOPS [3-(N-morpholino)propanesulfonic acid], 50 mM sodium acetate, 10 mM EDTA Loading buffer (10 ml): 4.8 ml deionized formamide, 1.07 ml of 10× MOPS buffer, 1.73 ml (37%) formaldehyde, 0.533 ml glycerol, 1.033 ml bromphenol blue (saturated solution), 0.834 ml water Rinse solution*: 75 mM NaOH, 100 mM NaC1 100 mM Tris-Cl (pH 7.5)
Procedure 1. Melt 1.3 g agarose (Bio-Rad, Richmond, CA) in 74 ml of sterile water and add 10 ml of 10× MOPS buffer. Cool to 50° and add 16.2 ml of 37% formaldehyde solution (2.2 M final), then pour the contents into a 10 by 15 cm gel casting tray with the comb in position. Formaldehyde gels are cast in a fume hood and allowed to set for 30-45 min. We do not use ethidium bromide in gels which are subjected to Northern transfer. 2. Remove the comb gently and cover the gel with 1 × MOPS buffer. 3. To prepare RNA samples for electrophoresis, vacuum-dry aliquots containing 12-15/~g of total RNA to decrease the volume to 5-10/~1. Do not dry completely. Add 20/~1 of loading buffer to each sample and heat to 65° for 10 min followed by quick cooling on ice. 4. Load samples and electrophorese at 50 V until the bromphenol blue migrates a distance of approximately 7 cm from the well. 5. After electrophoresis, rinse the gel with deionized water and soak in the rinse solution for 40 min at room temperature on a rocking platform. 6. Neutralize the gel with 100 mM Tris-Cl (pH 7.5) for 45-60 min on a rocking platform with one change of buffer and set up the Northern transfer as described below. * Sterile ultrapure deionizedwater can be used to make these solutions instead of DEPCtreated water.
230
ASSAY OF STRESS GENES/PROTEINS
[18]
It is usually not necessary to electrophorese samples in ethidium bromide-containing gels since RNA prepared by the AGPC method is usually undegraded, provided that proper care has been taken during isolation and electrophoresis. However, if RNA isolation is carried out for the first time, an aliquot of 2/zg of total RNA can be electrophoresed in 1% agarose gels containing ethidium bromide using 0.5 × TBE bufferJ 5 Northern Transfer: Capillary Blot Procedure. We use Gene Screen nylon membranes (NEN Du Pont, Dreieich, Germany) for Northern transfer and use the manufacturer's conditions for transfer and hybridization. These are described briefly here.
Solutions Transfer buffer (pH 6.5)*: 25 mM Na2HPO4, 25 mM NaH2PO4
Procedure 1. Place a sponge (small pore, 2 × 15 × 19 cm) in a RNase-free plastic container and add enough transfer buffer to soak the sponge. D o n o t submerge the sponge in buffer. Air bubbles can be easily removed by poking the sponge with an RNase-free pipette or glass rod. 2. Place a Whatman (Clifton, N J) paper 3MM (prewet in transfer buffer), cut to a size intermediate between the gel size and the sponge size, on the sponge. Remove entrapped air bubbles by rolling a pipette or glass rod over the paper. Overlay with two more wet pieces of Whatman 3MM paper. Care should be taken to remove air bubbles at each overlay step. 3. Place the gel with the well bottom facing away from the sponge and overlay with a wet piece of Gene Screen membrane cut to the same size as the gel. The nylon membrane should be wetted in transfer buffer for at least 15 min prior to use. 4. Overlay the nylon membrane with one piece of wet and two pieces of dry Whatman 3MM paper cut to the same size as the gel. Place Parafilm strips along the edges of the gel to prevent short-circuiting during transfer. 5. Place a stack of paper towels (8-10 cm high) cut to the same size as the gel and a 1-kg lead weight on top of the stack and leave for at least 16 hr for transfer. 6. Rinse the membrane with transfer buffer to remove residual agarose, place the blot on a Whatman 3MM paper with the transferred RNA side facing upward, and air dry. Bake the blot at 80° for 2-4 hr. After this step the nylon membrane can be stored at room temperature until use.
Hybridization of RNA. We probe the Northern blot for the HO-1 mRNA with the large EcoRI fragment (1000 bp) of a full-length cDNA
[18]
INDUCTION OF H O - I AS OXIDATIVE STRESS MARKER
231
clone (clone 2/107). The full-length cDNA is inserted into the EcoRI sites of the phagemid pBluescript SK (MI3-), and we maintain it in the Stratagene (LaJolla, CA) E. coli host XL1-Blue {endA1, hsdR17 (rk-, mk+), supE44, thi-1, k-, recA1, gyrA96, relA1, lac- [F', proAB, laclqZ A M15, TnlO (TetR)]}. This plasmid DNA (15-20 /xg) is digested to completion with EcoRI and electrophoresed in a 0.6% low melt agarose gel. The 1000-bp fragment is cut out from the gel and used directly for labeling. This cDNA clone is available from our laboratory. To control for the variation between RNA samples in the same gel, we reprobe each membrane for GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA, 2° levels of which are not affected by UVA or other oxidants. Probes for other constitutive mRNA species can also be used as loading controls. The cDNA probes are labeled with [a-32p]dCTP or [ot-32p]dATP using a Random Primed labeling kit supplied by Boehringer Mannheim (Mannheim, Germany). We purify the 3Zp-labeled fragment on a Elutip-d column (Schleicher and Schuell, Dassel, Germany) according to the procedure recommended by the manufacturer.
Solutions Prehybridization buffer (recommended for Gene Screen nylon membranes by manufacturer): 50% formamide, v/v (deionized), 0.2% polyvinylpyrrolidone, 0.2% bovine serum albumin (BSA), 0.2% Ficoll, 50 mM Tris-HCl (pH 7.5), 1.0 M NaC1, 0.1% sodium pyrophosphate, 1.0% SDS (w/v), 10% dextran sulfate (optional), 100 /zg/ml denatured salmon sperm DNA 20 x SSC*: 3 M sodium chloride, 0.3 M sodium citrate Washing solution 1": 2× SSC, 0.1% SDS Washing solution 2*: 0.1x SSC, 0.1% SDS Procedure 1. For prehybridization, prewet the blot in 6x SSC and place in a plastic bag. Remove as much liquid as possible. Add 12 ml ofprehybridization buffer at 42° to the bag, remove air bubbles, and seal the bag. Leave the bag in a 42 ° water bath with constant agitation for at least 4 hr (can be left for 24 hr). Two filters can be placed in each bag, with the sides containing the RNA facing away from one another. 2. For denaturing the probe, boil in a water bath for 10 rain and quick cool on ice for 10 rain. Add the denatured probe (specific activity 108 cpm/ mg) to the prehybridization bag (2-3 x 10 6 cpm/ml, 10-50 ng/ml), remove 20 p. Fort, L, Marty, M. Piechaczyk, S. E. Sabrouty, C. Dani, P. Jeanteur, and J. M. Blanchard, Nucleic Acids Res. 13, 1431 (1985).
232
ASSAY OF STRESS GENES/PROTEINS
[18]
air bubbles, seal the bag, and leave for hybridization at 42 ° for 16--24 hr. The probe purified on an Elutip-d column is in a volume of 500 ~1, and we add this directly to the prehybridization solution. 3. Collect the hybridization solution and store at - 2 0 °. The solution can be reused once; to do so, remove all buffer from the bag after prehybridization and add the hybridization solution which has been boiled in a water bath for 10 min. 4. After hybridization, wash the membrane with constant agitation as follows: (a) once in 100 ml of 2× SSC for 5 min at room temperature; (b) once in 100 ml of washing solution 1 for 30 min at 50°; and (c) once in 100 ml of washing solution 2 for 30 rain at 65 °. For the HO-I cDNA probe, the third washing step (c) is usually not necessary. 5. Air dry the membrane, wrap it in Saran wrap and expose to a preflashed film at - 7 0 °. 15Because the membranes are usually rehybridized with the GAPDH probe, they should be left slightly damp. 6. Prior to rehybridization, the labeled probe is stripped from the membrane by boiling in washing solution 2 for 40 min with one solution change. It is recommended that the labeled probe be stripped off from the membrane soon after autoradiography if rehybridization is to be carried out. The stripped Northern blot can be air dried and stored between sheets of Whatman 3MM paper at room temperature. A typical Northern blot analysis of total cellular RNA isolated by the AGPC method and probed for the HO-1 and GAPDH mRNA species is shown in Fig. 1. For this experiment cell cultures were exposed to a range of fluences of UVA radiation in the presence of 5 mM N-acetylcysteine. Quantification of Northern Analysis. Because there is a fluence-dependent change in HO-1 mRNA levels, it is necessary to carry out autoradiography for different periods of time. Autoradiographs in which the control sample (see Fig. 1) which has the lowest signal intensity is clearly visible and the band with the highest signal intensity is still in the linear range MJm N-AC
-2
0 -
0.1 +
0.2 +
-
0.3 +
0.4 +
-
0.5 +
-
+
HO
GAPDH
FIG. 1. Human skin fibroblasts (FEK 4) were irradiated with increasing Iluences of UVA in the presence (+) or absence ( - ) of 5 mM N-acetylcysteine (N-AC). Total RNA was subjected to Northern blotting and probed first for HO- 1 mRNA and then for GAPDH mRNA.
[18]
INDUCTION OF H O - 1 AS OXIDATIVE STRESS MARKER
233
should be used for densitometric analysis. A detailed densitometric analysis of autoradiographs with different exposure times needs to be done for one experiment, after which it is possible to judge the signal intensity visually. Radioactive signals are quantified in our laboratory by densitometry using an Elscript 400 (Hirschmann) densitometer with evaluation software. We usually scan vertically down each track rather than scanning the band of interest horizontally along the gel. The area under the curve of the band of interest is integrated with the evaluation software supplied with the densitometer. Areas determined by densitometry of autoradiographs in Fig. 1 are shown in Table I. Calculation o f Increase in HO m R N A The HO-1 mRNA signal intensity increases severalfold over basal levels in response to oxidative stress, and this response can be modulated by chemical agents such as N-acetylcysteine (see Fig. 1). The first step in the quantification is to normalize for the variation in loading between samples on the same gel using the GAPDH mRNA signal (or any other constitutive mRNA) on autoradiographs. A GAPDH ratio is calculated as follows: 1. Calculate the mean areas under the curves of all the GAPDH bands on a gel from the autoradiograph.
TABLE I AREAS DETERMINED BY DENSITOMETRY FOR HO-1 m R N A AND G A P D H m R N A
Lane number a 1 2 3 4 5 6 7 8 9 10 11 12
HO-1 area
GAPDH area
GAPDH ratio
GAPDH ratio x HO-1 area
3.21 8.01 8.86 6.19 15.57 13.23 53.22 30.13 109.17 55.04 93.16 73.72
33.89 26.42 23.82 30.63 18.03 19.41 20.52 23.51 48.72 33.52 33.65 43.21
0.87 1.12 1.24 0.97 1.64 1.53 1.44 1.26 0.61 0.88 0.88 0.69
2.79 8.97 10.99 6.00 25.53 20.24 76.64 37.96 66.59 48.44 81.98 50.87
a See Fig. 1; lanes are n u m b e r e d from left to right.
234
ASSAY OF STRESS GENES/PROTEINS
[18]
2. Divide the mean by the GAPDH area of each sample to obtain a GAPDH ratio for each lane. If equal amounts of total RNA were loaded in all tracks, then the signal intensity would be the same in all the tracks. This is usually not the case and thus the need for this normalization step. 3. Multiply the area under the curve of each HO-1 band by the corresponding GAPDH ratio. The numbers thus obtained are normalized for the variation between samples on the same gel. 4. The calculated area of the control sample (Table I, lane 1) is used as a unit to calculate the increase(-fold) in HO-1 mRNA above basal levels. A plot of relative increase in HO- 1 mRNA levels as a function of fluence is shown in Fig. 2A.
30 t "--
u) 25
"
t
°
15
OT 2.0
.
,
•
,
•
,
•
,
•
,
"
B 0 1.5
-,-~ 1.0 o~ o
0.0 0.0
0.1
0.2
Fluence
0.3
0.4
0.5
(MJm -2 )
FIG. 2. The HO-I and GAPDH mRNA signals were quantified by densitometry. The GAPDH signal was used as an internal control to normalize for the variation in loading between samples. (A) Normalized HO-1 mRNA signals are expressed as the relative increase above basal levels and plotted as a function of fluence. (©) Control cells, (0) cells treated with N-acetylcysteine. (B) The ratios of the normalized HO-1 mRNA signal areas in samples treated with 5 mM N-acetylcysteine (N-AC) and control samples are plotted as a function of fluence.
[19]
REGULATION OF GAP JUNCTIONAL COMMUNICATION
235
The increase(-fold) of HO-1 mRNA varies between experiments and thus error bars are large; however, the pattern is consistent for any agent. To quantify the effect of a test agent it is useful to calculate the ratios of the HO-1 mRNA areas (normalized) in untreated and treated samples of at least three experiments. Such ratios with their standard deviations are plotted against fluence in Fig. 2B. Transient enhancement of HO- 1 mRNA accumulation appears to be an extremely sensitive marker of oxidative stress, and it has been employed in assays in several laboratories. In this chapter we have described detailed methodology based on the Northern blot procedure for the comparative measurement of HO-1 mRNA levels and provide an example in which increased accumulation of HO- I mRNA after a mild oxidative stress (UVA radiation) is suppressed by a free radical scavenging antioxidant (N-acetylcysteine). Acknowledgments The studiesdescribedherein have been supportedby the SwissNational ScienceFoundation (FN 31-30880-91)and the Swiss League against Cancer. N-Acetylcysteinewas a kind gift of Imphazarrn S.A. (Cadempino, Switzerland).
[ 1 9] A s s a y s for R e g u l a t i o n o f G a p J u n c t i o n a l C o m m u n i c a t i o n and Connexin Expression by Carotenoids
By
JOHN S. BERTRAM and LI-XIN
ZHANG
Introduction Compelling epidemiologic evidence has shown that certain carotenoids have cancer chemopreventive activities.1 However, their mode of action in this respect is unclear. Carotenoids as a group are considered to possess antioxidant properties,2 and a limited number act as provitamin A sources in mammals.3 In view of the postulated role of oxidative damage in carcinogenesis 4'5 and the known activity ofretinoids as cancer chemopreventives,6 I j. S. Bertram, L. N. Kolonel, and F. L. M e y s k e n s , Cancer Res. 47, 3012 (1987). 2 N. I. Krinsky, Free Radical Biol. Med. 7, 617 (1989). 3 j. A. Olson, J. Nutr. 119, 105 (1989). 4 p. A. Cerutti, Science 227, 375 (1985). 5 L. H. Breimer, Mol. Carcinog. 3, 188 (1990). 6 R. C. Moon, J. Nutr. 119, 127 (1989).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994by AcademicPress, Inc. All rights of reproduction in any form reserved.
[19]
REGULATION OF GAP JUNCTIONAL COMMUNICATION
235
The increase(-fold) of HO-1 mRNA varies between experiments and thus error bars are large; however, the pattern is consistent for any agent. To quantify the effect of a test agent it is useful to calculate the ratios of the HO-1 mRNA areas (normalized) in untreated and treated samples of at least three experiments. Such ratios with their standard deviations are plotted against fluence in Fig. 2B. Transient enhancement of HO- 1 mRNA accumulation appears to be an extremely sensitive marker of oxidative stress, and it has been employed in assays in several laboratories. In this chapter we have described detailed methodology based on the Northern blot procedure for the comparative measurement of HO-1 mRNA levels and provide an example in which increased accumulation of HO- I mRNA after a mild oxidative stress (UVA radiation) is suppressed by a free radical scavenging antioxidant (N-acetylcysteine). Acknowledgments The studiesdescribedherein have been supportedby the SwissNational ScienceFoundation (FN 31-30880-91)and the Swiss League against Cancer. N-Acetylcysteinewas a kind gift of Imphazarrn S.A. (Cadempino, Switzerland).
[ 1 9] A s s a y s for R e g u l a t i o n o f G a p J u n c t i o n a l C o m m u n i c a t i o n and Connexin Expression by Carotenoids
By
JOHN S. BERTRAM and LI-XIN
ZHANG
Introduction Compelling epidemiologic evidence has shown that certain carotenoids have cancer chemopreventive activities.1 However, their mode of action in this respect is unclear. Carotenoids as a group are considered to possess antioxidant properties,2 and a limited number act as provitamin A sources in mammals.3 In view of the postulated role of oxidative damage in carcinogenesis 4'5 and the known activity ofretinoids as cancer chemopreventives,6 I j. S. Bertram, L. N. Kolonel, and F. L. M e y s k e n s , Cancer Res. 47, 3012 (1987). 2 N. I. Krinsky, Free Radical Biol. Med. 7, 617 (1989). 3 j. A. Olson, J. Nutr. 119, 105 (1989). 4 p. A. Cerutti, Science 227, 375 (1985). 5 L. H. Breimer, Mol. Carcinog. 3, 188 (1990). 6 R. C. Moon, J. Nutr. 119, 127 (1989).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994by AcademicPress, Inc. All rights of reproduction in any form reserved.
236
ASSAY OF STRESS GENES/PROTEINS
[19]
both activities could contribute to the observed epidemiologic findings. In studies conducted with a highly characterized system of in vitro carcinogenesis, 7 we have demonstrated a novel carotenoid action: that of upregulation of gap junctional communication (GJC). 8 This action is highly correlated with the chemopreventive properties but not with the provitamin A status or lipid-phase antioxidant potency of a series of dietary carotenoids. 8 We have further demonstrated that this induction of GJC by carotenoids is achieved by increasing the expression of connexin 43 (Cx43), 9 o n e of a family of genes coding for a structural subunit of the gap junction. 10 Here we describe the measurement of gap junctional communication at the functional level by means of assays of dye transfer. Because junctions may be in the open or closed state, any observed increased dye transfer can be the result of an increased number of junctions and/or an increased number of open junctions. To distinguish between these possibilities we have included assays for the amount and cellular location of connexin 43. Gap junctions are clusters of hydrophilic channels traversing the plasma membranes of coupled cells. The channels when open allow the passive diffusion of inorganic ions and small water-soluble compounds up to about 1000 Da, such as small metabolites, and second messengers. 11:2 Each gap junctional channel is made of two tightly joined hemichannels, each donated by one of the two participating cells. Each hemichannel is termed a hemiconnexon, and two connexons in register are paired to form a continuous aqueous channel by end-to-end joining. Each hemiconnexon is composed of six transmembrane protein subunits termed connexins. A family of connexins has been described with tissue and cell type specificity) 3 Connexin 43 (Cx43), one member of the connexin family, is a major gap junctional protein expressed in many cell types. In the mouse C3H/ 10T1/2 cell culture system (10T1/2) used here, Cx43 appears to be the only connexin expressed) 4 This cell line has been extensively used for studies of carcinogenesis, growth control, and differentiation. 7 7 j. S. Bertram, IARC Sci. Publ. 67, 77 (1985). 8 L.-X. Zhang, R. V. Cooney, and J. S. Bertram, Carcinogenesis (London) 12, 2109 (1991). 9 L.-X. Zhang, R. V. Cooney, and J. S. Bertram, Cancer Res. 52, 5707 (1992). l0 E. C. Beyer, D. L. Paul, and D. A. Goodenough, J. Membr. Biol. 116, 187 (1990). ii W. R. Loewenstein, Cell (Cambridge, Mass.) 48, 725 (1987). lz j. C. Sfiez, J. A. Connor, D. C. Spray, and M. V. L. Bennett, Proc. Natl. Acad. Sci. U.S.A. 86, 2708 (1989). 13 M. Tornomura, K. Kadomatsu, S. Matsubara, and T. Muramatsu, J. Biol. Chem. 265, 10765 (1990). i4 M. Rogers, J. M. Berestecky, M. Z. Hossain, H. Guo, R. Kadle, B. J. Nicholson, and J. S. Bertram, Mol. Carcinog. 3, 335 (1990).
[19]
REGULATION OF GAP JUNCTIONAL COMMUNICATION
237
Methods
Cell Cultures 10T1/2 cells, which can be obtained from the ATCC (Rockville, MD), are grown in Eagle's basal medium supplemented with 5% serum. 15 We have found that iron-supplemented calf serum (HyClone, Logan, UT) can substitute for the increasingly expensive fetal serum previously used for growth of this cell line. Substitution results in an approximate 6-fold reduction of costs and the opportunity to purchase larger lots of serum which reduces the need for frequent screening of new lots of serum. Cells are used between passages 5 and 15 and are cultured at 37° in 5% CO2 in air. Because of the requirement for low passage cells it is essential that multiple vials of cells be frozen in liquid N2 as soon as possible. We have found that 10% (v/v) dimethyl sulfoxide (DMSO) in complete medium is a superior cryopreservative to glycerol. For communication assays cells are normally seeded at a density of 104 cells/60-mm culture dish; under these conditions cells should be confluent within 7 days.
Drug Delivery Extensive research in vitro with carotenoids was previously obstructed by difficulty in delivering these highly lipophilic compounds to target cells. We have overcome this problem by the use of tetrahydrofuran (THF) as a delivery vehicle. This allows the formation of pseudosolutions of carotenoids in cell culture medium at high concentrations. 16,17Carotenoids are dissolved in T H F [99.9% pure containing 0.025% (w/v) butylated hydroxytoluene (BHT) as preservative, Aldrich Chemical Co., Milwaukee, WI]. Solutions must be kept in dark bottles, sealed with a neoprene septum, under inert gas at - 20°. Solutions can be withdrawn by hypodermic glass syringe without addition of atmosphere, as described in detail previously. 17The final concentration of THF in the culture medium should not be above 0.5%. Routinely, 2 mM stock solutions of carotenoids are prepared. Under these conditions, solutions of canthaxanthin can be kept for at least 1 month, and most carotenoids can be stored for up to 2 weeks without degradation. Lycopene should be kept for only 1 week under these conditions. 15 C. A. Reznikoff, J. S. Bertram, D. W. Brankow, and C. Heidelberger, Cancer Res. 33, 2339 (1973). ~6j. S. Bertram, A. Pung, M. Churley, T. J. I. Kappock, L. R. Wilkins, and R. V. Cooney, Carcinogenesis (London) 12, 671 (1991). 17 R. V. Cooney and J. S. Bertram, this series, Vol. 214, p. 55.
238
ASSAY OF STRESS GENES/PROTEINS
[19]
Cultures should be mildly agitated after addition of the drugs. Readers are strongly recommended to consult Cooney and Bertram 17 for details of the preparation and handling of carotenoid solutions.
Measurement of Gap Junctional Communication by Dye Microinjection To minimize interexperimental variations cells should be confluent prior to treatment with carotenoids and should have an identical history of plating and refeeding. In particular, increasing age of the cultures results in progressively increasing communication, whereas recent (within 24-48 hr) refeeding causes decreases in communication. Cells are injected with the fluorescent dye Lucifer Yellow CH (LH) (Sigma, St. Louis, MO), which has been widely utilized for the measurement of gap junctional communication. LH is a substituted naphthalimide with two sulfonate groups (Mr 472.2). It has found wide usage in the measurement of gap junctional communication because (1) it is water soluble and sufficiently small to be able to travel through gap junctions but, since it is charged, cannot diffuse through the surface of cell membranes; (2) it does not become protein-bound and thus remains mobile; (3) its excitation and fluorescence spectra are similar to fluorescein and thus can be visualized by the usual fluorescence optics; and (4) it is nontoxic at useful concentrations. Junctional permeability is assayed by microinjection of LH into a single cell with the aid of an Eppendorf microinjector. Micropipettes are prepared from glass capillary tubes with a Flaming-Brown micropipette puller, Model P-80/PC (Sutter Instrument Co., San Rafael, CA). The correct parameters for pulling a good micropipette needle must be determined empirically, since each heating filament and each batch of glass tubes require individual settings. Once determined the parameters can be stored in the electronic memory of the instrument. The puller allows for up to 10 programs to be stored. As a guide we use tubes of borosilicate glass (outer diameter 1 mm, inner diameter 0.78 mm) from Sutter Instruments. The tip after pulling should have a diameter of about 0.2/~m with an acute taper. These parameters result in a robust, fairly rigid pipette which causes a rapidly sealing puncture in the cell membrane. The micropipette is backfilled with a few microliters of 10% Lucifer Yellow solution in 0.33 M LiC1 through a 10-/~1glass syringe, then attached to a Zeiss micromanipulator connected to an Eppendorf microinjector. The fluorescent dye is microinjected under consistent pressure of Nz into the perinuclear region a single cell identified under phase-contrast or Nomarsky optics. A final magnification of at least 200 × is required. Injection pressures must be determined empirically.
[19]
REGULATION OF GAP JUNCTIONAL
COMMUNICATION
239
We normally inject 15 to 20 cells in a line across a dish over a 10-rain time period; at the end of this time the first injected cell is relocated under fluorescence optics and the image recorded on videotape. All cells are similarly recorded in sequence over a 10-min interval. The total number of fluorescent cells adjacent to each dye-injected cell is then determined and serves as an index of gap junctional communication. Measurements of gap junctional communication are performed in duplicate cultures for each treatment. Figures 1 and 2A,B show the induction of gap junctional communication by carotenoids using this method.
Immunofluorescent Detection of Gap Junctional Plaques Cells are seeded on Permanox plastic slides (Nunc Inc., Naperville, 1L) for fluorescent microscopy. Conventional plastic dishes will themselves fluoresce and interfere with the fluorescein isothiocyanate (FITC) signal, whereas cells grown on glass have a different morphology from those grown on plastic. Treatment with carotenoids is carried out as described above. After treatment, the slides are briefly washed in 37° PBS + [phosphate-buffered 40-
-~ C)
321
t--
h-~
24-
O c-
E E
o o
16 g
0
0 0
7
14
21
28
Days of T r e a t m e n t FIG. 1. Carotenoid effects on gap junctional communication in 10T1/2 cells. The cells were seeded at 1000 cells/60-mm dish and treated with 10 -s M carotenoids dissolved in THF when confluent on the seventh day and weekly thereafter. Junctional communication was indexed as the number of cells to which Lucifer Yellow was transferred within 10 min of injection into a test cell. Data points are the means -+ SE of 15 microinjections performed in two dishes each with the following carotenoids: Q, 0.5% THF control; V, fl-carotene; ¢,, canthaxanthin; II, lutein; A, lycopene. (From Zhang e t al., s with permission.)
240
ASSAY OF STRESS GENES/PROTEINS
[19]
[19]
REGULATION OF GAP JUNCTIONAL COMMUNICATION
241
saline (PBS) as defined below, plus 0.01% MgSO4 and 0.01% CaCI 2] and fixed with cold methanol ( - 20°). If slides are to be stored they should be first transferred to room temperature methanol before drying. If they are to be immediately processed they should be hydrated for 10 min in cold PBS (0.8% NaC1, 0.115% Na2HPO 4, 0.02% KH2PO4, and 0.02% KC1), followed by blocking for 1 hr in PBS containing 5% bovine serum albumin (BSA) and 0.2% sodium azide at 4 °. The primary antibody is then applied diluted 1 : 30 with the blocking solution. We use a polyclonal antiserum raised in rabbits by immunization against a synthetic polypeptide corresponding to the C-terminal region (residues 368-382) of the predicted sequence of connexin 43.14 Slides are incubated at 4 ° for 60 min in a humidified chamber, then rinsed with cold PBS and washed for I hr in high salt PBS (0.5 M NaC1 in PBS) followed by a final wash in PBS. FITC-conjugated second antibody (goat anti-rabbit IgG F(ab)z fragment (Sigma) is diluted 1 : 40 in blocking solution, placed on the slides, and incubated for 60 min at 4 °. Slides are washed as above and mounted under coverslips using 0.233 g 1,4-diazobicyclo[2.2.2]octane (DABCO), 800/zl water, 200 ~1 1 M Tris-HC1, pH 8.0, and 9.0 ml glycerol to retard photobleaching (P. Lichter, personal communication, 1992). The slides are finally sealed with clear nail polish. Immunofluorescent microscopy of the slides is performed using a Zeiss Axioplan microscope (Thornwood, NY). Gap junctions are characterized as fluorescent plaques in junctional regions of the cell membrane (Fig. 2).
Western Blot Analysis of Connexin Proteins Isolation of Membrane Proteins. Cultures of 10T1/2 cells are grown in 150-mm dishes and treated with carotenoids as described above. After removal of culture medium dishes are washed with ice-cold PBS containing I mM NaF and 1 mM phenylmethylsulfonyl fluoride (PMSF, dissolved in ethanol), and then cells are scraped from the dish with a rubber policeman into 10 ml of the same solution. After centrifugation at 3000 rpm for 10
FIG. 2. Increased gap junctional communication is accompanied by an increased number and size of immunofluorescent plaques in regions of cell-cell contact. Confluent cultures were treated with 10 -5 M canthaxanthin for 7 days. (A, B) Photomicrographs of dye transfer. (A) Solvent control; (B) canthaxanthin. Junctional permeability was assayed by microinjection of the fluorescent dye Lucifer Yellow CH (10% in 0.33 M LiC1) into a single cell as described. 21(C-F) Immunofluorescent detection of junctional plaques induced by canthaxanthin. (C, E) Control, fluorescent and respective phase-contrast image; (D, F) canthaxanthintreated cells, fluorescent and respective phase-contrast image. 10T1/2 cells were seeded on Permanox plastic slides and treated as in Fig. 3A,B. Arrows indicate junctional plaques surrounding a single cell. (From Zhang et al., 9 with permission.)
242
ASSAY OF STRESS GENES/PROTEINS
[19]
min, cell pellets are either frozen at - 7 0 ° (up to 2 weeks) or lysed in 30-50/xl of lysis buffer [1% Tergitol NP-40, 50 mM iodoacetamide, 10 mM PMSF, 1 mM EDTA, 1 p.M leupeptin, 2/zg/ml aprotinin, 0.7 ~g/ml pepstatin in borate buffer] for 1 hr at 4 °. Lysates are then centrifuged at 10,000 g for 15 min at 4 ° and the supernatant assayed by Western blotting for connexin 43. Prior to electrophoresis, the protein content is assayed using the BCA Protein Assay Reagent (Pierce Chemical Co., Rockford, IL) according to the manufacturer's instructions. Protein Electrophoresis and Western Blotting. Protein samples are separated by conventional sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) according to Laemmli. ~8 Cell lysates containing equal amounts of protein (30-60 ~g/lane) are mixed well with 2 x SDS sample buffer containing 10% of 2-mercaptoethanol. After at least 15 min of reaction at room temperature without boiling (the samples must be reduced with 2-mercaptoethanol but not heated~4), proteins are electrophoresed on a 10% SDS-polyacrylamide minigel (Bio-Rad, Richmond, CA) at 200 V. Prestained SDS-PAGE protein standards (Bio-Rad) are also loaded for the indication of both molecular weight and transference efficiency. After electrophoresis, the proteins in the gel are transferred to an Immobilon membrane (Millipore, Bedford, MA) at 100 V and 4 ° with stirring for 35-40 min. After transfer, the gel can be stained with 0.25% Coomassie blue R solution [methanol-glacial acetic acid-water, 50 : 10 : 40 (v/v)] and destained with 5% methanol/9% acetic acid for checking protein transference. Detection of Connexin 43. After Western blotting, the membrane is blocked with blotto for 1 hr or overnight at 4° and then reacted with the same primary anti-Cx43 antibody as used in the immunofluorescence studies diluted 1 : 200 in blotto-borate buffer (1 : 2, v/v) with gentle shaking. (blotto is 5% nonfat dry milk in PBS; borate buffer contains 1.03% boric acid, 0.785% NaCI, and 0.11% NaC1, pH 8.0). After incubation, the blot is washed with excess borate buffer for 1 hr with 3-4 changes of the buffer. ~25I-labeled protein A is diluted 1:2 in blotto-borate buffer to achieve a concentration of 106 cpm/I 0 ml of solution and the blot incubated with this solution for 1 hr at room temperature with shaking, followed by a 1-hr wash in borate buffer with three changes of washing buffer. The blot is then gently pressed against absorbent paper to remove excess liquid, wrapped in plastic film, and exposed to Kodak (Rochester, NY) X-Omat AR film with double intensifying screens at - 7 0 ° for 2-3 days. Figure 3 demonstrates the use of this method to detect induction of Cx43 by carotenoids. I8 U. K. Laemmli, Nature (London) 227, 680 (1970).
[19]
R E G U L A T I O N OF GAP J U N C T I O N A L C O M M U N I C A T I O N
A
1
234
0
1
5
6
4
7
243
78
Cx43 B
2
3
Cx43 treatment (days)
Cx43 0 0.31.03.010 canthaxanthin (uM) FIG. 3. Induction of Cx43 in 10T1/2 cells by carotenoids and the synthetic retinoid tetrahydrotetramethylnapthalenylpropenylbenzoic acid Hoffmann-La Roche (Nutley, NJ) (TTNPB) but not ~-tocopherol. (A) Structure-activity studies. Cultures were treated with 10-8 M TTNPB for 3 days, 10-5 M fl-carotene or canthaxanthin for 7 days, or 10-5 M lycopene or a-tocopherol for 7 days with retreatment every 3 days. Lane 1, Solvent control; 2, a-tocopherol; 3, canthaxanthin; 4, lycopene; 5, fl-carotene; 6, TTNPB. Lanes 7 and 8 show results of a separate experiment in which cultures received solvent control (lane 7) or 10-5 M methylbixin for 7 days (lane 8). (B) Time course. Cultures were treated with 10-5 M canthaxanthin and analyzed at the indicated times. (C) Dose-response. Cultures were treated with the stated concentrations of canthaxanthin for 7 days then analyzed. (From Zhang e t al., 9 with permission.)
Chemiluminescent Detection. As an alternative detection m e t h o d , we have f o u n d that a c o m m e r c i a l n o n i s o t o p i c kit, W e s t e r n - L i g h t Chemilumin e s c e n t D e t e c t i o n S y s t e m ( T R O P I X , B e d f o r d , MA), is m o r e sensitive and c o n v e n i e n t than the a b o v e m e t h o d o f radioactive detection. Using the kit, only a few minutes is required to identify i m m u n o r e a c t i v e proteins immobilized on m e m b r a n e s , in c o n t r a s t to the 2 or 3 d a y s n e e d e d for the isotopic detection m e t h o d . A f t e r processing, the m e m b r a n e should be stained for a b o u t 10 min with 0.25% C o o m a s s i e R solution as used for gel staining, destained with 50% m e t h a n o l / 2 5 % acetic acid, and p h o t o g r a p h e d . This serves as an internal c o n t r o l for normalizing for e a c h sample the a m o u n t o f total proteins loaded and blotted.
244
ASSAY OF STRESS GENES/PROTEINS
[20]
Analysis and Quantitation. Analysis and quantification can best be carried out using digital image analysis which allows the subtraction of background and the direct comparison of treated and control bands. We use the National Institutes of Health Image program, a public domain program written for the Macintosh.
[20] O n e - D a y N o r t h e r n Blotting for D e t e c t i o n of m R N A : N D G A Inhibits t h e I n d u c t i o n of M n S O D m R N A b y Agonists of T y p e 1 T N F R e c e p t o r
By GRACE H. W. WONG and DAVm V. GOEDDEL Introduction Production of tumor necrosis factor (TNF) by cells is transient, tightly regulated, and can be triggered by infection and by oxidative stress. 1 In addition to its originally defined tumoricidal activity, TNF has been shown to act on many cell types and mediate pleiotropic activities2 including protection against various types of oxidative stress. 3 TNF also induces the expression of manganous superoxide dismutase (MnSOD), an enzyme that scavenges superoxide radical (02 ~) in mitochondria.4 The induction of MnSOD expression occurs at the transcriptional level and requires very low levels of TNF (0.1 ng/ml). The induction of MnSOD mRNA is rapid (less than 1 hr), substantial (more than 10-fold), and direct because it can be blocked by actinomycin D but not by cycloheximide.5 This chapter describes a 1-day Northern blotting method to examine the regulation of MnSOD mRNA. B. Beutler, ed., "Tumor Necrosis Factors: The Molecules and Their Emerging Role in Medicine." Raven, New York, 1992. 2 W. Fiers, F E B S Lett. 285, 199 (1991). 3 G. H. W. Wong, A. Kamb, L. A. Tartaglia, and D. V. Goeddel, "Molecular Biology of Free Radical Scavenging Systems," pp. 69-96. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 1992. 4 G. H. W. Wong and D. V. Goeddel, Science 242, 941 (1988). 5 G. H. W. Wong, A. Kamb, J. H. Elwell, L. W. Oberley, and D. V. Goeddel, in "Tumor Necrosis Factors: The Molecules and Their Emerging Role in Medicine" (B. Beutler, ed.), p. 473. Raven, New York, 1992.
METHODSIN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress,Inc. All rightsof reproductionin any formreserved.
244
ASSAY OF STRESS G E N E S / P R O T E I N S
[20]
Analysis and Quantitation. Analysis and quantification can best be carried out using digital image analysis which allows the subtraction of background and the direct comparison of treated and control bands. We use the National Institutes of Health Image program, a public domain program written for the Macintosh.
[20] O n e - D a y N o r t h e r n Blotting for D e t e c t i o n o f m R N A : N D G A Inhibits t h e I n d u c t i o n of M n S O D m R N A b y Agonists of T y p e 1 T N F Receptor
By GRACE H. W. WONG and DAVID V. GOEDD~L Introduction Production of tumor necrosis factor (TNF) by cells is transient, tightly regulated, and can be triggered by infection and by oxidative stress. ~ In addition to its originally defined tumoricidal activity, TNF has been shown to act on many cell types and mediate pleiotropic activities2 including protection against various types of oxidative stress. 3 TNF also induces the expression of manganous superoxide dismutase (MnSOD), an enzyme that scavenges superoxide radical (02:) in mitochondria.4 The induction of MnSOD expression occurs at the transcriptional level and requires very low levels of TNF (0.1 ng/ml). The induction of MnSOD mRNA is rapid (less than 1 hr), substantial (more than 10-fold), and direct because it can be blocked by actinomycin D but not by cycloheximide.5 This chapter describes a 1-day Northern blotting method to examine the regulation of MnSOD mRNA. B. Beutler, ed., "Tumor Necrosis Factors: The Molecules and Their Emerging Role in Medicine." Raven, New York, 1992. 2 W. Fiers, FEBS Lett. 285, 199 (1991). 3 G. H. W. Wong, A. Kamb, L. A. Tartaglia, and D. V. Goeddel, "Molecular Biology of Free Radical Scavenging Systems," pp. 69-96. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 1992. 4 G. H. W. Wong and D. V. Goeddel, Science 242, 941 (1988). 5 G. H. W. Wong, A. Kamb, J. H. Elwell, L. W. Oberley, and D. V. Goeddel, in "Tumor Necrosis Factors: The Molecules and Their Emerging Role in Medicine" (B. Beutler, ed.), p. 473. Raven, New York, 1992.
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[20]
NORTHERN BLOTTING FOR DETECTION OF m R N A
245
One-Day Northern Blotting A. Isolation of Total Cytoplasmic RNA 1. Harvest adherent cultured cells either by trypsinization or with a cell scraper. Transfer cells (106-108) to an autoclaved siliconized l-ml Eppendorf tube and wash once with cold phosphate-buffered saline (PBS). Autoclave tubes, tips, water, and all buffers before use. Wear plastic gloves throughout. 2. Fast centrifuge the cells at room temperature (12,000 rpm, 5 sec). To avoid cell death, never centrifuge for longer than 20 sec. 3. Remove the supernatant and keep the cell pellet on ice. 4. Add 300/zl ice-cold lysis buffer A [10 mM Tris-HC1, pH 7.5, 0.1 M NaC1, 5 mM MgCI2, 0.5% Nonidet P-40 (NP-40)] containing 10 mM of the RNase inhibitor vanadyl-ribonucleoside complexes (VRC, from BioRad, Richmond, CA). 5. Vortex (maximum speed on a bench-top mixer) and incubate on ice for 30 sec (never more than 2 min; otherwise the buffer will lyse the nuclear membrane and the cytoplasmic RNA will be contaminated with DNA and nuclear RNA). 6. Centrifuge for 1 min to pellet nuclei (save intact nuclei for DNA extraction); RNA is in the supernatant. 7. Add 100/zl sodium dodecyl sulfate (SDS) buffer B (10 mM TrisHCI, pH 7.5, 0.1 M NaC1, 5 mM EDTA, 0.1% SDS) containing fresh 0.1% 2-mercaptoethanol (2-ME) and 0.1% diethyl pyrocarbonate (DEPC, Sigma, St. Louis, MO) to the RNA-containing supernatant. Add 700/zl phenol-chloroform (50:50, v/v) and mix thoroughly. (The RNA can be stored in this mixture at - 2 0 ° without degradation for several years.) Precaution: When samples are removed from storage at - 2 0 °, open the cap immediately to avoid explosion during warming and add 10/zl of a mixture of 2-ME and DEPC to prevent RNA degradation. Work in a fume hood and wear protective clothing. 8. Extract the RNA with 700/zl phenol-chloroform at least twice. [Phenol is equilibrated with water, 8-hydroxyquinoline (0.1%), and 2-ME (0. I%) to avoid oxidation and also provide yellow color indication.] If no aqueous phase can be recovered, remove the organic phase (yellow layer) from the bottom and extract with chloroform alone to obtain a clear aqueous phase and proceed. 9. Precipitate the RNA with 800/zl of 95% ethanol and 30/zl of 3 M sodium acetate (pH 7.5) at - 2 0 ° for 20 rain. 10. Collect the RNA pellet by microcentrifugation (15 rain, 14,000 rpm, cold room).
246
ASSAY OF STRESS GENES/PROTEINS
[20]
1 I. Rinse the pellet with cold 75% ethanol and drain dry. 12. Redissolve the pellet in 100 /zl water containing 0.1% sarkosyl (sodium N-lauroylsarcosinate, 30%, from Pfalz and Bauer Inc., Waterbury, CT). 13. Dilute 5/~1 in 500/~1 water to determine A260 and A280. Check the quality of the RNA by analytical electrophoresis on a vertical 1.2% agarose formaldehyde gel before preparation of poly(A) + RNA. Store the RNA at - 2 0 ° until use.
B. Monitoring Quality of RNA by Vertical Agarose Formaldehyde Gel Electrophoresis 1. Prepare running gel buffer 1 × MEN, pH 7.5 [10 x MEN is 200 mM 3-(N-morpholino)propanesulfonic acid (MOPS), 50 mM sodium acetate, 10 mM EDTA; the 10 x MEN buffer will be yellow after autoclaving]. 2. Prepare the 1.2% agarose gel (1.8 g SeaKern HGT agarose (FMC BioProducts, Rockland, ME), 15 ml of 10 × MEN, 110 ml water). 3. Microwave for 3 rain to melt the agarose. 4. Heat the taped vertical gel plates in an oven (80°) for 5 min (because agarose solidifies quickly on cold plates). 5. Add 25 ml formaldehyde to the melted agarose (the temperature drops to approximately 70 °) and prepare to pour the gel (do not wait for it to cool). 6. Transfer the agarose mix into a 50-ml blue-cap tube and pour into the vertical gel plates carefully (without bubbles). 7. Insert a gel comb into the agarose and let it solidify for 30 min inside a fume hood. 8. Denature 5 ~g RNA at 65 ° for 10 rain using 15 p~lof RNA denaturing buffer per sample (stock contains 1000/A formamide, 200/A of 10 × MEN, 320/~1 formaldehyde, and 480/A water). The RNA in the denaturing buffer can be stored a t - 20° without degradation for at least I year, and the denaturing buffer is good for at least 6 months. 9. Remove the gel comb very carefully or the wells may break. Remove the comb as slowly as possible and constantly apply water into each well and make sure the binder clips are off before bending the comb forward and backward. 10. Always prerun the gel to ensure that the buffer and the gel are working before loading the RNA samples. 11. Add 3 /~1 of loading dye [50% glycerol, 1 mM EDTA (pH 8.0), 0.25% xylene cyanol FF] to each sample (15/~1); spin, mix, and carefully load the sample into the vertical gel. 12. Run the gel at a constant voltage of 150 V for 30 min to complete.
['90]
NORTHERN BLOTTING FOR DETECTION OF m R N A
247
13. Detect the RNA by UV shadowing without staining the gel. Wrap the gel with Saran wrap and place the gel on a fluorescence precoated silica gel thin-layer chromatography (TLC) plate (Plate 60 F254 from Merck, Darmstadt, Germany, or E. M. Science, Cherry Hill, N J). Hold the UV lamp over the wrapped gel, and two sharp bands should appear, the 28 S and 18 S ribosomal RNA. I f R N A is intact, then proceed to make poly(A) ÷ RNA. If a smear of degraded RNA appears, stop and discard the RNA samples. The quality of RNA samples can also be checked even before the phenol-chloroform extraction. Spin the mixture of RNA and phenol-chloroform from Step A.7 (14,000 rpm, 3 min, room temperature (RT)), and mix 5/zl of RNA-containing supernatant (clear top phase; phenol is in the yellow bottom layer) with 10/xl denaturing buffer (from Step B.8) to denature the RNA at 65 ° for I0 min. Run samples on the gel and check the RNA by UV shadowing as described in steps B.8 to B. 13. C. Isolation o f Poly(A) ÷ R N A Preparation of Oligo(dT)-Cellulose 1. Dissolve 1 g of oligo(dT)-cellulose (type III, Collaborative Research, Bedford, MA) in 50 ml water and let swell for 3 min in a 50 ml screw-cap tube. 2. Spin (tabletop centrifuge; 1000 rpm for 10 sec at RT) and remove the water. 3. Clean the oligo(dT)-cellulose with 0.1 M NaOH (50 ml) by inverting and mixing thoroughly for 10 sec. 4. Centrifuge (1000 krpm, 10 sec at RT) to remove the supernatant. 5. Wash once with sterile autoclaved water and centrifuge to remove the supernatant. 6. Wash once with 50 ml of 10 x oligo(dT) buffer (4 M NaC1, 0.1 M Tris, pH 7.5, 10 mM EDTA). 7. Centrifuge (1000 rpm, 10 sec, RT); remove the supernatant. 8. Wash with 50 ml of 1 x oligo(dT) buffer until the pH is neutral (approximately pH 7.5), testing with pH paper. 9. Centrifuge (1000 rpm, 10 sec, RT) to remove the supernatant. 10. Store the oligo(dT)-cellulose pellet at - 2 0 ° until use. Enrichment ofPoly(A) ÷RNA. Always check the quality of RNA before the preparation of poly(A) ÷ RNA by analyzing 5/xl of total RNA on a vertical 1.2% agarose formaldehyde gel as described in Section B. 1. Dissolve the RNA in 300/zl water and incubate at 65 ° for 10 min in order to denature. Transfer to ice. 2. Add 30/zl sarkosyl (sarkosyl is preferable to SDS because sarkosyl does not precipitate at 0 °) to inactivate RNase.
248
ASSAY OF STRESS GENES/PROTEINS
[20]
3. Add 30 ~I of 10 x oligo(dT) buffer. 4. Dissolve the RNase-free oligo(dT)-cellulose pellet in 15 ml of I x oligo(dT) buffer. 5. Take 300 tzl of this mixture and add to the RNA samples (from Step 3). 6. Shake at room temperature in a capped Eppendorf tube for at least 10 min to ensure that the poly(A) ÷ RNA has bound to the oligo(dT)-cellulose. 7. Centrifuge (14,000 rpm, 1 min, RT) to separate the poly(A)- RNA in the supernatant. Save the supernatant until certain that the poly(A) ÷ RNA has been recovered. 8. Rinse the pellet with 1 x oligo(dT) buffer. 9. Elute the poly(A) + RNA by adding 300/zl water containing 0.1% sarkosyl to the pellet; mix thoroughly and incubate at 65 ° for 30 sec. 10. Mix, invert, and centrifuge (14,000 rpm, 1 min, RT). 11. Precipitate the poly(A) ÷ RNA from the supernatant with 800/zl ethanol (95%) and 30/zl sodium acetate (3 M, pH 7.5) at - 20° for 20 min. 12. Centrifuge (14,000 rpm, 10 min, 4 °) and carefully remove the supernatant. 13. Wash the pellet with cold 75% ethanol. 14. Dry and redissolve the poly(A)+ RNA in 20/xl water containing 0.1% sarkosyl and store the poly(A) + RNA at - 2 0 ° until use.
D. Hybridization with Radioactive D N A Probes I. Check the quality of the poly(A) + RNA by UV shadowing as described in Step B. 13. Two sharp bands should be seen indicating that 28 S and 18 S ribosomal RNA remaining in the poly(A) + RNA preparation. The ribosomal RNAs are coprecipitated and usually a sign that goodquality poly(A) + RNA has been made. 2. Transfer the RNA from the thin gel onto a nitrocellulose membrane filter using 20 x SSC. Complete transfer takes 20 rain. The 28 S and 18 S ribosomal RNAs will disappear from the gel after transfer (check with the UV shadowing technique). 3. Bake the filter in a vacuum oven at 80° for 10 min. 4. Prehybridize the filter at 42° for 10 rain in a hybridization box containing 50 ml prehybridization mix (see below). 5. Add the denatured 3ZP-labeled probe (I00 ° for 5 rain, then ice for 1 min) to the prehybridization mix (see below), add dextran sufate to 10%, and further hybridize for at least 3 hr at 42°. 6. Rinse the filter 3 times in 0.2 x SSC plus 0.1% SDS at room temperature for 1-2 min.
[20]
249
NORTHERN BLOTTING FOR DETECTION OF m R N A
7. Wash in 0.2 x SSC plus 0.1% SDS at 42° with constant shaking for 20 min. 8. Transfer the filter onto the blotting paper and wrap the filter with Saran wrap immediately (never let the filter completely dry). 9. Expose the filter to X-ray film at - 70° (normally, 10 min is sufficient to detect/3-actin signal and 1-3 hr is required to detect MnSOD mRNA signal if 3/zg of poly(A) + RNA is used per lane).
Prehybridization Mix 500ml 250ml 50ml 50 ml 10 ml I0 ml 10 ml 120 ml 1000ml
Formamide 20 x SSC (3 M NaC1, 0.3 M sodium citrate, pH 7.0) 100 x Denhardt's solution [2% (w/v) each of Ficoll, polyvinylpyrrolidone, and bovine serum albumin (BSA)] 1 M Sodium phosphate, pH 6.5 3 M Tris-HC1, pH 8.5 SDS (10%) Salmon sperm DNA (Sigma) (10 mg/ml, boiled for 5 min then chilled in ice) Water (pH8.0-8.5)
E. Labeling Overlapping Oligonucleotides by Filling in with 32P-Labeled Nucleotides I. Add 1 ~1 of each oligonucleotide (concentration of 2.5OD260/ml) to 35 ~lwater.
,
A ,
3,111/I/3' B
Sequence of MnSOD 66-mer A: 5'-GGG GAG TTG CTG GAA GCC ATC AAA CGT GAC TTT GGT TCC TTT GAC AAG TTT AAG GAG AAG CTG ACG-3' Sequence of MnSOD 63-mer B: 5'-GAA ACC AAG CCA ACC CCA ACC TGA GCC TTG GAC ACC AAC AGA TGC AGC CGT CAG CTT CTC CTT-3' /3-Actin sequence (45-mer) A: 5'-GCC TCT GGC CGT ACC ACT GGC ATC GTG TAG GAC TCC GGT GAC GGG-3'
250
ASSAY OF STRESS GENES/PROTEINS
[20]
fl-Actin sequence (45-mer) B: 5'-CCC CTC GTA GAT GGG CAC AGT GTG GGT GAC CCC GTC ACC GGA GTC-3' The underlined regions of sequences A and B are complementary. 2. Add 5/zl of 10x Klenow polymerase buffer (0.5 M Tris-Cl, 0.1 M MgCI2, pH 7.5). 3. Incubate at 55 ° for 10 min. 4. Further incubate at 25 ° for 10 rain, then put on ice. 5. Add 3/zl of dGTP (2 mM) and 3/zl of dTTP (2 mM). 6. Add 10/xl of a-a2p-labeled dATP (Amersham, Arlington Heights, IL, 10204, 10 mCi/ml) and 10/zl of ot-32p-labeled dCTP (Amersham 10205, 10 mCi/ml). 7. Add 2 units of Klenow polymerase. 8. Incubate for 15 min at 25°. 9. Further incubate for 15 min at 37°. 10. Add 200/zl water and ethanol precipitate with 30/xl sodium acetate (3 M), 700/zl ethanol (95%), and 1 ~1 of carrier tRNA or DNA (10 mg/ml), or purify the probe with a Bio-Spin chromatography column (Bio-Rad). 11. Redissolve the DNA probe in 50 ~1 water. The probe should contain at least about 108 cpm and is ready for use (use the entire probe per 50 ml in a hybridization box). Store the labeled probe at 4°; it may be reused for at least 3-4 weeks (by reheating to 70 ° for 5 min and chilling in ice for 3 min). Example
Figure 1 shows an example of Northern blot analysis using the 1-day procedure. 6-9 Both MnSOD and fl-actin probes were used. Agonist antibodies against the type 1 TNF receptor (TNF-R1) but not the type 2 TNF receptor (TNF-R2) were found to induce high levels of MnSOD mRNA, whereas/3-actin mRNA remained unchanged. Thus, TNF induces MnSOD expression by activation of TNF-R1 and is not involved in intracellular events per se. The intracellular mediators of the activation signal are unknown. Surprisingly, induction of MnSOD by both TNF and the antiTNF-R1 antibodies can be blocked by nordihydroguaiaretic acid (NDGA), 6 y . Beck, R. Oren, B. Arnit, A. Levanon, M. Gorecki, and J. R. Hartman, Nucleic Acids Res. 15, 9076 (1987). 7 p. Ponte, S. Y. Ng, J. Engel, P. Gunning, and L. Kedes, Nucleic Acids Res. 12, 1687 (1984). 8 G. H. W. Wong, L. A. Tartaglia, M. S. Lee, and D. V. Goeddel, J. Immunol. 149, 3350 (1992). 9 L. A. Tartaglia and D. V. Goeddel, J. Biol. Chem. 267, 4304 (1992).
[20]
NORTHERN BLOTTING FOR DETECTION OF m R N A
251
fl I
8
II
I
.,,, -
- M n S O D (4kb)
I~-Actin-- m - M n S O D (lkb) 1 2 3 4 5 6 7 8 FIG. 1. Nordihyguaiaretic acid (NDGA) but not indomethacin (INDO) inhibits the induction of MnSOD mRNA by TNF and by anti-TNF-Rl antibodies. Confluent cultures of A549 human lung carcinoma cells were treated with TNF-a (0.1 /xg/ml) or with a 1/100 dilution of polyclonal antibodies against TNF-R1 or TNF-R2 for 1 hr at 4°, washed three times with medium, and further incubated with or without 10 mM of either NDGA or INDO for 12 hr. Poly(A) ÷ RNA (3/zg/lane) was hybridized to both 32p-labeled MnSOD and/3-actin probes. 6-8 The specific activity of human TNF-~ is 4 × 107 units/rag, and the rabbit antisera against human TNF-R1 or TNF-R2 have been described previously. 9-j°
an inhibitor of lipooxygenase. Indomethacin (INDO), an inhibitor of cyclooxygenase, does not block the induction (Fig. 1). These results suggest that leukotrienes or some other product(s) of the lipoxygenase pathway may mediate the induction of MnSOD mRNA by TNF. The 1-day Northern blotting method can be used to isolate pure and undegraded poly(A) + RNA from a large number of samples for hybridization with various DNA probes. The method is simple, inexpensive, rapid, and reliable and is suitable for gene regulation studies using cultured cells. Enrichment of poly(A) + RNA by the batch method in an Eppendorf tube allows one to enrich a large sample of poly(A) + RNA within a short time. This I-day Northern blotting method is not suitable for isolating RNA from tissues. The intact nuclei isolated during the preparation of cyto-
252
ASSAY OF STRESS G E N E S / P R O T E I N S
[21]
plasmic R N A can also be used for nuclear runon/runoff experiments and for isolating D N A . The U V shadowing m e t h o d can be conveniently used to detect and p h o t o g r a p h 28 S and 18 S R N A without staining or other manipulation of the gel. Acknowledgments We thank the manufacturing group at Genentech (San Francisco, CA) for providing pure recombinant human TNF-a; Greg Bennett for preparation of rabbit anti-TNF-R1 and antiTNF-R2 antibodies; Louis Tamayo for art work; Tracey Rivas and Sunita Sohrabji for help in preparing the manuscript; and Drs. Alan Harris, Susanne Baumhueter, and Louis Tartaglia for critical comments. G.H.W.W. thanks Dr. Linus Pauling, Linda Pauling-Kamb, and Alexander David Kamb for support.
[21] E v a l u a t i o n
of Biomolecular
Damage
by Ozone
B y CARROLL E. CROSS and BARRY HALLIWELL
Introduction O z o n e (O3), an important toxic c o m p o n e n t of photochemical air pollution, is believed to exert its toxic effects via its strong oxidizing capacity. 1-4 Although its pathophysiological effects on the respiratory tract have been extensively studied, 1,5 and m u c h is k n o w n concerning its chemical reactivity with a variety of biological molecules studied in i s o l a t i o n , 2-4,6-8 the precise biomolecular m e c h a n i s m s of inhaled 03 toxicity are not fully understood. The respiratory tract lining fluids ( R T L F s ) represent the first biological substances coming into contact with inhaled 03. As such, constituents of the R T L F s m a y well represent an important first line of respiratory tract defense, p r e s u m a b l y b y absorbing, reacting with, and detoxifying 03. Indeed, it has been suggested that, at concentrations expected to be present in polluted air, inhaled 03 and NO2 will react p r i m a r i l y with the oxidizable biomolecules present in the R T L F s and that these inhaled I D. B. Menzel, 13, 183 (1984). 2 M. A. Mehlman and C. Borek, Environ. Res. 42, 36 (1987). 3 M. G. Mustafa, Free Radical Biol. Med. 9, 245 (1990). 4 W. A. Pryor, Am. J. Clin. Nutr. 53, 702 (1991). 5 M. Lippmann, J. Air Pollut. Control Assoc. 39, 692 (1989). 6 j. B. Mudd, R. Leavitt, A. Ongun, and T. T. McManus, Atmos. Environ. 3, 669 (1969). 7 R. S. Oosting, M. M. J. Van Greevenbroek, J. Verheof, L. M. G. Van Golde, and H. P. Haagsman, Am. J. Physiol. 261, L77 (1991). 8 W. A. Pryor, B. Das, and D. F. Church, Chem. Res. Toaicol. 4, 341 (1991).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All fights of reproduction in any form reserved.
252
ASSAY OF STRESS G E N E S / P R O T E I N S
[21]
plasmic R N A can also be used for nuclear runon/runoff experiments and for isolating D N A . The U V shadowing m e t h o d can be conveniently used to detect and p h o t o g r a p h 28 S and 18 S R N A without staining or other manipulation of the gel. Acknowledgments We thank the manufacturing group at Genentech (San Francisco, CA) for providing pure recombinant human TNF-a; Greg Bennett for preparation of rabbit anti-TNF-R1 and antiTNF-R2 antibodies; Louis Tamayo for art work; Tracey Rivas and Sunita Sohrabji for help in preparing the manuscript; and Drs. Alan Harris, Susanne Baumhueter, and Louis Tartaglia for critical comments. G.H.W.W. thanks Dr. Linus Pauling, Linda Pauling-Kamb, and Alexander David Kamb for support.
[21] E v a l u a t i o n
of Biomolecular
Damage
by Ozone
B y CARROLL E. CROSS and BARRY HALLIWELL
Introduction O z o n e (O3), an important toxic c o m p o n e n t of photochemical air pollution, is believed to exert its toxic effects via its strong oxidizing capacity. 1-4 Although its pathophysiological effects on the respiratory tract have been extensively studied, 1,5 and m u c h is k n o w n concerning its chemical reactivity with a variety of biological molecules studied in i s o l a t i o n , 2-4,6-8 the precise biomolecular m e c h a n i s m s of inhaled 03 toxicity are not fully understood. The respiratory tract lining fluids ( R T L F s ) represent the first biological substances coming into contact with inhaled 03. As such, constituents of the R T L F s m a y well represent an important first line of respiratory tract defense, p r e s u m a b l y b y absorbing, reacting with, and detoxifying 03. Indeed, it has been suggested that, at concentrations expected to be present in polluted air, inhaled 03 and NO2 will react p r i m a r i l y with the oxidizable biomolecules present in the R T L F s and that these inhaled I D. B. Menzel, 13, 183 (1984). 2 M. A. Mehlman and C. Borek, Environ. Res. 42, 36 (1987). 3 M. G. Mustafa, Free Radical Biol. Med. 9, 245 (1990). 4 W. A. Pryor, Am. J. Clin. Nutr. 53, 702 (1991). 5 M. Lippmann, J. Air Pollut. Control Assoc. 39, 692 (1989). 6 j. B. Mudd, R. Leavitt, A. Ongun, and T. T. McManus, Atmos. Environ. 3, 669 (1969). 7 R. S. Oosting, M. M. J. Van Greevenbroek, J. Verheof, L. M. G. Van Golde, and H. P. Haagsman, Am. J. Physiol. 261, L77 (1991). 8 W. A. Pryor, B. Das, and D. F. Church, Chem. Res. Toaicol. 4, 341 (1991).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All fights of reproduction in any form reserved.
[21]
EVALUATION OF BIOMOLECULAR DAMAGE BY OZONE
253
gases are unlikely to reach the underlying respiratory tract epithelial cells (RTECs). 9-1° It follows that the toxicity of O3 to R T E C s may not be mediated by direct O 3 attack, but by cytotoxic products generated when 03 interacts with the R T L F s . Respiratory tract lining fluids are difficult to obtain and variable in composition. To examine the spectrum of reactions that can be expected to occur when complex biological fluids (such as R T L F s ) are exposed to 03, we have used plasma as a target. 11 Plasma is easier to obtain than R T L F s and contains a wide range of antioxidants (both aqueous and in the lipid phase). 12-14 Of course, it is not an ideal model (e.g., mucin and surfactant are missing). Injury to the lung, for example, by high concentrations of oxidants, however, can be expected to cause transudation of plasma constituents into the R T L F s , 15 that is, they become closer to plasma in composition, Indeed, it has been proposed 12 that this transudation, often used as a marker of injury, is also beneficial in that it will increase the total antioxidant capacity of the R T L F s . Hence, exposure of plasma to 03, followed by measurements of consumption of antioxidants in relation to oxidative protein modification and the appearance of lipid hydroperoxides,11 provides important clues in understanding 03 reactions with other complex biological fluids (such as RTLFs) and cell culture media. Often cells in culture are exposed to 03 without considering the reactions of 03 with constituents of the culture medium. Experimental Design Fresh heparinized human plasma (or other biological fluid or culture medium) is obtained, and aliquots are placed into open plastic Falcon dishes contained within a closed fully humidified chamber which is exposed to a constantly monitored and maintained level of O3 in 5% CO2/ 9 E. M. Postlethwait, S. D. Langford, and A. Bidani, Toxicol. Appl. Pharamcol. 109, 464 (1991). 9a E. M. Postlethwait, S. D. Langford, and A. Bidani, Toxicol. Appl. Pharamcol. 125, 77 (1994). i0 W. A. Pryor, Free Radical Biol. Med. 12, 83 (1992). 11C. E. C r o s s , P. A. Motchnik, B. A. Bruener, D. A. Jones, H. Kauf, B. N. Ames, and B. Halliwell, FEBS Lett. 298, 269 (1992). 12B. Halliwell, Biochem. Pharmacol. 34, 569 (1988). 13B. Halliwell and J. M. C. Gutteridge, Arch. Biochem. Biophys. 2~0, 1 (1990). 14B. Frei and B. N. Ames, in "Molecular Biology of Free Radical Scavenging Systems" (J. Scandalios, ed.), p. 23. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1992. 15C. G. A. Persson, I. Erjefalt, N, Alkner, C. Baumgarden, L. Greiff, B. Gustafsson, A. Luts, U. Pipkorn, F. Sundler, C. Svensson, and P. Wollmer, Clin. Exp. Allergy 21, 17 (1991).
254
ASSAY OF STRESS GENES/PROTEINS
[2 II
95% air at 37°, control plasma being exposed to an identical gas stream without O3.11'16'17 Our system is designed to maintain an ambient 03 concentration at a designated level by maintaining a flow to volume ratio of approximately 10/min, varying the 03 flow to maintain a constant 03 level in the chamber. However, the exposure system can be modified to monitor 03 "uptake" into the exposure solution by precise determinations of both intake and outflow 0 3 concentrations. ~6'17 The overall research strategy is to determine the oxidative changes in the medium of interest in relation to levels of 0 3 , time of 03 exposure, and depletions of antioxidants. Antioxidants that can be measured include those in the aqueous phase [ascorbic acid, uric acid, reduced glutathione (GSH), cysteine] and lipid-soluble antioxidants (ubiquinol, a-tocopherol, and fl-carotene and other carotenoids). Oxidative damage to proteins can be measured by several methods, including the carbonyl assay 18,19 and formation of fluorescent products from tyrosine, 2° including bityrosine. 21 Damage to lipids can be measured by a multiplicity of methods,22 especially useful being the high-performance liquid chromatography (HPLC)-based measurement of hydroperoxides with chemiluminescence detection. 23'24 In addition, products of free radical attack on antioxidants can be measured [e.g., dehydroascorbic acid, oxidized glutathione (GSSG), ubiquinone, allantoin and other degradation products of uric acid, 25 and a-tocopherol quinone]. • Figure 1 illustrates a typical experiment showing the loss of antioxidants and the occurrence of oxidative damage when fresh human plasma is exposed to 03.11.26 There is a rapid depletion of both uric acid and ascorbic acid, important aqueous antioxidants, whereas protein SH thiol groups are lost more slowly. By contrast, there is little change in the 16 D. C. Bolton, B. K. Tarkington, Y. C. Zhee, and J. W. Osebold, Environ. Res. 27, 466 (1982). 17 B. K. Tarkington, T. R. Duvall, and J. A. Last, this volume [22]. 18 R. L. Levine, D. Garland, C. N. Oliver, A. Amici, I. Climent, A. G. Lenz, B. W. Ahn, S. Shaltiel, and E. R. Stadtman, this series, Vol. 186, p. 464. 19 A. Z. Reznick, and L. Packer, this series, Vol. 233 [38]. 2o C. A. O'Neill, A. van der Vliet, M. L. Hu, H. Kaur, C. E. Cross, S. Louie and B. HaUiwell, J. Lab. Clin. Med. 122, 497 (1993). 21 C. Giulivi and K. J. A. Davies, J. Biol. Chem. 268, 8752 (1993). 22 B. Halliwell and S. Chirico, Am. J. Clin. Nutr. 57, 7155 (1993). 23 B. Frei, R. Stocker, and B. N. Ames, Proc. Natl. Acad. Sci. U.S.A. 85, 9748 (1988)• 24 C. E. Cross, T. Forte, R. Stocker, S. Louie, Y. Yamamoto, B. N. Ames, and B. Frei, J. Lab. Clin. Med. 115, 396 (1990). 25 H. Kaur and B. HaUiwell, Chem.-Biol. Interact. 73, 235 (1991). 26 C. E. Cross, A. Z. Reznick, L. Packer, P. A. Davis, Y. J. Suzuki, and B. Halliwell, Free Radical Res. Commun. 15, 347 (1991).
[21]
EVALUATION OF BIOMOLECULAR DAMAGE BY OZONE
255
"
o
".~_~
1 8o
O
~.)
~
60 4o
~e..,
20
•
0
0
1
2 3 Time (hours)
FIG. 1. Depletion of selected antioxidants and protein thiol groups in human plasma exposed to 16 ppm O3 .II Results are expressed as the percentage of antioxidants initially present corrected for loss in air-exposed controls. A, a-Tocopherol; II, protein thiols; ©, ascorbic acid; [2, uric acid.
major lipid-soluble antioxidant a-tocopherol. Unlike what is found with 03 exposures of isolated lipids, little lipid hydroperoxide formation could be demonstrated (< 1 /zM), and no shifts in lipoprotein electrophoretic mobility could be demonstrated. On the other hand, it was readily possible to demonstrate protein oxidation, as revealed by the protein thiol oxidation and by the increase in protein carbonyls (Fig. 2). 1.75 ¢X,
e~o
1.50
o
1.25'
r-.,
1.00" O
(..)
/
0.75 O
r.)
0.50 0
2
4
Time (Hours) FIo. 2. Oxidative damage to human plasma proteins exposed to 16 Plasma + air; II, plasma + ozone.
ppm
0 3 .26 A ,
256
ASSAY OF STRESS GENES/PROTEINS
[21]
Comments Experimental approaches designed to characterize the mechanisms of 03 reactions with constituents of the respiratory tract should ideally include a systematic evaluation of the entire spectrum of 03 reactions occurring in the RTLFs. The described methodology provides a strategy for evaluating the bioreactivity of 03 in an extracellular fluid that has many, but not all, of the constituents present in RTLFs. Extrapolation of our results to the in vivo situation should bear the following in mind: (1) the role of"reactive absorption" at the fluid surface9; (2) the packaging and antioxidant potentials of RTLF mucin 27-29and surfactant and its constituents3°-32; (3) the turnover of RTLF constituents; (4) the reactions of 0 3 to give other reactive oxygen species33,34; (5) the role of transudative/inflammatory processes, including production of reactive oxygen species by activated phagocytic cellslS; (6) the role of metal chelators such as ceruloplasmin, transferrin, and lactoferrin, 13,35known to be present in RTLFs36'37; (7) the role of antioxidant enzymes known to be present in RTLFs and presumably active in vivo, such as catalaseaS; and (8) the cytopathological effects of products of 03 reaction in the RTLFs on the underlying epithelial cells (e.g., such as effects on cytokine and eicosanoid production).
27 C. E. Cross, B. Halliwell, and A. Allen, Lancet 1, 1328 (1984). 28 M. B. Grisham, C. yon Ritter, B. F. Smith, J. T. LaMont, and D. N. Granger, Am. J. Physiol. 253, G93 (1991). 29 H. Hiraishi, A. Terano, S. Ota, H. Mutoh, T. Sugimoto, T. Harada, M. Razandi, and K. J. Ivey, J. Lab. Clin. Med. 121, 570 (1993). 30 S. Matalon, B. A. Holm, R. R. Baker, M. K. Whitfield, and B. A. Freeman, Biochim. Biophys. Acta 1035, 121 (1990). 31 B. Rustow, R. Haupt, P. A. Stevens, and D. Kunze, Am. J. Physiol. 265, L133 (1993). 32 G. E. Hatch, in "Treatise on Pulmonary Toxicology" (R. A. Parent, ed.), Vol. 1, p. 617. CRC Press, Boca Raton, Florida, 1992. 33 W. H. Glaze, Environ. Health Perspect. 69, 151 (1986). 34 j. R. Kanofsky and P. Sima, J. Biol. Chem. 266, 9039 (1991). 35 j. M. C. Gutteridge and B. HaUiwell, in "Atmospheric Oxidation and Antioxidants" (G. Scott, ed.), Vol. 3, pp. 71-99. Elsevier, Amsterdam, 1993. 36 E. R. Pacht and W. B. Davis, J. Appl. Physiol. 64, 2093 (1988). 37 W. B. Davis and E. R. Pacht, in "Lung Injury" (R. G. Crystal and J. B. West, eds.), p. 61. Raven Press, New York, 1992. 38 A. M. Cantin, G. A. Fells, R. C. Hubbard, and R. G. Crystal, J. Clin. Invest. 86, 962 (1990).
[22]
OZONE EXPOSURE OF CULTURED CELLS AND TISSUES
257
[22] O z o n e E x p o s u r e o f C u l t u r e d Cells a n d T i s s u e s By BRIAN K. TARKINGTON, TIMOTHY R. D U V A L L , a n d JEROLD A . LAST
Introduction The effects of oxidant air pollutants such as ozone (03) and nitrogen dioxide (NO2) on the respiratory tract have been studied using in vivo inhalation exposures, ~'2 organ culture or explant systems, 3,4 and cell culture systems. 5'6 With in vitro systems, a major problem is design of the exposure system to deliver constant and reproducible concentrations of ozone to replicate culture dishes, since ozone is so highly reactive chemically. The other challenge is to mimic in vivo exposure conditions where the luminal surfaces of respiratory epithelial cells are exposed almost directly to the inspired air, except for a mucous or surfactant layer of variable thickness. Exposure of organ or cell cultures through a stationary liquid layer is undesirable because relatively insoluble gases have little effect except at high concentrations. 7'8 In addition, the mechanism of action may be different if an oxidant first reacts with the components of a relatively thick liquid layer rather than reacting directly with the cell or its surface layer. 9 For the latter problem, a biphasic cell culture system of respiratory epithelial cells has recently been developed. 10In this biphasic culture, epithelial cells are maintained between air and the liquid medium so that a direct exposure of epithelial cells to ozone is feasible. In vitro systems for exposure of respiratory epithelial cells or explants to ozone have been developed and used successfully by several groups at the Air Pollution Exposure Facility of the California Regional Primate D. W. Wilson, C. G. Plopper, and D. L. Dungworth, Am. J. Pathol. 116, 193 (1984). 2 j. A. Last, in "Air Pollution, the Automobile and Public Health," (A. Y. Watson, R. R. Bates, and D. Kennedy, eds.), p. 415. National Academy Press, Washington, DC, 1988. 3 K. J. Nikula, D. W. Wilson, D. L. Dungworth, and C. G. Plopper, Toxicol. Appl. Pharmacol. 93, 394 (1988). 4 W. C. Eisenberg, K. Tyler, and L. J. Schiff, Experientia 40, 514 (1984). 5 R. E. Rasmussen, J. Toxicol. Environ. Health 13, 397 (1984). 6 D. C. Bolton, B. K. Tarkington, Y. C. Zee, and J. W. Osebold, Environ. Res. 27, 466 (1982). 7 D. M. Pace, P. A. LandoR, and B. T. Aftonomes, Arch. Environ. Health 18, 165 (1969). s W. L. Hagar, W. E. Sweet, and F. Sweet, J. Air Pollut. Control Assoc. Notebook 31, 993 (1981). 9 D. G. Wenzel and D. L. Morgan, Drug Chem. Toxicol. 5, 201 (1982). i0 j. M. Whitcutt, K. B. Adler, and R. Wu, In Vitro Cell. Dev. Biol. 24, 420 (1988).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
258
ASSAY OF STRESS GENES/PROTEINS
i
Oxyoe. ~v, I
A
B c,
-
H
r- I N
~
v6 D
[22]
~
II I
~
r~
N
L
L Carbon dioxide I FIG. 1. Schematic diagram of small in vitro exposure system. A, Ozonizer; B, ozone bypass valve; C, incubator; D, humidificationbottles; E, exposure vessel; F, control vessel; G, pressure gauges; H, ozone analyzer; I, thermometer; J, heating tape; K, thermometer; L, thermal insulation. (From Tarkington et al?2).
Research Center (University of California, Davis). 3,11-13 The systems are designed to generate and monitor consistent, reproducible levels of ozone o v e r a range of concentrations in a humidified atmosphere, allowing exposure times greater than those previously used by others .4,5 Several replicate culture vials or plates can be equally exposed to the experimental atmosphere, and the epithelial cells are exposed directly with only a thin aqueous film o v e r the epithelial surface. Control cells or explants may be simultaneously exposed to the experimental atmosphere without ozone.
Materials and Methods Small in Vitro Exposure System The atmospheres used are filtered air with 5% by volume carbon dioxide added, all saturated with water vapor and held at 37.5 °. A schematic diagram of the system is shown in Fig. 1. One vessel for ozone exposure and one control vessel are used. Most of the components of the system H K. J. Nikula and D. W. Wilson, Fundam. Appl. Toxicol. 15, 121 (1990). lz B. K. Tarkington, R. Wu, W. Sun, K. J. Nikula, D. W. Wilson, and J. A. Last, Toxicology 88, 112 (1994). 53 j. M. Cheek, A. R. Buckpitt, C. Li, B. K. Tarkington, and C. G. Plopper, Toxicol. Appl. Pharmacol. 125, 59 (1994).
[22]
OZONE EXPOSURE OF C U L T U R E D CELLS A N D TISSUES
259
with which the ozone comes in contact are made from Teflon. A few items are glass or type 316 stainless steel. To ensure very stable flows, three stages of pressure regulation are used on all gases. Oxygen, air, or carbon dioxide are introduced to the associated control valves at a constant pressure of 0.21 kg/cm 2. Valve V1 and rotameter R1 are used to deliver a flow of medical grade oxygen of 5 liters/min through the electrical discharge ozonizer (Model IV, Erwin Sander Elektroapparatebau G.m.b.H., Uetze-Eltze, Germany). This flow rate is used to provide adequate cooling of the ozonizer. With the unit set to maximum output, as much as 700 ppm ozone can be produced in the oxygen. After closure of the ozone bypass valve slightly, valve V2 and rotameter R2 are used to meter 50, 100, or 150 ml/min of the ozone in oxygen stream into the exposure portion of the system. Flow rate selection is dependent on the maximum ozone concentration desired. Finer adjustments to ozone production are electrical. Flow rates of 2.8, 2.75, or 2.7 liters/min of filtered air are added through valve V3 and rotameter R3. Dry 99.8% carbon dioxide is introduced through valve V4 and rotameter R4. The carbon dioxide flow rate is 150 ml/min. By adjusting the total flow rate of the air and oxygen streams to 2.85 liters/min, a 5% by volume carbon dioxide atmosphere is maintained in a total flow rate of 3 liters/min through the exposure vessel. This provides excess flow for the ozone analyzer (Model 1003-AH, Dasibi Environmental Corporation, Glendale, CA), which samples at 2 liters/rain and prevents recirculation of the sampled gas. Except for the lack of provisions for ozone introduction, the atmosphere stream for the control vessel is produced in the same way with the air flow rate of 2.85 liters/min controlled by valve V5 and rotameter R5, and the carbon dioxide flow rate of 150 ml/min controlled by valve V6 and rotameter R6. Lines 6.4 mm (1/4 inch) in diameter pass through a port in the incubator and convey the gas streams to the exposure vessel and control vessel. The gas streams are humidified by bubbling through 1000-ml bottles containing sterile distilled water. Production of complete saturation necessitates slight heating on the outside of the bottles. Heating mats with input controlled by ac power supplies are used, and the outside of the bottles are wrapped with thermal insulation. The humidified gases are then conveyed to 320ml (or l-liter)jars (Savillex Corporation, Minnetonka, MN) that serve as the exposure and control vessels. The jars can be mounted on a rocking platform if needed. The platform is used for tracheal explants so that they are regularly bathed with culture medium. Each vessel (Fig. 2) can contain five glass culture vials. Inside each vessel the gases enter at the top through a jet oriented so that the atmosphere is injected tangentially to the wall and swirls across the tops of the culture vials. This is to promote mixing
260
ASSAY OF STRESS GENES/PROTEINS EXHAUST ATMOSPHERE
III
[22]
INLET ATMOSPHERE
--
GLASS VIAL CONTAINING MEDIUM AND TRACHEA
FIG. 2. Side view of exposure vessell for small in vitro exposure system. Details are given in the text.
and even exposure among each of the five vials. Exhaust from each vessel is taken from the bottom in the center well below the tops of the vials. Humidification of the atmospheres is adjusted so that the 2 ml of medium contained in each culture vial neither loses nor gains volume during a 24hr period. Exhaust lines 9.5 mm (3/8 inch) in diameter exit from the vessels. Near each vessel a T is provided to connect a pressure line and a remotely mounted gauge so that the pressure in each vessel is monitored. The vessels are operated at a pressure of 1.5 cm (water gauge) above that of the laboratory to prevent any influx of biological contaminants in the event of a leak. Changes in pressure are also used to monitor the system for leaks. The exposure vessel exhaust line passes through a port in the incubator and connects to the ozone analyzer inlet and exhaust so that 2 liters/min of this atmosphere is circulated through and sampled by the analyzer. Because the gases exiting the incubator are saturated with moisture and at 37.5 ° , all exhaust lines, pressure lines, sample lines, gauges, and valves are wrapped with heating tape and thermal insulation. Electrical power to the heating tape is controlled so that the humid gases are held at about 40 ° until they are vented to a fume hood. This prevents the considerable condensation in the system that would occur if the gases were allowed to cool to the laboratory temperature of 22°. In addition, condensation on the optics of the ozone analyzer is avoided by partially blocking off the cooling air intake to increase the internal temperature of the instrument from the normal 34° to about 42° to 43 °. This is below the maximum recommended ambient operating temperature of 45 °. The internal temperature of the analzyer is monitored with a remote thermistor probe thermometer affixed to the absorption chamber near the sample exit.
[22]
OZONE EXPOSURE OF CULTURED CELLS AND TISSUES
261
Large in Vitro Exposure System A larger scale in vitro exposure system enables the simultaneous exposure of cells to three ozone or nitrogen dioxide levels with a filtered air control. Pollutant levels in the three exposure vessels are independently adjustable. This is the preferred system for performing dose-response studies. Also, more material can be exposed in each vessel. The arrangement of components is similar to that used for the small in vitro exposure system except for the addition of two more flow channels for exposure vessels, larger tubing diameters to accommodate higher flow rates, and solenoid valves to control which vessel is sampled by the ozone analyzer. Standard 86 by 127 mm culture plates, available in a wide variety of well configurations, containing the cells on inserts, are exposed to ozone, nitrogen dioxide, or filtered air in specially designed cylindrical glass vessels 3.66 liters in volume (Fig. 3). Atmospheres saturated with water vapor at 37.5 ° and containing 95% air and 5% carbon dioxide by volume flow through each vessel at a total rate of 15 liters/rain. Incorporated in the lid of each vessel is a tangential mixing dome followed by a diffuser plate with 19 symmetrically located holes, each 1.6 mm in diameter, to distribute the flow evenly. Exhaust is taken from a central point below a perforated desiccator plate on which the culture plate is placed. Vessel
FIG. 3. Exposure vessel for large in vitro exposure system. Details are given in the text.
262
ASSAY OF STRESS GENES/PROTEINS
[22]
geometry and flow patterns are critical in assuring that a homogeneous pollutant concentration is maintained within and that each culture receives the same exposure.
Ozone Monitoring Ozone concentrations in each vessel are continuously monitored during exposures with an ultraviolet ozone analyzer. In the large system with three exposure vessels and a control vessel, the analyzer is connected to computer-controlled valves that cycle from vessel to vessel. Typically, each is sampled for 2 min during a total cycle of 8 min. All materials in contact with the sampled ozone are Teflon or glass. Concentration data from the ozone analyzer are collected with a computer-based data acquisition system (PC-AT, IBM Corporation, Boca Raton, FL, and System 4000, ADAC Corporation, Woburn, MA) that is also used to produce statistical reports on the exposure conditions. Calibration of the analyzer is performed according to the national reference method 14and is traceable to a National Institute of Standards and Technology absolute ozone photometer. It is necessary to prevent condensation from the atmospheres saturated with moisture at 37.5 ° on the optics of the analyzer by raising the internal temperature to about 42° to 43°. Since the response of the instrument to a given ozone concentration is inversely proportional to the absolute temperature of the sample, ~5 calibrations are performed at the elevated temperature to eliminate this temperature difference as a small source of error. During exposure, stable ozone concentrations are easily maintained with few adjustments. Table I shows a portion of the computerized data report for a typical exposure. The desired concentration of 0.5 ppm in this case was achieved with a mean level for the 24-hr period of 0.49 ppm and a standard deviation of about 2% of the mean. High peak excursions in concentration, which could in theory influence any biological effects independently of the mean level, did not occur.
Testing Uniformity of Exposure To ensure that each of the culture vials in the exposure apparatus receive the same amount of ozone, 4 ml of 2% neutral buffered potassium iodide solution ~6is pipetted into each. In a typical series of tests, the vial 14 U.S. Code of Federal Regulations, Prot. Environ. 40, 667 (1988). 15 W. B. DeMore, J. C. Romanovsky, M. Feldstein, W. J. Haming, and P. K. Mueller, in "Calibration in Air Monitoring," p. 131. Am. Soc. Test. Mater., Philadelphia, 1976. 16 B. E. Saltzman, in "Selected Methods for the Measurement of Air Pollutants" (M. Storlazzi and S. Hochheiser, eds.), p. D-1. Public Health Service, Durham, North Carolina, 1965.
[22]
OZONE EXPOSURE OF C U L T U R E D CELLS A N D TISSUES
7 O N
0
0
z
0
<
.N
Z < N
<
~
[-..
o
~
~
263
264
ASSAY OF STRESS GENES/PROTEINS
[22]
T A B L E II EXPOSURE OF CULTURE VIALS TO VARIOUS CONCENTRATIONS OF OZONE a
Ozone concentration (ppm)
Duration of exposure
Coefficient of variation (% of absorbance)
Nominal
Actual
Hr
Min
Absorbance (at 352 nm)
0.10 0.10 0.10
0.091 + 0.007 0.091 -+ 0.003 0.084 +- 0.008
1 1 1
40 40 40
0.285 - 0.016 0.303 - 0.028 0.315 +- 0.020
5.6 9.2 6.4
0.25 0.25 0.25
0.244 -+ 0.025 0.256 +- 0.017 0.245 - 0.011
0 0 0
40 40 40
0.599 --- 0.071 0.526 --- 0.027 0.458 --_ 0.018
11.8 5.1 3.9
0.50
0.498 -+ 0.019
0
50
0.702 -+ 0.016
2.3
1.0 1.0 1.0
0.991 +- 0.082 0.984 -+ 0.098 0.978 +- 0.117
0 0 0
10 10 10
0.784 - 0.032 0.724 +_ 0.021 0.706 -+ 0.023
4.1 2.9 3.3
a To groups of four replicate culture vials 35 mm (diameter) by 10 mm deep was added 4 ml of 2% neutral buffered KI. Results are presented as mean values -+ i SD for
absorbance readings after the indicated exposure to ozone. Each trial was performed independently of the others, with freshly prepared batches of KI. The experiment with 0.50 ppm of ozone used five replicate vials in the 320-ml jars, whereas the other experiments used the apparatus with l-liter jars. (From Tarkington et al.~2)
contents are exposed to 0.1 to 1.0 ppm of ozone for up to 100 min, and the resulting color development is measured with a spectrophotometer set at 352 nm. The mean absorbance, for example, of five vials exposed to 0.5 ppm of ozone for 50 min was 0.702 with a standard deviation of 0.016 (coefficient of variation less than 5%). The test is repeated several times with various concentrations of ozone for various durations of exposure. Table II shows typical results, and several points should be noted. First, the coefficient of variation observed in the groups of four or five replicate vials ranged from about 2.3 to 11.8% in 10 experiments. The mean (--- 1 SD) variance observed was 5.5 --- 3.0%. Not surprisingly, there was relatively greater variation in the data from vials exposed to the lower concentrations of ozone, presumably reflecting, at least in part, the limitations of the neutral buffered KI method. An improved boric acidbuffered KI method 17 is said to have better reproducibility and more accurately defined stoichiometry. Second, the apparatus was able to de17 D. L. Flamm, Environ. Sci. Technol. 1L 978 (1977).
[22]
OZONE EXPOSURE OF CULTURED CELLS AND TISSUES
265
liver, accurately and reproducibly, constant amounts of ozone to several replicate vials in the chambers over a 10-fold range of ozone concentraitons and exposure durations. Multiple exposures at the same concentration, even when performed on different days with different batches of reagent, gave very similar results. Third, if the results of the two most variable trials are eliminiated from the data analysis, we find that for eight exposures to ozone at 0.1 to 1.0 ppm the mean variance was 4.2 + 1.4%, indicating a very reproducible procedure with little difference in the observed variance from group to group despite the large differences in ozone concentration and exposure duration. Conclusion These exposure systems allow the generation and monitoring of consistent, reproducible levels of ozone over a range of concentrations from 0.1 to 3 ppm or greater, and they allow the investigator to expose equally preparations of cultured cells or tracheal explants to ozone for 24 hr or longer. There are no problems encountered in maintaining the exposure atmospheres at these concentrations and durations. Ozone is generated from pure oxygen to avoid contamination with nitrogen oxides as occurs at low concentrations when air is the substrate for production of ozone. Ozone does not react with the materials used for construction. There are a large number of potential uses for these in vitro exposure systems in answering important unresolved questions with regard to effects of ozone on the respiratory system. The systems allow exposure of significant quantities of material that can be harvested for biochemical assays and for microscopic analysis. The exposure systems are also flexible, allowing the addition of other agents to either the test atmosphere or the medium. These systems should be easily fabricated and allow other laboratories to perform similar exposures relatively inexpensively and conveniently. Acknowledgments This work was supported by research grants from the U.S. National Institutes of Health, ES-00628, ES-05707, and RR-00169. Helpful comments of Dr. C. E. Cross, technical assistance of R. R. Proulx, artwork by C. F. Sarason, and glassware by T. Adams are gratefully acknowledged.
[23]
DETERMINATION
OF ANTIOXIDANTS
IN HUMAN
PLASMA
269
[23] M e a s u r e m e n t o f A n t i o x i d a n t s in H u m a n B l o o d P l a s m a By PAUL A. MOTCHNIK, BALZ FREI, and BRUCE N. AMES Introduction Increased antioxidant intake has been associated with reduced incidence of certain human diseases such as atherosclerosis and cancer.l-3 Free radicals are believed to be involved in the etiology of these diseases. 4 Consequently, there is an interest in the mechanisms of antioxidant protection against free radical-induced injury. Human blood plasma is often used as a model for studying free radical-induced damage 5-7 because it contains critical targets of oxidative damage, such as lipoproteins, includes many important antioxidants, and represents a physiologically relevant milieu. In addition, plasma antioxidant levels are important in assessing the nutritional antioxidant status of an individual. To study the relationship between diet, antioxidants, and the involvement of free radicals in human diseases, reliable methods of measuring plasma antioxidants are required. A number of analytical techniques have been developed to measure antioxidants, and a survey of these methods is available. 8 The antioxidants in plasma can be classifted into two groups: the watersoluble antioxidants of the aqueous phase of plasma, and the lipid-soluble antioxidants associated with lipoproteins. The water-soluble antioxidants include ascorbic acid (AA), or vitamin C, uric acid (UA), protein thiols, and bilirubin. Plasma also contains very low levels of glutathione, which is a major antioxidant in cells. The lipid-soluble antioxidants comprise a- and y-tocopherol, ubiquinol, lycopene, /3-carotene, and some other carotenoids and oxycarotenoids. In this chapter we describe assays for National Regional Council, "Diet and Health: Implications for Reducing Chronic Disease Risk." National Academy Press, Washington, D.C., 1989. 2 E. B. Rimm, M. J. Stampfer, A. Ascherio, E. Giovannucci, G. A. Colditz, and W. C. Willet, N. Engl. J. Med. 328, 20 (1993). 3 G. Block, B. Patterson, and A. Subar, Nutr. Cancer 18, 1 (1992). 4 B. Halliwell and J. M. C. Gutteridge, this series, Vol. 186, p. 1. 5 D. D. M. Wayner, G. W. Burton, K. U. Ingold, and S. Locke, FEBS Lett. 187, 33 (1985). 6 D. D. M. Wayner, G. W. Burton, K. U. Ingold, L. R. C. Barclay, and S. J. Locke, Biochim. Biophys. Acta 924, 408 (1987). 7 B. Frei, R. Stocker, and B. N. Ames, Proc. Natl. Acad. Sci. U.S.A. 85, 9748 (1988). 8 j. K. Lang, M. Schillaci, and B. Irvin, in "Modern Chromatographic Analysis of Vitamins" (A. P. De Leenheer, W. E. Lambert, and H. J. Neils, eds.), p. 153. Dekker, New York, 1992.
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
270
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[23]
measuring plasma antioxidants and some of the problems associated with each assay.
Quantitation of Water-Soluble Antioxidants Ascorbic Acid and Uric Acid Uric acid and ascorbic acid are important biological antioxidants capable of scavenging a wide variety of different oxidants. 9'1° In human blood plasma the concentrations of UA and AA are about 160-450 /zM and 30-150/~M, respectively. 10A variety of assays for AA combine colorimetric reactions with the oxidation of AA to dehydroascorbic acid or the reverse reduction reaction. 11 However, these methods lack sensitivity and specificity, or are time-consuming. A number of new, sensitive, and specific assays have been developed that use paired-ion, reversed-phase high-performance liquid chromatography with electrochemical detection (HPLC-EC) to measure UA and AA. The method described here is a modification the H P L C - E C assays of Behrens and Madere 12and Kutnink et al. 13 Sample Preparation and Stability. Ascorbic acid in blood plasma from normal individuals is present almost exclusively in the reduced form. ~4A primary concern in the measurement of AA in plasma is the artifactual formation of dehydroascorbic acid, the two-electron oxidation product of AA. Although relatively stable in whole b l o o d , 15 AA will slowly oxidize in plasma and serum.14 Dehydroascorbic acid can be reduced back to AA with reducing agents such as dithiothreitol (DTT) or 2,3-dimercapto-1propanol. However, dehydroascorbic acid can also undergo spontaneous hydrolysis to form 2,3-diketogulonic acid, a step that is essentially irreversible. The potential for AA oxidation requires that the samples be stabilized with acid or analyzed within several hours after collection. Unlike AA, UA is quite stable in plasma, diluted samples, and standard solutions.13 All standards and samples are kept on ice or at 4° during preparation. Standards are prepared from frozen ( - 7 0 °) stock solutions of 250/zM AA 9 A. Bendich, L. J. Machlin, and O. Scandura, Adv. Free RadicalBiol. Med. 2, 419 (1986). 10 R. Stocker and B. Frei, in "Endogenous Antioxidant Defences in Human Blood Plasma" (H. Sies, ed.), p. 213. Academic Press, London, 1991. i1 p. W. Washko, W. O. HartzeU, and M. Levine, Anal. Biochern. 181, 276 (1989). 12 W. A. Behrens and R. Madere, Anal. Biochem. 165, 102 (1987). 13 M. A. Kutnink W. C. Hawkes, E. E. Schaus, and S. T. Omaye, Anal. Biochem. 166, 424 (1987). i4 K. R. Dhariwal, W. O. Hartzell, and M. Levine, Am. J. Clin. Nutr. 54, 712 (1991). 15 W. Lee, K. A. Davis, R. L. Rettmer, and R. F. Labb6, Am. J. Clin. Nutr. 48, 286 (1988).
[23]
D E T E R M I N A T I O N O F A N T I O X 1 D A N T S IN H U M A N P L A S M A
271
in 5% (w/v) metaphosphoric acid (MPA) containing 0.54 mM Na2EDTA. All MPA solutions should be freshly prepared. Uric acid standards (1-10 mM) are prepared in 25 mM potassium phosphate, pH 7.4, 0.9% (w/v) NaCI (phosphate-buffered saline, PBS). The UA solution may need to be warmed to 37° to completely dissolve the UA. This solution is then diluted with 5% MPA, 0.54 mM NazEDTA, to a final concentration of 250/xM. The 250/xM stock solutions of AA and UA can be stored at -70 ° for several months. Stock solutions are diluted with mobile phase (see below) to produce working standards containing 2.5/xM AA or UA. To generate standard curves 5-25/xl of the standard solutions are injected. Plasma (25-50/zl) is mixed in an Eppendorf tube with an equal volume of 10% (w/v) MPA containing 0.54 mM NazEDTA, agitated in a vortex mixer, and centrifuged at full speed for 3 min to pellet the precipitated plasma proteins. The supernatant is removed and can be stored at -70 ° until analyzed. To measure AA and UA, the samples are diluted 50-fold with mobile phase immediately prior to injection of 25/xl. Total AA (AA plus dehydroascorbic acid) is determined by diluting the MPA-stabilized samples 50-fold with mobile phase containing 1.0 mM cysteine and incubating for 5 min at 4 °. 12,13Dehydroascorbic acid concentrations are calculated from the difference between AA and total AA. Chromatography. For the separation of UA and AA, a Waters Associates (Milford, MA) Model 510 solvent delivery system equipped with an SSI (State College, PA) pulse damper, a Rheodyne (Cotati, CA) 7125 injector equipped with a 200-/.d sample loop, and 3/zm Supelcosil (Supelco, Bellefonte, PA) LC-18-DB guard (1.5 cm × 4.6 mm) and analytical (7.5 cm × 4.6 mm) columns are used. The mobile phase consists of filtered and degassed 40 mM sodium acetate buffer, pH 4.75, 0.54 mM NaEEDTA, and 1.5 mM dodecyltriethylammonium phosphate (Regis Chemical Co., Morton Grove, IL) used as the ion-pairing agent. The separations are performed at a flow rate of 1.5 ml/min with a back pressure of approximately l l00 psi. Both the column and the detector require 2-3 hr for equilibration. Columns should be washed routinely in methanol and water (50: 50, v/v). The detection system consists of an LC-4B amperometric detector with a glassy-carbon working electrode and an AglAgC1 reference electrode. The applied potential is set at +0.5 V with a sensitivity setting of 50 nA. The data from HPLC analysis are digitized by a Nelson 760 analytical interface (Cupertino, CA) and processed by Perkin-Elmer/Nelson Analytical Turbochrom data acquisition software on an IBM PS/2 Model 70 computer. As shown in Fig. 1, analysis of freshly prepared human plasma following HPLC separation produces two prominent peaks which correspond
272
[23]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
I
UA 5nA
cO t~9
nO
I..IJ
AA
I
I
1
I
I
2
I
I
I
3
Time (min)
FIG. 1. Electrochemical chromatogram of freshly prepared human blood plasma following ion-pairing reversed-phase HPLC. The plasma was diluted with an equal volume of 10% (w/v) MPA containing 0.54 mM Na2EDTA, centrifuged, and the supernatant diluted 1/50 with mobile phase. The injection volume was 25/zl. Peaks indicated are uric acid (UA) and ascorbic acid (AA). Other details are as described in the text.
to UA (retention time 1.9 min) and AA (2.9 min). The slight transient loss in current at 1.6 min is produced by the added MPA. Chromatograms of samples treated with cysteine in order to reduce dehydroascorbic acid to AA are similar except that an off-scale peak eluting near the column void is present because of the electrochemical oxidation of cysteine (chromatogram not shown). Standard peak areas are used to construct calibration curves in the range of 10-65 pmol for both UA and AA. The amount of AA and UA in samples is determined using the standard curves. The standard calibration curve is linear over a range of 0.6 to 80 pmol for UA and 1.0 to 150 pmol for AA. The detection limit for UA and AA is 0.6 and 1.0 pmol, respectively. The coefficient of variation is 2%. Loss of resolution, peak tailing, or broadening of peaks can usually be corrected by washing the column with methanol and water (50:50, v/v). If this treatment is not successful the guard column may be changed. Increases in baseline noise indicate a dirty electrode or contaminants in the mobile phase.
[23]
DETERMINATION OF ANTIOXIDANTS IN HUMAN PLASMA
273
Bilirubin
Bilirubin has been proposed to have biological antioxidant activity of potential importance.16 Under physiologically relevant conditions, bilirubin can scavenge peroxyl radicals with high efficiency. 17Bilirubin in human blood plasma is present mainly bound to albumin at concentrations ranging from about 5 to 20/xM, 1°and it can be measured using HPLC with spectrophotometric detection at 460 nm as described by McDonagh et al. 18 An aliquot (50-250 t~l) of plasma is added to 4 volumes of ice-cold methanol in an Eppendorf tube and mixed in a vortex mixer. The precipitated protein is pelleted by centrifugation, and 50-100/zl of the supernatant is analyzed by HPLC. The HPLC system consists of a Waters Associates solvent delivery system equipped with a Rheodyne injector as described above for AA and UA except that no pulse damper is necessary. A 3/zm LC-18-DB column (15 cm × 4.6 ram) is used with a mobile phase consisting of 0.1 M dioctylamine (Aldrich, Milwaukee, WI) in methanol and water (96:4, v/v) at a flow rate of 1 ml/min. Optical detection utilizes a Kratos (Westwood, NJ) Model 773 UV detector set to 460 nm. The retention time for bilirubin is 11.8 min (chromatogram not shown). Bilirubin (Sigma, St. Louis, MO) is used as a standard without further purification and the area of the peak used for quantitation. The limit of detection is 30 pmol. The linear range is 30-1200 pmol and the coefficient of variation 3%. Protein Thiols In plasma most of the protein-associated thiol groups are found on albumin. 19 Because albumin is present in very high concentrations in plasma, and has a half-life of only about 20 days, oxidation of its thiol groups may have no significant biological consequences. 10Therefore, albumin has been suggested to act as a sacrifical antioxidant. 2° Plasma usually contains 400-600/zM protein thiols, l0 Protein thiols can be measured with a spectrophotometric method using dithionitrobenzene (DTNB). 21 ~6R. Stocker, Y. Yamamoto, A. F. McDonagh, A. N. Glazer, and B. N. Ames, Science 235, 1043 (1987). i7 R. Stocker, A. N. Glazer, and B. N. Ames, Proc. Natl. Acad. Sci. U.S.A. 84, 5918 (1987). 18 A. F. McDonagh, L. A. Palma, F. R. Trull, and D. A. Lightner, J. Am. Chem. Soc. 104, 6865 (1982). i9 R. Radi, K. M. Bush, T. P. Cosgrove, and B. A. Freeman, Arch. Biochem. Biophys. 286, 117 (1991). 20 B. Halliwell, Biochem. Pharmacol. 37, 569 (1988). 21 G. L. Ellman, Arch. Biochem. Biophys. 82, 70 (1959).
274
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[23]
The following solutions should be warmed to 37° and mixed in an Eppendorf tube: 900/zl of 0.2 M Na2HPO4 containing 2 mM Na2EDTA, 100/zl plasma, and 20/zl of 10 mM DTNB in 0.2 M Na2HPO4. The solution is mixed in a vortex mixer and transferred to a cuvette, and the absorbance is measured at 412 nm against air. The reaction takes about 5 min to go to completion, during which there is an increase in the absorbance. The maximum value is recorded. After reaching a maximum, the absorbance may decrease at a slow and steady rate. A sample blank should be measured because bilirubin, fl-carotene, and other plasma constituents that absorb at 412 nm can interfere with protein thiol measurement. For the sample blank, the absorbance of a mixture of 920/zl of 0.2 M Na2HPO4 containing 2 mM Na2EDTA and 100/zl plasma is measured against air. The contribution of the reagents alone to the absorbance at 412 nm also has to be determined. The reagent blank consists of 900 /xl of 0.2 M Na2HPO4 containing 2 mM Na2EDTa, 100/zl PBS, and 20/xl of 10 mM DTNB in 0.2 M Na2HPO4. The absorbance for sample and reagent blanks are subtracted from plasma absorbance values to obtain the corrected values. A calibration curve is produced using glutathione dissolved in PBS. The protein thiol concentration in plasma is determined from the standard curve using the corrected absorbance values for plasma. The detection limit is 3.5 nmol, and the linear range is 3.5-120 nmol (35-1200 /zM in plasma). The coefficient of variation is 2.1%. Quantification of Lipid-Soluble Antioxidants: a-Tocopherol, 7-Tocopherol, Lycopene, fl-Carotene, Ubiquinol, and Ubiquinone The lipid-soluble antioxidants present in lipoproteins of human blood plasma are a- and 7-tocopherol (15-40/xM), lycopene (0.5-1.0 txM), flcarotene (0.3-0.6/xM), and ubiquinol 10 (0.4-1.0/xM), the reduced form of coenzyme Q10. A number of analytical techniques that use HPLC with electrochemical detection have been developed for the quantitative determination of these lipid-soluble antioxidants. A single-step analysis for the simultaneous measurement of ubiquinol and ubiquinone (the twoelectron oxidation product of ubiquinol) was established using HPLC with both UV (for ubiquinone) and electrochemical (for ubiquinol) detection32 This method, which was modified by Lang e t al. 23 to include tocopherol determination, is useful for tissues that contain relatively high concentrations of ubiquinone, but in plasma the sensitivity of UV detection is 22 S. Ikenoya, M. Takada, T. Yuzuriha, K. Abe, and K. Katayama, Chem. Pharm. Bull. 29, 158 (1981). 23 j. K. Lang, K. Gohil, and L. Packer, Anal. Biochem. 157, 106 (1986).
[23]
DETERMINATION OF ANTIOXIDANTS IN HUMAN PLASMA
275
inadequate for ubiquinone detection. To measure total ubiquinone in plasma (ubiquinol plus ubiquinone) using HPLC with electrochemical detection, the samples can be analyzed before and after treatment with sodium borohydride, which reduces ubiquinone to ubiquinol. 24 The content of ubiquinone is then determined as the difference between the ubiquinol in untreated samples (which measures ubiquinol) and reduced samples (which measures total ubiquinone). The disadvantage of this method is that two extractions and two HPLC analyses are required. The use of a single electrochemical detector for ubiquinol, ubiquinone, and tocopherol was developed by Edlund. 25 In this system the reversible oxidation-reduction of ubiquinones was exploited by using a series of inline, coulometric electrochemical cells to achieve postcolumn oxidation-reduction-oxidation. In this way, only reversible redox couples were detected at the last electrode. We have modified this method to electrochemically reduce ubiquinone 10 to ubiquinol 10, and to measure plasma a- and y-tocopherol, lycopene, fl-carotene, ubiquinol 10, and ubiquinone 10 (as the quinol) by oxidation in a single chromatographic step. Preparation of Standards. a-Tocopherol (E ~%= 75.8 liter mol-I cm-1; 292 nm), lycopene (E 1% = 3450; 472 nm), /3-carotene (E 1% = 2620; 453 nm), and ubiquinone 10 (E 1~ = 165; 275 nm) are dissolved individually in chloroform and diluted in ethanol to a final concentration of about 1/~M, and the concentrations are determined by the absorbances using the above extinction coefficients. Ubiquinol 10 is freshly prepared from the corresponding ubiquinone homolog by adding I0/zl of a 0.25% (w/v) solution of sodium borohydride in methanol to 1 ml of a 100 /xM ubiquinone solution. The solution is vortexed and incubated for 30 min on ice in the dark. After the incubation 1 ml of water is added, the solution is vortexed, and 4 ml of hexane is added to extract the ubiquinol. The mixture is briefly centrifuged and the hexane layer removed. The extraction is repeated and the hexane extracts pooled and dried under a stream of nitrogen. The resulting residue is dissolved in ethanol, and the ubiquinol concentration is determined spectrophotometrically using an extinction coefficient (E 1%) of 46.4 liter tool -l cm -~ at 290 nm. 26 Sample Preparation. All standards and samples are kept on ice or at 4° during preparation. Plasma (0.25 ml) is added to 1 ml of ethanol and mixed briefly, and 5 ml of hexane is added. The solution is mixed on a vortex mixer for 3 min, centrifuged for 15 min at 1000 g, and the upper 24 T. Okamoto, Y. Fukunaga, Y. Ida, and T. Kishi, J. Chromatogr. 430, 11 (1988). 25 p. O. Edlund, J. Chromatogr. 425, 87 (1988). 26 y. Hatefi, in "Coenzyme Q (Ubiquinone)" (F. F. Nord, ed.), p. 275. Wiley, New York, 1963.
276
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[23]
hexane layer transferred to a glass tube. Another 5 ml of hexane is added and the extraction procedure repeated. The hexane extracts are pooled and dried at room temperature under a stream of nitrogen, and the resulting residue is dissolved in either 0.5 ml of methanol/ethanol (1 : 1, v/v) (for Method 1) or methanol/ethanol/2-propanol (22.5:73.6:3.9, v/v/v) (for Method 2). Samples should be dried and dissolved just prior to analysis. The dried extracts should be mixed vigorously with the appropriate solvent to ensure that the extracted material is completely dissolved. Brief sonication (30 sec) may be used to dissolve the sample residue completely. Chromatography. The HPLC system consists of a Waters Associates solvent delivery system equipped with a Rheodyne injector as described above for AA and UA. Two HPLC methods are available for the simultaneous measurement of a- and y-tocopherol, lycopene,/3-carotene ubiquinol, and ubiquinone, Method I uses electrochemical detection for a-tocopherol, lycopene, /3-carotene, ubiquinol, and ubiquinone. This is accomplished by on-line, postcolumn electrochemical reduction of ubiquinone to ubiquinol, followed by detection of all antioxidants by oxidation. Method 2 uses electrochemical detection for a-tocopherol, carotenoids, and ubiquinol measurement. To measure ubiquinone using Method 2, the sample extract is analyzed before and after being treated with sodium borohydride to reduce ubiquinone to ubiquinol. The amount of ubiquinone is obtained by the difference between total and reduced ubiquinol. Method 1. Three micron particle size Supelcosil (Supelco) LC-8-DB guard (1.5 cm x 4.6 mm) and analytical (15 cm x 4.6 mm) columns are used. The mobile phase consists of 20/zM lithium perchlorate in methanol and water (96 : 4, v/v) and is delivered at a flow rate of 2 ml/min. Electrochemical detection is accomplished by using an ESA (Bedford, MA) Model 5100 Coulochem detector equipped with a 5021 conditioning cell and a 5011 analytical cell. The conditioning cell is placed between the column outlet and the analytical cell. The conditioning cell potential is set at -0.55 V (for reduction). The potentials for the analytical cell are set at +0.01 and +0.35 V for electrodes 1 and 2, respectively. The signal produced by electrode 2 is used for detection of antioxidants. This configuration allows for separation of ubiquinone from ubiquinol on the column, followed by electrochemical reduction of ubiquinone to ubiquinol and subsequent detection by oxidation at electrode 2. In this way all antioxidants are determined simultaneously with a single electrode. As shown in Fig. 2A, analysis of lipid extracts of human blood plasma by this method produces several peaks which coelute with authentic standards of y-tocopherol (retention time 3.4 min), a-tocopherol (3.7 min), lycopene (8.2 min), a-carotene (10.2 min),/3-carotene (11.3 min), ubiquinol 10 (20.8 min), and ubiquinone 10 (34.4 rain). This method has an advantage
A (Z-TC
t
5nA
)-C 0 0~ tO
UB-ol
0
nO W
(z-C UB-one
LP I
!
5
10
20
15
25
30
35
Time (min)
B I
10nA
0~-TC
LP
(~ o] c O El., (/) O
UB-ol
nO W
~-TC
i
I
I
I
I
i
i
i
i
5
i
10
I
I
1
i
15
Time (min) FIG. 2. Electrochemical chromatogram of lipid extracts of human blood plasma following reversed-phase HPLC. (A) An LC-8-DB column was used with a mobile phase consisting of 20 mM lithium perchlorate in methanol and water (96 : 4, v/v) at a flow rate of 2 ml/min. The plasma was extracted with hexane and dissolved in methanol/ethanol (1 : 1, v/v). (B) An LC-18-DB column was used with a mobile phase consisting of 20 mM lithium perchlorate in methanol/ethanol/2-propanol (22.5 : 73.6 : 3.9, v/v/v) at a flow rate of 1 ml/min. The plasma was extracted with hexane and dissolved in mobile phase. Peaks indicated are ~/-tocopherol (7-TC), a-tocopherol (a-TC), lycopene (LP), a-carotene (a-C),/3-carotene (/3-C), ubiquinol 10 (UB-ol), and ubiquinone 10 (UB-one). Other details are as described in the text.
278
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[23]
over Method 2 in that it gives improved resolution of plasma antioxidants, is about 10-fold more sensitive, and requires only a single analysis to measure ubiquinol and ubiquinone. However, the run time is longer, and the retention times using this method are very sensitive to small changes in the composition of the mobile phase. The changes in retention times can be prevented by always preparing the mobile phase by adding 40 ml of water to 960 ml of methanol. It is also important to note that the sample solvent in this system (methanol/ethanol, 1 : l, v/v) is a stronger eluant than the mobile phase. This can lead to peak broadening or even splitting with larger injection volumes. To avoid this problem the injection volume should be kept to a minimum. We have found that volumes up to 25/zl can be injected with no apparent degradation in peak shape. The limit of detection for a-tocopherol and ubiquinol is 0.1 pmol with a coefficient of variation of 2.7 and 8.9%, respectively. Method 2. Three micron particle size Supelcosil LC-18-DB guard (1.5 cm × 4.6 ram) and analytical (15 cm x 4.6 mm) columns are used. The mobile phase consists of 20 mM lithium perchlorate in methanol/ethanol/ 2-propanol (22.5:73.6:3.9, v/v/v) delivered at a flow rate of 1 ml/min. the electrochemical detection system consists of an LC-4B amperometric detector with a glassy-carbon working electrode and an AglAgCl reference electrode. The applied potential is set at +0.6 V with a sensitivity setting of 50 nA. Analysis of human plasma lipid extracts by HPLC (Fig. 2B) produces peaks corresponding to y-tocopherol (retention time 5.3 min), a-tocopherol (5.8 min), ubiquinol 10 (11.3 min), lycopene (13.1 min), and fl-carotene (14.7 min). The peaks for lycopene and fl-carotene in both Methods 1 and 2 are not completely resolved from other electrochemically active carotenoids present in blood plasma (i.e., a-carotene). Interference can hamper quantitation of lycopene and fl-carotene if the other carotenoids are present in relatively high concentrations. There are several other methods available for the separation and detection of individual carotenoids.27-29 The detection limit for a-tocopherol and ubiquinol for Method 2 is 1 pmol with a coefficient of variation of 3.1 and 8.3%, respectively. The wide range of concentrations of these lipid-soluble antioxidants in plasma often presents a problem in quantitation using these methods. The high concentration of a-tocopherol (25/xM) relative to ubiquinol (0.7 /zM) makes it difficult to keep tocopherol on scale and obtain a measurable peak for ubiquinol. The problem can be minimized by reducing the baseline 27 K. W. Miller and C. S. Yang, Anal. Biochem. 145, 21 (1985). 28 D. B. Milne and J. Botnen, Clin. Chem. (Winston-Salem, N.C.) 32, 874 (1986). 29 N. E. Craft, S. A. Wise, and J. H. Soares, J. Chromatogr. 589, 171 (1992).
[24]
ANTIOXIDANT
STATUS IN PLASMA AND BODY FLUIDS
279
noise of the electrochemical detector as much as possible to ensure accurate integration of small peaks. Much of the baseline noise produced by electrochemical detectors is caused by fluctuations in the back pressure. Use of a pulse damper and low pulse pumps will help in reducing the noise. Summary The concentration of antioxidants in human blood plasma is important in investigating and understanding the relationship between diet, oxidant stress, and human disease. The H P L C - E C technique combines selectivity with high sensitivity for measuring both water- and lipid-soluble antioxidants. The excellent sensitivity of the methods described here allows one to measure a panel of antioxidants in a small volume of plasma. Acknowledgments This work was supported by National Institutes of Health Grant CA39910 to B.N.A. and by NationalInstituteof EnvironmentalHealth SciencesCenterGrant ES01896. P.A.M. was supportedby NationalInstitutesof HealthTrainingGrantESO 7075. B.F. was supported by Grant 13-528-901fromthe AmericanHeartAssociation,MassachusettsAffiliate,Inc., and a FutureLeaderAwardfromthe InternationalLifeSciencesInstitute-NutritionFoundation.
[24] T o t a l A n t i o x i d a n t S t a t u s in P l a s m a a n d B o d y F l u i d s By CATHERINE RICE-EVANS and NICHOLASJ. MILLER Introduction Methods that have been developed for the measurement of the antioxidant activity of fluids are all essentially inhibition methods: a free radical species is generated, there is an end point by which the presence of the radical is detected, and the antioxidant activity of the added sample inhibits the end point by scavenging the free radical. Methods vary greatly as to the radical that is generated, the reproducibility of the generation process, and the end point that is used (Fig. 1). An important consideration that has been often overlooked is the reliability and practicability of the method: much of the literature in this field has been based on extremely small sample numbers, using methods that are too complex to generate easily significant estimates of the analytical precision. METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[24]
ANTIOXIDANT
STATUS IN PLASMA AND BODY FLUIDS
279
noise of the electrochemical detector as much as possible to ensure accurate integration of small peaks. Much of the baseline noise produced by electrochemical detectors is caused by fluctuations in the back pressure. Use of a pulse damper and low pulse pumps will help in reducing the noise. Summary The concentration of antioxidants in human blood plasma is important in investigating and understanding the relationship between diet, oxidant stress, and human disease. The H P L C - E C technique combines selectivity with high sensitivity for measuring both water- and lipid-soluble antioxidants. The excellent sensitivity of the methods described here allows one to measure a panel of antioxidants in a small volume of plasma. Acknowledgments This work was supported by National Institutes of Health Grant CA39910 to B.N.A. and by NationalInstituteof EnvironmentalHealth SciencesCenterGrant ES01896. P.A.M. was supportedby NationalInstitutesof HealthTrainingGrantESO 7075. B.F. was supported by Grant 13-528-901fromthe AmericanHeartAssociation,MassachusettsAffiliate,Inc., and a FutureLeaderAwardfromthe InternationalLifeSciencesInstitute-NutritionFoundation.
[24] T o t a l A n t i o x i d a n t S t a t u s in P l a s m a a n d B o d y F l u i d s By CATHERINE RICE-EVANS and NICHOLASJ. MILLER Introduction Methods that have been developed for the measurement of the antioxidant activity of fluids are all essentially inhibition methods: a free radical species is generated, there is an end point by which the presence of the radical is detected, and the antioxidant activity of the added sample inhibits the end point by scavenging the free radical. Methods vary greatly as to the radical that is generated, the reproducibility of the generation process, and the end point that is used (Fig. 1). An important consideration that has been often overlooked is the reliability and practicability of the method: much of the literature in this field has been based on extremely small sample numbers, using methods that are too complex to generate easily significant estimates of the analytical precision. METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
280
[24]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
OPD/H202 [7] Peroxylradicalsfrom HRP/HzOz \ , . ABAP[1,6] 12 Ferrylrnyoglooin [ ] ~ ~ radicals and // A/APH[8] CuZ~Hz02[9,~ ] ~ \ ABTS [13] / / AMVN[5]Lipoperoxidesin / / brainhornogenates CuZ+/cumenehydroperoxide t/ [2,3] / [10] Dopaoxidation[11] ~ superoxide[4]
Chemiluminescence [5,6,12] ~
"
Ox, oon
uptake[1] ,
or
.0.e
.
change [7,13]
e
~
endpoint
/
\
inhibition
ofendpoint
J
)
/ Cr~ellrphology [4] COproduction [11] Fro. 1. Methods for the measurement of antioxidant activity. HRP, Horseradish peroxidase; OPD, o-phenylenediamine; ABTS, 2,2'-azinobis(3-ethylbenzothiazoline 6-sulfonate); ABAP, 2,2'-azobis(2-amidinopropane) hydrochloride; AAPH, 2,2'-azobis(2-amidinopropane) dihydrochloride; AMVN, 2,2'-azobis(2,4-dimethylvaleronitrile); TBA-RS, thiobarbituric acid-reactive substances; and CO, carbon monoxide. Numbers within brackets refer to literature cited in the text.
The TRAP assay of Wayner e t a l . ~ has to date been the most widely used assay of antioxidant activity. Whereas the earlier methods 2-4 were based on inhibition of spontaneous tissue autoxidation by antioxidants, these authors took advantage of their discovery that thermal decomposition of the water-soluble azo compound 2,2'-azobis(2-amidopropane) hy1 D. D. M. Wayner, G. W. Burton, K. U. Ingold, and S. Locke, FEBS Lett. 187, 33 (1985). 2 A. A. Barber, Arch. Biochem. Biophys. 92, 38 (1961).
[24]
ANTIOXIDANT STATUS IN PLASMA AND BODY FLUIDS
281
drochloride (ABAP) yields peroxyl radicals at a known and constant rate. Each molecule of Trolox (Trolox®, Hoffman-LaRoche, Basel, Switzerland), a-tocopherol, or other phenolic antioxidant traps two peroxyl radicals, giving Trolox a stoichiometric factor of 2.0 in the TRAP assay. Other pure antioxidants have different stoichiometric factors (ascorbate, 1.5; urate, 1.7), and these must be taken into account when extrapolating back to molar concentrations from TRAP values. The procedure for a serum sample is that it is mixed with linoleic acid to prevent the termination reaction of two peroxyl radicals occurring, although it has been suggested that this step is not necessary 5 and that the observed problems had been caused by minute changes in the incubation pH, rather than termination reactions. Then an aliquot (25-50/zl) is added to the air-saturated buffer in the chamber of an oxygen electrode (Clark-Collip rate amperometric electrode). The reaction, started by the addition of ABAP, is monitored until oxygen uptake is maximal. After 50% of the oxygen has been taken up, Trolox is added to calibrate the system and the oxygen uptake monitored for a further period. On the recorder trace the lag time before the start of oxygen uptake is the response which is measured. The total assay time for a single sample varies according to its antioxidant capacity, but that for a serum sample is approximately 90 rain. A reference interval of 571-1284/zM was derived for stored serum samples (n = 45), and this was not found to be significantly different when fresh plasma was used (n = 6).1 A major problem with the original TRAP assay method lies in the oxygen electrode end point: an oxygen electrode will not maintain its stability over the period of time required (up to 2 hr per sample). Such instability will be more pronounced with plasma samples than with pure solutions. Hence a high degree of imprecision is inherent. The authors of the original method did not present a multipoint dose-response curve or any indication of the analytical imprecision of the response at different doses. Calibration is by a single mid-concentration dose point. The assay itself is too lengthy to carry out adequate precision monitoring, or to permit analysis of large numbers of samples. The TRAP assay was modified by Metsa-Ketela 6 with respect to the end point; luminol-enhanced chemiluminescence was used, which is more
3 j. Stocks, J. M. C. Gutteridge, R. J. Sharp, and T. L. Dorrnandy, Clin. Sci. Mol. Med. 47, 215 (1974). 4 T. Ogasawara and M. Kan, Tohoku J. Exp. Med. 144, 9 (1984). 5 T. Mets~i-Ketel~i and A.-L. Kirkkola, Free RadicalRes. Commun. 16, Suppl. 1 ; 215 (1992). 6 T. Mets~i-Ketel~i, in "Bioluminescence and Chemiluminescence: Current Status" (P. E. Stanley and L. J. Kricka, eds.), pp. 389-392. Wiley, Chichester, 1991.
282
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[24]
precise and lends itself to a degree of automation. Thus significant numbers of samples can be processed. Production of peroxyl radicals enhances the chemiluminescent reaction in this system: the addition of an antioxidant to the reaction extinguishes the chemiluminescence, the duration of which is directly proportional to the radical trapping ability of the antioxidant sample. It is claimed this end point produces an assay of considerably better precision than the original TRAP assay. The slope of the response curve is 131.7 sec (lag time before the start of oxygen uptake) per nanomole of Trolox. A further modification has been the introduction of an assay for antioxidants in the lipid phase 5 using 2,2'-azobis(2,4-dimethylvaleronitrile) (AMVN) as the source of peroxyl radicals. The nonpolar AMVN will generate radicals within the lipid phase of a mixture, as opposed to ABAP, which will generate peroxyl radicals within the aqueous phase. Again using luminol and a luminometric end point, this assay has been applied to the measurement of the antioxidant capacity of human low density lipoprotein (LDL) preparations. The o-phenylenediamine (OPD) method of Nakamura e t al. 7 u s e s this substance as a peroxidase substrate. On conversion to the activated (free radical) form, OPD will exhibit an increase in absorbance at 430 nm. In this assay 500/zl of serum is mixed with OPD and H202 in citrate-phosphate buffer, pH 5.0. The change in absorbance over a 2-hr period is measured, and the results are expressed as the absorbance change. No antioxidant standard was reported as being used, but a number of antioxidant drugs have been examined. Sera from a number of patients have been measured in the system, and correlations made with clinical parameters of disease states, erythrocyte sedimentation rate, etc. In the phycoerythrin assay of DeLange and Glazer, 8 antioxidants that react rapidly with peroxyl radicals (such as ascorbate, urate, and a-tocopherol) protect phycoerythrin from damage by such radicals, and hence inhibit the quenching of its characteristic fluorescence. 2,2'Azobis(2-amidinopropane) dihydrochloride (AAPH), which thermally decomposes to yield peroxyl radicals on being heated to 37°, is used as a constant source of peroxyl radicals./3-Phycoerythrin is used as the detector agent; it is normally highly fluorescent (excitation 540 nm; emission 565 nm), but it loses fluorescence intensity when damaged by peroxyl radicals. Phycoerythrin reacts unusually slowly with peroxyl radicals: 100-fold slower than ascorbate or a-tocopherol, and 60-fold slower than other protein molecules. Mixing phycoerythrin with AAPH results in a 7 K. Nakamura, H. Endo, and S. Kashiwazaki, Int. J. Tissue React. 9, 307 (1987). 8 R. J. DeLange and A. N. Glazer, Anal. Biochem. 177, 300 (1989).
[24]
ANTIOXIDANT STATUS IN PLASMA AND BODY FLUIDS
283
slow, progressive loss of phycoerythrin fluorescence over a period of hours. Addition of a serum sample or an antioxidant solution inhibits the loss of fluorescence intensity, and this inhibition is proportional to the antioxidant capacity of the added sample. The phycoerythrin assay uses 20-/~1 sample volumes and is carried out in triplicate. Trolox is used as an antioxidant standard. The incubation time is 90 min, and the percent inhibition of fluorescence loss is plotted against the dose of added antioxidant. However, the kinetics of phycoerythrin fluorescence quenching is not zero-order for all pure antioxidant substances analyzed, and under those circumstances the percent inhibition is derived from initial rate readings of fluorescence loss. The assay is suitable for kinetic analysis, and can hence be performed using a commercial microprocessor-controlled fluorescence microplate reader. At the time of writing there is no further information in the literature as to the results obtained with this assay. The thiobarbituric acid-reactive substances (TBARS) method of Arshad et al. 9 w a s designed to measure the susceptibility of the plasma to Cu2+/H202-stimulated peroxidation. Plasma is incubated with cupric acetate and hydrogen peroxide solutions for 1 hr at 37°, and the millimolar concentration of lipid peroxides formed is estimated with thiobarbituric acid (TBA). Two milliliters of plasma is required for the technique, and unless this can be reduced by at least an order of magnitude the test would be impractical to carry out on any wide-ranging scale. Evaluation of samples from nine diabetics and nine controls shows the much greater peroxidation potential of diabetic sera. As well as having higher basal TBA levels the diabetic sera show approximately 25 times the increment in TBA over basal levels, and this can presumably be related to a lower antioxidant status in diabetics. A fluorescence assay based on the lipophilic fluorophore cis-parinaric acid, a substance that is fluorescent in a lipid environment, has been developed by McKenna et al. 10cis-Parinaric acid fluorescence (excitation 324 nm, emission 413 nm) is destroyed on reaction with free radicals, and the rate constant of the fluorescence loss can be decreased by addition of an antioxidant such as a-tocopherol, cis-Parinaric acid was in use as a fluorescent fatty acid probe when it was noted that it formed nonfluorescent products with free radicals in a first-order, steady-state reaction, and an assay was established to evaluate the antioxidant potential of the lazaroid group of drugs. The reaction is initiated by copper sulfate, fol9 M. A. Q. Arshad, S. Bhadra, R. M. Cohen, and M. T. R. Subbiah, Clin. Chem. (WinstonSalem, N.C.) 37, 1756 (1991). 10 R. McKenna, F. J. Kezdy, and D. E. Epps, Anal. Biochem. 196, 443 (1991).
284
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[24]
lowed by cumene hydroperoxide, and the loss of fluorescence is monitored for 30 min. The method of Cooper and Engel,11 measuring carbon monoxide production from DOPA, has also been applied to drugs and pure antioxidant solutions, rather than the extracellular fluid. 3,4-Dihydroxyphenylalanine (DOPA) oxidation will proceed with the release of carbon monoxide, if there are no antioxidants in the incubation mixture. The CO measurements are made using gas-liquid chromatography. No results for serum samples are, as yet, available for this method. Finally, in the method of Whitehead et al. 12free radicals generated by horseradish peroxidase and hydrogen peroxide or perborate are passed first of all to p-iodophenol, and then to luminol, which can be monitored by chemiluminescence. Addition of antioxidants thus inhibits the luminescence of the incubation mixture. This new luminometric technique can be calibrated with Trolox, for which a stoichiometric factor of 2 is ascribed; serum samples must be deproteinized for assay. General Principle for Measuring Antioxidant Activity We have developed a new method for measuring antioxidant activity 13 (Fig. 2), based on the inhibition by antioxidants of the absorbance of the radical cation of 2,2'-azinobis(3-ethylbenzothiazoline 6-sulfonate) (ABTS.+), which has a characteristic long-wavelength absorption spectrum showing maxima at 660, 734, and 820 nm. The ABTS +. radical cation is formed by the interaction of ABTS (150 IzM) with the ferrylmyoglobin radical species, generated by the activation of metmyoglobin (2.5 /zM) with H202 (75 ~M). Antioxidant compounds suppress the absorbance of the ABTS +. radical cation to an extent and on a time scale dependent on the antioxidant capacity of the substance under investigation. 14,15 In terms of assay design, several different analytical strategies are apparent, for example: (a) decolorization assay, (b) inhibition assay (fixed time point), (c) inhibition assay (reaction rate), and (d) lag phase measurement. For the decolorization assay (a), the reaction between ABTS and hydrogen peroxide could be allowed to proceed until the color of the incubaII M. J. Cooper and R. R. Engel, Clin. Chim. Acta 202, 105 (1991). 12T. P. Whitehead, G. H. G. Thorpe, and S. R. J. Maxwell, Anal. Chim. Acta 266, 265 (1992). 13 C. Rice-Evans and M. J. Davies, U.K. Patent 91,242,727 (1991). 14 N. J. Miller, C. Rice-Evans, M. J. Davies, V. Gopinathan, and A. D. Milner, Clin. Sci. 84, 407 (1993). t5 K. D. Whitburn, J. J. Shieh, R. M. Sellers, M. Z. Hoffman, and I. A. Taub, J. Biol. Chem. 257, 1860 (1982).
[24]
ANTIOXIDANT STATUS IN PLASMA AND BODY FLUIDS H X - Fem
+
285
H202
metmyoglobin
%%% %%
%%%
V "X- [Fe Iv
=
A O]
ABTS
ferryl rnyoglobin t •
/
s
ABTS +
+
H X - F e m. . . . . . . . . . . . . . .
blue-green radical cation
FIG. 2. Formation of the ABTS "+radical cation from activated myoglobin.
tion mixture is stable. This reaction requires the presence of myoglobin, acting as a peroxidase, via the formation of the ferrylmyoglobin radical, to which ABTS donates an electron, forming ABTS +.. When an aliquot of plasma is added to the reaction mixture, there is a degree of decolorization owing to the presence of plasma antioxidants which reverse the formation of the ABTS radical cation. The percent loss of color (blank - test, measuring at 734 nm), or the percentage of color remaining at a given time point, can then be used as an index of plasma antioxidant activity. For the inhibition assay with a fixed time point (b), ABTS, myoglobin, and a sample are mixed, and the reaction is initiated by the addition of hydrogen peroxide. The reaction may also be initiated by the addition of metmyoglobin, with the hydrogen peroxide added at an earlier time point. After a fixed time the absorbance of the solution is read, along with a buffer blank (which will have a higher absorbance value than a solution containing plasma). The blank absorbance minus the test absorbance, divided by the blank absorbance (expressed as a percentage), is the percentage inhibition of the reaction. This value, B - T / B x 100 (the percent % Inhibition
= A b S b l a n k -- A b S t e s t
AbSblank
286
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[24]
inhibition), defines the response of the system and is proportional to the antioxidant capacity of the plasma sample. Under stable conditions the antioxidant capacity of the sample is inversely proportional to the test absorbance at 734 nm. For the inhibition assay measuring reaction rates (c), the procedure outlined in (b) is followed, with all the reagents added together and the reaction started with hydrogen peroxide, but the reaction rates of the test and buffer blank are monitored. A result is thus derived by comparison of reaction rates rather than absorbance at a fixed time point. It might thus be possible to derive the result at an earlier point in the reaction than by using a fixed time method, and the linearity range of the assay might simultaneously be extended. For the lag phase measurement assay (d), sample, metmyoglobin, ABTS, and hydrogen peroxide are mixed at time zero, and the time is noted for the development of color in the cuvette to be initiated. The length of time of the lag phase before the reaction starts is proportional to the concentration of antioxidant in the sample. Method for Determination of Total Antioxidant Activity The strategy followed in the development of the assay protocol has been that outlined in (b) above, using hydrogen peroxide as starter. [The assay design used in the TRAP assay ~is that described in (d) above, i.e., measurement of lag time or induction time.] Measurement of absorption maxima at the red end of the spectrum overcomes the potential problem of direct spectral interference from myoglobin itself, or other heme proteins, whose absorptivity decreases greatly above 580 nm. Similarly, direct interference from bilirubin, which absorbs strongly in the region of 460 nm, should be minimal at high wavelengths. Materials
Metmyoglobin (equine) is purchased from Sigma (St. Louis, MO), 2,2'azinobis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS) from Aldrich (Milwaukee, WI), and hydrogen peroxide (Aristar) from BDH (Poole, UK). Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid) is purchased from Aldrich. The antioxidants N-acetylcysteine, albumin (human fraction V), L-ascorbic acid, bilirubin, glutathione (reduced form), DL-a-tocopherol, and uric acid are obtained from Sigma, and so is heparin. D-Mannitol, Nonidet P-40 (NP-40), urea, potassium ferricyanide, and buffer salts are obtained from BDH. The buffer used is isotonic 5 mM phosphate-buffered saline, pH 7.4 (PBS).
[24]
A N T I O X I D A N T STATUS IN PLASMA A N D BODY F L U I D S
287
Reagent Preparation 1. Trolox (2.5 mM; Rmm 250.29) is prepared by dissolving 0.15641 g of Trolox in 250 ml of PBS. At this pH the solution is near the upper limit of Trolox solubility, and gentle ultrasonication is required to dissolve the crystals. Experiments have shown that frozen Trolox at this concentration is stable for more than 6 months. Fresh working standards (0.5, 1.0, 1.5, 2.0 mM) are prepared daily by mixing 2.5 mM Trolox with PBS. 2. Working solutions of hydrogen peroxide are prepared from stock Aristar H202 (BDH) after an initial dilution to a concentration of 500 mM in PBS. Aristar H202 is supplied as a 30% solution, with a specific gravity of 1. I0 (1.099-1.103); Since the Rmmis 34, a 500 mM solution is prepared by diluting 515/zl of H202 to 10 ml in PBS (solution A). A working solution of 450/zM is used for the manual antioxidant assay, and this is prepared by diluting 45 /zl of solution A to 50 ml with PBS. For the automated antioxidant assay a hydrogen peroxide working solution of 1.08 mM is used, and this is prepared by diluting 108/xl of solution A to 50 ml with PBS. Stock and working hydrogen peroxide solution are freshly prepared prior to use. 3. ABTS (5 mM; Rmm 548.68) is prepared by dissolving 0.02743 g of ABTS in 10.0 ml of PBS buffer. ABTS dissolved in this way has a millimolar extinction coefficient (end) of 38.8 at 340 nm. 4. o~-Tocopherol (Rmm 430.7) is a viscous oil at room temperature, and solutions of known concentration are prepared by taking an aliquot, dissolving it in ethanol, and reading the absorbance of dilutions of the stock solution (emMof a-tocopherol is 3.26 at 292 nm). For introduction into the incubation mixture an emulsion of a-tocopherol is produced by making appropriate dilutions of the ethanolic stock solution in buffer containing 2% NP-40, followed by vigorous vortex mixing. 5. Bilirubin (2.0 mM; Rmm 584.7) is prepared by adding 0.5 ml of 0.1 M KOH to 0.00585 g of bilirubin, followed by 4.5 ml of PBS with 2% NP-40 added. Assaying a blank of KOH/buffer/2% NP-40 confirmed that this solvent system had no antioxidant activity. 6. Uric acid (2.0 mM; Rmm168.1) is prepared by dissolving 0.03362 g in 100 ml of 0.5 g/liter lithium carbonate solution (which has no antioxidant activity). 7. Metmyoglobin is purified prior to use on a 35 × 2.5 cm Sephadex G-15-120 column in phosphate-buffered saline, pH 7.4. The concentration of myoglobin in the column eluate is calculated from the extinction coefficients, and aliquots of metmyoglobin in PBS are stored frozen until required for use. A 400/xM solution of metmyoglobin (0.0752 g in 10 ml of PBS) is prepared; 0.0244 g of potassium ferricyanide is dissolved in 100
288
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[24]
TABLE I TOTAL ANTIOXIDANT ACTIVITY: PROTOCOL FOR COBAS BIO CENTRIFUGAL ANALYZER
Step Transfer
Mix Read initial absorbance at 734 nm Transfer Mix ( = T0)
Incubate for 6 min at 30°
Read final absorbance at 734 nm T6 min Plot
Component Sample Water ABTS Metmyoglobin
Volume or Comment
Concentration -Flushes probe, 180 ~ M ~ 3.0 p,M J
3.0/xl 30 p,l 300/zl of mixed reagent
H202 H202 added to all cuvettes by centrifugal force: reaction starts Sample ABTS Metmyoglobin H202 All cuvettes
1.074 mM
25/zl Final volume 358 ttl
0.84% 150/zM 2.5 p.M 75/zM
Final concentration in incubation mixture
Trolox standards
Concentration versus AA734nm
Logit/log4 curve fit
All cuvettes
ml of PBS, and then 10 ml of the ferricyanide solution is added to 10 ml of metmyoglobin solution. The mixture is applied to the column and eluted with PBS. The first fraction is collected and its absorbance read at 490, 560, 580, and 700 nm. The absorbance reading at 700 nm is subtracted from the readings at 490, 560, and 580 nm to correct for background absorbance. Calculation of the relative proportions of the different forms of myoglobin can be carried out using deconvolution procedures or applying the Whitburn algorithms based on the extinction coefficients at 490, 560, and 580 nmlS: [MetMb] = 146A490 - 108A560 + 2.1A580 [FerrylMb] = -62A490 + 242A560 - 123A580 [MbO2] = 2.8A490 - 127A560 + 153A580 Myoglobin prepared in this way should only be used if it constitutes more than 94% of the total heme species present. The purified myoglobin is diluted with buffer to a concentration of 140/zM (if necessary), divided into aliquots, and stored frozen until use.
[24]
ANTIOXIDANT STATUS IN PLASMA AND BODY FLUIDS
289
TABLE II TOTAL ANTIOXIDANT ACTIVITY: PROTOCOL FOR MANUAL METHOD USING SPECTROPHOTOMETER
Step Transfer
Mix Transfer Start clock Mix Transfer into cuvette at 30° Incubate for 6 min at 30° Read final absorbance at 734 nm T6min Plot
Component Sample Buffer, pH 7.4 Metmyoglobin ABTS
Concentration
Volume or Comment
To make 1.0 ml 70/xM 500 tzM
8.4/zl 489/d 36/xl 300 ~1
HzO2
450/xM
167/zl
Sample ABTS Metmyoglobin H202 All cuvettes
0.84% 150/~M 2.5/zM 75/~M
Trolox standards
Final concentration in incubation mixture
Concentration versus AA734 nm
Automated Antioxidant Assay Using Premixed Reagent In the protocol devised for the Cobas Bio centrifugal analyzer (Roche Diagnostic System Inc., Branchburg, NJ), 300 /xl of ABTS/myoglobin reagent in PBS is mixed with 3/zl of sample, the probe flushed with 30 /zl of diluent, and then hydrogen peroxide (25/zl) added last as a starter. The incubation volume is thus 358/zl; reagents are prepared so that on dilution into this incubation volume they are at the desired concentration (2.5/zM metmyoglobin, 150/xM ABTS, 75/xM HzO 2 , and 0.84% sample fraction). This is performed by mixing 2.143 ml of 140/zM metmyoglobin with 3.6 ml of 5 m M ABTS and diluting the mixture to 100 ml with PBS. Working hydrogen peroxide for the Cobas analyzer is prepared at a concentration of 1.08 AM. The protocol is summarized in Table I. Timing and temperature control require careful evaluation when the analysis is first set up: 6 rain represents the point at which the inhibition of the reaction in the cuvette containing the top standard (2.5 m M Trolox) has been overcome, and color is starting to develop (i.e., the lag phase has ended). Minor interinstrumental variations may necessitate adjusting the time of taking the final reading.
290
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[24]
TABLE III RESULTS FROM SIXTEEN SEPARATE CALIBRATION CURVES
Trolox (mM)
A734nm
% Inhibition of blank
S.D.
CV (%)
0.0 0.5 1.0 1.5 2.0 2.5
0.690 0.564 0.426 0.282 0.139 0.020
-18.2 38.3 59.1 79.9 97.1
0.033 1.64 2.34 3.97 4.55 1.41
4.8 8.9 6.1 6.7 5.7 1.5
Manual Total Antioxidant Assay For a manual procedure (Table II) the following concentrations of reagents are used per tube: sample, 0.84%; metmyoglobin, 2.5/zM; ABTS, 150/~M; H202, 75/xM; buffer to a total volume of 1.0 ml. The reagents are mixed as follows: 8.4/zl of sample, 489/xl of buffer, 36/xl of 70/zM metmyoglobin, and 300/xl of 5 mM ABTS, with vortexing. The reaction is started by the addition of 167/zl of 450/zM H202, the clock started, and the tube revortexed. The reaction mixture is transferred into a 1-cm cuvette in a spectrophotometer. A quantitative relationship exists between the absorbance at 734 nm at 6 min and the antioxidant activity of the sample or standard. Using the manual method, timing and temperature are difficult to control to within the required limits.
Results
Estimates of Analytical Imprecision The dose-response curve for Trolox concentration versus absorbance at 734 nm is highly reproducible under the conditions described for the TABLE IV INTRAASSAY PRECISION OF TOTAL ANTIOXIDANT ASSAY Parameter
High control
Low control
Number Mean (mM) S.D. CV (%)
25 1.52 0.008 0.54
25 0.83 0.013 1.59
[24]
ANTIOXIDANT STATUS IN PLASMA AND BODY FLUIDS
291
TABLE V INTERASSAY PRECISION OF TOTAL ANTIOXIDANT ASSAY
Parameter
High control
Low control
Number Mean (mM)
20 1.52 0.055 3.6
20 0.84 0.050 6.1
S.D. CV (%)
automated assay. Table III shows the results obtained for calibration curves derived from 16 sequential and separately prepared Trolox stock solutions, assayed over a 4-month period on the Cobas Bio centrifugal analyzer. Further repeated reanalysis of frozen 2.5 mM Trolox in PBS suggests that solutions of that concentration are in fact stable at - 20 ° for more than 6 months. Aliquots from two serum pools (reconstituted freeze-dried controls) were included in each batch of samples. Table IV shows the results from the repeated reanalysis of these pools in a single run. Table V shows the accumulated data from 20 sequential runs, illustrating the high degree of reproducibility which can be obtained from day-to-day running of the assay.
Results for Pure Antioxidant Substances Solutions of antioxidant substances can be compared to Trolox solutions by means of this assay, and hence to one another on the basis of T A B L E VI RESULTS FOR SOLUTIONS OF PURE ANTIOXIDANTS
Substance
TEAC a
Bilirubin Urate Ascorbate a-Tocopherol Albumin Glutathione N-Acetylcysteine
n
S.D.
1.50
3
1.02 0.99 0.97 0.63 0.90 1.43
5 5 3 3 3 3
0.12 0.06 0.04 0.01 0.02 0.03 0.08
a TEAC is the millimolar concentration of a Trolox solution having the antioxidant capacity equivalent to a 1.0 m M solution of the sub-
stance under investigation.
292
[24]
ANTIOXIDANT CHARACTERIZATION AND ASSAY T A B L E VII TOTAL PLASMA ANTIOXIDANT ACTIVITY IN PREMATURE AND TERM BABIES Time of assay At birth At 5 days
Premature
Term
1.21 +- 0.08 (n = 16) 1.41 +- 0.09 (n = 14)
1.46 -+ 0.07 (n = 18) 1.50 -+ 0.14 (n = 15)
molar antioxidant activity. A mean figure is derived for the antioxidant capacity per mole of substance, or TEAC (Trolox equivalent antioxidant capacity), based on results derived from several separate assays. The TEAC value is defined as the millimolar concentration of a Trolox solution having the same antioxidant capacity as a 1.0 mM solution of the substance. Derivation of a TEAC value gives a direct comparative measure of the antioxidant capacity among groups of substances (Table VI), for example, drugs that are being evaluated as potential antioxidants. The plasma antioxidants urate, ascorbate, and a-tocopherol have the same antioxidant capacity as Trolox, as expected. Bilirubin has an antioxidant capacity 50% greater (on a mole for mole basis) than these substances. Albumin has a lower antioxidant capacity. The thiol compounds exhibit widely varying activities, with N-acetylcysteine being more effective than glutathione. Urea and heparin (measured at 10,000 IU) do not respond in the assay (as expected) and heparin can therefore be used for the collection of plasma without influencing the total antioxidant value of the sample. Glucose and mannitol do not act as antioxidants in this system, although they have been reported to react with the hydroxyl radical under specific conditions; neither does ethanol, which can therefore be used to solubilize substances for inclusion in the assay. This method for measuring a TEAC value is dependent on the substance being water soluble, or on its being solubilized or emulsified with a detergent. The activity of a-tocopherol can be determined by dissolving it in ethanol and then emulsiT A B L E VIII RELATIONSHIP OF TOTAL PLASMA ANTIOXIDANT ACTIVITY OF BABIES AT BIRTH WITH MATERNAL LEVELS Source Premature babies at birth Mothers of premature babies Term babies at birth Mothers of term babies
Total antioxidant activity (mM) 1.17 1.25 1.46 1.41
-+ 0.06 -+ 0.08 -+ 0.07 -+ 0.07
(n (n (n (n
= = = =
6) 6) 18) 17)
[24]
ANTIOXIDANT STATUS IN PLASMA AND BODY FLUIDS
293
fying with 2% aqueous NP-40 (a nonionic detergent) before introducing it into the reaction system.
Reference Interval for Adult Human Plasma An adult reference interval (95th percentile range) of 1.32-1.60 mM (1.46 + 0.14 raM) was derived for the method, based on 312 samples. No significant difference was noted in the ranges for plasma and serum.
Cfinical Applications Because the neonatal period is a time of oxidative stress, and since premature neonates are known to be antioxidant deficient with respect to o~-tocopherol, the plasma antioxidant status of term infants (37-42 weeks gestation) was investigated and compared with the antioxidant status of premature infants (25-30 weeks gestation). Applying this method, premature infants were shown to have significantly lower plasma antioxidant activity than term infants (Table VII). After 5 days the preterm infants had stabilized their plasma antioxidant activity to within the adult reference interval. An important additional finding was that the mothers of premature infants have a lower plasma antioxidant activity than the mothers of term infants, that is, the levels of the babies reflected those of the mothers in terms of antioxidant activity (Table VIII). Thus, although samples from term infants and their mothers fall within the adult reference interval for plasma total antioxidant activity, premature infants of less than 32 weeks gestation and their mothers show evidence of deficient antioxidant activity. This may contribute significantly to the morbidity and mortality among such infants, who require active oxygen therapy at birth owing to the immaturity of their lungs. Acknowledgments We thank the British Technology and Hoffman LaRoche (Basel, Switzerland) for funding the development of this work.
294
ANTIOXIDANT
CHARACTERIZATION
AND ASSAY
[25]
[25] A n a l y s i s o f V i t a m i n E H o m o l o g s in P l a s m a a n d T i s s u e : High-Performance Liquid Chromatography
By WILLY SCHUEP and ROSEMARIE RETTENMAIER Introduction Vitamin E, one of the essential micronutrients, occurs in nature in a series of compounds called tocopherols and tocotrienols. The four naturally occurring tocopherols (T), R,R,R-a-, R,R,R-fl-, R,R,R-7-, and R,R,R-8-tocopherol, differ in the number and position of methyl groups on the chromanol ring (Fig. 1). Parallel to these, four derivatives with an unsaturated side chain are also found in nature. All the eight forms mentioned have been isolated from vegetable oils and other plant materials, ~ which represent the richest natural source of these compounds. This is due to their special function as antioxidants, as stabilizers, and as one of the factors responsible for the initiation of flowering. Because humans and most animals are unable to synthesize vitamin E, they have to rely on obtaining it from other sources. Whereas the various homologs are synthesized by many plants and can be taken up via the normal food/feed chain, a-T is the predominant component in human and animal tissue and blood, followed by 7-T. Under certain circumstances there is an increased need for this micronutrient. Supplementation of foods and feeds is then an alternative to provide an adequate vitamin E supply. Usually the enrichment is performed with special formulated preparations of all-rac-a-tocopheryl acetate or other esters. The biological activities as determined by a rat resorption test vary depending on the vitamer and range from 100% for R,R,R-a-T to 57% for R,R,Rfl-T, 31% for R,R,R-7-T, and 1.4% for R,R,R-8-T. The biopotency of R-a-tocotrienol was found to be 30%. 2 The determination of naturally occurring as well as added tocopherols is desirable in order to assess the nutritional value of the foodstuffs. In blood plasma and tissues, such as liver, the a-T level reflects the vitamin E status in humans and animals. Thus, a possible state of deficiency is recognizable in due time, and therapeutic measures can be taken. This is of great importance especially in cases of cardiovascular diseases and
l p. B. McCay and M. M. King, in "Vitamin E: A Comprehensive Treatise" (L. J. Machlin, ed.), p. 289. Dekker, New York, 1980. 2 B. J. Weimann and H. Weiser, A m . J. Clin. Nutr. 53, 1056S-1060S (1991).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All fights of reproduction in any form reserved.
[25]
H P L C METHODS FOR ANALYSIS OF VITAMIN E
HO2~ Rz
R1
R1 ] R_t2 H CH3 3 CH3
R1
R1 CH3 CH3 H H
I
~
I
3'
7'
11'
R2 CH3 H CH3 H
295
a-Tocopherol [$-Tocopherol 7-Tocopherol 5-Tocopherol
a-Tocotrienol 13-Tocotrienol 7-Tocotrienol 5-Tocotrienol
FIG. 1. Structures of vitamin E active components.
cancer. In animal nutrition vitamin E deficiency leads to muscular degeneration, liver necrosis, encephalomalazia, etc. The vitamin E concentration in tissues is rather low. Having a lipophilic character it occurs together with the other cell lipids, which might cause in certain cases severe analytical problems. Working examples for the analysis of vitamin E in tissue have been reported earlier, 3,4 and numerous methods have been developed and described. In 1971 a review on the methodology was published by BunnellJ A comprehensive and very detailed survey was written in 1992 by Bourgeois. 6 The determination of vitamin E in food and tissues by chemical analysis includes lipid extraction followed by saponification. A simpler procedure with the saponification carried out first has been preferred by many investigators. The unsaponifiable matter has to be extracted and purified for the colorimetric or fluorometric estimation. For example, the extract obtained after saponification is subjected to two-dimensional thin-layer chromatographic (TLC) separation. The tocopherol zones have to be eluted and quantified photometrically after a modified Emmerie-Engel reaction. 7'8 The quantification can 3 j. L. Burtiss and A. T. Diplock, this series, Vol. 105, p. 131. 4 I. D. Desai, this series, Vol. 105, p. 138. 5 R. H. Bunnell, Lipids 6, 245 (1971). 6 C. Bourgeois, "Determination of Vitamin E and Tocotrienol." Elsevier, Barking, England, 1992. 70. Isler and G. Brubacher, in "Vitamine I, FettlOsliche Vitamine," p. 127 and pp. 142-143. Thieme, Stuttgart and New York, 1982. s "Official Methods of Analysis," 15th Ed., Vol. 2, Sect. 971.30. Association of Official Analytical Chemists, Arlington, Virginia, 1990.
296
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[25]
also be accomplished by densitometric scanning directly on the plate. 9 High-performance liquid chromatography (HPLC) has been applied widely, and this technique offers clear advantages over the other abovementioned methods. Unlike the assay in tissue, the determination of vitamin E active compounds in plasma does not require a saponification step. Tocopherol and tocotrienol analyses by HPLC can be performed either in the direct phase or the reversed-phase mode. The straight-phase chromatography has the advantage of a baseline separation of all the tocopherols and tocotrienols, whereas with the reversed-phase systems the separation of/3- and y-T and/3- and y-tocotrienol is not possible. The choice of chromatographic system depends on several factors, such as the availability of instruments and chemicals, the number of samples to be assayed in a period of time, accuracy, reproducibility, and sensitivity. The methods described below were developed ~°,11 to assay a large number of samples within a short time. They are already in use and have been proved to be robust, efficient, and accurate. Materials and Methods
Chemicals and Solutions Solvents and chemicals are of analytical grade. Methanol, ethanol, 1,4-dioxane, tetrahydrofuran, ammonium acetate, and 2,6-di-tert-butyl-pcresol (BHT) are from Fluka (Buchs, Switzerland). Acetonitrile, HPLCgrade is from S. Rathburn Chemicals (Walkerbrun, Scotland). Potassium hydroxide, n-hexane, and toluene are from Merck (Darmstadt, Germany). Ascorbic acid, all-rac-a-, all-rac-/3-, all-rac-y-, and all-rac-8-tocopherol and all-rac-a-, all-rac-/3-, all-rac-7-, and all-rac-8-tocotrienol are of high purity and are synthesized on a laboratory scale in our chemical department (F. Hoffman-La Roche, Basel, Switzerland). These compounds are commercially available from Merck (Darmstadt, Germany).
Sample Preparation Precautions (brown glass and dimmed light) have to be taken to avoid decomposition of the vitamers during sample processing and in the autosampler. Specimens can be stored at - 2 0 ° until analysis. 9 L. Dostalova, L. Salmenper/L V. V,'lclavincovfi, P. Heinz-Erian, and W. Schiiep, Nestle Nutr. Workshop Ser. 16, 275-298 (1988). l0 j. p. Vuilleumier, H. E. Keller, D. Gysel, and F. Hunziker, Int. J. Vitam. Nutr. Res. 53, 265 (1983). H D. Hess, H. E. Keller, B. Oberlin, R. Bonfanti, and W. Schtiep, Int. J. Vitam. Nutr. Res. 61, 232 (1990).
[25]
H P L C METHODS FOR ANALYSIS OF VITAMIN E
297
Homogenates of the liver samples are prepared with a mincer after partial thawing. Any other tissue is finely cut. For analysis 10 g ofhomogenate is weighed accurately in duplicate into 250-ml round-bottomed flasks. One gram of ascorbic acid is added as an antioxidant, followed by 1 ml of methanol and 10 ml ofmethanolic potassium hydroxide (375 g of potassium hydroxide is dissolved in 750 ml of water then 450 ml of methanol added). The samples, while stirring, are heated for 30 min under reflux in a boiling water bath under a nitrogen atmosphere. After cooling the saponified mixtures are transferred quantitatively into 100-ml volumetric flasks using a solution of 35% ethanol in water and filled to the mark with it. Three milliliters of the obtained mixture is extracted in duplicate with 3 ml of a solvent mix of n-hexane/toluene (1 : 1) in 10-ml centrifuge tubes, by shaking for 10 rain at 250 strokes/min. After centrifugation at 4200 g for 10 min at 0° the clear, sometimes yellow-colored organic extract is transferred into HPLC vials for chromatographic separation of the tocopherol vitamers. Plasma or serum samples are thawed and mixed well. A 250-/~I sample is diluted in a 4-ml centrifuge tube with 250/~1 of doubly distilled water and deprotonized by adding 500/zl of ethanol followed by mixing for 10 sec on a vortex mixer. From a dispenser (Hamilton Microlab M, Bonaduz, Switzerland) 1000/~1 of n-hexane is added and the tubes closed with a polyethylene stopper. After mechanical shaking for 10 min the tubes are centrifuged at a force equivalent to 2000 g at 4 ° for 10 min. For straightphase HPLC the supernatant can be used directly. For reversed-phase HPLC 400 /~1 is pipetted into Eppendorf tubes which are placed in a Speed-Vac concentrator (Savant Instrument, Farmingdale, NY) and evaporated to dryness at ambient temperature under reduced pressure and flushed with nitrogen. The residue is dissolved completely in 100/~1 of ethanol/dioxane (1 : 1) with shaking for 10 min in a mixer for Eppendorf tubes; then 150/~1 of acetonitrile is added and the samples mixed again. The sample extract is now ready for injection into the reversed-phase HPLC system. The concentration of the solutions used as external standards is checked photometrically prior to the addition of the antioxidant. The 1% t-'litTl%cmvalue in n-hexane for a-tocopherol is 85 at 297 nm. The E1 cm values in ethanol used are as follows: a-T 75.8 (292 nm), a-tocotrienol 91 (292.5 nm), fl-T 89.4 (296 nm), fl-tocotrienol 87.3 (294 rim), y-T 91.4 (298 nm), y-tocotrieno190.5 (296 nm), 8-T 87.3 (298 nm), 8-tocotrieno188.1 (297 nm). 7 Routinely the quantification is done using calibration of the readily available a-T, but with the necessary correction of the response factor of the corresponding homologs. These factors can be estimated by injecting the corresponding external standards onto the HPLC system and setting the obtained results in relation to the a-T value. The factors are different
298
A N T I O X I D A NCHARACTERIZATION T AND ASSAY
[25]
TABLE I CHROMATOGRAPHICCONDITIONSFOR DETERMINATIONOF TOCOPHEROLS Component Column Stationary phase
Mobile phase
Column temperature Flow Pressure Injection Detection
Integrator
Calculation Standard
Injection Retention time
Run time/sample
Straight-phase HPLC
Reversed-phase HPLC
Hibar RT (125 x 4 mm) LiChrosorb Si 60, 5/zm, combined with guard column 20 × 4 mm Si 60, 5/zm (Stagroma, Wallisellen, Switzerland) 3% 1,4-Dioxane in n-hexane
Stainless steel (250 × 4.6 mm) Ultrasphere ODS 5/xm (Beckman No. 235329, San Ramon, CA)
Ambient 1.6 ml/min Approximately 50 bar 10-80/xl Excitation: 295 nm Emission: 330 nm Spectrofluorometer 650-10 LC (Perkin-Elmer, Norwalk, CT) Spectra Physics SP 4270 (San Jose, CA) or Multichrom, VG Laboratory Systems (Altrincham, Cheshire, England) External standard method, peak area all-rac-c~-Tocopherol, 1/~g/ml in n-hexane containing 0.01% BHT as antioxidant 20/zl t~-Tocopherol 4.3 miD a-Tocotrienol 5.8 miD /3-Tocopherol 6.9 miD /3-Tocotrienol 9.4 miD y-Tocopherol 7.7 miD y-Tocotrienol 10.5 rain 8-Tocopherol 11.7 miD &Tocotrienol 16.3 miD 25 miD
Acetonitrile/tetrahydrofuran/ methanol/l% ammonium acetate (684 : 220 : 68 : 28) 28° 1.5 ml/min Approximately 50 bar 100 ~1 Excitation: 298 nm Emission: 328 nm Spectrofluorometer LS40 (PerkinElmer) Multichrom, VG Laboratory Systems
External standard method, peak area all-rac-a-Tocopherol,10 ~g/ml in n-hexane containing 0.01% BHT as antioxidant 100/~1 ct-Tocopherol 7.2 miD a-Tocotrienol 4.1 miD fl-Tocopherol 6.4 miD /3-Tocotrienol 3.7 miD y-Tocopherol 6.4 miD 3,-Tocotrienol 3.7 miD ~-Tocopherol 5.6 miD &Tocotrienol 3.4 miD 22 miD
f r o m s y s t e m to s y s t e m a n d d e p e n d o n the c o l u m n , d e t e c t o r , a n d m o b i l e p h a s e utilized. T h e r e s u l t s for t o c o p h e r o l s a n d t o c o t r i e n o l s are g i v e n in m i c r o g r a m s p e r g r a m o f tissue or m i l l i g r a m s p e r liter o f p l a s m a , t a k i n g into c o n s i d e r a t i o n the s a m p l e a m o u n t a n d the d i l u t i o n u s e d . F o u r c a l i b r a t i o n r u n s w i t h a - T are r o u t i n e l y e x e c u t e d . F o r the p l a s m a a s s a y the c a l i b r a t i o n is
[25]
H P L C METHODS FOR ANALYSIS OF VITAMIN E
299
600 -
5001 I
4ool ~
300
20O ~00 0.0
2.5
5.0
7.5
t0,0 Time
t2.5
t5.0
'i71. 5
(minuLesl
FIG. 2. Separation of a standard mixture by straight-phase HPLC.
performed by treating the standard solution like a plasma sample. The calculation is based on the external standard method using area counts. Results and Discussion The chromatography was monitored using fluorescence. For the quantification of vitamin E this detection mode is superior to using UV for selectivity and sensitivity. In the same straight-phase HPLC system the separation of a-T, /3-T, y-T, and 8-T as well as the a-, /3-, y-, and &tocotrienols can be achieved. Typical chromatograms from straightphase and reversed-phase HPLC of tocopherol and tocotrienol standards as well as a spiked liver tissue using fluorescence detection are shown in Figs. 2-5. According to our findings vitamin E is not distributed homogeneously in liver. Thus, it is absolutely essential to secure a homogenate specimen for reliable analysis. These findings forced us to choose a rather large specimen size. As the antioxidant, ascorbic acid was preferred in place of the frequently used pyrogallol or butylated hydroxytoluene. 12,13Saponification before solvent extraction was proved to release natural tocopherols more efficiently from feedstuffs than extraction before saponification. ~4 12 C. F. Bourgeois, J. Assoc. Off. Anal. Chem. 71, 12 (1988), 13 A. Deschuytere and H. Deelstra, Fresenius' Z. Anal. Chem. 324, 1 (1986). 14 C. H. McMurray, W. J. Blanchflower, and D. A. Rice, J. Assoc. Off. Anal. Chem. 63, 1258 (1980).
300
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[25]
28~ ~x
24(]
mc~
20O
i
160 120
,
,0
,
m
2.0
,
,
41.0
t
,
~ 610
,
I
6.0 Time
,
,
,
I
t0.0
,
,
,
I
12.0
,
,
,
I
t4.0
I
I
I
I
t6.0
,
,
t
,
t6,0
(minutes]
FIG. 3. Separation of a standard mixture by reversed-phase HPLC.
Reproducibility. In tissue the coefficient of variation (CV) for a-T was found to be within 4.63% on 126 double determinations with a concentration range of 0.6 to 60 /zg/g of liver tissue. Livers of various species (calf, cattle, swine, and chicken) were used to determine the CV. It was noted that the CV of calf (-+7.74%) and some chicken livers (-+6.21%) was higher than that of bovine livers (-+3.72%). Obviously the efficiency of extraction depends on the lipid pattern of the liver tissue. Nishiyama et al., 15 also describe a dependency on the amount of lipids present in the analyte and the reproducibility of the extraction of a-T after saponification. The CV of the reversed-phase HPLC method for plasma was found to be -+2.3% for ot-T (n = 14) and -+2.4% for y-T (n = 14) for the multiple determination of a human plasma on the same day. Frozen plasma was assayed on 35 different days to test the day-to-day reproducibility of the method. The following CV values were obtained: -+4.6% for ot-T (n = 35) and -+6.0% for y-T (n = 35). The CV of the straight-phase HPLC method for plasma was determined with a-T only. Day-to-day variation was checked in duplicate with 109 plasma samples within a range of 2-18 mg/ liter and was found to be -+ 1.8%. Recoveries. To determine possible losses occurring during the assay procedure 5 g of rat liver homogenate was spiked prior to the analysis with all-rac-a-T, all-rac-C3-T, all-rac-y-T, all-rac-8-T, all-rac-a-tocotrienol, and all-rac-y-tocotrienol in ethanol. After performing the described assay 15 j. Nishiyama, E. G. Ellison, G. R. Mizuno, and J. R. Chipault, J. Nutr. Sci. Vitaminol. 21, 355 (1975).
[25]
H P L C METHODS FOR ANALYSIS OF VITAMIN E
6°°I --
301
.
6001
a_
J 200
m
.,( uJ m
-
i-
i-
o-~
,
m
z
~Lu
.I
< c.~
m
~ .j
t00
,
0.0
.j
~
•
~
~
2.5
5.0
~
7.5
t0.0 Time
t2.5
t5.0
t7.5
(minutes)
FIG. 4. Separation of a spiked liver sample by straight-phase HPLC.
on straight-phase HPLC, the content of the vitamers and rates of recoveries were calculated. For ot-T recovery was 91.7-100% at a range of 4.64-17.45 /xg/g, for a-tocotrienol 85.8-102.6% at a range of 2.04-8.92 /xg/g, for /3-T 87.8-96.4% at a range of 5.50-15.88 /xg/g, for y-T 93.0-101.8% at a range of 4.73-13.5 ~g/g, for y-tocotrienol 96.4-100.6% at a range of 5.35-16.04 /xg/g, and for 8-T 98.2-103.9% at a range of 4.36-12.45 /xg/g. The recoveries at higher supplementation levels were close to 100%, which is surprising in view of the single extraction step conducted directly from the diluted saponification mixture. Obviously the solvent mixture used for the extraction has an optimal composition.
280
240
200 mo t60
c
12(3
, .0
2.0
,
I 4.0
w
,,,
,
6.0
,
i
I
fl.O Time
,
,
,
I
10.0
,
,
,
I
t2.0
,
,
,
I
t4.0
.
,i
I
t6.0
(minutes)
FIG. 5. Separation of a spiked liver sample by reversed-phase HPLC.
,
,
,
|
1fl.0
302
ANTIOXIDANT
CHARACTERIZATION
AND ASSAY
[26]
The rat liver tissue used for the recovery experiment was not deficient in a-T and y-T; fl-T and 8-T were found only in traces, whereas a-tocotrienol and 3,-tocotrienol could not be detected. The values of 0.22 /zg/g found for fl-T and 0.28/zg/g for 8-T also represent the detection limit under the described conditions. Conclusion Simple, efficient, and robust methods for the simultaneous determination of tocopherols and tocotrienols in tissue and plasma have been developed and evaluated. The vitamers can be quantified in the extract using HPLC running in the straight-phase mode or reversed-phase mode with fluorescence detection. The described analytical methods have proved to be very effective and fast, enabling the processing of large numbers of samples within a short period of time. The methods are highly specific and sensitive, covering a wide range of different concentrations in the samples. It may therefore be applied in routine assays of samples such as tissue or plasma having greatly varying concentrations of the vitamins. Acknowledgments We thank Mrs. C. Brodhag, Mrs. C. Grunenwald, Mrs. D. Hess, and Mr. K. Steiner for technical assistance in development of the assays.
[26] S e p a r a t i o n o f t h e E i g h t S t e r e o i s o m e r s o f
all-rac-a-Tocopherol f r o m T i s s u e s a n d P l a s m a : C h i r a l Phase High-Performance Liquid Chromatography and Capillary Gas Chromatography
By GEORGES RISS, ALFRED W. KORMANN, ERNST GLINZ, WILLI WALTHER, and URS B. RANALDER Introduction Natural a-tocopherol consists of a single stereoisomer, 2R,4'R,8'R-atocopherol (R,R,R-a-TOH). Owing to the chiral centers, totally synthetic tocopherols are composed of equal parts of RRR, RRS, RSR, RSS, SSS, SSR, SRS, and SRR stereoisomers. Synthetic tocopherols, particularly esters such as all-rac-a-tocopheryl acetate (all-rac-~-TAc), are widely used as food and feed additives. METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
302
ANTIOXIDANT
CHARACTERIZATION
AND ASSAY
[26]
The rat liver tissue used for the recovery experiment was not deficient in a-T and y-T; fl-T and 8-T were found only in traces, whereas a-tocotrienol and 3,-tocotrienol could not be detected. The values of 0.22 /zg/g found for fl-T and 0.28/zg/g for 8-T also represent the detection limit under the described conditions. Conclusion Simple, efficient, and robust methods for the simultaneous determination of tocopherols and tocotrienols in tissue and plasma have been developed and evaluated. The vitamers can be quantified in the extract using HPLC running in the straight-phase mode or reversed-phase mode with fluorescence detection. The described analytical methods have proved to be very effective and fast, enabling the processing of large numbers of samples within a short period of time. The methods are highly specific and sensitive, covering a wide range of different concentrations in the samples. It may therefore be applied in routine assays of samples such as tissue or plasma having greatly varying concentrations of the vitamins. Acknowledgments We thank Mrs. C. Brodhag, Mrs. C. Grunenwald, Mrs. D. Hess, and Mr. K. Steiner for technical assistance in development of the assays.
[26] S e p a r a t i o n o f t h e E i g h t S t e r e o i s o m e r s o f
all-rac-a-Tocopherol f r o m T i s s u e s a n d P l a s m a : C h i r a l Phase High-Performance Liquid Chromatography and Capillary Gas Chromatography
By GEORGES RISS, ALFRED W. KORMANN, ERNST GLINZ, WILLI WALTHER, and URS B. RANALDER Introduction Natural a-tocopherol consists of a single stereoisomer, 2R,4'R,8'R-atocopherol (R,R,R-a-TOH). Owing to the chiral centers, totally synthetic tocopherols are composed of equal parts of RRR, RRS, RSR, RSS, SSS, SSR, SRS, and SRR stereoisomers. Synthetic tocopherols, particularly esters such as all-rac-a-tocopheryl acetate (all-rac-~-TAc), are widely used as food and feed additives. METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[26]
SEPARATION OF EIGHT ~-TOCOPHEROL STEREOISOMERS
303
Rat resorption/gestation tests with individual a-TAc stereoisomers revealed biopotencies which ranged from 100% for RRR to 21% for SSR. Furthermore, tissues of rats fed RRR-a-TAc-d6 and SRR-a-TAc-d3 (1 : 1) contained different levels of these stereoisomers, that is, they displayed biodiscrimination. 2 Therefore, a suitable method to distinguish all eight a-TOH stereoisomers in biological samples was required. Accordingly, we complemented our capillary gas chromatography (GC) method for determination of a-TOH stereoisomer pairs I with a chiral phase high-performance liquid chromatography (HPLC) system) This combination was applied to evaluate patterns of all eight a-TOH stereoisomers in rat tissues and plasma after oral all-rac-a-TAc treatment. 4 It involved extraction of a-TOH, acetylation of a-TOH to yield a-TAc, and separation by HPLC with a home-made chiral phase into four stereoisomer pairs (RSR + RSS, RRR + RRS, SSS + SSR, and SRS + SRR). Then, the a-TAc of the HPLC peaks was converted to a-tocopheryl methyl ether (a-T-ME). Finally, the a-T-ME samples were separated by capillary GC into eight individual stereoisomers. In addition, several HPLC purifications had to be applied between these steps. Subsequently, these time-consuming procedures were substantially improved and simplified, in particular by omitting the acetylation of a-TOH and by applying a commercially available HPLC chiral phase. 5 Scheme I summarizes the new method. Sources of Materials and Equipment Vitamin E standard compounds are obtained from the departments of Quality Control or of Vitamin Chemistry at Hoffmann-LaRoche (Basel, Switzerland), and 2,2,5,7,8-pentamethyl-6-chromanol is kindly supplied by H. Schneider and Dr. R. K. M~iller (Hoffmann-La Roche). Reagents and solvents of analytical grade or HPLC grade are purchased from E. Merck (Darmstadt, Germany) or Fluka (Buchs, Switzerland), Reactivials from Pierce/Kontron (Ziarich, Switzerland), and Acrodisc LC13 filters (for the removal of solids in extracts and solutions) from Gelman/ Skan (Basel, Switzerland). i H. Weiser and M. Vecchi, Int. J. Vitam. Nutr. Res. 52, 351 (1982). 2 K. U. Ingold, G. W. Burton, D. O. Foster, L. Hughes, D. A. Lindsay, and A. Webb, Lipids 22, 163 (1987). 3 M. Vecchi, W. Walther, E. Glinz, T. Netscher, R. Schmid, M. Lalonde, and W. Vetter, Heir. Chim. Acta 73, 782 (1990). 4 H. Weiser, G. Riss, and A. W. Kormann, Ann. N . Y . Acad. Sci. 669, 393 (1992); H. Weiser, G. Riss, and A. W. Kormann, J. Nutr. 1994 (in preparation). 5 E. Glinz, W. Walther, R. Maurer, and U. B. Ranalder, poster presented at " H P L C '91 ," Basel, Switzerland, June 3-9, 1991.
304
ANTIOXIDANT CHARACTERIZATION AND ASSAY Step 1:
Extraction of t~-TOH from samples with known a-TOH contents
Step 2:
Purification of extracted c~-TOH by semipreparative HPLC
Step 3:
Conversion of purified c~-TOH to t~-T-ME
Step 4:
Separation by HPLC with Chiralcel OD into five peaks
$
RRS RRR RSR RSS
Step 5:
[26]
(SSR + SSS + SRS + SRR)
Separation by capillary GC SSR SSS SRS SRR
SCHEME I. Determination of all eight individual c~-tocopherol stereoisomers.
Chromatographic equipment is obtained from Dani (Monza, Italy), E. Merck-Hitachi, Petrarch Systems [Bristol, PA (Roth, Basel, Switzerland)], Alltech/Socolabo (Pully, Switzerland), Daicel/Stehelin (Basel, Switzerland), Hewlett-Packard (Basel, Switzerland), Kontron, MachereyNagel (Oensingen, Switzerland), Phase-Sep/Brunschwiler (Basel, Switzerland), Stagroma (Wallisellen, Switzerland), Varian (Basel, Switzerland), and Waters/Brechbiihler (Schlieren, Switzerland).
Methods
Determination of o~-Tocopherol Contents of Biological Samples To obtain optimal results, purifications of tissue o~-TOH for stereoisomer analyses have to be carried out with known amounts of o~-TOH. Procedures for isolation of a-TOH from tissue homogenates with sodium dodecyl sulfate (SDS) and application of 2,2,5,7,8-pentamethyl-6-chromanol as an internal standard for HPLC analysis of ct-TOH levels are adapted from published methods. 6,7 Alternatively, other methods can be applied to isolate a-TOH and to determine its levels in samples. 8 We use an HPLC system with equipment from Kontron: Control unit 450MT2, pump 420, autosampler 460, UV detector 430 at 290 nm, and fluorescence detector SFM 25 (excitation 290 nm, emission 330 nm). Aliquots of 50/zl extract are injected onto a precolumn with Lichrosorb Si60, 5/z, 0.4 × 2.5 cm (Merck), and a column with Nucleosil-NH2, 3/z, 6 G. W. Burton, A. Webb, and K. U. Ingold, Lipids 20, 29 (1985). 7 T. Ueda and O. Igarashi, J. Micronutr. Anal. 3, 185 (1987). 8 j. K. Lang, M. Schillaci, and B. Irvin, in "Modern Chromatographic Analysis of Vitamins" (A. P. De Leenheer, W. E. Lambert, and H. J. Nelis, eds.), 2nd Ed., pp. 153-195. Dekker, New York and Basel, 1992.
[26]
S E P A R A T I O N O F E I G H T Ot-TOCOPHEROL S T E R E O I S O M E R S
305
0.45 x 25 cm (Macherey-Nagel). The mobile phase is n-hexane/tert-butyl methyl ether/2-propanol (94 : 6 : 0.05, v/v/v) at a flow rate of 2 ml/min. Recovery of the internal standard was 97.5 - 3.7% (n = 4). The mean coefficient of variation of 12 duplicate a-TOH analyses was 1.18%. Calculations of a-TOH present in peaks are based on an ,~.1~ ~ cm value of 87 at 290 rim. Extraction o f Tissue or Plasma a-Tocopherol Extraction procedures for determinations of a-TOH stereoisomers in rat tissues and plasma (Scheme I, step 1) are initiated with a total amount of 3-40/xg a-TOH per sample, that is, tissue weights and plasma volumes are chosen according to established a-TOH levels (see above). Procedures are given below for liver, brain, adipose tissue, and plasma with normal a-TOH levels. If these levels are low (e.g., in vitamin E-deficient animals), tissues or plasma from the same experimental group have to be pooled. In that case, sample weights or aliquots and volumes of reagent solutions have to be adapted accordingly. Extraction. A 0.50-g aliquot of tissue homogenate (liver, brain, or adipose tissue) or 1.0 ml plasma is mixed with 1.0 ml of 0.1 M aqueous SDS in a 30-ml glass tube, and the contents are vigorously stirred (e.g., vortex mixing) for 1 min. One milliliter methanol containing 0.01% butylated hydroxytoluene (BHT) is added, the contents are briefly homogenized (e.g., Polytron homogenizer, Kinematica, Luzern, Switzerland), and the tube is rinsed with 1.0 ml methanol. Lipids are extracted with 5.0 ml n-heptane containing 0.01% BHT by vigorous shaking for 15 min, and solids are removed by low-speed centrifugation (10 min, room temperature). The supernatant is collected, solids are washed with n-heptane, then centrifuged again, and the combined n-heptane fractions are dried with Na2SO 4. Solvents are evaporated under N2, the residue is dissolved in 1.0 ml methanol/tetrahydrofuran (1 : 1, v/v), and extract aliquots are further purified by repeated HPLC (Step 2) prior to conversion of a-TOH to a-T-ME. In some cases, for example, for lipid-rich tissues, the volumes of methanol/tetrahydrofuran have to be increased to avoid problems with viscosity or solubility. These procedures for a-TOH isolation from rat tissues need to be modified for analyses of pig liver and feed samples because extraction with SDS leads to low yields of a-TOH. A 2 g aliquot is homogenized in 20 ml methanol with 2% L-ascorbic acid, 8 ml 50% KOH are added, and the sample is hydrolyzed at 70° for 30 min. The hydrolysate is cooled, filtered through glass wool, and transferred to an Extrelut 20 column (Merck). a-TOH is eluted by 2 x 120 ml iso-octane, the eluant is dried by Na2SO 4 and evaporated under N2. The residue is dissolved in n-hexane
306
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[26]
with 2% ethyl acetate and transferred to a column with 5 g Kieselgel Si60 (size 0.063-0.2 mm; Merck) previously washed with the same solvent. Impurities are removed by 20 ml n-hexane with 2% ethyl acetate, and ot-TOH is eluted by 50 ml n-hexane with 5% ethyl acetate. 9 The eluant is evaporated under Nz, the pellet is dissolved and purified by HPLC (see Scheme I, step 2). In contrast to the first extraction procedure (see above), only one HPLC purification is required, and the total time for all procedures up to chiral phase HPLC (see Scheme I, step 4) is reduced by approximately 50%.
Purification of a-Tocopherol Extracts by High-Performance Liquid Chromatography Aliquots of a-TOH extracts (-0.1 ml) are purified (Scheme I, step 2) by the following HPLC system: Control unit Anacomp 220, pumps 414-T, autosampler 660T or manual injection, UV detector Uvikon 722LC at 290 nm (all Kontron); column Radial-Pak C18, 10 ~, 0.8 x 10 cm with /xBondapak C18 Guard-Pak from Waters, and methanol as mobile phase at a flow rate of 2 ml/min. Peaks containing a-TOH are collected and pooled, and the solvent is evaporated under Nz. The HPLC step can be repeated or complemented with other systems if a-TOH samples are not sufficiently pure for the subsequent derivatization (Scheme I, step 3).
Derivatization of Purified ot-Tocopherol to a-Tocopherol Methyl Ether In a 1-ml screw-capped tube equipped with a valve (e.g., Reacti-vial, Pierce), approximately 10/xg a-TOH is dissolved in 50/~1 monoglyme (ethylene glycol dimethyl ether). The solution is stirred during the whole procedure. Twenty-five microliters of 50% KOH is added dropwise, the vial is closed and flushed with argon, and then 30/zl dimethyl sulfate is added dropwise. After 1 hr at room temperature, 15/xl dimethyl sulfate is added. After an additional 2 hr, solvents are evaporated under N2, 100 /xl doubly distilled water is added, a-T-ME is extracted with twice 500/.d n-hexane and centrifugation, and the n-hexane in the combined extracts is evaporated under N 2 . Derivatization (Scheme I, step 3) of a typical a-TOH tissue extract normally leads to a yield of approximately 40% of a-T-ME. This can be checked by a GC system which separates a-TOH (retention time 40.6 min), a-T-ME (39.4 min), and a-TAc (41.4 min). The following conditions are used: GC unit Dani 3800, column coated with PS 086 (12-15% phenylpolymethylsiloxane; Petrarch), length 15 m, inner diameter 0.3 mm. The 9 A. W. K o r m a n n , G. Riss, and A. Rippstein, J. Lipid Res. 21, 780 (1980).
[26]
S E P A R A T I O N O F E I G H T Ot-TOCOPHEROL S T E R E O I S O M E R S
307
column temperature is programmed from 50° to 330° with a gradient of 5°/min. Analytes are dissolved in ethyl acetate, 1/~l is injected onto a programmed temperature vaporizer, and the flame ionization detector is set at 330°. Peaks are integrated by a Varian DS 654 data system.
Separation of a-Tocopherol Methyl Ether Stereoisomers by Chiral Phase High-Performance Liquid Chromatography To protect the chiral phase, it is advisable to include a purification step prior to the chiral separation. This can be achieved with a HPLC system consisting of a 25 x 0.4 cm column of Spherisorb S3-W (PhaseSep) and a mobile phase of n-hexane containing 0.5% methyl tert-butyl ether at a flow rate of 1 ml/min. Chiral phase HPLC (Scheme I, step 4) is performed with the following system: Pump L-6200 from Merck-Hitachi and a septum injector; column 25 x 0.46 cm with Chiralcel OD from Daicel with n-hexane as mobile phase at a flow rate of 1.5 ml/min; detection at 220 nm by a UV-Vis detector L-6200 from Merck-Hitachi and integration by Varian DS 654. Figure 1 shows a typical separation of a-T-ME stereoisomers derived from a rat liver extract. HPLC of all-rac-a-T-ME should yield peaks with 50.0% for peak 1 (all 2S stereoisomers) and 12.5% for the four peaks with 2R stereoisomers. Four injections of all-rac-a-T-ME led to a mean of 50.2% with an interassay variation of 0.7% for peak 1, and means of 12.3 to 12.5% and interassay variations of 0.7 to 1.4% for the other peaks. For the evaluation of intra-assay variation, a mixture of 4.0 mg allrac-~-TOH and 6.0 mg RRR-u-TOH was derivatized to a-T-ME. This mixture should result in peaks with 20% for (all 2S), 65% for RRR, and 5% each for RRS, RSR, and RSS stereoisomers. Ten injections led to means of 19.6%, 66.1%, and 4.7 to 4.8%, respectively, and the variation coefficients for all peaks were -<0.9%. Alternatively, a 25 x 0.4 cm column of Nucleosil 1000-5 (MachereyNagel) coated with ( + )-poly(triphenylmethyl methacrylate) can be used. 3
Capillary Gas Chromatography for Separations of 2S-o~-Tocopherol Methyl Ether Stereoisomers The GC system for separation of 2S-o~-T-ME stereoisomers (Scheme I, step 5) has been described previously. 3 It consists of a GC unit HP 5890 (Hewlett-Packard), an injector operating in splitless mode at 260°, and a glass capillary of 100 m length and 0.3 mm inner diameter, dynamically coated with Silar 10C (Alltech). The column is operated isothermally at 165°, the flame ionization detector at 220°. Analytes dissolved in 1/xl ethyl
308
[26]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
50-
4 45-
> E
(D C 0 C~
35--
3O-
1 25-
2O-
15--
i0
20
30
Time
[min]
40
50
FIo. 1. Separation of ~-tocopheryl methyl ether (a-T-ME) stereoisomers from a purified rat liver extract by HPLC with Chiralcel OD. Peaks were identified by cochromatography with standard ~-T-ME stereoisomers: (1) all four 2S stereoisomers (i.e., SSR + SSS + SRS + SRR); (2) RSS; (3) RRS; (4) RRR; (5) RSR. Extractions, purifications, and derivatizations were carried out as indicated in the text (Scheme I, steps I to 3), and chromatographic conditions were as described in the section on separation of a-T-ME stereoisomers by chiral phase HPLC (Scheme I, step 4).
acetate are injected, and peaks are integrated by a Varian DS 654 data system. Figure 2 demonstrates the separation of the four 2S-a-T-ME stereoisomers after collection of peak 1 from HPLC with Chiralcel OD (Fig. 1). The products described above for chiral phase HPLC were also used to evaluate the performance of the GC system. All-rac-t~-T-ME should
[26]
309
S E P A R A T I O N OF E I G H T Ot-TOCOPHEROL STEREOISOMERS
[sis] [SRRI
[SSS]
[SSR]
r
I
I
I
I
I
I
122
124
126
128
130
132
rain
FIG. 2. Capillary GC of the four 2S-a-tocopheryl methyl ether stereoisomers after separation from a rat liver extract with HPLC on Chiralcel OD. Peak 1 from the Chiralcel OD column as shown in Fig. 1 was collected and separated by capillary GC as described in the text (Scheme I, step 5). Peaks were identified by cochromatography with a-T-ME stereoisomer standards. yield four peaks with 25% each. Ten injections led to peaks with means ( - S D ) of 24.81 to 25.22 (--- 0.28 to 0.41)%. The mixture with 4.0 mg allrac-a-T-ME and 6.0 mg RRR-a-T-ME should result in peaks of 70% for (RRR + SSS), and 10% each for (RRS + SRR), (RSR + SRS), and (RSS + SRR). With nine injections, means of 69.44 (---0.81)% for (RRR + SSS), and 9.87 to 10.42 (-+0.40 to 0.67)% for the other three peaks were obtained. Alternatively, a 100 m x 0.25 m m column, supplied ready coated with cyanosilicone oil Silar 10C by Quadrex Corp. (Woodbridge, CT) can be used. 10 Concluding R e m a r k s In c o m p a r i s o n to the initial assay, 1the new method as described a b o v e offers major advantages. All separation steps are performed with a - T - M E derivatives, that is, only one a - T O H derivative is required. Chiralcel OD, a commercially available chiral H P L C phase, is able to separate all four 2 R - a - T - M E stereoisomers individually from a peak containing the four 2S stereoisomers (Fig. 1). This information will be sufficient for m a n y ~0N. Cohen, C. G. Scott, C. Neukom, R. J. Lopresti, G. Weber, and G. Saucy, Heir. Chim. Acta 64, 1158 (1981).
310
ANTIOXIDANT
CHARACTERIZATION
AND ASSAY
[27]
purposes, for example, to check the origin of a dietary supplement. In any case, individual 2S forms can be measured by the same GC system as in the older method if the distribution pattern of all eight a-TOH stereoisomers has to be known (see Fig. 2 and Weiser et al.4). Ueda et al. 11 developed a chiral phase HPLC method with a Chiralpak O P ( + ) column which separates all-rac-a-TAc into four peaks containing (all 2R), ( S S S + S S R ) , S R R and S R S stereoisomers. Therefore, the system of Ueda et al. 11 yields only limited information with regard to a-TOH stereoisomer distribution, whereas our combination of HPLC and GC methods permits to determine all eight a-TOH stereoisomers individually. As mentioned, Ingold et al. used a GC-mass spectrometry (GC-MS) method to differentiate between R , R , R - a - T O H - d 6 and S,R,R-a-TOH-d3 in rat tissues. 2 It is not feasible to follow this experimental protocol for an evaluation of all-rac forms because it would require at least eight differently deuterated a-TAc stereoisomers and a very complex GC-MS setup to distinguish all individual stereoisomers. it T. Ueda, H. Ichikawa, and O. Igarashi, J. Nutr. Sci. Vitam. 39, 207 (1993).
[27] T o c o p h e r o n e a n d E p o x y t o c o p h e r o n e P r o d u c t s o f Vitamin E Oxidation B y D A N I E L C . LIEBLER
Introduction a-Tocopherol (vitamin E) is the principal chain-breaking antioxidant in most biological membranes and in lipoproteins.l,2 As a chain-breaking antioxidant, it reacts primarily with peroxyl radicals, which are the principal chain-carrying species in biological lipid peroxidation 3 [Eq. (1)]. The a-TH + LOO. ~ a-T. + LOOH
(1)
reaction yields the tocopheroxyl radical (a-T0, a resonance-stabilized phenoxyl radical that does not readily propagate radical chain reactions. I L. J. Machlin, in "Handbook of Vitamins" (L. J. Machlin, ed.), 2nd Ed., p. 99. Dekker, New York, 1991. 2 L. Packer and J. Fuchs, eds., "Vitamin E in Health and Disease." Dekker, New York, 1993. 3 G. W. Burton and K. U. Ingold, Acc. Chem. Res. 19, 194 (1986).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
310
ANTIOXIDANT
CHARACTERIZATION
AND ASSAY
[27]
purposes, for example, to check the origin of a dietary supplement. In any case, individual 2S forms can be measured by the same GC system as in the older method if the distribution pattern of all eight a-TOH stereoisomers has to be known (see Fig. 2 and Weiser et al.4). Ueda et al. 11 developed a chiral phase HPLC method with a Chiralpak O P ( + ) column which separates all-rac-a-TAc into four peaks containing (all 2R), ( S S S + S S R ) , S R R and S R S stereoisomers. Therefore, the system of Ueda et al. 11 yields only limited information with regard to a-TOH stereoisomer distribution, whereas our combination of HPLC and GC methods permits to determine all eight a-TOH stereoisomers individually. As mentioned, Ingold et al. used a GC-mass spectrometry (GC-MS) method to differentiate between R , R , R - a - T O H - d 6 and S,R,R-a-TOH-d3 in rat tissues. 2 It is not feasible to follow this experimental protocol for an evaluation of all-rac forms because it would require at least eight differently deuterated a-TAc stereoisomers and a very complex GC-MS setup to distinguish all individual stereoisomers. it T. Ueda, H. Ichikawa, and O. Igarashi, J. Nutr. Sci. Vitam. 39, 207 (1993).
[27] T o c o p h e r o n e a n d E p o x y t o c o p h e r o n e P r o d u c t s o f Vitamin E Oxidation B y D A N I E L C . LIEBLER
Introduction a-Tocopherol (vitamin E) is the principal chain-breaking antioxidant in most biological membranes and in lipoproteins.l,2 As a chain-breaking antioxidant, it reacts primarily with peroxyl radicals, which are the principal chain-carrying species in biological lipid peroxidation 3 [Eq. (1)]. The a-TH + LOO. ~ a-T. + LOOH
(1)
reaction yields the tocopheroxyl radical (a-T0, a resonance-stabilized phenoxyl radical that does not readily propagate radical chain reactions. I L. J. Machlin, in "Handbook of Vitamins" (L. J. Machlin, ed.), 2nd Ed., p. 99. Dekker, New York, 1991. 2 L. Packer and J. Fuchs, eds., "Vitamin E in Health and Disease." Dekker, New York, 1993. 3 G. W. Burton and K. U. Ingold, Acc. Chem. Res. 19, 194 (1986).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[27]
VITAMINE OXIDATIONPRODUCTS
311
HO~~,,,,,,
/
~
"0"
016H33
"016H33
/
o~-tocopherol
tocopheroxylradical
Reduction of a-T. to a-tocopherol by biochemical reductants may complete a one-electron redox cycle, which is proposed to account for apparent antioxidant synergism between a-tocopherol and other antioxidants. 4,5 The a-T. molecules that do not complete redox cycles may react with a second peroxyl radical to form other products [eq. (2)]. In consuming
(2)
a-T. + LOO. --~ products
a second peroxyl radical, reaction (2) contributes to the antioxidant effects of a-tocopherol. Reaction (2) removes a-T. from the system, leads to a-tocopherol oxidative turnover, and may yield marker products for a-tocopherol function in biological systems. This chapter describes methods to characterize and analyze 8a-substituted tocopherones and epoxytocopherones, which are the principal products formed by peroxyl radicalmediated oxidations of a-tocopherol. Oxidation of a-Tocopherol by Peroxyl Radicals Studies in homogeneous solutions and lipid bilayers indicate that reaction (2) yields several products. Peroxyl radicals add to the 8a position of a-T. in a radical-radical termination reaction to form 8a-(alkyldioxy)toc o p h e r o n e s ( 1 ) , 6-8 which are structurally analogous to adducts formed
C16H33
016H33 1
2
4 E. Niki, Chem. Phys. Lipids 44, 227 (1987). 5 L. Packer and V. E. Kagan, in "Vitamin E in Health and Disease" (L. Packer and J. Fuchs, eds.), p. 179. Dekker, New York, 1993. 6 R. Yamauchi, T. Matsui, Y. Satake, K. Kato, and Y. Ueno, Lipids 24, 204 (1989). 7 j. Winterle, D. Dulin, and T. Mill, J. Org. Chem. 49, 491 (1984). 8 D. C. Liebler, P. F. Baker, and K. L. Kaysen, J. Am. Chem. Soc. 112, 6995 (1990).
312
[27]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
between peroxyl radicals and other phenolic antioxidants. 9,1° 8a-(Alkyldioxy)tocopherones (1) may hydrolyze to 8a-hydroxytocopherone (2), which in turn rearranges to ot-tocopheryl quinone (a-TQ)." Intermediate 2 also may be formed directly from a-T-, possibly by electron transfer or disproportionation reactions of o~-T..128a-Hydroperoxytocopherone(3), a
16H33, ,,c
C16H33 3
(x-tocopherylquinone
minor product, is formed by oxygen addition to a-T., followed by hydrogen abstraction. 8,'3 Tocopherones 1-3 typically account for only about 30-50% of o~-tocopherol consumed. 8,14 The remaining products, 4a,5-epoxy-8ahydroperoxytocopherone (4) and 7,8-epoxy-8a-hydroperoxytocopherone
C16H33 '
4
00~ c ....
6
C16H33
O'[ ~ H
OOH
16H33
5
00~ C
16H33
7
(5), are formed by an as yet unidentified mechanism that may involve an 8a-peroxyl radical intermediate.8 The epoxides readily hydrolyze to the corresponding epoxyquinones, 2,3-epoxy-a-tocopherylquinone (6) and 5,6-epoxy-ot-tocopherylquinone (7). 8 9 C. E. Boozer, G. S. Hammond, C. S. Hamilton, and J. N. Sen, J. Am.Chem. Soc. 77, 3233 (1955). 10 E. C. Horswill and K. U. Ingold, Can. J. Chem. 44, 263 (1966). " D. C. Liebler, K. L. Kaysen, and T. A. Kennedy, Biochemistry 28, 9772 (1989). 12 D. C. Liebler and J. A. Burr, Biochemistry 31, 8278 (1992). 13 M. Matsuo, S. Matsumoto, Y. Itaka, and E. Niki, J. Am. Chem. Soc. 111, 7179 (1989). 14 D. C. Liebler, K. L. Kaysen, and J. A. Burr, Chem. Res. Toxicol. 4, 89 (1991).
[27]
VITAMIN E OXIDATIONPRODUCTS
313
In addition to these products, oxidations of a-tocopherol and its model compound 2,2,5,7,8-pentamethylchroman-6-ol in hexane ~5 or benzene 16 also yielded dimer and trimer products. The dimer and trimer products are not formed in significant amounts by peroxyl radical oxidations in polar organic solvents 8 or in oxidations in lipid bilayers, 14and their analysis is not described here.
Preparation of 8a-Alkyldioxytocopherones and Epoxytocopherones 8a-(Alkyldioxy)tocopherones can be prepared easily from a-tocopherol by oxidation in homogeneous solution by azo initiators such as 2,2'azobis(2,4-dimethylvaleronitrile) (AMVN). 6-8 If another oxidizable substrate is not included, the alkyl portion of the 8a-alkyldioxy moiety is derived from the initiator (as in product 1). Procedure
In a typical reaction, a-tocopherol (1 mg, 2.35/xmol) and AMVN (10.8 mg, 43.6/zmol) are dissolved in 7 ml oxygenated acetonitrile and heated in a sealed screw-capped tube for 3 hr at 50°. The reaction mixture is then cooled and concentrated under nitrogen to approximately 0.5-1.0 ml. Epoxytocopherones and 8a-(alkyldioxy)tocopherones are then isolated by preparative high-performance liquid chromatography (HPLC) on a Spherisorb ODS-2, 5/~m, 4.6 × 250 mm column eluted with methanol/1 N sodium acetate, pH 4.25 (93 : 7, v/v), at 1.5 ml/min. Epoxytocopherones 4 and 5, which elute between 8 and 12 min, and 8a-(alkyldioxy)tocopherones 1, which elute between 22 and 30 min, are detected by UV absorbance at 254 nm. Smaller amounts of 8a-hydroperoxytocopherone (3) may elute at approximately 15-17 min, just prior to a-tocopherol ( - 1 8 min). Four diastereomers of 1 comprise the 8a-(alkyldioxy)tocopherone fraction, whereas multiple diastereomers of both 4 and 5 comprise the epoxytocopherone fraction. Products may be isolated from the HPLC mobile phase by extraction with hexane. Diastereomers of I may be resolved completely by HPLC on a Spherisorb silica, 5/~m, 4.6 × 250 mm column eluted with hexane/2-propanol (99.5 : 0.5, v/v) at 1 ml/min. Resolution of 4 and 5 on this normal-phase HPLC system had not been reported but appears to be feasible.
15 D. C. Liebler, J. A. Burr, S. Matsumoto, and M. Matsuo, Chem. Res. Toxicol. 6, 351 (1993). I6 R. Yamauchi, T. Matsui, K. Kato, and Y. Ueno, Agric. Biol. Chem. 12, 3257 (1989).
314
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[27]
Comments Treatment of the reaction mixture with 1 volume of ethanol and 0.5 volume of 1 N HCI for 30 rain at room temperature, followed by extraction of the products with hexane and preparative HPLC as described above, offers a convenient method of preparing standards of a-TQ and epoxyquinones 6 and 7. 8a-(Alkyldioxy)tocopherones derived from other peroxyl radicals may be prepared by azo compound-initiated oxidations provided that another oxidizable substrate is present in approximately 100- to 1000fold molar excess over a-tocopherol. For example, methyl linoleate peroxyl radicals were isolated from AMVN oxidation of a-tocopherol in neat methyl linoleate. 17 The HPLC analysis of 8a-(alkyldioxy)tocopherones with large, hydrophobic 8a-substituents requires a methanol/ethyl acetate mobile phase. The AMVN-initiated oxidations in alcohols yield 8a-(alkyldioxy)tocopherones 2 and 8a-alkoxy derivatives of epoxides 4 and 5. 6 Similarly, oxidations in aqueous/organic mixtures yield 8a-(alkyldioxy)tocopherones 1, epoxides 4 and 5, and the respective hydrolysis products a-TQ and epoxyquinones 6 and 7. 8 Epoxides 4 and 5 appear to be more labile to solvolysis reactions than 8a-(alkyldioxy)tocopherones 1. Standards of 8a-hydroxytocopherone (2) and 8a-hydroperoxytocopherone (3) may be prepared by chemical oxidation 18 and photooxidation, 19 respectively. 8a-Substituted tocopherones 1-3 and epoxytocopherones 4 and 5 display characteristic UV absorbance maxima at approximately 240 and 255 nm, respectively. 6'8 In electron ionization or chemical ionization mass spectrometry of 8a-(alkyldioxy)tocopherones and epoxytocopherones, molecular ions are typically weak or absent. 6'8 Prominent ions at m/z 445 and 430 result from loss of the 8a-substituents from compounds 1 and 4/5, respectively. The absolute configuration at C-8a in 8a-(alkyldioxy)tocopherones derived from RRR-a-tocopherol may be deduced from the 1H nuclear magnetic resonance (NMR) chemical shift of the 2methyl protons. 6 In the C-8a(S) configuration, in which the 8a-alkyldioxy moiety and the 2-methyl protons are trans with respect to the ring system, the 2-methyl protons occupy a shielding environment above the dienone ring and resonate at approximately I. 1 ppm. In the C-8a(R) configuration, the 8a-alkyldioxy moiety forces the 2-methyl group away from the dienone, and these protons resonate instead at approximately 1.3 ppm.
17 R. Yamauchi, T. Matsui, K. Kato, and Y. Ueno, Lipids 25, 152 (1990). 18 W. Durckheimer and L. A. Cohen, J. Am. Chem. Soc. 86, 4388 (1964). 19 R. L. Clough, B. G. Yee, and C. S. Foote, J. Am. Chem. Soc. 101, 683 (1979).
[27]
VITAMIN E OXIDATIONPRODUCTS
315
Reduction of 8a-Substituted Tocopherones to a-Tocopherol Facile reduction of 8a-substituted tocopherones 1-3 to a-tocopherol by ascorbate or nordihydroguaiaretic acid (NDGA) provides an indirect method of measuring tocopherone formation. 1t,12Although both 8a-substituted tocopherones and a-T. are reduced to a-tocopherol in organic/aqueous mixtures, tocopherones are reduced only below pH 4, whereas a-T. is most readily reduced at approximately pH 7. 2o Reduction to a-tocopherol may be the most convenient means of measuring 8a-hydroxytocopherone (2), which readily rearranges to a-TQ. 11'12a-TQ itself is not reduced to a-tocopherol, but instead yields a-tocopheryl hydroquinone. Procedure
A 0.5-ml sample of tissue homogenate or microsomal, mitochondrial, or liposome suspension containing approximately 0.5 nmol a-tocopherol and/or a-tocopherol oxidation product(s) is added to a borosilicate glass test tube containing 1.125 ml ethanol, ! /zmol deferoxamine, 20 nmol butylated hydroxytoluene (BHT), and 200/zmol ascorbate in 0.5 ml of 300 mM sodium formate, pH 3. After 30 min at room temperature, 0.5 nmol 8-tocopherol (an internal standard for HPLC) and 0.5 ml each of ethanol and water are added, and the mixture is extracted with three 1-ml portions of hexane. The extracts are evaporated under nitrogen and analyzed by reversed-phase HPLC as described above. Sensitive detection of tocopherols is achieved by electrochemical detection with the detector set at +0.35 V. Comment
The addition of BHT to the reduction reactions improves a-tocopherol recovery, probably by suppressing any secondary a-tocopherol oxidation that may occur. Deferoxamine is added to chelate transition metals that may cause adventitious oxidation of a-tocopherol or its oxidation products. NDGA (50-200/zmol) may be substituted for ascorbate; other phenolic reductants may also suffice. To determine whether a-tocopherol is regenerated from an 8a-substituted tocopherone, rather than a-T., a parallel reaction can be done in which a 50 mM Tris-HCl, pH 7, buffer is substituted for the formate buffer. Preferential reduction at pH 3 suggests that the reducible species is an 8a-substituted tocopherone, rather than a-T.. The pH of the ascorbate/buffer solution should be checked and 20 K. Mukai, M. Nishimura, and S. Kikuchi, J. Biol. Chem. 266, 274 (1991).
316
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[28]
readjusted (if necessary) after adding the ascorbate. Because ascorbate oxidizes readily, this solution should be prepared immediately before use. Acknowledgments I thank Jeanne A. Burr and Kathy Kaysen for contributions to this work, which was supported in part by U.S. Public Health Service Grant CA49743.
[28] L i g h t - I n d u c e d G e n e r a t i o n o f V i t a m i n E R a d i c a l s : Assessing Vitamin E Regeneration
By
VALERIAN E . KAGAN a n d LESTER PACKER
Introduction Vitamin E is the major-lipid-soluble chain-breaking antioxidant of blood plasma and of membranes, l However, the vitamin E concentration in membranes and lipoproteins is extremely low relative to the concentrations of phospholipids in membranes or lipids in lipoproteins. 2 Efforts to explain the very high efficiency of vitamin E (tocopherols) in protecting polyunsaturated lipids in membranes and lipoproteins against oxidative LOO. + Toc-OH--> LOOH + Toc-O.
(1)
damage by peroxyl (alkoxyl) radicals via reaction (1) resulted in the concept of " a free radical reductase," a hypothetical enzymatic system capable of regenerating vitamin E by catalyzing the reduction of its phenoxyl free radicalrcductasc
Toc-O. + Red-H
> Toc-OH + Red.
(2)
radical [reaction (2)]. 3-5 Direct measurements of the vitamin E phenoxyl radical and its reduction can be performed by electron spin resonance (ESR) spectroscopy. Different enzymatic and nonenzymatic oxidation systems have been successfully used for generating steady-state concentrations of the vitamin E phenoxyl radical high enough to be detected by l G. W. Burton, A. Joyce, and K. U. Ingold, Arch. Biochem. Biophys. 221, 281 (1983). 2 H. H. Draper, in "Vitamin E in Health and Disease" (L. Packer and J. Fuchs, eds.), p. 53. Dekker, New York, 1993. 3 p. B. McCay, D. D. Gibson, K. L. Fong, and K. R. Hornbrook, Biochim. Biophys. Acta 431, 459 (1976). 4 R. F. Burk, M. J. Trumble, and R. A. Lawrence, Biochim. Biophys. Acta 618, 35 (1980). 5 j. M. C. Reddy, R. W. Scholz, C. E. Thomas, and E. J. Massaro, Life Sci. 31, 571 (1982).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
316
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[28]
readjusted (if necessary) after adding the ascorbate. Because ascorbate oxidizes readily, this solution should be prepared immediately before use. Acknowledgments I thank Jeanne A. Burr and Kathy Kaysen for contributions to this work, which was supported in part by U.S. Public Health Service Grant CA49743.
[28] L i g h t - I n d u c e d G e n e r a t i o n o f V i t a m i n E R a d i c a l s : Assessing Vitamin E Regeneration
By
VALERIAN E . KAGAN a n d LESTER PACKER
Introduction Vitamin E is the major-lipid-soluble chain-breaking antioxidant of blood plasma and of membranes, l However, the vitamin E concentration in membranes and lipoproteins is extremely low relative to the concentrations of phospholipids in membranes or lipids in lipoproteins. 2 Efforts to explain the very high efficiency of vitamin E (tocopherols) in protecting polyunsaturated lipids in membranes and lipoproteins against oxidative LOO. + Toc-OH--> LOOH + Toc-O.
(1)
damage by peroxyl (alkoxyl) radicals via reaction (1) resulted in the concept of " a free radical reductase," a hypothetical enzymatic system capable of regenerating vitamin E by catalyzing the reduction of its phenoxyl free radicalrcductasc
Toc-O. + Red-H
> Toc-OH + Red.
(2)
radical [reaction (2)]. 3-5 Direct measurements of the vitamin E phenoxyl radical and its reduction can be performed by electron spin resonance (ESR) spectroscopy. Different enzymatic and nonenzymatic oxidation systems have been successfully used for generating steady-state concentrations of the vitamin E phenoxyl radical high enough to be detected by l G. W. Burton, A. Joyce, and K. U. Ingold, Arch. Biochem. Biophys. 221, 281 (1983). 2 H. H. Draper, in "Vitamin E in Health and Disease" (L. Packer and J. Fuchs, eds.), p. 53. Dekker, New York, 1993. 3 p. B. McCay, D. D. Gibson, K. L. Fong, and K. R. Hornbrook, Biochim. Biophys. Acta 431, 459 (1976). 4 R. F. Burk, M. J. Trumble, and R. A. Lawrence, Biochim. Biophys. Acta 618, 35 (1980). 5 j. M. C. Reddy, R. W. Scholz, C. E. Thomas, and E. J. Massaro, Life Sci. 31, 571 (1982).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
[28]
U V LIGHT-INDUCED VITAMIN E RADICALS
317
ESR. They include peroxidase plus hydrogen peroxide, lipoxygenase plus unsaturated fatty acid, azo initiators of peroxyl radicals plus stable organic radicals like diphenylpicrylhydrazyl, etc. 6 Using these methods, it has been demonstrated that ascorbate and reduced thiols (e.g., dihydrolipoic acid and glutathione) were efficient in quenching the vitamin E phenoxyl radical ESR signal in liposomes, low density lipoproteins, and membranes (liver microsomes, mitochondria, and submitochondrial particles).7-9 In addition, enzyme-dependent mechanisms were identified in membranes which were able to reduce the vitamin E phenoxyl radical (to quench the ESR signal of the vitamin E radical generated by oxidation systems). 10-12In particular, the electron transport enzymatic systems (NADPH- and NADH-dependent enzymes in microsomes as well as NADH- or succinate-dependent enzymes in mitochondria) prevented an appearance of the vitamin E phenoxyl radical signal in the ESR spectra until these substrates were fully consumed. It was concluded that several pathways operate to regenerate vitamin E, rather than a single "free radical reductase" as initially hypothesized. However, there is an inherent problem in the interpretation of these ESR results because they do not provide direct evidence for reduction of the vitamin E phenoxyl radical. Quenching and delay in the appearance of the vitamin E phenoxyl radical ESR signal in the presence of an oxidation system (e.g., lipoxygenase + linolenic acid) may be due to the interaction of the reductant with peroxyl radicals [reaction (3)], and not to the reduction of the vitamin E radical in reaction (2). As a result, the ESR measurements of the vitamin E phenoxyl radical quenching do not proROO. + Red-H ~ ROO-H + Red.
(3)
vide for an unequivocal interpretation when peroxyl radicals or other oxidizing radicals are involved in the generation of vitamin E radicals in reaction (1).
6 R. J. Melhorn, J. Fuchs, S. Sumida, and L. Packer, this series, Vol. 186, p. 197. 7 E. Niki, J. Tsuchiya, R. Tanimura, and Y. Kamiya, Chem. Lett. p. 789 (1982). 8 M. A. Scarpa, M. Rigo, M. Maiorino, F. Ursini, and C. Gregolin, Biochim. Biophys. Acta 8111, 215 (1984). 9 V. E. Kagan, E. A. Serbinova, T. Forte, G. Scita, and L. Packer, J. Lipid Res. 33, 385 (1992). l0 L. Packer, J. J. Maguire, R. J. Melhorn, E. A. Serbinova, and V. E. Kagan, Biochem. Biophys. Res. Commun. 159, 229 (1989). II j. j. Maguire, V. E. Kagan, E. A. Serbinova, B. A. C. Ackrell, and L. Packer, Arch. Biochem. Biophys. 292, 47 (1992). lz V. E. Kagan, E. A. Serbinova, and L. Packer, Arch. Biochem. Biophys. 282, 221 (1990).
318
ANTIOXIDANT CHARACTERIZATION AND ASSAY
1 I
lO GAUSS
[28]
1
I
FIG. 1. Vitamin E phenoxyl radical ESR signal generated by UVB-irradiation of vitamin E (a-tocopherol) in methanol-water dispersion (trace I) and in dioleoylphosphatidylcholine (DOPC) liposomes (trace 2). Conditions: (1) a-Tocopherol (3.0 mM) was dispersed in a methanol-water mixture (4:1, v/v); (2) the DOPC concentration was 20 mM and the atocopherol concentration 1.8 mM.
Ultraviolet Light-Induced Generation of Vitamin E Phenoxyl Radical Vitamin E (a-tocopherol) has an a b s o r b a n c e m a x i m u m at 295 nm. Thus vitamin E itself m a y absorb U V light (UVB light) and b e c o m e a free radical (the p h e n o x y l radical): 6 T o c - O H + hv---->Toc-O. + e - + H +
(4)
In this w a y the vitamin E p h e n o x y l radical can be generated in the absence of any other oxidizing agent. The advantage of the U V B induction of vitamin E radicals is the absence of other radicals (peroxyl, alkoxyl) in the s y s t e m which m a y interact with the reducing agents. 13'14 A n o t h e r advantage is that the U V B - i n d u c e d generation of the vitamin E radicals and termination of the vitamin E oxidation can be p e r f o r m e d by switching the light on or off. The t o c o p h e r o x y l radical E S R signal appears immediately on U V B irradiation of m e t h a n o l - w a t e r suspensions and dioleoylphosphatidylcholine (DOPC) liposomes containing vitamin E (a-tocopherol) (Fig. 1). In 13 V. E. Kagan, E. Witt, R. Goldman, G. Scita, and L. Packer, Free Radical Res. C o m m u n . 16, 51 (1992).
14V. E. Kagan, A. Shvedova, E. A. Serbinova, S. Khan, C. Swansson, R. Powell, and L. Packer, Biochem. Pharmacol. 44, 1637 (1992).
[28]
UV LIGHT-INDUCEDVITAMINE RADICALS
319
the dark, there was no ESR signal in the region of the tocopheroxyl radical in these preparations. UVB-irradiation of a suspension of human low density lipoproteins (LDL) in the ESR spectrometer cavity resulted in the appearance of a characteristic vitamin E phenoxyl radical ESR signal (Fig. 2). The phenoxyl radical was generated in LDL from endogenous vitamin E and was not detected in the dark. The magnitude of the vitamin E radical signal was a function of the LDL concentration in the suspensions and the vitamin E content in the LDL. UVB-induced ESR signals of the phenoxyl radical generated from endogenous vitamin E could be enhanced by the addition of exogenous a-tocopherol or its homologs. The signals from exogenous tocopherols had the same characteristic parameters in the ESR spectra. Similarly, when membrane suspensions containing either high endogenous vitamin E concentrations (owing to dietary supplementation) or exogenously added vitamin E were UVB-irradiated, the signal of the vitamin E phenoxyl radical could be readily detected in the ESR spectrum, and its reactions with reductants could be studied. Interactions of Reductants with Ultraviolet Light-Induced Vitamin E Phenoxyl Radical Reductants which are able to reduce the vitamin E phenoxyl radical should prevent the appearance of its signal in the ESR spectra. Addition
I IOGAUSS
]
J FIG. 2. ESR signals of the vitamin E phenoxyl radical and ascorbyl radical induced in human low density lipoproteins (LDL) by UVB-irradiation. The ascorbyl radical signal was recorded 2 min after the addition 0.5 mM ascorbate to human LDL (15 mg protein/ml; endogenous vitamin E content 6 nmol/mg protein).
320
ANTIOXIDANT
CHARACTERIZATION AND ASSAY
[29]
of ascorbate to LDL results in disappearance of the endogenous UVBinduced vitamin E radical ESR signal and in the appearance of the ascorbyl radical signal (Fig. 2). Similarly, in liposomes containing a-tocopherol suspended in a solution of ascorbate, there was a delay in the appearance of the tocopheroxyl signal on UVB-irradiation. During this delay the ascorbyl radical signal could be detected. Neither the ascorbyl radical signal nor the vitamin E phenoxyl radical signal was detected in the ESR spectra in the dark. In hairless mouse skin homogenates with exogenously added vitamin E, UVB-irradiation produced a characteristic ESR signal of the ascorbyl radical signal generated from endogenous ascorbate. This signal decayed over time and was substituted by the vitamin E phenoxyl radical ESR signal. Similarly, the UVB-induced vitamin E phenoxyl radical signal generated in rat liver microsomes was quenched by NADPH and reappeared only after NADPH oxidation. In all these cases, the UVB light-induced vitamin E phenoxyl radical was the only radical species that could interact with endogenous or exogenous reductants. This provided direct and unequivocal evidence for the regeneration of vitamin E via reduction of its phenoxyl radical by enzymatic or nonenzymatic pathways.
[29] A n t i o x i d a t i v e A c t i v i t y of T o c o t r i e n o l in H e t e r o g e n e o u s S y s t e m : Indication of Restriction within M e m b r a n e b y Fluorescence Measurement B y MASAKAZU YAMAOKA a n d K A N K I KOMIYAMA
Introduction Antioxidative activities of the vitamin E homologs tocotrienol (Toc3) and tocopherol (Toc) have been compared in both homogeneousI-5 and 1 A. Seher and St. A. Ivanov, Fette, Seifen, Anstrichm. 75, 606 (1973). 2 N. Yanishlieva-Maslarova, A. Popov, A. Seher, and S. A. Ivanov, Fette, Seifen, Anstrichm. 79, 357 (1977). 3 M. Nogala-Kalucka, M. Gogolewski, and A. Luczynski, Rocz. Akad. Roln. Poznaniu 141, 105 (1982). 4 M. Yamaoka, A. Tanaka, and A. Kato, Yukagaku 34, 120 (1985). 5 M. Yamaoka, Yukagaku 34, 184 (1985).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
320
ANTIOXIDANT
CHARACTERIZATION AND ASSAY
[29]
of ascorbate to LDL results in disappearance of the endogenous UVBinduced vitamin E radical ESR signal and in the appearance of the ascorbyl radical signal (Fig. 2). Similarly, in liposomes containing a-tocopherol suspended in a solution of ascorbate, there was a delay in the appearance of the tocopheroxyl signal on UVB-irradiation. During this delay the ascorbyl radical signal could be detected. Neither the ascorbyl radical signal nor the vitamin E phenoxyl radical signal was detected in the ESR spectra in the dark. In hairless mouse skin homogenates with exogenously added vitamin E, UVB-irradiation produced a characteristic ESR signal of the ascorbyl radical signal generated from endogenous ascorbate. This signal decayed over time and was substituted by the vitamin E phenoxyl radical ESR signal. Similarly, the UVB-induced vitamin E phenoxyl radical signal generated in rat liver microsomes was quenched by NADPH and reappeared only after NADPH oxidation. In all these cases, the UVB light-induced vitamin E phenoxyl radical was the only radical species that could interact with endogenous or exogenous reductants. This provided direct and unequivocal evidence for the regeneration of vitamin E via reduction of its phenoxyl radical by enzymatic or nonenzymatic pathways.
[29] A n t i o x i d a t i v e A c t i v i t y of T o c o t r i e n o l in H e t e r o g e n e o u s S y s t e m : Indication of Restriction within M e m b r a n e b y Fluorescence Measurement B y MASAKAZU YAMAOKA a n d K A N K I KOMIYAMA
Introduction Antioxidative activities of the vitamin E homologs tocotrienol (Toc3) and tocopherol (Toc) have been compared in both homogeneousI-5 and 1 A. Seher and St. A. Ivanov, Fette, Seifen, Anstrichm. 75, 606 (1973). 2 N. Yanishlieva-Maslarova, A. Popov, A. Seher, and S. A. Ivanov, Fette, Seifen, Anstrichm. 79, 357 (1977). 3 M. Nogala-Kalucka, M. Gogolewski, and A. Luczynski, Rocz. Akad. Roln. Poznaniu 141, 105 (1982). 4 M. Yamaoka, A. Tanaka, and A. Kato, Yukagaku 34, 120 (1985). 5 M. Yamaoka, Yukagaku 34, 184 (1985).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[29]
FLUORESCENCE ASSAY FOR VITAMIN E HOMOLOGS
321
TABLE I INHIBITION OF PHOSPHOLIPID OXIDATION BY TOCOTRIENOLS AND TOCOPHEROLSa
(/zM)
DLPC b (mM)
AAPH c (raM)
Induction
Antioxidant
Concentration
ot-Toc d a-Toc e a-Toc3 e a-Toc3 d a-Toc d a-Toc e a-Toc3 e a-Toc 3 d y-Toc d y-Toc e y-Toc3 e 3,-Toc3 d
0.73 0.73 0.73 0.73 1.48 1.47 1.47 1.48 4.04 3.95 3.95 4.30
0.30 0.30 0.30 0.30 0.30 0.30 0.30 0.30 1.93 1.93 1.93 1.93
0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 10.0 10.0 10.0 10.0
2940 4800 5280 9600 3390 6840 7200 15,000 72 443 480 780
period (s)
a Adapted from Yamaoka et al.12 b Dilinoleoylphosphatidylcholine. c 2,2'-Azobis(2-amidinopropane) dihydrochloride. d Distributed from water to the liposomes. e Mixed with the liposomal membrane.
heterogeneous 6-13 systems. In a homogeneous system below 60 °, no significant difference was observed in antioxidative activities. 4,5 This can be explained by the similar chemical structures of Toc3 and Toc, except for the unsaturated side chain of Toc3, and their similar intrinsic reactivities with peroxyl radicals. 12 In a heterogeneous system at physiological temperature, however, Toc3 shows a greater antioxidative activity than the corresponding T o c . 6-8A0'13 Our results (Table I) also revealed that the antioxidative activity of Toc3 was greater than that of Toc when Toc3 and Toc were distributed from water to the model membrane, while both were the same when the compounds were in the membrane from the beginning. 11,12It was also 6 T. Tatsuta, Bitamin 44, 185 (1971). 7 M. MinD, S. Nakagawa, and T. Wakabayashi, Jpn. Kokai Tokkyo Koho JP 58 96,021(1983). 8 T. Inada, Osaka Ika Daigaku Zasshi 44, 140 0985). 9 M. Nakano, K. Sugioka, T. Nakamura, and T. Oki, Biochim. Biophys. Acta 619, 274 (1980). l0 K. Komiyama, K. Iizuka, M. Yamaoka, H. Watanabe, N. Tsuchiya, and I. Umezawa, Chem. Pharm. Bull. 37, 1369 (1989). El M. Yamaoka and K. Komiyama, J. Jpn. Oil Chem. Soc. 38, 478 (1989). 12 M. Yamaoka, M. J. H. Carrillo, T. Nakahara, and K. Komiyama, J. Am. Oil Chem. Soc. 68, 114 (1991). 13 E. Serbinova, V. Kagan, D. Hart, and L. Packer, Free Radical Biol. Med. 10, 263 (1991).
322
[29]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
TABLE II CONSUMPTION OF ANTIOXIDANTS DURING INDUCTION PERIOD a Concentration
Induction period
Consumption d
Antioxidant
(/zM) b
(S) c
(%) c
~-Toc3 e a-Toc e y-Toc3 ~
13.5 13.5 14.9 13.5 13.3 13.2 12.5 13.2
2400 180 1800 210 990 720 1230 1050
82 23
y-Toc e a-Toc3 f ~-Toc f y-Toc3 f T-Toc f
81 5 74 70 58 55
a Adapted from Yamaoka et a l ) 2 b Initial concentration of vitamin E. ¢ Values are averages of duplicate runs. The spreads in the induction periods and in the consumption were -+ 120 sec and +5%, respectively. d Percentage of antioxidant consumed during the induction period.
e Distributed from water to the liposomes. f Mixed with the liposomal membrane.
observed that both Toc3 and Toc in a model membrane were not completely consumed during the induction period, and that the longer the induction period, the greater the consumption (Table II).12 In that model membrane, otherwise mixed with phospholipid before multilamellar liposome formation, vitamin E was distributed from water phase to the membrane, incorporated into the membrane, and consumed during the oxidation of the liposomes initiated by the water-soluble oxidant. Although the estimated concentration of o~-Toc was greater than that of o~-Toc3 (Table III14), judging by a one-sided test of statistical hypothesis (significance level 0.05) a-Toc3 was consumed more than o~-Toc (Table II). There have been some reports on the relatively poor antioxidative activity of a-Toc in membranes, ~5-19 and the influence of ot-Toc on the properties of membranes .20,21 We have proposed that owing to some inter14 M. Yamaoka and M. J. H. Carrillo, Chem. Phys. Lipids 55, 295 (1990). 15 M. E. Leibowitz and M. C. Johnson, J. Lipid Res. 12, 662 (1971). 16 E. Niki, M. Takahashi, and E. Komuro, Chem. Lett., p. 1573 (1986). 17 T. Doba, G. W. Burton, and K. U. Ingold, Biochim. Biophys. Acta 835, 298 (1985). 18 R. J. Mehlhorn, S. Sumida, and L. Packer, J. Biol. Chem. 264, 13448 (1989). 19 V. Kagan and P. J. Quinn, Eur. J. Biochem. 171, 661 (1988). 2o K. Fukuzawa, H. Chida, A. Tokumura, and H. Tsukatani, Arch. Biochem. Biophys. 206, 173 (1981). 21 K. Fukuzawa, H. Ikeno, A. Tokumura, and H. Tsukatani, Chem. Phys. Lipids 23, 13 (1979).
[29]
F L U O R E S C E N C E ASSAY F O R V I T A M I N E H O M O L O G S
323
TABLE III DISTRIBUTION OF ANTIOXIDANTS IN LIPID PHASE a
Coct/f°ct c
Antioxidant
Xmax (nm)b
(%)
a-Toc3 a-Toc y-Toc3 y-Toc
292 292.5 297 297
95.60 -+ 0.31 96.21 -+ 0.43 95.10 -+ 0.57 95.12 -+ 0.67
a Adapted from Yamaoka and Carrillo/4 b Maximum wavelength in l-octanol. c Mean _+ standard deviation (n = 4) of the percentage of the residual concentration of the antioxidant in 1-octanol after partitioning between 1-octano! and water (Coct)to the initial concentration (C~ct). action b e t w e e n the m e m b r a n e and vitamin E, Toc3 and Toc are inhibited in the reaction with phospholipid peroxides within the m e m b r a n e . As a first step to verify this speculation, the change in the physicochemical property of the m e m b r a n e by vitamin E partitioning was measured via fluorescence quenching and fluorescence polarization of the probes. 14 In this chapter the procedures used for these m e a s u r e m e n t s are described and results are discussed in relation to the interaction between the m e m brane and the distributed vitamin E, as well as the different antioxidative activities of Toc3 and Toc.
Methods
Measurement of Fluorescence Quenching and Polarization An aliquot of a dioleoylphosphatidylcholine ( D O P C ) - c h l o r o f o r m solution is transferred to an amber-colored round-bottomed flask, and the solvent is r e m o v e d at r o o m t e m p e r a t u r e under reduced pressure with a rotary evaporator. The residual thin film of D O P C is kept under a stream of nitrogen, and 5 m M H E P E S - N a O H buffer solution (pH 7.3) is added. The final concentration of D O P C is 0.414 mM. The flask containing the mixture is packed with nitrogen, sealed with a stopper, shaken for 1 min, and sonicated for 60 min with a Sharp UT-52 bath-type sonicator (50 W) below 10°. The average diameter of the liposomes is around 200 nm as measured by a Coulter N4 submicron particle analyzer.
324
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[29]
The change in the fluorescence intensity of 3,3'-di-propylthiocarbocyanine [diS-C3-(5)] due to the addition of vitamin E is measured by the modified method of Kurihara et al. 22-24 An aliquot of the prepared DOPC liposome solution is diluted with 5 mM H E P E S - N a O H buffer solution (pH 7.3) to prepare liposomes of various concentrations. To a 3-ml aliquot of the dilute liposome solution is added 3/xl of diS-C3-(5), and the mixture is stirred continuously for 30 min at 30 °. The final concentration of diS-C3-(5) is 0.91/xM. The fluorescence intensity is measured using a FP 770 Jasco spectrofluorophotometer (excitation at 622 nm and emission at 680 nm); the cell holder is kept at 30° during the measurement. After measuring the initial fluorescence intensity (I0) of the liposome solution, vitamin E (as 10 mM ethanol solution) is added to the solution while stirring at 30°. To ensure that Toc3 or Toc is uniformly distributed, the mixture is kept standing at 30° for 5 min before measuring the fluorescence intensity (IQ). The interval is determined in advance by measuring the time course of the fluorescence intensity every minute during the 5-min interval after each addition of a-Toc. The change in the fluorescence polarization of 1,6-diphenyl- 1,3,5-hexatriene (DPH) caused by the addition of vitamin E is measured by modified method of Shinitzky and Barenholz. 25 A solution of DPH in tetrahydrofuran (1 mM) is diluted with water to make a 0.1 mM aqueous solution and is added to the DOPC liposome solution. The mixture is continuously stirred for 3 hr at 30°. The final molar ratio of DPH to DOPC is 1 : 1,000. The polarization is measured at 30° by a FP770 Jasco spectrofluorophotometer (excitation at 360 nm and emission at 428 nm with the spectrum bandwidth set at 10 nm) equipped with diffraction gratings (1800 lines/ mm) and FP 2010 polarizers at both the excitation and emission sides. At the emission side, wavelengths below 390 nm are cut off and are not detected with this equipment. Using the following relationship, the polarization (P) of the DPH fluorescence is calculated: P = (III- I±)/(III + I±)
(I)
where Irl and I Lare the fluorescence intensities detected through the polarizer oriented parallel and perpendicular, respectively, to the direction of polarization of the excitation beam.
z2 T. Nomura and K. Kurihara, Biochemistry 26, 6135 (1987). 23 T. Nomura and K. Kurihara, Biochemistry 26, 6141 (1987). 24 T. Kumazawa, T. Nomura, and K. Kurihara, Biochemistry 27, 1239 (1988). 25 M. Shinitzky and Y. Barenholz, J. Biol. Chem. 249, 2652 (1974).
[29]
FLUORESCENCE ASSAY FOR VITAMIN E HOMOLOGS
3
_o
325
t
2
10-6
10 -s
10-4
Concentration of Antioxidant(M) FIG. l. Changes in fluorescence intensity (Io/IQ) versus concentration of antioxidant. Values are averages of four measurements for ((3) a-Toc, (e) a-Toc3, ([2) y-Toc, and (11) y-Toc3. The concentration of DOPC was 0.051 mM, and that of diS-C3-(5) was 0.91 /xM. (Data adapted from Yamaoka and Carrillo/4)
Fluorescence Quenching and Changes in Fluorescence Polarization by Vitamin E Distribution to Membranes Plots of the change in fluorescence intensity of diS-C3-(5 ) (IO/IQ) versus the concentrations of Toc3 and Toc are shown in Fig. 1. The degree of the change in the fluorescence intensity decreased in the order a-Toc > a-Toc3.3,-Toc also decreased the fluorescence intensity of diS-Ca-(5) more than y-Toc3. The change in fluorescence polarization of DPH by vitamin E partitioning is shown in Table IV. The results suggest that vitamin E tends to make the DOPC liposomes less fluid. The degree of change in fluorescence polarization by Toc is larger than that with Toc3. For each type of vitamin E homolog, the physiochemical property of the membrane changed in relation to the concentration of the antioxidant, although the number of methyl groups at the chromanol nucleus is also related to the change in membrane-forming properties. It was reported that the change in fluorescence intensity of diS-C3-(5) in the model membrane reflected the potential-dependent partition of the dye between the inside and the outside of the vesicle. 26 The change in membrane potential caused by odorants was measured using diS-C3-(5) 26 A. S. Waggoner, C. H. Wang, and R. L. Tolles, J. Membr. Biol. 33, 109 (1977).
326
[29]
ANTIOXIDANT CHARACTERIZATION AND ASSAY T A B L E IV CHANGES IN FLUORESCENCE POLARIZATION OF DIPHENYLHEXATRIENE BY ADDITION OF VITAMIN E a Antioxidant
(Concentration (/zM)
a-Toc a-Toc a-Toc3 a-Toc3 y-Toc y-Toc y-Toc3 7-Toc3
3.3 16.7 3.3 16.6 3.4 16.7 3.2 9.5
P/Pob 1.247 1.502 1.080 1.357 1.267 1.581 1.070 1.283
++ -+ -+ -+ -+ -+ -+
0.179 0.228 0.185 0.028 0.182 0.313 0.207 0.131
a A d a p t e d from Y a m a o k a and Carrillo. TM b M e a n - standard deviation of four m e a s u r e m e n t s (n = 4) of the ratio o f the c h a n g e d polarization in the presence of D P H (P) to the initial polarization (P0).
and compared with the data from fluorescence polarization of D P H . 22-24 The odorants might induce the conformational change within the membrane, alter the orientation of the fixed charges and dipoles within the membrane, and change the phase boundary potential at the cis side of the membrane. On the other hand, the increased fluorescence polarization of DPH reflects the decreased fluidity of the membrane. Is there any relationship between the fluorescence quenching and the increase in fluorescence polarization? By considering that the change in fluorescence intensity of diS-C3-(5) caused by odorants was not always related to the change in fluorescence polarization of DPH, the sites where the fluorescence intensity of diS-C3-(5) changed and where the polarization of the fluorescence of DPH changed may be different. 22-24 We can estimate, however, that the partitioning of vitamin E causes some conformational changes in phospholipid membranes as well as a decrease in the membrane fluidity of DOPC liposomes.
Conclusion From these results, we have concluded that the difference in the antioxidative activities among Toc3 and Toc in the membrane can be explained in the following way. The partitioning of Toc3 and Toc causes some conformational change in the phospholipid membrane which can be observed as changes in the physicochemical properties of the membrane,
[30]
VITAMIN E COMPOUNDS IN PLATELETS AND RED CELLS
327
and the extent of the changes decreases in the order Toc > Toc3. Through these processes, Toc3 and Toc may show different inhibition when reacting with peroxide within the membrane.
[30] D e t e r m i n a t i o n o f T o c o p h e r o l s a n d T o c o p h e r o l q u i n o n e in H u m a n R e d B l o o d Cell a n d P l a t e l e t S a m p l e s B y G O V I N D T . VATASSERY
Introduction Prolonged vitamin E deficiency in the rat results in a variety of hematological problems that include higher platelet and reticulocyte counts, increased platelet aggregation in vitro, and normocytic anemia.l'2 The susceptibility of red cells to hemolysis in the presence of dialuric acid was proposed as a laboratory test for the assessment of vitamin E nutritional status by Gyorgy and Rose. 3 Even though this test is rarely used, the importance of vitamin E for maintaining the biological integrity of red cells is well established. Platelets have been employed in investigations dealing with assessment of nutritional status in humans. 4'5 One of the oxidation products of vitamin E is tocopherolquinone which exhibits interesting biological activities. Tocopherolquinone has been reported to be an antisterility factor in the vitamin E-deficient male hamster and rat. 6 Mackenzie et al. 7 have shown that a-tocopherolquinone is active in alleviating muscular dystrophy in vitamin E-deficient rabbits. Tocopherol as well as its oxidation products have also been found to inhibit platelet aggregation under in vitro conditions. 8'9 We have attempted to use the concentrations of tocopherol and tocopherolquinone as an index of the oxidative status of membranes with respect 1 L. Machlin, R. F. Filipski, A. L. Willis, D. C. Kuhn, and M. Brin, Proc. Soc. Exp. Biol. Med. 149, 275 (1975). 2 G. F. Combs, Proc. Nutr. Soc. 40, 187 (1981). 3 p. Gyorgy and C. S. Rose, Ann. N.Y. Acad. Sci. 52, 231 (1949). 4 j. Lehmann, Am. J. Clin. Nutr. 34, 2104 (1981). 5 G. T. Vatassery, A. M. Krezowski, and J. H. Eckfeldt, Am. J. Clin. Nutr. 37, 1020 (1983). 6 S. I. Mauer and K. E. Mason, J. Nutr. 105, 491 (1975). 7 j. B. Mackenzie, H. Rosenkrantz, S. Ulick, and A. T. Milhorak, J. Biol. Chem. 183, 655 (1950). 8 A. C. Cox, G. H. R. Rao, J. M. Gerrard, and J. G. White, Blood 55, 907 (1980). 9 R. Mower and M. Steiner, Prostaglandins 24, 137 (1982).
METHODS IN ENZYMOLOGY.VOL. 234
Copyright © 1994by AcademicPress. Inc. All rights of reproductionin any form reserved.
[30]
VITAMIN E COMPOUNDS IN PLATELETS AND RED CELLS
327
and the extent of the changes decreases in the order Toc > Toc3. Through these processes, Toc3 and Toc may show different inhibition when reacting with peroxide within the membrane.
[30] D e t e r m i n a t i o n o f T o c o p h e r o l s a n d T o c o p h e r o l q u i n o n e in H u m a n R e d B l o o d Cell a n d P l a t e l e t S a m p l e s B y G O V I N D T . VATASSERY
Introduction Prolonged vitamin E deficiency in the rat results in a variety of hematological problems that include higher platelet and reticulocyte counts, increased platelet aggregation in vitro, and normocytic anemia.l'2 The susceptibility of red cells to hemolysis in the presence of dialuric acid was proposed as a laboratory test for the assessment of vitamin E nutritional status by Gyorgy and Rose. 3 Even though this test is rarely used, the importance of vitamin E for maintaining the biological integrity of red cells is well established. Platelets have been employed in investigations dealing with assessment of nutritional status in humans. 4'5 One of the oxidation products of vitamin E is tocopherolquinone which exhibits interesting biological activities. Tocopherolquinone has been reported to be an antisterility factor in the vitamin E-deficient male hamster and rat. 6 Mackenzie et al. 7 have shown that a-tocopherolquinone is active in alleviating muscular dystrophy in vitamin E-deficient rabbits. Tocopherol as well as its oxidation products have also been found to inhibit platelet aggregation under in vitro conditions. 8'9 We have attempted to use the concentrations of tocopherol and tocopherolquinone as an index of the oxidative status of membranes with respect 1 L. Machlin, R. F. Filipski, A. L. Willis, D. C. Kuhn, and M. Brin, Proc. Soc. Exp. Biol. Med. 149, 275 (1975). 2 G. F. Combs, Proc. Nutr. Soc. 40, 187 (1981). 3 p. Gyorgy and C. S. Rose, Ann. N.Y. Acad. Sci. 52, 231 (1949). 4 j. Lehmann, Am. J. Clin. Nutr. 34, 2104 (1981). 5 G. T. Vatassery, A. M. Krezowski, and J. H. Eckfeldt, Am. J. Clin. Nutr. 37, 1020 (1983). 6 S. I. Mauer and K. E. Mason, J. Nutr. 105, 491 (1975). 7 j. B. Mackenzie, H. Rosenkrantz, S. Ulick, and A. T. Milhorak, J. Biol. Chem. 183, 655 (1950). 8 A. C. Cox, G. H. R. Rao, J. M. Gerrard, and J. G. White, Blood 55, 907 (1980). 9 R. Mower and M. Steiner, Prostaglandins 24, 137 (1982).
METHODS IN ENZYMOLOGY.VOL. 234
Copyright © 1994by AcademicPress. Inc. All rights of reproductionin any form reserved.
328
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[30]
to vitamin E. l°'ll This involves determination of the two compounds in the same sample. A few publications dealing with the analysis of tocopherolquinone can be found in the literature. 12-15 The method recommended in this report is the latest modification of our earlier procedure. 16
Assay Method Samples of red cells or platelets are saponified, and the tocopherol derivatives are extracted with hexane. The isolated compounds are then separated and quantitated using high-performance liquid chromatography (HPLC). Reagents. All chemicals and solvents used are reagent or chromatography grade from standard suppliers. Absolute ethanol is redistilled prior to use. Standard tocopherols and quinone are from Eastman Organic Chemicals (Rochester, NY). Solutions. The following solutions are used for the assay: butylated hydroxytoluene (BHT), 0.025% (w/v) in redistilled ethanol; ascorbic acid, 30% (w/v), and pyrogallol, 25% (w/v), in deionized water. Stock solutions (100 mg/100 ml) of a-, Y-, and 8-tocopherols, tocopherolquinone, and tocol are made in redistilled ethanol. The stock standards are stable for months if kept in the freezer at - 40°. Dilute working solutions of standards can be made in ethanol containing 0.025% BHT, and the diluted stock solutions are stable in the refrigerator for several weeks. Collection o f Blood and Isolation o f Red Cells and Platelets. It should be kept in mind that platelet function and biochemical properties are very susceptible to a number of drugs. Therefore special care has to be taken to ascertain that blood is obtained in a drug-flee state. Factors that influence the properties of platelets such as drug treatment, the method of drawing blood, and storage conditions have been discussed in an earlier volume in this series.17 Blood samples for the isolation of platelets are drawn into tubes containing sodium citrate (1.25 ml of 3.8% sodium citrate per 12 ml of blood). The blood is mixed gently for a few minutes and transferred into polypropylene tubes (16 × 100 ram) and centrifuged at 250 g for 15 min. The l0 G. T. Vatassery, Biochim. Biophys. Acta 926, 160 (1987). 11 G. T. Vatassery, Lipids 24, 299 (1989). i2 S. K. Howeland and Y. M. Wang, J. Chromatogr. 227, 174 (1982).
13D. D. Stump, E. E. Roth, Jr., and H. S. Gilbert,J. Chromatogr. 306, 371 (1984). 14G. A. Pascoe, C. T. Duda, and D. Reed, J. Chromatogr. 414, 440 (1987). 15M. E. Murphy and J. P. Kehrer, J. Chromatogr. 421, 71 (1987). 16G. T. Vatassery and W. E. Smith, Anal. Biochem. 167, 411 (1987). 17M. B. Zucker, this series, VoL 169, p. 117.
[30]
VITAMIN E COMPOUNDS IN PLATELETS AND RED CELLS
329
platelet-rich plasma (PRP) is then drawn out with a Pasteur pipette, mixed with EDTA (1 ml of 0.2 M disodium EDTA per 10 ml of PRP), and centrifuged at 2000 g for 15 min. The platelet pellet is then washed twice with phosphate-buffered saline (PBS) at pH 7.4 containing 4 mM EDTA. Platelets are resuspended in the appropriate medium required for the specific oxidation experiments in vitro. Blood samples for collection of red blood cells are drawn into Vacutainers (purple top, 16 x 100 mm; evacuated blood collection tubes made by Becton Dickinson and Company, Rutherford, NJ) containing disodium EDTA as anticoagulant and centrifuged at 2600 g for 15 min. Plasma and the buffy coat on top of the red cells are removed. The cells are washed twice with PBS, pH 7.4. The washed red cells can then be used for analysis. Saponification and Sample Treatment. Two milliliters ethanol containing 0.025% BHT, 0.1 ml of 30% ascorbic acid, and 0.2 ml of 25% pyrogallol are pipetted into saponification tubes (16 x 125 mm Teflon-capped culture tubes). The red cell or platelet samples are then added. With red cell samples it is preferable to add the samples to tubes containing the saponification medium while being vortexed. One milliliter of 10% (w/v) potassium hydroxide solution is added, and the mixture is saponified at 60° for 30 min. The tubes are cooled and 2 ml of water is added. Two milliliters of hexane containing 0.025% BHT is then added. Tocopherols and quinone are extracted into the hexane phase by vortexing the mixture for 1 min. The hexane phase is separated out and evaporated under a stream of nitrogen. The residue is redissolved in mobile phase and injected on the chromatographic column. Chromatography of Tocopherol and Tocopherolquinone. Instruments used for HPLC consist of the following components: Model 126 pump, Model 507 autosampler, and System Gold software (Beckman Instruments, San Ramon, CA) supported by an IBM Model 80-386 computer. Tocopherol compounds are detected electrochemically with a Coulochem 5100 A detector (ESA Inc., Bedford, MA) using the various cells and potentials as follows: 5011 analytical cell with detector 1 at -0.25 V and detector 2 at +0.55 V and 5021 conditioning cell at -0.75 V. These electrochemical conditions are modified from the recommendations of Murphy and Kehrer ~5and involve reduction of tocopherolquinone at one electrode followed by oxidation of the resultant species at the second electrode. Chromatographic conditions are as follows: column, ultrasphere ODS, 5 tzm, 4.6 x 250 mm from Beckman Instruments; mobile phase, 6% buffer, 12% acetonitrile, and 82% methanol with the mixture containing 7.5 mM NaH2PO 4 . H20 (final concentration); flow rate, 3 ml/ min. Representative retention times of the various tocopherol derivatives under these conditions are the following: ~-tocopherol, 15.6 min; y-tocoph-
330
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[30]
erol, 12.5 min; 8-tocopherol, 10.4 min; tocopherolquinone, 9.4 rain; and tocol (internal standard), 7.5 min. A typical chromatogram obtained during analysis of a sample of platelets is shown in Fig. 1. General Comments
All operations involving platelets are performed at room temperature (20 °) using siliconized glassware or plastics. Resuspension of platelets is done gently, employing plastic Pasteur pipettes. It is better not to subject the platelet samples to vortex mixing. Red cells may be processed at 4 °. Detection limits for the assay depend largely on the chromatographic detector used. Electrochemical detectors are the most sensitive (detection limit 25 pg injected on the column). Tocopherols can also be estimated using fluorescence (295 nm excitation and 340 nm emission) or UV absorption at 295 nm and tocopherolquinone by UV absorption at 265 nm.
A
3.6-
E i
(~ ¢-)
E
3.4
~ 3.2
i r~ 2.8
. . . .
;
. . . .
. . . .
1;'
'
'
Time (minutes) FIG. l. Chromatogram obtained during the assay of human platelets. A sample of blood was obtained from a human volunteer, and the platelets were isolated and assayed for tocopherols and tocopherolquinone using the recommended procedure. Peak A, Tocol; B, a-tocopherolquinone; C, 8-tocopherol; D, y-tocopherol; E, a-tocopherol.
[31]
URIC AND ASCORBATE/DEHYDROASCORBATERATIO
331
Tocopherols in red cell samples tend to be very labile, and a mixture of the three antioxidants is necessary to avoid losses during sample processing. With platelet samples a mixture of BHT and ascorbate is sufficient. Interestingly, ascorbate is necessary to avoid losses of tocopherols as well as quinone during assays. Use of 8-tocopherol as an internal standard is recommended in some publications. However, 8-tocopherol may occur naturally in small amounts in many samples. For this reason tocol is the preferred internal standard. Finally, Burton et al. z8 have recommended an extraction procedure using sodium dodecyl sulfate (SDS) as a detergent solubilizer of the biological samples. When the proposed method involving saponification was compared with the SDS extraction procedure, the results obtained by both methods were not significantly different with either platelet or red cell samples. is G. W. Burton, A. Webb, and K. U. Ingold, Lipids 20, 29 (1985).
[31] V i t a m i n C, D e h y d r o a s c o r b a t e , a n d U r i c A c i d in T i s s u e s and Serum: High-Performance Liquid Chromatography B y G. BARJA and A. HERNANZ
Introduction Ascorbic acid and uric acid 1 are low molecular weight water-soluble antioxidants present in significant amounts in the tissues and body fluids of humans and other mammals. Their measurement in biological samples is of current interest in biomedical research since they can provide protection against many pathologies implicating free radicals at various stages of their progression. High-performance liquid chromatography (HPLC) offers reliable methods for the measurement of ascorbic and uric acid that should replace less specific colorimetric or electroanalytic ,3 techniques. Two HPLC methods specifically designed for the measurement of these
1 B. N. Ames, R. Cathart, E. Schwiers, and P. Hochstein, Proc. Natl. Acad. Sci. U.S.A. 78, 6858 (1981). 2 p. j. Garry, G. M. Owen, D. W. Lashley, and P. C. Ford, Clin. Biochem. 7, 131 (1974). 3 E. L. McGowen, M. G. Rusnack, M. Lewis, and J. A. Tillotson, Anal. Biochem. 119, 55 (1982).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[31]
URIC AND ASCORBATE/DEHYDROASCORBATERATIO
331
Tocopherols in red cell samples tend to be very labile, and a mixture of the three antioxidants is necessary to avoid losses during sample processing. With platelet samples a mixture of BHT and ascorbate is sufficient. Interestingly, ascorbate is necessary to avoid losses of tocopherols as well as quinone during assays. Use of 8-tocopherol as an internal standard is recommended in some publications. However, 8-tocopherol may occur naturally in small amounts in many samples. For this reason tocol is the preferred internal standard. Finally, Burton et al. z8 have recommended an extraction procedure using sodium dodecyl sulfate (SDS) as a detergent solubilizer of the biological samples. When the proposed method involving saponification was compared with the SDS extraction procedure, the results obtained by both methods were not significantly different with either platelet or red cell samples. is G. W. Burton, A. Webb, and K. U. Ingold, Lipids 20, 29 (1985).
[31] V i t a m i n C, D e h y d r o a s c o r b a t e , a n d U r i c A c i d in T i s s u e s and Serum: High-Performance Liquid Chromatography B y G. BARJA and A. HERNANZ
Introduction Ascorbic acid and uric acid 1 are low molecular weight water-soluble antioxidants present in significant amounts in the tissues and body fluids of humans and other mammals. Their measurement in biological samples is of current interest in biomedical research since they can provide protection against many pathologies implicating free radicals at various stages of their progression. High-performance liquid chromatography (HPLC) offers reliable methods for the measurement of ascorbic and uric acid that should replace less specific colorimetric or electroanalytic ,3 techniques. Two HPLC methods specifically designed for the measurement of these
1 B. N. Ames, R. Cathart, E. Schwiers, and P. Hochstein, Proc. Natl. Acad. Sci. U.S.A. 78, 6858 (1981). 2 p. j. Garry, G. M. Owen, D. W. Lashley, and P. C. Ford, Clin. Biochem. 7, 131 (1974). 3 E. L. McGowen, M. G. Rusnack, M. Lewis, and J. A. Tillotson, Anal. Biochem. 119, 55 (1982).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
332
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[31]
substances in tissues 4 or blood 5 are described. They can be performed with basic HPLC equipment, since elution is isocratic and detection is based in ultraviolet absorption in both cases. The tissue method 4 is adapted here to the simultaneous measurement of ascorbate, dehydroascorbate, and uric acid in the same sample in two consecutive HPLC runs. Measurement of Ascorbic Acid, Uric Acid, 4 and Dehydroascorbate in Tissues
Principle. Ascorbic and uric acid can be efficiently and rapidly separated on standard reversed-phase columns with isocratic elution after addition of appropriate concentrations of a counterion to a buffered waterbased mobile phase. This solves many problems associated with ionexchange chromatography. As maximums of absorption are found around 267 nm for ascorbic acid and around 292 nm for uric acid, the choice of an intermediate wavelength (280 nm) allows the simultaneous measurement of both antioxidants in the same chromatogram with enough sensitivity for most mammalian tissues using a conventional UV detector. 4 Standards. Prepare a solution containing 0.4 mM ascorbic acid and 89 /zM uric acid in 50 mM perchloric acid. The solution is maintained in the cold (5°), protected from light, and promptly injected in the HPLC system every day that samples are processed. Sample Preparation. Tissue samples (100 mg/ml) are homogenized in cold 50 mM perchloric acid, centrifuged at 3000 g for 10 min at 5 °, and filtered through 0.5/xm pore diameter membranes, and 20/zl is directly injected in the HPLC system. The use of other acids with higher acidity and ionic strength for extraction and stabilization of ascorbic and uric acid produces strong reductions of retention times on successive injections and will greatly shorten the useful life of reversed-phase columns. This problem is absent with perchloric acid at the concentration cited without compromising extraction efficiency from the tissue or the stability of the substrates. Chromatography Conditions. The mobile phase is acetonitrile/water (12.5 : 87.5, v/v), pH 5.5, containing 4.3 mM disodium hydrogen phosphate and 1.07 mM myristyltrimethylammonium bromide as counterion. When the mobile phase is isocratically pumped at 0.75 ml/min through a C~8 reversed-phase Nucleosil 7/zm (4.6 × 100 mm) column (Machery-Nagel, Oensingen, Switzerland), retention times for uric and ascorbic acid are around 5.5 and 9 min, respectively (column temperature 25°). Peaks are 4 G. Barja de Quiroga, M. L6pez-Torres, R. P6rez-Campo, and C. Rojas, Anal. Biochem. 199, 81 (1991). 5 R. E. Hughes, Biochem. J. 64, 203 (1956).
[3 II
URIC AND ASCORBATE/DEHYDROASCORBATE RATIO
333
detected at 280 nm for both substances. I f a programmable UV detector is available, further sensitivity can be gained setting the wavelength at 292 nm during the first part of the chromatogram and automatically changing to 267 nm after uric acid has eluted. A total of 25-35 min between injection of two consecutive samples is convenient for ,many tissues tested. It is essential to control carefully the pH of the mobile phase since, as is shown in Table I, an increase in this parameter strongly reduces the retention time of ascorbic acid without affecting that of uric acid. This characteristic of the mobile phase affects more intensely the differential retention of both acids than the acetonitrile or counterion concentration. Thus, pH should be the first parameter to vary when direct application of the method described here to a different sample does not initially allow good resolution from interfering peaks. Comments. The method described here is widely applicable to tissue samples; good, fast chromatographic resolutions have been obtained using rat brain, lung, and liver; pigeon liver; mouse liver; guinea pig liver and lung; and trout liver. A detailed study performed in mouse live# showed 93.9-96.2% stability of both substances in the perchloric acid extract during 8 hr at 5° in the dark, 94.3-96.5% recovery, good reproducibility, and linearity of response in homogenates containing from 1.5 to 30 mg protein/ml. In addition to retention times, peak purity can be assessed with a conventional UV detector after incubating the homogenates with specific degradative enzymes. Thus, total disappearance of the ascorbic acid peak is obtained after 5 min of incubation of a freshly prepared mouse liver homogenate (pH 6.4 at 25°C) with 500 IU/ml of ascorbate oxidase prior to the addition of perchloric acid. Similarly, incubation of a mouse TABLE I EFFECT OF pH OF MOBILE PHASE ON DIFFERENTIAL CHROMATOGRAPHIC SEPARATION OF ASCORBIC AND URIC ACIDa Retention time (min) Compound
pH 6.4
pH 5.9
pH 5.5
Uric acid Ascorbic acid
5.4 5.5
5.4 7.0
5.4 9.1
a HPLC conditions: mobile phase, acetonitrile/water (12.5 : 87.5, v/v) containing 4.3 mM disodium hydrogen phosphate and 1.07 mM myristyltrimethylammonium bromide pumped at 0.75 ml/min; column, Cm Nucleosil 7/xm Macherey-Nagel (4.6 x 100 mm); temperature, 25° -+ 1°.
334
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[31]
liver homogenate (pH 7.4, 25 °) with uricase (5 IU/ml) leads to total disappearance of the uric acid peak. a Dehydroascorbate Analysis. Information about the dehydroascorbate content of the tissues can be obtained by performing the method of Hughes, 5 adapted to HPLC of liver tissue as described here. For this purpose the centrifuged and filtered perchloric acid extract is divided into two aliquots. The first is diluted by a factor of 10 with 50 mM perchloric acid and injected in the chromatograph to measure the reduced form of ascorbate as described above. The second aliquot is carefully neutralized to pH 7.0 using a pH micrometer by adding around 100/xl of 45% K2HPO 4 per milliliter of sample. A portion of the neutralized sample is immediately added to a vial containing a previously weighed amount of DL-homocysteine (e.g., 4 mg). The amount of sample added is that needed to get a final DL-homocysteine concentration of I% (e.g., add 0.4 ml of sample to 4 mg of DL-homocysteine). Shake and incubate for 15 min at 25°. This procedure fully reduces the dehydroascorbate present in the sample to ascorbate (DL-homocysteine is fully effective only at neutral pH). At the end of the 15-min period the sample is diluted by a factor of 10 with 50 mM perchloric acid and is immediately injected in the chromatograph (direct injection of the undiluted samples containing 1% homocysteine causes strong reductions in retention times and shortens the column half-life; these problems are eliminated by decreasing the homocysteine concentration to 0.1%). Now the ascorbate peak represents total ascorbate (reduced plus oxidized forms). Dehydroascorbic acid is calculated by subtracting the amount of reduced ascorbate from total ascorbic acid, and the ratio dehydroascorbic/ascorbic can thus be calculated. To obtain optimal results the reduced ascorbate sample must also be diluted by a factor of 10 with 50 mM perchloric acid before injection in the chromatograph. This is especially important in tissues such as liver where 90% or more of the ascorbic acid is present in the reduced form. We have obtained good, reproducible results with this method using guinea pig liver. Incubation of the "total ascorbate" sample with ascorbate oxidase (see "Comments" above) at the optimum pH of the enzyme leads to total disappearance of the "ascorbate plus dehydroascorbate" peak. This demonstrates that all the increase in peak area from the untreated to the treated (homocysteine) sample is due to reduction of dehydroascorbate and not to reduction of other substances present in the tissue. Thus, this method allows the simultaneous measurement of the three substances, ascorbate, dehydroascorbate, and uric acid, in the same tissue sample in two consecutive HPLC runs. The sensitivity of a conventional UV detector is much higher than needed for accurate detection of these substances in many tissues using the described method.
[31]
URIC AND ASCORBATE/DEHYDROASCORBATE RATIO
335
Measurement of Ascorbic and Uric Acids in Serum 6
Principle. Different high-performance liquid chromatography methods have been used to determine ascorbic acid in serum from humans. 7-9 However, few HPLC methods using UV spectrophotometric detection have been shown to be simple and sensitive. Here, a simple, rapid, and sensitive reversed-phase HPLC procedure for ascorbic and uric acid determination in 0.5 ml serum (or less) using paired-ion chromatography with UV spectrophotometric determination is described. The effects of solvent pH and ascorbic acid oxidation are also examined. With this method analysis of clinical samples in order to investigate presymptomatic decreases in the ascorbic acid as well as antioxidant status can be carried out without difficulty. Standards. Prepare stock solutions of ascorbic acid (6 raM) and uric acid (1 mM) in 50 ml of 0.3 mM trifluoroacetic acid in the presence of 10 mM 1,4-dithioerythritol (DTE) as antioxidant. This solution is stable for 1 week when frozen at - 2 5 °. Ascorbic and uric acid working standards are prepared flesh daily by diluting the stock solutions (1 : 10) with the mobile phase. Samples. Serum samples are prepared from blood collected from the antecubital vein in the presence of 1 mM DTE as antioxidant. A 0.5-ml serum sample is mixed with 0.1 ml of 1.2 M trifluoroacetic acid to obtain a protein-flee extract after centrifugation. To eliminate trifluoroacetic acid, supernatants are desiccated under vacuum with centrifugation, then stored at - 2 5 ° under N2 until analysis. Immediately before being analyzed the serum extracts are reconstituted with 0.2 ml of mobile phase and passed through a 0.45-nm HV filter (Millipore, Bedford, MA); 5-20/zl is injected in the HPLC system. Chromatography Conditions. The mobile phase is 5 mM ammonium formate buffer, pH 6.0, containing 5 mM tetrahexylammonium chloride (Fluka, Ronkonkoma, NY) as a paired-ion reagent in water/methanol (65 : 35, v/v). The addition of tetrahexylammonium chloride to the mobile phase allows ionic compounds, such as ascorbic acid, to be separated on C~8 reversed-phase columns, eliminating problems of precise pH, temperature control, reproducibility, and short column life associated with ion 6 A. Hernanz, J. Clin. Chem. Clin. Biochem. 26, 459 (1988). 7 W. Lee, P. Hamernyik, M. Hutchinson, V. A. Raisys, and R. F. Labb6, Clin. Chem. (Winston-Salem, N.C.) 28, 2165 (1982). 8 T. Iwata, M. Yamaguchi, S. Hara, and M. Nakamura, J. Chromatogr. Biomed. Appl. 344, 351 (1985). 9 A. J. Speek, J. Schrijver, and W. H. P. Schreurs, J. Chromatogr. Biomed. Appl. 305, 53 (1984).
336
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[31]
T A B L E II EFFECT OF 1,4-DITHIOERYTHRITOL ON ASCORBIC ACID LEVELS IN SERUM a Ascorbic acid (t~mol/liter) Sample
With DTE
Without DTE
Serum Aqueous standard
63.4 --- 12.2 55.7 4- 3.5
14.5 -4- 12.2 18.6 - 1.2
Determination in duplicate, with and without 1,4-dithioerythritol (DTE), of ascorbic acid levels (txmol/liter, mean -+- SD) in serum from fifteen healthy adults and in ten different aliquots of an aqueous standard of 58/xmol/ liter ascorbic acid was performed. HPLC conditions: mobile phase, methanol/water (35:65, v/v) containing 5 m M ammonium formiate buffer, pH 6.0, and 5 m M tetrahexylammonium chloride pumped at 0.7 ml/min; column, C18 Novapak 4 /zm Waters (3.6 x 150 mm); temperature, 25 -4- 1°.
exchange. Further filtration with a 0.5-nm FHUP filter (Millipore) was performed. The flow rate of the mobile phase is adjusted to 0.7 ml/min. When the mobile phase is isocratically pumped through a C18 reversedphase Novapak (15 cm × 3.6 mm i.d.) column (Waters, Milford, MA), retention times for ascorbic and uric acids are around 8 and 10 min, respectively (column temperature 25°). Peaks are detected at 265 nm for both substances, or at 254 nm if a filter photometer is available. It is essential to control carefully the pH of the mobile phase because a decrease in mobile phase pH significantly reduces the absorption at 265 nm of ascorbic acid. Comments. The method described here has wide application to serum samples if ascorbic acid oxidation is avoided. With this method it is possible to measure without any risk of oxidation serum ascorbic acid amounts as low as 10 pmol (2 ng), which corresponds to 0.5 txmol/liter if 10/xl of serum extract is injected. Correlation between peak responses and injected ascorbic acid concentrations was found to be linear from 10 to 600 pmol (2 to 110 ng). The day-to-day reproducibility of the total HPLC procedure was determined by measuring in duplicate separate portions of a serum pool on 10 consecutive days. The coefficient of variation obtained was 7. I% for the serum pool having a mean concentration of 50 tzmol/liter. Precision was calculated by measuring a serum pool 10 times in a single run. The coefficient of variation was 5.4%. Analytical recovery was carried out by adding 50/xl of 30/zmol/liter ascorbic acid to 0.5 ml of a serum
[31]
URIC AND ASCORBATE/DEHYDROASCORBATE RATIO
337
pool and measuring the total ascorbic acid on 10 different days. The recovery obtained was 94.5 -+ 6.0 (mean % m SD). In addition to the retention times, peak purity can be assessed by incubating the serum extracts with specific degradative enzymes. Total disappearance of the ascorbic acid peak is obtained after incubation with ascorbate oxidase or H20: . Similarly, incubation with uricase leads to total disappearance of the uric acid peak. 4 The reliable measure of dehydroascorbic content in serum samples is controversial. Whereas some authors 1° have shown that when 1,4-dithiothreitol is added to urine samples the amount of ascorbic acid increased, owing to reduction of dehydroascorbic acid to ascorbic acid, others ll also using 1,4-dithiothreitol have shown minimum amounts of dehydroascorbic acid (only 3% of total ascorbic acid) in human plasma. In agreement with these last authors, Levine et al.~2 have also demonstrated the absence of dehydroascorbic acid in human blood. Schmidt et al. 13have indicated that dehydroascorbic acid is unstable at neutral pH and in the presence of mild biological oxidants, as occurs in many vital biological processes in the living organism, and it is rapidly metabolized. As noted in Table II serum samples, obtained from blood collected without 1,4-dithioerythritol as the antioxidant, as well as standard samples stored without 1,4-dithioerythritol, present lower values of ascorbic acid after being subjected to the entire HPLC method on the same day as blood collection. This sample oxidation has not been described when UV spectrophotometric9 or electrochemical7 detection was used.
10 L. W. Donner and K. B. Hicks, Anal. Biochem. 115, 225 (1981). 11 M. Okamura, Clin. Chim. Acta 103, 259 (1980). 12 M. Levine, K. R. Dhariwal, P. W. Washko, J. DeB Butter, R. W. Welch, Y. Wang, and P. Bergsten, Am. J. Clin. Nutr. 54, 11575 (1991). 13 K. Schmidt, H. Oberritter, G. Bruchelt, V. Hagmaier, and O. Hornig, Int. J. Vitam. Nutr. Res. 53, 77 (1983).
338
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[32]
[32] I n V i v o D e t e r m i n a t i o n o f S u p e r o x i d e a n d V i t a m i n C Radicals Using Cytochrome c and Superoxide Dismutase Derivatives By MASAYASU INOUE and KEIKO KOYAMA
Introduction Most reactive oxygen species rapidly react with various molecules and interfere with cellular functions. 1 Based on experiments using antioxidant enzymes and scavengers for reactive oxygen species and free radicals, such as Cu/Zn-type superoxide dismutase (SOD), glutathione, ascorbic acid, and dimethyl sulfoxide (DMSO), critical roles of these reactive species in the pathogenesis of various diseases have been postulated. 2 However, these antioxidants react with a wide variety of compounds and are involved in various metabolic pathways. Thus, even if these scavengers inhibited tissue injury, the corresponding reactive oxygen species may not always be involved in their pathogenesis. To get direct evidence for the involvement of reactive oxygen species in the pathogenesis of various diseases, these species should be determined quantitatively in vivo.
Cytochrome c (Cyt c) has been used for determining superoxide radicals in vitro. 3 Because cytochrome c reacts not only with superoxide radicals but also with other compounds with reducing activity and serves as a substrate for cytochrome c-reductase and cytochrome c-oxidase, acetylated Cyt c (AC) has been used for the detection of superoxide radicals, particularly in complex biological systems. 4 However, both Cyt c and AC rapidly undergo glomerular filtration and disappear from circulation with a half-life of 2-3 min. Thus, in practice it is difficult to use Cyt c and AC for in vivo detection of superoxide radicals and related metabolites. To overcome such frustrating situations, an acetylated Cyt c derivative that circulates bound to albumin with a prolonged in vivo half-life was synthesized. i H. Sies, ed., " O x i d a t i v e S t r e s s . " A c a d e m i c Press, Orlando, Florida, 1985. 2 I. Emerit, L. Packer, and C. Auclair, eds., " A n t i o x i d a n t s in T h e r a p y and Preventive M e d i c i n e . " P l e n u m , N e w York, 1990. 3 j. M. M c C o r d and I. Fridovich, J. Biol. Chem. 244, 6049 (1969). 4 S. Minakami, K. Titani, and H. Ishikawa, J. Biochem. (Tokyo) 45, 341 (1958).
METHODS 1N ENZYMOLOGY,VOL. 234
Copyright © 1994by AcademicPress, Inc. All rights of reproduction in any form reserved.
[32]
LONG-ACTING CYTOCHROME C AND S O D DERIVATIVES
339
SM NH2 H2N~
NH2
H?-- CO0-
H2N~NH2
NH, NH2
HC--C~ 0 ~'0 HC--C~o
Cyt.c S
HCH
NH2~ NH2
C~-cH 0 II
H E="l
H2N ~'F~• J~ H SM-Cyt.c NH2
~
NH2
NH2
mm' 2-3
I
CH~-C~
CH3--CO ~ 1 ~
I
NH
°
'
L
=
0 ..~-~'-""~...~
~ NH2
/
II ~
^,. ..... ~,
. NH I
CH3--CO
NH I
CO
SMAC
I
CH3 FIG. I. Synthesis of a cytochrome c derivative that circulates bound to albumin with prolonged in vivo half-life. Cytochrome c was incubated with the anhydride of half-butylesterified poly(styrene-co-maleic acid) (SM). The SM-Cyt c thus formed was further reacted with acetic anhydride. The resulting SMAC forms a dissociable complex with albumin and escapes from being filtered by the glomerulus.
Synthesis of Long-Acting Cytochrome c Derivative We previously reported that superoxide dismutase covalently linked with half-butyl-esterified poly(styrene-co-maleic acid) (SM, molecular weight 1600) circulates bound to albumin and has a prolonged in vivo half-
340
ANTIOXIDANT CHARACTERIZATION AND ASSAY
REACTIVITY
OF CYTOCHROME
TABLE I SMAC
C AND
WITH
VARIOUS
[32]
COMPOUNDS
a
Rate b Antioxidant
Concentration (/~M)
Cytochrome c
SMAC
30 60 10 400 30 500
0.036 0.065 nd nd nd nd
0.003 0.005 nd nd nd nd
Ascorbic acid Reduced glutathione Uric acid Bilirubin Bilirubin + albumin
Incubation mixtures contained, in a final volume of 1 ml, 0.1 M phosphate buffer, pH 7.4, varying concentrations of low molecular weight compounds, and 20/xM cytochrome c or SMAC. The reaction was started by adding the reducing agents at 25°. The change in absorbance at 550 nm was monitored. b Change in absorbance at 550 nm/min, nd, Below detectable levels (<0.001).
100
A w
°
\
10
L.
0
10
Time
20
(rain)
FIG. 2. Fate of Cyt c and SMAC in the circulation. Under pentobarbital anesthesia, animals received intravenous injections of 2 ~mol/kg of either Cyt c (open circles) or SMAC (closed circles) into the tail vein. At the indicated times after administration, plasma levels of Cyt c and SMAC were determined.
[32]
LONG-ACTING CYTOCHROME C AND S O D DERIVATIVES
0
Time
I
I
5
10
(min)
341
FIG. 3. Reduction of SMAC in the circulation of intact and paraquat-treated rats. Animals received intravenous injections of either 0.1 ml of saline (open symbols) or 10 mg/kg of SMSOD (closed symbols). Five minutes after injection, 2 tzmol/kg of SMAC was administered. At the indicated times, plasma samples were obtained, and time-dependent reduction of SMAC was determined spectrophotometrically. SMAC was also administered with 50 mg/ kg of paraquat (squares).
life. 5,6 Experiments with various proteins and peptides revealed that their half-lives in circulation are increased by covalently linking SM to the biopolymers. To increase the half-life of Cyt c, an SM-conjugated and acylated Cyt c derivative was synthesized (Fig. 1) essentially as described for the synthesis of SM-conjugated SOD (SM-SOD). 5 The incubation medium contains, in a final volume of 10 ml, 0.1 M borate buffer, pH 8.0, 5 mM Cyt c, and 6 mM SM. The reaction is started by adding SM dissolved in dimethyl sulfoxide; the final concentration of dimethyl sulfoxide is 10%. After incubation at 25 ° for 3 hr, the reaction mixture is dialyzed against 5 liters of distilled water. Under these conditions, 1 mol of SM is covalently linked per mole of Cyt c. The dialyzed 5 T. Ogino, M. Inoue, Y. Ando, M. Awai, and Y. Morino, Int. J. Pept. Protein Res. 32, 464 (1988). 6 M. Inoue, I. Ebashi, and N. Watanabe, Biochemistry 28, 6619 (1989).
342
ANTIOXIDANT CHARACTERIZATIONAND ASSAY
[32] 2
E e-
0.04
Oxi 0.02
SMAC/~I
k.
d••
1
< 0
I
0
I
Time
I
2
I
(min)
I
4
FIG. 4. Reduction of SMAC in fresh plasma. A freshly isolated plasma sample was incubated with 20/zM SMAC at 25°, and time-dependent changes in absorbance at 550 nm were determined (curve 1). At the indicated times, 2 units/ml of ascorbate oxidase was added to the incubation mixture (curve 2). Incubation was also performed with plasma samples that were preincubated with ascorbate oxidase (2 units/ml) at 25° for 30 min (curve 3).
SM-conjugated C y t c (SM-Cyt c, 50 mg/ml) is diluted with the same volume of saturated sodium acetate solution. Acetic anhydride is added with vigorous stirring at 0 ° in 10 portions to give a final concentration of 200 mM. After 30 min, the reaction mixture is dialyzed at 4 ° for 20 hr against 5 liters o f 10 m M p h o s p h a t e buffer, p H 7.4, containing 0.15 M NaCI with two changes of the buffer solution. U n d e r these conditions, a b o u t 70% of the TNBS-titratable amino groups of SM-Cyt c are acetylated. The acetylated SM-Cyt c (SMAC) solution thus obtained is concentrated o v e r an A m i c o n (Danvers, MA) ultrafiltration m e m b r a n e (PM10). Storage o f S M A C at - 20 ° for at least 3 months does not affect its biochemical properties. Results of Using Derivatives to Study Superoxide and Ascorbyl Radicals in Vivo Although S M A C does not serve as a substrate for c y t o c h r o m e creductase and c y t o c h r o m e c-oxidase, it retains 8 and 80% of the Cyt c activity to react with ascorbyl and superoxide radicals, respectively. 7 7 R. Kunitomo, Y. Miyauchi, and M. Inoue, J. Biol. Chem. 267, 8732 (1992).
[33]
U B I Q U I N O L S AS A N T I O X I D A N T S
343
Reactivity of SMAC with other reducing agents such as cysteine and glutathione is negligible (Table I). When injected intravenously into rats, SMAC circulated bound to albumin with a half-life of 130 min, which showed a marked contrast to the short half-life of Cyt c (Fig. 2). The SMAC was rapidly reduced in the circulation of intact animals. Treatment of animals with paraquat markedly enhanced the rate of reduction (Fig. 3). Although long-acting SM-SOD effectively dismutates superoxide radicals and inhibited the superoxide-dependent reduction of SMAC in vitro, the reduction of the circulating SMAC was not affected by SM-SOD in normal and paraquat-treated animals. Administration of ascorbate oxidase markedly enhanced the rate of SMAC reduction in the circulation and in freshly isolated plasma samples. However, preincubation of plasma with ascorbate oxidase completely inhibited the reduction of SMAC, suggesting that ascorbyl radicals might be responsible for the reduction of SMAC in plasma (Fig. 4) and in the circulation and that the rate of ascorbyl radical generation might be increased by paraquat. The rate of SMAC reduction was also enhanced by alloxan, a diabetogenic agent. 8 In contrast to the experiments with paraquat, the rate of reduction of the circulating SMAC was markedly inhibited by SM-SOD. Thus, superoxide radicals might appear in the circulation of animals which were treated with alloxan. The combined use of SMAC and SM-SOD might permit quantitative studies on the in vivo generation of ascorbyl and superoxide radicals and their role in the pathogenesis of oxidative tissue injury. 8 K. Koyama, K. Takatsuki, and M. Inoue, Arch. Biochem. Biophys. in press (1994).
[33] Assay o f U b i q u i n o n e s a n d U b i q u i n o l s as A n t i o x i d a n t s B y VALERIAN E. KAGAN, ELENA A. SERBINOVA, D. A . STOYANOVSKY, S. KHWAJA, a n d LESTER PACKER
Introduction A diverse range of biological electron-transfer membranes are known to contain a complement of ubiquinones that can undergo redox changes. 1 In membranes of animals, the role of ubiquinones in electron transport is unclear with the notable exception of ubiquinones in mitochondria. In mitochondria, ubiquinones act as mobile distributors of reducing equivalents between the NADH dehydrogenase, succinate dehydrogenase, and 1 T. Ramasarma, in "Coenzyme Q" (G. Lenaz, ed.), p. 67. Wiley, New York, 1985.
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[33]
U B I Q U I N O L S AS A N T I O X I D A N T S
343
Reactivity of SMAC with other reducing agents such as cysteine and glutathione is negligible (Table I). When injected intravenously into rats, SMAC circulated bound to albumin with a half-life of 130 min, which showed a marked contrast to the short half-life of Cyt c (Fig. 2). The SMAC was rapidly reduced in the circulation of intact animals. Treatment of animals with paraquat markedly enhanced the rate of reduction (Fig. 3). Although long-acting SM-SOD effectively dismutates superoxide radicals and inhibited the superoxide-dependent reduction of SMAC in vitro, the reduction of the circulating SMAC was not affected by SM-SOD in normal and paraquat-treated animals. Administration of ascorbate oxidase markedly enhanced the rate of SMAC reduction in the circulation and in freshly isolated plasma samples. However, preincubation of plasma with ascorbate oxidase completely inhibited the reduction of SMAC, suggesting that ascorbyl radicals might be responsible for the reduction of SMAC in plasma (Fig. 4) and in the circulation and that the rate of ascorbyl radical generation might be increased by paraquat. The rate of SMAC reduction was also enhanced by alloxan, a diabetogenic agent. 8 In contrast to the experiments with paraquat, the rate of reduction of the circulating SMAC was markedly inhibited by SM-SOD. Thus, superoxide radicals might appear in the circulation of animals which were treated with alloxan. The combined use of SMAC and SM-SOD might permit quantitative studies on the in vivo generation of ascorbyl and superoxide radicals and their role in the pathogenesis of oxidative tissue injury. 8 K. Koyama, K. Takatsuki, and M. Inoue, Arch. Biochem. Biophys. in press (1994).
[33] Assay o f U b i q u i n o n e s a n d U b i q u i n o l s as A n t i o x i d a n t s B y VALERIAN E. KAGAN, ELENA A. SERBINOVA, D. A . STOYANOVSKY, S. KHWAJA, a n d LESTER PACKER
Introduction A diverse range of biological electron-transfer membranes are known to contain a complement of ubiquinones that can undergo redox changes. 1 In membranes of animals, the role of ubiquinones in electron transport is unclear with the notable exception of ubiquinones in mitochondria. In mitochondria, ubiquinones act as mobile distributors of reducing equivalents between the NADH dehydrogenase, succinate dehydrogenase, and 1 T. Ramasarma, in "Coenzyme Q" (G. Lenaz, ed.), p. 67. Wiley, New York, 1985.
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
344
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[33]
cytochrome b-c~ segment of the electron transport chain and as participants of the protonmotive Q cycle responsible for the transfer of protons across the coupling membrane. 2,3 Ubiquinones are stoichiometrically in excess of electron-transfer chains, and it was suggested that in addition to their role in mitochondrial respiration, ubiquinones in their reduced form, ubiquinols, may act as free radical scavengers. In this way ubiquinones/ubiquinols would prevent oxidative damage in mitochondria and in other electron-transporting membranes as well as in lipoproteins. 4-6 Two alternative mechanisms of ubiquinol antioxidant function have become the topic of vigorous discussion. One possibility is that ubiquinols act independently as chain-breaking antioxidants, providing hydrogen atoms to reduce peroxyl and/or alkoxyl radicals. 7,8 An alternative mechanism suggests a redox interaction between ubiquinol and another lipidsoluble antioxidant, vitamin E, in its one-electron oxidized form, vitamin E phenoxyl radical. 9,~° As a result of the reduction of the vitamin E phenoxyl radical by ubiquinol, regeneration of vitamin E occurs. The second mechanism suggests that reduction of ubiquinone by electron transport would thus drive ubiquinol-dependent regeneration of vitamin E, whereas the direct chain-breaking effect of ubiquinol may be less significant. The importance and contribution of each of these mechanisms into the overall antioxidant protection of membranes and lipoproteins provided by ubiquinols and by vitamin E should depend on the local concentrations of antioxidants, their mutual availability and their availability to oxidizing radicals, their mobility and uniformity of distribution in the membrane, etc. Comparison of Radical Scavenging Activity of Ubiquinol and Vitamin E In chemical systems, vitamin E (a-tocopherol) has a higher reactivity toward peroxyl radicals than ubiquinols. The rate constants for the interaction of ubiquinols Q9 and Q6 with the radicals generated by thermal decomposition of an azo initiator of peroxyl radicals, azodiisobutyronitrile, in B. Chance and B. G. Hollunger, J. Biol. Chem. 236, 1534 (1961). s p. Mitchell, J. Theor. Biol. 62, 327 (1976). 4 L. Landi, L. Cabrini, M. Sechi, and P. Pasquali, Biochem. J. 222, 463 (1984). 5 V. E. Kagan, E. A. Serbinova, G. Koynova, S. Kitanova, V. Tyurin, T. Stoytchev, P. Quinn, and L. Packer, Free Radical Biol. Med. 9, 117 (1990). 6 R. Stocker and B. Frei, Proc. Natl. Acad. Sci. U.S.A. 88, 1646 (1991). 7 B. Frei, M. Kim, and B. Ames, Proc. Natl. Acad. Sci. U.S.A. 87, 4879 (1990). 8 p. Forsmark, F. Aberg, B. Norling, K. Nordenbrand, G. Dallner, and L. Ernster, FEBS Lett. 285, 39 (1991). 9 A. Mellors and A. L. Tappel, J. Biol. Chem. 241, 4353 (1966). l0 V. E. Kagan, E. A. Serbinova, and L. Packer, Biochem. Biophys. Res. Commun. 169,
851 (1990).
[33]
UBIQUINOLS AS ANTIOXIDANTS
345
ethylbenzene were reported to be 3.2 x 105 and 3.4 x 105 M -1 sec -1, respectively. 11 Under the same conditions, a-tocopherol was about 10 times more reactive.~Z This suggests that a-tocopherol may preferentially interact with peroxyl radicals if its concentration in the membrane is higher than or comparable with that of ubiquinol. However, in ordered systems (liposomes, membranes, lipoproteins) not only chemical reactivity and concentration but also mobility and uniformity of distribution may be crucial for the efficiency of radical scavenging. To characterize the membrane radical scavenging activities of a-tocopherol and ubiquinol Ql0, their efficiencies in inhibiting lipid peroxidation in liposomes and membranes were compared. Effects on Oxidation of cis-Parinaric Acid. cis-Parinaric acid, a fluorescent fatty acid probe, can be used for measurements of antioxidant efficiency in membranes. 13'14 cis-Parinaric acid derives its fluorescence from four conjugated double bonds which can be readily oxidized by a variety of free radicals to yield nonfluorescent products. The probe is strongly fluorescent only when partitioned into a lipophillic environment, such as cell membranes. A lipid-soluble azo initiator of peroxyl radicals, 2,2'-azobis(2,4-dimethylvaleronitrile) (AMVN), caused fluorescence decay of cis-parinaric acid incorporated into dioleoylphosphatidylcholine liposomes (Fig. 1). The decay was inhibited by ubiquinol and a-tocopherol in a concentration-dependent manner (Fig. 2). We found that a-tocopherol was about 100 times more efficient in inhibiting AMVN-induced oxidation of cis-parinaric acid than was ubiquinol Q10: the concentrations producing 50% inhibition were 0.9 and 108 /xM, respectively. Thus, in liposomal membranes a-tocopherol has remarkably higher radical scavenging efficiency than ubiquinol Ql0. However, this difference might be due to specific interactions of a-tocopherol and ubiquinol, both with parinaric acid 15 and with AMVN. To avoid these possible specific interactions, we compared the antioxidant efficiencies of a-tocopherol and ubiquinol in liver microsomal membranes using a different oxidation system. Effects on Lipid Peroxidation in Rat Liver Microsomes. Ubiquinol Q9 added to rat liver microsomes produced 50% inhibition of Fe 2+ plus ascorbate-induced lipid peroxidation at a concentration of 7.6 x 10 -5 M, whereas a-tocopherol gave a half-maximal effect at a much lower 11 y . A. Zaslavsky, N. G. Khrapova, S. F. Terekhova, and E. B. Burlakova, Biofizika 22, 359 (1977). 12 V. V. Naumov and N. G. Khrapova, Biofizika 28, 730 (1987). x3 F. A. Kuypers, J. J. M. van den Berg, C. Schalkwijk, C. Roelofsen, and J. A. F. Op den Kamp, Biochim. Biophys. Acta 921, 266 (1987). 14 R. McKenna, F. J. Kezdy, and D. E. Epps, Anal. Biochem. 196, 443 (1991). t5 V. E. Kagan and P. J. Quinn, Eur. J. Biochem. 171, 661 (1988).
346
500-
A N T I O X I D A NCHARACTERIZATION T AND ASSAY
•~ ,~
"-]~ ~
[33]
10 la.MTocopherol
\ -
c~l~~l~,," I 10
1
I
20 30 Time (Minutes)
I 40
FIG. 1. Azo initiator-induced fluorescence decay of cis-parinaric acid incorporated into dioleoylphosphatidylcholine (DOPC) liposomes. Incubation conditions: cis-parinaric acid (2.5/zM) incorporated into DOPC liposomes (0.2 mg/ml) in phosphate buffer (50 mM, pH 7.4 at 40°), AMVN (1.0 mM). Ubiquinol Ql0 or a-tocopherol was incorporated into the DOPC liposomes by sonication.
concentration (0.2 × 10 -5 M). 5 Thus, in liver microsomal membranes exogenously added ubiquinol Q9 also exerted much lower antioxidant activity as compared with a-tocopherol. However, there are some uncertainties with the in vitro experiments in which ubiquinols or tocopherols dissolved in organic solvents are then added to membrane suspensions. One uncertainty concerns the completeness of the incorporation of the antioxidants into membrane. This can be overcome at least in part by sonicating membranes in the presence of exogenously added antioxidants. Sonication caused a significant increase in the antioxidant efficiency of both ubiquinol Q9 and a-tocopherol (decrease in concentrations producing half-maximal inhibition). However, a-tocopherol retained a remarkably higher antioxidant efficiency as compared with ubiquinol Q9 (Table I). Correlations between Ubiquinol/a-Tocopherol Content and Lipid Peroxidation in Liver Microsomal Membranes in Vivo. Based on the results of in vitro studies one would assume that in vivo vitamin E (a-tocopherol) may be more efficient than ubiquinols as a chain-breaking antioxidant, if the concentrations in membranes are comparable. To test this hypothesis, we tried to find correlations between the levels of endogenous ubiquinols or tocopherols, on the one hand, and the endogenous concentrations
[33]
UBIQUINOLS AS ANTIOXIDANTS
347
100 • 80
i®
i '° 0 .01
' .1
1
10
~ 1000
100
A l p h a - T o c o p h e r o l o r U b i q u f u o l QIO, pM
Fn6.2. Effect of a-tocopherol and ubiquinol Ql0 on cis-parinaric acid oxidation in DOPC liposomes induced by azo-initiators of peroxyl radicals. O, Ubiquinol (AMVN); O, ubiquinol (AAPH); A, tocopherol (AMVN); A, tocopherol (AAPH). Incubation conditions: AMVN or AAPH (1 mM); other conditions were as given in the legend to Fig. 1.
TABLE I INHIBITION OF Ee 2÷ PLUS ASCORBATE-INDUCED LIPID PEROXIDATION IN RAT LIVER M1CROSOMES BY t~-ToCOPHEROL AND UBIQUINOL Q9 Concentration producing 50% inhibition (/zM) Antioxidant
Before sonication
After sonication"
a-Tocopherol Ubiquinol Q9
2.0 76.0
0.8 12.8
Rat liver microsomes (1.0 mg/ml) in 0.1 M phosphate buffer (pH 7.4) were sonicated in the presence of exogenously added a-tocopherol or ubiquinol Q9 (90 sec at 4°), after which lipid peroxidation was induced by adding ascorbate (0.5 mM) and Fe 2+ (20/zM).
348
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[33]
of lipid peroxidation products or susceptibility of membranes to lipid peroxidation, on the other. In our experiments, after 8 weeks of dietary manipulations with ubiquinone Q~0 (Q10-supplemented diet) and vitamin E (vitamin E-deficient or vitamin E-supplemented diets) several groups of animals were obtained having more than a 50-fold difference in the a-tocopherol content and about a 10-fold difference in the ubiquinone/ ubiquinol content in their liver microsomal fractions (Table II). In these microsomes we assessed the levels of endogenous fluorescent lipid peroxidation products and the susceptibility to in vitro lipid peroxidation induced by a lipid-soluble azo initiator of peroxyl radicals, AMVN. We found a strong inverse correlation between the vitamin E content and both endogenous (r = 0.98) and induced lipid peroxidation (r = 0.86). No significant correlation was revealed between ubiquinone Q9+ 10 or ubiquinol Q9+10 concentrations and endogenous or induced lipid peroxidation (r = 0.3-0.4). These cumulative data suggest that vitamin E may be more efficient than ubiquinols in the membrane antioxidant protection via the pathway of direct radical scavenging. This does not rule out direct antioxidant effects of ubiquinols in membrane compartments deficient in vitamin E. However, in the presence of sufficient vitamin E concentrations, an alternative ubiquinol-dependent antioxidant mechanism may be operative, namely, ubiquinol-dependent reduction of the vitamin E phenoxyl radical. Ubiquinol-Dependent Regeneration of Vitamin E
Sparing Effects of Ubiquinols against Vitamin E Oxidation. If the reduction of the vitamin E phenoxyl radical by ubiquinols occurs in membranes, then the vitamin E oxidation should be prevented partially or completely in the presence of ubiquinol. In other words, ubiquinol should produce a lag period in tocopherol oxidation whose duration should be dependent on the ubiquinol concentration. In fact, protective effects of ubiquinols against a-tocopherol oxidation have been reported. 7'16 However, a concurrent autoxidation of ubiquinols by oxygen under experimental conditions used in these studies interfered with the reaction of ubiquinols with the vitamin E radicals, thus complicating the interpretation of the results. Two approaches can be used to cope with this problem: (1) performing ubiquinol/a-tocopherol oxidation under anaerobic conditions using a source of carbon-centered (alkyl) radicals, or (2) dissolving ubiquinol in a slightly "acidified" organic solvent. This simple procedure prevents 16 y . Yamamoto, E. Komuro, and E. Niki, J. Nutr. Sci. Vitarninol. 36, 505 (1990).
[33]
UBIQUINOLS
Z *e
"~
349
AS ANTIOXIDANTS
~ 0
0
Z < /-
~g g~
o'~
°~.~
~2 Z
Z
~E
9~
<
~ z
Z
°t~:
e.,
;>'8
.o
350
A N T I O X I D A N T
30"
-~ 0
A
0
20
t.
C H A R A C T E R I Z A T I O N
30"
- AAPI-I
6
AND
[33]
ASSAY
- AAPI-I
B
20
¢
d |
o o
10"
0
= .m =
+AAPH
10'
.r,
t <
,L
0
. 0
. . 20
. . 40
. 60
Time, min
, 80
0 0
, 20
.
,
40 Time,
-
,
60
.
,
80
min
FIG. 3. Azo initiator-induced oxidation of a-tocopherol or ubiquinol Q2 in DOPC liposomes under aerobic and anaerobic conditions. (©, I ) +02; (A, A) +N2. Incubation conditions: DOPC (5.0 mg/ml), AAPH (10 mM), in phosphate buffer (50 mM, pH 7.4 at 40°).
ubiquinol autoxidation both during its storage and during incorporation into liposomes. 17 We used both approaches to study the oxidation of ubiquinols (Q2 and Ql0) and a-tocopherol incorporated into dioleoylphosphatidylcholine liposomes in the presence of the azo-initiator 2,2'azobis(2-amidinopropane) dihydrochloride (AAPH). Incubation of dioleoylphosphatidylcholine liposomes with incorporated ubiquinols Q2 or Ql0 or with incorporated a-tocopherol in the presence of AAPH at 37° resulted in a linear oxidation of the antioxidants, with the rate dependent on the AAPH concentration (Fig. 3). Both atocopherol and ubiquinols were oxidized at the same rate under aerobic or anaerobic conditions. If, in the equimolar mixture of o~-tocopherol and ubiquinol, AAPH-induced oxidation of the antioxidants occurs independently, one would expect that the consumption of both ct-tocopherol and ubiquinol would take place with a rate two times lower than the rate of the oxidation of each separately. Experimental results, however, showed that under both anaerobic and aerobic conditions in the mixture with atocopherol, ubiquinols were oxidized to ubiquinones at the same rate as when they were oxidized in the absence of a-tocopherol. In contrast, in a mixture the oxidation of a-tocopherol did not begin until ubiquinol was 17 A. J. Swallow, in "Function of Quinones in Energy Conserving Systems" (B. L. Trumpover, ed.), p. 59. Academic Press, New York, 1982.
[33]
UBIQUINOLS
30
AS ANTIOXIDANTS 30
A
351
B
e4 20
.~
20
10"
lO
°m
-= 0 0
30 60 Time, rain
90
0
30 Time,
60 min
90
FIG. 4. Azo initiator-induced oxidation of a mixture of c~-tocopherol and ubiquinol Q2 in DOPC liposomes under anaerobic (A) or aerobic (B) conditions. O, a-tocopherol; A, ubiquinol Q2 ; [3, ubiquinone Q2. Incubation conditions: DOPC (5.0 mg/ml), AAPH (10 mM), in phosphate buffer (50 mM, pH 7.4 at 40°).
completely oxidized. Following this lag period, the oxidation proceeded at the same rate as for ubiquinol or tocopherol alone (Fig. 4). These high-performance liquid chromatography (HPLC) results directly demonstrate that ubiquinols and a-tocopherol in the mixture interact in such a way that ubiquinol completely prevents the a-tocopherol oxidation, presumably via a mechanism of reducing its phenoxyl radical. In fact, the reactivity of ubiquinols toward the o~-tocopherol phenoxyl radical is known to be rather high (second-order rate constants 2.2 and 3.7 × 105 M-~ sec-1 at 25° in ethanol and benzene, respectively), 18,~9which is significantly higher than their reactivity toward peroxyl radicals. 12This conclusion can be supported by direct electron spin resonance (ESR) measurements of the ubiquinol/a-tocopherol phenoxyl radical interactions. Reduction of Vitamin E Phenoxyl Radical by Ubiquinols. In the presence of the enzymatic oxidation system of lipoxygenase plus linolenic acid, which generates peroxyl radicals, vitamin E incorporated into dioleoylphosphatidylcholine liposomes gives a characteristic pentameric (ESR) signal arising from the phenoxyl radical with component g values of 2.0122, 2.0092, 2.0061, 2.0028, and 1.9932°'21 (Fig. 5). Under the conditions 18 K. Mukai, S. Kikuchi, and S. Urano, Biochim. Biophys. Acta 1035, 77 (1990). 19 K. Mukai, S. Itoh, and H. Morimoto, J. Biol. Chem. 267, 22277 (1992). 20 L. Packer, J. J. Maguire, R. J. Melhorn, E. A. Serbinova, and V. E. Kagan, Biochem. Biophys. Res. Commun. 159, 229 (1989). 21 V. E. Kagan, E. A. Serbinova, and L. Packer, Arch. Biochem. Biophys. 282, 221 (1990).
352
A N T I O X I D A NCHARACTERIZATION T AND ASSAY
[33]
UBIQUINOL QlO
UBIQtaNOL Q10
FIG. 5. Electron spin resonance spectra of lipoxygenase plus linolenic acid-induced vitamin E phenoxyl radicals in DOPC liposomes. Incubation conditions: DOPC liposomes (20 mg/ml) in phosphate buffer (0.1 M, pH 7.4 at 25°), soybean lipoxygenase (3 U/~I), linolenic acid (200 ~M), ~-tocopherol (1.5 raM), ubiquinol Q10 (0.5 mM).
used, the signal was persistent for 15-20 min. Addition of ubiquinol Ql0 resulted in immediate and complete disappearance of the vitamin E phenoxyl radical ESR signal. In separate experiments, the vitamin E phenoxyl radical was generated by UV-irradiation of dioleoylphosphatidylcholine liposomes containing a-tocopherol. The UV-induced ESR signal was also quenched by ubiquinol Q10- Thus, ubiquinol Ql0 is able to reduce the vitamin E phenoxyl radical in liposomes. Similarly in rat liver microsomes enriched with vitamin E by dietary supplementation, the ESR signal of the endogenous vitamin E phenoxyl radical could be observed in the presence of lipoxygenase plus linolenic acid (Fig. 6). Addition of NADPH to microsomes resulted in a decrease (but not complete disappearance) of the ESR signals of endogenous atocopherol-derived phenoxyl radicals. The decrease was transient, however, and after some delay the magnitude of the ESR signal increased and subsequently followed characteristic decay kinetics, probably owing to irreversible oxidation of vitamin E. Simultaneous addition of NADPH plus ubiquinone Q1 to microsomes resulted in complete elimination of the ESR signal. The elimination of the signal was also transient, and the duration of the lag period was linearly dependent on the concentration of ubiquinone Ql added. In the absence of NADPH, ubiquinone Q1 did not cause the disappearance of the vitamin E radical signal. In mitochondrial suspension, addition of NADH caused a transient disappearance of the ESR signal of the vitamin E phenoxyl radical generated by lipoxygenase plus linolenic acid. In the presence of both NADH and ubiquinone Q1, the lag period, during which the ESR signal of the
[33]
UBIQUINOLS AS ANTIOXIDANTS
353
12"
6
+ Q1
8
4"
+ NADPH
r13
ee ;t3 t~
rm
0
0
+ NADPH + Q1
L
.!
10
20
30
Time, min FIG. 6. Time course of endogenous vitamin E phenoxyl radical ESR signal induced by lipoxygenase plus linolenic acid in rat liver microsomes in the presence of NADPH and ubiquinone QI. Incubation conditions: vitamin E-enriched microsomes (14.5 nmol a-tocopherol/mg protein, 15 mg protein/ml) in phosphate buffer (0.1 M, pH 7.4 at 25°), lipoxygenase (3 U/p.l), linolenic acid (0.8 raM), ubiquinone Qt (0.5 mM), NADPH (6.0 mM). Vitamin Eenriched liver microsomes were isolated from vitamin E-supplemented rats [20 g (_+) atocopheryl acetate/kg diet].
vitamin E phenoxyl radical could not be detected, was significantly longer, although ubiquinone Q] by itself did not cause a decrease in or elimination of the ESR signal. On the contrary, in the presence of ubiquinone Q1 the magnitude of the signals was greater. Thus ubiquinone Q] interacting with NADH- or NADPH-dependent electron carriers in mitochondria and microsomes synergistically increases the lag time of the appearance of the vitamin E phenoxyl radicals formed in the reaction of chromanols with lipoxygenase-generated peroxyl radicals. HPLC measurements showed that ubiquinone Q1 exerts a sparing effect decreasing the level of vitamin E consumption in the course of incubation.l° In studies with mitochondrial membranes, it was found that a similar effect was produced when ubiquinone was reduced by succinate-dependent electron transport: the lipoxygenase plus linolenic acid-induced vitamin E phenoxyl radical ESR signal was quenched and the vitamin E consumption was prevented. By incorporating isolated succinate-ubiquinone reductase [mitochondrial complex II; succinate dehydrogenase (ubiquinone)] into dioleoylphosphatidylcholine liposomes in the presence of a-tocopherol and ubiquinone Ql0, it was possible to determine that enzyme-linked reduction of ubiquinone can prevent vitamin E phenoxyl
354
ANTIOXIDANT
CHARACTERIZATION
AND ASSAY
[34]
radical accumulation and a-tocopherol consumption. 22 Succinate-ubiquinone reductase incorporated into liposomes showed no direct reduction of the a-tocopherol phenoxyl radical and seems to be a poor candidate for the agent responsible for reduction of the vitamin E radical under physiological conditions. If succinate-ubiquinone reductase can reduce the vitamin E radical, the rate is substantially lower than when the same reaction occurs in the presence of reduced quinone. In the presence of quinone, either as ubiquinone Ql or ubiquinone Ql0, succinate plus succinate-ubiquinone reductase prevents accumulation of the a-tocopherol phenoxyl radical and a-tocopherol consumption. The interaction of the a-tocopherol phenoxyl radical with reduced ubiquinone thus appears important in cellular recycling of mitochondrial vitamin E. 22 Conclusion The ubiquinone/ubiquinol redox couple may act efficiently as a mediator in the regeneration of vitamin E by electron transport in cellular membranes. Does this mean that in the absence of vitamin E ubiquinols may not be important for antioxidant protection? It is possible that, in addition to a concerted antioxidant function of ubiquinols via the vitamin E recycling mechanism, direct radical scavenging effects may be necessary in the compartments devoid oftocopherols. It has been suggested that ubiquinols may be necessary for reducing the vitamin E phenoxyl radical to prevent prooxidant effects of vitamin E in low density lipoproteins (LDL). 23 22j. j. Maguire, V. E. Kagan, B. A. C. Ackrell, E. A. Serbinova, and L. Packer, Arch. Biochem. Biophys. 292, 47 (1992). 23 V. W. Bowry, K. U. Ingold, and R. Stocker, Biochem. J. 288, 341 (1992).
[34] A n t i o x i d a n t P r o p e r t i e s o f a - T o c o p h e r o l and a-Tocotrienol By ELENA A. SERBINOVA a n d LESTER PACKER
Introduction Vitamin E was discovered at the University of California at Berkeley in 1922 in the laboratory of Herbert M. Evans. From its initial discovery as an antifertility agent, it was given the name of tocopherol, meaning child birth (tokos) to carry (pherein). Indeed, it was the development of the rat fetal reabsorption biological assay that was used as the basis to METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, lnc, All rights of reproduction in any form reserved,
354
ANTIOXIDANT
CHARACTERIZATION
AND ASSAY
[34]
radical accumulation and a-tocopherol consumption. 22 Succinate-ubiquinone reductase incorporated into liposomes showed no direct reduction of the a-tocopherol phenoxyl radical and seems to be a poor candidate for the agent responsible for reduction of the vitamin E radical under physiological conditions. If succinate-ubiquinone reductase can reduce the vitamin E radical, the rate is substantially lower than when the same reaction occurs in the presence of reduced quinone. In the presence of quinone, either as ubiquinone Ql or ubiquinone Ql0, succinate plus succinate-ubiquinone reductase prevents accumulation of the a-tocopherol phenoxyl radical and a-tocopherol consumption. The interaction of the a-tocopherol phenoxyl radical with reduced ubiquinone thus appears important in cellular recycling of mitochondrial vitamin E. 22 Conclusion The ubiquinone/ubiquinol redox couple may act efficiently as a mediator in the regeneration of vitamin E by electron transport in cellular membranes. Does this mean that in the absence of vitamin E ubiquinols may not be important for antioxidant protection? It is possible that, in addition to a concerted antioxidant function of ubiquinols via the vitamin E recycling mechanism, direct radical scavenging effects may be necessary in the compartments devoid oftocopherols. It has been suggested that ubiquinols may be necessary for reducing the vitamin E phenoxyl radical to prevent prooxidant effects of vitamin E in low density lipoproteins (LDL). 23 22j. j. Maguire, V. E. Kagan, B. A. C. Ackrell, E. A. Serbinova, and L. Packer, Arch. Biochem. Biophys. 292, 47 (1992). 23 V. W. Bowry, K. U. Ingold, and R. Stocker, Biochem. J. 288, 341 (1992).
[34] A n t i o x i d a n t P r o p e r t i e s o f a - T o c o p h e r o l and a-Tocotrienol By ELENA A. SERBINOVA a n d LESTER PACKER
Introduction Vitamin E was discovered at the University of California at Berkeley in 1922 in the laboratory of Herbert M. Evans. From its initial discovery as an antifertility agent, it was given the name of tocopherol, meaning child birth (tokos) to carry (pherein). Indeed, it was the development of the rat fetal reabsorption biological assay that was used as the basis to METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, lnc, All rights of reproduction in any form reserved,
[34]
ot-TOCOPHEROL AND Ot-TOCOTRIENOL
355
evaluate vitamin E activity. One international unit for vitamin E activity was based on the amount of vitamin E that was needed to prevent reabsorption of the fetus in a pregnant rat. At least eight compounds have been isolated from plant sources which have vitamin E activity. All have a six-membered chromanol ring structure and a side chain. The tocols have a phytol chain, and the tocotrienols have a similar structure with double bonds at the 3', 7', and 11' portions of the side chain. Both tocols and tocotrienols occur as a variety of isomers which differ by the number and location of methyl groups on the chromanol ring.~ The amount of methyl substituents in the chromanol nucleus gives rise to a, fl, y, and 8 isomers. 2,3 Tocopherols and tocotrienols are present in various components of the human diet. Tocopherols are found in polyunsaturated vegetable oils and in the germ of cereal seeds, whereas tocotrienols are found in the aleurone and subaleurone layers of cereal seeds and in palm oil. Although the tocopherols and tocotrienols are closely related chemically, they have widely varying degrees of biological effectiveness. The potency of atocotrienol evaluated by gestation-resorption assays have been shown to be 32% of the potency of a-tocopherol. 4 Even though the mechanism of physiological activity of vitamin E is not clearly understood, it is likely that at least some of the biological activities that have been demonstrated are due to its antioxidant function. The biological activity of vitamin E is generally believed to be due to its antioxidant action of inhibiting lipid peroxidation in biological membranes by scavenging chain-propagating peroxyl radical (ROO-)3'5'6: ROO. + Toc-OH--~ ROOH + Toc-O.
(1)
The antioxidant function of vitamin E per se is localized in the chromanol nucleus, whereas the phenolic hydroxy group donates an H atom to quench lipid radicals. 7'8 There appears to be widespread confusion concerning both the relative and the absolute antioxidant effectiveness in vitro of the individual tocoph1 L. J. Machlin, in "Handbook of Vitamins: Nutritional, Biochemical and Clinical Aspects" (L. J. Machlin, ed.), pp. 245-265. Dekker, New York, 1984. 2 j. C. Bauernfidd, in "Vitamin E " (L. J. Machlin, ed.), pp. 99-135. Dekker, New York, 1984. 3 L. A. Witting, in "Free Radicals in Biology" (W. A. Pryor, ed.), Vol. 4, pp. 295-319. Academic Press, New York, 1980. 4 j. I . McHale, J. Green, S. Marcinkiewicz, Br. J. Nutr. 15, 253 (1961). 5 G. W. Burton and K. U. Ingold, J. Am. Chem. Soc. 103, 6472 (1981). 6 G. W. Burton, A. Joyce, and K. U. Ingold, Arch. Biochem. Biophys. 221, 281 (1983). 7 A. L. Tappel, Vitam. Horm. (N. Y.) 20, 493 (1962). 8 G. W. Burton and K. U. Ingold, Acc. Chem. Res. 19, 194 (1986).
356
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[34]
erols that make up vitamin E. It was suggested 9 that relative effectiveness depends on the experimental conditions. In homogeneous solution the reaction rate constants of chromanols with peroxyl radicals [reaction (1)] do not depend on the length or unsaturation of the hydrocarbon chain, but are mainly dependent on the number of methyl groups in the benzene ring of the chromanol nucleus. ~0,~1However, in microdomains of heterogeneous membranous systems vitamin E owes its antioxidant potency not solely to the chemistry as in reaction (1) but also to its mobility and accessibility within the membrane. 12-15 It is known that the molecular mobility ofpolyenoic lipids in membrane bilayers is much higher than that of saturated lipids. 16Thus we may predict that tocotrienols with their unsaturated hydrocarbon side chain should be more mobile and less restricted in their interactions with lipid radicals in membranes than tocopherols. As a result, the antioxidant potency of tocotrienols in membranes is expected to be higher than that of tocopherols. Indeed, there is indirect evidence of higher antioxidant activity of tocotrienols in comparison with tocopherols. Tatsuta 17reported in studies on hemolysis that o~-tocotrienol showed higher efficiency in protecting red blood cells than ot-tocopherol in vitro, reverse to the result observed in vivo. Tocotrienols have been reported to possess higher protective activities against cardiotoxicity of the antitumor redox cycling drug adriamycin. The cardiotoxicity is believed to be caused by free radicals generated by adriamycin. 18It was also found that o~-tocotrienol showed higher inhibitory effects on lipid peroxidation induced by adriamycin in rat liver and murine microsomes than o~-tocopherhol. 19 However, in liposomes prepared from 9 j. R. Chipault, in "Autoxidation and Antioxidants" (W. O. Lundberg, ed.), Vol. 2, pp. 477-542. Wiley (Interscience), New York, 1962. 10 G. W. Burton, T. Doba, and E. J. Gabe, J. Am. Chem. Soc. 107, 7053 (1985). 11 y . B. Burlakova, Y. E. Kuchtina, I. P. Ol'khovskaya, I. K. Sarycheva, Y. B. Sinkina, and N. G. Khrapova, Biofizika 24, 989 (1980). 12 E. Niki, A. Kawakami, Y. Yamamoto, and Y. Kamiya, Bull. Chem. Soc. Jpn., Suppl. 58, 1971 (1985). 13 E. Niki, A. Kawakami, T. Saito, Y. Yamamoto, J. Tsuchiya, and Y. Kamiya, J. Biol. Chem. 260, 2191 (1985). 14 E. Niki, Chem. Phys. Lipids 44, 227 (1985). 15 E. Niki and E. Komuro, Basic Life Sci. 49, 561 (1987). 16 M. Shinitzky, in "Physiology of Membrane Fluidity" (M. Shinitzky, ed.), Vol. 1, p. 1-51. CRC Press, Boca Raton, Florida, 1984. 17 T. Tatsuta, Vitamin 44, 18S (1971). 18 K. Komiyama, K. Iizuka, M. Yamaoka, H. Watanabe, N. Tsuchiya, and I. Umezawa, Chem. Pharm. Bull. 37, 1369 (1989). i9 A. Kato, M. Yamaoka, A. Tamaka, K. Komiyama, I. Umezawa, Abura Kagaku 34, 37S (1985).
[34]
O~-TOCOPHEROL AND Ot-TOCOTRIENOL
357
microsomal phospholipids or from dipalmitoylphosphatidylcholine, tocopherol and tocotrienol were equally efficient in inhibiting iron-induced cholesterol 5-a-hydroperoxide decomposition. 2° Interaction of a-Tocopherol and a-Tocotrienol with Superoxide Radicals Initiation of lipid peroxidation often involves an oxygen-activation stage during which reactive oxygen species are formed. Among these, superoxide anion, the product of one-electron oxygen reduction, is considered to be important. 2~ Although vitamin E is a lipid-soluble antioxidant, the hydroxy group of its chromanol nucleus is exposed to the aqueous phase. Hence, interaction of vitamin E with oxygen radicals seems to be feasible. To compare the efficiencies of a-tocopherol and a-tocotrienol in scavenging superoxide radicals, one can measure their effects on lucigeninsensitized chemiluminescence of superoxide radicals generated by xanthine oxidase plus xanthine. Incubation Conditions. Incubation medium (2 ml) contains DOPC (dioleoylphosphatidylcholine) liposomes (254 /zM) in Tris-HC1 buffer, pH 7.5 at 37°; 100/~M lucigenin (bis-N-methylacridinium nitrate); 100/xM Desferal; 50/zM xanthine. The reaction is started by adding 14 mU xanthine oxidase. Measurements are done on LKB-Pharmacia (Piscataway, N J) 1250 Luminometer. Chromanols and DOPC have to be first dissolved in chloroform, dried under nitrogen, and resuspended in Tris buffer by sonicating for 10 min at 22 kHz, 4°. Tocopherol and a-tocotrienol incorporated into DOPC liposomes were equally efficient in quenching the superoxide-evoked chemiluminescence of lucigenin (Fig. 1). Thus, a-tocopherol and a-tocotrienol do not differ in the efficiency of their interaction with superoxide radical. 22 Radical Scavenging Activity of a-Tocopherol and a-Tocotrienol in Hexane To compare quantitatively the radical scavenging activity of a-tocopherol with that of a-tocotrienol, their interactions with peroxyl radicals can be measured. The assay proposed is performed at 40° using a hydrophobic azo initiator of radicals, 2,2'-azobis(2,4-dimethylvaleronitrile) [CH(CH3)2--CHzC(CHa)--CN--N=N--C(CHa)CN--CHE--CH(CH3)2] 2o M. Nakano, K. Sugioka, T. Nakamura, and T. Oki, Biochim. Biophys. Acta 619, 274 (1980). 21 I. Fridovich, Arch. Biochem. Biophys. 247, 1 (1986). 22 E. Serbinova, M. Tsuchiya, S. Goth, V. Kagan, and L. Packer, in "Vitamin E in Health and Disease" (L. Packer and J. Fuchs, eds.), pp. 235-241. Dekker, New York, 1992.
358
ANTIOXIDANT CHARACTERIZATIONAND ASSAY
[34]
60"
o~ i
a-T0c0pher01 / R^2 = 0.981
50
'-~°°'o 3040 20 /
~-Tocotrieno
R^2=0.978
o 0
'
0
i
20
"
•
i
'
i
40 60 Chromanols, p.M
•
I
80
FIG. I. Inhibitionof superoxide-dependentlucigeninchemiluminescencebyct-tocopherol and a-tocotrienolin DOPCliposomes.
(AMVN) (Polysciences Inc., Warrington, PA), and a polyunsaturated fatty acid, cis-parinaric acid, in hexane. 23 AMVN produces reactive peroxyl radicals at a constant rate by thermal decomposition as follows: R--N~.-~-N~R ~ [R. N 2 • R] ~ 2eR. + (1 - e ) R ~ R + N: R.+ O2~ROO" The AMVN-induced peroxyl radicals oxidize c&-parinaric acid, which is monitored by a decay of its characteristic fluorescence (excitation at 304 nm and emission at 421 nm). Following the addition of AMVN, the fluorescence intensity of cis-parinaric acid decreases linearly. When a-tocopherol or ct-tocotrienol is added to the system, a protection of cis-parinaric acid against the peroxidation is observed as measured by the inhibition of the fluorescence decay (Fig. 2). The duration of the protection period is the same for a-tocopherol and ct-tocotrienol. Thus, the radical scavenging activity of both antioxidants with peroxyl radicals is similar in hexane, where random collisions of antioxidant molecules with peroxyl radicals occur. 22 23M. Tsuchiya, G. Scita, H.-J. Freistleben, V. Kagan, and L. Packer, this series, Vol. 190 (1992).
[34]
359
Ot-TOCOPHEROLAND ot-TOCOTRIENOL
z cd a: >I-¢,/) Z ILl I.-Z UJ (..) Z UJ ¢..) ¢t) UJ n-
O :ZD _J LL
100
8 ~mNM~
~ ' - ~
6 4 0 i-
100 p.M l cis-
~PARINARIC
0~ 0
I
160 p . ~ . . ~ et-TOCOPHEROL / ¢x-TOCOTRIENOL
/I 4
EX 304 nm, EM 421
I 8
I,
I 12
I
I 16
I
I 20
nm I 24
TIME IN MINUTES FIG. 2. Comparison of antioxidant activity between tx-tocopheroI and tx-tocotrienol in hexane.
Incubation Conditions. The reaction mixture (3 ml) contains AMVN (100 mM) and cis-parinaric acid (30/zM) in hexane, a-Tocopherol and a-tocotrienol (160/zM each) have to be initially dissolved in chloroform and added to the incubation medium during the course of AMVN-induced fluorescence loss of cis-parinaric acid. Radical Scavenging Activity of a-Tocopherol and a-Tocotrienol in Liposomes: Azo Initiator-Based Assay Addition of A M V N to a suspension of diolcoylphosphatidylcholinc (DOPC) liposomes in the presence of luminol produces a characteristic chcmilumincsccnt response. This response is not observed in the absence of liposomes, indicating that the recorded chemiluminescence represents the reaction of AMVN-dcrivcd pcroxyl radicals with luminol in D O P C liposomal membranes. Liposomcs with incorporated a-tocophcrol or o~-tocotricnol inhibit AMVN-induced luminol-sensitized chemiluminescence in a concentration-dependent fashion22 (Fig. 3). The antioxidant efficiency of a-tocopherol and a-tocotricnol is different. The concentrations of a-tocophcrol and a-tocotricnol producing 50% inhibition of AMVN-induccd chemiluminescence arc 7.5 and 5/~M, respectively. Thus in liposomcs a-tocotricnol is 1.5 times more efficientscavenger ofpcroxyl radicals than o~-tocopherol. These data are in agreement with the results
360
[34]
ANTIOXlDANT CHARACTERIZATION AND ASSAY A DOSE DEPENDENCY
B REACTION CURVE 200
lOOq
A
s=
80
mTOCOPHEROL
160
.._1
i,.,--
0 re, Z
60
120
O O
3 p.M c~ TOCOTRIENOL
ii
O
4
'I,-" z u.I O
80
ua o :z
4
2 ~-TOCOTRIENOL
"._..= -' ,~= O
0 0
I
2
CONCENTRATION(gM)
0
4
8
TIME IN MINUTES
FIG. 3. Radical scavenging activity of a-tocopherol and a-tocotrienol incorporated in DOPC liposomes.
reported by Yamaoka and Komiyama24 on the antioxidant activity of a-tocopherol and a-tocotrienol in the 2.2'-azobis(2-aminopropane) dihydrochloride (AAPH)-initiated oxidation of dilinoleoylphosphatidylcholine (DLPC) liposomes, a-Tocotrienol added after liposome formation shows higher antioxidant activity than a-tocopherol. Incubation Conditions. Incubation medium (2 ml) contains DOPC liposomes (2.5 mM), luminol (150/zM), and a-tocopherol (or a-tocotrienol) in various concentrations in Tris-HCl buffer, pH 7.4. The reaction starts at 40° by the addition of AMVN (2.5 mM). Chromanols and DOPC must be initially dissolved in chloroform, dried under nitrogen, and resuspended in Tris buffer by sonicating for 10 rain at 22 kHz, 4°.
Antioxidant Activity of a-Tocopherol and a-Tocotrienol in Membranes Comparison of the antioxidative properties of different tocopherols and tocotrienols in preventing the oxidation of lard showed that tocotrienols are more active than the corresponding tocopherols. 25The antioxidative efficiency of tocotrienol isomers measured at 1I0 ° in the dark increases in the following order: c~- >/3- > y- > 8-tocotrienol. 24 M. Yamaoka and K. Komiyama, J. Jpn. Oil Chem. Soc. 38, 478 (1989). 25 V. A. Seher and S. A. Ivanov, Fette, Seifen, Anstrichm. 7S, 606-9 (1973).
[34]
o~-TOCOPHEROL AND Ot-TOCOTRIENOL
361
TABLE I CONSTANTS FOR 50% INHIBITION OF LIPID PEROXIDATION BY ot-ToCOPHEROL AND ot-ToCOTRIENOL IN RAT HEART MICROSOMES AND MITOCHONDRIA a
Constants (M) Microsomes Antioxidant a-Tocopherol a-Tocotrienol
Fe(II) + N A D P H 3.8 × 10 -5 0.2 × 10 -5
Fe(II) + a s c o r b a t e
Mitochondria Fe(lI) + ascorbate
7.1 × 10 _6 0 . 4 × 10 _6
2.8 × 10 -6 0.3 × 10 -6
a Timeof lipid peroxidation: 5 min. Incubationmediumcontained0.5 mg protein/ml, 10 mM F e S O 4 • 7H20, and 0.5 mM NADPH or ascorbate in 0.1 M potassium, sodium phosphate buffer, pH 7.4, at 37°. Quantitative comparison of the antioxidant potencies of c~-tocopherol and a-tocotrienol in a more physiological system [Fe(II) plus ascorbateor Fe(II) plus NADPH-induced lipid peroxidation in rat liver microsomes] has shown that a-tocotrienol exerts much higher antioxidant activity than a-tocopherol. The concentrations of a-tocopherol producing 50% inhibition (Ks0) are 40 and 60 times higher than those for a-tocotrienol for Fe E+ plus NADPH- and Fe E+ plus ascorbate-dependent lipid peroxidation, respectively. 26 Similar results are obtained for rat heart mitochondria and microsomes (Table I). I n c u b a t i o n C o n d it io n s . Incubation medium contains NADPH or ascorbate (0.5 mM), FeSO4" 7 H20 (10/zM), and protein (0.3 mg/ml) in 0.1 M potassium, sodium phosphate buffer, pH 7.4 at 37°. Secondary lipid peroxidation products, interacting with 2-thiobarbituric acid, are determined spectrophotometrically. Chromanols must first be dissolved in ethanol. To prevent the effect of ethanol on the accumulation of lipid peroxidation products the final ethanol concentration in the reaction mixture has to be less than 0.5%. Methods for Generation of Chromanoxyl Radicals and Recycling Efficiency of a-Tocopherol and a-Tocotrienol The steady-state concentrations of vitamin E in membranes are determined by (1) the efficiency of incorporation into membranes following transfer from blood lipoproteins and (2) its metabolism in membranes. The main intramembrane metabolic pathway of vitamin E is believed to 26E. A. Serbinova, V. E. Kagan, D. Han, and L. Packer, Free Radical Biol. Med. 10, 263 (1991).
362
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[34]
be scavenging of lipid radicals in the course of initiation and propagation of lipid peroxidation. 3'8'27 The vitamin E free radical is formed during reaction (1). It has been generally believed that great mobility of vitamin E in the lateral plane of the membrane and its exact positioning in the membrane are extremely important for this reaction. Vitamin E is anchored in the hydrocarbon part of membrane bilayer by the phytol tail which is 13 carbons long, just the right length to position the chromanol nucleus, which possesses the antioxidant activity, at the membrane interface. There, vitamin E through its phenolic hydroxyl group quenches free radicals, in the process becoming the (phenoxyl or chromanoxyl) tocopheroxyl radical. Tocopheroxyl free radicals are less reactive than other lipid radicals (peroxyl or alkoxyl radicals) generated in membranes and thus serve to break the chain of free radical reactions in lipid peroxidation. However, the free radical, tocopheroxyl form of vitamin E is susceptible to oxidation or to destruction by reacting with itself or by other reactions that cause it to decompose as a result of radical-initiated reactions. Unless reduced (regenerated) to its original antioxidant form, it will be lost before prooxidant reactions occur. This type of vitamin E action has been suspected for a long time. In fact it was proposed in 1966 by Mellors and Tappel that the ubiquinone components of membranes may serve to protect vitamin E against loss by interacting with it. 2s It is also known from in vitro studies that ascorbate (vitamin C) regenerates vitamin E from its free radical form, but whether this was an important activity of membranes was not known. Studies in the laboratory of Burke revealed that reduced glutathione (GSH), the primary preventative water-soluble antioxidant in most aerobic cells, protects against lipid perodixation of microsomal membranes in vitro. However, glutathione does not exhibit this activity if membranes are prepared from vitamin E-deficient animals. From this finding Bast, McCay, and others have suggested the existence of a vitamin E free radical reductase activity, that is, some enzyme or enzyme systems capable of specifically regenerating vitamin E, and that glutathione may be one of the substrates for this type of activity.29'3° However no direct experimental data on enzymatic regeneration of vitamin E was obtained. The only way to elucidate these types of membrane reactions of vitamin E is to follow 27 G. 28 A. 29 A. 30 p.
W. Burton and K. U. Ingold, Acc. Chem. Res. 19, 194 (1986). Mellors and A. L. Tappel, J. Biol. Chem. 214, 4353 (1966). Bast and G. R. M. M. Haenen, Biochim. Biophys. 574, 537 (1988). B. McCay, Annu. Reo. Nutr. 5, 323 (1985).
[34]
t~-TOCOPHEROL AND ot-TOCOTRIENOL
363
dynamically the reactions of the vitamin E tocopheroxyl radicals in natural membranes using highly sensitive electron spin resonance (ESR) techniques.26, 31-33
Generation of Chromanoxyl Radicals Generation of Chromanoxyl Radicals in Microsomes, Mitochondria, and Low Density Lipoproteins. Chromanoxyl radicals from a-tocotrienol and a-tocopherol are generated using (1) an enzymatic oxidation system (soybean 15-1ipoxygenase plus linolenic acid), (2) a hydrophobic azo-initiator of peroxyl radicals, AMVN, and (3) UV i r r a d i a t i o n . 26'31-33 When the enzymatic oxidation system is used the incubation medium (100/~1) contains low density lipoprotein (LDL) (9-13 mg protein/ml), microsomes or mitochondria (30-50 mg protein/ml), or liposomes (30 mg lipids/ml) in 50 mM phosphate buffer, pH 7.4 at 25°. The concentration of exogenously added chromanols is 80 nmol/mg protein. Linolenic acid (1.4 mM) plus lipoxygenase (10 U//A) and chromanols are subsequently added to the LDL suspension. Chromanols are added in ethanolic solution. With the azo-initiator the incubation medium is essentially the same, but AMVN (5.0 mM) is added instead of lipoxygenase plus linolenic acid and the reaction carried out at 40 °. Generation of Chromanoxyl Radicals in Erythrocyte Ghosts. Chromanoxyl radicals from a-tocopherol and homologs are generated in erythrocyte ghosts using an enzymatic oxidation system consisting lipoxygenase and arachidonic a c i d . 34 The reaction medium (50/zl) contains erythrocyte membrane suspension (5 mg/ml protein), arachidonic acid (1.97 mM), lipoxygenase (4.5 U//xl), and chromanols (9 mM) in 50 mM phosphate buffer, pH 7.4 at 20°. The reaction is started with lipoxygenase addition. Incorporation of Chromanols in Membranes. Membranes are preincubated with chromanols (added from ice-cold ethanolic solution) for 20 min at 25 ° (for microsomal and mitochondrial membranes) or at 20° (for erythrocyte ghosts). The suspensions are centrifuged for 60 min at 105,000 g at 4 ° and for 20 min at 30,000 g at 4° for microsomal and mitochondrial membranes or erythrocyte ghosts, respectively. Under the above conditions the amount of incorporated a-tocopherol/a-tocotrienol in all types of membranes is 84-90%. 31 L. Packer, J. J. Maguire, R. J. Mehlhorn, E. A. Serbinova, and V. E. Kagan, Biochem. Biophys. Res. Commun. 159, 229 (1989). 32 V. Kagan, E. Serbinova, T. Forte, G. Scita, and L. Packer, J. Lipid Res. 33, 385 (1992). 33 V. Kagan, E. Serbinova, and L. Packer, Arch. Biochem. Biophys. 280, 33 (1990).
364
ANTIOXIDANT CHARACTERIZATION AND ASSAY
40°C
CONTROL
40°C
+AMVN
40°C
+UV
40°C
+UV + AMVN
25°C t,.,~
~
[34]
+LIPOXYGENASE+ LINOLENICACID
g = 2.00 FIG. 4. Electron spin resonance spectra of chromanoxyl radicals generated in LDL from endogenous vitamin E.
Irradiation
Irradiation is achieved by a solar simulator (Solar Light Co., Model 14S), whose output closely matches the solar spectrum in the wavelengths 290-400 nm. The samples are illuminated directly in the ESR resonator cavity; the distance between the light source and the sample is
~~x~ MICROSOMES
~ ~ ~ LIPOSOMES
FIG. 5. Electron spin resonance spectra of chromanoxyl radicals generated from atocopherol and a-tocotrienol by an enzymatic oxidation system (lipoxygenase plus linolenic acid) in microsomes or liposomes.
[34]
O~-TOCOPHEROL AND O~-TOCOTRIENOL
365
T A B L E II RECYCLING EFFICIENCY AND DELAY TIME FOR REAPPEARANCE OF CHROMANOXYL RADICALS FROM ot-TocOPHEROL AND ot-ToCOTRIENOL IN RAT LIVER MICROSOMES Antioxidant
Delay time (min) a
Recycling efficiency b
(+)-~-Tocopherol (+)-a-Tocotrienol
1.0 - 0.2 ¢ 3.0 -+ 0.3 c
0.23 - 0.02 c 0.37 - 0.04 c
a The delay time was measured after addition of ascorbyl palmitate. b The recycling efficiency was measured after addition of NADPH. c Values were averaged from 5 data points.
30 cm. The power density of the light at the sample surface in the spectral region 310-400 nm is 1.5 mW/cm 2 and drops to 10% of this value at 290 nm.
Electron Spin Resonance Measurements The ESR measurements are made on a Varian E 109E or Brucker IBM ER 200 D-SRC spectrometer in gas-permeable Teflon tubings (0.8 mm internal diameter, 0.013 mm thickness) obtained from Zeus Industrial Products (Raritan, NJ). The gas-permeable tube ( - 8 cm in length) is filled with 60 ~1 of a mixed sample, folded into quarters, and placed in an open 3.0 mm internal diameter ESR quartz tube such that all of the sample is within the effective microwave irradiation area. ESR spectra are recorded either in the dark or under continuous UVAB-irradiation by the solar simulator in the ESR cavity. Spectra are recorded at 100 mW power and 2.5 gauss modulation, and 25 gauss/min scan time. Spectra are recorded at room temperature under aerobic conditions by flowing oxygen through the ESR cavity. Chromanoxyl and ascorbyl radical ESR signals are recorded at 3245 gauss magnetic field strength, scan range 100 gauss, and time constant 0.064 sec.
Efficiency of Chromanoxyl Radical Reduction To evaluate quantitatively the efficiency of chromanoxyl radical reduction the recycling efficiency coefficient (Re) can be calculated:
Re
=
(A_re d -
A+red)/Are d
where A_red and A +red are the magnitudes of ESR signals of chromanoxyl radicals in the absence and presence of a reductant, respectively. The
366
ANTIOXlDANT CHARACTERIZATION AND ASSAY
[34]
values of recycling efficiency vary from 1 to 0, which correspond to 100% reduction (complete transient disappearance of ESR signal) and 0% reduction (no effect on ESR signal), respectively. 34 Under the conditions described chromanoxyl radicals from vitamin E and its homologs are generated using all three methods (Fig. 4). 33 Both a-tocopherol and a-tocotrienol produce characteristic pentameric chromanoxyl radical signals with component g values of 2.0122 2.0092, 2.0061, 2.0028, and 1.9993 both in microsomes and in liposomes (Fig. 5). 26'35'36 It has been reported that ~-tocotrienol radical ESR signals are significantly higher than those of ~-tocopherol in the presence of either microsomes or liposomes. 26 Addition of NADPH to the microsomal suspension results in a decrease of the magnitude (but not complete disappearance) of the ESR signals of a-tocopherol (or a-tocotrienol). The NADPH-dependent decrease of ESR signals is much more pronounced for a-tocotrienol than for a-tocopherol. The data presented on Table II show that in microsomes NADPH-supported recycling efficiency (Re) for o~-tocotrienol is higher than for a-tocopherol. Also, the delay time of chromanoxyl radical ESR signal reappearance after addition of ascorbyl palmitate is greater for o~-tocotrienol than for a-tocopherol (Table II). These results show that o~-tocotrienol has a higher recycling efficiency than a-tocopherol. The higher recycling efficiency of a-tocotrienol must be contributing to its higher antioxidant activity compared to a-tocopherol. However, whereas the recycling efficiency and the delay time for a-tocopherol are only about 1.6 and 2.5-3 times less than those for o~-tocotrienol, the concentrations exerting 50% inhibition of lipid peroxidation differ by 40-60 times. This indicates that the higher antioxidant activity of a-tocotrienol in vitro must result from the contribution of other factors in addition to its higher recycling efficiency, such as more uniform distribution in membrane bilayer and stronger disordering of membrane lipids compared to c~-tocopherol.26 Conclusion It is recognized that differences in vivo in the antioxidant activity of tocopherols and tocotrienols may depend greatly on their pharmacokinetics. However ot-tocotrienol may have higher antioxidant activity in oioo under conditions of oxidative stress owing to its more effective antioxidant potency in membranes. 34 A. Constantinescu, D. Han, and L. Packer, J. Biol. Chem. 268(15), 10906 (1993). 35 K. Mukai, K. Takamatsu, and K. Ishizu, Bull. Chem. Soc. Jpn. 57, 3507 (1984). 36 K. Mukai, N. Tsuzuki, S. Ouchi, and K. Fukuzawa, Chem. Phys. Lipids 30, 337 (1982).
[35]
HPLC DETERMINATIONOF GSSG IN BLOOD
367
[35] D e t e r m i n a t i o n o f O x i d i z e d G l u t a t h i o n e in B l o o d : High-Performance Liquid Chromatography B y M I G U E L ASENSI, JUAN SASTRE, FEDERICO V . PALLARDO,
JOSE M. ESTRELA, and JOSE Vlr~A Introduction The measurement of glutathione status 1 is important in determining oxidative stress 2 in tissues and biological fluids. The ratio of reduced to oxidized glutathione (GSH/GSSG) is thus a good indicator of the oxidative stress that may occur under physiological and pathological conditions. 3'4 Changes in GSSG levels have been considered as intracellular signals able to modulate enzyme activity. 5,6 Thus, it is important to have accurate methods to determine GSSG in biological fluids and in cells. In many cases, it is possible to use tissues such as liver, muscle, or brain to determine GSH/GSSG. However, especially in human studies, samples from these tissues are not readily available, and measurement of blood samples is required. A major problem is the measurement of GSSG in the presence of GSH because spontaneous or catalyzed GSH oxidation must be prevented. Indeed, assuming that glutathione reductase (GR) is at equilibrium, we calculated7 that the ratio of GSH to GSSG must be about 105. Even if this is not the case, that is, if GR is not at equilibrium and if the GSH/GSSG ratio is about 100, a 2% oxidation of GSH will cause a 100% change in GSSG. Thus, oxidation of GSH must be kept to a minimum if measurements of GSH/GSSG are to be meaningful. Several methods have been devised to measure GSH and GSSG. 8-11 l N. S. Kosower and E. M. Kosower, Int. Rev. Cytol. 54, 109 (1978). 2 H. Sies, Angew. Chem. 25, 1058 (1986). 3 j. Vifia, ed., "Glutathione: Metabolism and Physiological Functions." CRC Press, Boca Raton, Florida, 1990. 4 A. Meister, in "Metabolism and Functions of Glutathione" (D. Dolphin, R. Poulson, and O. Avramovic, eds.), pp. 367-474. Wiley, New York, 1989. 5 H. F. Gilbert, J. Biol. Chem. 257, 12086 (1982). 6 H. F. Gilbert, this series, Vol. 107, p. 330. 7 j. Vifia, R. Hems, and H. A. Krebs, Biochem. J. 170, 627 (1978). 8 M. W. Fariss and D. J. Reed, this series, Vol. 143, p. 101. 9 T. P. M. Akerboom and H. Sies, this series, Vol. 77, p. 373. 10 F. A. M. Redegeld, A. S. Koster, and W. P. van Bennekom, in "Glutathione: Metabolism and Physiological Functions" (J. Vifia, ed.), pp. 11-20. CRC Press, Boca Raton, Florida, 1990. ti R. C. Fahey and G. L. Newton, this series, Vol. 143, p. 85.
METHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
368
ANTIOXlDANT CHARACTERIZATION AND ASSAY
[35]
Most methods rely on chelating thiols in order to prevent GSH oxidation. An excellent, probably the best, thiol chelator is N-ethylmaleimide (NEM).12 However, it must be removed before GSSG measurements because it inactivates the enzymes used for such measurements .9 Thus, highperformance liquid chromatography (HPLC) methods are preferred to measure glutathione. One of the most widely used HPLC methods was des.cribed by Reed and colleagues in the early 1980s. a We have used this method and found oxidation of internal standards of GSH added to extracts of liver or kidney ranging from about 3 to 4%. When blood was used, however, oxidation was up to 30%. This appears to have been the case in other studies in which GSSG was measured in blood using this method. 13 In this chapter, an HPLC method to determine GSSG in blood is described which minimizes GSH oxidation to about 0.25%. Assay Procedure
Principle. The accurate measurement of GSSG in the presence of GSH relics on rapid and effective GSH quenching. To prevent GSH oxidation during sample preparation, several quenching agents, such as N-ethylmaleimide (NEM), 2-vinylpyridine, and iodoacetic acid, have been used to alkylate thiol groups. Among them, NEM is preferred because of its rapid reaction rate (completion within I min), in contrast to 2-vinylpyridine (20-50 rain) or iodoacetic acid (5-15 rain)) ° Furthermore, treatment with NEM can be achieved on ice and in acidic medium, which minimize GSH oxidation, whereas quenching by 2-vinylpyridine or iodoacetic acid occurs at room temperature and in a neutral or basic medium. Thus, GSH quenching by NEM under acidic conditions is the most convenient because it prevents GSH autoxidation that may occur in a neutral or basic medium. According to our method, blood samples are treated with pcrchloric acid (5% final concentration) containing NEM (20 mM final concentration) and bathophenanthrolinedisulfonic acid (1 mM final concentration) as metal chelator. Then, blood samples are derivatized and analyzed by HPLC to determine GSSG. Reagents 12% Perchloric acid (PCA) containing 40 mM N-ethylmaleimide (NEM) and 2 mM bathophenanthrolinedisulfonic acid (BPDS) 1 mM 3,-Glutamylglutamate (Glu-Glu) prepared in 0.3% perchloric acid 12j. p. Richie and C. A. Lang, Anal. Biochem. 163, 9 (1987). 13 K. Gohii, C. Viguie, W. C. Stanley, G. A. Brooks, and L. Packer, J. Appl. Physiol, 64, 115 (1988).
[35]
HPLC DETERMINATIONOF GSSG IN BLOOD
369
2 M Potassium hydroxide (KOH) containing 0.3 M 3-(N-morpholino)propanesulfonic acid (MOPS) 1% 1-Fluoro-2,4-dinitrobenzene (FDNB) dissolved in ethanol. Mobile phase A: 80% methanol (HPLC grade), 20% water (HPLC grade) Mobile phase B: Prepared by adding 800 ml of a stock sodium acetate solution to 3.2 liters of solvent A; the stock sodium acetate solution is prepared by adding 1 kg sodium acetate (HPLC grade) and 448 ml of water (HPLC grade) to 1.39 liter of glacial acetic acid (HPLC grade) 8
Sample Preparation 1. Add 0.5 ml of whole blood to 0.5 ml of ice-cold 12% PCA containing 40 mM NEM and 2 mM BPDS. Blood samples must be treated with PCA immediately after extraction from the animal or subject. Mix thoroughly. 2. Centrifuge at 15,000 g for 5 min at 4°. 3. Take 0.5 ml of the acidic supernatant and keep it on ice until derivatization. Samples can also be stored frozen at - 2 0 ° for 1 week.
Derivatization 1. Add 50 tzl of 1 mM glutamylglutamate and 10/xl of a pH indicator solution to 500 tzl of acidic supernatant. 2. Adjust to pH 8.0-8.5 with 2 M KOH containing 0.3 M MOPS to prevent excessive alkalinization. Check the pH after neutralization with a pH meter. 3. Centrifuge samples at 15,000 g for 5 min. 4. Add an aliquot of 25/zl of each supernatant to 50/zl of 1% 1-fluoro2,4-dinitrobenzene in a small glass tube. After 45 min of incubation in the dark at room temperature, the derivatized samples are desiccated under vacuum and stored at - 20° in the dark until injection. Samples processed in this way are stable for several weeks.
Analysis by Chromatography. Samples processed as mentioned above are dissolved in 50/~1 of 80% methanol (mobile phase A) and injected onto the HPLC system. A Spherisorb NH2 column (20 x 0.4 cm, 5 tzm particles) is used. An NH2-~Bondapak column is also suitable for this method. The flow rate is 1.0 ml/min during the procedure. The mobile phases and the gradient are as followsS: solvent A is 80% methanol, and solvent B is 0.5 M sodium acetate in 64% methanol, prepared as described by Fariss and Reed. 8 After injection of 25 ~1 of derivatized solution, the mobile phase is held at 80% A, 20% B for 5 min followed by a 10-min linear gradient up to 1% A, 99% B. 8 The mobile phase is held
370
[35]
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
8.0
6°0
.1 0
4.0
2.0 i
0o0
'
0.0
'
I 2.0
'
'
' 4.0
'
I
'
'
i
'
6.0
I 8.0
'
'
' lO.O
I
'
I 12.0
'
~
I 14.0
MINUTES
FIo. 1. Chromatogram of the N-dinitrophenyl derivatives of a blood sample processed and analyzed as described in the text. The retention time for the N-dinitrophenyl derivative of GSSG was 11.78 rain. The G S H - N E M adduct decomposes and appears as three peaks for the corresponding N-dinitrophenyl derivatives (retention times: 3.16, 5.39, 7.02 rain). Ten volts on the y axis is equivalent to 0.05 AOD units at 365 nm.
at 99% B until GSSG has eluted. Using this method, chromatograms such as that shown in Fig. 1 are obtained. The G S H - N E M adduct decomposes and appears as three peaks (see Fig. 1). Thus, GSH must be measured by an enzymatic method, such as the one using glyoxalase7 or glutathione transferasefl 4 in an aliquot to which no NEM has been added. 14 R. Brigelius, C. Muckel, T. P. M. Akerboom, and H. Sies, Biochem. Pharmacol. 32, 2529 (1983).
'
[36]
ASSAY ANTIOXIDANTS WITH ciS-PARINARIC ACID
371
Applications When we measured GSH and GSSG levels in blood samples following the method described by Fariss and Reed, 8 we obtained remarkably high GSSG levels. Indeed, GSH/GSSG ratios were 10 or lower for all blood samples assayed. When we added 0.5 ml of 12% PCA containing 2 mM BPDS and 1/zM GSH to 0.5 ml of whole blood and assayed the samples following the method by Fariss and Reed, 8 we found that 23 -- 6% (n = 5) of the GSH present originally in the standard solution was oxidized. When our method was used, the percent of oxidation of an standard GSH solution (1084 nmol/ml) was 0.22 - 0.23% (n = 4). High GSSG levels can be erroneously obtained owing to oxidation of GSH during sample preparation, especially with blood samples. This may lead to erroneous conclusions concerning the pathophysiological changes of glutathione status. For instance, Gohil et al.13 reported GSH/GSSG ratios in blood of about 1. These artifactual increases in GSSG may be important when trying to assess oxidative status. For instance, Gohil et al. 13reported that in humans changes in GSSG in blood during exhaustive exercise were not related to changes in blood lactate. When we repeated these experiments, but using the present method for GSSG determination, we observed an excellent linear relationship between GSSG/GSH and lactate/pyruvate ratios.~5 15 j. Sastre, M. Asensi, E. Gasc6, F. V. Pallard6, J. A. Ferrero, T. Furukawa, and J. Villa, Am. J. Physiol. 32, R992 (1992).
[36] A n t i o x i d a n t A c t i v i t y o f a - T o c o p h e r o l , f l - C a r o t e n e , a n d U b i q u i n o l in M e m b r a n e s : c i s - P a r i n a r i c Acid-Incorporated Liposomes By MASAHIKO TSUCHIYA, VALERIAN E. K A G A N , HANS-JOACHIM FREISLEBEN, MASANOBU MANABE, and LESTER PACKER
Introduction Owing to their high reactivities, oxygen free radicals, which are generated by various biological and chemical processes in vivo, are potentially dangerous to living cells. These radicals can induce oxidative destruction of the polyunsaturated fatty acyl chains of membrane lipids by the processes known as lipid peroxidation. The resultant loss of membrane integMETHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress.Inc. All rightsof reproductionin any formreserved.
[36]
ASSAY ANTIOXIDANTS WITH ciS-PARINARIC ACID
371
Applications When we measured GSH and GSSG levels in blood samples following the method described by Fariss and Reed, 8 we obtained remarkably high GSSG levels. Indeed, GSH/GSSG ratios were 10 or lower for all blood samples assayed. When we added 0.5 ml of 12% PCA containing 2 mM BPDS and 1/zM GSH to 0.5 ml of whole blood and assayed the samples following the method by Fariss and Reed, 8 we found that 23 -- 6% (n = 5) of the GSH present originally in the standard solution was oxidized. When our method was used, the percent of oxidation of an standard GSH solution (1084 nmol/ml) was 0.22 - 0.23% (n = 4). High GSSG levels can be erroneously obtained owing to oxidation of GSH during sample preparation, especially with blood samples. This may lead to erroneous conclusions concerning the pathophysiological changes of glutathione status. For instance, Gohil et al.13 reported GSH/GSSG ratios in blood of about 1. These artifactual increases in GSSG may be important when trying to assess oxidative status. For instance, Gohil et al. 13reported that in humans changes in GSSG in blood during exhaustive exercise were not related to changes in blood lactate. When we repeated these experiments, but using the present method for GSSG determination, we observed an excellent linear relationship between GSSG/GSH and lactate/pyruvate ratios.~5 15 j. Sastre, M. Asensi, E. Gasc6, F. V. Pallard6, J. A. Ferrero, T. Furukawa, and J. Villa, Am. J. Physiol. 32, R992 (1992).
[36] A n t i o x i d a n t A c t i v i t y o f a - T o c o p h e r o l , f l - C a r o t e n e , a n d U b i q u i n o l in M e m b r a n e s : c i s - P a r i n a r i c Acid-Incorporated Liposomes By MASAHIKO TSUCHIYA, VALERIAN E. K A G A N , HANS-JOACHIM FREISLEBEN, MASANOBU MANABE, and LESTER PACKER
Introduction Owing to their high reactivities, oxygen free radicals, which are generated by various biological and chemical processes in vivo, are potentially dangerous to living cells. These radicals can induce oxidative destruction of the polyunsaturated fatty acyl chains of membrane lipids by the processes known as lipid peroxidation. The resultant loss of membrane integMETHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress.Inc. All rightsof reproductionin any formreserved.
372
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[36]
rity and function is implicated in such pathological conditions as inflammation, diabetes, reperfusion injury, radiation damage, cancer, aging, and neurological diseases. However, in healthy living cells, oxidative processes can be ingeniously intercepted by a network of interacting antioxidants. a-Tocopherol, a lipid-soluble membrane constituent, has been demonstrated to be an essential factor in the cellular antioxidant defense system.2 By donating a hydrogen atom, it functions as an efficient chain-breaking antioxidant that blocks lipid peroxidation. 3 Other biological compounds, including carotenoids and ubiquinones, appear also to play a role in protection of biological membranes against oxygen free radicals. Carotenoids inhibit peroxidation of methyl linoleate and microsomal lipids,4'5 as well as peroxyl radical-initiated fatty acid peroxidation in hexane. 6 Mitochondrial membranes depleted of ubiquinones appear to be more sensitive to oxidative damage by lipid peroxidation inducers,7 and, functioning in its reduced or semireduced form, ubiquinone prevents human low density lipoprotein (LDL) oxidation. 8 Although these natural antioxidants may react directly with oxygen free radicals, it has been suggested that they eliminate radicals indirectly via a recycling process that conserves ct-tocopherol stores. 9 This process may be a very important antioxidant defense system in vivo. The fundamental chemistry of these reactions has been investigated in detail, mainly in homogeneous solutions and/or aqueous dispersions, and is now believed to be fairly well understood. 3A°Such chemical studies, however, are not sufficient to understand antioxidant reactions in vivo, since vital antioxidants function in connection with biomembranes, which compartmentalize the reactions. We have previously demonstrated that antioxidant activity is highly dependent on the environment. 6 The effects of membranes on antioxidant reactions must therefore be taken into consideration. Approaches to this problem include biochemical studies with natural materials such as membranous fractions, whole cells, or tissues,
1 B. Halliwell and J. M. C. Gutteridge, "Free Radicals in Biology and Medicine," 2nd Ed. Oxford Univ. Press (Clarendon), Oxford, 1989. 2 L. Packer, Am. J. Clin. Nutr. 53, 1050S (1991). 3 G. W. Burton and K. U. Ingold, J. Am. Chem. Soc. 103, 6472 (1981). 4 j. Terao, Lipids 24, 659 (1989). 5 p. Palozza and N. I. Krinsky, Free Radical Biol. Med. U , 407 (1991). 6 M. Tsuchiya, G. Scita, H. J. Freisleben, V. E. Kagan, and L. Packer, this series, Vol. 213, p. 460. 7 V. C. Joshi, J. Jayaraman and T. Ranasarma, Biochem. J. 88, 25 (1963). 8 R. Stocker, V. W. Bowry, and B. Frei, Proc. Natl. Acad. Sci. U.S.A. 88, 1646 (1991). 9 V. E. Kagan, E. A. Serbinova, and L. Packer, Arch. Biochem. Biophys. 282, 221 (1990). l0 E. Niki, T. Saito, A. Kawakami, and Y. Kamiya, J. Biol. Chem. 259, 4177 (1984).
[36]
ASSAY ANTIOXIDANTS WITH ciS-PARINARIC ACID
373
that is, systems that mimic well in vivo reactions, u,12 Even these methods, however, cannot fully clarify the details of the reactions owing to the lack of reliable methodology under these complex conditions, and there is still an experimental gap between fundamental chemical studies and biochemical studies aimed at understanding the reactions of vital antioxidants in the presence of membranes. One technique that has helped to close this gap is electron spin resonance (ESR) spectroscopy, which can sensitively detect and elucidate short-duration antioxidant reactions with free radicals in the presence of membranes. A drawback is that ESR spectroscopy requires special large-scale equipment. 9,13 Recently, it was demonstrated that a fluorescent polyunsaturated fatty acid, cis-parinaric acid, can be used as a sensitive and reliable reporting molecule for peroxidation in membranes.14 It shows potent fluorescence, as well as high susceptibility to peroxidation by various oxygen free radicals. Attenuation of its fluorescence is a good index of the oxidative stress at the site where cis-parinaric acid is present. In addition, this lipid is readily and easily incorporated into membranes and causes little unfavorable disturbance of the lipid bilayer. 14 Thus, it is considered to be a sensitive and almost ideal probe that allows direct continuous monitoring of oxidative stress in membranes. Moreover, cis-parinaric acid-incorporated liposomal membranes provide a good model for in vivo reactions, although this is a quite simple system. In the present study, we have used this model to investigate the antioxidant activity of a-tocopherol and other natural compounds against peroxyl radicals, major oxygen free radicals that damage biomembranes. Materials and Methods
Chemicals 2,2'-Azobis(2,4-dimethylvaleronitrile) (AMVN) is purchased from Polysciences, Inc. (Warrington, PA) and cis-parinaric acid from Molecular Probes (Junction City, OR). Butylated hydroxytoluene (BHT), fl-carotene, dioleoylphosphatidylcholine (DOPC), isoluminol, and methyl linoleate ubiquinol 10 are from Sigma Chemical Company (St. Louis, MO). 11 R. Ferrari, O. Vesioli, C. Guarnieri, and M. Caldarera, Acta Vitaminol. Enzymol. 5, 11 (1983). 12 C. Michielis, M. Race, and J. Remacle, Arch. Int. Physiol. Biochim. 94, S13 0986). 13 V. E. Kagan, E. A. Serbinova, T. Forte, G. Scita, and L. Packer, J. Lipid Res. 33, 385 0992). 14 F. A. Kuypers, J. J. M. van den Berg, C. Schalkwijk, B. Roelofsen, and J. A. F. Op den Kamp, Biochim. Biophys. Acta 921, 266 (1987).
374
[36]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
a-Tocopherol and 2,2,5,7,8-pentamethyl-6-chromanol (PMC) are gifts from Henkel Corp. (LaGrange, IL) and Eisai Co. (Tokyo, Japan), respectively. Other reagents are commercial products of analytical grade.
Rate of Radical Generation by Azo Initiator AMVN, [CN(CH3)z--CH2--C(CH3)CN--N=N--C(CH3)CN-CHz--CH(CH3)z], a diazo compound, is thermally decomposed without enzymes or biotransformation to yield peroxyl radicals as follows15: (I
-
R - - N ~ N - - R - - * R. + N2 + R. --~ 2eR-
e)R--R
+ N2
R. + 02--~ ROO. where e is the efficiency of free radical production. The rate constant, K~, of peroxyl radical generation from AMVN is obtained by the method of Barclay and Ingold 16 and Niki et al. 17 In brief, a known concentration of a chain-breaking phenolic antioxidant, BHT or PMC (both of which have been shown to react with exactly two peroxyl radicals and terminate oxidation chains) is added to a system where oxidation is induced only by the decomposition of AMVN, and the induction period, t~nh, is measured during which oxidation is suppressed. The rate constant kl is given by Eq. (1)16'17: k~ = 2[IN]/tinh[AMVN]
(1)
where [IN] and [AMVN] are the concentrations of PMC (or BHT) and AMVN, respectively, in the system. The oxidation of a sample is measured by monitoring the oxygen consumption of the reaction mixture with an oxygen electrode. The reaction mixture contains 1.67 mM AMVN, 100 mM methyl linoleate, and various concentrations of PMC (or BHT) in 10 mM Triton X-100 solution at 45 °. Concentrations of PMC and BHT are confirmed by spectrophotometric measurement at 292 nm [log(extinction coefficient) = 3.54] and 277 nm [log(extinction coefficient) = 3.34], respectively. 18
Measurement of Fluorescence of cis-Parinaric Acid in Liposomes The DOPC liposomes with incorporated cis-parinaric acid and antioxidants are prepared by sonication of 6 txM cis-parinaric acid, various antiox~5 E. Niki, this series, Vol. 186, p. 100. 16 L. R. C. Barclay and K. U. Ingold, J. Am. Chem. Soc. 103, 6478 (1981). 17 E. Niki, M. Saito, Y. Yoshikawa, Y. Yamamoto, and Y. Kamiya, Bull. Chem. Soc. Jpn. 59, 471 (1986). is R. C. Weast, ed., "Handbook of Chemistry and Physics," 52nd Ed. Chem. Rubber Publ. Co., Cleveland, OH, 1971.
[36]
ASSAY ANTIOXIDANTS WITH ciS-PARINARIC ACID
375
idants, and 1.3 mM DOPC dispersion in 20 mM phosphate buffer (pH 7.4) under nitrogen gas at 4°. The oxidation is started by quickly raising the sample temperature to 45 ° just after the incorporation of 300/xM AMVN into the DOPC liposomes by further sonication. The fluorescence intensity is then monitored at an excitation wavelength of 324 nm with a 5 nm slit and an emission wavelength of 413 nm with a 5 nm slit at 45 °, using a Perkin-Elmer (Norwalk, CT) MPF-44A spectrofluorometer, a-Tocopherol and/3-carotene concentrations are determined by high-performance liquid chromatography (HPLC). Ubiquinol 10 is obtained by the reduction of ubiquinone 10 with NaBH 4 in hexane/ethanol solution, and the concentration of the reduced form is spectrophotometrically determined r~-I% ~ L ' I cm = 46.4 at 290 nm). ~9 Concentrations of ascorbate and cis-parinaric acid are determined using the extinction coefficients of 1.45 × 10 3 M - l c m -1 at 265 nm z° and 80 x 10 3 M -~ c m - l at 303 nm (or 74 x 103 M -1 at 318 nm), 2~ respectively.
Measurement of Lipid Hydroperoxides Resulting from Peroxidation of cis-Parinaric Acid with Chemiluminescence Chromatography System A hexane extract of the above DOPC liposomes after 60 min of reaction with AMVN is analyzed by a chemiluminescence HPLC system that can sensitively detect lipid hydroperoxide products. The measurement is performed according to the method developed by Ames and others. 2z The HPLC system consists of a chemiluminescence detector S-3400 (Soma Optic Ltd., Japan), a reversed-phase octadecylsily column No. 235329 (Beckman, San Ramon, CA), a pulseless low pressure pump LPP297 (Lazar Research Laboratories, California) for isoluminol, an HPLC pump 114 M (Beckman) for mobile phase, and an HPLC injector 7125 (Rheodyne, Coati, CA). Samples are chromatographed using methanol as the mobile phase; eluted lipid hydroperoxides are detected by the chemiluminescence reaction with isoluminol. Results
Rate of Radical Generation by Azo Initiator The rate constant of radical generation by AMVN, ki, at 45 ° was obtained from Fig. 1B by using Eq. (I). The experimental value of ~9F. 20 G. 21 L. 22 y .
L. Crane and R. Barr, this series, Vol. 18C, p. 137. R. Buettner, Free Radical Res. Comrnun. 10, 5 (1990). A. Sklar, B. S. Hudson, and R. D. Simoni, J. Supramol. Struct. 4, 449 (1976). Yamamoto, M. H. Brodsky, J. C. Baker, and B. N. Ames, Anal. Biochem. 160, 7 (1987).
376
ANTIOXIDANT CHARACTERIZATION AND ASSAY A
120
[36]
B
+AMVN
60
-
80
40
~"
+AMVN +BHT
0
40
20
.~,oo //-~.VN-I[ ~ 0
0
~
I ~
/
7
"PMC l OBHT
~
I
20 40 60 80 0 2 4 6 8 Time (min) 10 3[IN]/[AMVN]
0
FIG. 1. (A) Rate of oxygen uptake induced by AMVN and induction period by BHT during the oxidation of 100 mM methyl linoleate dispersion at 45° in the presence of l0 mM Triton X-100, 1.67 mM AMVN. (B) Plot of induction period as a function of the ratio of PMC (or BHT) concentration [IN] to AMVN concentration [AMVN]. 4.41 × 10 -6 sec -~ is in agreement with that previously reported by B raughler and Pregenzer at 37 °.23 Thus, the rate of peroxyl radical production by A M V N in this experiment is 1.32 × 10 - 9 M sec -].
Effects of Dioleoylphosphatidylcholine and cis-Parinaric Concentration on Fluorescence The relation between the fluorescence intensity of cis-parinaric acid and the concentration of DOPC is shown in Fig. 2A. The intensity reached a plateau at concentrations greater than 0.8 m M DOPC, indicating incorporation o f whole fractions of probe and minimal collisional interactions of the probe throughout this range. The fluorescence intensity increased linearly with increasing cis-parinaric acid concentration from 0.5 to 6.5 M when the DOPC concentration was in the plateau range (Fig. 2B). This enables estimation of the probe concentration under these conditions.
Time Course and Spectra of Fluorescence Decay of cis-Parinaric Acid Induced by Azo Initiator and Concomitant Lipid Hydroperoxide Production The fluorescence of cis-parinaric acid (excitation at 324 nm and emission at 413 nm) was stable in the absence of A M V N while the measurement 23 j. M. Braughler and J. F. Pregenzer, Free Radical Biol. Med. 7, 125 (1989).
[36]
ASSAY ANTIOXIDANTS WITH
A
ciS-PARINARIC ACID
377
B
S 1oo 80
60 t'-
40
3
zo
w_ 0
I
0.0
I
1.0
2.0
DOPC(rr~)
I
3.0
i
O..0
2.0
i
4.0
6).0
cis-Padnaric Acid (/aM)
FIG. 2. Fluorescence intensity of cis-parinaric acid in DOPC liposomes as a function of DOPC (A) and cis-parinaric acid concentration (B). The reaction mixture contained 6/xM cis-parinaric acid and DOPC as indicated (A), or 1.3 mM DOPC and cis-parinaric acid as indicated (B) in 20 mM phosphate buffer (pH 7.4).
was performed (Fig. 3). With AMVN-incorporated liposomes, the fluorescence decreased to about 20% of maximal intensity in 60 min. The analogous decrease in the fluorescence spectra indicates that the decrease in fluorescence intensity is a result of the actual decomposition and disappearance of cis-parinaric acid. The peroxidation of cis-parinaric acid was further confirmed by chemiluminescence HPLC (Fig. 4). Lipid hydroperoxide with a retention time of 8.9 min was eluted from samples of the fluorescence experiment during the 60-min reaction. It was completely undetected in the absence of AMVN.
Effects of Antioxidants on Azo Initiator-Induced Fluorescence Decrease of cis-Parinaric Acid Various natural antioxidants, such as a-tocopherol,/3-carotene, and ubiquinol 10 (reduced form of ubiquinone 10), that had previously and individually been incorporated into DOPC liposomes prevented cis-parinaric acid fluorescence decay by AMVN in a concentration-dependent
378
A N T I O X I D A NCHARACTERIZATION T AND ASSAY A
B I
I
C ~
I
I
I
I
"~ 100
I
I
/ I
"~ D "~
! ~
I
O0 "~
nin
"~ D
0 min ")
80
..0 80
.*2_
i
CO 60 E nin
C.I +AMVN ~ m 40 C +{x_t ocopherol/X Q) 0to 20 0
--~ U_
[36]
t
/
cco
.*2_ 60 rm
0 40 Q) C GO 20 CO o G :3
+AMVN o
o
0
I
I
20 40 Time (rain)
60
I
270
I
I
I
I
I
I
310 350 380 440 Wavelength (nm)
I
Ii -
-
500
FIG. 3. (A) Time course of cis-parinaric acid fluorescence decay induced by A M V N at 45° (excitation at 324 nm and emission at 413 nm). The reaction mixture consisted of 1.3 mM DOPC liposomes containing 6 i~M cis-parinaric acid and 300/zM AMVN (and 1.3 ~M a-tocopherol if needed) in 20 mM phosphate buffer (pH 7.4). (B) Fluorescence excitation and (C) emission spectra of cis-parinaric acid corresponding to each point of the time course recording. Excitation spectra were scanned with 413 nm emission wavelength, and emission spectra with 324 nm excitation wavelength.
manner (Fig. 3). The protective effect was further confirmed by measuring the generated lipid hydroperoxides by chemiluminescence HPLC (Fig. 4). A lipid hydroperoxides peak (retention time 8.9 min) from cis-parinaric acid peroxidation was progressively inhibited by increasing doses of atocopherol. A large negative peak (retention time 10.8 min) was identified as ot-tocopherol remaining in the sample. Thus, the decrease in the hydroperoxide peak was not due to direct quenching of the chemiluminescence HPLC reaction by residual free ot-tocopherol. The ratio of the initial decay rate of cis-parinaric acid fluorescence in the absence of an antioxidant to that in the presence of an antioxidant, V/VA, is related to the rate constant for peroxyl radical quenching by the antioxidant and to the concentration of the antioxidant by Eq. (2).24: 24y. Kono, M. Takahashi, and K. Asada, Arch. Biochem. Biophys. 174, 454 (1976).
[36]
ASSAY
ANTIOXIDANTS
WITH
¢i$-PARINARIC
4-300/./M AMVN I
,~ ~
ACID
379
+300//MAMVN . +0.3WM~-tocopneroI +300pMAMVN +?pM ~-tocopherol
~
-AMVN I
I
I
I
I
I
I
I
0
2
4
6
8
10
12
14
Time (min) FIG. 4. Chemiluminescencechromatograms of the samplesof Fig. 3 in a 60-rain reaction that contained 6 l.tM cis-parinaric acid, various concentrations of a-tocopherol as indicated, 300/~M AMVN, and 1.3 mM DOPC liposomes in 20 mM phosphatebuffer (pH 7.4). Detailed conditions were described in the Materials and Methods section.
V/V A = 1 + (kA/kpnA[PnA])[A]
(2)
where kA, kpnA, [PnA], and [A] are, respectively, the second-order rate constant for the reaction between the antioxidant and peroxyl radicals, the rate constant for the reaction between cis-parinaric acid and peroxyl radicals from AMVN, the concentration of cis-parinaric acid, and the concentration of the antioxidant. Thus, the slope of the line of V/VA against antioxidant concentration indicates the scavenging potency of the incorporated antioxidant for peroxyl radicals, with cis-parinaric acid as a standard. Of the three antioxidants tested, a-tocopherol was the most effective peroxyl radical scavenger (Fig. 5), which supports previous reports that a-tocopherol is a major and potent natural antioxidant in membranes. 2 From the slopes of the lines, the rate constant for quenching of peroxyl radicals by a-tocopherol was calculated to be 4.4 times larger than that for quenching by fl-carotene, and 4.8 times larger than that for quenching by ubiquinol 10.
380
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[36]
4.0
<
> >
3.0
2.0
1.0
, 0.0
, 1.0
, 2.0
3.0
Concentration (~M) FIG. 5. Relationship between V~VA and antioxidant concentration. V is the rate of initial fluorescence decay of cis-parinaric acid in the absence of antioxidants; VA, the rate of initial fluorescence decay of cis-parinaric acid in tile presence of the antioxidants indicated. For experimental conditions, see text. ([]) ~-Tocopherol; (A) fl-carotene; and (11) Q10 (ubiquinol 10, the reduced form of ubiquinone 10).
Interaction of a-Tocopherol with Other Antioxidants Because this system is simple and allows easy estimation of kinetics in the presence of membranous components, it was used to investigate the interaction of ct-tocopherol with other natural antioxidants, which is still a major question in in vivo antioxidant reactions. The addition of ascorbate or ubiquinol 10 to the a-tocopherol-incorporated liposomes increased the slope of V/VA, indicating that the apparent rate constant of ot-tocopherol for quenching peroxyl radicals became larger, whereas the addition of/3-carotene did not change the slope, indicating that/3-carotene did not alter the apparent rate constant of o~-tocopherol (Fig. 6). Discussion
Use of DOPC liposomes possessing cis-parinaric acid as a probe for oxidative stress, AMVN as a lipid-soluble radical initiator, and an antioxidant has provided a good model for the reaction of antioxidants with oxygen free radicals in biological membranes. This assay system, which enables continuous reaction monitoring with simple equipment, has potentially wide applications for investigation of fundamental antioxidant reactions.
[36]
ASSAY ANTIOXIDANTS WITH ciS-PARINARIC ACID A
C
B I
I
I
i
+1.3 uM I}-carotene
3.0-
381
T Jj/
Q1O
/J.
2.0
1.0~ 0 I0
'
'
015
1.0
I
0.0
0.5
I
I
1.0
0.0
0.S
I
1.0
1.5
c~-Tocopherol (HM) FIo. 6. Effect of (A) ascorbate, (B)/3-carotene, and (C) Ql0 (ubiquinol 10, the reduced form of ubiquinone 10) on the plot of V~VA against a-tocopherol concentration. Experimental conditions are described in the text.
The natural lipid-soluble compounds a-tocopherol, r-carotene, and ubiquinol 10 are expected to function differently as antioxidants in vivo, according to the differences in molecular structure (Fig. 7). Because it possesses a phenol group, ct-tocopherol scavenges free radicals by donating a hydrogen atom to them. 3 Ubiquinol 10 is expected to function in a similar manner. 25 The scavenging mechanism of/3-carotene is still unclear. 26 Besides, several fragmented products that may be harmful to membranes have been reported after the reaction of/3-carotene with free radicals. 27 Tsuchiya et al. 6 have reported that the conjugated double bond structure between rings is important for scavenging activity. The results presented here confirm that/3-carotene and ubiquinol, as well as a-tocopherol, are able to prevent oxidation of membranous components. However, the effects of r-carotene and ubiquinol are approximately 5 times weaker than those of a-tocopherol. It is surprising that a-tocopherol efficiently protects membranes against oxidation despite its low concentration in membranes. 28 There have been 25 V. Kagan, E. Serbinova, and L. Packer, Biochem. Biophys. Res. Commun. 169, 851 (1990). 26 N. I. Krinsky, Free Radical Biol. Med. 7, 617 (1989). 27 G. J. Handelman, F. J. G. M. van Kuijk, A. Chatterjee, and N. I. Krinsky, Free Radical Biol. Med. 10, 427 (1991). z8 j. j. Maguire, V. Kagan, B. A. C. Ackrell, E. Serbinova, and L. Packer, Arch. Biochem. Biophys. 292, 47 (1992).
382
[36]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
o~-Tocopherol CH 3 ]
H~
HO"
~
_ CH3
CH 3
CH3
.CH3
v
CH 3
I~-Carotene H3C
H3c\/cH3 I
/
CH 3
/
CH3
11
II
__ I
__ L
~.,,.,,,,';~CH 3
H3C
H3C
1
n3c cn3
Ubiquinol 10 OH
CH30~CH3.~
°.30- y
o.
Iv[
_ I'"
~.3j, °
FIG. 7. Molecular structures of a-tocopherul,/3-carotene, and ubiquinol 10.
several reports suggesting that some antioxidant such as ascorbate or ubiquinol can regenerate tocopherol that has been oxidized by reacting with free radicals and thereby maintain the apparent concentration of tocopherol (recycling tocopherol). 9A°'25'29 The potent protective effect of tocopherol as an antioxidant in biomembranes may be partly explained by such a recycling mechanism. The apparent increase in the rate constant of a-tocopherol by ascorbate or ubiquinol 10 provides evidence for atocopherol recycling. Ubiquinol 10 affected V/VA even at the absence of t~-tocopherol but ascorbate did not, which suggests that ubiquinol 10 scavenges radicals, as well as recycles a-tocopherol in membranes, but ascorbate only recycles a-tocopherol. On the other hand, E-carotene did not change the apparent rate constant of a-tocopherol, which suggests 29 j. j. M. van den Berg, F. A. Kuypers, B. Roelofsen, and J. A. F. Op den Kamp, Chem. Phys. Lipids 53, 309 (1990).
[36]
ASSAY ANTIOXIDANTS WITH ciS-PARINARIC ACID
383
the absence of interaction. These two antioxidants probably function independently in membranes. It can be supposed that the production of peroxyl radicals in the presence of excess cis-parinaric acid leads to the first-order decrease in fluorescence in the initial stage of the reaction, under the assumption that the chain reaction is negligible at this stage since interaction of each molecule is highly restricted by incorporation into the membranes. 3° Based on this assumption, the apparent first-order rate constant was obtained from the inverse half-time of fluorescence decay, and the second-order rate constant for the reaction, calculated from the dependence of the first-order rate constants on the concentration of cis-parinaric acid, was estimated to be 1.32 × 103 M -1 s e c -1. Using Eq. (2) to describe the competition of cis-parinaric acid and antioxidants for peroxyl radicals (Fig. 5), secondorder rate constants for the reaction of the antioxidants a-tocopherol,/3carotene, and ubiquinol I0 with peroxyl radicals were estimated to be 8.76 × 103, 1.97 × 103, and 1.84 × 103 M -1 sec -1, respectively. These values are smaller than those previously reported in the literature (e.g., ranging from 0.8 to 230 × 105 M -1 sec -~ for a-tocopherol), which were chemically determined in solution or suspension. 31 Niki et al. 32 reported a similar phenomenon, namely, that the rate constant of a-tocopherol in liposomal membranes was approximately 50 to I00 times less than that in homogeneous solution. They report that a possible reason for this difference is the restriction of antioxidant mobility resulting from the incorporation of tocopherol into membranes. 32 Using analogous compounds, we also found that mobility has a significant impact on the activity of antioxidants in membranes. 33Thus, the effect of membranes on antioxidant behavior appears to be another important factor for evaluating antioxidant activity, although the pure chemical reactivity of an antioxidant with radicals is certainly of great importance. Acknowledgments This work was supported by a grant from the National Institutes of Health (CA 47597) and by the Foundation for Total Health Promotion (1992). The authors thank Dr. Kozo Utsumi (Kochi Medical School) for valuable suggestions.
3o M. Takahashi and K. Asada, J. Biochem. (Tokyo) 91, 889 (1982). 3t E. Niki, Chem. Phys. Lipids 44, 227 (1987). 32 E. Niki, M. Takahashi, and E. Komuro, Chem. Lett. p. 1573 (1986). 33 E. Serbinova, V. Kagan, D. Han, and L. Packer, Free Radical Biol. Med. 10, 263 (1991).
384
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[37]
[37] S i n g l e t O x y g e n Q u e n c h i n g b y C a r o t e n o i d s By ALFRED
R.
SUNDQUIST, KARLIS BRIVIBA,
and
H E L M U T SIES
Introduction The quenching of singlet molecular oxygen [Oz(1Ag),abbreviated below as IO2] by carotenoids was first demonstrated by Foote and Denny ~using a chemical competition technique which measured the inhibition by/3carotene of the 102-dependent photooxygenation of 2-methyl-2-pentene. Since then considerable attention has been given to the antioxidant activities of carotenoids and oxycarotenoids (xanthophylls), 2,3 and a variety of additional chemical and physical techniques to evaluate 102 quenching have been described. Other chemical competition techniques determine the degree to which the test carotenoid inhibits the autosensitized 4'5 or methylene blue-sensitized 6 photooxidation of rubrene, or the chlorophyll-sensitized photooxidation of soybean oil. 7 Bleaching of the quencher has been used to measure the 10 2 quenching activity of the pigments crocin8'9 and bixin) ° Physical techniques are based on spectral properties of the excited species involved in the quenching process. For example, quenching can be measured as a decrease in the level or lifetime of 102 photoemission [reaction (I)] using 102 ~ 02 + hv (1270 nm)
(1)
sensitive infrared detectors. 11With some methods, 102 is produced continously either with a chemicaP 2 or photochemicaP 3 source, and the effect I C. S. Foote and R. W. Denny, J. Am. Chem. Soc. 9t), 6233 (1968). 2 N. I. Krinsky, Free Radical Biol. Med. 7, 617 (1989). 3 p. Di Mascio, M. E. Murphy, and H. Sies, Am. J. Nutr. 53, 194S (1991). 4 S. R. Fahrenholtz, F. H. Doleiden, A. M. Trozzolo, and A. A. Lamola, Photochem. Photobiol. 20, 505 (1974). s D. J. Carlsson, T. Suprunchuk, and D. M. Wiles, J. Polym. Sci. Part B U , 61 (1973). 6 M. M. Mathews-Roth, T. Wilson, E. Fujimori, and N. I. Krinsky, Photochem. Photobiol. 19, 217 (1974). 7 S.-H. Lee and D. B. Min, J. Agric. Food Chem. 38, 1630 (1990). 8 W. Bors, C. Michel, and M. Saran, Biochim. Biophys. Acta 796, 312 (1984). 9 p. Manitto, G. Speranza, D. Monti, and P. Gramatica, Tetrahedron Lett. 28, 4221 (1987). l0 G. Speranza, P. Manitto, and D. Monti, J. Photochem. Photobiol., B. 8, 51 (1990). II A. A. Gorman and M. A. J. Rodgers, J. Photochem. Photobiol,, B 14, 159 (1992). 12 p. Di Mascio, S. Kaiser, and H. Sies, Arch. Biochem. Biophys. 274, 532 (1989). 13 E. Oliveros, P. Murasecco-Suardi, A. M. Braun, and H.-J. Hansen, this series, Vol. 213, p. 420.
METHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
[37]
SINGLET OXYGEN QUENCHING BY CAROTENOIDS
385
of the carotenoid on the steady-state level of photoemission is determined. In another method, the carotenoid-dependent decrease in the lifetime of a pulse of photochemically generated 10 2 is measured by time-resolved luminescence spectroscopy. 14 Rate constants have been also determined by spectrophotometrically monitoring the triplet excited carotenoid (3C) formed during the quenching process [reaction (2)]. 15-17 1 0 2 q- C ~
(2)
0 2 q- 3C
In this chapter we describe in more detail the technique 12which makes use of a germanium photodiode to monitor 10 2 photoemission and a thermodissociable endoperoxide to generate a steady-state level of 10 2.18
Method
Reagents The endoperoxide of 3,3'-(l,4-naphthylidene) dipropionate (NDPO2) decomposes at moderate temperatures to yield the parent compound, NDP, and molecular oxygen, a portion of which (-50%) is in the singlet excited state [reaction (3)]. 19,20A convenient method to synthesize NDPOz in high yield is to incubate a solution of NDP 2°'21with HEOz in the presence
2No00C~ooc~ ~~ "~ 37°Ci ~aO0~ No NaOOC NDP02
'-'1"- 02 + 102 (3)
NDP
of sodium molybdate, 19,2° afterward precipitating the NDPO 2 with acid. Samples of carotenoids have been generously provided by Dr. J. Bausch (F. Hoffmann-La Roche, Basel, Switzerland). 14 p. F. Conn, W. Schalch, and T. G. Truscott, J. Photochem. Photobiol., B 11, 41 (1991). 15 A. Farmilo and F. Wilkinson, Photochem. Photobiol. 18, 447 (1973). 16 F. Wilkinson and W.-T. Ho, Spectrosc. Lett. 11, 455 (1978). i7 M. A. J. Rodgers and A. L. Bates, Photochem. Photobiol. 31, 533 (1980). 18 p. Di Mascio, A. R. Sundquist, T. P. A. Devasagayam, and H. Sies, this series, Vol. 213, p. 429. 19 j. M. Aubry, J. Am. Chem. Soc. 107, 5844 (1985). 2o p. Di Mascio and H. Sies, J. Am. Chem. Soc. 111, 2909 (1989). 21 R. Saint-Jean and P. Cannone, Bull. Soc. Chim. Ft., p. 3330 (1971).
386
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[37]
Assay Carotenoid solutions are freshly prepared in amber-colored vials with Nz-purged chloroform and kept on ice. Stock solutions of NDPO2 are made by neutralizing the free acid (see above) with NaOH in DzO18; the solution is stored at - 7 0 ° as small aliquots and thawed as needed. For reasons of solubility a mixed assay solvent [chloroform-ethanol, 1 : 1 (v/v)] is preferred. The 10 z quenching assays are conducted with a liquid nitrogen-cooled germanium diode (North Coast Scientific Co. Model EO-817L, Santa Rosa, CA) attached to a sample chamber.18 The assays are carried out in a cuvette (35 × 6 × 55 mm) placed in a thermostatted (37 °) holder. The signal is processed with a lock-in amplifier and monitored continuously with a chart recorder. After recording the baseline with the assay solvent, NDPO2 is added to the cuvette and the photoemission followed until a maximum (So) is reached; immediately thereafter the carotenoid is added, and the resulting level of photemission (S) is recorded. The assays are repeated over a range of carotenoid concentrations (e.g., 0.2-10 tzM).
Calculations Quenching rate constants are determined graphically by plotting the degree of quenching (i.e., the ratio So~S)versus the carotenoid concentration. The slope of this plot (Stern-Volmer) is equivalent to (kq + kr) ~', namely, the sum of the rate constants for physical quenching and chemical reaction multiplied by the lifetime of l o 2 (33 /zsec in the present assay solvent). The contribution of chemical reaction to the quenching of 10 2 by carotenoids is minor (<0.05% in the case of fl-carotene 22) and can be ignored. Singlet Oxygen Quenching Rate Constants The values obtained for a variety of biological carotenoids and oxycarotenoids (xanthophylls) using this technique are listed in Table I. For comparison, a range of values reported by other authors using different techniques is also presented. As is especially apparent in the cases of flcarotene and lutein, the rate constants that have been determined for a given carotenoid vary considerably. Although the rate of 10 2 quenching by carotenoids may be influenced by the assay solvent, this appears to account only partly for the disparate values; rather, systematic differences 22 A. A. Krasnovskii, Jr., and L. I. Paramoval
Biophysics 28, 769 (1983).
[37]
SINGLET OXYGEN QUENCHING BY CAROTENOIDS
387
TABLE I RATE CONSTANTS FOR QUENCHING OF SINGLET OXYGEN BY CAROTENOIDS, OXYCAROTENOIDS, AND SOME RELATED SYNTHETIC COMPOUNDS kq + kr (109 M -I sec-1) a
Compound Carotenoids Lycopene y-Carotene a-Carotene fl-Carotene Oxycarotenoids (xanthophylls) Astaxanthin Canthaxanthin Zeaxanthin Lutein Cryptoxanthin Synthetic polyenes C28-Polyene-tetrone C28-Polyene-tetrone diacetal C40-Epiisocapsorubin
Present method b
Literature values c
9 7 6 4
14-19 0.8 8 5-23
7 6 3 2 2
14 13-18 3-12 6-21
16 9 8
Because the rate constant of chemical reaction, kr, is relatively small, overall rate constants are equivalent to the physical quenching rate constant, kq.22 b Determined as described in the text with the endoperoxide of 3,3'-(1,4-naphthylidene) dipropionate as IO2 source and a germanium diode to measure IO2 photoemission. Values from Di Mascio et al. TM and Devasagayam e t al. 24 c For a listing of individual values with citations, see Di Mascio et al. 12
in the various techniques appear to be more significant. TM Nevertheless, it is clear from these studies that carotenoids very efficiently quench 102; furthermore, of the carotenoids and oxycarotenoids tested, lycopene exhibits the highest activity. 12'14 Since carotenoids can occur in biological systems as mixtures of the all-trans and cis isomers, it is also of interest to compare the IO2 quenching activity of the geometric isomers of a given carotenoid. Using the technique of time-resolved luminescence, Truscott and colleagues14 found that the quenching rate constants of the 9-cis and 15-cis isomers of fl-carotene were somewhat lower (10-40%) than that of all-trans-fl-carotene. In hu-
388
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[38]
man tissues and serum the cis isomers of B-carotene typically account for less than 5% of the total B-carotene content. 23 Rate constants for a number of synthetic polyenes were also determined using the technique described here, 24 and selected values are also presented in Table I. C28-Polyene-tetrone was found to quench ~O2 more rapidly than the carotenoids. Acknowledgments Our studies were supported by the National Foundation for Cancer Research, Bethesda, Maryland. A.R.S. is a Fellow of the Alexander yon Humboldt Foundation, B o n n - - B a d Godesberg, Germany. 23 W. Stahl, W. Schwarz, A. R. Sundquist, and H. Sies, Arch. Biochem. Biophys. 294, 173 (1992). 24 T. P. A. Devasagayam, T. Werner, H. Ippendorf, H.-D. Martin, and H. Sies, Photochern. Photobiol. 55, 511 (1992).
[38] S e p a r a t i o n o f G e o m e t r i c a l I s o m e r s o f fl-Carotene and Lycopene By W I L H E L M S T A H L a n d H E L M U T SIES
Introduction
Carotenoids are widespread in nature, occurring in bacteria, algae, fungi, higher plants, and in several animal species. More than 500 different structures have been identified. Because carotenoids are synthesized only in plants and lower organisms, those occurring in animals or humans are taken up from the diet. 1-5 Some carotenoids are precursors of vitamin A, playing an important role in the supply of retinoids for mammals. 6'7 Carotenoids can be divided into two major groups: hydrocarbon carotenoids and oxycarotenoids (xanthophylls), the latter containing at least one oxygen atom. The provitamin i O. Isler, ed., "Carotenoids." Birkhaeuser, Basel, 1971. 2 j. Gross, "Pigments in Vegetables." Van Nostrand-Reinhold, New York, 1991. 3 j. C. Bauernfeind, "Carotenoids as Colorants and Vitamin A Precursors." Academic Press, New York, 1981. 4 T. W. Goodwin, "The Biochemistry of the Carotenoids." Chapman & Hall, London, 1980. 5 H. Pfander, "Key to Carotenoids." Birkhaeuser, Basel, 1987. 6 j. A. Olson and M. R. Lakshman, this series, Vol. 189, p. 425. 7 L. E. Gerber and K. L. Simpson, this series, Vol. 189, p. 433.
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
388
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[38]
man tissues and serum the cis isomers of B-carotene typically account for less than 5% of the total B-carotene content. 23 Rate constants for a number of synthetic polyenes were also determined using the technique described here, 24 and selected values are also presented in Table I. C28-Polyene-tetrone was found to quench ~O2 more rapidly than the carotenoids. Acknowledgments Our studies were supported by the National Foundation for Cancer Research, Bethesda, Maryland. A.R.S. is a Fellow of the Alexander yon Humboldt Foundation, B o n n - - B a d Godesberg, Germany. 23 W. Stahl, W. Schwarz, A. R. Sundquist, and H. Sies, Arch. Biochem. Biophys. 294, 173 (1992). 24 T. P. A. Devasagayam, T. Werner, H. Ippendorf, H.-D. Martin, and H. Sies, Photochern. Photobiol. 55, 511 (1992).
[38] S e p a r a t i o n o f G e o m e t r i c a l I s o m e r s o f fl-Carotene and Lycopene By W I L H E L M S T A H L a n d H E L M U T SIES
Introduction
Carotenoids are widespread in nature, occurring in bacteria, algae, fungi, higher plants, and in several animal species. More than 500 different structures have been identified. Because carotenoids are synthesized only in plants and lower organisms, those occurring in animals or humans are taken up from the diet. 1-5 Some carotenoids are precursors of vitamin A, playing an important role in the supply of retinoids for mammals. 6'7 Carotenoids can be divided into two major groups: hydrocarbon carotenoids and oxycarotenoids (xanthophylls), the latter containing at least one oxygen atom. The provitamin i O. Isler, ed., "Carotenoids." Birkhaeuser, Basel, 1971. 2 j. Gross, "Pigments in Vegetables." Van Nostrand-Reinhold, New York, 1991. 3 j. C. Bauernfeind, "Carotenoids as Colorants and Vitamin A Precursors." Academic Press, New York, 1981. 4 T. W. Goodwin, "The Biochemistry of the Carotenoids." Chapman & Hall, London, 1980. 5 H. Pfander, "Key to Carotenoids." Birkhaeuser, Basel, 1987. 6 j. A. Olson and M. R. Lakshman, this series, Vol. 189, p. 425. 7 L. E. Gerber and K. L. Simpson, this series, Vol. 189, p. 433.
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[38]
SEPARATION OF B-CAROTENE AND LYCOPENE ISOMERS
389
FIG. 1. Structures of B-carotene and lycopene.
A, B-carotene, and the nonprovitamin A, lycopene, are the most abundant hydrocarbon carotenoids in the human organism and account for up to 50% of total carotenoids in human blood. 8-H The main sources are yellow, green, and red vegetables, notably carrots for B-carotene and tomatoes for lycopene, z
Geometrical
Isomers
B-Carotene and Lycopene B-Carotene and lycopene (Fig. 1), like other carotenoids, are composed of isoprene subunits arranged in a system of conjugated double bonds responsible for their color. The absorption maxima of carotenoids depend on the number of conjugated double bonds but is also influenced by solvent effects. Typical hmax values for B-carotene are 450 nm and 472 nm for lycopene (solvent hexane). The absorption in the visible range and the high extinction coefficients (e.g., 140 mM -1 cm -1 for B-carotene) permit selective and sensitive detection of a large number of important carotenoids with UV-Vis detection after separation by liquid chromatography, lZ 8 H. H. Schmitz, C. L. Poor, R. B. Wellman, and J. W. Erdman, J. Nutr. 121, 1613 (1991). 9 L. A. Kaplan, J. M. Lau, and E. A. Stein, Clin. Physiol. Biochem. 8, 1 (1990). ~0W. Stahl, W. Schwarz, A. R. Sundquist, and H. Sies, Arch. Biochem. Biophys. 294, 173 (1992). ll R. S. Parker, J. Nutr. 119, 101 (1989). 12 L. Zechmeister, "Cis-Trans Isomeric Carotenoids, Vitamin A and Arylpolyenes." Springer-Verlag, Wien, 1962.
390
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[38]
Geometrical Isomers of Carotenoids Owing to the presence of double bounds, carotenoids show the phenomenon of cis-trans isomer formation, also called stereomutation.12 The cis configuration [or Z (zusammen] is defined as the one with the highest priority groups on the same side and the trans configuration [E (entgegen)] as that with highest priority groups on opposite sides of the carbon-carbon bond. In early classic research papers the prefix neo together with a letter was used to assign carotenoid geometrical isomers of unknown structure. Referring to their chromatographic properties, the main fl-carotene Zisomers were named neo-fl-carotene U and B, later identified as 9-cis (neo-U) and 13- and/or 15-cis for neo-/3-carotene B. The number of possible stereoisomers of a carotenoid depends on the number of double bonds. However, of the numerous theoretically possible isomers only few are sterically unhindered and therefore are formed preferentially. Twenty of 272 theoretically possible forms of fl-carotene and only 72 of 1052 geometrical isomers of lycopene are likely to exist. 1,12 With some exceptions the all-trans isomers of these hydrocarbon carotenoids are predominantly found in carotenoid-producing organisms, accompanied by low amounts of the respective cis isomers. 2 all-trans-flCarotene and some mono-cis isomers are shown in Fig. 2.
Absorption Spectra of Cis Isomers The UV-Vis absorption spectra of all-trans-[3-carotene and different fl-carotene cis isomers are presented in Fig. 3. Compared to the all-trans spectrum, carotenoid cis isomers possess a characteristic subsidiary band in the near-UV region, referred to as the "cis-band." 12 It is located about 140 nm hypsochromic from the absorption maximum of the respective all-trans-carotenoid. The intensity of the cis-band depends on the position of the double bond and reaches a maximum when the double bond is situated in the center of the molecule (e.g., 15-cis isomer in the case of fl-carotene). 1,12-~5 The absorption maximum of the mono-cis isomers undergo a hypsochromic shift between 2 and 8 nm (poly-cis somewhat higher) concomitant with a decrease in the extinction coefficient. 16-18 13 G. Englert and M. Vecchi, Helu. Chim. Acta 63, 1711 (1980). 14 U. Hengartner, K. Bernhard, K. Meyer, G. Englert, and E. Glinz, Helo. Chim. Acta 75, 1848 (1992). t5 j. A. Haugan, G. Englert, E. Glinz, and S. Liaaen-Jensen, Acta Chem. Scand. 46, 389 (1992). t6 K. Tsukida, K. Saiki, T. Takii, and Y. Koyama, J. Chromatogr. 245, 359 (1982). t7 N. H. Jensen, A. B. Nielsen, and R. Wilbrandt, J. Am. Chem. Soc. 104, 6117 (1982). 18 L. Zechmeister, A. L. LeRosen, W. A. Schroeder, A. Polgar, and L. Pauling, J. Am. Chem. Soc. 65, 1940 (1943).
[38]
SEPARATION OF t-CAROTENE AND LYCOPENE ISOMERS
391
all-transfi-Carotene
15-cis13-Carotene
~'~
FIG. 2. Structures of all-trans and mono-cis isomers of/3-carotene.
Up to 30% lower extinction coefficients have been described for different B-carotene cis isomers compared to the all-trans form. This should be taken into consideration when quantifying cis isomers on the basis of alltrans standards) 6 The UV-Vis absorption characteristics are useful for identification of carotenoid stereoisomers, although other spectroscopic methods must be applied for unequivocal assignment.
392
ANTIOXIDANT CHARACTERIZATION AND ASSAY
C O
....................all-trans - 15-cis ........ 13-cis . . . . . . 9-cis
[38]
r~... i',~i~ i'~ ~'., ~ ~ . t
"
(~. L. O I/] rl
<
>
b,= O
[
I
I
300
400
500
Wavelength (nm) Fro. 3. Absorption spectra (UV-Vis) of aU-trans and mono-cis isomers of/3-carotene.
Induction of lsomerization B-Carotene and lycopene undergo stereoisomerization as soon as they are dissolved. The rate of stereoisomerization depends on the solvent, being rather slow at room temperature in solvents such as benzene or hexane, when the sample is protected from light; B-carotene is less sensitive to isomerization than the open-ring analog lycopene. Elevation of the temperature increases the rate of isomerization, which proceeds to a quasiequilibrium of geometrical isomers. In general, the cis isomers of a carotenoid are more sensitive to thermally induced isomerization than the all-trans forms. 19 The light-induced isomerization of carotenoids is most effective on irradiation with light of the wavelength corresponding to the main absorption bands. The most widespread method to induce stereoisomerization of carotenoids is irradiation with light in the presence of iodine as catalyst.'2 Even at room temperature, carotenoid isomers are rapidly formed, and a very similar pattern of isomers is obtained regardless of which isomer was used as a starting molecule. 19As in the case of most other carotenoids, the alltrans geometrical form of B-carotene and lycopene is the most stable and therefore is predominant. L~2,16
19 C. A. Pesek, J. J. Warthesen, and P. S. Taoukis, J. Agric. Food Chem. 38, 41 (19~)0).
[38]
SEPARATION OF fl-CAROTENE AND LYCOPENE ISOMERS
393
Separation of fl-Carotene and Lycopene Geometrical Isomers Many methods have been described for the analysis of carotenoids, 2°-22 but only a few concern separation of geometrical isomers, fl-Carotene and lycopene geometrical isomers have been analyzed in various systems using liquid chromatography. Analysis included isomer compositions obtained in model reactions and the identification and quantification of carotenoid geometrical isomers in biological samples such as food, blood, or tissues. In model reactions, mainly heat or light/iodine-induced formation of trans and cis isomers was used to obtain sample material for further identification of isomers by UV-Vis, nuclear magnetic resonance (NMR), infrared (IR), or mass spectroscopy. Standard mixtures of different fl-carotene geometrical isomers have been used in carotenoid analysis for the optimization of chromatographic systems, testing different column materials for their suitability to resolve carotenoid isomers. Model Systems
Carotenoid samples obtained in model systems are rather clean (free of interfering contaminants) and can often be analyzed without sample processing, thereby avoiding further artifacts. Care must be taken with sample handling, because light and temperature might influence the isomer pattern. The first separation of geometrical isomers of/3-carotene by liquid chromatography was described by Zechmeister and Polgar 23 who isolated 8 geometrical isomers from photoisomerized/3-carotene samples. Owing to limitations in spectroscopic techniques at the time, identification of the isolated compounds was not possible. The most complete resolution of /3-carotene isomers was obtained using inorganic stationary phases such as A1203 or Ca(OH) 2 . Vecchi et al. 24 separated 11 and isolated 8 cis isomers of/3-carotene obtained in a model reaction with an AI20 3 phase and hexane as eluent. This system is very sensitive to moisture and temperature, however, and sophisticated equipment is necessary to achieve reproducible results. Tsukida et al.~6 applied Ca(OH)2 stationary phases to investigate solutions of/3-carotene photoisomers and resolved 17 peaks, 10 of which were assigned as geometrical isomers. 16,25 2o N. I. Krinsky and S. Welankiwar, this series, Vol. 105, p. 155. 21 H. J. C. F. Nelis and A. P. De Leenheer, Anal. Chem. 55, 270 (1983). 2z L. R. Cantilena and D. W. Nierenberg, J. Micronutr. Anal. 6, 127 (1989). 23 L. Zechmeister and A. Polgar, J. Am. Chem. Soc. 64, 1856 (1943). 24 M. Vecchi, G. Englert, R. Maurer, and V. Meduna, Heir. Chim. Acta 64, 2746 (1981). 25 y . Koyama, M. Hosomi, A. Miyata, H. Hashimoto, S. A. Reames, K. Nagayama, T. Kato-Jippo, and T. Shimamura, J. Chromatogr. 439, 417 (1988).
394
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[38]
Several methods using reversed-phase high-performance liquid chromatography (HPLC) techniques have been applied to the analysis of carotenoid isomers in model systems. Different HPLC methods with reversedphase column materials have been compared to Ca(OH) 2 for effectiveness in resolving all-trans-fl-carotene from cis analogs. The Vydac C~8 201 TP (Vydac, Hesperia, CA) column was the most effective reversed-phase column, resolving all-trans-, 9-cis-, and 13-cis-fl-carotene. The Ca(OH)2 material, however, was described as superior in selectivity for carotenoid isomers. 26 We have introduced Suplex pKb 100 columns for HPLC analysis of lycopene and fl-carotene isomers in serum and tissues, 27 achieving good resolution of fl-carotene mono-cis isomers. CH3CN is necessary as a component of the eluent to separate fl-carotene from a-carotene and lycopene in biological samples. It can be omitted for the resolution ofphotoisomerized fl-carotene standard mixtures (see Fig. 4). Cyclodextrin stationary phases with nonpolar eluents have been tested, and separation of all-trans- from 15-cis-fl-carotene was obtained. This column produces an elution sequence of carotenoids very similar to that found with silica gel materials. 28 A highly efficient method for the separation of lycopene isomers obtained after thermal stereoisomerization of all-trans-lycopene was described by Hengartner et al. 14 With Nucleosil 300-5 (Macherey-Nagel, Diiren, Germany) as stationary phase and hexane/N-ethyldiisopropylamine (2000 : 1, v/v) as mobile phase, several mono- and poly-cis isomers of lycopene were separated. A few were identified by spiking experiments with synthetic reference compounds or after semipreparative isolation of lycopene isomers identified by spectroscopic methods including NMR.14
Biological Samples In contrast to model isomer mixtures, studying biological materials requires sample processing to obtain carotenoids. This mainly includes a homogenization step (food or tissues) and an extraction step into organic nonpolar solvents. During these procedures the samples might be exposed to light or temperatures which might induce the formation of geometrical isomers, potentially shifting the original isomeric pattern. Dim light and
26 C. A. O'Neil, S. J. Schwartz, and G. L. Catignani, J. Assoc. Off. Anal. Chem. 74, 36 (1991). 27 W. Stahl, A. R. Sundquist, M. Hanusch, W. Schwarz, and H. Sies, Clin. Chem. (WinstonSalem, N.C.) 39, 810 (1993). 28 A. M. Stalcup, H. L. Jin, D. W. Armstrong, P. Mazur, F. Derguini, and K. Nakanishi, J. Chromatogr. 499, 627 (1990).
A
w C i
? O}
B-Carotene
j
.~
Isomer Mixture
B
+ 9-cis
C
+ 13-cis
D
+ 15-cis . . . . . . . . .
b
~
.....
J
......
10 time (min)
FIG. 4. Chromatogram of a mixture of t~-¢arotcne isomers obtained by iodine-induced isomcfization. HPLC conditions: column, Suple× pKb-100 (Supelco, Bellefont¢, PA); detection, 460 nm; flow rate, ] mVmin; eluent, methanoVdichioromethane (%:4, v/v).
396
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[38]
low temperatures are helpful in preventing artifactual formation of cis isomers. Details of extraction procedures or special modifications used to isolate carotenoids from biological samples are beyond the scope of this chapter. It should be mentioned, however, that saponification or enzymatic digestion procedures are applied to improve extraction efficiency or resolution on HPLC. Like other standard extraction procedures, such steps pose an additional risk for isomerization and might shift the isomeric pattern in the sample as well. Therefore, control experiments appropriate for the procedure are recommended to exclude the possibility of isomer formation during sample processing. Useful controls are samples spiked with known amounts of a pure isomer, processed as the unknowns.
Fruits and Vegetables Inorganic liquid chromatography column material based on Ca(OH)2 proved to be effective in separating carotenoid isomers in fruits and vegetables. It was found that food processing like cooking but especially canning increased the amount of cis isomers in food.29-31 Unequivocal identification of the geometrical isomers, however, was not described. They were assigned according to their behavior on elution from the column and from UV-Vis spectra. The main problem with lime-based or similar columns is that only few commercially available lime products generate efficient separations. The analysis requires meticulous control of conditions, and prepacked columns to obtain reproducible results with defined materials are not available. Therefore, reversed-phase HPLC was used in analysis of carotenoid geometrical isomers. Quackenbush introduced the Vydac 201 TP column for the analysis of carotenoid geometrical isomers and separated all-trans-, 9-cis-, and 13-cis-[3-carotene in iodine isomerization mixtures but also in kale and carrot samples. 32 At least 6 lycopene geometrical isomers as obtained by iodine-induced isomerization were separated. Khachik et al. 33 found the Vydac 201 TP material to be superior, especially when analyzing hydrocarbon carotenoid geometrical isomers. On the other hand, spherical particle material (with small pore sizes) was more suitable for the separation of oxycarotenoid (xanthophyll) geometrical isomers (Microsorb CI8 5/~m/100 ,~).33 29 j. p. Sweeney and A. C. Marsh, J. Assoc. Off. Anal. Chem. 53, 937 (1970). 3o T. Panalaks and T. K. Murray, J. Inst. Can. Technol. Aliment. 3, 145 (1970). 3t L. A. Chandler and S. J. Schwartz, J. Food Sci. 52, 669 (1987). 32 F. W. Quackenbush, J. Liq. Chromatogr. 10, 643 (1987). 33 F. Khachik, G. R. Beecher, and W. R. Lusby, J. Agric. Food Chem. 37, 1465 (1989).
[38]
SEPARATION OF fl-CAROTENE AND LYCOPENE ISOMERS
397
Craft and co-workers tested five commercially available C18 column materials. 34 In this study, a mixture of/3-carotene isomers was separated by the wide-pore polymeric C18 columns but not by the monomeric C18 columns. Based on the size of the rigid fl-carotene molecule (33 A), the narrow pore size of the monomeric column material was suggested to limit accessibility of the molecule to the bonded phase inside the pores. Other authors that reported the separation of trans-fl-carotene from cis isomers by reversed-phase liquid chromatography achieved the separations with wide-pore C~8 modified silica synthesized from trichlorosilanes (polymeric synthesis) rather than silica modified with monochlorosilanes (monomeric synthesis). Further, it should be mentioned that supercritical fluid chromatography (SFC) was applied to separate/3-carotene geometrical isomers in food materials. 35 The influence of stationary phases, temperature, and organic modifiers on the separation w a s t e s t e d . 36,37 Although SFC is a promising tool, it appears that in routine analysis normal HPLC will dominate due to more simple equipment.
Blood Several methods have been described for the analysis of human blood or blood components for carotenoid composition. Few, however, deal with information on the presence of carotenoid cis isomers, in particular those of fl-carotene or lycopene. Reversed-phase HPLC techniques were used to separate all-trans-fl-carotene from its cis isomers, but peak identification always appeared to be difficult. 38-4° Even with two Vydac 201 TP 54 columns in series, /3-carotene cis isomers could not be resolved reproducibly. 41 A step gradient technique with reversed-phase columns improved the separation of oxycarotenoid cis isomers; however, identifi-
34 N. E. Craft, L. C. Sander, and H. F. Pierson, J. Micronutr. Anal. 8, 209 (1990). 35 H. H. Schmitz, W. E. Artz, C. L. Poor, J. M. Dietz, and J. W. Erdman, J. Chromatogr. 479, 261 (1989). 36 E. Lesellier, A. Tchapla, M. R. Pechard, C. R. Lee, and A. M. Krstulovic, J. Chromatogr. 557, 59 (1991). 37 M. C. Aubert, C. R. Lee, A. M. Krstulovic, E. Lesellier, M. R. Pechard, and A. Tchapla, J. Chromatogr. 557, 47 (1991). 38 C. D. Jensen, T. W. Howes, G. A. Spiller, T. S. Pattison, J. H. Whittam, and J. Scala, Nutr. Rep. Int. 35, 413 (1987). 39 A. L. Sowell, D. L. Huff, E. W. Gunter, and W. J. Driskel, J. Chromatogr. 431,424 (1988). 40 j. G. Bieri, E. D. Brown, and J. C. Smith, J. Liq. Chromatogr. 8, 473 (1985). 41 W. G. Rushin, G. L. Catignani, and S. J. Schwartz, Clin. Chem. (Winston-Salem, N.C.) 36, 1986 (1990).
398
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[38]
0 C
eu 0 tO 0 ¢: o
o a
¢3
.m
o
'
I u~
ooi *15-cis-~-Cor0tene
._1
* 9-cis-~-Car0tene
unspiked 110 Time (min)
2()
FIG. 5. Chromatogram of a human plasma sample. HPLC conditions: see Fig. 4; eluent, methanol/acetonitrile/2-propanol (54 : 44 : 2, v/v).
cation of lycopene or B-carotene cis isomers was not achieved. 42,43 For the more polar carotenoids nitrile-bounded column material proved to be superior for the resolution of geometrical isomers. 44 We have shown that the main cis isomer in human plasma is either 13or 15-cis-B-carotene, and that lycopene consists of at least four geometrical isomers in human plasma, using a RP- 18 end-capped material with isocratic elution.l° With an improved method using another reversed-phase column material (Suplex pKb-100), we were able to separate further the geometrical isomers of B-carotene and lycopene (see Fig. 5). 27 As indicated in Fig. 5, the mono-cis isomers of B-carotene are clearly resolved and identified by spiking experiments. The main B-carotene cis isomer in human serum is 13-cis, contributing about 5% to total B-carotene. At least six lycopene isomers are detectable but have not yet been assigned. 42 N. I. Krinsky, M. D. Russett, G. J. Handelman, and D. M. Snodderly, J. Nutr. 1211, 1654 (1990). 43 G. J. Handelman, B. Shen, and N. I. Krinsky, this series, Vol. 213, p. 336. 44 F. Khachik, G. R. Beecher, M. B. Goli, W. R. Lusby, and J. C. Smith, Anal. Chem. 64, 2111 (1992).
[38]
SEPARATION OF B-CAROTENE AND LYCOPENE ISOMERS
399
== A
o
t~
c
o
.-
~
.m._m
+15-cis-13-Corotene
.9-cis-[3-Cor0tene
unspiked
6
i'o Time
26
(rain) tn e-
'-.c_. o
~
~x
0
E
j
e"
o
-~ ,o
I-
I
0
20
y,=
t
40
Time (rain) FIG. 6. HPLC chromatograms of human tissue samples (testis). HPLC conditions: see Fig. 4; (A) eluent, methanol/acetonitrile/2-propanol (54:44:2, v/v); (B) eluent, methanol/ acetonitrile/2-propanol/water (10 : 40 : 40 : 10, v/v).
400
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[38]
Tissue
With the same methodology, the composition of B-carotene and lycopene isomers in human tissues was investigated. To avoid changes during sample processing, special care is taken as follows. The tissue samples are powdered under liquid nitrogen before homogenization using mortar and pestle to shorten homogenization times. The HPLC techniques were sufficient (especially with the Suplex pKb-100 column) to separate alltrans from the mono-cis isomers of B-carotene and to demonstrate that all these compounds are detectable in human tissue (see Fig. 6A). Human tissues clearly differ from blood with respect to the isomeric pattern of B-carotene. In contrast, only small differences in the lycopene isomer patterns in tissues and blood were detected. Further separation of the geometrical isomers of lycopene can be achieved using a different mobile phase (see Fig. 6B). This, however, doubles retention times and increases the time for an HPLC run up to 40 min. Analyses of tissue material for the geometrical isomers of/3-carotene have been published by Ben Amotz et al. 45'46 They used Vydac 201 TP 54 material and analyzed the alltrans- and 9-cis-B-carotene composition of rat and chicken liver after administration of a mixture of these geometrical isomers in the diet. 45'46 Concluding Remarks Progress has been made in determining carotenoid geometrical isomers. For chemically obtained samples (as opposed to biological material), for example, mixtures of isomers obtained by isomerization of standard carotenoids, lime columns still give the best results. Their use, however, is limited because of technical problems regarding reproducibility. More easily handled C18 reversed-phase HPLC materials have been introduced for the analysis of biological material such as food and blood or tissue samples; these HPLC methods are capable of resolving the all-trans and mono-cis geometrical isomers of B-carotene and lycopene. More sophisticated techniques will improve separations of carotenoid geometrical isomers in biological samples, such as employing new column material or liquid chromatography techniques like supercritical fluid chromatography (SFC). Acknowledgments This work was supported by the National Foundation for Cancer Research, Bethesda, Maryland; the Ministerium fiir Wissenschaft und Forschung, Nordrhein-Westfalen; and the Bundesministerium ftir Forschung and Technologie (Bonn). 45 A. Ben Amotz, S. Mokady, S. Edelstein, and M. Avron, J. Nutr. 119, 1013 (1989). 46 S. Mokady, M. Avron, and A. Ben Amotz, J. Nutr. 120, 889 (1990).
[39]
ANTIOXIDANT ACTIVITY OF VITAMIN A
401
[39] L , i p o p e r o x y l R a d i c a l - S c a v e n g i n g A c t i v i t y o f V i t a m i n A a n d A n a l o g s in H o m o g e n e o u s S o l u t i o n By MARIA A. LIVREA and LUISA TESORIERE
Introduction The ability of vitamin A and its analogs to act as antioxidants has been proposed by a number of investigators. This finding has been derived both from in vitro inhibition studies of lipid peroxidation in subcellular systems ~-5 and from in vivo studies 6 in which animals were treated with very high doses of vitamins. Our observations with an in vivo rat model 7 have shown that accumulation of vitamin A in cell membranes in the nanomolar range makes the membranes resistant to peroxidative stress both induced in vivo by acute administration of doxorubicin and produced in vitro by iron/ascorbate. This suggests that vitamin A may be considered a physiological antioxidant. Investigations carried out with a number of pure chemical systems have shown that retinoids possess peroxyl radical-scavenging activity, 8-H thereby confirming that these compounds could behave as effective inhibitots of lipid peroxidation by acting as chain-breaking antioxidants. These studies also show that both the assay system and the types of peroxyl radicals generated are important in determining the reactivity of retinoids. A method is reported in this chapter for the assay of the reactivity of alltrans-retinol, as well as other natural and synthetic retinoids, with radicals C. Nicotra, M. A. Livrea, and A. Bongiorno, IRCS Med. Sci. 3, 141 (1975). 2 0 . Halevy and D. Sklan, Biochim. Biophys. Acta 918, 304 (1987). 3 G. F. Vile and C. C. Winterbourn, FEBS Lett. 238, 353 (1988). 4 N. P. Das, J. Neurochem. 52, 585 (1989). 5 M. A. Livrea, M. Valenza, A. Bongiorno, M. Ciaccio, and L. Tesoriere, Free Radical Res. Commun. 16, Suppl. 1, 12.28 (1992). 6 V. N. R. Kartha and S. Krishnamurthy, J. Vitam. Nutr. Res. 47, 394 (1977). 7 M. Ciaccio, M. Valenza, L. Tesoriere, A. Bongiorno, R. Albiero, and M. A. Livrea, Arch. Biochem. Biophys. 301, 103 (1993). 8 M. Tsuchiya, G. Scita, D. F. T. Thompson, V. E. Kagan, M. A. Livrea, and L. Packer, in "Retinoids: Progress in Research and Clinical Applications" (M. A. Livrea and L. Packer, eds.), p. 525. Dekker, New York, 1993. 9 M. Hiramatsu and L. Packer, this series, Vol. 190, p. 273. t0 M. Tsuchiya, G. Scita, H.-J. Freisleben, V. E. Kagan, and L. Packer, this series, Vol. 213, p. 460. II H.-J. Freisleben and L. Packer, Biochem. Soc. Trans. 21, 325 (1993).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
402
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[39]
generated by reaction of the lipid-soluble azo initiator 2,2'-azobis(2,4dimethylvaleronitrile) (AMVN) with methyl linoleate. General Principles for Measuring Antioxidant Activity The oxidation of linoleic acid methyl ester (LAME) is the simplest model for studying the oxidation of polyunsaturated lipids, and it has been widely adopted to evaluate antioxidant activity.12-~6 Because linoleic acid has two double bonds, peroxidation occurs at the bisallylic hydrogens and generates conjugated diene hydroperoxides quantitatively. ~7 Therefore, lipid oxidation can be followed quantitatively by UV measurements of the conjugated dienes formed. To obtain a constant rate of chain initiation, which is essential to kinetic studies, azo initiators are commonly used. Under these conditions, the peroxidation reaction proceeds via the mechanism exemplified in Eqs. (1)-(5): Initiation R- + O2--~ ROO" R O O . + L H ~ ROOH + L .
(1) (2)
Propagation
L.
+
0 2~
LOO"
(3)
kp
LOO. + LH
~ LOOH + L-
(4)
~ nonradical products
(5)
Termination
2LOO"
kt
where R. is the azo initiator radical, LH represents linoleic acid, L., LOO-, and LOOH are, respectively, the alkyl, alkylperoxyl radical, and hydroperoxide generated, and kp and kt are the rate constants for propagation and termination of the radical chain. Because the rate of initiation is 12 E. Niki, T. Saito, A. Kawakami, and Y. Kamiya, J. Biol. Chem. 259, 4177 (1984). 13j. M. Braughler and J. Pregenzer, Free Radical Biol. Med. 7, 125 (1989). 14 y. Yamamoto, E. Komuro, and E. Niki, J. Nutr. Sci. Vitaminol. 36, 505 0990). 15 N. Noguchi, Y. Yoshida, H. Kaneda, Y. Yamamoto, and E. Niki, Biochem. Pharmacol. 44, 39 (1992). 16 N. Gotoh, K. Shimizu, E. Komuro, J. Tsuchiya, N. Noguchi, and E. Niki, Biochim. Biophys. Acta 1128, 147 (1992). 17y . Yamamoto, E. Niki, Y. Kamiya, and H. Shimasaki, Biochim. Biophys. Acta 795, 332 (1984).
[39]
ANTIOXIDANT ACTIVITY OF VITAMIN A
403
constant, the steady-state treatment can be applied, and the rate of oxidation (RD) is given by the rate of production of hydroperoxides. When a conventional chain-breaking antioxidant (IH) is added to the peroxidation system, LOOH formation is inhibited as a result of the scavenging of chain-carrying peroxyl radicals, and the antioxidant is degraded. An inhibition period (tinh), whose duration depends on the concentration of the antioxidant, is observed, after which oxidation proceeds at a rate similar to that measured in the absence of the antioxidant. The rate of disappearance of the chain-breaking antioxidant closely approximates the initiation rate. The latter can be calculated from the lag period (/inh) produced by a known amount of antioxidant, according to Eq. (6)18: R i = n[IH]/tin h
(6)
where n is the stoichiometric factor for the antioxidant, that is, the number of LOO- trapped by each molecule. The rate of chain initiation is usually calculated from the inhibition period produced by o~-tocopherol, whose stoichiometric factor is 2.19 The rate of oxidation during the inhibition period (Rinh), in the presence of a chain-breaking antioxidant, is expressed by Eq. (7)20: Rin h = k p [ L H ] R i / n k i n h [ I H ]
(7)
where kin h is the rate constant that expresses how rapidly the antioxidant will react with peroxyl radicals. Antioxidant parameters can be quantitatively obtained from the lipid model system if substrate and reaction conditions are carefully controlled. Whether or not contaminating oxidation products of methyl linoleate are present should be determined so that conjugated diene hydroperoxides produced after methyl linoleate oxidation can be accurately quantified. When necessary, methyl linoleate must be purified by column chromatography on Florisil (Floridin, New York, NY) as reported. 21 Similarly, the purity of retinoids must be ascertained. Methods Assays measuring peroxidation of methyl linoleate are performed under dim red light to avoid interfering with the peroxidation process by 18 C. E. Boozer, G. S. Hammond, C. E. Hamilton, and J. N. Sen, J. Am. Chem. Soc. 77, 3233 (1955). 19j. R. C. Barklay and K. U. Ingold, J. Am. Chem. Soc. 103, 6478 (1981). 2o G. W. Burton and K. U. Ingold, J. Am. Chem. Soc. 103, 6472 (1981). 2x j. Terao, A. Nagao, D. Park, and B. P. Lim, this series, Vol. 213, p. 454.
404
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[39]
light e x p o s u r e 22'23and to prevent photoreactions of retinoids. Linoleic acid methyl ester (315 mM; Sigma, St. Louis, MO) and micromolar amounts of retinoids (variable from 2 to 16/zM, depending on the retinoid utilized, in a 1.0 ml final volume of methanol) are allowed to equilibrate in a water bath at 37° for 60 sec, under air. Appropriate amounts of azo initiator in methanol are then added, and the incubation is started. Portions of the mixture (10 ~1) are taken at 5-rain intervals and injected onto a reversedphase high-performance liquid chromatography (HPLC) column (LC-18, particle size 5/xm; 1.6 m m × 25 cm; Supelco, Tokyo, Japan), equilibrated and then eluted with methanol at a flow rate of 1.0 ml/min. In such a system, linoleic acid hydroperoxides, as revealed from UV absorption at 234 nm, have a retention time of 4.0 rain, and unreacted linoleic acid methyl ester has a retention time of 6.5 rain. Quantitation of linoleic acid hydroperoxides is by reference to a standard curve made with known amounts of linoleic acid methyl ester hydroperoxide (Sigma). Alternatively, fractions containing lipid conjugated diene hydroperoxides are collected, dried under nitrogen, and redissolved in several milliliters of cyclohexane. Absorption spectra are then recorded from 200 to 300 nm, and spectrophotometric quantitation is done at 234 nm, using an absorbance coefficient of 26,000 M - 1 cm- 1.24
Kinetic Parameters for Reaction of all-trans-Retinol with Linoleic Acid-Derived Peroxyl Radicals Table I summarizes the effects of all-trans-retinol (Sigma) on the oxidation of methyl linoleate in methanol, initiated with AMVN at 37° under air. all-trans-Retinol is shown to be an effective chain-breaking antioxidant. Under all conditions utilized, Rinh and tinh can be accurately measured, and the inhibition rate constant can be calculated from Eq. (8): kinh =
kp[LH]/Ri,hti,h
(8)
According to the experimental conditions, the calculated kinh value varies from 3.3 × l05 to 3.7 × l05 M - l s e c -1 when the absolute rate constant for the oxidation of methyl linoleate at 37 ° is assumed to be 100 M-1 sec-i.x2 The stoichiometric factor for all-trans-retinol, calculated from Eq. (6), varies from 0.15 to 0.36. 22 V. E. Kagan, A. A. Shvedova, K. N. Novikov, and Yu.P. Kozlov, Biochim. Biophys. Acta 330, 76 (1973). 23 T. Hiramitsu and D. Armstrong, Ophthalmol. Res. 23, 196 (1991). 24 R. O. Recknagel and E. A. Glende, this series, Vol. 105, p. 331.
[39]
A N T I O X I D A N T ACTIVITY OF V I T A M I N A
--.,
Z I-
_
~-g
0 Z <
z
.o 0 o 0
.1 < [-.
::L
<
o z
~ ~:~
0 Z
,',.~
o
o
g
-:~
-~
0 Z
_o iN
Z~
o
"~ o~ Z
2~
<~ ~r..)
405
406
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[39]
Figure 1 shows the elution pattern of radioactivity, as resolved by reversed-phase HPLC on a C18 column as above, when 10% aliquots of incubation mixtures (containing 315 mM LAME, 1.5 mM AMVN, and 16 I~M all-trans-[aH]retinol, 1.0 mCi/mmol, in a total volume of 1.0 ml) are assayed at various incubation time intervals. It appears that about 100% of the radioactivity associated with the peak of all-trans-retinol disappears within 20 min, in accordance with the lag time observed with that concentration of antioxidant (Table I). A number of radioactive metabolites, which may represent oxidation products arising from all-trans-retinol, are recovered and increase in concentration with incubation time. Although it is evident that retinoids are able to scavenge chain-propagating lipoperoxyl radicals, the mechanism of their antioxidant activity is not yet entirely known. Samokyszyn and Marnett 25'26presented evidence that 13-cis-retinoic acid reacts preferentially with peroxyl radicals by addition but not by hydrogen atom abstraction reactions. Analogous to addition reactions of antioxidants with conjugated polyunsaturation, 27'28 peroxyl radical addition to retinoids generates a resonance-stabilized carbon-centered radical ( R - - O O L .)29 that may then react either with another lipoperoxyl radical to give rise to a nonradical product [Eq. (9)] or with O2, reversibly, to generate a retinoid-derived peroxyl radical [Eq. (10)]: kinh
R - - O O L . + LOO. > nonradical products R - - O O L . + O2 ~- O O L - - R - - O O .
(9) (10)
Because 5,6-epoxide, as well as other unidentified oxidation products, are released during peroxidative degradation of retinoids,25'3° a third pathway [Eq. (1 1)] must be taken into account for the carbon-centered radical: R--OOL.---> retinoid 5,6-epoxide + LO-
(11)
This can also be considered an autoxidative pathway for retinoid oxidation, because it does not lead to a net radical consumption. The antioxidant activity of retinoids is seen as the net effect of all these competing reactions. 25 V. M. Samokyszyn and L. J. Marnett, J. Biol. Chem. 262, 14119 (1987). 26 V. M. Samokyszyn and L. J. Marnett, this series, Vol. 190, p. 282. 27 G. W. Burton and K. U. Ingold, Science 224, 569 (1984). R. Stocker, Y. Yamamoto, A. F. McDonagh, A. N. Glazer, and B. N. Ames, Science 235, 1043 (1987) V. M. Samokyszyn and L. J. Marnett, Free Radical Biol. Med. 8, 491 (1990). 30 R. Yamauchi, N. Miyake, K. Kato, and Y. Ueno, Biosci. Biotechnol. Biochem. 56, 1529 (1992).
[39]
ANTIOXIDANT
ACTIVITY OF VITAMIN
25"
.......
407
A
-100
I I. . . . . . . . . . .
-"J
I I I
#1 II II I
12.5"
Z
-50
AB
I1','
30
O m
4b
s'o
"T 0
60
.,A g
15-
,..........
100
O Z
u._
-r
? o p
7.5-
i¢ //i
fl
,
1
,50
I" IM
X
~E
In
el_ u
0 15-
_
I Jl 30
40
.. o
50 ,- . . . . . . . . . .
60
Z Ill U
100
C
ZS-
0
30 FRACTION
,50
40
5'0
0 60
NUMBER
FIG. 1. Reversed-phase HPLC separation of all-trans-[3H]retinol (solid bar) and metabolites (open bars) generated during the reaction of 16 i x M all-trans-[3H]retinol with linoleic acid-derived peroxyl radicals. The elution system consisted of a linear methanol-water gradient from 30 to 90% methanol over 30 min, followed by isocratic elution with 90% methanol for 20 min, then with 100% methanol for additional 10 min, at a flow rate of 1.0 ml/min. Fractions of 1.0 mi are collected and the radioactivity measured. (a) Zero time; (b) 10 min incubation; (c) 20 min incubation.
408
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[39]
Partitioning of the retinoid carbon-centered intermediate between competing antioxidant and autoxidative pathways may also govern the prooxidant and antioxidant activity of all-trans-retinol, since the shift of the equilibrium of reaction (10) in favor of the formation of retinoid-derived peroxyl radical will actually increase the steady-state concentrations of peroxyl radicals. Under the experimental conditions described, autoxidation reactions of all-trans-retinol with Oz do not seem to occur nor to affect the antioxidant efficacy of all-trans-retinol. Given the autocatalytic nature of the peroxidative chain reactions that would be set in motion by the retinoid-derived peroxyl radical, compared to the 1 : 1 stoichiometry of the inhibition reaction [Eq. (9)], the occurrence and linear dependence of lag times on antioxidant concentrations (Table I) could not be observed. On the other hand, since the measured stoichiometric factor is less than 1, it appears evident that even when lag times are linearly dependent on the retinol concentration, some retinol is consumed without a net radical consumption, possibly via epoxide formation. Kinetics of Peroxyl Radical-Scavenging Activity of Natural and Synthetic Retinoids Oxidation of linoleic acid methyl ester in methanol may be employed to measure the lipoperoxyl radical-scavenging activity of vitamin A analogs. Table II shows the inhibition kinetic parameters obtained with some natural and synthetic retinoids. Retinoids are chosen and assayed on the basis of the solubility in methanol, all-trans-Retinyl palmitate and all-trans-
TABLE II INHIBITION KINETIC PARAMETERSOF OXIDATION OF LINOLEIC ACID METHYL ESTER IN METHANOL BY VITAMIN A ANALOGS
Inhibitor
all-trans-Retinyl palmitate all-trans-Retinoic acid 13-cis-Retinoic acid Acitretin Etretinate Temarotene Ro 15-1570
Concentration (/~M)
LAME (raM)
AMVN (mM)
10 2 2 10 10 10 10
315 315 315 315 315 315 315
3.0 3.0 3.0 3.0 3.0 3.0 3.0
tinh (sec)
780 1320
108 Rinh (M sec -I)
10 -5 kinh (M-Isec -I) n
No inhibition* 5.5 6.3 0.7 28.8 No inhibition No inhibition No inhibition No inhibition
2.7 4.0
* Since the Sigma compound contains 1% butylated hydroxytoluene (BHT), reference assays must be performed with the same amount of BHT. When corrected for the effect of BHT itself, no antioxidant activity of all-trans retinyl palmitate was evident.
[39]
A N T I O X I D A N T ACTIVITY OF VITAMIN A
409
"13
¢5 e-
0.)
:.=
).,
e'~
,.d "e
z <
~'~ ra
0
~
rz
"7"7"7-6-6-6-6
0o~'~ e~
o
O
o
.~" ,.~
z <
~
~.~ ,. ,-
"Ua
u~
CheL
.x2_ 0 e~ ",~
._.9.
2.o-6 -ta
410
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[40]
retinoic acid are from Sigma, whereas synthetic analogs of retinoic acid have been kindly supplied by Hoffman-La Roche (Basel, Switzerland). Under the assay conditions employed, 13-cis-retinoic acid is the most effective retinoid, followed by all-trans-retinoic acid. Other analogs of retinoic acid are ineffective in inhibiting oxidation of methyl linoleate. Because as reported above, the antioxidant efficiency of retinoids may be due to the effective steady-state concentration of the retinoid-derived carbon-centered radical, the higher antioxidant efficiency displayed by 13-cis-retinoic acid as compared to other retinoids may be the result of a relatively lower susceptibility of the carbon-centered intermediate from 13-cis-retinoic acid to autoxidative reactions. As has been pointed out, the antioxidant efficiencyof retinoids depends largely on the phase in which they scavenge radicals. The retinoids assayed here reveal very different potencies when assayed in other artificial systems, such as dioleoyl phosphatidylcholine (DOPC) liposomes. Table III summarizes the known stoichiometric factors for the reaction between peroxyl radicals and retinoids, as measured under various conditions.
[40] N a t u r a l l y O c c u r r i n g Flavonoids: S t r u c t u r e , C h e m i s t r y , a n d H i g h - P e r f o r m a n c e Liquid C h r o m a t o g r a p h y M e t h o d s for Separation and Characterization By D I P A K K . D A S
Introduction The flavonoids are naturally occurring low molecular weight aryl chromones that are ubiquitous in the photosynthesizing cells of plants. They participate in the light-dependent phase of photosynthesis to catalyze electron transport, l thus regulating the ion channels involved in photophosphoregulation.2 After the photosynthesizing cells cease to exist, flavonoids are released and appear in plant juice and resin. Many members of the flavonoid family possess attractive colors, and as a consequence they play a vital role in the ecology of plants by making the flowers and fruits attractive to bees and birds. There is no evidence that flavonoids t j. C. Cantley, Jr., and G. G. H a m m e s , Biochemistry 15, 1 (1976). 2 H. Mukohato, S. Nakabayaski, and M. Higashida, FEBS Lett. 85, 215 (1978).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproduction in any formreserved.
410
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[40]
retinoic acid are from Sigma, whereas synthetic analogs of retinoic acid have been kindly supplied by Hoffman-La Roche (Basel, Switzerland). Under the assay conditions employed, 13-cis-retinoic acid is the most effective retinoid, followed by all-trans-retinoic acid. Other analogs of retinoic acid are ineffective in inhibiting oxidation of methyl linoleate. Because as reported above, the antioxidant efficiency of retinoids may be due to the effective steady-state concentration of the retinoid-derived carbon-centered radical, the higher antioxidant efficiency displayed by 13-cis-retinoic acid as compared to other retinoids may be the result of a relatively lower susceptibility of the carbon-centered intermediate from 13-cis-retinoic acid to autoxidative reactions. As has been pointed out, the antioxidant efficiencyof retinoids depends largely on the phase in which they scavenge radicals. The retinoids assayed here reveal very different potencies when assayed in other artificial systems, such as dioleoyl phosphatidylcholine (DOPC) liposomes. Table III summarizes the known stoichiometric factors for the reaction between peroxyl radicals and retinoids, as measured under various conditions.
[40] N a t u r a l l y O c c u r r i n g Flavonoids: S t r u c t u r e , C h e m i s t r y , a n d H i g h - P e r f o r m a n c e Liquid C h r o m a t o g r a p h y M e t h o d s for Separation and Characterization By D I P A K K . D A S
Introduction The flavonoids are naturally occurring low molecular weight aryl chromones that are ubiquitous in the photosynthesizing cells of plants. They participate in the light-dependent phase of photosynthesis to catalyze electron transport, l thus regulating the ion channels involved in photophosphoregulation.2 After the photosynthesizing cells cease to exist, flavonoids are released and appear in plant juice and resin. Many members of the flavonoid family possess attractive colors, and as a consequence they play a vital role in the ecology of plants by making the flowers and fruits attractive to bees and birds. There is no evidence that flavonoids t j. C. Cantley, Jr., and G. G. H a m m e s , Biochemistry 15, 1 (1976). 2 H. Mukohato, S. Nakabayaski, and M. Higashida, FEBS Lett. 85, 215 (1978).
METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproduction in any formreserved.
[40]
NATURALLY OCCURRING FLAVONOIDS
411
are produced in animal cells; they appear in animal and human cells following consumption of vegetables. The average daily diet of humans contains about 1 g of flavonoids, a surprising high amount and enough to bring the flavonoid level to pharmacologically significant concentrations in many tissues. Toxicity of flavonoids for animals and humans is extremely rare; no residual flavonoids accumulate in the body. Flavonoids are likely to play a variety of roles in mammalian cell function. At one point some members of flavonoid family were named vitamin P because of their ability to protect a cell against enhanced capillary permeability. 3 However, they are no longer considered as vitamins because further studies failed to fulfill all the criteria for a vitamin. Nevertheless, many flavonoids possess potent antioxidant, 4-6 anti-inflammatory,7 antiallergic,8 and antihemorrhagic 9 properties. A number of flavonoids have also been found to scavenge free radicals directly 1° and to inhibit lipid peroxidation. 1~ They can inhibit several important enzymes in cellular systems, including ATPase, ~2 phospholipase, ~3 prostaglandin cyclooxygenase, 14 lipoxygenase, 15 aldose reductase, 16 and hexokinase) 7 Flavonoids can also induce several enzymes, namely, aryl hydroxylase and epoxide hydroxylase. 18Thus, flavonoids seem to possess several pharmacologic properties which make them excellent agents to serve as natural biological response modifiers. 3 j. Kiihnau, World Rev. Nutr. Diet. 24, 117 (1976). 4 M. J. Laughton, B. Halliwell, P. J. Evans, and J. R. S. Hoult, Biochem. Pharmacol. 38, 2859 (1989). 5 0 . P. Sharma, Biochem. Pharmacol. 25, 1811 (1976). 6 M. Das and P. K. Roy, Biochem. Int. 17, 203 (1988). 7 M. Gabor, in "Handbook of Experimental Pharmacology" (J. R. Vane and S. H. Ferreira, eds.), pp. 215-231. Springer-Verlag, New York, 1980. 8 M. Sasajima, S. Nakane, R. Saziki, H. Saotome, K. Hatayana, K. Kyogoku, and I. Tanaka, Folia Pharmacol. Jpn. 74, 897 (1978). 9 R. J. Gryglewski, R. Korbut, J. Robak, and J. Swies, Biochem. Pharmacol. 36, 317 (1987). l0 j. Robak and R. J. Gryglewski, Biochem. Pharmacol. 37, 837 (1988). I1 j. Baumann, F. V. Bruchhausen, and G. Wurm, Prostaglandins 2,0, 627 (1980). t2 C. M. S. Fewtrell and B. D. Gomperts, Biochim. Biophys. Acta 469, 52 (1977). 13 j. R. Vane, Nature (London), New Biol. 231, 232 (1971). 14 M. J. Laughton, P. J. Evans, M. A. Moroney, J. R. S. Hoult, and B. Halliwell, Biochem. Pharmacol. 42, 673 (1991). 15 y . T. Yoshimoto, M. Furukawa, S. H. T. Yamamoto, and S. Watanabe-Kohno, Biochem. Biophys. Res. Commun. 116, 612 (1983). 16 S. Almeda, D. H. Bing, R. Laura, and P. A. Friedman, Biochemistry 20, 3731 (1981). 17 y. Graziani, Biochim. Biophys. Acta 460, 364 (1977). i8 M. T. Huang, F. F. Johnson, V. Miller-Eberhand, D. R. Koop, M. J. Coon, and A. H. Canney, J. Biol. Chem. 256, 10897 (1981).
412
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[40]
Structure and Chemistry Flavonoids are a class of naturally occurring organic phytochemicals derived from 2-phenylchromone. The immediate family members include flavone, isoflavone, 3-hydroxyflavone or flavonol, and the 2,3-dihydro derivatives of flavone, namely, flavanones, which are interconvertible with the isomeric chalcones. Flavanones undergo a series of transformations affecting the heterocyclic C ring to give rise to flavonoids. Anthocyanins and catechin, other members of the family, possess a benzopyrylium nucleus and a tetrahydropyranol ring, respectively. Other members from within each group originate from differences in oxygenation of rings A and C and from derivatization reactions such as O-methylation and O- and C-glycosylation. Flavone and flavonol glycosides are widely distributed in plants, the sugar residue being either glucose or rhamnose, occasionally a biose sugar. An extension of the flavonoid family includes flavoproteins, which are a group of proteins having riboflavin in the molecular portion of the prosthetic group. Included are Warburg's yellow enzyme and cytochrome reductase, both of which have riboflavin phosphate (i.e., flavin mononucleotide) as the prosthetic group, and diaphorase. The prosthetic group of diaphorase is flavin adenine dinucleotide, which has riboflavin phosphate and adenylic acid joined through the phosphate groups. Flavoproteins act as oxygen carriers in biological systems. The flavonoid aglycones possess a benzene ring (ring A) condensed with a six-membered heterocyclic ring (ring B) which carries a phenyl ring (ring C) as a substituent at the 2 position. Variations in the sixmembered ring and type of substituents give rise to the different members of the flavonoid family. For example, when the C ring is a y-pyrone the compounds are flavonols and isoflavonols and flavones and isoflavones; when the C ring forms the corresponding dihydro derivative the compounds are flavanones and isoflavanones (Fig. 1). When the benzenoid substituent is in the 2 position, the compounds are called flavone(s), and when it is in the 3 position, they are called isoflavonoids. A hydroxyl group at the 3 position distinguishes flavonols from other members of the family. Furthermore, C-hydroxylation and C-methoxylation, as well as C- and O-methylation, methoxylation, and glycosylation, proceed in a selective fashion, leading to formation of the entire flavonoid family. With the notable exception of isoflavones, all members of the flavonoid family have the same basic skeleton; the oxidation levels of the various carbons in the heteroxyclic ring are the key feature determining the structure type. The origin of the flavonoid family from flavanone is described in Fig. 2. An important structural feature that is most likely to be related to the antioxidant properties of many flavonoid compounds is the presence of
[40]
NATURALLYOCCURRINGFLAVONOIDS
413
c
0 Flovonot Isoflavonot
Ftavanone
Isoftavanone
0 Ftovone
Iso f tavone
FIG. 1. Main classes of the flavonoidfamily.
phenolic groups. In addition, the presence of a hydroxyl group on both aromatic rings seems to be helpful, but not essential, for the antioxidative action of flavonoids. With over 4000 known flavonoid compounds, the structures can vary considerably, and structure-activity relationships are expected to play an important role in determining the antioxidative functions of flavonoids. Isolation and Separation of Flavonoids by High-Performance Liquid Chromatography Extraction
Because of the wide variety of biological functions performed by flavonoids, there has been a growing interest in isolating and separating the flavonoids from plant products. Extraction procedures may vary considerably depending on the source of flavonoids. Flavonoids can be extracted
414
[40]
ANTIOXIDANT CHARACTERIZATION AND ASSAY H
11iii i1~01~
~
~11
0
Flavanones
~
OH
Chalcones
~
IIIII~OH
Ar
O
O Flavones
~
H
H ~Ar
H
H
H
~T~
~OH
0 Dihydroflavonois
O Aurones
O
Isoflavone$ FIG. 2. Origin of the main flavonoid family from flavanones (Reproduced with permission from D. Barton and W. D. Ollis, "Comprehensive Organic Chemistry: The Synthesis and Reactions of Organic Compounds." Pergamon, New York, 1979.)
[40]
NATURALLY OCCURRING FLAVONOIDS
415
from dried Ginkgo biloba leaves (600 mg) by refluxing with 60% aqueous acetone (15 ml) for 15 min. 19 Extracts should be filtered through a 0.45/xm membrane filter before the high-performance liquid chromatography (HPLC) separation. Dried flowers, leaves, or other plant parts may also be extracted with 60% methanol, 19'2° dichloromethane, 21 70% ethanol, 2z chloroform and warm methanol, 23'24hexane, light petroleum fraction and 70% aqueous ethanol, z5 warm ethanol, 26 as well as acetonitrile/water (80 : 20, v/v) containing 0.4% Triton X-100. 27 If the extracts contain any unwanted contaminants, they may be dried to a powder under a stream of N2 or under vacuum, then dissolved in water or a water/methanol mixture for reversed-phase (C18) solid-phase extraction (SPE), 19-21'27-3° and eluted with water/methanol or methanol alone. Reversed-phase C18 SPE can also be used to extract flavonoids from plasma. 3~,32
Separation Although several methods are available for the separation of flavonoids, including paper chromatography, 33 thin-layer chromatography (TLC) ,34and high-performance TLC,35 the most popular and useful method is analysis by H P L C . 19'23'25'27'28'36 Several years ago, S n y d e r 36 described an HPLC method for the separation of methylated flavonoid aglycones. Since then the method has been modified and refined considerably. Although normal-phase chromatography has been used for the separation
19 p. Pietta, P. Mauri, and A. Bruno, J. Chromatogr. 553, 223 (1991). 2o K. Eskins and H. J. Dutton, Anal. Chem. 51, 1885 (1991). 2J p. G. Pietta, P. L. Mauri, C. Gardana, and A. Bruno, J. Chromatogr. 547, 439 (1991). 22 p. Pietta, P. Mauri, C. Gardana, R. M. Facino, and M. Carini, J. Chromatogr. 537, 449 (1991). 23 T. Vogt and P.-G. Gulz, J. Chromatogr. 537, 453 (1991). 24 T. Z. Dzido, E. Soczewinski, and J. Gudej, J. Chromatogr. 550, 71 (1991). 25 F. A. Blouin and Z. M. Zarins, J. Chromatogr. 441, 443 (1988). 26 p. Li, T. Y. Zhang, X. Hua, and Y. Ito, J. Chromatogr. 538, 219 (1991). 27 D. Treutter, J. Chromatogr. 436, 490 (1988). 28 H. Schulz and G. Albroscheit, J. Chromatogr. 442, 353 (1988). 29 p. Pietta, E. Manera, and P. Ceva, J. Chromatogr. 357, 233 (1986). 3o p. Pietta, A. Bruno, P. Maufi, and A. Rava, J. Chromatogr. 593, 165 (1992). 31 y . Wakui, E. Yanagisawa, E. Ishibashi, Y. Matsuzaki, S. Takeda, H. Sasaki, M. Aburada, and T. Oyama, J. Chromatogr. 575, 131 (1992). 32 A. Marzo, E. A. Martelli, and G. Bruno, J. Chromatogr. 535, 255 (1990). 33 F. J. Francis, J. Food Sci. 50, 1640 (1985). 34 A. Hiermann, J. Chromatogr. 174, 478 (1979). 35 D. Heimler, J. Chromatogr. 366, 407 (1986). 36 L. R. Snyder, J. Chromatogr. Sci. 16, 223 (1978).
416
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[40]
of flavonoids,37'38 it is now generally agreed that reversed-phase HPLC is the method of choice for the separation of a wide variety of flavonoid compounds. Generally, such separations are rapid and provide high resolution and sensitivity. Because of the similarity in structure of the wide variety of flavonoid families, relatively uniform conditions can be used to yield satisfactory separations. For example, reversed-phase Cs or Cls columns, in conjunction with methanol-water or acetonitrile-water as eluting solvent, can be used to separate many flavonoid compounds effectively. 26,39 Several modifications may enhance the separation process. The addition of acetic, phosphoric, or formic acid to the above eluents has been found to enhance s e p a r a t i o n . 25,27,28,4°,41 Reversed-phase chromatography yields sharper peaks at lower pH along with better separation of peaks. Modification of the elution process simply by switching from isocratic elution42to gradient elution 25 can help achieve sharper separation. The particle size of the column packing also seems to be important. A smaller particle size such as 3 /zm can cause better peak resolution with lower run times. 21,25,27 Although several investigators used elevated column temperatures, 28'4° most researchers found equally good separations at ambient column temperature. 19,23,25,272-Propanol and tetrahydrofuran have also been employed in the mobile phase to improve flavonoid separation. 19,32 The most effective form of detection has been UV detection in the 330-365 nm range,19'21-23'28'29'4°'43'44which allows the use of photodiode array detection for peak identification) 9,21,22,2s The presence of a catechol moiety in flavonoids makes the compound electrochemically active and oxidizable, allowing the use of sensitive and selective electrochemical detector. 3~,32However, it is important to remember that electrochemical detection requires a low organic content in the mobile phase. As discussed above, organic solvents such as acetonitrile, methanol, 2-propanol, or tetrahydrofuran are required for most reversedphase separations of flavonoids. Whenever possible, the collection of chromatographic peaks for identification by mass spectral or nuclear magnetic resonance (NMR) analysis 44 would certainly be advantageous. Fi37 S. Hara and S. Ohnishi, J. Liq. Chromatogr. 7, 59 (1984). 3s M. V. Piretti and P. Doghieri, J. Chromatogr. 514, 334 (1990). 39 S. V. Ting, R. L. Rouseff, M. H. Dougherty, and J. A. Attaway, J. Food Sci. 44, 69 (1979). 4o I. Molnar, K. H. Gober, and B. Christ, J. Chromatogr. 550, 39 (1991). 41 j. p. Goiffon, M. Brun, and M. J. Bourrier, J. Chromatogr. 537, 101 (1991). 42 E. Moran, R. O'Kennedy, and R. D. Thornes, J. Chromatogr. 416, 165 (1987). 43 L. Zeng, R. Y. Zhang, T. Meng, and Z. C. Lou, J. Chromatogr. 513, 247 (1990). 44 L. Qimin, H. V. D. Heuvel, O. Delorenzo, J. Corthout, L. A. C. Pieters, A. J. Vlietinck, and M. Claeys, J. Chromatogr. 562, 435 (1991).
[40]
NATURALLY OCCURRING FLAVONOIDS
417
nally, the use of an approach such as that suggested by Snyder et al., 45 where methanol, acetonitrile, tetrahydrofuran, and water are blended so that an optimum resolution of the particular flavonoids in question can be achieved in a systemic fashion, is also advisable. The use of computer simulation software can achieve the optimum resolution of particular flavonoids.24,4°
Chromatography A typical HPLC chromatogram obtained from the extract of Ginkgo biloba leaves is shown in Fig. 3. Fifteen different varieties of flavonoids present in the extract are identified by retention times and by comparing the UV spectra of the peaks with those of the corresponding standards. An example of the UV spectra can be seen in Fig. 4. The Chromatogram in Fig. 3 was obtained by injecting a 10-/xl volume of the filtered flavonoid extract onto a C8 Aquapore RP 300 column (7 /~m particle size; 250 mm x 4 mm i.d.) (Brownlee Labs, Santa Clara, CA) in a Waters Associates (Milford, MA) chromatograph equipped with a Model U6K universal injector, two Model 510 pumps, and a Model 680 automated gradient controller. UV detection at 360 nm was obtained with a Model HP 1040 A photodiodearray detector (Hewlett-Packard, Waldbronn, Germany), where the acquisition of UV spectra was automatic at the apex, both inflection points, and the base of all peaks (230-430 nm, 2 nm steps). The flow rate was adjusted to 2 ml/min. Eluent A was water/2-propanol (95:5, v/v), and eluent B was 2-propanol/tetrahydrofuran/water (40 : 10 : 50, v/v/v). A linear gradient was run from 20% B to 60% B over a 40-min period. The UV spectra of each peak, after subtraction of the corresponding UV base spectrum, were computer normalized and the plots were superimposed. Peaks were considered to be homogeneous when there was exact correspondence among the corresponding spectra (match factor >990). It should be clear from the above discussions that extraction and separation methods can vary widely depending on the nature of the flavonoid compounds. The following represents an outline of a method which should work for a wide variety of flavonoids. However, it is important to remember that this is a suggested starting point for an unknown flavonoid compound, and the method may need modification. 1. Extract the plant products twice with hot methanol. 2. Combine the methanol extracts and then evaporate to dryness under N 2 at 40 °. 45 L. R. Snyder, J. L. Glajch, and J. J. Kirkland, "Practical HPLC Method Development," pp. 100-106. Wiley, New York, 1988.
A
xllt
vJ
8 P1 v
XI
E
VI
IX IV
II
VII
0 .L.,.J . . . . . . . . . . . . . . . . . . . . . . . .
5
10 15 Time (rain)
.
20
,
..,
.
25
B 70 P2
G
60 50 E) rr 40 E 30
B
10. VIII
20
30
40
50
Time (min)
Fl~. 3. Typical chromatogram from Ginkgo biloba leaves. Column, 7/zm C8 Aquapore RP-300; eluent, (a) water 2-propanol (95:5, v/v), (b) 2-propanoi/tetrahydrofuran water (40: 10:50, v/v/v), linear gradient from 20 to 60% B in 40 min; flow rate, 1 ml/min. There is an approximately 10-fold decrease in UV sensitivity from P1 to P2. Peaks are labeled as follows: (I) rutin; (II) isoquercitrin; (III) quercetrin; (IV) quercetin-3-O-[6"-p-coumaroylglucosyl-(l --* 2)-rhamnoside]; (V) quercetin-3-O-rhamnosyl-(l ---*2)-rhamnosyl-(l ~ 6)-glucoside; (VI) kaempferol-3-O-rutinoside; (VII) astragalin; (VIII) kaempferol-3-O-[6"-p-coumaroyl-(1 ---* 2)-rhamnoside]; (IX) kaempferol-glycoside; (X) kaempferol-3-O-rhamnosyl-(1 2)-rhamnosyl-(1 ---* 6)-glucoside; (XI) isorhamnetin-3-O-rutinoside; (B) bilobetin; (6) gintigetin; (IG) isoginkgetin; and (S) sciadopitxsin. (Reproduced with permission from Petta et al.19
[40]
419
NATURALLY OCCURRING FLAVONOIDS 100
Vlll
90 80 70 60 -~ 50 co 40
30 20 10 0
250
300 350 Wavelength (nm)
400
FIG. 4. Absorption spectra (UV) of kaempferol-3-O-rutinoside (VI), kaempferol-glycoside (IX), q u e r c e t i n - 3 - O - [ 6 " - p - c o u m a r o y l g l u c o s y i - ( l ~ 4)-rhamnoside] (IV), and kaempferol-3O-[6"-p-coumaroylglucosyl-(l ~ et a/. 19)
4)-rhannoside] (VIII). (Reproduced with permission from
Pietta
3. Resuspended the pellets in 30% (v/v) methanol and apply to SepPak Cls cartridge (Waters Associates), previously activated with methanol (3 ml), water (5 ml), and 30% methanol (5 ml). 4. Wash the cartridge with 50% methanol. Flavonoids are then eluted with methanol. 5. The elute may be dried under N2 at 40 ° and resuspended in methanol at a set volume (200-400 ~1) in order to concentrate the flavonoids. 6. Filter the sample through microfilterfuge tubes (0.2/zm Nylon-66 membrane filters, Rainin Instrument Co., Inc., Woburn, MA) prior to HPLC injection. 7. Inject approximately 25/.d of the filtered sample onto a Beckman Ultrasphere ODS Cls (3 ~m particle size, 7.5 × 4.6 mm i.d.) column (Rainin) in a Waters chromatograph equipped with a Millennium 2010 Chromatography Manager full control computer system, Satellite Wisp Model 700 injector, Model 996 photodiode array UV detector (set at 360 nm and set to scan the region 200-450 nm at peak maximum), two Model 510 pumps, a/xBondapak C18 Guard-Pak precolumn, and a Gilson (Middleton, WI) FC-203 fraction collector. Adjust the flow rate to 1 ml/min at ambient temperature.
420
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[41]
8. Run the sample under seven different isocratic gradients for 30 min in order to obtain information concerning the resolution of the flavonoids. The seven isocratic mobile phases are as follows: (1) methanol/ water/acetic acid (25:74: I, v/v/v); (2) acetonitrile/water/acetic acid (15:84: 1, v/v/v); (3) tetrahydrofuran/water/acetic acid (I 1:88: 1, v/v/v); (4) methanol/acetonitrile/water/acetic acid (25 : 15 : 59: 1, v/v/v/v); (5) acetonitrile/tetrahydrofuran/water/acetic acid 915 : 11 : 63 : 1, v/v/v/v); (6) methanol/tetrahydrofuran/water/acetic acid (25:11:63:1, v/v/v/v); (7) methanol/acetonitrile/tetrahydrofuran/water/acetic acid (25 : 15 : 11 : 48 : 1, v/v/v/v/v). This information should lead to the development of an optimum isocratic mobile phase. 9. If some ofthe flavonoids remain unresolved, then a gradient elution should be attempted. I0. Once the flavonoids are separated, the UV scans along with gas chromatography-mass spectrometry (GC-MS) should identify the various flavonoids.
[41] F l a v o n o i d A n t i o x i d a n t s : R a t e C o n s t a n t s for R e a c t i o n s with Oxygen Radicals B y W O L F BORS, CHRISTA M I C H E L ,
and
M A N F R E D SARAN
Introduction Flavonoids are plant secondary metabolites having a polyphenol structure, occurring mostly as glycosides or methoxylated derivatives, but also as aglycones) It has been assumed for years that they act as antioxidants, primarily based on the fact that they extend the shelf-life of fat-containing foodstuffs. 2 In contrast, an antioxidative function in plants themselves is still a matter of debate, 3 even though protective effects during plant photooxidative processes have been reported. 4'5 I j. B. Harborne, "The Flavonoids: Advances in Research Since 1980." Chapman & Hall, London, 1988. 2 j. Kiihnau, World Rev. Nutr. Diet. 24, 117 (1976). 3 j. B. Harborne, in "Plant Flavonoids in Biology and Medicine" (V. Cody, E. Middleton, and J. B. Harborne, eds.), p. 15. Alan R. Liss, New York, 1986. 4 U. Takahama, Photochern. Photobiol. 38, 363 (1983). 5 G. R. Wagner, R. J. Youngman, and E. F. Elstner, J. Photochem. Photobiol., B 1, 451 (1988).
METHODSIN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
420
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[41]
8. Run the sample under seven different isocratic gradients for 30 min in order to obtain information concerning the resolution of the flavonoids. The seven isocratic mobile phases are as follows: (1) methanol/ water/acetic acid (25:74: I, v/v/v); (2) acetonitrile/water/acetic acid (15:84: 1, v/v/v); (3) tetrahydrofuran/water/acetic acid (I 1:88: 1, v/v/v); (4) methanol/acetonitrile/water/acetic acid (25 : 15 : 59: 1, v/v/v/v); (5) acetonitrile/tetrahydrofuran/water/acetic acid 915 : 11 : 63 : 1, v/v/v/v); (6) methanol/tetrahydrofuran/water/acetic acid (25:11:63:1, v/v/v/v); (7) methanol/acetonitrile/tetrahydrofuran/water/acetic acid (25 : 15 : 11 : 48 : 1, v/v/v/v/v). This information should lead to the development of an optimum isocratic mobile phase. 9. If some ofthe flavonoids remain unresolved, then a gradient elution should be attempted. I0. Once the flavonoids are separated, the UV scans along with gas chromatography-mass spectrometry (GC-MS) should identify the various flavonoids.
[41] F l a v o n o i d A n t i o x i d a n t s : R a t e C o n s t a n t s for R e a c t i o n s with Oxygen Radicals B y W O L F BORS, CHRISTA M I C H E L ,
and
M A N F R E D SARAN
Introduction Flavonoids are plant secondary metabolites having a polyphenol structure, occurring mostly as glycosides or methoxylated derivatives, but also as aglycones) It has been assumed for years that they act as antioxidants, primarily based on the fact that they extend the shelf-life of fat-containing foodstuffs. 2 In contrast, an antioxidative function in plants themselves is still a matter of debate, 3 even though protective effects during plant photooxidative processes have been reported. 4'5 I j. B. Harborne, "The Flavonoids: Advances in Research Since 1980." Chapman & Hall, London, 1988. 2 j. Kiihnau, World Rev. Nutr. Diet. 24, 117 (1976). 3 j. B. Harborne, in "Plant Flavonoids in Biology and Medicine" (V. Cody, E. Middleton, and J. B. Harborne, eds.), p. 15. Alan R. Liss, New York, 1986. 4 U. Takahama, Photochern. Photobiol. 38, 363 (1983). 5 G. R. Wagner, R. J. Youngman, and E. F. Elstner, J. Photochem. Photobiol., B 1, 451 (1988).
METHODSIN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
[41]
SCAVENGING OF PEROXYL RADICALS BY FLAVONOIDS
421
The biochemical background for the antioxidative effect of flavonoids is inhibition of lipid peroxidation, which has been observed on numerous occasions. 6-1° Owing to the polyphenol structure, this inhibition can be brought about either by chelating of transition metals 11'12or by scavenging of free radicals with the formation of less reactive flavonoid aroxyl radicals. 12.13At present, radical scavenging is clearly the favored mechanism as evidenced by the lopsided ratio of reports on scavenging versus chelating properties of flavonoids. Most reports, however, compare only the relative radical scavenging efficiencies. In the case of oxygen radicals, this concerns the four representative species superoxide anions (O27), 8'14-18 hydroxyl radicals ('OH), 19'2° peroxyl radicals (RO2.), 21'22and alkoxyl radicals (RO.). However, O2-, in contrast to HO2., is incapable of initiating lipid peroxidation by itself, 23 and freely diffusible .OH radicals exist unequivocally only in irradiated aqueous solutions. Thus, both of these radical species are considered to be of minor relevance to lipid peroxidation processes, whereas peroxyl and alkoxyl radicals are the principal chain-initiating and chain-propagating intermediates. In the quest for quantitative structure-activity relationships (QSAR) of radical-scavenging properties of flavonoids, it is imperative that absolute rate constants for these reactions be known. For this purpose, sources 6 A. K. Ratty and N. P. Das, Biochem. Med. Metab. Biol. 39, 69 (1988). 7 j. Robak, F. Shridi, M. Wolbis, and M. Krolikowska, Pol. J. Pharmacol. Pharm. 40, 451 (1989). 8 M. T. Meunier, E. Duroux, and P. Bastide, Plant. Med. Phytother. 23, 267 (1989). 9 A. Mora, M. Payfi, J. L. Rios, and M. J. Alcaraz, Biochem. Pharmacol. 40, 793 (1990). l0 G. T. Liu, T. M. Zhang, B. E. Wang, and Y. M. Wang, Biochem. Pharmacol. 43, 147 (1992). 11 C. A. B. Clemetson and L. Andersen, Ann. N.Y. Acad. Sci. 136, 339 (1966). 12 I. B. Afanas'ev, A. I. Dorozhko, A. V. Brodskii, V. A. Kostyuk, and A. I. Potapovitch, Biochem. Pharmacol. 38, 1763 (1989). 13 N. Cotelle, J. L. Bernier, J. P. H6nichart, J. P. Catteau, E. Gaydou, and J. C. Wallet, Free Radical Biol. Med. 13, 211 (1992). 14 j. Baumann, G. Wurm, and F. von Bruchhausen, Arch. Pharm. (Weinheim, Ger.) 313, 330 (1980). 15 j. C. Monboisse, P. Braquet, A. Randoux, and J. P. Borel, Biochem. Pharmacol. 32, 53 (1983). 16j. Robak and R. J. Gryglewski, Biochem. Pharmacol. 37, 837 (1988). 17 A. I. Huguet, S. Manez, and M. J. Alcaraz, Z. Naturforsch., C: Biosci. 45C, 19 (1990). 18 G. Sichel, C. Corsaro, M. Scalia, A. J. di Bilio, and R. P. Bonomo, Free Radical Biol. Med. 11, 1 (1991). 19 S. R. Husain, J. Cillard, and P. Cillard, Phytochemistry 26, 2489 (1987). 20 A. Puppo, Phytochemistry 31, 88 (1992). 21 j. Torel, J. Cillard, and P. Cillard, Phytochemistry 25, 385 (1986). 22 T. Ariga and M. Hamano, Agric. Biol. Chem. 54, 2499 (1990). 23 B. H. J. Bielski, R. L. Arudi, and M. W. Sutherland, J. Biol. Chem. 258, 4759 (1983).
422
A N T I O X l D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[41]
o f o x y g e n radicals w h i c h c a n b e u t i l i z e d for k i n e t i c s t u d i e s a n d e x a m p l e s o f specific r e s u l t s a r e p r e s e n t e d in this c h a p t e r .
Methods for Oxygen Radical Generation Radiolytic, photolytic, chemical, and enzymatic systems m a y bc used as sources of oxygen radicals. Although the usefulness for kinetic studies declines in the above order, each of the systems has bccn employed at Mast for comparative studies. Radiolysis, or its derivative method pulse radiolysis, represents the cleanest and most selective production system for oxygen radicals24-26;the method has bccn described repeatedly in this series.27-29 Photolysis is capable of producing radicals from sufficiently labile compounds, for example, hydro- or endopcroxidcs yielding alkoxyl radicals) °-32 Moreover, its efficacy can occasionally bc enhanced by the inclusion of sensitizer molecules which themselves absorb light energy and transfer it to the radical-yielding molecule) 3 Chemical systems to generate oxygen radicals arc k n o w n for each of the four principal species. Sources for . O H radicals comprise Fenton systems in aqueous solutions 2°'34or specific generation systems in organic solvents) 5,36The Fenton system, however, is ambiguous 37'3sas reactions of frcc . O H radicals or the strongly oxidizing fcrryl species cannot bc discriminated on the basis of kinetic or analytical data.39 Chemical sources for 02-=arc less convenient 16'4°and likewise lack the specificityof radiolytic 24w. Bors, W. Hcller,C. Michel, and M. Saran, thisseries,Vol. 186,p. 343. M. Erbcn-Russ, W. Bors, and M. Saran, Int. J. Radiat. Biol. 52, 393 (1987). 26I. Gy6rgy, S. Antus, A. Blazovics,and G. F61diak,Int. J. Radiat. Biol. 61, 603 0992). 27K. D. Asmus, thisseries,Vol. I05,p. 167. M. G. Simic,thisseries,Vol. 186,p. 89. 29M. Saran and W. Bors, thisseries,Vol. 233 [2]. 3oE. Pcllc,D. Macs, G. A. Padulo,E. K. Kim, and W. P. Smith,Arch. Biochem. Biophys. 283, 234 (1990). 31 Z. F61des-Papp, G. Gerber, R. St6sser, and G. Schneider, J. Prakt. Chem. 333, 293 (1991). 32W. Bors, C. Michel, and K. Stettmaier, J. Chem. Soc., Perkin Trans. 2, p. 1513 (1992). 33D. Griller, K. U. Ingold, and J. C. Scaiano, J. Magn. Reson. 38, 169 (1980). 34I. E. Blasig, H. Loewe, and B. Ebert, Biomed. Biochim. Acta 47, $252 (1988). 35R. D. Grant, E. Rizzardo, and D. H. Solomon, J. Chem. Soc., Chem. Commun., p. 867 (1984). 36T. Tezuka, in "The Role of Oxygen in Chemistry and Biochemistry" (W. Ando and Y. Moro-oka, eds.). Elsevier, Amsterdam; Stud. Org. Chem. 33, 157 (1988). 37H. Yamamoto, H. Takei, T. Yamamoto, and M. Kimura, Chem. Pharm. Bull. 27, 789 (1979). 38j. D. Rush and W. H. Koppcnol, J. Biol. Chem. 261, 6730 (1986). 39j. D. Rush, Z. Maskos, and W. H. Koppenol, this series, Vol. 186, p. 148. 40 I. Ueno, M. Kohno, K. Haraikawa, and I. Hirono, J. Pharm. Dyn. 7, 798 (1984).
[41]
SCAVENGING OF PEROXYL RADICALS BY FLAVONOIDS
423
sources. Peroxyl radicals result from oxygen attachment to alkyl radicals formed after thermolysis of a number of azo compounds. 4~ This steadystate source has already been applied to study flavonoid antioxidants. 22,42 It has to be considered, however, that substituent effects and/or diffusion limitations owing to the bulky structures may result in completely different reactivities as compared to lipid-derived peroxyl radicals. Therefore, "controlled" lipid peroxidation initiated by such a z o c o m p o u n d s 43,44 would seem to be the only reasonable chemical source of peroxyl radicals, spontaneous autoxidation being too slow and too inefficient a radical source. Chemical generation of alkoxyl radicals via the organic Fenton reaction is fraught as well by ambiguity owing to the multitude of radicals formed 45 and is thus a poor system for kinetic studies (and never with flavonoids). Enzymatic sources, finally, are xanthine o x i d a s e 13A6,34,46 o r the NADPH oxidase of phagocytes,13,47-5° which generate predominantly 02". Enzymatic 02- production, however, is accompanied to a certain extent by H20 2 production via the dismutation of 02". These unknown H202 levels may be liable to unpredictable admixtures of .OH radicals arising from Fenton reactions via trace metal contamination. Lipoxygenase, in contrast, generates fatty acid peroxyl radical intermediates almost exclusively, 5~ but as these are produced in close proximity of the prosthetic site, they may not be freely diffusible and would thus not be capable of reacting stoichiometrically with antioxidant compounds. A fUrther complication arises from the fact that the bleaching of carotenoids, cooxidized 41 E. Niki, this series, Vol. 186, p. 100. 42 V. M. Darley-Usmar, A. Hersey, and L. G. Garland, Biochem. Pharmacol. 38, 1465 (1989). 43 L. R. C. Barclay, S. J. Locke, J. M. MacNeil, and J. VanKessel, J. Am. Chem. Soc. 106, 2479 (1984). 44 E. A. Lissi, M. Salim-Hanna, M. Faure, and L. A. Videla, Xenobiotica 21, 995 (1991). 45 T. L. Greenley and M. J. Davies, Biochim. Biophys. Acta 1116, 192 (1992). 46 R. Salvayre, P. Braquet, T. Perruchot, and L. Douste-Blazy, in "Flavonoids and Bioflavonoids 1981" (L. Farkas, M. Gabor, F. Kallay, and H. Wagner, eds.). Elsevier, Amsterdam; Stud. Org. Chem. 11, 437 (1982). 47 C. Pagonis, A. I. Tauber, N. Pavlotsky, and E. R. Simons, Biochem. Pharmacol. 35, 237 (1986). 48 M. Damon, F. Michel, C. Le Doucen, and A. Crastes de Paulet, Bull. Liaison--Groupe Polyphenols 13, 569 (1986). 49 M. Lonchampt, B. Guardiola, N. Sicot, M. Bertrand, L. Perdrix, and J. Duhault, Arzneim.Forsch. 39, 882 (1989). 5o B. A. 't Hart, T. R. A. M. Ip Vai Ching, H. van Dijk, and R. P. Labadie, Chem.-Biol. Interact. 73, 323 (1990). 51 j. F. G. Vliegenthart, Chem. Ind. (London), p. 241 (1979).
424
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[41]
during the catalytic activity of lipoxygenase, 52'53can be effectively inhibited by flavonoids :4 yet the enzyme activity could also be directly inhibited by flavonoids. TM
Procedures
Preparation of Solutions for Pulse Radiolysis Experiments Aqueous solutions at pH 8-8.5, saturated separately with either N20 or 02 ; concentration of aliphatic (fatty acids, alcohols, etc.) peroxyl radical precursors 1 mM, concentration of flavonoids 10-100/~M. To facilitate the dissolution of flavonoids, stock solutions should be prepared at higher pH (pH I0-I 1); in this case absence of oxygen should be ensured to avoid autoxidation reactions. 56,57Owing to the higher solubility of N20 in water (25.5 mM) as compared to 02 (1.4 mM), all hydrated electrons (e~q) would be converted to •OH radicals rather than to 027 [see reaction (1)], provided the mixing ratio of N20- to O2-saturated solutions remains greater than I. Aqueous solutions at pH 11, containing 10 mM NaN3 and saturated with N20; fatty acid hydroperoxides at 1 mM and flavonoids at 10-100/xM. Pulse doses of 10 Gy provide radical concentrations of about 6/zM (G value, i.e., radiolytic yield of 5.6 molecules per 100 eV absorbed energy, equivalent to 0.57/zmol/J). The observation parameter is the buildup of the flavonoid aroxyl radical absorption 24 as a pseudo-first-order reaction which, plotted against the flavonoid concentration, directly yields the second-order scavenging rate constant. Competition Kinetics. As the expensive instrumentation of pulse radiolysis limits this method to very few laboratories, an altogether more accessible method, namely, competition kinetics, may be employed pro52 A. Ben Aziz, S. Grossman, I. Ascarelli, and P. Budowski, Phytochemistry 10, 1445 (1971). 53 W. Grosch and G. Laskawy, Biochim. Biophys. Acta 575, 439 (1979). 54 j. Oszmianski and C. Y. Lee, J. Agric. Food Chem. 38, 688 (1990). 55 M. J. Laughton, P. J. Evans, M. A. Moroney, J. R. S. Hoult, and B. Halliwell, Biochem. Pharmacol. 42, 1673 (1991). 56 W. F. Hodnick, E. B. Milosavljevic, J. H. Nelson, and R. S. Pardini, Biochem. Pharamcol. 37, 2607 (1988). 57 A. T. Canada, E. Giannella, T. D. Nguyen, and R. P. Mason, Free Radical Biol. Med. 9, 441 (1990).
[41]
SCAVENGING OF PEROXYL RADICALS BY FLAVONOIDS
425
vided specific steady-state sources of oxygen radicals (e.g., X- or Xirradiation, photolysis) are available. Based on the equation ACo/AC = Vo/V = 1 + (kc/ks)[C]/[S ]
a plot of the ratio of concentration or rate changes in the absence versus the presence of a competing substance (C, e.g., an antioxidant) against the ratio of the initial concentration of the competitor versus reference substance (S) yields a straight line intersecting the ordinate at unity. The use of a reference substance, whose absolute rate constant with the radical in question is known, allows the calculation of absolute rate constants for the tested substances from the slope, 58,59yet care has to be taken to check for secondary radical reactions which might distort the scavenging reaction rate constant) 8 As an example, the determination of the reactivity of photolytically generated tert-butoxyl radicals 6° has been described in more detail. 59 The same "crocin assay" has also been employed to determine the reactivity of peroxyl radicals derived from azo initiators. 59,61For these radicals, absolute rate constants are still unavailable as the rate constant(s) with crocin has yet to be evaluated. To observe the competitive inhibition of the reaction of the reference substance with the radicals, other methods can also be employed: enhanced chemiluminescence (especially after oxygen radical generation by phagocytes 13'62'63 and electron paramagnetic resonance (EPR) spectroscopy in combination with spin trapping. 34'4°'64 Specific Results Obtained by Pulse Radiolysis The majority of absolute rate constants for flavonoids have been obtained after generating different types of oxygen radicals by pulse radiolysis and observing the temporal absorption changes due to aroxyl radical formation or substrate depletion by rapid kinetic spectroscopy. 24-26,65As the determination of absolute rate constants of flavonoids with .OH and O5 have been previously described in this series, 24 in this chapter the 58 W. Bors, C. Michel, and M. Saran, in "CRC Handbook of Methods for Oxygen Radical Research" (R. A. Greenwald, ed.), p. 181. CRC Press, Boca Raton, Florida, 1985. 59 W. Bors, C. Michel, and M. Saran, in "Lipid-Soluble Antioxidants: Biochemistry and Clinical Applications" (A. S. H. Ong and L. Packer, eds.), p. 52. Birkhaeuser, Basel, 1992. 6o W. Bors, C. Michel, and M. Saran, Biochim. Biophys. Acta 796, 312 (1984). 61 M. Coassin, F. Ursini, and A. Bindoli, Arch. Biochem. Biophys. 299, 330 (1992). 62 N. Suzuki, A. Goto, I. Oguni, S. Mashiko, and T. Nomoto, Chem. Express 6, 655 (1991). 63 C. Pascual and C. Romay, J. Biolumin. Chemilumin. 7, 123 (1992). 64 S. Kitagawa, H. Fujisawa, and H. Sakurai, Chem. Pharm. Bull. 40, 304 (1992). 65 W. Bors, W. Heller, C. Michel, and M. Saran, in " F r e e Radicals and the Liver" (G. Csomos and J. Feher, eds.), p. 77. Springer-Verlag, Berlin, 1992.
426
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[41]
reaction with pulse-radiolytically generated peroxyl radicals is emphasized. In irradiated aqueous solutions, two principal conversion systems may be employed. 25'65 First, in N20-saturated solutions all hydrated electrons (eaq) are converted to .OH radicals: eaq + N20 + H20--* "OH + O H - + N 2
(1)
The .OH radicals (and to a lesser extent the remaining 10% H atoms) abstract hydrogen atoms from aliphatic structures such as fatty acids or alcohols: •OH + R--CH2--R'--COOH---~ R - - . C H - - R ' - - C O O H + H20 •OH + R--CH2--OH----> R - - . C H - - O H + H20
(2) (3)
Owing to the high reactivity of .OH radicals, such reactions occur quite randomly, 25,66,67 and it is understandable that the subsequent addition of O2 to these alkyl radicals by a diffusion-controlled reaction 68'69 leads to various isomeric peroxyl radicals: R - - . C H - - R ' - - C O O H + O2--->R - - C ( H ) O O - - - R ' - - C O O H R--.CH---OH + O2 ~ R - - C ( H ) O O . - - O H
(4) (5)
Aside from being scavenged by antioxidants, peroxyl radicals can undergo competing decay reactions, if concentration and/or reactivity of the antioxidant are low. Fatty acid peroxyl radicals have been calculated to decay in bimolecular reactions with rate constants of 2-4 x 108 M -~ sec-~. 25 The fate of isopropyl peroxyl radicals is more complicated and depends on pH and on the radical concentration (i.e., pulse dose). Both a spontaneous first-order decay and a base-catalyzed reaction lead to the liberation of 02- (and acetone): (CH3)aC(OH)OO" ~ 02- + H + + ( C H 3 ) 2 C = O + O H - --~ O2~ + H 2 0 + ( C H 3 ) 2 C - ~ - O
(CH3)zC(OH)OO"
(6) (7)
albeit with widely differing rate constants, k6 being 6.5-7.0 x 102 sec -~ and k7 being 5 x l0 9 M -I s e c - l . 70,71 The same products are formed in 66 M. G. J. Heijman, A. J. P. Heitzman, H. Nauta, and Y. K. Levine, Radiat. Phys. Chem. 26, 83 (1985). 67 K. D. Asmus, H. M6ckel, and A. Henglein, J. Phys. Chem. 77, 1218 (1973). 68 G. E. Adams and R. L. Willson, Trans. Faraday Soc. 65, 2981 (1969). 69 A. Marchaj, D. G. Kelley, A. Bakac, and J. H. Espenson, J. Phys. Chem. 95, 4440 (1991). 70 y . Ilan, J. Rabani, and A. Henglein, J. Phys. Chem. 80, 1558 (1976). 71 E. Bothe, G. Behrens, and D. Schulte-Frohlinde, Z. Naturforsch. B: Anorg. Chem., Org. Chem. 32B, 886 (1977).
[41]
SCAVENGING OF PEROXYL RADICALS BY FLAVONOIDS
a reaction catalyzed by phosphate buffer (Ilan M-I
et
427
a/.7°; k8 = 1.1 x 107
sec-1):
(CH3)zC(OH)OO" + HPO42- ~ 027 + H2PO 4- + (CH3)2C=O
(8)
In addition, second-order decay may also occurT°: (CH3)2C(OH)OO-
0 2 q"- 2 ( C H 3 ) z C ( O H ) O . O 2 + (CH3)zC(OH)--OO--C(OH)--(CH3)
2
(9) Because peroxyl radicals lack a distinctive absorption spectrum, the scavenging reaction by flavonoids can only be determined from the buildup of the absorption of the flavonoid aroxyl radical. 25 Using this approach, absolute rate constants of flavonoids with isopropyl peroxyl radicals 65 and with linoleyl peroxyl radical isomers 25 have been determined (Table I). An alternative generation system for peroxyl radicals has been established to determine whether the various peroxyl radical isomers, produced by the random attack of .OH radicals at linoleic acid [reactions (2) and (4)], are kinetically equivalent, z5 In NzO-saturated solutions containing
TABLE I RATE CONSTANTS OF SELECTED FLAVONOIDS WITH PEROXYL RADICALSa
Rate constant (x 108 M -1 sec -I) Substrate b Flavonol Kaempferol (3,5,7,4') Quercetin (3,5,7,3',4') Flavone Apigenin (5,7,4') Luteolin (5,7,3' ,4')
R(OH)OO.
n-LOO.
13-LOO-
19.5 2.1
0.34 0.18
0.42 0.15
n.d. c n.d.
n.d. n.d.
0.77 1.85
a Kinetic evaluation of aroxyl radical formation after pulse-radiolytic generation of peroxyl radicals: R(OH)OO., isopropylperoxyl radicals at pH 8.5-9.065, n-LOO., linoleic acid peroxyl radical isomers in N20/O2 system, pH 11.5; 13-LOO., 13-hydroperoxyl linoleic acid radical in N20/N 3- system, pH 11.5. 25 b Numbers in parentheses denote positions of hydroxyl groups. c n.d., Not determined.
428
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[41]
sodium azide, all .OH radicals obtained in reaction (1) are converted to azide radicals: •O H + N 3- ~
0o)
"N 3 + O H -
Azide radicals are highly electrophilic but are more restricted in their reactivity than .OH radicals. Thus, they are unable to abstract aliphatic H atoms from fatty acids at an appreciable rate 72 and react preferentially by electron transfer. 73 In the specific case of enzymatically produced fatty acid hydroperoxides, the dissociation of the hydroperoxy group with a pK value above 12TM facilitates a direct oxidation of this group by .N 3 radicals in strongly alkaline solutions25: R--C(H)OO---CH2--COO-
+ "N 3 R--C(H)OO'--CH2--COO-
+ N 3-
(11)
These conditions furthermore allow the use of millimolar concentrations of fatty acids in homogeneous solutions, as at lower pH micelle formation and consequently lower solubility prevails. 72It turned out that these selectively generated 13-hydroperoxyllinoleic acid radicals were kinetically quite similar to the random mixture of peroxyl radical isomers, obtained in reactions (2) and (4). 25 This obviates the need of using the latter, more complicated and more limited, system.
Conclusions The determination of the reactivities of flavonoid antioxidants as well as the identification of the corresponding radical structures 24,75are worthwhile endeavors as several open questions beckon to be answered. First, do flavonoids also function as antioxidants in plant tissue itself? Second, should secondary reactions of the flavonoid aroxyl radicals occur with lipid moieties (on time scales too slow for pulse-radiolytic observation), such an initiation of chain reactions would diminish the apparent antioxidao tive potential of these polyphenols. A third area of potential surprises concerns the unresolved interactions of flavonoids or their aroxyl radicals 72 M. Erben-Russ, W. Bors, R. Winter, and M. Saran, Radiat. Phys. Chem. 27, 419 (1986). 73 Z. B. Alfassi and R. H. Schuler, J. Phys. Chem. 89, 3359 (1985). 74 L. S. Silbert, in "Organic Peroxides" (D. Swern, ed.), Vol. 2, p. 637. Wiley (Interscience), New York, 1970. 75 W. Bors, W. Heller, C. Michel, and K. Stettmaier, in "Free Radicals: From Basic Science to Medicine" (G. Poli, M. Albano, M. U. Gianzani, eds.), p. 374. Birkhaeuser, Basel (1993).
[42]
ASSAYS FOR C O N D E N S E D T A N N I N S
429
with other biological cofactors, such a s ascorbate 76,77 or glutathione. A reminder of the complexities of flavonoid structure-activity studies is the fact that similar structural requirements seem to exist for various functions of flavonoids, some of which may not involve radical intermediates at all. 65 76 A. Bentsath, S. Rusznyak, and A. Szent-Gy6rgyi, Nature (London) 139, 326 (1937). 77 A. Negre-Salvayre, A. Affany, C. Hariton, and R. Salvayre, Pharmacology 42, 262 (1991).
[42] A s s a y o f C o n d e n s e d T a n n i n s or F l a v o n o i d O l i g o m e r s a n d R e l a t e d F l a v o n o i d s in P l a n t s
By ANN E. HAGERMAN and LARRY G. BUTLER Introduction The flavonoids are a diverse group of plant phenolics based on a 15carbon skeleton. The chemistry of the simple flavonoids is well known, and reviews of their structures, reactions, and distribution in higher plants are available. 1 The discussion in this chapter is restricted to the flavonoid oligomers and polymers known as condensed tannins, and a few closely related flavonoids. The condensed tannins (I) are formally polymers of flavan-3-ols such as catechin (II). The probable biosynthetic precursors of the condensed tannins are flavan-3,4-diols, or leucoanthocyanidins, such as catechin4fl-ol (111).2 Oxidative degradation of condensed tannins yields the corresponding anthocyanidins, such as cyanidin (IV), so that the condensed tannins are sometimes called proanthocyanidins. The pattern of hydroxyl-
oLO. vOH v ~OH I
8
6~
°"
O H OH 4
II
I R. J. Grayer, Methods Plant Biochem. 1, 283 (1989). 2 H. A. Stafford, "Flavonoid Metabolism." CRC Press, Boca Raton, Florida, 1990.
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All fights of reproduction in any form reserved.
[42]
ASSAYS FOR C O N D E N S E D T A N N I N S
429
with other biological cofactors, such a s ascorbate 76,77 or glutathione. A reminder of the complexities of flavonoid structure-activity studies is the fact that similar structural requirements seem to exist for various functions of flavonoids, some of which may not involve radical intermediates at all. 65 76 A. Bentsath, S. Rusznyak, and A. Szent-Gy6rgyi, Nature (London) 139, 326 (1937). 77 A. Negre-Salvayre, A. Affany, C. Hariton, and R. Salvayre, Pharmacology 42, 262 (1991).
[42] A s s a y o f C o n d e n s e d T a n n i n s or F l a v o n o i d O l i g o m e r s a n d R e l a t e d F l a v o n o i d s in P l a n t s
By ANN E. HAGERMAN and LARRY G. BUTLER Introduction The flavonoids are a diverse group of plant phenolics based on a 15carbon skeleton. The chemistry of the simple flavonoids is well known, and reviews of their structures, reactions, and distribution in higher plants are available. 1 The discussion in this chapter is restricted to the flavonoid oligomers and polymers known as condensed tannins, and a few closely related flavonoids. The condensed tannins (I) are formally polymers of flavan-3-ols such as catechin (II). The probable biosynthetic precursors of the condensed tannins are flavan-3,4-diols, or leucoanthocyanidins, such as catechin4fl-ol (111).2 Oxidative degradation of condensed tannins yields the corresponding anthocyanidins, such as cyanidin (IV), so that the condensed tannins are sometimes called proanthocyanidins. The pattern of hydroxyl-
oLO. vOH v ~OH I
8
6~
°"
O H OH 4
II
I R. J. Grayer, Methods Plant Biochem. 1, 283 (1989). 2 H. A. Stafford, "Flavonoid Metabolism." CRC Press, Boca Raton, Florida, 1990.
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All fights of reproduction in any form reserved.
430
[42]
ANTIOXIDANT CHARACTERIZATION AND ASSAY OH
HO
HO
"~H T
"OH
0 OH
OH
III
IV
ation on the A and B rings of the flavonoids is variable, yielding structural diversity among the condensed tannins) Structural complexity also results from the diversity of positions for polymerization of the parent flavan-3,4-diols, and from the possible stereochemical variations in the interflavan bonds. For example, simple linear condensed tannins such as the 4 ~ 8-1inked procyanidin B-1 (I) are common in some plants, but species rich in 5-deoxyflavan-3-ols produce branched chain polymers such as the profisetinidins (e.g., V). Some compounds with multiple interflavan bonds, such as proanthocyanidin A-2 (VI), have been isolated. Stereochemical diversity is illustrated by the series of procyanidin dimers B-1 (I), B-2 (VII), B-3 (VIII) and B-4 (IX). The condensed tannins commonly isolated from plant tissues and used in most studies of the biological effects of tannins contain not only oligomers like those depicted here but also less well-characterized polymers. Degrees of polymerization as high as 50 have been claimed for some condensed tannins.4 OH
O
OH
OH
o.
0~ ]
0H
OH
HO
V
VI
3 R. W. Hemingway, in "Chemistry and Significance of Condensed Tannins" (R. W. Hemingway and J. J. Karchesy, eds.), p. 83. Plenum, New York, 1989. 4 V. M. Williams, L. J. Porter, and R. W. Hemingway, Phytochemistry 22, 569 (1983).
[42]
ASSAYS FOR CONDENSED TANNINS
431
r~"~OH
HOe"IT OH VII
OH
H
O
~
~
OH
OH
OH
VIil
IX
Condensed tannins can be isolated from a wide variety of plant species and tissues. Among the most common sources of tannins used in laboratory studies are grain from Sorghum bicolor and extracts from the bark of quebracho (Schinopsis spp.). The tannin from Sorghum grain is largely comprised of 4 --~ 8-1inked procyanidin polymers similar to B-1 (I), and that from quebracho is a complex mixture of profisetinidins (V). The condensed tannins are structurally and chemically distinct from the hydrolyzable tannins (gallotannins and ellagitannins),5 so the commercial gallotannin known as tannic acid is an inappropriate material to use as a standard in studies of condensed tannins.
Extraction and Purification of Condensed Tannins Fresh, frozen, or lyophilized tissue can be extracted, although in general fresh material yields the best results. Sample treatment does affect phenolic extractability, so all samples should be treated similarly. 6 To obtain sufficient quantities of tannin for routine characterization, about 2 g dry weight of the tissue is needed. The method can be scaled down for analytical work. The ground or homogenized plant tissue is extracted with 10 volumes of 70% acetone (acetone/water, 70:30, v/v) and is sonicated for 30 min at 4°. The sample is centrifuged and the supernatant is stored at 4 °. The extraction is repeated three more times, and the extracts are pooled to obtain an overall recovery of about 90% of the total extractible phenolics in the tissue. The first two extracts usually contain about 75% of the total extractable phenolics. Tannins are usually purified by taking advantage of their sorption by Sephadex LH-20. 7 The acetone is removed from the crude plant extract 5 E. Haslam, "Plant Polyphenols: Vegetable Tannins Revisited." Cambridge Univ. Press, Cambridge, 1989. 6 A. E. Hagerman, J. Chem. Ecol. 14, 453 0988). 7 D. H. Strumeyer and M. J. Malin, J. Agric. Food Chem. 23, 909 (1975).
432
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[42]
and replaced by a smaller volume of ethanol before applying the sample to the Sephadex, which is equilibrated in ethanol. Nontannin phenolics are eluted with ethanol while monitoring absorbance at 280 nm. The tannin is then recovered by elution with aqueous acetone. The adsorption can be done either batchwise, for purification of large quantities of tannin, or in a column, for separation of individual tannin polymers. This method can be used either to purify tannin from the plant source of interest, or to purify well-characterized condensed tannins that are to be used as analytical standards. Quebracho tannin is commercially available as a crude powder which contains 30-50% nontannin materials. The material can be obtained in bulk from Trask Chemical Corporation [3200 W. Somerset Court, Marietta, GA 30067; (404) 955-9190] or from Tannin Corporation [60 Pulaski Street, P.O. Box 606, Peabody, MA 01960; (617) 532-4010]. Small amounts of quebracho tannin to be used for analytical standards can be obtained from Dr. Ann Hagerman (Department of Chemistry, Miami University, Oxford OH 45056). Sorghum tannin is often a more useful standard than quebracho tannin, because it gives higher color yields in several of the methods used to determine tannin. Quebracho tannin and Sorghum tannin can be purified by a modification of the Sephadex LH-20 method described above. 8,9
General Phenolic Assays M a n y methods for determining total phenolics have been described. These methods rely on the chemistry of the phenolic group and are thus not specific for tannins, but they provide insight into the levels of total (tannin plus nontannin) phenolics in plant extracts. A m o n g the most comm o n general phenolic methods are a variety of spectrophotometric redox assays, which detect the easily oxidized phenolic group. Redox methods are subject to interference by many compounds; of particular importance are the antioxidants often added to plant extracts, such as ascorbic acid. One widely used rcdox method is the Folin (Lowry, Folin-Denis) method) ° W e prefer the Prussian blue method, H since it is less subject to interference from nonphenolic compounds than is the Folin method.~2 For routine analysis using the Prussian blue method, dispense 0.10ml samples into 125-mi Erlenmeyer flasks and add 50.0 ml distilledwater. To each sample, add 3.0 rnl of 0.10 M FeNH4(SO4)2 in 0. I M HCI and s T. N. Asquith and L. G. Butler, J. Chem. Ecol. 11, 1535 (1985). 9 A. E. Hagerman and L. G. Butler, J. Agric. Food Chem. 28, 947 (1980). 10T. Swain and W. E. Hillis, J. Sci. Food Agric. 10, 63 (1959). 11 M. P. Price and L. G. Butler, J. Agric. Food Chem. 25, 1268 (1977). t2 A. E. Hagerman and L. G. Butler, J. Chem. Ecol. 15, 1795 (1989).
[42]
ASSAYS FOR CONDENSED TANNINS
433
swirl. Additions should be timed; 1-min intervals are convenient. Exactly 20 min after the addition of the FeNH4(SO4) 2, start timed additions (1min intervals) of 3.0 ml of 8 mM K3Fe(CN)6 , mixing after each addition. Exactly 20 min after the addition of the K3Fe(CN) 6 , read the absorbance at 720 nm, making readings at 1-min intervals. The spectrophotometric blank should undergo the same timed reaction but substituting sample solvent for the sample; the solvent can influence the reaction progress, so standards should be dissolved in the same solvent as the unknowns. Any phenolic compound can be used as a standard for the Prussian blue assay; gallic acid or catechin are convenient. When interpreting the results obtained with any redox assay, including the Prussian blue assay, it must be recognized that color yield is dependent on phenolic structure and the redox potential of the phenolic groups. F u n c t i o n a l G r o u p Assays
Anthocyanidin Formation A characteristic reaction of condensed tannins is oxidative cleavage of the interflavan bond to produce anthocyanidin pigments such as cyanidin (IV). This reaction has long been used as the basis for determining condensed tannins, but the chemistry of the reaction has only recently been elucidated. 13The method in which color yield and reproducibility of the assay have been optimized is described here. 13 In a screw-capped culture tube add 6.0 ml of acid butanol reagent (950 ml of n-butanol with 50 ml concentrated HCI) to a 1.0-ml aliquot of the sample. Add 0.2 ml of iron reagent [2% FeNH4(SO4) 2 in 2 N HCI] and vortex the sample. Cap the tube loosely and put it in a boiling water bath for 50 min. Cool the tube and read the absorbance at 550 nm. The spectrophotometric blank should contain solvent, acid butanol reagent, and iron reagent. The color yield in the assay is dependent on the stability of the interflavan bond. Bond stability is determined by tannin structure. For example, 5-deoxyproanthocyanidins such as quebracho tannin give substantially less color than simple procyanidins such as Sorghum tannin) Furthermore, the presence of water in the reaction decreases color formation substantially. Samples dissolved in nonaqueous solvents will give the greatest color, but the assay is sensitive enough to give acceptable results even when substantial amounts of water are present if similar amounts of water are included in the standards. 13L. J. Porter, L. N. Hrstich, and B. C. Chan, Phytochemistry 25, 223 (1986).
434
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[42]
If the samples are colored before heating (owing to flower pigments or chlorophyll) the absorbance obtained before heating can be subtracted from the absorbance obtained after heating. For chlorophyll-containing samples this correction is only approximate. Because chlorophyll is partially destroyed during heating, the absorbance obtained before heating is larger than the absorbance due to chlorophyll in the heated sample. Better results can be obtained by adsorbing the polyphenols on polyvinylpyrrolidone before developing the color. 14 If flavan-3,4-diols are present, they can be assayed by initially developing the color at room temperature. 14
Reaction with Substituted Benzaldehydes In the vanillin reaction an aromatic aldehyde such as vanillin reacts with the meta-substituted ring of flavanols to yield a red adduct. Although the vanillin reaction has been widely used to determine condensed tannin, the reaction is not specific for condensed tannins. Any appropriately substituted flavanol reacts in the assay. Thus flavan-3-ols such as catechin and epicatechin also react with vanillin to yield a red-colored adduct. Furthermore, because the vanillin reacts only with meta-substituted flavonoids, condensed tannins composed principally of 5-deoxyflavonoids, such as quebracho tannin, produce little color in the vanillin reaction. The vanillin method described here was devised to eliminate problems with reproducibility. 15 The vanillin reagent must be prepared daily by mixing 1 part of 1% (w/v) vanillin in methanol with 1 part of 8% HC1 in methanol (8 ml concentrated HC1 brought to 100 ml with absolute methanol). A solution of 4% HC1 in methanol is also prepared (4 ml concentrated HC1 brought to 100 ml with absolute methanol). The vanillin reagent and the 4% HC1 solution are brought to 30°, and 1.0-ml aliquots of the samples are dispensed into culture tubes and brought to 30°. Each sample must be run in duplicate, with one of the pair used for the reaction and the other for the background. At exactly 1-min intervals 5.0 ml of the vanillin reagent is added to one set of samples, and 5.0 ml of the 4% HCI solution is added to the second set of samples. The samples are left in the water bath for exactly 20 min, and the absorbance at 500 nm is read. The spectrophotometric blank contains vanillin reagent with sample solvent. The absorbance of each background sample is subtracted from the absorbance of the corresponding reaction tube. The background absorbance is substantial if the samples contain pigments. The vanillin reaction is very sensitive to the presence of water. Even a small amount of water in the reaction mixture will substantially quench i4 j. j. Watterson and L. G. Butler, J. Agric. Food Chem. 31, 41 (1983). t5 M. L. Price, S. Van Scoyoc, and L. G. Butler, J. Agric. Food Chem. 26, 1214 (1978).
[42]
ASSAYS FOR CONDENSED TANNINS
435
color yield. All standards should be prepared in anhydrous organic solvents (usually methanol). If water must be present in the samples to be analyzed, the same amount of water should be added to the standards. Catechin is commonly used to standardize the vanillin reaction, but interpretation of the results ("catechin equivalents") obtained with this standard is difficult. Under the normal conditions for the vanillin assay (methanol solvent), condensed tannins and catechin both react with vanillin, but the rates of reaction of the polymer and the monomer are quite different.~5 In general, the absorbances obtained with catechin are lower than the absorbances obtained with similar amounts of condensed tannin. Use of an alternate solvent to overcome these problems has been described. 16
Protein Precipitation Methods The characteristic reaction of tannins is their ability to precipitate protein. Many of the biological effects of tannins have been attributed to their ability to precipitate protein, but it is not possible at this time to predict biological activity reliably based on any of the many methods available for measuring protein precipitation. Among the factors to be considered when evaluating protein precipitation are the effects of pH, ~7 protein structure, TM tannin structure, ~9 the tannin to protein ratio, 2° and nonprotein modifiers in the reaction mixture (e.g., detergents2~). Of the many methods that have been developed for measuring protein precipitation, we summarize two of the simplest and most generally applicable methods here. Other methods have been compared in a more extensive review.12
Protein Precipitable Phenolics In the following assay, a standard protein is used to precipitate tannin, and the amount of tannin precipitated is assessed using a spectrophotometric method based on the formation of colored iron-phenolate complexes.~7 The method can be combined with methods for determining precipitated protein 2z to evaluate both components of the precipitate. 16 L. G. Butler, M. L. Price, and J. E. Brotherton, J. Agric. Food Chem. 30, 1087 (1982). 17 A. E. Hagerman and L. G. Butler, J. Agric. Food Chem. 26, 809 (1978). 18 A. E. Hagerman and L. G. Butler, J. Biol. Chem. 256, 4494 (1981). 19 T. N. Asquith and L. G. Butler, Phytochemistry 25, 1591 (1986). 2o A. E. Hagerman and C. T. Robbins, J. Chem. Ecol. 13, 1243 (1987). 21 M. M. Martin, D. C. Rockholm, and J. S. Martin, J. Chem. Ecol. 11, 485 (1985). 22 A. E. Hagerman and L. G. Butler, J. Agric. Food Chem. 28, 944 (1980).
436
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[42]
Dispense 2.0 ml of 1 mg/ml bovine serum albumin (BSA) dissolved in buffer (0.20 M acetic acid, 0.17 M NaC1, pH 4.9) into tubes. Up to 1 ml of the tannin-containing sample is added with mixing, and the mixtures are incubated for 15 min at room temperature (purified tannin) or for 24 hr at 4° (plant extracts).2° The samples are centrifuged and the supernatants discarded. The pellets are redissolved by vigorous agitation in 4.0 ml SDS/ TEA [ 1% (w/v) sodium lauryl sulfate, 5% (v/v) triethanolamine]. The color is developed by adding 1.0 ml of 10 mM FeCI 3 in I0 mM HCI and vortexing immediately. After about 15 min the absorbance is recorded at 510 nm. The spectrophotometric blank is a mixture containing reagents but no sample. Even traces of acetone inhibit precipitation of protein by phenolics, 2° so all acetone must be removed from plant extracts before attempting the method.
Radial Diffusion Assay In the radial diffusion assay, the tannin-containing sample diffuses through a protein-containing agar gel, and an opaque ring forms as the tannin precipitates protein in the gel. The area (or diameter squared) of the ring is proportional to the amount of tannin used. The response of the method is dependent in part on the molecular size of the tannin, since diffusion is a function of molecular size. For samples containing chemically similar tannins, the method provides a rapid, simple way to determine tannins, z3 To make eight diffusion plates (suitable for analysis of 32 samples), dissolve 1.0 g agarose [Type I, low EEO, gel point 36° (e.g,, Sigma, St. Louis, MO, A-6013)] in 100 ml of buffer (50 mM acetate containing 60 /zM ascorbic acid, pH 5.0) by heating with continuous stirring. Allow the agarose solution to cool, with occasional stirring, to 45 °, and then add 0.10 g BSA while gently stirring. The protein should be completely dissolved without allowing the solution to cool further. Using a serological pipette with a large opening at the tip, dispense 9.5 ml of solution into 10-cm plastic petri dishes. Allow the solution to gel while the dishes are on a level surface, and then cover each dish and seal with a strip of Parafilm. The plates can be stored at 4° for up to a week before use. Wells are punched in the plates using a 4.0-mm punch (e.g., Bio-Rad, Richmond, CA, 170-4029), with the wells arranged so there are four evenly spaced wells per plate. The plugs of agarose are removed from the wells with gentle suction. Up to 8/xl of sample is then added to each well, and the dishes are resealed and incubated on a level surface at 30° for 96 hr. 23 A. E. H a g e r m a n , J. Chem. Ecol. 13, 437 (1987).
[43]
ROLE OF FLAVONOIDS AND IRON CHELATION
437
A ruler is then used to record the diameter of the ring that has formed. The plates can be stored at 4° for several weeks after development of the rings. The sample can be dissolved in any solvent that is convenient, but we find that aqueous organic mixtures (50% methanol or 70% acetone) are most convenient. If the tannin-containing extract is very dilute, 8/xl may not be sufficient for a response. Larger volumes can be applied to a well by dispensing repetitive 8-/xl samples, but the well must not become completely dry between successive aliquots that are to be added.
[43] R o l e o f F l a v o n o i d s a n d I r o n C h e l a t i o n in A n t i o x i d a n t A c t i o n
By
ISABELLE MOREL, GI~RARD LESCOAT, PIERRE CILLARD,
and JOSIANE CILLARD Introduction The antioxidant effect of flavonoids has been of interest for a considerable time) -4 The potential of flavonoids to inhibit lipid peroxidation in biological models is supposed to reside mainly in their free radical scavenging capacity rather than in their iron chelating activity. 5,6This last property has often been considered as a minor mechanism in the antioxidant action, since it has not been clearly established in biological systems. The assessment of a relationship between the antioxidant effect and the iron chelating capacity of flavonoids is subsequently of interest. 7 For this purpose, we used rat hepatocyte cultures as a biological model where lipid peroxidation was induced by iron [Fe(III)] in its complexed form with nitrilotriacetic
T. F. Slater and N. N. Eakins, in " N e w Trends in the Therapy of Liver Diseases" (A. Bertelli, ed.), p. 84. Karger, Basel, 1975. 2 W. Bors and M. Saran, Free Radical Res. Cornmun. 2, 289 (1987). 3 W. Bors, W. Heller, C. Michel, and M. Saran, this series, Vol. 186, p. 343. 4 j. Torel, J. Cillard, and P. Cillard, Phytochemistry 2,5, 383 (1986). 5 C. G. Fraga, V. S. Martino, G. E. Ferraro, J. F. Coussio, and A. Boveris, Biochem. Pharmacol. 36, 717 (1987). 6 A. K. Ratty and N. P. Das, Biochem. Med. Metab. Biol. 39, 69 (1988). 7 I. Morel, G. Lescoat, J. Cillard, P. Cogrel, O. Sergent, N. Pasdeloup, P. Brissot, and P. Cillard, Biochem. Pharmacol. 45, 13 (1993).
METHODS IN ENZYMOLOGY,VOL.234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
[43]
ROLE OF FLAVONOIDS AND IRON CHELATION
437
A ruler is then used to record the diameter of the ring that has formed. The plates can be stored at 4° for several weeks after development of the rings. The sample can be dissolved in any solvent that is convenient, but we find that aqueous organic mixtures (50% methanol or 70% acetone) are most convenient. If the tannin-containing extract is very dilute, 8/xl may not be sufficient for a response. Larger volumes can be applied to a well by dispensing repetitive 8-/xl samples, but the well must not become completely dry between successive aliquots that are to be added.
[43] R o l e o f F l a v o n o i d s a n d I r o n C h e l a t i o n in A n t i o x i d a n t A c t i o n
By
ISABELLE MOREL, GI~RARD LESCOAT, PIERRE CILLARD,
and JOSIANE CILLARD Introduction The antioxidant effect of flavonoids has been of interest for a considerable time) -4 The potential of flavonoids to inhibit lipid peroxidation in biological models is supposed to reside mainly in their free radical scavenging capacity rather than in their iron chelating activity. 5,6This last property has often been considered as a minor mechanism in the antioxidant action, since it has not been clearly established in biological systems. The assessment of a relationship between the antioxidant effect and the iron chelating capacity of flavonoids is subsequently of interest. 7 For this purpose, we used rat hepatocyte cultures as a biological model where lipid peroxidation was induced by iron [Fe(III)] in its complexed form with nitrilotriacetic
T. F. Slater and N. N. Eakins, in " N e w Trends in the Therapy of Liver Diseases" (A. Bertelli, ed.), p. 84. Karger, Basel, 1975. 2 W. Bors and M. Saran, Free Radical Res. Cornmun. 2, 289 (1987). 3 W. Bors, W. Heller, C. Michel, and M. Saran, this series, Vol. 186, p. 343. 4 j. Torel, J. Cillard, and P. Cillard, Phytochemistry 2,5, 383 (1986). 5 C. G. Fraga, V. S. Martino, G. E. Ferraro, J. F. Coussio, and A. Boveris, Biochem. Pharmacol. 36, 717 (1987). 6 A. K. Ratty and N. P. Das, Biochem. Med. Metab. Biol. 39, 69 (1988). 7 I. Morel, G. Lescoat, J. Cillard, P. Cogrel, O. Sergent, N. Pasdeloup, P. Brissot, and P. Cillard, Biochem. Pharmacol. 45, 13 (1993).
METHODS IN ENZYMOLOGY,VOL.234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
438
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[43]
acid (NTA). The F e - N T A complex is known to induce a rapid accumulation of iron inside the cells. 8-11 Materials and Methods Ferric Nitrilotriacetate Solution
Nitrilotriacetic acid (NTA) is used to maintain ferric iron in a soluble state; it is a low-affinity iron chelator. The ferric nitrilotriacetate solution is prepared according to the following method12: to 47 mg of nitrilotriacetic acid disodium salt (Sigma Chimie, St. Quentin Fallavier, France), dissolved in 10 ml of sterile water, is added 20 mg of ferric ammonium citrate (Merck, Darmstadt, Germany). The final concentration of ferric ion is 10 mM, and the molar ratio ferric iron to NTA is 1 : 2. The solution is sterilized before use by filtration through a 0.22-/xm filter. Flavonoids
Based on solubility, the flavonoid to be tested is dissolved in either water or dimethyl sulfoxide (DMSO). A constant concentration of DMSO is maintained in control samples (2%), a dose that does not affect control rates of lipid peroxidation or iron mobilization. Cell Isolation and Culture
Adult rat hepatocytes are isolated from 2-month-old Sprague-Dawley male rats by cannulating the portal vein and perfusing the liver with a coUagenase solution. 13 The cells are collected in Leibovitz medium containing, per milliliter, 1 mg bovine serum albumin (BSA) and 5/zg bovine insulin. The cell suspension is filtered through gauze and allowed to sediment for 20 rain in order to eliminate cell debris, blood, and sinusoidal cells. The cells are then washed three times by centrifugation at 50 g for 1 min at 4 °, tested for viability, and counted. The hepatocytes are then suspended in a mixture of 75% (v/v) Eagle's minimum essential medium 8 B. R. Bacon, A. S. Tavill, G. M. Brittenham, C. H. Park, and R. O. Recknagel, J. Clin. Invest. 71, 429 (1983), 9 B. Desvergne, G. Baffet, P. Loyer, M. Y. Rissel, G. Lescoat, C. Guguen-Guillouzo, and P. Brissot, Eur. J. Cell Biol. 49, 162 (1989). i0 p. Brissot, J. Farjanel, D. Bourel, J. P. Campion, A. Guillouzo, A. Rattner, Y. Deugnier, B. Desvergne, B. Ferrand, M. Simon, and M. Bourel, Dig. Dis. Sci. 32, 620 (1987). i1 S. Shedlofsky, H. L. Bonkowsky, P. R. Sinclair, W. J. Bement, and J. J. Pomeroy, Biochem. J. 212, 321 (1983). 12 G. P. White and A. Jacobs, Biochim. Biophys. Acta 543, 217 (1978). 13 C. Guguen, A. Guillouzo, M. Boisnard, A. Le Cam, and M. Bourel, Biol. Gastroenterol. 8, 223 (1975).
[43]
ROLE OF FLAVONOIDS AND IRON CHELATION
439
and 25% medium 199, supplemented with 10% fetal calf serum (FCS) and containing, per milliliter, the following: streptomycin (50/~g), penicillin (7.5 IU), bovine insulin (5/~g), bovine serum albumin (1 mg), and NaHCO3 (2.2 mg). Usually, according to the experimental procedure, either 2.5 x 10 6 hepatocytes are plated in 25-cm 2 Nunclon flasks (Roskilde, Denmark), or 1.5 x 105 hepatocytes are suspended in I ml of medium in multiwell tissue culture plates. The medium is changed 3-4 hr later and renewed the day after with the same medium as above but without serum and supplemented with 10 -7 M dexamethasone.
Lipid Peroxidation Chromatography Procedure. Free malondialdehyde (MDA) quantification ~4 is performed using a high-performance liquid chromatography (HPLC) system [Laboratory Data Control (LDC) Finnigan, Orsay, France] which is equipped with a Spherogel-TSK G 1000 PW size-exclusion column (7.5 mm i.d. x 30 cm, Cluzeau, Ste. Foy, France). This sizeexclusion column separates compounds using a gel-permeation technique which allows a diffusion of the smallest molecules such as MDA inside the gel and then permits their separation from larger compounds which are eluted earlier. The retention time (56 min for MDA) is, however, greatly increased in comparison with classic C~8 columns. The use of this size-exclusion column is nevertheless highly recommended, especially in biological systems where many interferences may occur. The eluant is composed of 0.1 M disodium phosphate buffer, pH 8, at a flow rate of 1 ml/min. The absorbance of free MDA is monitored at 267 nm, using the maximal sensitivity (0.001 absorbance unit full scale). The injections (250 ~1) are performed by an autosampling injector (Promis, LDC) and the data are recorded and treated using a chromatography software (Thermochrom, LDC). Preparation of Free MDA Standard. Five microliters of 1,1,3,3-tetramethoxypropane (Sigma) are hydrolyzed in 5 ml of 0.1 N HC1 during 5 min in boiling water. This is performed in a closed vial in order to prevent evaporation. The solution is then diluted 1/ 1000 in 10 mM Na2HPO4 buffer, pH 7.45, which corresponds to a final concentration of 6/~M of MDA. Because the stability of diluted solutions of MDA is very weak, a new solution must be prepared daily. The concentration of MDA in samples is calculated using a standard curve of free MDA. Preparation of Samples for Analysis. During the first day of experimentation (day 1), the cultures are maintained in the presence of ferric 14 A. S. Csallany, D. M. Guan, J. D. Manwaring, and P. B. Addis, Anal. Biochem. 142, 277 (1984).
440
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[43]
iron nitrilotriacetate (Fe-NTA) in order to obtain final iron concentrations of I00/.,M. More precisely, each sample with F e - N T A is compared to control cultures without any supplementation and to cultures supplemented with 200/zM NTA, which is the amount of NTA added to the cultures in the experimental procedure. After 24 hr of incubation, the medium is renewed with the same medium supplemented or not with iron with or without DMSO, and the flavonoids are added. After another 24-hr period of incubation (day 2), culture media are collected and hepatocytes are washed twice with 10 mM phosphate buffer, pH 7.45. They are resuspended in 1 ml of the same buffer. The cells are lysed using an ultrasonic homogenizer during 1 min in ice. An aliquot was stored at - 1 7 ° until protein content is estimated. The samples (culture media or cell homogenates) are filtered through a 500-Da membrane ultrafilter (Amicon, Danvers, MA) in a 10-ml Amicon cell pressurized at 4 bars with nitrogen gas. The filtrate is used for the HPLC procedure. Free MDA is quantified separately in culture media and in cell homogenates (see Fig. 1).
MDA
MDA
A
0
i6
B
6
i6
FIG. 1. Chromatograms of free MDA in hepatocytes (A) and in the medium (B) of cultures which have been supplemented with F e - N T A (100/zM) for 48 hr. The size-exclusion HPLC system allowed good separation of free MDA (retention time 56 min) from other biological components.
[43]
ROLE OF FLAVONOIDS AND IRON CHELATION
441
The results may be expressed in nanograms per milligram of protein for free MDA in the cells and in nanograms per milliliter of culture medium for extracellular free MDA. For convenience, they can also be expressed as total free MDA (nanomolar) present in each culture sample (MDA in the cells plus MDA in the culture medium).~5 However the absolute level of free MDA in biological models may vary from one experiment to another, especially when the cells are provided by different animals. Moreover, variations in culture medium composition or in incubation times before treatment of the cells may induce considerable modifications in MDA levels. Thus, the standardization of culture conditions is of major importance to obtain reproducible values. In the hepatocyte culture model where all precautions have been taken, absolute values for total free MDA levels are 1724 --- 32 nM in cultures supplemented only with 100 ~M F e - N T A , whereas the basal level of free MDA in hepatocyte cultures without supplementation is 37.3 - 3.6 nM. To avoid experimental variations, total free MDA may also be expressed as a percentage of control values. In our experiments, where the antioxidant activity was investigated, 100% of MDA recovery corresponded to MDA level in cultures supplemented with iron only. 16 Another source of variation in experimental values is the weak stability of free MDA, which results in a loss of MDA in the samples. Stability experiments have shown that the samples must be subjected to ultrafiltration as soon as possible to stop lipid peroxidation and to avoid MDA linkage to biological substrates such as proteins or DNA, and that MDA level in the ultrafiltrates has to be evaluated within 24 hr, unless samples are frozen in liquid nitrogen. When working with frozen samples, two main indications should be kept in mind: first, ultrafiltration is necessary before freezing since MDA levels decrease in nonultrafiltrated frozen samples; second, a standard solution of MDA must be frozen simultaneously to prevent any variations resulting from a small loss in the samples. However, direct analysis of MDA on fresh ultrafiltrates remains the best and the simplest way to obtain accurate results (Fig. 2). Protein content is determined on thawed cell homogenates according to the Bradford reaction, 17performed on a Cobas-Bio automatic analyzer (Roche, Neuilly/Seine, France) using the Bio-Rad protein assay reagent diluted with phosphate-buffered saline (1 : 4, v/v) and using bovine serum albumin as a standard (0-700/xg/ml). 15 I. Morel, G. Lescoat, J. Cillard, N. Pasdeloup, P. Brissot, and P. CiUard, Biochem. Pharmacol. 39, 1647 (1990). 16 I. Morel, J. Cillard, G. Lescoat, O. Sergent, N. Pasdeloup, A. Z. Ocaktan, M. A. Abdallah, P. Brissot, and P. Cillard, Free Radical Biol. Med. 13, 499 (1992). iv M. M. Bradford, Anal. Biochem. 72, 248 (1976).
442
[43]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
<
[-0
Hosmetin "cetin I
400 FIG. 2. Antioxidant activity of flavonoids at various concentrations in biological s y s t e m . Total free M D A in hepatocyte cultures was e x p r e s s e d as a percentage o f control cultures s u p p l e m e n t e d with iron (100/xM F e - N T A ) for 48 hr. In the samples, the addition of flavonoids was performed 24 hr after iron loading, and the following incubation for 24 hr was carried out in the p r e s e n c e of iron (100/~M) and a flavonoid.
TABLE I IRON MOBILIZATION BY FLAVONOIDS a 55Fe
Sample Control (55Fe) Catechin Quercetin Diosmetin Desferal
Intracellular (cpm//xg o f proteins) 233 107 184 238 32
+ + + -+ +
21 5 17 21 3
Extracellular (cpm/ml) 710 2594 938 761 5697
-+ + + + +
63 498 62 74 413
" T h e hepatocyte cultures were loaded with 55Fe (1 /~M) on day 1 and were treated for the n e x t 2 days with (or without) flavonoid or desferrioxamine (Desferal).
[44]
FLAVONOIDS AND COUMARINS ON EICOSANOIDS
443
Iron-55 Mobilization from Iron-Loaded Hepatocyte Cultures To compare the chelating activity of the flavonoids in cell cultures, the hepatocytes in multiwell tissue culture plates (well area 1.7 cm 2) are loaded during 24 hr with 1/xM [55Fe]ferric chloride (specific activity 1.5 mCi/mg Fe; Radiochemical Center, Amersham, UK). TM Culture medium is then renewed (day 1) with the same medium but without iron and with or without I00/zM of flavonoid or desferrioxamine B (Desferal) serving as a reference for iron chelation. After 2 days of incubation (day 3), which is the minimal time required to obtain significant values, the concentrations of 55Fe are determined extracellularly and intracellularly using a/3 counter. Extracellular radioactivity is counted directly in 1 ml of culture medium, whereas for counting intracellular activity, the cells must first be rinsed with phosphate-buffered saline and then sonicated in 1 ml of the same buffer. Control cultures are supplemented during 24 hr with iron and the following 2 days with or without DMSO, which is used in the samples to dissolve the flavonoids. The results shown in Table I are expressed in counts per minute (cpm) per milliliter of medium or microgram of protein. Conclusion The establishment of a relationship between the antioxidant activity of flavonoids and their iron chelating capacity is of major importance in explaining their mechanism of action. These investigations should be undertaken with other antioxidants presenting structural characteristics allowing iron chelation, such as polyphenolic structures. 18 p. Jego, N. H u b e r t , I. M o r e l , N. P a s d e l o u p , A. O c a k t a n , M. A b d a l l a h , P. B r i s s o t , a n d G. L e s c o a t , Biochem. Pharmacol. 43, 1275 (1992).
[44] A c t i o n s o f F l a v o n o i d s a n d C o u m a r i n s on L i p o x y g e n a s e and Cyclooxygenase
By J. R. S. HOULT, MICHELE A. MORONEY, and MIGUEL PAY/~ Introduction Flavonoids and coumarins are polyphenolic compounds containing multiple substituents on the benzo-y-pyrone nucleus (Tables I and II) and are widely distributed in the plant kingdom. They have low mammalian toxicity and are present in large amounts in the diet of humans because METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[44]
FLAVONOIDS AND COUMARINS ON EICOSANOIDS
443
Iron-55 Mobilization from Iron-Loaded Hepatocyte Cultures To compare the chelating activity of the flavonoids in cell cultures, the hepatocytes in multiwell tissue culture plates (well area 1.7 cm 2) are loaded during 24 hr with 1/xM [55Fe]ferric chloride (specific activity 1.5 mCi/mg Fe; Radiochemical Center, Amersham, UK). TM Culture medium is then renewed (day 1) with the same medium but without iron and with or without I00/zM of flavonoid or desferrioxamine B (Desferal) serving as a reference for iron chelation. After 2 days of incubation (day 3), which is the minimal time required to obtain significant values, the concentrations of 55Fe are determined extracellularly and intracellularly using a/3 counter. Extracellular radioactivity is counted directly in 1 ml of culture medium, whereas for counting intracellular activity, the cells must first be rinsed with phosphate-buffered saline and then sonicated in 1 ml of the same buffer. Control cultures are supplemented during 24 hr with iron and the following 2 days with or without DMSO, which is used in the samples to dissolve the flavonoids. The results shown in Table I are expressed in counts per minute (cpm) per milliliter of medium or microgram of protein. Conclusion The establishment of a relationship between the antioxidant activity of flavonoids and their iron chelating capacity is of major importance in explaining their mechanism of action. These investigations should be undertaken with other antioxidants presenting structural characteristics allowing iron chelation, such as polyphenolic structures. 18 p. Jego, N. H u b e r t , I. M o r e l , N. P a s d e l o u p , A. O c a k t a n , M. A b d a l l a h , P. B r i s s o t , a n d G. L e s c o a t , Biochem. Pharmacol. 43, 1275 (1992).
[44] A c t i o n s o f F l a v o n o i d s a n d C o u m a r i n s on L i p o x y g e n a s e and Cyclooxygenase
By J. R. S. HOULT, MICHELE A. MORONEY, and MIGUEL PAY/~ Introduction Flavonoids and coumarins are polyphenolic compounds containing multiple substituents on the benzo-y-pyrone nucleus (Tables I and II) and are widely distributed in the plant kingdom. They have low mammalian toxicity and are present in large amounts in the diet of humans because METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
444
ANTIOXIDANT
CHARACTERIZATION
AND
[44]
ASSAY
~D cD
t:h
2 m
0
[.... <
e..
A
z Z
e
0
Fu~ Z in
.)
0
Z <
~5
a~ 0
~5
<
r..)
z Z <
i~llll :0~ =© : =0= 0=
I
0
io=o=fo=i
o
~lllll
©
..q
_=
0
> o
0 0 0 0 0 0
Z @ > < M
o=o= io=o=o=
0;$
E
m, L.)
~ N N N N N N 0 0 0 0 0 0 0
e-~ e.. 0~
Q
.1 < N m
o
£.)
o r..)
O
t-, •~ •
~ .~
"~ .= ~ ~
o ,.~
t;x0
[44]
FLAVONOIDS
AND
COUMARINS
. 0~
ON
A
445
EICOSANOIDS
AA
0
=
o
AA
A
,I Z
o u.l Z u~
d
===
==
0 0 0 0
0
15
O0
t
II
o~ <
Z <
.o
E
< 0
;>
0
Ira l-
~=' ~
~= 0
.~
~.~ .o~
.
<
~844~84
0
, 0 ~ 0 0 0 0
v
446
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[44]
common foods of plant origin contain from trace quantities to several grams per kilogram fresh weight. More than 4000 such compounds have been identified and are responsible for much of the coloring in nature, although it is not clear whether they have precise functions in plants. 1,2 Reviews have emphasized that flavonoids and coumarins possess numerous biological and pharmacological properties, some of potential therapeutic interest. 3-s Indeed various flavonoids and coumarins which possess anti-inflammatory actions in bioassays have been identified in extracts prepared from plants used for this purpose in traditional folk medicines .6-9 In biochemical terms, many of the plant-derived polyphenols can act as antioxidants, 1° and this is considered elsewhere in this volume. Those flavonoids and coumarins containing o-dihydroxy functions might also be expected to chelate transition metal ions and/or be able to reduce them, and thus to be potentially prooxidant. In fact, flavonoids and coumarins have been shown both to scavenge reactive oxygen species and to promote their formation, depending on experimental conditions. 11-13 The interactions of these natural compounds with reactive oxygen species and with inflammation has also prompted many studies on their effects on the formation of proinflammatory eicosanoids derived from the cyclooxygenase and lipoxygenase pathways of arachidonate m e t a b o l i s m ) 2,14,15 This is because both these enzymes catalyze controlled stereospecific free radical 1 j. B. Harborne, ed., "The Flavonoids: Advances in Research Since 1980." Chapman & Hall, London, 1988. z R. D. H. Murray, J. M6ndez, and R. A. Brown, "The Natural Coumarins." Wiley, New York, 1982. 3 B. Havsteen, Biochem. Pharmacol. 32, 1141 (1983). 4 E. Middleton, Prog. Clin. Biol. Res. 213, 493 (1986). 5 D. Egan, R. O'Kennedy, E. Moran, D. Cox, E. Prosser, and R. D. Thornes, Drug Metab. Rev. 22, 503 (1990). 6 A. Villar, M. A. Gasc6, and M. J. Alcaraz, J. Pharm. Pharmacol. 36, 820 (1984). 7 H. Otsuka, S. Fujioka, T. Komiya, E. Mizuta, and M. Takamoto, Yakugaku Zasshi 102, 162 (1982). 8 A. Della Loggia, A. Tubaro, P. Dri, C. Zilli, and P. del Negro, Prog. Clin. Biol. Res. 213, 481 (1986). 9 K. Sekiya, H. Okuda, and S. Arichi, Biochim. Biophys. Acta 713, 68 (1982). l0 R. A. Larson, Phytochemistry 27, 969 (1988). ii M. J. Laughton, B. Halliwell, P. J. Evans, and J. R. S. Hoult, Biochem. Pharmacol. 38, 2859 (1989). 12 M. J. Laughton, P. J. Evans, M. A. Moroney, J. R. S. Hoult, and B. HaUiwell, Biochem. Pharmacol. 42, 1673 (1991). 13 M. Payfi, B. Halliwell, and J. R. S. Hoult, Biochem. Pharmacol. 44, 205 (1992). 14 M. J. Alcaraz and M. L. Ferrfindiz, J. Ethnopharmacol. 21, 209 (1987). I5 M.-A. Moroney, M. J. Alcaraz, R. A. Forder, F. Carey, and J. R. S. Hoult, J. Pharm. Pharmacol. 40, 787 (1988).
[44]
FLAVONOIDS AND COUMARINS ON EICOSANOIDS
447
peroxidation of arachidonic acid at their active sites, and polyphenols might be expected to interfere with this process. This chapter describes methods used for the analysis of the actions of flavonoids and coumarins on the generation of eicosanoids by stimulated leukocytes. These cells are well suited for this purpose as they readily generate eicosanoids after engulfing opsonized particles or as a result of activation by soluble stimulants such as calcium ionophores, chemotactic peptides and arachidonic acid itself. Moreover, they are directly involved in the acute inflammatory process. Both rat mixed peritoneal leukocytes and human polymorphonuclear neutrophil leukocytes (PMNs) have been used, although both cell types yield essentially similar results. The eicosanoid products of arachidonate metabolism are measured by radioimmunoassay or by radiometric (radio-TLC) methods, both of which are described here. Methods
Preparation of Mixed Leukocytes from Rat Peritoneal Cavity: Induction of Peritonitis Mixed peritoneal leukocytes are elicited from female Wistar rats by an intraperitoneal injection of 10 ml of a solution of 6% oyster glycogen in saline. This induces a leukocyte-rich acute peritonitis. As an alternative, methylcellulose (10 ml of a 1% solution) or casein (10 ml of a 2% solution) may also be used. Sixteen to twenty hours later the rats are sacrificed by cervical dislocation, and 40 to 60 ml ice-cold modified Hanks' balanced salt solution (HBSS) free of C a 2+ and Mg 2+ is injected into the peritoneal cavity. After about 90 sec vigorous massage, the peritoneal washing is removed and centrifuged at 4° in polypropylene tubes for 10 min at 800 g (2100 rpm, MSE Coolspin, Fisons, Loughborough, U.K.), and the contaminating erythrocytes are lysed. This is done by resuspending the pelleted cells in a small volume (2 to 5 ml) of HBSS and adding 9 volumes of isotonic Tris-buffered ammonium chloride (0.83%, pH 7.2) for 10 min at 37 °. After further centrifugation, the cells are resuspended in 10 ml modified HBSS and a portion taken for counting using a hemocytometer. After adding further modified HBSS and a final centrifugation, the cells are resuspended at the desired density (generally 2.5 x 106 or 5 X 106 cells/ ml) in HBSS containing 1.26 mM Ca 2+ and 0.9 mM Mg 2+. Cell smears are prepared using a Shandon Cytospin 2 (Life Sciences UK, Basingstoke, U.K.) according to conventional procedures with successive staining with hematoxylin followed by Chromotrope 2R. This shows that 70-80% of the cells are PMN neutrophils, the remainder being
448
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[44]
mainly monocytes. For example, two rats injected with glycogen as above yielded 2.42 x 108 cells, of which 79.1% were PMNs, 19.6% monocytes, and 1.3% eosinophils as determined by this staining procedure. Cell viability based on trypan blue exclusion is greater than 95%. Cells may be stored for up to 3 hr on ice without loss of function. However, viability declines more rapidly if the cells are kept in HBSS at 37 °. If desired, pure PMN populations may be prepared by resuspending the mixed cell suspension in 2 to 5 ml modified HBSS after the red cell lysis stage and loading the cells onto 5 ml of Histopaque (Sigma Chemical Company) 1077 Or 1083. Samples are then centrifuged for 30 min at room temperature at 400 g, with monocytes sedimenting to the interface and neutrophils to the bottom of the tube. The pellet should be resuspended in 5 to 10 ml modified HBSS and centrifugation/resuspension cycles repeated at least twice to remove the Histopaque. Smears of cells prepared according to this method show over 95% PMNs. As a general precaution, it is advisable to avoid contact of rat or human leukocytes with glass surfaces, especially when dispensing from syringes.
Preparation of Suspensions of Human Polymorphonuclear Leukocytes Blood from healthy nonmedicated adult donors is drawn by venipuncture of the antecubital vein using a 19-gauge butterfly cannula into a sterile 60-ml polypropylene syringe and dispensed into one-tenth volume of 3.15% (w/v) trisodium citrate, ensuring satisfactory mixing. Volunteers must give informed consent, and ethics committee guidelines should be followed. The anticoagulated blood is then pooled in 50-ml polycarbonate tubes and centrifuged at 200 g (1500 rpm, MSE Chilspin 2) for 15 min at room temperature. The upper platelet-rich layer is removed, and the residual blood is combined with an equal volume of dextran (prepared by dissolving 6 g of dextran T-500 plus 2.7 g of NaCI in 300 ml distilled water). After several inversions to ensure mixing, the blood is left at room temperature for 45-60 min to permit the erythrocytes to sediment. The upper PMN-rich phase is then collected and concentrated by centrifugation at 200 g for 15 min at room temperature. Contaminating erythrocytes may be removed by hypotonic lysis by resuspending the cells in 27 ml icecold distilled water for 20 sec followed by the addition of 3 ml of 10 x concentrated modified HBSS (free of Ca 2+ and Mg 2÷) to restore tonicity and a repeat centrifugation. Care must be taken not to exceed this incubation time unduly, as loss of leukocyte viability and reduction of yield may result. The cell pellets are gently resuspended in 10 ml ice-cold modified HBSS, and a solution of Histopaque 1083 is then carefully layered under
[44]
FLAVONOIDS AND COUMARINS ON EICOSANOIDS
449
the cell suspension to form a discontinuous gradient. This is achieved by dispensing 5 ml through a needle placed to the bottom of the centrifuge tubes, which should then be spun for 40 min at 400 g at room temperature. Neutrophils sediment to the bottom. After two cycles of washing and centrifugation, the cell pellets are finally resuspended in HBSS containing 1.26 mM Ca 2÷ and 0.42 mM Mg 2÷, so as to achieve a concentration of 2.5 × 106 or 5 × 106 cells/ml. These preparations contain more than 95% PMN, and viability is greater than 95%, as established by exclusion of trypan blue. Cells should be kept at room temperature for not longer than 2-3 hr.
Stimulation of Release of Eicosanoids and Radioimmunoassay Triplicate aliquots of 0.5 ml leukocytes at 2.5 × 106 or 5 x 106 cells/ ml are placed in 3.0-ml round-bottomed polypropylene tubes and preincubated at 37° for 10 min with the compound of interest or its vehicle (generally methanol or dimethyl sulfoxide), which should be added in 5/xl or less. To the cell suspension is then added 2.5 or 5 /xl of stimulant to activate eicosanoid biosynthesis from endogenous precursors. In the majority of our more recent experiments we have used the nonphysiological calcium ionophore A23187 (final concentration 1/zM), dissolved in dimethyl sulfoxide (equivalent volumes of vehicle should be added to "basal" unstimulated control tubes lacking ionophore), for a further 10 min of incubation. The cells are then pelleted by centrifugation at 1500 g for 10 min at 4 ° (2500 rpm, MSE Coolspin), and the supernatants are decanted and subjected to radioimmunoassay (RIA) or else kept frozen at - 20°. Aliquots of 2-25/~1 of the thawed samples are subjected to radioimmunoassay for prostaglandins (PG) PGE 2, PGF2~, 6-keto-PGFl~, thromboxane B2 (TXB2), or leukotriene B 4 (LTB4) as described previously) 5 The assays can be performed directly on the supernatants without need for time-consuming extraction of the eicosanoids. A detailed review of the procedures for obtaining antibodies to eicosanoids and for performing immunoassays using charcoal separation methods or double-antibody precipitation is available) 6 Necessary precautions for the validation ofradioimmunoassays should be taken. These include the following: (1) testing in each assay one or more "external standard solutions" kept for this purpose (to check the consistency of the assay from day to day); (2) verifying that the drugs used as inhibitors do not interfere with the assay (either add appropriate 16 E. Granstr6m and H. Kindahl, Adv. Prostaglandin Thromboxane Res. 5, 119 (1978).
450
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[44]
amounts of the drug to known standards used for the calibration curve, or add the drug to stimulated cells after the incubation and compare results with effects observed if it is added before); and (3) checking the identity of the displacing ligand present in the samples with that of the known standards using dilution curves (perform tests at several dilutions on a selection of samples and check for "parallelism" with the standard curve).
Measurement of Eicosanoid Metabolism by Radiochromatography using [14C]Arachidonic Acid as Substrate Aliquots of 0.5 ml leukocytes at 2 x l 0 7 cells/ml are placed in 7-ml polypropylene tubes and preincubated with drugs or vehicle as described in the previous section. After this, 5-10/zl of stimulant (generally A23187) or its vehicle is added together with 5-10/~1 of [14C]arachidonic acid (e.g., Amersham International, Aylesbury, UK; specific activity 58 mCi/mmol), containing 0.125/zCi (yielding 0.66 ~g substrate or 4.3 ~M at this specific activity) and incubated for 10 min. The eicosanoids generated from the arachidonate require extraction from the incubation medium and chromatographic separation using thin-layer chromatography (TLC). This is achieved by adding 10 /xl of 0.5 M citric acid to bring the pH to 3.5, followed by extraction into two successive portions of 1.0 ml ethyl acetate. The organic fractions are transferred to 3-ml flat-bottomed glass vials and the solvent removed in an air stream after placing the vials in a multiplace heating block set at 37°. The residues are taken up in 15 to 25/~1 methanol and spotted quantitatively using capillary tube droppers (e.g., Drummond microcaps) or Hamilton microsyringes onto flexible 20 x 20 cm silica gelcoated TLC plates. Thirteen 1.5-cm wide bands are drawn in pencil on the sheets for this purpose, with the origin placed 2 cm from the bottom. Known standards are applied to the outer lanes. Careful preliminary testing is needed in order to find out what kind of extraction conditions, chromatography solvent, and TLC plates are optimal.17 For eicosanoid generation from arachidonic acid in leukocytes we have chosen Merck (Darmstadt, Germany) aluminum foil-backed sheets, type 5554 (layer thickness 0.2 mm), and the solvent is ethyl acetate-formic acid (80 : 1, v/v), as set out in Moroney et al.~5 After 60 min of development time in a preequilibrated glass chamber, there is about 15 cm solvent migration, affording very satisfactory separation of all the major leukocyte eicosanoid metabolites. This is shown in Fig. 1 for an experiment using ionophore-stimulated rat peritoneal leukocytes. After chromatography, the positions of radioactive peaks may be verified by 17 M.-A. Moroney, Ph.D. Thesis, London University (1992).
[44]
451
FLAVONOIDS AND COUMARINS ON EICOSANOIDS
B
A
!
1
d~
D
0
1 J to
IJ
U to
t6
FIG. 1. Radiochromatograms of arachidonate metabolites in ionophore-stimulated rat peritoneal neutrophils. In (A) the cells were incubated with arachidonate alone, whereas 1 /~M A23187 was also added to (B), (C), and (D). Part (B) shows the spectrum of metabolites formed. In (C) the cells were pretreated with 10 /xM indomethacin (which reduces all cyclooxygenase metabolites), whereas in (D) the cells were pretreated with the specific thromboxane synthetase inhibitor dazoxiben at 5/xM.
452
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[44]
scanning the plates using a Panax Model RTLS-1A radiochromatogram scanner, and the relevant portions of the chromatogram are sectioned and counted for radioactivity. Five to 25/~g quantities of the reference standards are applied to the outer bands and may be localized by scanning (for radioactive standards) or visualized under UV light [e.g., 5-hydroxyeicosatetraenoic acid (5-HETE)], or by cutting off the lanes and placing them in an iodine chamber for 2 to l0 min (e.g., for the cyclooxygenase products). Typical Rf values obtained under these conditions are as follows: arachidonic acid (AA), 0.68; 5-HETE, 0.57; LTB4, 0.46; TXB 2, 0.28; PGE 2, 0.23; 6-keto-PGFl~, 0.12; with phospholipids (PL) at the origin and di-/tri-glycerides (D/TG) at 0.78. An alternative TLC sheet which is satisfactory for many other applications involving separation of prostaglandins is the type 13181 plasticbacked sheet from Kodak (Rochester, NY) (layer thickness 0.1 mm), although solvent migration is not so rapid. Nevertheless, it should be noted that there are discernible differences not only in the absolute R e values but also in the relative positions of different prostaglandins when their chromatographic characteristics are compared using the same solvent on the two kinds of TLC sheets.~S These differences can be exploited to advantage for confirming substance identity in experiments using TLC on the two kinds of sheets but in the same solvent. Moreover, both sheets can be cut up into small sections and the pieces transferred directly into scintillation vials for counting without need for scraping off the silica gel. This improves sample throughput, reduces labor, and, importantly, allows each lane to be cut up easily into many small segments, thus improving resolution and allowing multiple metabolites to be quantitated accurately. It should be noted that for the leukocyte experiments, alternative sample acidification and extraction procedures used by other authors were tested and rejected.17 These included the use of larger solvent volumes, formic acid rather than citric acid, and ether rather than ethyl acetate. However, the method described here appeared to be optimal in terms of safety, practicability, and ease of application as well as in terms of efficiency of recovery. The use of tracer eicosanoids showed that sample recovery through the extraction procedure was about 90% for all of the eicosanoid products tested, as well as for phospholipids and di-triglycerides. Comments These techniques have been used to investigate the inhibitory effects of flavonoids and coumarins on the enzymes of arachidonate metabolism. Is j. R. S. Hoult, K. B. Bacon, D. J. Osborne, and C. Robinson, Biochem. Pharrnacol. 37, 3591 (1988).
[44]
FLAVONOIDS AND COUMARINS ON EICOSANOIDS
453
Representative data for the effects of the plant-derived polyphenols are given in Tables I and II, and an example of the profile of arachidonic acid metabolites generated by ionophore-stimulated rat mixed peritoneal leukocytes is shown in Fig. 1. It is clear from this profile that the rat leukocytes generate metabolites via both the 5-1ipoxygenase and cyclooxygenase pathways and thus are a useful and simple model for studying the effects of drugs on these enzymes and their regulation. 15'19Human PMNs generate similar amounts of 5-1ipoxygenase products but appear to express cyclooxygenase at only very low activity, and they are therefore less useful for drug inhibition studies. The results obtained using the radiochromatography method have been substantitated by data obtained using specific radioimmunoassays with each of the metabolic products. This shows, for example, that TXB2 is the major cyclooxygenase product (greater than PGE2 or 6-keto-PGFl~). The radioimmunoassay approach is clearly the preferred method for inhibitor screening studies on account of its much higher sample throughput capacity. It also does not require addition of arachidonic acid, which may itself perturb the system) 5 However, validation using radiochromatography is desirable. The results of the tests on flavonoids and coumarins confirm that both types of compounds are capable of inhibiting the 5-1ipoxygenase and cyclooxygenase enzymes, as described by us and several other authors) 4,15'2° However, the structural features required for inhibition are not the same: 5-1ipoxygenase inhibition is optimal in polyhydroxylated compounds, especially if they possess a vicinal diol (Tables I and II) and an additional lipophilic substituent elsewhere (Table 1I). 20 It appears that 5-1ipoxygenase inhibition may depend on a combination of iron ion-reducing and iron ion-chelating abilities, and is not dependent on lipid peroxyl scavenging.12 These properties are optimal in the polyhydroxylated and potentially prooxidant catechols such as quercetin, myricetin, daphnetin, and esculetin. 11-13Cyclooxygenase inhibition does not appear to be dependent on these aspects of the polyphenol molecular structure, and the determinants for this remain to be firmly established) 4:5 In conclusion, the ability of dietary polyphenolic compounds to inhibit the conversion of arachidonic acid to proinflammatory eicosanoids may be of therapeutic interest. Work has stressed the anti-inflammatory activity of various purified plant-derived phenolics, and the finding that this may depend on reducing eicosanoid generation at the site of inflammation. 21 19 M. A. Moroney, R. A. Forder, F. Carey, and J. R. S. Hoult, Br. J. Pharmaeol. 101, 128 (1990). 2o T. Horie, M. Tsukayama, H. Kourai, C. Yokoyama, M, Furukawa, T. Yoshimoto, S. Yamamoto, S. Watanabe-Kohno, and K. Ohata, J. Med. Chem. 29, 2256 (1986). 21 M. L. Ferr~ndiz and M. J. Alcaraz, Agents Actions 32, 283 (1991).
454
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[45]
However, it has not always proved possible to correlate this directly with in oitro potency. 14,21.22This may reflect pharmacokinetic aspects including metabolism, or the fact that other properties are also important in vivo.
This was emphasized recently in an important epidemiological study showing correlation between dietary consumption of flavonoids and protection against coronary heart disease mortality and myocardial infarction in elderly Dutch men.23The authors suggested that the flavonoids might act through both antioxidant and eicosanoid inhibitory mechanisms. Further laboratory and clinical studies are thus needed to determine whether plant polyphenolics have beneficial actions on human health. Acknowledgments M.P. thanks the Spanish Ministry of Education and Science for a traveling Research Fellowship. 22 R. M. McMillan, A. J. Millest, K. E. Proudman, and K. B. Taylor, Br. J. Pharmacol. 87, 53P (1986). 23 M. G. L. Hertog, E. J. M. Feskens, P. C. H. Hollman, M. B. Katan, and D. Kromhout, Lancet 342, 1007 (1993).
[45] D e t e r m i n a t i o n of S t r u c t u r e - A n t i o x i d a n t A c t i v i t y Relationships of Dihydrolipoic Acid B y YUICHIRO J . SUZUKI, MASAHIKO TSUCHIYA, a n d LESTER PACKER
Introduction
a-Lipoic acid (thioctic acid; 6,8-dithio-n-octanoic acid; molecular weight 206.32) is an essential cofactor for metabolism in a-keto acid dehydrogenase reactions. This vitamin-like substance has been supplemented orally for health benefits and also has been used as a therapeutic agent in a variety of diseases including liver and neurological disorders as well as mushroom poisoning. Patients diagnosed with diabetes mellitus and atherosclerosis have also been found to contain a decreased level of endogenous a-lipoic acid, and possible therapeutic use of a-lipoic acid in these diseases has been considered. In physiological systems, a-lipoic acid usually exists as lipoamide covalently bound to a lysine residue of the enzyme complexes, and it functions in the transfer of the two-carbon fragment resulting from decarboxylation of pyruvate from a-hydroxyethylthiamin pyrophosphate to acetyl-CoA, itself being reduced in the process. The METHODSIN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
454
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[45]
However, it has not always proved possible to correlate this directly with in oitro potency. 14,21.22This may reflect pharmacokinetic aspects including metabolism, or the fact that other properties are also important in vivo.
This was emphasized recently in an important epidemiological study showing correlation between dietary consumption of flavonoids and protection against coronary heart disease mortality and myocardial infarction in elderly Dutch men.23The authors suggested that the flavonoids might act through both antioxidant and eicosanoid inhibitory mechanisms. Further laboratory and clinical studies are thus needed to determine whether plant polyphenolics have beneficial actions on human health. Acknowledgments M.P. thanks the Spanish Ministry of Education and Science for a traveling Research Fellowship. 22 R. M. McMillan, A. J. Millest, K. E. Proudman, and K. B. Taylor, Br. J. Pharmacol. 87, 53P (1986). 23 M. G. L. Hertog, E. J. M. Feskens, P. C. H. Hollman, M. B. Katan, and D. Kromhout, Lancet 342, 1007 (1993).
[45] D e t e r m i n a t i o n of S t r u c t u r e - A n t i o x i d a n t A c t i v i t y Relationships of Dihydrolipoic Acid B y YUICHIRO J . SUZUKI, MASAHIKO TSUCHIYA, a n d LESTER PACKER
Introduction
a-Lipoic acid (thioctic acid; 6,8-dithio-n-octanoic acid; molecular weight 206.32) is an essential cofactor for metabolism in a-keto acid dehydrogenase reactions. This vitamin-like substance has been supplemented orally for health benefits and also has been used as a therapeutic agent in a variety of diseases including liver and neurological disorders as well as mushroom poisoning. Patients diagnosed with diabetes mellitus and atherosclerosis have also been found to contain a decreased level of endogenous a-lipoic acid, and possible therapeutic use of a-lipoic acid in these diseases has been considered. In physiological systems, a-lipoic acid usually exists as lipoamide covalently bound to a lysine residue of the enzyme complexes, and it functions in the transfer of the two-carbon fragment resulting from decarboxylation of pyruvate from a-hydroxyethylthiamin pyrophosphate to acetyl-CoA, itself being reduced in the process. The METHODSIN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
[45]
DIHYDROLIPOICACID REDUCED FORM OF ~.-LIPOIC AICD
~ HS
(dihydrolipoic acid,
455 DHLA)
COOH SH
REDUCED FORM OF BISNORLIPOIC ACID
~ HS
-~COOH SH
REDUCED FORM OF TETRANORLIPOIC ACID
~ HS
COOH SH
REDUCED FORM OF METHYLESTER OF ct-LIPOIC ACID
~ HS
~~--..,~COOCH3 SH
FIG. 1. Structure of dihydrolipoic acid homologs.
reduced form of t~-lipoic acid is dihydrolipoic acid (DHLA; see Fig. 1 for structure) which contains a disulfhydryl structure. As the redox potential of the DHLA/a-lipoic acid couple is - 0 . 3 2 V,l DHLA is a potent reductant. DHLA has been found to exert some antioxidant actions. It has been demonstrated that DHLA prevents microsomal lipid peroxidation. 2'3 Bast and Haenen 4 have proposed that DHLA prevents lipid peroxidation by reducing glutathione which in turn recycles vitamin E. We have demonstrated that DHLA scavenges oxygen free radicals (superoxide and hydroxyl radicals) using electron spin resonance (ESR) spin trapping techniques. 5 DHLA has also been shown to reduce 1 R. 2 A. 3 H. 4 A. (I. Y.
L. Searls and D. R. Sanadi, J. Biol. Chem. 235, 2485 (1960). Bast and G. R. M. M. Haenen, Biochim. Biophys. Acta 963, 558 (1988). Scholich, M. E. Murphy, and H. Sies, Biochim. Biophys. Acta 1001, 256 (1989). Bast and G. R. M. M. Haenen, in "Antioxidant in Therapy and Preventive Medicine" Emerit, L. Packer, and C. Auclair, eds.), p. 111. Plenum, New York, 1990. J. Suzuki, M. Tsuchiya, and L. Packer, Free Radical Res. Commun. 15, 255 (1991).
456
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[45]
peroxyl, ascorbyl, and chromanoxyl radicals, 6 and to quench singlet oxygen. 7 Understanding the structural consequences of a compound in its action significantly contributes to the accelerated development of therapeutic agents. Assessment of the structure-activity relationships of antioxidants, however, is difficult in that a complex set of information is necessary to cope with the wide range of actions of an antioxidant in physiological systems. Such a wide range of activity necessitates considerations of different physicochemical characteristics including hydrophobicity and electrostatic behaviors. This chapter describes the methodologies used to understand the structure-antioxidant activity relationships of D H L A using various structural homologs (see Fig. 1 for structures). 8 Generation of Dihydrolipoic Acid from a-Lipoic Acid Preparation of a-Lipoic Acid Solution
a-Lipoic acid may be introduced to assay systems by dissolving in alcohol such as ethanol and benzyl alcohol. Alternatively, a-lipoic acid (e.g., 20 mM) can be dissolved in buffer solutions (e.g., pH 7.4) either by mild heat or by high pH. Mild Heat Method. 9 Add 4.12 mg/ml a-lipoic acid in buffered solution, pH 7.4 (such as potassium phosphate, Tris). Dissolve slowly at 45 ° while stirring. Readjust to pH 7.4 with K O H at 37°. High p H Method. Add 4.12 mg/ml a-lipoic acid in buffered solution, pH 7.4. Dissolve slowly at high pH (pH 9-10) while maintaining the pH with K O H until all solutes dissolve. Readjust to pH 7.4 with HC1. Some homologs such as tetranorlipoic acid and bisnorlipoic acid can be dissolved in aqueous solution more readily, whereas more hydrophobic derivatives such as the methyl ester derivative and sulfoxide of lipoic acid (fl-lipoic acid) need to be dissolved in organic solution such as chloroform. Enzymatic Formation by Lipoamide Dehydrogenase
Lipoamide dehydrogenase (EC 1.8.1.4, dihydrolipoamide dehydrogenase; type III, from porcine heart; Sigma Chemical Co., St. Louis, MO) can catalyze the NADH-dependent reduction of a-lipoic acid to generate 6 v. E. Kagan, A. Shvedova, E. Serbinova, S. Khan, C. Swanson, R. Powell, and L. Packer, Biochem. Pharmacol. 44, 1637(1992). 7T. P. A. Devasagayam, M. Subramanian, D. S. Pradhan, and H. Sies, Chem.-Biol. Interact. 86, 79 (1993). 8y. j. Suzuki, M. Tsuchiya, and L. Packer, Free Radical Res. Commun. 18, 115 (1993). 9 L. Mtiller and H. Menzen, Biochim. Biophys. Acta 1052, 386 (1990).
[451
DIHYDROLIPOICACID
457
B
A 3 t==
E
I//I ///I ////
E
i i i / / / # /
IIII I/// / 1 1 1 / / 1 # / # # /
I/II l/If / 1 1 #
i l i i i11i I I I I III1 / / / I i l l I I I I l l I l i II11 I I I
I
l I i l I
IIit Illi f11I I I I I
0
.&2
8@R
I//Z
Nq,
FIG. 2. Lipoamide dehydrogenase activity.
DHLA. We have found that the reduction mediated by this enzyme is specific in that only R-oMipoic acid (physiological enantiomer) but not Sa-lipoic acid, tetranorlipoic acid, bisnorlipoic acid, or the sulfoxide of lipoic acid can be reduced by this enzyme. The following mixture has been used to obtain the results shown in Fig. 2. Lipoamide dehydrogenase activity is determined by measuring the increase of sulfhydryl content using DTNB (see below). For the reaction, mix the following in a l-ml cuvette: KH2PO4-KOH, pH 7.4 (150 mM), 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB) (200/xM), NADH (I00/zM), o~-lipoic acid (100 #M), and lipoamide dehydrogenase (21 /xg/ml). Chemical Synthesis by Sodium Borohydride
Reduced forms of a-lipoic acid and homologs can be prepared by reduction with sodium borohydride in NaOH. I° Dissolve a-lipoic acid in NaOH, t0 F. Bonomi and S. Pagani, Eur. J. Biochem. 155, 295 (1986).
458
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[45]
and add excess sodium borohydride. Wait for 30 rain, then carefully add 6 M sulfuric acid in order to destroy the excess sodium borohydride. Add an equal volume of dichloromethane and mix well. Transfer the organic phase to a test tube. Add excess sodium sulfate and filter. Evaporate the solvent to give a pale yellow oil. Detection of Dihydrolipoic Acid Spectrophotometric Assay The sulfhydryl content of DHLA can be estimated by monitoring the reduction of Ellman's reagent [5,5'-dithiobis(2-nitrobenzoic acid), DTNB] at 412 nm using an extinction coefficient of 13,600 M -1 c m - l . . For example, the following conditions have been used by Suzuki et al. 5 The reaction mixture in a 1-ml cuvette contains KH2PO4-KOH, pH 7.4 (150 mM), DTNB (200/~M), and the sample to be measured. When reactive oxygen generators are used in the assay system, caution must be taken to prevent nonspecific oxidation of the reaction product, p-nitrobenzoic acid, by hydrogen peroxide. In this case, 100 U/ml of catalase is added to the mixture.12 Nuclear Magnetic Resonance Spectra of a-Lipoic Acid and Dihydrolipoic Acid The IH nuclear magnetic resonance (NMR) spectra of a-lipoic acid and DHLA in chloroform are shown in Fig. 3. The following protocol is used. Dissolve the sample (50 rag) in 600/zl chloroform-d (Sigma, 99.8 atom% D). Place in 5-mm NMR tube. Obtain the spectrum at pulse width of 22/xsec at 300 MHz at room temperature. Chemical shifts resulting from the reduction may be useful in validating the synthesis of the reduced form as well as in in vioo detection of the reduced form. Similar chemical shifts are observed by the reduction of a-lipoic acid homologs such as tetranorlipoic acid, bisnorlipoic acid, and the methyl ester of a-lipoic acid. Antioxidant Properties of Dihydrolipoic Acid Determination of Second-Order Kinetic Rate Constant for Reaction with Superoxide Anion Radicals The apparent second-order kinetic rate constant for the reaction of DHLA with superoxide anion radical (O2-) is determined from the compel1 G. L. Ellman, Arch. Biochem. Biophys. 82, 70 (1959). t2 y . Suzuki, V. Lyall, T. U. L. Biber, and G. D. Ford, Free RadicalBiol. Med. 9, 479 (1990).
[45]
DIHYDROLIPOIC ACID
Lipoic
459
Acid
jL
DHLA I
I
I
I
I
I
I
6.0
5.0
4.0
3.0
2.0
1.0
0.0
PPM FIG. 3. NMR spectra of tx-lipoic acid and dihydrolipoic acid.
tition between DHLA and epinephrine for Oz-:. The oxidation of epinephrine by xanthine plus xanthine oxidase is followed by monitoring the increase of adrenochrome product at 480 nm at 25°. Measurement of epinephrine oxidation by 02- is preferred over other conventional 02detection assays such as those using cytochrome c and nitro blue tetrazolium (NBT) because of direct reduction of cytochrome c and NBT by DHLA. A representative assay system8 contains the following in a 1ml cuvette: KH2PO4-KOH, pH 7.4 (100 raM), epinephrine (0.5 raM), xanthine (100 /zM), xanthine oxidase (EC 1.1.3.22, from cow's milk; Boehringer Mannheim, Indianapolis, IN, 40 mU/ml), and various concentrations of DHLA. Calculation. Following the determination of the DHLA concentration at which 50% of O2--induced epinephrine oxidation is inhibited, the second-order kinetic rate constant for the reaction between O2-: and DHLA (kDHLA)can be computed13 according to [DHLA]kDHLA
= [epinephrine]kepinephrin e
where kepinephrin e is the second-order kinetic rate constant for the reaction between 02 ~ and epinephrine, 4.0 × 1 0 4 M -1 s--l. 14 13 B. Halliwell, Free Radical Res. Commun. 9, 1 (1990). 14 K. Asada and S. Kanematsu, Agric. Biol. Chem. 40, 1891 (1976).
460
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[45]
Fluorescence-Based Assay for Peroxyl Radical Scavenging Activity in Aqueous Phase and Membranes Peroxyl radical scavenging activities of D H L A in both aqueous solutions and liposomes can be assayed by monitoring the fluorescence of oxidizable probes. Generally, B-phycoerythrin is used in hydrophilic environments ~5 and cis-parinaric acid (Molecular Probes, Junction City, OR) in hydrophobic environments ~6 as fluorescence probes. Peroxyl radical diazo initiators ~7 should also be selected according to solubility characteristics, for example, 2,2'-azobis(2-amidinopropane)dihydrochloride (AAPH; Polysciences, Inc., Warrington, PA) for assay in solution and 2,2'azobis(2,4-dimethylvaleronitrile) (AMVN, Polysciences) for membranes. These radical initiators generate peroxyl radicals at a constant rate by thermal decomposition at 40 °, 17causing the fluorescence decay of probes. Prevention of this fluorescence decay by DHLA or homologs is interpreted as the peroxyl radical scavenging activity. The degree of protection can be evaluated as the ratio of the initial fluorescence decay rate in the absence of antioxidant to that in the presence of antioxidant, V/VA. A greater value indicates that the rate constant for peroxyl radical is higher. In Solution. The reaction mixture contains Tris-HCl, pH 7.4 (20 mM), B-Phycoerythrin (0.5 riM), AAPH (25 raM; added in methanol), and D H L A or homolog. The reaction is started by a quick increase of the sample temperature to 40 °. Fluorescence is monitored at 540 nm excitation and 575 nm emission with a 5-nm slit. The concentration of B-phycoerythrin is corrected by a molar extinction coefficient of 2.41 × 10 6 M -~ cm-l at 545 rim.18 In Membranes. Dioleoylphosphatidylcholine (DOPC) liposomes with incorporated cis-parinaric acid and antioxidant are made by sonication of 6 izM cis-parinarie acid, DHLA, and 1.3 mM DOPC dispersed in 20 mM Tris-HCl (pH 7.4) under nitrogen gas at 4 °. The reaction is started by a quick rise of the sample temperature to 40°just after the incorporation of 300 ~ M AMVN into DOPC liposomes by sonication. The fluorescence intensity is monitored at 324 nm excitation and 413 nm emission with a 5-nm slit. The concentration of cis-parinaric acid is corrected by a molar extinction coefficient of 8.0 x l04 M - l cm- ~at 303 nm, or 7.4 × 104 M - 1 cm -l at 318 rim. 19 15 A. N. Glazer, FASEB J. 2, 2487 (1988). 16M. Tsuchiya, G. Scita, H. J. Freisleben, V. E. Kagan, and L. Packer, this series, Vol. 213, p. 460. 17 E. Niki, this series, Vol. 186, p. 100. 18 A. N. Glazer and C. S, Hixson, J. Biol. Chem. 252, 32 (1977). 19 L. A. Sldar, B. S. Hudson, and R. D. Simoni, J. Supramol. Struct. 4, 449 (1976).
[45]
DIHYDROLIPOICA C I D
461
Concluding Remarks As experimental evidence accumulates, it is conceivable that alipoic acid/DHLA is effective as a therapeutic agent. In many cases, the effects of a-lipoic acid appear to be linked to its ability to influence oxidative stress-mediated processes. For example, ct-lipoic acid dietary supplementation successfully prevents myocardial damage induced by ischemia-reperfusion2°; a-lipoic acid prevents macromolecular alterations induced by high concentrations of glucose21; preincubation of cultured human T cells with a-lipoic acid inhibits activation of NFrB, a transcription factor responsible for the activation of human immunodeficiency virus ( H I V ) . 22 Such findings on the effects of a-lipoic acid may relate to the antioxidant properties of DHLA, and thus the development of DHLA-based antioxidants may have significant clinical ramifications. Success in such a strategy may require a more comprehensive approach based on the structure-function relationships of a compound, involving (1) determination of structure-activity relationships, (2) application of concepts derived from structural studies in cellular systems to elucidate the mechanism of action, and finally (3) verification of information in intact animals. Acknowledgments The work was supported by the National Institutes of Health (CA47597) and ASTA Medica. This work was done during the tenure of a Research Fellowship from the American Heart Association, Califomia Affiliate, to Y. J. S.
20 E. Serbinova, S. Khwaja, A. Z. Reznick, and L. Packer, Free Radical Res. Commun. 17, 49 (1992). 21 y. j. Suzuki, M. Tsuchiya, and L. Packer, Free Radical Res. Commun. 17, 211 (1992). 22 y. j. Suzuki, B. B. Aggarwal, and L. Packer, 8iochem. Biophys. Res. Commun. 189, 1709 (1992).
462
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[46]
[46] A n t i o x i d a n t Action of Ginkgo biloba E x t r a c t E G b 761 B y L U C I A MARCOCCI, LESTER PACKER, MARIE-THERESE D R O Y - L E F A I X , A B D E L H A F I D SEKAKI, and M O N I Q U E GARDI~S-ALBERT
Introduction Extracts of Ginkgo biloba leaves are complex titrated, standardized mixtures of active ingredients introduced into medical practice as polyvalent therapeutic agents. In particular, they appear to be beneficial in diseases characterized by impairment of the circulation, and they have been prescribed in diseases in which free radicals are probably implicated, namely, in psychic and behavioral disorders of the elderly, in peripheral vascular deficiency, and in retinal deficiency of ischemic origin. Ginkgo biloba extracts have been reported to reduce the free radicalinduced lipid peroxidation generated by NADPH-Fe 3÷ systems in rat microsomes,2 and to protect human liver microsomes from lipid peroxidation induced by cyclosporin A. 3 The extracts inhibit reactive oxygen production in human leukocytes on stimulation with phorbol myristate acetate. 4 In vivo studies on animal models appear to confirm the antiradical properties of the extract. Extracts may reduce adriamycin-induced inflammation of the rat hindpaw, 5 prevent damage in the retina of rats made diabetic by injection of alloxan, 6 and provide effective protection against functional disorders observed in cerebral and myocardial ischemia models. 7,8 I F. V. DeFeudis, ed., "'Ginkgo biloba Extract (EGb 761): Pharmacological Activities and Clinical Applications." Elsevier, Paris, 1991. 2 j. Pincemail, C. Deby, Y. Lion, P. Hans, K. Drieu, and R. Goutier, in "Flavonoids and Bioflavonoids" (L. Farkas, M. Gabor, and F. Kallay, eds.), p. 423. Szeged, Hungary, 1985. 3 S. A. Barth, G. Inselmann, R. Engemann, and H. T. Heidemann, Biochem. Pharmacol. 41, 1521 (1991). 4 j. Pincemail, A. Thirion, M. Dupuis, P. Braquet, K. Drieu, and C. Deby, Experientia 43, 181 (1987). 5 A. Etienne, M. Chapelat, M. Braquet, P. Clostre, K. Drieu, and F. V. DeFeudis, in "Cerebral Ischemia" (A. Bes, P. Braquet, R. Paoletti, and B. K. Siesjo, eds.), p. 379. Excerpta Medica, Amsterdam, 1974. 6 M. T. Droy-Lefaix, B. Bonhomme, and P. Braquet, in "R6kan Ginkgo biloba: Recent Result in Pharmacology and Clinic" (E. W. Funfgeld, ed.), p. 83. Springer-Verlag, Berlin, 1988. 7 S. Spinnewyn, N. Novel, and F. Clostre, in "R6kan Ginkgo biloba: Recent Result in Pharmacology and Clinic" (E. W. Funfgeld, ed.), p. 143. Springer-Verlag, Berlin, 1988. s j. M. Guillon, L. Rochette, and L. Baranes, in "Rokan Ginkgo biloba: Recent Result in Pharmacology and Clinic" (E. W. Funfgeld, ed.), p. 153. Springer-Vedag, Berlin, 1988.
METHODSIN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
[46]
ANTIOXIDANTACTIONOF EGb 761 in Vitro
463
Different extracts can be produced from a given plant depending on the manufacturing process employed. EGb 761, prepared from Ginkgo biloba leaves according to a special process, is a standardized mixture of several different chemical constituents, prepared as a dry powder. EGb 761 contains two groups of major substances: flavonoid glycosides and terpenoids. L9 The flavonoid fraction is composed of three flavonols: quercetin, kaempferol, and isorhamnetin which are linked to a sugar, l The terpenoid fraction is composed of ginkgolides and bilobalides, l°'H Some organic acids are also present in EGb 761 and play a role in the solubility of the extract in water. 9 This chapter reports the radical scavenging properties of EGb 761, as studied in several in vitro systems. First, the interaction of EGb 761 with superoxide and hydroxyl radicals generated by using the water radiolysis method is presented. The method of radiolysis is useful because it specifically generates superoxide and hydroxyl radicals in known quantities, thus reducing the risk of possible interferences. To evaluate the role of terpenes contained in EGb 761, a comparative study is described using CP 202, a similar mixture lacking the terpene fraction. Second, the interaction of EGb 761 with peroxyl radicals is assayed by a chemiluminescent method. Third, the interaction of EGb 761 with nitric oxide is assessed by two different spectrophotometric methods. Fourth, the effect of EGb 761 on xanthine oxidase activity is described.
Materials Standard EGb 761 (batch K 923) and CP 202 are obtained as dry powders from IPSEN Laboratories (Paris, France) and used without purification; 2,2'-azobis(2,4-dimethylvaleronitrile) (AMVN) is from Polysciences, Inc. (Warrington, PA); 5-amino-2,3-dihydro-l,4-phthalazinedione (luminol), dioleoyl-L-a-phosphatidylcholine, (DOPC), sodium nitroprusside, hydroxylamine, glucose, xanthine, sulfanilamide, naphthylethylenediamide, bovine hemoglobin, glucose oxidase from Aspergillus niger (Type II-S), and catalase (C40) are from Sigma (St. Louis, MO); Sephadex G-25 is from Pharmacia (Uppsala, Sweden). All other products are reagent grade.
9 K. Drieu, Presse M~d. 15, 1455 (1986). l0 K. Weinges, M. Hepp, and H. Jaggy, Liebig's Ann. Chem., p. 521 (1987). H K. Drieu, in " R f k a n , Ginkgo biloba: Recent Results in Pharmacology and Clinic" (E. W. Fiinfgeld, ed.), p. 32. Springer-Verlag, Berlin, 1988.
464
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[46]
Scavenging Effect against Superoxide Anion Radical The reaction of EGb 761 or CP 202 with superoxide is tested by monitoring the absorption spectra before and after exposure to the reactive oxygen species. Superoxide is produced by the method of steady-state radiolysis of water by irradiation of aqueous solution of the extracts with y rays.
Preparation of Reagents Ginkgo biloba Extract Solutions. Solutions of EGb 761 or CP 202 at concentrations between 25 and 200/zg/ml are prepared in 100 mM sodium phosphate buffer, pH 7. Irradiation Conditions Ginkgo biloba extract solutions are irradiated with 3/rays produced from a 6°Co-irradiator. The activity of the irradiator is 80 Ci (dose rate 3.4 x l0 -2 Gy/sec). Irradiation doses range between 0 and 300 Gy. To obtain selective production of superoxide during irradiation, the solutions are saturated with oxygen and I00 mM sodium formate. Spectrophotometric Analysis of Irradiated Solutions The absorption spectra of the solutions are analyzed before and after irradiation by using a Beckman DU70 spectrophotometer (optical path 1 cm). Absorption spectra are recorded from 240 to 500 nm. The Beer-Lambert law is verified in all the wavelength ranges studied and for all the concentrations of EGb 761 or CP 202 used.
Results Irradiation of EGb 761 solutions with increasing doses of y rays increased their absorption across a broad wavelength range (spectra not shown). The addition of 0.3 /zM superoxide dismutase to the solutions before irradiation prevented the spectral change of EGb 761 (data not shown), thus proving superoxide to be responsible for the observed effect. At 266 nm the change in absorbance varied linearly with the irradiation doses (Fig. 1). The slope of the line increased with EGb 761 concentration between 25 and 100 tzg/ml, and remained constant at higher EGb 761 concentrations (data not shown). These data suggest that EGb 761 completely scavenges superoxide at concentrations above 100/zg/ml, whereas at lower concentrations the
ANTIOXIDANTACTIONOF EGb 761 in Vitro
[46]
465
0.44 0.42
¢q 0.40 0.38 O
0.36 0.34
!
0
20
!
40
• '"'1
60
qU
80
-=
I
|
100
120
Irradiation d o s e (Gy)
F~G. 1. Effectof superoxide generated by water radiolysison EGb 761 absorbance at 266 nm. EGb 761 solution (25/xg/ml) in 100 mM phosphate buffer, pH 7, containing100 mM sodiumformate and saturated with oxygenwas irradiated with increasingdoses of y rays. Increasingirradiationcorrespondsto increasingquantitiesof superoxidereactingwith the solute. antioxidant competes with spontaneous dismutation of the radical. In contrast, when solutions of CP 202 (200 t~g/ml), which lacks the terpene components, were exposed to increasing doses of irradiation, no change in absorbance was observed across the broad wavelength range (data not shown). Hence, the terpene fraction in EGb 761 could play a role in the scavenging of superoxide radicals. Scavenging Effect against Hydroxyl Radical The reaction of EGb 761 or CP 202 with hydroxyl radicals is tested by monitoring the absorption spectra before and after exposure to this reactive oxygen species. Hydroxyl radical is produced by the method of steady-state radiolysis of water by irradiation of aqueous solution of the extracts with 3' rays.
Preparation of Reagents Ginkgo biloba Extract Solutions. Solutions of EGb 761 or CP 202 at concentrations between 25 and 200 t~g/ml are prepared in 100 mM sodium phosphate buffer, pH 7.
466
[46]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
Irradiation Conditions Ginkgo biloba extract or CP 202 solutions are irradiated with T rays produced from a 6°Co-irradiator. The activity of the irradiator is 80 Ci (dose 3.4 × 10-2 Gy/sec). Irradiation doses range between 0 and 300 Gy. To obtain selective generation of hydroxyl radicals, the solutions are deaerated and saturated with NeO. Spectrophotometric Analysis of Irradiated Solutions The absorption spectra of the solutions are analyzed before and after irradiation by using a Beckman DU70 spectrophotometer (optical path I cm). Absorption spectra are recorded from 240 to 500 nm. The Beer-Lambert law is verified in all the wavelength ranges studied and for all the concentrations of EGb 761 or CP 202. The absorptions (from 240 to 500 nm) of EGb 761 or CP 202 solutions of the same concentration are very similar (differential absorptions < 10%).
Results The absorbance of EGb 761 and CP 202 solutions across a broad wavelength range decreased with increasing doses of irradiation. Figure 2 shows this decrease at 266 nm for both EGb 761 than CP 202 at 50 0.8
0.7 a °o
0.6 0.5 .4
m
0
50
u
100
m
150
k
200
I r r a d i a t i o n d o s e (Gy) FIG. 2. Effectof hydroxylradical generatedby water radiolysison EGb 761 and CP 202 absorbance at 266 nm. EGb 761 solution(50 ~g/ml, O) or CP 202 (50 ~g/ml, O) in 100 mM phosphate buffer, pH 7, saturated with N20 was irradiatedwith increasingdoses ofy rays. Increasingirradiationcorrespondsto increasingquantitiesof hydroxylradicalreactingwith the solute.
[46]
ANTIOXIDANTACTIONOF EGb 761 in Vitro
467
/zg/ml. In the presence of EGb 761, the absorbance decrease is linearly dependent on irradiation dose only for doses below 80 Gy (filled circle). In contrast, in the presence of CP 202 the absorbance decrease is linearly dependent on the range of irradiation doses (open circles). The change of absorbance versus irradiation dose is independent of the EGb 761 concentrations analyzed (25 and 200/xg/ml) (data not shown), thus implying that the lowest concentration of antioxidant was sufficient for scavenging all hydroxyl radical produced by gamma radiolysis. In contrast, the change of absorbance versus irradiation dose depended on the concentrations of CP 202. It increased with increasing concentrations of CP 202 and reached a plateau at concentrations above 100/zg/ml (data not shown). This indicates that there is competition for hydroxyl free radicals between CP 202 and another scavenger, perhaps radiolytically produced H202 . Scavenging Activity against Peroxyl Radical The peroxyl radical scavenging activity of EGb 761 is determined by monitoring the effect of the antioxidant on the chemiluminescence produced, in the presence of DOPC liposomes, from the reaction of luminol with peroxyl radicals. 12Peroxyl radical is generated from a hydrophobic azo initiator, AMVN, on thermal decomposition. 13
Preparation of Reagents Liposomes. Put 2.5 ml of a chloroform solution of DOPC (20 mg/ml) in a test tube and dry under a stream of N2. Add 5 ml of 20 mM Tris-HC1 buffer, pH 7.4. Vortex for 5 min and then sonicate for 20 min in a probe sonicator. Incubate at room temperature for 30 min. Azo Initiator Solution. Dissolve 500 mM AMVN in ethanol and store on ice to prevent decomposition. Luminoi Solution. Weigh 7 mg of luminol and add 10 ml of 20 mM Tris-HC1 buffer, pH 7.4. Vortex for 5 min then filter with 0.45-/zm Costar (Cambridge, MA) filter unit. Measure the optical density of the filtered solution at 347 nm, and calculate the concentration of luminol by using a extinction coefficient of 7630. Dilute the luminol in 20 mM Tris-HC1 buffer, pH 7.4 to obtain a final concentration of 2 mM. Ginkgo biloba Extract Solution. Dissolve EGb 761 at concentrations from 50 to 500/zg/ml in 20 mM Tris-HCl buffer, pH 7.4.
12 V. E. Kagan, H. J. Freisleben, M. Tsuchiya, T. Forte, and L. Packer, Free Radical Res. Commun. 15, 265 (1991). 13 E. Niki, this series, Vol. 186, p. 100.
468
ANTIOXIDANT CHARACTERIZATION AND ASSAY
o
[46]
40
!2o i
10
i~
-I~ 0
5
I0
15
20
T i m e (mln) FIG. 3. Scavenging effect of EGb 761 against peroxyl radicals. A 150/~M luminoi solution was exposed to 12/zM AMVN in 20 mM Tris-HCl, pH 7.4, in the presence of 2 mg/ml DOPC liposomes and different concentrations of EGb 761 : 0, O; II, 5/zM; 0 , 9/zM; and &, 15/xM. An experiment representative of three trials is reported.
Assay Method Put 1.35 ml of 20 mM Tri-HCI buffer, pH 7.4, in a chemiluminescence tube; add 400/A of DOPC and 50/zl of EGb 761 solutions. Warm at 40 ° for 15 min. Add 150/zl luminol solution. Start the reaction by adding 50 /A of AMVN solution. Record the induction of chemiluminescence for at least 10 min.
Results Addition of AMVN to luminol solution in the presence of DOPC liposomes produces chemiluminescence that reaches a maximum in 3-4 min and then is stable for at least 10 min. In the presence of EGb 761 the maximum of the chemiluminescence signal is lower and is reached after a longer time interval. The effect is dependent on the concentration of the antioxidant (Fig. 3). EGb 761 at 5/zg/ml reduces the chemiluminescence maximum by 50%, and at 15 ~g/ml antioxidant, no chemiluminescence is observed. Direct interference of EGb 761 with the chemiluminescence signal from luminol can be excluded; luminol emittance is around 450 nm, TM and the absorbance of EGb 761 is negligible at this wavelength. 14 E. Cadenas and H. Sies, this series, Vol. 105, p. 221.
[46]
ANTIOXIDANTACTIONOF EGb 761 in Vitro
469
Xanthine Oxidase Activity Compounds that interact with xanthine oxidase can affect the oxidation of xanthine to uric acid. The reaction can be recorded spectrophotometrically at 295 nm) 5
Preparation of Reagents Xanthine Oxidase Solution. Dilute xanthine oxidase to a final concentration of 250 mU/ml in 0.1 mM phosphate buffer, pH 7.4. Xanthine Solution. Weigh 1.74 mg of xanthine, add 10 ml water, then boil for 5 min to completely dissolve the powder. Ginkgo biloba Extract Solution. Prepare EGb 761 solutions at concentrations between 10 and 500/zg/ml in 0.1 mM phosphate buffer, pH 7.4. Assay Method Pipette 0.75 ml of 0.1 mM phosphate buffer, pH 7.4, into a quartz cuvette. Add 250/xl EGb 761 and 15/zl xanthine to both the sample and the reference. Start the reaction by adding 10/zl xanthine oxidase to the sample cuvette. Record the absorbance at 295 nm for 3 min at room temperature.
Results In the presence of 15/zM xanthine, EGb 761 inhibited xanthine oxidase activity in a dose-dependent manner up to 50/xg/ml (Fig. 4). The maximum of inhibition (20-30% residual activity) was obtained in the presence of EGb 761 at concentrations higher than 50/xg/ml. Scavenging Activity against Nitric Oxide: Oxyhemoglobin Methods Oxyhemoglobin is a well-known scavenger of nitric oxide, which oxidizes the protein to methemoglobin.16 Scavengers of nitric oxide compete with oxyhemoglobin for the radical, thus protecting the protein from the oxidation. The rate of oxidation of oxyhemoglobin to methemoglobin can be measured at 576 nm. ~7 Nitric oxide is produced in the reaction of catalase-Complex I with hydroxylamine.18 Catalase-Complex I is formed 15 I. Fridovich, in"Handbook of Methods for Oxygen Radical Research" (R. A. Greenwald, ed.) p. 51. CRC Press, Boca Raton, Florida, 1985. 16 M. P. Doyle and J. W. Hoekstra, J. lnorg. Biochem. 14, 351 (1981). 17 E. E. Di Iorio, this series, Vol. 76, p. 57. 18 D. Keilin and P. Nicholis, Biochim. Biophys. Acta 29, 302 (1958).
470
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[46]
120 ¢J al o
I00 80
60 40-
ig 2o 0
I
0
30
!
!
I
J
60
90
120
150
EGb 761 [lag/ml) FIG. 4. Effect of EGb 761 on xanthine oxidase activity. The activity of 2.5 mU/ml xanthine oxidase was measured at 295 nm by monitoring the formation of uric acid from 15 /~M xanthine at room temperature in 100 mM phosphate buffer, pH 7.4. Average values from three experiments (--+SD) are reported.
during the reaction of catalase with hydrogen peroxide, which is generated from oxidation of glucose by glucose oxidase.
Preparation of Reagents Ginkgo biloba Extract Solution. Prepare EGb 761 in 20 mM phosphate buffer, pH 7.4, at final concentrations ranging between 50 and 500/zg/ml. Enzymatic System for Nitric Oxide Production. Dissolve 1 M glucose in 20 mM phosphate buffer, pH 7.4. Dissolve 0.45 mg/ml glucose oxidase in 1 ml in 20 mM phosphate buffer, pH 7.4. Dissolve 20 mg/ml catalase in 1 ml 20 mM phosphate buffer, pH 7.4. Dissolve 2.5 mM hydroxylamine in 20 mM phosphate buffer, pH 7.4. Oxyhemogiobin Solution. Dissolve 50 mg of hemoglobin in I ml of 20 mM phosphate buffer, pH 7.4. Add 5 mg dithionite and mix. Dithionite reduces the methemoglobinpresent in the commercial solution of hemoglobin. To remove dithionite, the solution is passed through a Sephadex G-25 column. Determine the concentration of oxyhemoglobin by reading the absorbance at 540 nm, using an extinction coefficient for oxyhemoglobin of 54,000 M -~ cm -]. Dilute the solution in 20 mM phosphate buffer, pH 7.4, to obtain oxyhemoglobin solutions of 0.3, 0.6, 1.8, 2.5, and 3.6 raM. Assay Method Put 1 ml of 20 mM phosphate buffer in a spectrophotometric silica cuvette and warm at 37° for 10 min. Add 10/~1 of oxyhemoglobin solution,
[46]
ANTIOXIDANT ACTION OF EGb 761 in Vitro
471
20/zl EGb 761 solution, 10/zl catalase, 10/zl glucose oxidase, and 10/zl glucose. Record the decrease in the absorbance at 576 nm for 1 min against blank containing buffer. Add I0/zl hydroxylamine and record the change in the absorbance for an additional 5 min. Results
EGb 761 can be a scavenger of nitric oxide. The extract inhibits the oxidation o f oxyhemoglobin induced by nitric oxide in a dose-dependent manner. As hemoglobin is itself a scavenger of nitric oxide, the protective effect of E G b 761 is higher at low concentration of hemoglobin (Fig. 5); 80/xM E G b 761 reduces the oxidation of 3, 6, and 18/xM hemoglobin by 70, 60, and 20%, respectively. Antioxidant at 8 0 / z M does not protect 36 /xM hemoglobin (data not shown). No effect of EGb 761 on glucose oxidase activity, as measured by the polarographic method of oxygen consumption, was observed; the extract did not affect the catalase activity, as measured spectrophotometrically, and thus any interference of the extract on the enzymatic systems used to produce nitric oxide can be excluded (data not shown).
100
8o .~
6o
~
40
0
~
20 o
|
0
20
!
40
|
60
i
80
I
100
EGB 761 (~g/ml)
FIG. 5. Effect of EGb 761 on oxyhemogiobinoxidation by nitric oxide generated from an enzymatic system. Oxyhemoglobin at 6/zM (0) or 25/~M (O) was exposed to nitric oxide in 20 mM phosphate buffer, pH 7.4, at 37°, in the presence of different concentrations of EGB 761. Nitric oxide was produced enzymaticallyby the reaction of 200/zg/ml catalase with 23/~M hydroxylamine, in the presence of hydrogen peroxide produced from the oxidation of 10 mM glucose by 4.5/xg/mi glucose oxidase. Average values from four experiments (-+SD) are reported.
472
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[46]
Scavenging Activity against Nitric Oxide: Nitrite Detection Method Nitric oxide interacts with oxygen to produce stable products, nitrite and nitrate. Scavengers of nitric oxide compete with oxygen, leading to a reduced production of nitrite. The concentration of nitrite in aqueous solution can be assayed spectrophotometrically by using the Greiss reagent, with which nitrite reacts to give a stable product which absorbs at 542 nm. 19 This method is less sensitive than the oxyhemoglobin method, and high concentrations of nitrite have to accumulate before being detected. To produce a constant flux of nitric oxide over a period of hours, a chemical source of nitric oxide is preferable. Sodium nitroprusside, which spontaneously produces nitric oxide when dissolved in aqueous solution at physiological pH, 2° is used.
Preparation of Reagents Sodium Nitroprusside Solution. Immediately before the experiment, dissolve I0 mM sodium nitroprusside in 20 mM phosphate buffer, pH 7.4, previously bubbled with argon. Ginkgo biloba Extract Solution. Dissolve EGb 761 in 20 mM phosphate buffer, pH 7.4, to obtain solutions ranging between 50 and 500/xg/ml. Greiss Reagent Solution. Prepare solution A containing 2% (% w/v) sulfanilamide and 4% (% w/v) H3PO 4 . Prepare solution B containing 0.2% (% w/v) naphthylethylenediamide. The solutions are stable for months at 4°. Add 50 ml of solution A to 50 ml of solution B just before the assay. Incubation Conditions Dilute 0.5 ml of EGb 761 solutions of various concentrations with 0.5 ml of sodium nitroprusside solution and incubate at 25° for 150 min. Make control samples by incubating EGb 761 with 20 mM phosphate buffer, pH 7.4, without sodium nitroprusside.
Nitrite Assay At the end of the incubation, add 1 ml of Greiss reagent to each sample. Read the absorbance at 542 nm of the samples containing sodium nitroprusside against the absorbance of the control samples. Calculate the
t9 L. C. Green, D. A. Wagner, J. Glogowski, P. L. Skipper, J. S. Wishnok, and S. R. Tannenbaum, Anal. Biochem. 126, 131 (1982). 2o M. Feelisch and E. A. Noack, Eur. J. Pharmacol. 139, 19 (1987).
ANT1OXIDANT ACTION OF EGb 761 in Vitro
[46]
473
nitrite concentration by referring to the absorbance of standard solutions of potassium nitrite.
Results A linear time-dependent production of nitrite is obtained for at least 4 hr from 5 m M sodium nitroprusside (data not shown). In the presence of EGb 761, the concentration of nitrite detected after 150 min of incubation is decreased depending on the antioxidant concentration (Fig. 6). EGb 761 at 1 m M did not affect the calibration curves obtained from standard solutions of potassium nitrite, thus indicating that the antioxidant does not react directly with nitrite. Concluding Remarks The extract E G b 761 from Ginkgo biloba leaves appears to be an efficient and multifunctional antioxidant compound for protection against oxidative stress. The gamma radiolysis method we describe indicates that EGb 761 can react with hydroxyl radicals and superoxide anions. The scavenging effect against hydroxyl radical appears to be dose-independent for concentrations of EGb 761 higher than 25 tzg/ml. H o w e v e r , a dosedependent effect of EGb 761 was obtained against superoxide radical up to 150/xM antioxidant. The different dose effects are in agreement with
120 "
9O
-¢ z
60 30
0
1
0
50
i
I
100
150
i
200
EGb 761 (~g/ml)
FIG. 6. Effect of EGb 761 on nitrite production from sodium nitroprusside. Sodium nitroprusside (5 raM) was incubated in 20 mM phosphate buffer, pH 7.4, with different concentrations of EGb 761 at 25° for 150 rain. The production of nitrite was assayed by using the Greiss reagent. Average values from four experiments (-SD) are reported.
474
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[46]
the poor reactivity of superoxide free radicals (which may dismuate if the rate with a specific scavenger is weak), as well as with the high reactivity of hydroxyl free radicals (which usually have diffusion-controlled rates). The superoxide scavenging action of EGb 761 is characterized by an increase in UV absorbance, whereas the action of hydroxyl radicals on EGb 761 involves a decrease in UV absorbance. Thus, attack at different molecular sites by hydroxyl radical or superoxide free radicals on EGb 761 can be postulated. CP 202 (i.e., EGb 761 without terpenes) reacts with hydroxyl free radicals, but it does not seem to scavenge superoxide anion. This important difference can be attributed to the absence of the terpenic fraction in CP 202. To our knowledge terpenes do not directly scavenge superoxide free radicals. Hence, their presence in EGb 761 could contribute indirectly to the reaction. For example, terpenes could react with a superoxide adduct or a superoxide-induced free radical product formed by the action of superoxide anions on a specific molecular site of EGb 761. Such interactions could result in the transformation of the terpenic fraction, which would give UV-visible absorbance modifications. The presence of an iron superoxide dismutase in Gingko biloba extracts has been reported. 2~ We cannot exclude the possibility that the superoxide scavenging properties of EGb 761 Ginkgo biloba extract is due to the presence of this protein, and that the activity of the enzyme is lost during the preparation of the terpene-free compound, CP 202. The effect of EGb 761 on chemiluminescence production in the reaction of luminol with AMVN-generated peroxyl radical in the presence of DOPC liposomes indicates that EGb 761 can scavenge peroxyl radicals, thus explaining the reported ability of EGb 761 to protect membranes from peroxidative degradation. 22 EGb 761 not only interacts directly with reactive oxygen species, but it could interfere with their generation. EGb 761 inhibits the activity of xanthine oxidase, a cellular enzymatic source of superoxide and hydrogen peroxide, whose concentration has been shown to increase under ischemia-reperfusion conditions. 23 The scavenging effect of EGb 761 on reactive oxygen species as well as the inhibition of xanthine oxidase activity could explain the beneficial effect of the extract on functional disorders
21 M. V. Duck and M. L. Salin, Arch. Biochem. Biophys. 243, 305 (1985). 22 A. Tosaki and M. T. Droy-Lefaix, in "Ginkgo biloba Extract (EGb 761) as a Free Radical Scavenger" (C. Ferradini, M. T. Droy-Lefaix, and Y. Christen, eds.), p. 141. Elsevier, Paris, 1993. 23 j. McCord, N. Engl. J. Med. 312, 159 (1985).
[46]
ANTIOXIDANT ACTION OF E G b 761 in Vitro
475
linked to oxidative stress as in brain, cochlear, cardiac, and retinal ischemia-reperfusion.24-26 The radical scavenging effects of EGb 761 also extend to reaction with nitric oxide. In fact, EGb 761 is able to compete with hemoglobin for reaction with the radical. The scavenging effect of EGb 716 on nitric oxide is consistent with the nitrite assay. However, we have to point out that this assay is less specific for nitric oxide detection than the method involving oxyhemoglobin. Nitrite is a final product of the reaction of nitric oxide with oxygen through intermediates such as NO2, N204, and N203; as a consequence, the decrease in nitrite production we observed in the presence of EGb 761 could be due not only to the reaction of EGb 761 with nitric oxide, but also to the reaction of the extract with the other nitrogen oxides. If this were the case, EGb 761 appears to be an effective antioxidant agent against oxidative damage mediated by nitrogen oxides in general. It is difficult to define the biological consequences of the interaction of EGb 761 with nitric oxide, taking into account the multifunctional role of this radical. Nitric oxide, identified as being produced in different types of mammalian cells, 27 has been reported to be involved in oxidative reactions. The interaction of nitric oxide with ferritin resulting in iron release and consequent catalysis of lipid peroxidation reactions has been described. 28 Furthermore, products from the reaction of nitric oxide with superoxide such as peroxynitrite anion (or its conjugate acid peroxynitrous acid), hydroxyl radical, and nitrogen dioxide, by themselves or in association with one another, have also been shown to oxidize sulfhydryls or initiate lipid peroxidation reactions. 29,3° On the other hand, nitric oxide appears to have a toxic bactericidal effect, to account for the activity of endothelium-derived relaxing factor, to act as neurotransmitter, and to prevent platelet aggregation. 27 Nitric oxide by promoting vasodilation has been reported to exert a protective effect in ischemia-reperfusion models. 31 Thus, the scavenging action of EGb 761 against nitric oxide could modulate the biological responses of nitric oxide. 24 M. Culcasi, S. Piestri, I. Carrirre, P. d'Arbigny, and K. Drieu, in "Ginkgo biloba Extract (EGb 761) as a Free Radical Scavanger" (C. Ferradini, M. T. Droy-Lefaix, and Y. Christen, eds.), p. 1-23. Elsevier, Paris, 1993. 25 B. Spinnewyn, N. Blavet, and F. Clostre, Presse Med. 15, 1511 (1986). 26 M. E. Szabo, M. T. Droy-Lefaix, M. Doly, and P. Braquet, Ophthalmol. Res. 25, 1 (1993). 27 S. Moncada, R. M. J. Palmer, and E. A. Higgs, Pharmacol. Rev. 48, 109 (1991). 28 D. W. Reif and R. D. Simmons, Arch. Biochem. Biophys. 283, 537 (1990). 29 R. Radi, J. S. Beckman, K. M. Bush, and B. A. Freeman, J. Biol. Chem. 266, 4244 (1991). 3o B. Halliwell, M. L. Hu, S. Louie, T. R. Duvall, B. K. Tarkington, P. Motchnik, and C. E. Cross, FEBS Lett. 313, 62 (1992). 31 G. Johnson, P. S. Tsao, D. Mulloy, and A. M. Lefer, Crit. Care Med. 19, 244 (1990).
476
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[47]
[47] E b s e l e n : A G l u t a t h i o n e P e r o x i d a s e M i m i c B y HELMUT SIES
Introduction Ebselen, 2-phenyl-1,2-benzisoselenazol-3(2H)-one, exhibits activity as an enzyme mimic. The reaction catalyzed is that of a glutathione (GSH) peroxidase, namely, the reduction of a hydroperoxide at the expense of thiol. The specificity for substrates ranges from hydrogen peroxide and smaller organic hydroperoxides to membrane-bound phospholipid and cholesterol hydroperoxides. In addition to glutathione, the thiol reductant cosubstrate can be dithioerythritol, N-acetylcysteine, or dihydrolipoate, or other suitable thiol compounds. Ebselen also has properties of a free radical and singlet oxygen quencher. Model experiments in vitro with liposomes, microsomes, isolated cells, and organs show that the protection against oxidative challenge afforded by ebselen can be explained largely by the activity as GSH peroxidase mimic. Whether this also explains preliminary results in clinical settings is yet unknown. Studies of the metabolism and disposition of ebselen showed that the selenium is not bioavailable, explaining the extremely low toxicity observed in animal studies. In 1984, the glutathione peroxidase-like activity of a novel biologically active selenoorganic compound, ebselen (then called PZ 51), was described. 1,2 Since then, extended research on this interesting molecule has ranged from pulse-radiolytic studies on radical reactivity through its biological properties in cells and organs all the way to clinical settings. 3 Basically, the compound is considered capable of contributing to the antioxidant defense in tissues, so that a potential pharmacological application becomes of interest. Early on, it became clear that many of the biological properties of ebselen were related to its property as an enzyme mimic, carrying out the function of the selenoenzyme glutathione peroxidase (GPx) and, as found by Maiorino et al.,4 of phospholipid hydroperoxide GSH peroxidase (PHGPx). This chapter is based on an overview) It is obvious that the biological functions of ebselen and its derivatives might I A. Mfiller, E. Cadenas, P. Graf, and H. Sies, Biochem. Pharmacol. 33, 3235 (1984). 2 A. Wendel, M. Fausel, H. Safayhi, G. Tiegs, and R. Otter, Biochem. Pharmacol. 33,
3241 (1984). 3 H. Sies, Free Radical Biol. Med. 14, 333 (1993). 4 M. Maiorino, A. Roveri, and F. Ursini,Arch. Biochem. Biophys. 295, 404 (1992). METHODS IN ENZYMOLOGY.VOL. 234
Copyright © 1994by Academic Press, Inc. All fights of reproductionin any formreserved.
[47]
EBSELEN
477
extend to other reactions in addition to the reduction of hydroperoxides, with the role of selenium in biology and medicine not yet being fully known. Parnham and Graf 5 reviewed a number of selenoorganic compounds as potential therapeutic agents.
Synthesis The compound ebselen, 2-phenyl-l,2-benzisoselenazol-3(2H)-one, also called PZ 51, was synthesized at Nattermann & Cie. GmbH (Cologne, Germany) under the following patents: European Patent 44,971; German Patent 3,027,073; Japanese Patent K-8256,427; U.S. Patent 4,352,799. The synthesis is based on those described by Lesser and Weiss 6 and Weber and Renson7: 2-chlorocarbonylbenzeneselenenyl chloride is reacted with aniline in a Schotten-Baumann reaction as described by Fischer and Dereu. 8 Several analogs and derivatives have also been synthesized. T M
Reaction Scheme for Glutathione Peroxidase-Like Activity The reaction cycle for the enzymatic catalysis is thought to proceed in three main steps, involving the enzyme-bound selenocysteine, E-CysSell, present as the selenol or more likely as the selenolate (see Fischer and Dereu 8 for a detailed discussion of this and of the intermediate complexes involved). In the first step [reaction (1)], the organic hydroperoxide ROOH reacts to yield the selenic acid, E-Cys-SeOH, and the corresponding alcohol, ROH. The following two steps consist of the sequential reduction by thiol, GSH; reaction (2) gives the selenodisulfide and water, and reaction (3) regenerates the selenol and the disulfide, oxidized glutathione (GSSG). The overall reaction is that of GPx or PHGPx [reaction (4)]. M. J. Parnham and E. Graf, Prog. Drug Res. 36, 9 (1991). 6 R. Lesser and R. Weiss, Chem. Ber. 57, 1077 (1924). 7 R. Weber and M. Renson, Bull. Soc. Chim. Fr. 2, 1124 (1976). 8 H. Fischer and N. Dereu, Bull Soc. Chim. Belg. 96, 757 (1987). 9 R. Cantineau, G. Thiange, A. Plenevaux, L. Christiaens, and M. Guillaume, J. Labelled Compd. Radiopharm. 23, 59 (1986). to C. Lambert, R. Cantineau, L. Christiaens, J. Biedermann, and N. Dereu, Bull. Soc. Chim. Belg. 96, 383 (1987). 11 L. Engmann and A. HaUberg, J. Org. Chem. 54, 2964 (1989). 12 S. R. Wilson, P. A. Zucker, R.-R. C. Huang, and A. Spector, J. Am. Chem. Soc. 111, 5936 (1989). it I. A. Cotgreave, R. Morgenstern, L. Engman, and J. Ahokas, Chem.-Biol. Interact. 84, 69 (1992). 14 K. A. CaldweU and A. L. Tappel, Biochemistry 3, 1643 (1964).
478
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[47]
E - C y s - S e H + ROOH---~ E-Cys-SeOH + ROH E-Cys-SeOH + GSH---~ E-Cys-Se-SG + H20 E-Cys-Se-SG + GSH---~ E - C y s - S e H + GSSG
(1) (2) (3)
ROOH + 2GSH---~ ROH + H20 + GSSG
(4)
In chemical analyses of the GPx-like reactivity of ebselen, Fischer and Dereu 8 proposed that the reduction and oxidation of ebselen can be summarized in two cycles, explaining the catalytic reduction of hydroperoxides in the presence of thiol, differing from the reactions proposed for the enzyme GPx and for selenocysteine ~4-16in that it is independent of a selenol intermediate. For ebselen, stabilization of the intermediate selenenic acid is thought by these authors to be achieved through reaction with the N H function of the benzamide to yield a cyclic benzoyl selenenamide, benzisoselenazole. Fischer and Dereu 8 suggest that this mode of stabilization of a selenenic acid through cyclization might also occur in the enzyme, GPx, as peptidic N H bonds from tryptophan (Trp-148; highly conserved in different peroxidases) as well as the N H 2 moiety of glutamine (Gin70), both in proximity of the selenocysteine (SeCys-35), as shown by x-ray crystallography of the enzyme, i7'18 According to Fischer and Dereu 8 and further work, ~9 the diselenide would react with the hydroperoxide to yield the parent compound and water. It was also noted 8 that the selenenyl sulfide might constitute a storage form of ebselen and eventually be responsible for transport of the drug. However, in a kinetic study of the catalysis of the GPx reaction by ebselen, Maiorino et al. 2° concluded that the mechanism appeared kinetically identical to that of the enzyme reaction, a Ter Uni Ping-Pong mechanism. Carboxymethylation of ebselen by iodoacetate to an inactive derivative suggested that a selenol moiety is involved. Cotgreave et al.13 have devised a method to detect ebselen selenol by its reaction with 1-chloro-2,4-dinitrobenzene. In applying this method, it was concluded 21 that the selenol is the predominant molecular species responsible for the GSH- or dithiothreitol-dependent peroxidase activity of ebselen. Thus, it appears that the reaction scheme of the catalysis of the GPx reaction by 15 K. A. Caldwell and A. L. Tappel, Arch. Biochem. Biophys. 112, 196 (1965). 16 K. Yasuda, H. Watanabe, S. Yamazaki, and S. Toda, Biochem. Biophys. Res. Commun. 96, 243 (1980). 17 R. Ladenstein, O. Epp, K. Bartels, A. Jones, R. Huber, and A. Wendel, J. Mol. Biol. 134, 199 (1979). 18 O. Epp, R. Ladenstein, and A. Wendel, Eur. J. Biochem. 133, 199 (1983). 19 G. R. M. M. Haenen, B. M. de Rooij, N. P. E. Vermeulen, and A. Bast, Mol. Pharmacol. 37, 412 (1990). 20 M. Maiorino, A. Roveri, M. Coassin, and F. Ursini, Biochem. Pharmacol. 37, 2267 (1988). 21 R. Morgenstern, I. A. Cotgreave, and L. Engman, Chem.-Biol. Interact. 84, 77 (1992).
[47]
EBSELEN
479
ebselen occurs in analogy to the mechanism of GPx enzyme catalysis as initially deduced by Wendel et al. z Unlike the enzyme-catalyzed reaction with binding sites conferring specificity for GSH, ebselen can utilize other thiols in addition to GSH, for example, dithioerythritol, 22 N-acetylcysteine,23 or dihydrolipoate. 24 Detection o f Glutathione Peroxidase-Like Activity o f Ebselen
The first demonstration of the ability of ebselen to carry out the reduction of hydroperoxides at the expense of thiol-reducing equivalents (i.e., the reaction carried out by GSH peroxidase) was performed using a coupled enzymatic assay; the compound, then called PZ 51, was compared to the sulfur analog, PZ 25, which did not exhibit activity. The data, together with further work on the antioxidant properties of ebselen, were published by Miiller et al. 1 Wendel et al. 2 reported GPx-like activity by measuring the loss of GSH and extended the kinetic analysis. The activation energy for the ebselen-catalyzed reaction is 55 kJ/mol per Kelvin, in comparison to 36.5 kJ/mol per Kelvin for the enzyme-catalyzed reaction for GPx. 2 The following assay systems for the detection of GPx-like activity have been described. (1) In the coupled enzymatic assay, as in usual coupled enzymatic assays, the test reaction (here GPx) is coupled to an indicator reaction operated such that the test reaction is rate-limiting. The indicator reaction is that of glutathione disulfide (GSSG) reductase, and the loss of reduced nicotinamide-adenine dinucleotide phosphate (NADPH) is monitored continuously by absorbance spectrophotometry. 1'2° (2) Assay of GSH removal can be done by stopping the reaction and then assaying for remaining GSH (e.g., by the formation of thionitrobenzoate from Ellman's reagent 2 or of a monobromobimane adduct24). (3) For assay of hydroperoxide removal, a standard method using the iron-thiocyanate complex has been employed, for example, z4 Radical Scavenging
The reactivity of ebselen and related selenoorganic compounds with
1,2-dichloroethane radical cations and halogenated peroxyl radicals was studied by pulse radiolysis. 25The rate constant for the reaction of ebselen 22 A. Miiller, H. Gabriel, and H. Sies, Biochem. Pharmacol. 34, 1185 (1985). 23 I. A. Cotgreave, M. S. Sandy, M. Berggren, P. M. Moldeus, and M. T. Smith, Biochem. Pharmacol. 36, 2899 (1987). 24 I. A. Cotgreave, P. Moldrus, R. Brattsand, A. Hallberg, C. M. Andersson, and L. Engman, Biochem. Pharmacol. 43, 793 (1992). 25 C. Sch6neich, V. Narayanaswami, K.-D. Asmus, and H. Sies, Arch. Biochem. Biophys. 282, 18 (1990).
480
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[47]
with trichloromethyl peroxyl radicals was determined to be 2.9 x 108 M -~ sec -1, whereas the sulfur analog 2-phenyl-l,2-benzisothiazol-3(2H)-one was oxidized at much lower rates. The rate constant observed for ebselen is comparable to that of o~-tocopherol under similar conditions.
Singlet Oxygen Quenching Reactivity of ebselen with singlet molecular oxygen is only sluggish, the rate constant being 2.5 x 106 M -1 s e c - l . 26 The sulfur analog exhibits a rate constant an order of magnitude lower than this.
Protection against Lipid Peroxidation The protective effect of ebselen against iron-ADP-induced lipid peroxidation in hepatic microsomal fractions using ascorbate or NADPH as reductant has been amply documented.l,27,28 In intact cells, it was shown that the protective effect depends on the presence of GSH. Hepatocytes depleted of GSH could not be protected by ebselen, whereas normal cells were protected, 22 suggesting that the GPx-like activity of ebselen is involved in the protection of intact cells against lipid peroxidation. In these experiments, several parameters of lipid peroxidation were followed, including the generation of low-level chemiluminescence and the formation of alkanes (ethane, n-pentane) and malondialdehyde. In two studies, 4,29 it was concluded that the major role of ebselen in protecting against different types of oxidative attack is to act in its capacity as GPx mimic and, in particular, as PHGPx mimic. Maiorino et al. 2° had observed that ebselen acted particularly well with the phospholipid hydroperoxides and with cholesterol hydroperoxide and cholesterol ester hydroperoxides. 4 This conclusion came from experiments with photooxidized liposomes and peroxidized low density lipoproteins. The radical scavenging activity of ebselen has little, if any, effect. This was shown by competition kinetics based on free radical-dependent bleaching of crocin4 or the lack of reaction with 2,2-diphenyl-l-picrylhydrazyl (DPPH) or the lack of suppression of oxidation of methyl linoleate induced by the radical generators, 2,2'-azobis(amidinopropane) dihydrochloride (AAPH) or 2,2'-azobis(2,4-dimethylvaleronitrile) (AMVN). 29 26 R. Scudock, M. Roug6e, R. V. Bensasson, M. Evers, and N. Dereu, Photochem. Photobiol. 54, 733 (1991). 27 M. Hayashi and T. F. Slater, Free Radical Res. Comrnun. 2, 179, (1986). 28 V. Narayanaswami and H. Sies, Free Radical Res. Commun. 10, 237 (1990). 29 N. Noguchi, Y. Yoshida, H. Kaneda, Y. Yamamoto, and E. Niki, Biochem. Pharmacol. 44, 39 (1992).
[47]
EBSELEN
481
Further, a role of ebselen in the sequestration of iron is considered minimal. 29
Biological Effects The biological effects of ebselen have been reviewed. 3 Especially in those spaces and compartments where the GPx and PHGPx enzymes are not present at high activity, the low molecular weight compound ebselen as an enzyme mimic might have pathophysiological interest. In numerous in vitro studies, beneficial (protective) effects of ebselen were observed, and sometimes compared to the sulfur analog which was not protective. The antiinflammatory effect was the initial property that attracted attention, 6'3°'31but the wide range of experimental conditions covered indicates a broader range of potential clinical interest. One common basis for many of the diverse effects is that ebselen, by its property as GPx and PHGPx mimic, can lower the peroxide tone, 32 important in the control of lipoxygenases and cyclooxygenases.
Metabolism and Disposition of Ebselen Ebselen metabolism has been studied in isolated perfused liver 33 and in intact rats, pigs, and human volunteers. 34,35 The detected metabolites share the characteristic that the isoselenazole ring is opened by cleavage of the S e - N bond. Apparently, the putative intermediate product, a selenodisulfide with glutathione, S(2-phenylcarbamoylphenylselenyl)glutathione, is labile. The metabolism after ring opening involves methylation to form 2-methylselenobenzanilide, or glucuronidation to form 2-glucuronylselenobenzanilide. Whereas the latter is released into bile, the former undergoes further metabolism. In pigs and humans, the dominant metabolite in plasma and urine is the selenoglucuronide. It is important to note that no unchanged ebselen was detectable in urine, plasma, or bile. 34 The facile ring opening of the isoselenazonone ring to form a selenosulfide is a probable basis for this. Whether ebselen is attacked by sulfhydryl compounds already in the stomach or intestine 3o M. J. Parnham, S. Leyck, N. Dereu, J. Winkelmann, and E. Graf, Adv. Inflammation Res. 10, 397 (1985). 3~ M. J. Parnham and E. Graf, Biochem. Pharmacol. 36, 3095 (1987). 32 M. E. Hemler, H. W. Cook, and W. E. M. Lands, Arch. Biochem. Biophys. 193,340 (1979). 33 A. Miiller, H. Gabriel, H. Sies, R. Terlinden, H. Fischer, and A. R6mer, Biochem. Pharmacol. 37, 1103 (1988). 34 H. Fische r, R. Terlinden, J. P. L6hr, and A. R6mer, Xenobiotica 18, 1347 (1988). 35 H. Sies, in "Selenium in Biology and Medicine" (A. Wendel, ed.), p. 153. SpringerVerlag, Heidelberg, 1989.
482
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[48]
before absorption or while transported through the mucosa is not yet known. In model studies using bovine serum albumin, rapid binding of ebselen only to albumin was observed. 36 In preliminary work on human plasma, we detected radioactivity from labeled ebselen only on albumin and negligible amounts with the globulin fractions on gel electrophoresis. It is assumed that ebselen binds to the reactive thiol group at cysteine34 in albumin, the location of bound cysteine and glutathione. 37 Thus, the current concept of transport of ebselen in the organism is that it is bound to proteins and that there is an interchange with low molecular weight thiols within cells and tissues. It is important to note that some of the experimental observations obtained with ebselen in oitro should be verified by employing ebselen in a protein-bound form (e.g., as the albumin complex). It is likely that some of the inhibitory effects of ebselen reported for isolated enzymes simply reflect the high reactivity of ebselen with protein thiols, and in vivo there was no detectable free ebselen. 33'34 Ebselen in the selenol form can be a substrate for pig liver flavin-containing monooxygenase. 38 An early observation by Wendel et al. 2 was that the selenium moiety in ebselen was not bioavailable; in selenium-deficient animals the selenoprotein GPx could not be augmented by ebselen, but it was by selenite. The property of not entering the body pool of selenium but rather being metabolized as explained earlier may explain the lack of toxicity observed in experimental studies. 36 H. Nomura, H. Hakusui, and T. Takegoshi, in "Selenium in Biology and Medicine" (A. Wendel, ed.), p. 189. Springer-Verlag, Heidelberg, 1989. 37 T. Peters, Jr., Ado. Protein Chem. 37, 161 (1985). 38 D. M. Ziegler, P. Graf, L. L. Poulsen, W. Stahl, a n d H . Sies, Chem. Res. Toxicol. 5, 163 (1992).
[48] N - A c e t y l c y s t e i n e B y PETER MOLDI~US and IAN A. COTGREAVE
Introduction N-Acetylcysteine (NAC) (Fig. 1) is a thiol-containing compound which has been used in clinical practice for nearly 40 years. Historically, NAC was introduced for the treatment of congestive and obstructive lung diseases, primarily those associated with hypersecretion of mucus, for example, chronic bronchitis and cystic fibrosis. NAC has also been used for over 20 years as the drug of choice in the treatment of paracetamol intoxiMETHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
482
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[48]
before absorption or while transported through the mucosa is not yet known. In model studies using bovine serum albumin, rapid binding of ebselen only to albumin was observed. 36 In preliminary work on human plasma, we detected radioactivity from labeled ebselen only on albumin and negligible amounts with the globulin fractions on gel electrophoresis. It is assumed that ebselen binds to the reactive thiol group at cysteine34 in albumin, the location of bound cysteine and glutathione. 37 Thus, the current concept of transport of ebselen in the organism is that it is bound to proteins and that there is an interchange with low molecular weight thiols within cells and tissues. It is important to note that some of the experimental observations obtained with ebselen in oitro should be verified by employing ebselen in a protein-bound form (e.g., as the albumin complex). It is likely that some of the inhibitory effects of ebselen reported for isolated enzymes simply reflect the high reactivity of ebselen with protein thiols, and in vivo there was no detectable free ebselen. 33'34 Ebselen in the selenol form can be a substrate for pig liver flavin-containing monooxygenase. 38 An early observation by Wendel et al. 2 was that the selenium moiety in ebselen was not bioavailable; in selenium-deficient animals the selenoprotein GPx could not be augmented by ebselen, but it was by selenite. The property of not entering the body pool of selenium but rather being metabolized as explained earlier may explain the lack of toxicity observed in experimental studies. 36 H. Nomura, H. Hakusui, and T. Takegoshi, in "Selenium in Biology and Medicine" (A. Wendel, ed.), p. 189. Springer-Verlag, Heidelberg, 1989. 37 T. Peters, Jr., Ado. Protein Chem. 37, 161 (1985). 38 D. M. Ziegler, P. Graf, L. L. Poulsen, W. Stahl, a n d H . Sies, Chem. Res. Toxicol. 5, 163 (1992).
[48] N - A c e t y l c y s t e i n e B y PETER MOLDI~US and IAN A. COTGREAVE
Introduction N-Acetylcysteine (NAC) (Fig. 1) is a thiol-containing compound which has been used in clinical practice for nearly 40 years. Historically, NAC was introduced for the treatment of congestive and obstructive lung diseases, primarily those associated with hypersecretion of mucus, for example, chronic bronchitis and cystic fibrosis. NAC has also been used for over 20 years as the drug of choice in the treatment of paracetamol intoxiMETHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
1"48]
N-ACETYLCYSTEINE
H
HO\
I
483
H
I
o C-C -c-s., H-N
H I
H3C/C">O FIG. 1. Chemical structure of N-acetylcysteine.
cation. More recently NAC has enjoyed a renaissance, and its applications now include use in the treatment of pulmonary oxygen toxicity, adult respiratory distress syndrome (ARDS), and, potentially, cases of human immunodeficiency virus (HIV-I) infections. The above diversity of pharmacological applications of NAC is rather unique and is due, in the main, to the multifaceted chemical properties of the cysteinyl thiol of the molecule. These include its nucleophilicity and redox activity, providing "scavenger" and antioxidant properties, and also its ability to undergo transhydrogenation or thiol-disulfide exchange (TDE) reactions with other thiol redox couples. Advances in the clinical use of NAC stem mainly from the development of suitable analytical techniques for the analysis of NAC and potential metabolites in biological systems. These techniques are being used to determine the human pharmacokinetic behavior of NAC, as well as its metabolic disposition. This chapter deals first with the analytical concepts which are at present available for the determination of NAC and metabolites in biological fluids. We then briefly detail the available data on the disposition of NAC in humans before reviewing the biological properties of NAC. Finally we review the present use and future possibilities of this thiol drug in clinical practice. Analysis of N-Acetyleysteine in Biological Systems In addition to the conventional restrictions of assay sensitivity, reproducibility, and precision, the assay of NAC in biological systems presents several additional problems to the analyst. First, the compound has few physical properties, apart from redox activity, which allow its direct detection. Second, as a typical thiol NAC may oxidize to disulfide species or undergo transhydrogenation reactions with other thiol redox couples, resulting in the potential introduction of artifacts during the manipulation of biological samples. Finally, biological systems contain low molecular weight thiols, such as cysteine and glutathione, which possess physical and chemical properties similar to those of NAC and are its primary
484
[48]
ANTIOXIDANT CHARACTERIZATION AND ASSAY TABLE I REAGENTS USED TO DERIVATIZE N-ACETYLCYSTEINE AND TO FACILITATE ITS ASSAY Reagent N-(1 -Pyrene)maleimide N-(7-Dimethylamino-4-methylcoumarinyl)maleimide 4-(Aminosulfonyl)-7-fluoro-2,1,3-benzoxadiazole Ammonium 7-fluoro-2,1,3-benzoxadiazole 2,4-Dinitro- l-fluorobenzene Monobromobimane o-Phthalaldehyde
Ref. 1 1 2 3 4 5 6
metabolities. Thus, the ideal analytical approach for NAC in complex biological systems must "trap" the reduced form of NAC as quickly as possible. This has been mostly accomplished by the quantitative chemical derivatization of NAC with a variety of electrophilic agents, resulting in stable adducts. These adducts are often more readily amenable to chromatographic separation than the parent compound, mostly on reversed-phase high-performance liquid chromatography (HPLC) columns, and exhibit physical properties, such as fluorescence, which facilitate detection and quantitation. Table I lists some of the reagents which have been used to derivatize NAC and form the basis of an analytical method for NAC in biological systems.l-6 Most of the reagents react with the thiolate anion of NAC, and the derivatizations are therefore conducted at alkaline pH. Unfortunately the rate of oxidation of NAC rapidly increases with increasing pH so that the derivatizing agent must react quantiatively and rapidly with the available NAC in the sample and preferably in situ without prior extraction of the thiol. Similarly, owing to the presence of endogenous low molecular weight and protein thiols in biological samples, enough of the reagent must be added to the sample to allow quantitative recovery of the NAC adduct in the presence of these other thiols. Finally, the physical properties of the adducts (e.g., fluorescence) should impart enough sensitivity to allow the accurate quantitation of NAC in body fluids during pharmacological usage. The reagents in Table I have all fulfilled the above criteria and been successfully applied to the analysis of reduced NAC in blood, plasma, or urine. However, each reagent presents different practical restrictions. I B. K~gedahl and M. K~llberg, J. Chromatogr. 229, 409 (1982). 2 T. Toyo6ka and K. Imai, J. Chromatogr. 282, 495 (1983). 3 T. Toyo6ka and K. Imai, Anal. Chem. 56, 2461 (1984). 4 p. A. Lewis, A. J. Woodward, and J. Maddock, J. Chromatogr. 327, 261 (1985). I. A. Cotgreave and P. Mold~us, Biopharm. Drug Dispos. 8, 365 (1987). 6 B. Gabard and H. Masher, Biopharm. Drug Dispos. 12, 343 (1991).
[48]
N-ACETYLCYSTEINE
485
Any complete assay protocol for NAC in biological systems must include the ability to determine the entire redox status of the thiol. This must include oxidized NAC in homo- and heterologous low molecular weight disulfides and NAC in mixed disulfides with protein. This has generally been accomplished by separating the soluble and proteinaceous portions of the sample, often by acid precipitation, and subsequent reduction of these components with reducing agents such as dithiothreitol (DTT), followed by derivatization of the "extra" NAC released from these oxidized forms. As an illustration of a typical complete analytical procedure for NAC, we discuss the methodologies used in our laboratories. 5
Analysis of Reduced and Oxidized N-Acetylcysteine in Biological Systems Using Monobromobimane The monobromobimane (mBBr) reagent offers a number of advantages over other reagents for the analysis of NAC in biological systems. First, the reagent reacts extremely rapidly, selectively, and stoichiometrically with thiolate anions to yield adducts which are very stable and highly fluorescent. Second, the reagent is equally reactive with endogenous thiols as it is with NAC, and the thiol-mBBr adducts are readily separated on conventional reversed-phase HPLC. This ensures the simultaneous analysis of NAC and potential thiol-containing metabolites. Finally, the reagent is membrane-permeable, allowing the in situ derivatization of NAC in aqueous-, cell-, and tissue-based samples using the same methodologies and without prior manipulation of the sample. Reduced N-Acetylcysteine. The sample (I00/zl of plasma, blood, urine, cell suspension, or tissue homogenate) is mixed with 100 /A of 8 mM mBBr in 50 mM N-ethylmorpholine buffer, pH 8.0, and allowed to react for 2 min in the dark. The sample is then acidified by the addition of 10 /zl of 100% aqueous trichloroacetic acid (TCA), the precipitated protein removed by centrifugation at 3000 g for 5 min at room temperature, and 25/zl of the supernatant applied to the HPLC column (150 × 4.5 mm, 3 /zm ODS reversed phase, Supelco, Bellefonte, PA). The mBBr derivatives of NAC, cysteine, and GSH are then eluted with isocratic 9% aqueous acetonitrile containing 0.25% acetic acid and 0.25% perchloric acid, pH 3.7. Typical HPLC traces from the analysis of reduced NAC in standards and in plasma following the intravenous administration of NAC are shown in Fig. 2. The absolute recovery of reduced NAC using these derivatization conditions is close to 100% over the range I/zM to 1 mM in plasma, and the absolute sensitivity of the method is 50 pmol NAC per sample or 6 pmol NAC on-column. The reproducibility of the assay is within 3-5%
486
ANTIOXIDANT CHARACTERIZATION AND ASSAY
a BE
C
b B
CDBE
I
I
9
18
~
[48]
B
B
!
i
9 18 Retention time (min)
B
I
I
|
0
9
18
FIG. 2. Typical HPLC traces obtained from the analysis of NAC standards and NAC in plasma with mBBR. (a) NAC (100/~M); (b) NAC, cysteine, and GSH (all 100/~M); (c) plasma obtained 10 min after the infusion of 200 mg NAC to a human volunteer. Peak C, Cysteine-mBBr adduct; D, GSH-MBBR adduct; B, reagent hydrolysis peaks; E, NAC-mBBr adduct.
over the range 1 ~M to 1 mM and the precision of assay 97% over this range. 5 Oxidized N-Acetylcysteine. Total NAC (reduced plus oxidized) can be determined following reduction of the sample (100/xl) with 100 mM DTT (5/zl) for 30 min at 37°. The total NAC pool is then derivatized with excess mBBr [100/zl of a 20 mM (or 30 mM in the case of plasma) solution in 50 mM N-ethylmorpholine buffer, pH 8.0] and analyzed as above. The reduction methods employed ensure quantitative recovery of NAC, 5 as well as of cysteine and glutathione, 7 from disulfides. The difference be7 I. A. Cotgreave and P. Mold6us, Biochem. Biophys. Methods 13, 231 (1986).
[48]
N-ACETYLCYSTE,INE
487
tween the free NAC levels and the total indicates the concentration of oxidized NAC in the sample, but gives no indication of the nature of the disulfides. Soluble disulfides containing NAC can be determined by the analysis of the free and total NAC levels in the supernatants of biological samples pretreated with TCA as follows. The sample (1 ml) is mixed with 50/zl of 100% TCA, vortexed, and centrifuged, and the supernatant is neutralized with NaHCO3 and analyzed for free and total NAC as above. Protein-NAC mixed disulfides can be determined from the precipitated protein by washing (with three 1-ml portions of 5% TCA) and redissolving the pellet in 1 ml of 1% sodium dodecyl sulfate (SDS) with neutralization. An aliquot (100/xl) is reduced with DTT and the released NAC determined as above. N-Acetylcysteine Pharmacokinetics in Humans The above methodologies provided the first controlled human pharmacokinetic data for NAC. 7-9 These data revealed that following a single intravenous infusion of 200 mg NAC the peak plasma level (-200 ~M) declined rapidly, and biphasically (aT1/aa~dbTl/2 values were 6 min and 40 min, respectively). NAC also rapidly formed disulfides in plasma following infusion which prolonged the existence of the drug in plasma to up to 6 hr. On the other hand, following oral ingestion of 200 mg NAC, the free thiol was undetectable, with low levels of oxidized NAC detected up to 3 hr after the dose. These data indicated that the drug was less than 5% bioavailable from the oral formulation and were quickly confirmed by other investigators in larger populations. 10,11This low oral bioavailability probably underlies the observation that free NAC is still undetectable and total NAC does not accumulate in plasma following repeated oral doses of 200 mg NAC to healthy volunteers. 12 Further pharmacokinetic data have been made available for different oral formulations of NAC and different dose regimes for the c o m p o u n d . 13-17 8 p. Mold6us, I. A. Cotgreave, and M. Berggren, Respiration 50, s31 (1986). 9 1. A. Cotgreave, R. Grafstrfm, and P. Mold6us, Bull. Eur. Physiopathol. Respir. 22, s263 (1986). I0 L. BorgstrOm, B. K~gedahl, and O. Paulsen, Eur. J. Clin. Pharmacol. 31, 217 (1986). IIB. Olsson, M. Johansson, J. Gabrielsson, and P. Bolme, Eur. J. CHn. Pharmacol. 34, 77 (1986). 12 I. A. Cotgreave, A. Eklund, K. Larsson, and P. Mold6us, Fur. J. Respir. Dis. 70, 73 (1987). 13 M. De Bernardi di Valserra, G. Mautone, E. Barindelli, P. Lualdi, F. Feletti, and M. R. Galmozzi, Eur. J. Clin. Pharmacol. 37, 419 (1989). 14 L. De Carro, A. Ghizzi, R. Costa, A. Longo, G. P. Ventresca, and E. Lodola, Arzneim. Forsch. 39, 382 (1989). 15 L. Borgstr6m and B. K~tgedahl, Biopharm. Drug Dispos. 11, 131 (1990). t6 W. A. Watson and P. E. McKinney, DICP, Ann. Pharmacther. 25, 1081 (1991). 17 M. R. Holdiness, Clin. Pharmacokinet. 20, 123 (1991).
488
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[48]
Metabolism of N-Acetylcysteine
It has been suggested that the low oral bioavallability of NAC is due to extensive first-pass metabolism in the body. Early work had demonstrated that the liver might be a component in this as NAC was effectively deacetylated by isolated hepatocytes and supported GSH synthesis in these cells. ~s Additionally, it was demonstrated that intestinal deacetylation of NAC was extensive in rats and that the major metabolites entering the hepatic portal circulation were cysteine, cystine, and sulfite) 9 This was subsequently confirmed in homogenates of rat intestine and by in situ peffusion of the rat small intestine by Sj6din e t al. 2° Homogenates of rat lung and liver also deacetylate NAC, as do human liver homogenates. The deacetylation of NAC is regiospecific, with the D isomer being poorly metabolized in rat and human tissues. Additionally, the homodisulfide of L-NAC is poorly deacetylated in rat and human tissues. E° Work has also shown that human endothelial cells are able to deacetylate NAC and utilize the liberated cysteine for the support of intracellular GSH biosynthesis. This indicates that the human endothelium may contribute to the metabolic clearance of intravenously administered NAC and may supply NAC-defived cysteine equivalents to other cell types in the underlying tissue or in the circulation. 2~ In addition, it was noted that the deacetylation of NAC during first-pass metabolism resulted in an elevation of circulatory levels of GSH in the rat. 19This phenomenon has since been demonstrated in humans by several groups following repeated administration of the drug.~2,22,23 N-Acetylcysteine as Antioxidant There is increasing evidence that oxidants play a major role in the development of a variety of human disease states. The sources of these oxidants vary and include activated inflammatory cells, cells undergoing redox cycling of xenobiotics, and environmental media such as cigarette smoke. Because NAC is used therapeutically in several disorders related to these sources it has been suggested to function as an antioxidant. Thus, NAC has been demonstrated to reduce effectively free radical species and other oxidants. The interaction with free radical species results in the 18 H. Thor, P. Mold~us, and S. Orrenius, Arch. Biochem. Biophys. 192, 405 (1979). t9 I. A. Cotgreave, M. Berggren, T. J. Jones, J. Dawson, and P. Mold~us, Biopharm. Drug Dispos. 8, 377 (1987). 2o K. Sj6din, E. Nilsson, A. Hallberg, and A. Tunek, Biochem. Pharmacol. 38, 3981 (1989). 21 I. A. Cotgreave, P. Mold~us, and I. Schuppe, Biochem. Pharmacol. 42, 13 (1991). 22 j. M. Burgunder, A. Varriale, and B. Lauterberg, Eur. J. Clin. Pharmacol. 36, 127 (1989). M. M. E. Bridgeman, M. Marsden, W. MacNee, D. C. Flenley, and A. P. Ryle, Thorax 46, 39 (1991).
[48]
N-ACETYLCYSTEINE
489
intermediate formation of NAC thiyl radicals, with NAC disulfide as the major end product. 8 Of special interest is the interaction of NAC with reactive oxygen intermediates such as the superoxide anion (O2~), hydrogen peroxide (H202), hydroxyl radical (.OH), and hypochlorous acid (HOC1), all of which are released by inflammatory cells. NAC has been shown to reduce •O H , H 2 0 2 , and H O C I . 8'24 NAC does not, however, appear to interact with 02 ~ to any major extent. The interaction of NAC with HzO 2 is fairly slow, whereas the interaction between .OH and NAC is extremely rapid, with a calculated rate constant of 1.36 x 101° M -1 sec -~. NAC is also a powerful scavenger of HOCI, and low concentrations of NAC have been demonstrated to protect cells from HOCl-induced toxicity. NAC has also been shown to protect against HOCl-induced contraction of guinea pig tracheal smooth muscle 25and to protect against HOCl-induced inactivation of arantiproteinase. 24 All of the studies in which NAC has been shown to interact directly with oxidants and protect against oxidant-induced toxicity have been performed in vitro, and the significance of a direct antioxidative effect of NAC in vivo is questionable, particularly in view of the pharmacokinetics of the reduced form of the drug. NAC may also exert its antioxidant effect indirectly by facilitating GSH biosynthesis and supplying GSH for glutathione peroxidase-catalyzed reactions. There are however, no data to support such a mechanism. An interesting possibility is to use NAC together with a synthetic glutathione peroxidase mimetic. We have, for instance, showed that NAC serves as a thiol substrate to the glutathione peroxidase mimic ebselen and that this combination protects isolated hepatocytes from oxidative damage. 26 Whether such a mechanism may occur in vivo remains to be established. N-Acetylcysteine as an Antimutagen and Anticarcinogen N-Acetylcysteine has been shown to be an antimutagen and anticarcinogen both in vitro and in vivo. NAC has thus been demonstrated to inhibit the mutagenicity of both directly acting carcinogens and procarcinogens in in vitro mutagenicity tests. 27 NAC has also been demonstrated to inhibit the in vivo DNA-adduct formation after administration of acetyl24 O. I. Aruoma, B. Halliwell, B. M. Hoey, and J. Butler, Free Radical Biol. Med. 6, 593 (1989). 25 A. Bast, G. R. M. M. Haenen, and C. J. A. Doelman, Am. J. Med. 91, s2 (1991). 26 I. A. Cotgreave, M. S. Sandy, M. Berggren, P. W. Mold6us, and M. T. Smith, Biochem. Pharmacol. 36, 2899 (1987). 27 S. De Flora, A. Izzotti, F. D'Agostini, and C. F. Cesarone, Am. J. Med. 91, s122 (1991).
490
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[48]
aminottuorene (AAF) or benzo[a]pyrene. 27 NAC is also able to protect certain enzymes involved in DNA replication and repair. 28 N-Acetylcystein has clear anticarcinogenic effects in various animal models and has been shown to decrease the formation of lung tumors in urethane-treated mice, 29prevent the formation of AAF-induced sebaceous squamocellular carcinomas of the zymbal glands 3°, and hydrazine-induced colon cancer in rats. 31 Present and Possible Future Therapeutic Use of N-Acetylcysteine Apart from the disorders detailed in the Introduction in which NAC is presently utilized, treatment with NAC may prove advantageous, based on animal experiments, in idiopathic pulmonary fibrosis, emphysema, and ischemia-repeffusion injury. Although NAC has proven efficacy in humans and animals in a number of diseases, the mechanism of the therapeutic action of NAC has only been elucidated in a few cases. In most of these cases decreased levels of GSH and thereby electrophilic/oxidative injury have been suggested or actually demonstrated, and the mechanism of the effects of NAC has, thus, been as a supplier of cysteine for GSH, thereby acting indirectly as an antioxidant or scavenger of electrophiles. Pharmacokinetic considerations suggest that direct interaction of NAC itself could probably only occur when NAC is given intravenously or by inhalation. One of the best studied examples of this mechanism of action of NAC is its activity as an antidote against acetaminophen toxicity through supporting GSH biosynthesis. Acetaminophen induces hepatic necrosis when taken in overdose. This toxicity is due to a reactive metabolite of acetaminophen, N-acetyl-p-benzoquinoneimine (NAPQI). NAPQI is normally detoxified by conjugation with GSH but in cases of overdose GSH becomes depleted, and there is a lack of cysteine for resynthesis which is compensated for by NAC. 32 The mechanism of action of NAC as a mucolytic agent in the treatment of chronic bronchitis and cystic fibrosis is established for NAC administered by inhalation. NAC is then present in high concentrations in the all'ways and readily reduces disulfide bridges between glycopolypeptides 28 C. F. Cesarone, L. Scarabelli, P. Giannoni, and M. Orunesu, Murat. Res. 245, 157 (1990). 29 S. De Flora, M. Astengo, D. Serra, and C. Bennicelli, Cancer Lett. 32, 235 (1986). 3o C. F. Cesarone, L. Scarabelli, M. Orunesu, M. Bagnasco, and S. De Flora, In Vivo 1, 85 (1987). 31 M. Wilpart, A. Speder, and M. Roberfroid, Cancer Lett. 31, 319 (1986). 32 p. Mold6us, in "Drug Reactions and the Liver" (M. Davis, J. M. Tredger, and R. Williams, eds.), pp. 114-156. Pitman Medical, London, 1981.
[48]
N-ACETYLCYSTEINE
491
in the protein complexes of mucus. 33 NAC has also proved to be effective when given orally, 34 but in this case the mechanism of action of NAC is not known. N-Acetylcysteine is currently being tested in the treatment of ARDS. It appears that a large part of the pathophysiology associated with ARDS is associated with neutrophil infiltration into the lung and the release of cytotoxic mediators, including oxidants. Both in animal models and in humans NAC has been shown to decrease significantly the severity of several pathophysiological effects and enhance survival; however, the mechanism of action is not clear. When NAC is given intravenously it has been suggested to act as an antioxidant. Indeed, plasma levels of reduced glutathione have been demonstrated to be decreased in ARDS patients, and, as stated above, treatment with NAC has been shown to result in increased plasma and red cell glutathione levels. 35 Potential therapeutic usages of NAC are in preventing the development of nitrate tolerance and as an immunomodulating agent. Patients treated with organic nitrates frequently develop tolerance to the effects of the drugs. This tolerance has been suggested to be due to depletion of intracellular sulfhydryl groups in vascular smooth muscle. There are numerous investigations where NAC, as well as other thiols, have been used in order to reverse this tolerance. 36-38Even though the results are somewhat contradictive, it is evident that NAC prevents the development of tolerance or at least partly restores the effect the organic nitrates. 38'39 The mechanism by which NAC reverses the tolerance is not known. Interestingly, only the L isomer of NAC is active. Because L-NAC is further metabolized to cysteine and GSH whereas D-NAC is not, 4° these data indicate that an enzymatic step is required for activity. The mechanism may also involve the intra- or extracellular formation of S-nitrosothiol and/or NO. 41 33 A. L. Sheffner, E. M. Medler, L. W. Jacobs, and H. P. Sarett, Am. Rev. Respir. Dis. 90, 721 (1964). 34 G. Boman, U. B~icker, S. Larsson, B. Melander, and L. W~hlander, Eur. J. Respir. Dis. 64, 405 (1983). 35 G. R. Bernard, Am. J. Med. 91, s54 (1991). 36 j. Torresi, J. D. Horowitz, and G. J. Dusting, J. Cardiooasc Pharmacol. 7, 777 (1985). 37 H. Tsuneyoshi, N. Akatsuka, M. Ohno, K. Hara, M. Ochial, and M. Moroi, Jpn. Heart J. 30, 733 (1989). 38 j. Abrams, Am. J. Med. 91, s106 (1991). 39 j. D. Horowitz, Am. J. Med. 91, sl13 (1991). 40 C. M. Newman, J. B. Wassen, G. W. Taylor, A. R. Boobis, and D. S. Davies, Br. J. Pharmacol. 99, 825 (1990). 41 H. L. Fung, S. Chong, E. Kowaluk, K. Hough, and M. Kakemi, J. Pharmacol. Exp. Ther. 245, 524 (1989).
492
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[49]
Additional studies have shown that intracellular sulfhydryl concentrations in peripheral blood mononuclear cells and monocytes are decreased in patients infected with HIV-1 and in individuals with immunodeficiencies. 42 It is also clear that T lymphocytes are dependent on supply of extracellular cysteine for intracellular cysteine and GSH biosynthesis since these cells exhibit poor uptake of cystine. 43 Furthermore, cysteine appears to play a pivotal role in the regulation of T-cell-mediated immune responses. Cysteine has, for instance, been found to inhibit the expression of NF-rB-dependent genes. 44 On the other hand, there are several other processes that are positively regulated by cysteine. Cysteine and NAC have also been shown to inhibit HIV-1 replication in infected cells as well as in normal peripheral blood mononuclear c e l l s . 45 These data indicate a cysteine deficiency in HIV-infected patients which may lead to enhanced HIV replication and overexpression of the NF-KB-dependent genes. Treatment with NAC may thus be considered for HIV-l-infected patients. 46 42 H.-P. Eck, H. Gmiinder, M. Hartmann, D. Ptzoldt, V. Daniel, and W. Dr6ge, Biol. Chem. Hoppe-Seyler 370, 101 (1989). 43 W. Dr6ge, H. P. Eck, H. Gmi~nder, and S. Mihm, Am. J. Med. 91, sl40 (1991). 44 S. Mihm, J. Ennen, U. Pessara, R. Kurth, and W. Dr6ge, AIDS (London) 5, 497 (1991). 45 M. Roederer, P. A. Raju, F. J. Staal, L. A. Herzenberg, and L. A. Herzenberg, AIDS Res. Hum. Retroviruses 7, 563 (1991). 46 M. Roederer, S. W. Ela, F. J. Staal, L. A. Herzenberg, and L. A. Herzenberg, AIDS Res. Hum. Retroviruses 8, 209 (1992).
[49] P r e p a r a t i o n a n d U s e o f G l u t a t h i o n e M o n o e s t e r s By MARY E. ArqDERSON, ELLEN J. LEVY, and ALTON MEISTER Introduction Because glutathione is not effectively transported into most cells, various derivatives of glutathione have been examined in the effort to identify compounds that are well transported and are converted to glutathione after transport.~-9 Glutathione monoesters are useful glutathione delivery 1 R. N. Purl and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 80, 5258 (1983). 2 M. E. Anderson, R. Powffe, R. N. Purl, and A. Meister, Arch. Biochem. Biophys. 239, 538 (1985). 3 V. P. Wellner, M. E. Anderson, R. N. Puff, G. L. Jensen, and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 81, 4732 (1984). 4 j. Martensson and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 86, 471 (1989).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproduction in any form reserved.
492
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[49]
Additional studies have shown that intracellular sulfhydryl concentrations in peripheral blood mononuclear cells and monocytes are decreased in patients infected with HIV-1 and in individuals with immunodeficiencies. 42 It is also clear that T lymphocytes are dependent on supply of extracellular cysteine for intracellular cysteine and GSH biosynthesis since these cells exhibit poor uptake of cystine. 43 Furthermore, cysteine appears to play a pivotal role in the regulation of T-cell-mediated immune responses. Cysteine has, for instance, been found to inhibit the expression of NF-rB-dependent genes. 44 On the other hand, there are several other processes that are positively regulated by cysteine. Cysteine and NAC have also been shown to inhibit HIV-1 replication in infected cells as well as in normal peripheral blood mononuclear c e l l s . 45 These data indicate a cysteine deficiency in HIV-infected patients which may lead to enhanced HIV replication and overexpression of the NF-KB-dependent genes. Treatment with NAC may thus be considered for HIV-l-infected patients. 46 42 H.-P. Eck, H. Gmiinder, M. Hartmann, D. Ptzoldt, V. Daniel, and W. Dr6ge, Biol. Chem. Hoppe-Seyler 370, 101 (1989). 43 W. Dr6ge, H. P. Eck, H. Gmi~nder, and S. Mihm, Am. J. Med. 91, sl40 (1991). 44 S. Mihm, J. Ennen, U. Pessara, R. Kurth, and W. Dr6ge, AIDS (London) 5, 497 (1991). 45 M. Roederer, P. A. Raju, F. J. Staal, L. A. Herzenberg, and L. A. Herzenberg, AIDS Res. Hum. Retroviruses 7, 563 (1991). 46 M. Roederer, S. W. Ela, F. J. Staal, L. A. Herzenberg, and L. A. Herzenberg, AIDS Res. Hum. Retroviruses 8, 209 (1992).
[49] P r e p a r a t i o n a n d U s e o f G l u t a t h i o n e M o n o e s t e r s By MARY E. ArqDERSON, ELLEN J. LEVY, and ALTON MEISTER Introduction Because glutathione is not effectively transported into most cells, various derivatives of glutathione have been examined in the effort to identify compounds that are well transported and are converted to glutathione after transport.~-9 Glutathione monoesters are useful glutathione delivery 1 R. N. Purl and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 80, 5258 (1983). 2 M. E. Anderson, R. Powffe, R. N. Purl, and A. Meister, Arch. Biochem. Biophys. 239, 538 (1985). 3 V. P. Wellner, M. E. Anderson, R. N. Puff, G. L. Jensen, and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 81, 4732 (1984). 4 j. Martensson and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 86, 471 (1989).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproduction in any form reserved.
[49]
PREPARATION AND USE OF GLUTATHIONE MONOESTERS
493
agents because they are readily transported into many types of cells and split intracellularly to form glutathione.* Treatment of cells (in suspension or culture) and of intact animals with glutathione itself may lead to some increase in cellular glutathione levels; such increases are usually due to extracellular breakdown of glutathione into its constituent amino acids and into dipeptides, then transport of these into cells followed by intracellular glutathione synthesis. Studies in which glutathione synthesis is completely inhibited, for example, by treatment with buthionine sulfoximine, show that administration of glutathione leads to little if any increase in cellular glutathione levels. 1-14 Glutathione monoesters are transported into many rodent tissues, including kidney, liver, pancreas, spleen, skeletal muscle, heart, lung, and, in neonatal animals, the lens and brain. Significant transport into adult brain does not occur, but transport into cerebrospinal fluid was found) 5 Oral administration of glutathione monoesters to mice was reported to increase glutathione levels in liver and kidney. 2 Monoesters of glutathione are also transported into human red blood cells, lymphocytes, and fibroblasts; in control studies, glutathione itself was not appreciably transported. 2,3 Administration of glutathione monoesters protects against toxicity due to such compounds as CdCI2 ,~6 HgCI 2 ,17 cisplatin,~8 acetaminophen,l mel5 j. Martensson, A. Jain, W. Frayer, and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 86, 5296 (1989). 6 j. Martensson, R. Steinherz, A. Jain, and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 86, 8727 (1989). 7 j. Martensson, A. Jain, and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 87, 1715 (1990). 8 j. Martensson, J. Han, O. W. Griffith, and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 90, 317 (1993). 9 A. Jain, J. Martensson, E. Stole, P. A. M. Auld, and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 88, 1913 (1991). * The esters discussed here refer to monoesters in which the glycine carboxyl group of giutathione is esterified; assignment was shown by nuclear magnetic resonance (NMR) spectroscopy.~ No preparation of the corresponding pure c~-ester of glutathione has apparently been published. l0 j. K. Dethmers and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 78, 7492 (1981). ii G. L. Jensen and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 80, 4714 (1983). 12 W. A. Abbott, R. J. Bridges, and A. Meister, J. Biol. Chem. 259, 15393 (1984). 13 A. Meister, Pharmacol. Ther. 51, 155 (1991), and references cited therein. 14 A. Meister, in " N e w Trends in Biological Chemistry" (T. Ozawa, ed.), pp. 69-79. Jpn. Sci. Soc. Press, Tokyo, Springer-Verlag, Berlin, 1991. 15 M. E. Anderson, M. Underwood, R. J. Bridge, and A. Meister, FASEB J. 3, 2527 (1989). 16 R. K. Singhal, M. E. Anderson, and A. Meister, FASEB J. 1, 220 (1987). 17 A. Naganuma, M. E. Anderson, and A. Meister, Biochem. Pharm. 40, 693 (1990). 18 M. E. Anderson, A. Naganuma, and A. Meister, FASEB J. 4, 3251 (1990).
494
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[49]
phelan, ~9 and cyclophosphamide,19 as well as against radiation. 2° Buthionine sulfoximine-induced cellular damage (e.g., mitochondrial destruction) is prevented by glutathione monoester treatment, but not, in control studies, by administration of glutathione. 4-9 The facile transport of glutathione monoesters into cells provides a valuable research approach. Glutathione monoesters are not yet available commercially, but simple methods for their synthesis have been published 1'16'21 and are given here. These methods have been used in our laboratory, and the products obtained have been applied in in vitro and in vivo studies.
Procedures
Glutathione Monoethyl Ester Hydrochloride 21
All procedures are carried out in a chemical fume hood. Appropriate precautions are taken for handling compressed hydrogen chloride gas and inflammable solvents. Hydrogen chloride gas is bubbled through an ethanol trap. (The flow of hydrogen chloride must be sufficient to prevent any backup of ethanol into the tank.) The outlet of the trap is connected with a short piece of Tygon tubing to a glass delivery tube, which is directed into a 500-ml round-bottom flask containing 50 ml of cold (00-4 °) anhydrous ethanol; HCI gas is bubbled slowly (about 10 min) into the flask until about 2 g of hydrogen chloride is dissolved in the cold ethanol. The flask and its contents are weighed (before and after) on a pan balance. Glutathione (5 g; 16.25 mmol) is added and dissolved by swirling. The glass stoppered (well greased) flask is allowed to stand on ice. The progress of the reaction is monitored by high-performance liquid chromatography (HPLC) or thinlayer chromatography (TLC) (see below); it is usually complete in 6 to 8 hr. After about 98% of the glutathione has disappeared, cold diethyl ether (400 ml) is added with swirling. This mixture is allowed to stand (4 to 18 hr) at 00-4 ° during which time a white viscous product forms. The ether is decanted, and the product is triturated with cold diethyl ether. A stream of nitrogen gas is blown over the precipitate in a hood to remove most of the residual ether. 19 B. A. Teicher, J. M. Crawford, S. A. Holden, Y. Lin, K. N. S. Cathcart, C. A. Luchette, and J. Flatow, Cancer (Philadelphia) 62, 1275 (1988). 20 M. B. Astor, M. E. Anderson, and A. Meister, Pharmacol. Ther. 39, 115 (1988). 21 M. E. Anderson and A. Meister, Anal. Biochem. 183, 16 (1989).
[49]
PREPARATION AND USE OF GLUTATHIONE MONOESTERS
495
The precipitate is dried overnight in a vacuum desiccator over KOH and P205 ; a dry white fluffy product is obtained. The product is crystallized from the minimal volume (about 5 ml) of warm (40°) distilled deionized water, and ethanol is added (final concentration, -50-60%) until crystallization begins. The mixture is allowed to stand at 4° (16-36 hr). The white crystals formed are washed on a sintered glass funnel with cold ethanol followed by cold ether. The crystals are dried in a vacuum desiccator over K O H and P205. The yield is 5.6 to 6.3 g (80 to 90%). The product contains 98-100% glutathione monoethyl ester (GSH-ME • HCI) as determined by HPLC and gives 100% of the theoretical value with 5,5'-dithiobis(2-nitrobenzoic acid (DTNB)n; little if any impurity (usually glutathione) is present. Glutathione Monoethyl Ester Hemihydrosulfate 21 All procedures with diethyl ether are carried out in a chemical fume hood. Sulfuric acid (2.74 ml; 51 mmol) is added with mixing to 100 ml of anhydrous ethanol in a l-liter flask. (The sulfuric acid should be fresh and of high purity; the ethanol preferably should be obtained from a newly opened bottle.) Glutathione (10 g; 32.5 retool) is added to the flask. The flask is stoppered (well-greased glass stopper) and mechanically shaken in a water bath (31°-34°). The glutathione usually dissolves within 10 rain. The reaction is followed by TLC or HPLC (see below). After 98% of the glutathione disappears (8 to 12 hr), the reaction mixture is cooled on ice. (The reaction can also be carried out at 20° to 25 ° for 16 to 20 hr.) Cold diethyl ether (700 ml) is then added with swirling. A white product forms immediately, and the reaction mixture is allowed to settle (for 4 to 18 hr) at 4 °. The supernatant is decanted, and the precipitate is washed with ether and a stream of dry nitrogen is blown over the product to remove most of the residual ether. The product is dried overnight in a vacuum desiccator over KOH and P205. The fluffy dry product is dissolved in the minimal amount of warm (40°) distilled deionized water (about 15 ml), and ethanol (about 25 ml) is slowly added with swirling until a white cloudy solution persists. After standing at 4° for 18 to 36 hr, the crystals are filtered on a sintered glass funnel and washed with cold ethanol, followed by cold ether. The crystals are dried in a vacuum desiccator over KOH and P 2 0 5 . The yield is I 1.2 to 12.6 g (80-90%). The product contains 98-100% glutathione monoethyl ester (GSH-ME • 0.5H2SO4) as determined by HPLC and gives 100% of 22 A. L. Ellman, Arch. Biochem. Biophys.
82,
70 (1959).
496
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[49]
the theoretical value with DTNB. The product may contain approximately 2% glutathione.
Free Base Glutathione Monoesters Salt-free preparations of glutathione monoester have been obtained after chromatography on IRC-50. 2 However, since the capacity of IRC50 is low, a high capacity lyophilizer is needed since large volumes must be handled. Standard methods for preparing free base amino acid esters 23 may be used with glutathione monoesters. For example, neutralization of the reaction mixture with triethylamine and standing at 4° for 3 to 18 hr yields free base ester; the yield is 70-80% and the thiol content (DTNB) is 95-100% of the theoretical value. Salt-free esters may be obtained by chromatography on a SM-4 (Bio-Rad, Richmond, CA) column; the yield is 80%, and the products give 93-98% of the theoretical reaction with DTNB. The free base also may be obtained from the ester salts by chromatography on Dowex-1 bicarbonate; the yields are about 70% and give 93-98% of the theoretical thiol value with DTNB. The free base has also been obtained by directly mixing the esterification reaction mixture with Dowex-1 (hydroxide).24 The free base form of glutathione monoester is more likely to oxidize on storage than the salt forms.
Recrystallization of Glutathione Monoesters The use of crystalline glutathione esters is strongly recommended; recrystallization may be desirable. Recrystallization of the esters is useful for removing impurities formed during synthesis. Since a small amount of glutathione tends to form during crystallization, the temperature should be kept below 40°. Crystallized preparations of glutathione monoethyl ester hydrochloride have not decomposed significantly when stored for 3 to 9 years in a vacuum desiccator over KOH or P205. Crystalline monoethyl ester hemihydrosulfate or hydrochloride have been obtained by placing the esterification mixture (after completion) at 4° for several days. Crystalline products have also been obtained when the reaction volume is reduced (about 50%) by flash evaporation after the reaction is complete.
Preparation of Other Glutathione Monoesters We have prepared 21 other glutathione monoesters, such as the n-propyl, 2-propyl, n-butyl, isobutyl, tert-butyl, and n-pentyl, by the procedures 23 j. p. Greenstein and M. Winitz, "The Chemistry of the Amino Acids." Wiley, New York, 1961. 24 E. Campbell and O. W. Griflith, Anal. Biochem. 183, 21 (1989).
[49]
PREPARATION AND USE OF GLUTATHIONE MONOESTERS
497
described above with the appropriate alcohol. In general, longer reaction times and lower yields are obtained with the higher alcohols. We have found the monoesters of methanol, ethanol, and 2-propanol to be effective in raising cellular glutathione levels; we have usually used the monoethyl and mono-2-propyl esters.
Thin-Layer Chromatography Thin-layer chromatography (TLC) may be used to give a rapid (-30-45 min) estimate of the progress of the esterification reaction) ,2,21Apply 0.5 or 1/zl each of the reaction mixture, glutathione (100 mM), and glutathione monoester (I00 mM) to TLC plates (MK6F; 1 x 3 inches). Dry the plate with air; do not use heat because heat will drive the reaction on the plate and give erroneous results. The chromatography is carried out in npropanol/acetic acid/water (16 : 3 : 5, v/v). The plate is dried and sprayed with ninhydrin (0.5% w/v in acetone). The colors develop after heating (100°-135 °) for 0.5-3 min. The colors and their intensity should be examined immediately since after longer times the intensities are not proportional to the amounts of compound. The Rf values and colors for glutathione, glutathione monoethyl ester, and glutathione diester are 0.22 (pink), 0.65 (pink), and 0.78 (yellow), respectively.
High-Performance Liquid Chromatography The progress of the esterification reaction, product purity, and amounts of thiols in biological samples may be analyzed by high-performance liquid chromatography (HPLC). Several modifications of the derivatization and chromatographic separation have been described. 21,25-29The following procedure is useful for monitoring the extent of esterification and product purity. Sample (2/zl reaction mixture) is added to 358/~1 of 5-sulfosalicyclic acid [4.31% (w/v) containing 0.5 mM diethylenetriaminepentaacetic acid (DTPA)] and 30 to 120 ~1 DTPA (4 mM; adjust the volumes of DTPA and glacial acetic acid so that the final volume is 600/.d). Monobromobimane (0.1 M in acetonitrile; 10/zl) is added. Tris-HCl (2 M; pH 9; 21°; 100/zl) 25 N. S. Kosower, E. M. Kosower, G. L. Newton, and H. M. Ranney, Proc. Natl. Acad. Sci. U.S.A. 76, 3382 (1979). 26 R. B. Fahey, G. L. Newton, R. L. Dorian, and E. M. Kosower, Anal. Biochem. l U , 357 (1981). 27 M. E. Anderson, this series, Vol. 113, p. 548. 28 M. E. Anderson, in "Coenzymes and Cofactors: Glutathione" (D. Dolphin, R. Poulson, and O. Avramovic, eds.), pp. 339-366. Wiley, New York, 1989. 29 E. Levy, M. E. Anderson, and A. Meister, unpublished (1993).
498
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[49]
is added to bring the pH to about 7.8. The reaction mixture is allowed to stand at room temperature in the dark for 20 min, then glacial acetic acid (10 to 90 ~1; pH -3.9) is added. The samples are stored at - 2 0 ° in the dark until analysis. Standards (glutathione and glutathione monoester) are prepared in 5-sulfosalicyclic acid containing DTPA and derivatized in the same way as the samples. Chromatography is carried out using an Ultrasphere Cls column (5/.~; 4.5 mm × 25 cm) (Beckman, San Remo, CA) with a 1 ml/min flow rate and fluorescent detection [o-phthaldialdehyde (OPA) filters)]. The thiol derivatives are chromatographed using gradients of buffer A (12.8% methanol, 0.25% acetic acid, 86.95% water, pH 3.9 with NaOH) and buffer B (90% methanol, 0.25% acetic acid, 9.75% water, no NaOH, pH -3.9). At sample injection the solvent changes from 100% A to 65% buffer A (35% buffer B) by use of a linear gradient over 25 min. Then, over the next 10 min a linear gradient is run to give 100% buffer B. After 5 min in buffer B, the column is regenerated for another injection by running buffer A for 8 min. The retention times for glutathione, glutathione monoethyl ester, and glutathione diethyl ester are 10, 22, and 35 min, respectively. Chromatography is carried out at room temperature; there are retention time variations with changes in temperature, buffer lots, Cls columns.
Metal Impurities Previously we noted that occasional preparations of glutathione monoester hydrochloride w e r e t o x i c . El Because it is likely that some iron from the hydrogen chloride tank was carded over into the cold alcohol, we added an "alcohol trap"; since this modification was adapted we observed no immediate lethal rodent toxicity. We have noted also that the quality of water used in glassware washing, in recrystallization, and in chromatography of the esters is of major importance. It is well known that glutathione forms complexes with many metals3°; thus, the use of distilled and deionized water is highly recommended. We have analyzed29 several batches of glutathione monoesters by atomic absorption analysis and determined that little metal is found when the "ethanol trap" procedure is used in the hydrogen chloride synthesis, but some metal, primarily copper, is found in some hydrosulfate preparations. In general, free base ester preparations obtained by adding Dowex-1 hydroxide to the reaction mixture were found to have the greatest metal contamination. 30 D. L. Rabenstein, in "Glutathione: Chemical, Biochemical and Medical Aspects" (D. Dolphin, O. Avramovic, and R. Poulsen, eds.), Part A, pp. 147-186. Wiley, New York, 1989.
[50]
GLUTATHIONE DIETHYL ESTER AND DERIVATIVES
499
It is useful to analyze batches of glutathione monoesters for metals by atomic absorption spectroscopy. A practical and highly sensitive biological indication of the presence of significant amounts of metal ions can be obtained by incubating erythrocyte suspensions with glutathione monoester preparations and measuring the level of thiol (DTNB) initially and after several hours. Typically the presence of significant amounts of metal ions is indicated by a decrease in thiol level. 29 Some preparations of glutathione monoester, which are nontoxic when given to mice, nevertheless produce a decrease in thiol in this and similar in vitro tests. Thus, it appears that small amounts of metal ions can be better tolerated in vivo than in vitro. Comments on in Vitro and in Vivo Experiments It is recommended that glutathione monoester salts be dissolved in distilled, deionized water and carefully adjusted to pH 6.5-7.2 immediately before use. The solutions usually contain a small amount of glutathione because the ester is slowly hydrolyzed. We have seen little oxidation of glutathione monoesters when prepared in distilled, deionized water at pH 6.5 to 6.8 and kept for several hours. However, when the neutralized monoesters are added to cell culture media containing serum and other components, more rapid oxidation occurs.
[50] P r e p a r a t i o n a n d P r o p e r t i e s o f G l u t a t h i o n e D i e t h y l E s t e r and Related Derivatives
By
E L L E N J. LEVY, MARY E. ANDERSON, a nd ALTON MEISTER
Introduction Glutathione diethyl ester, like glutathione mono(glycyl) esters, is an effective glutathione delivery compound. When mice are injected intraperitoneaUy with glutathione diethyl ester, the tissue levels of glutathione increase to about the same extent as found after injection of glutathione monoethyl ester. That essentially equivalent effects are found after administration of the monoester and the diester is explained by the finding that mouse blood plasma contains appreciable amounts of glutathione diester ot-esterase, which rapidly converts the diester to the monoester. On the other hand, human plasma does not have this esterase; for this reason, glutathione diethyl ester would be expected to be stable in human blood METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproduction in any formreserved.
[50]
GLUTATHIONE DIETHYL ESTER AND DERIVATIVES
499
It is useful to analyze batches of glutathione monoesters for metals by atomic absorption spectroscopy. A practical and highly sensitive biological indication of the presence of significant amounts of metal ions can be obtained by incubating erythrocyte suspensions with glutathione monoester preparations and measuring the level of thiol (DTNB) initially and after several hours. Typically the presence of significant amounts of metal ions is indicated by a decrease in thiol level. 29 Some preparations of glutathione monoester, which are nontoxic when given to mice, nevertheless produce a decrease in thiol in this and similar in vitro tests. Thus, it appears that small amounts of metal ions can be better tolerated in vivo than in vitro. Comments on in Vitro and in Vivo Experiments It is recommended that glutathione monoester salts be dissolved in distilled, deionized water and carefully adjusted to pH 6.5-7.2 immediately before use. The solutions usually contain a small amount of glutathione because the ester is slowly hydrolyzed. We have seen little oxidation of glutathione monoesters when prepared in distilled, deionized water at pH 6.5 to 6.8 and kept for several hours. However, when the neutralized monoesters are added to cell culture media containing serum and other components, more rapid oxidation occurs.
[50] P r e p a r a t i o n a n d P r o p e r t i e s o f G l u t a t h i o n e D i e t h y l E s t e r and Related Derivatives
By
E L L E N J. LEVY, MARY E. ANDERSON, a nd ALTON MEISTER
Introduction Glutathione diethyl ester, like glutathione mono(glycyl) esters, is an effective glutathione delivery compound. When mice are injected intraperitoneaUy with glutathione diethyl ester, the tissue levels of glutathione increase to about the same extent as found after injection of glutathione monoethyl ester. That essentially equivalent effects are found after administration of the monoester and the diester is explained by the finding that mouse blood plasma contains appreciable amounts of glutathione diester ot-esterase, which rapidly converts the diester to the monoester. On the other hand, human plasma does not have this esterase; for this reason, glutathione diethyl ester would be expected to be stable in human blood METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by Academic Press, Inc. All rights of reproduction in any formreserved.
500
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[50]
and could serve as a cellular glutathione delivery agent for human cells. Studies indicate that glutathione diethyl ester is effectively transported into human cells) It is evident that neither rats nor mice are good experimental models for studies on glutathione diester. However, studies on hamsters, guinea pigs, rabbits, and sheep have shown that plasma from these species lacks glutathione diethyl ester a-esterase; thus, these animals are potentially useful as experimental models for studies on the in vivo disposition of glutathione diethyl ester. Conversion of Glutathione Diester to Glutathione Monoester in Mouse Plasma The conversion reaction can be shown in vitro in mixtures containing mouse or rat blood plasma (or serum) and glutathione diester. The reaction mixtures (final volume, 0.4 ml) contain glutathione diethyl ester (8 mM), phosphate-buffered saline (pH 7.2) containing 16 mM EDTA, and mouse or rat blood plasma (I-20/zl). After incubation at 37° for 0-15 min, the samples are treated with 5-sulfosalicylic acid (final concentration, 3.3% w/v) and analyzed for thiols by high-performance liquid chromatography (HPLC).I Under these conditions, I/zl of mouse plasma splits 4 nmol of glutathione diester to mono(glycyl) ester per minute. This reaction can be demonstrated in vivo by injecting mice intraperitoneally with glutathione diethyl ester at a dose of 5 mmol/kg of body weight; after 15 min, no diester is found in the plasma, but very high levels (-1500 /zM) of glutathione mono(glycyl) ester are found. Plasma glutathione diester a-esterase thus far has been found only in mice and rats. Transport of Glutathione Diethyl Ester into Human Cells Glutathione diethyl ester is rapidly transported into a number of different human cell types, including erythrocytes, peripheral mononuclear cells, purified T cells, cultured skin fibroblasts, and cultured human ovarian cell lines. In each of these studies it was found that application of glutathione diethyl ester led to efficient transport of the ester, which was more rapid than found in comparable studies with glutathione monoethyl ester. Studes on uptake and efflux of the diester and monoester in erythrocytes showed that the esters can cross the cell membrane in both directions and that transport of the diester is substantially more rapid than that of the monoester. The transported diester is rapidly split intracellularly to the monoester. Thus, treatment of various human cells with glutathione I E. J. Levy, M. E. Anderson, and A. Meister, Proc. Natl. Acad. Sci. U.S.A. 90, 9171 (1993).
[50]
GLUTATHIONE DIETHYL ESTER AND DERIVATIVES
501
diesters led to very high levels ofintracellular monoester, which are hydrolyzed, leading to formation of increased cellular levels of glutathione. The effective uptake of diester and its cleavage to the monoester provide a system in which the transported glutathione diester is effectively trapped within the cells as the monoester, which can serve to supply the cell with glutathione over a period of time (Fig. I). Initial studies on hamsters have shown that intraperitoneal injection of glutathione diethyl ester at a dose of 5 mmol/kg body weight leads to a 3-fold increase of the glutathione level of the liver after 1 hr. Procedures Preparation o f Glutathione Diethyl Ester l
Procedures involving use of ether, formic acid, acetic acid, and acetic anhydride should be carried out in a chemical fume hood. Glutathione (GSH) (5 g; 16.25 mmol) is added to 100 ml of an ice-cold solution of anhydrous ethanol containing 5.6 ml of concentrated H 2 S O 4 . The resulting slurry is stirred briefly on ice to effect complete solution of GSH. The reaction is allowed to stand at 00-5 ° for 3-7 days. Formation of the diester will be about 50-70% complete; progress can be followed by HPLC. 1 Cold diethyl ether (10 volumes) is added to the reaction mixture. The white precipitate, which begins to form immediately, is allowed
(a) Extracellular: GSH-Diester
II
(b) 1l TransportRapid Intracellular: GSH-Diester
GSH-Monoester
(C) ]l TranspOrtslow (d) Rapid
• . GSH-Monoester (e) I Slow GSH
FIG. 1. Transport and cleavage of GSH diester in human cells. (a) GSH diethyl ester is rapidly split to GSH monoester in mouse and rat plasma, but not in plasma from humans, hamsters, rabbits, guinea pigs, and sheep. (b) GSH diethyl ester is rapidly transported into human cells and (d) is rapidly split within the cells to form GSH monoester, which is (more slowly) converted to GSH (e).
502
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[50]
to settle for several hours (on ice), after which time the product forms a white syrup at the bottom of the flask. The clear supernatant is decanted; residual ether is removed by drying the syrup under a stream of dry nitrogen. The syrup, which contains both monoester and diester, is dissolved in the minimal amount of cold water (about 25 ml), and the pH is carefully adjusted to pH 5 with 10 M NaOH. The solution is applied to a Chelex column (5 × 10 cm, Na + form, pH 5.5). [The Chelex column is prepared by washing Chelex resin (Na ÷ form) with 4-5 portions (total 5 volumes) of 0.5 M acetic acid in sodium acetate buffer, pH 5.5, and then removing the acetate by extensive washing with deionized distilled water.] The loaded column is first eluted with water; the eluent is monitored by reaction with 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB), 2 and the identity of the thiols is determined by HPLC. The monoester elutes first. After complete elution of the monoester, the solvent is changed to 0.2 M Na2SO4. This elutes the diester; the eluent is collected until it no longer gives a positive reaction with DTNB. The diester solution is lyophilized to dryness. The solid, which contains the diester and NazSO4, is extracted with cold absolute ethanol until the extracts are thiol-free as determined by lack of reaction with DTNB. The ethanolic extracts are combined and evaporated to very low volume (10-20 ml) under vacuum. The diester is precipitated by adding cold diethyl ether. The product, a white powder, is washed several times with diethyl ether, dried in air or briefly under dry N2, and then dried under vacuum over KOH. The yield is 20-25%. Analysis by HPLC shows that the product obtained is about 95% diester; impurities are monoester (-1%), cysteinylglycine ethyl ester (-2%), and an unknown by-product (-2%) eluting about 2-3 min after the diester on HPLC. An equivalent synthesis of the diester can be performed using 5 g ofGSH monoethyl ester. 0.5 H2SO4 as the starting material. 3 Comments. The glutamyl o~-carboxyl moiety of GSH is more slowly esterified than the glycine carboxyl, and thus GSH must be exposed for longer periods to relatively high concentrations of acid. The lability of the y-glutamyl moiety of GSH leads to formation of by-products that include cysteinylglycine ester and another product not yet identified (this may be a cyclized ester 4 or a diketopiperazine), which elutes on HPLC at 38 min; under these conditions, the retention times for cysteine, GSH, GSH monoethyl ester, cysteinylglycine ethyl ester, and GSH diethyl ester were, respectively, 7, 10, 22, 29, and 35 min. 1The present method consti2 G. L. Ellman, Arch. Biochem. Biophys. 82, 70 (1959). 3 M. E. Anderson, E. J. Levy, and A. Meister, this volume [49]. 4 p. K. Thornally, Biochem. J. 275, 535 (1991).
[50]
GLUTATHIONE DIETHYL ESTER AND DERIVATIVES
503
tutes a compromise in which formation of by-products is kept to a minimum at the expense of overall yield. The presence of metal ions as contaminants in some GSH ester preparations has been discussed.~,3 Copper is a frequent contamination found in both GSH monoester and GSH diester preparations. The presence of Cu in monoester preparations may lead to increased oxidation of cellular GSH (which is decreased by chelating agents), as found in studies on human erythrocyte suspensions; however, evidence was not found for such oxidation in vivo. Interestingly, GSH diethyl ester preparations that contained similar amounts of Cu did not promote oxidation of thiols in erythrocyte suspensions, nor were they toxic when tested in vivo in mice and hamsters.
Preparation of N-Acylglutathione and Glutathione Ester Derivatives The preparation of the N-acetyl derivatives of GSH, GSH monoethyl ester, and GSH diethyl ester, along with that of N-formyl-GSH and N-formyl-GSH mono(glycyl) ester, is given below. N-Acetylglutathione. GSH (5 g; 16.25 mmol) is dissolved in the minimal amount of water (about 30 ml)fl This solution is added over a period of 1 hr to a solution containing 50 ml of glacial acetic acid and 50 ml of acetic anhydride at 25°; cooling on ice is recommended. After the addition is complete the mixture is placed on ice, and an additional 25 ml of acetic anhydride is added. The mixture is allowed to warm gradually to room temperature and to stand overnight. Diethyl ether (at least 6-7 volumes) is added to the mixture. The white turbid solution, on standing overnight at 0°, gradually deposits the crystalline product as white needles. The yield is 60%; GSH is the only impurity [<10% by nuclear magnetic resonance (NMR) spectroscopy]. N-Acetylglutathione Monoethyl(glycyl) Ester. The acetylated monoester is prepared from the free base form of GSH monoethyl ester) After dissolving 5 g of the monoester in the minimal amount of water (about 30 ml), the synthesis proceeds identically to that of N-acetylglutathione. The yield is 30% and purity over 90% (by NMR). N-Acetylglutathione Diethyl Ester. To 31 ml of an ice-cold solution of absolute ethanol and 0.6 ml of concentrated H 2 S O 4 is added 0.7 g of N-acetylglutathione (see above). The solution is kept on ice and agitated occasionally until the starting material is dissolved. The reaction is complete after standing at 0°-4 ° overnight. To the mixture is added 10 g of Dowex-1 (HCO 3- form) in four portions; this is stirred under a stream of s E. J. Levy, M. E. Anderson, and A. Meister, Anal. Biochem. 214, 135 (1993).
504
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[50]
dry nitrogen for several minutes after each addition. The progress of sulfate removal should be monitored both by testing a drop of the solution with a drop of aqueous BaC12 (a white precipitate indicates SO42-) and by testing the solution with pH paper (the solution should be about pH 6 when removal of sulfate is complete). If necessary, additional portions (2-3 g) of the Dowex-1 may be added, in the same manner, until sulfate removal is complete. The resin is removed by filtration and then washed with several portions of absolute ethanol; the combined ethanolic solutions are evaporated to low volume (about 10 ml) under vacuum. The product is precipitated by adding diethyl ether; the white powder obtained is washed several times with diethyl ether and air-dried. The yield is about 50%, and the purity (HPLC), 90-95%. N-Formyiglutathione. To a cold solution of GSH in 99% formic acid (10 ml/g GSH) is added acetic anhydride (5 ml/g GSH). 5 The solution is agitated at room temperature for 30-60 min; formylation is complete after this time, as judged by absence of a color reaction with ninhydrin. The product is precipitated by addition of 5 volumes of diethyl ether. After standing overnight at 0°, the ether is decanted and the residual ether removed under a stream of dry N2. The hygroscopic solid is dried in a desiccator under reduced pressure over P205. The yield is 50-70%; the only impurity is GSH [<10% (NMR)]. The dry product can be crystallized by dissolving it in the minimal volume of water; after addition of 2-3 volumes of ethanol, ethyl acetate is added to the point of turbidity. Diethyl ether (4-5 volumes) is then added; on standing overnight at 0°, the product is obtained as a crystalline solid (white needles). N-Formylglutathione Monoethyl(Glycyl) Ester. Glutathione monoester in the free base form3 is treated as described above for preparation of N-formylglutathione. Precipitation with diethyl ether leads to a clear gelatinous precipitate. This semisolid is suction-filtered and then triturated with ether and filtered. A white powder is obtained, which is dried in air and then under vacuum over P205. The yield is 50-70% and purity (HPLC) over 95%. Comments. Previous reports 6-a erroneously identified the product obtained by treating GSH with formic acid and acetic anhydride as N-acetylGSH. However, this procedure formylates, rather than acetylates, GSH. 5 The product obtained by treating the monoethyl ester of GSH with formic acid and acetic anhydride is, similarly, N-formyl-GSH ethyl ester. We are 6 G. E. Utzinger, L. A. Strait, and L. D. Tuck, Experientia 19, 324 (1963). 7 D. Vander Jagt and L.-P. B. Han, Biochemistry 12, 5161 (1973). 8 M. E. Anderson, F. Powrie, R. N. Purl, and A. Meister, Arch. Biochem. Biophys. 239, 538 (1985).
[51]
ANALYSES OF PROBUCOL ACTIVITY IN SERUM
505
aware of only one other genuine acetylation of GSH9; in this preparation, glutathione disulfide was the starting material. Biological Applications of N-Acylglutathione Derivatives Only limited information is currently available on the biological applications of N-acylglutathione derivatives. Studies on suspensions of human erythrocytes have shown evidence for uptake of N-acetyl-GSH diethyl ester and for partial deesterification of this compound intracellularly; no cleavage of the N-acetyl moiety was found. Initial studies on hamsters indicate that administration of N-acetyl-GSH diethyl ester leads to some increase in the level of G S H in the liver and kidney of buthionine sulfoximine-treated animals. Administration of N-acetyl-GSH did not significantly increase liver G S H levels, whereas kidney G S H levels were greatly increased. In these studies no evidence was obtained for uptake of the compounds into brain. Acknowledgment This work was supported in part by DK 12034and AI 31804. 9 W.-J. Chen, G. F. Graminski, and R. N. Armstrong, Biochemistry 27, 647 (1988).
[5 1] A n t i o x i d a n t A c t i v i t y a n d S e r u m L e v e l s o f P r o b u c o l and Probucal Metabolites
By
S I M O N J. T . M A O , M A R K T . Y A T E S , a n d R I C H A R D L . J A C K S O N
Introduction Probucol [bis(3,5-di-tert-butyl-4-hydroxyphenylthio)propane] is a marketed cholesterol-lowering agent ~'2 that has been shown to be a potent antioxidant. 3,4 The antioxidant activity of probucol inhibits the formation of oxidatively modified low density lipoproteins (LDL) and reduces atherosclerosis in animals. 5-8 The uniqueness of probucol is its associa1R. E. Tedeschi, B. L. Martz, H. A.Taylor, and B. J. Cerimele, Artery 10, 22 (1982). 2 j. W. Barnhart, J. A. Sefranka, and D. D. McIntosh, Am. J. Clin. Nutr. 23, 1229(1970). 3 L. R. McLean and K. A. Hagaman, Biochemistry 28, 321 (1989). 4 R. L. Barnhart, S. J. Busch, and R. L. Jackson, J. Lipid Res. 30, 1703 (1989). 5T. E. Carew, D. C. Schwenke, and D. Steinberg, Proc. Natl. Acad. Sci. U.S.A. 84, 7725 (1987). 6 T. Kita, Y. Nagano, H. Yokode, K. Ishi, N. Kume, A. Ooshima, H. Yoshida, and C. Kawai, Proc. Natl. Acad. Sci. U.S.A. 84, 5928 (1987). METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by AcademicPress, Inc. All rights of reproduction in any form reserved.
[51]
ANALYSES OF PROBUCOL ACTIVITY IN SERUM
505
aware of only one other genuine acetylation of GSH9; in this preparation, glutathione disulfide was the starting material. Biological Applications of N-Acylglutathione Derivatives Only limited information is currently available on the biological applications of N-acylglutathione derivatives. Studies on suspensions of human erythrocytes have shown evidence for uptake of N-acetyl-GSH diethyl ester and for partial deesterification of this compound intracellularly; no cleavage of the N-acetyl moiety was found. Initial studies on hamsters indicate that administration of N-acetyl-GSH diethyl ester leads to some increase in the level of G S H in the liver and kidney of buthionine sulfoximine-treated animals. Administration of N-acetyl-GSH did not significantly increase liver G S H levels, whereas kidney G S H levels were greatly increased. In these studies no evidence was obtained for uptake of the compounds into brain. Acknowledgment This work was supported in part by DK 12034and AI 31804. 9 W.-J. Chen, G. F. Graminski, and R. N. Armstrong, Biochemistry 27, 647 (1988).
[5 1] A n t i o x i d a n t A c t i v i t y a n d S e r u m L e v e l s o f P r o b u c o l and Probucal Metabolites
By
S I M O N J. T . M A O , M A R K T . Y A T E S , a n d R I C H A R D L . J A C K S O N
Introduction Probucol [bis(3,5-di-tert-butyl-4-hydroxyphenylthio)propane] is a marketed cholesterol-lowering agent ~'2 that has been shown to be a potent antioxidant. 3,4 The antioxidant activity of probucol inhibits the formation of oxidatively modified low density lipoproteins (LDL) and reduces atherosclerosis in animals. 5-8 The uniqueness of probucol is its associa1R. E. Tedeschi, B. L. Martz, H. A.Taylor, and B. J. Cerimele, Artery 10, 22 (1982). 2 j. W. Barnhart, J. A. Sefranka, and D. D. McIntosh, Am. J. Clin. Nutr. 23, 1229(1970). 3 L. R. McLean and K. A. Hagaman, Biochemistry 28, 321 (1989). 4 R. L. Barnhart, S. J. Busch, and R. L. Jackson, J. Lipid Res. 30, 1703 (1989). 5T. E. Carew, D. C. Schwenke, and D. Steinberg, Proc. Natl. Acad. Sci. U.S.A. 84, 7725 (1987). 6 T. Kita, Y. Nagano, H. Yokode, K. Ishi, N. Kume, A. Ooshima, H. Yoshida, and C. Kawai, Proc. Natl. Acad. Sci. U.S.A. 84, 5928 (1987). METHODS IN ENZYMOLOGY,VOL. 234
Copyright © 1994by AcademicPress, Inc. All rights of reproduction in any form reserved.
506
ANTIOXlDANT CHARACTERIZATION AND ASSAY
[51]
tion with LDL; about 70% of the plasma probucol resides in LDL particles. 8 It has been hypothesized that such association directly protects LDL from oxidation inside the arterial wall. 8 For example, the watersoluble antioxidant vitamin C, which lacks an association with LDL, has not proved to be antiatherogenic. Similar to all chain-breaking antioxidants, probucol is oxidized, consumed, and depleted during oxidation. 4,8 Figure 1 shows the in vivo metabolic pathway of probucol as it undergoes oxidation. Three major metabolites, spiroquinone, diphenoquinone, and bisphenol, have been identified. 8 Interestingly, bisphenol is also an antioxidant, being oxidized to diphenoquinone when acting as such. Therefore, the antioxidant activity is continuously regenerated in vivo.
Here we describe the assay for the antioxidant activity of probucol in LDL and whole serum and the analytical procedures for the determination of probucol and probucol metabolites in serum. Determination of these metabolites in serum may provide insight as to the mode of action of probucol in inhibition of atherosclerosis.
Determination of Antioxidant Activity of Probucol Several factors must be considered for accurately measuring the antioxidant potency of probucol. Among these factors are the source of polyunsaturated lipids,9 the system used for free radical generation, 9-~2 solubility of the tested antioxidant, and the assay conditions.~3,14 Soybean phosphatidylcholine liposomes have been used for many antioxidant assays.~5 The protocol described here employs LDL as the lipid substrate, since suppression of oxidative modification of LDL is directly associated with the inhibition of atherosclerosis in vivo. 6,7 7 S. J. T. Mao, M. T. Yates, A. E. Rechtin, R. L. Jackson, and W. A. Van Sickle, J. Med. Chem. 34, 298 (1991). 8 S. J. T. Mao, M. T. Yates, R. A. Parker, E. M. Chi, and R. L. Jackson, Arterioscler. Thromb. U , 1266 (1991). 9 B. Halliwell and M. C. Gutteridge, Biochem. J. 219, 1 (1984). l0 y . Yamamoto, E. Niki, Y. Kamiya, and H. Shimasaki, Biochim. Biophys. Acta 795, 332 (1984). H L. R. C. Barclay, S. J. Locke, J. M. MacNeil, J. Van Kessel, G. W. Burton, and K. U. Ingold, J. Am. Chem. Soc. 106, 2479 (1984). 12j. W. Heinecke, H. Rosen, and A. Chait, J. Clin. Invest. 77, 757 (1984). 13 C. E. Thomas, Biochim. Biophys. Acta 1128, 50 (1992). t4 W. A. Pryor, T. Strickland, and D. F. Church, J. Am. Chem. Soc. 110, 2224 (1988). 15 E. Niki, E. Komuro, M. Takahashi, S. Urano, E. Ito, and K. Terao, J. Biol. Chem. 263, 19809 (1988).
[51]
ANALYSES OF PROBUCOL ACTIVITY IN SERUM
507
-
Oxidation
.o
2"-
CH3 Probucol
-'X
/
~-C H s ~ StobTlized Radical
iradical Coupling
HsC CHs Spiroquinone
Diphenoquinone
i ReducUon
H O ~ O H Bisphenol FI~. 1. Biotransformation pathway of probucol proposed in humans and rabbits. Detection of the metabolites in mouse serum suggests that a similar pathway exists in all animal species.
For oxygen radical generation CuSO4 (Cu2+)-induced free radicals, formed through the Fenton reaction, are used in the assay. The resulting LDL oxidation mimics that oxidation mediated by many cell types such as macrophages, fibroblasts, endothelial, and smooth muscle cells. 16 In addition, the lipid peroxides of LDL generated by Cu 2+ are almost identical to those produced by 2,2'-azobis(2-amidinopropane) dihydrochloride (AAPH), a hydrophilic, thermally decomposed radical initiator. 17 With 16 G. Ku, C. E. Thomas, A. L. Akeson, and R. L. Jackson, J. Biol. Chem. 2,67, 14183 (1992). 17 C. E. Thomas and R. L. Jackson, J. Pharmacol. Exp. Ther. 256, 1182 (1991).
508
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[51]
respect to solubility, probucol and most lipophilic antioxidants are not readily soluble in aqueous solution. Therefore, it is critical to dissolve probucol in an organic solvent that will not precipitate LDL. Dimethyl sulfoxide (DMSO), a polar aprotic solvent, is appropriate for this purpose. The pH of the assay mixture is also important as Cu 2÷ at high concentrations drastically lowers pH and thus precipitates LDL. This condition, however, can be avoided by adjusting the pH of the CuSO4 solution with Tris buffer (do not use sodium hydroxide or sodium phosphate) prior to use in the assay.
Determination of Antioxidant Activity of Probucol in Low Density Lipoproteins The following procedure has been used for assessing the antioxidant activity of probucol in LDL. 1. Dissolve probucol or the compound to be tested in DMSO to a final concentration of 1 mM (or 0.516 mg/ml of probucol). 2. Add 10 ~1 of the freshly prepared antioxidant described above to 50/~1 of LDL containing 250/zg protein (about 10/xM of apolipoprotein B) while gently shaking. 3. Incubate the mixture at 42 ° for 30 min. 4. Add 450/.d of 10 mM phosphate-buffered saline (PBS), gently mix (do not vortex), and thereafter add 10 t~l of 0.5 mM CuSO4 ; mix and incubate at 37°. The final Cu 2÷ concentration is 10/~M. 5. Take 250 ~1 from the reaction mixture at desired time points and stop the reaction by adding 1 ml of 20% trichloroacetic acid (TCA) and vortex. 6. Add 1 ml of 0.67% thiobarbituric acid (TBA) in 0.05 N NaOH; vortex and incubate at 80o-90 ° for 30 min to develop the color. 7. Centrifuge at 3000 rpm for 5 min at 25° and transfer 250/A of the supernatant into a 96-well microtiter plate. 8. Read the optical density at 540 nm using an enzyme-linked immunosorbent assay (ELISA) plate reader. 9. Standard solutions of malondialdehyde (MDA) containing 0-5 nmol MDA in 250/xl are prepared from a stock solution of malonaldehyde bis(dimethylacetal) (Aldrich, Milwaukee, WI) according to the method of Yagi. 18 Repeat steps 5-9. A typical assay using the procedure described above is shown in Fig. 2. It is possible to reduce the assay and reagent volumes in steps 5 and 18 K. Yagi, in "Lipid Peroxides in Biology and Medicine" (K. Yagi, ed.), p. 223. Academic Press, New York, 1982.
[51]
509
ANALYSES OF PROBUCOL ACTIVITY IN SERUM At -
~) 0
60
~
45
Control
L,
"
Bisphenol
.J
E
0
30
E
~5.
a ~
o o
5
10
!
!
15
20
T I M E (hours) FIG. 2. Inhibition of Cu2+-induced formation of thiobarbituric acid-reactive substances (TBARS) in LDL samples treated with probucol or bisphenoi. Suppression of lipid peroxidation from prolonged incubation with Cu 2+ (10/~M) was used to evaluate the antioxidant potency of tested compounds. Malondialdehyde (MDA) is used as a reference for TBARS.
6 for the TBA reaction, so that time course studies can be carried out. The procedure can also be adapted to determine the concentration of probucol that inhibits 50% of maximal LDL lipid peroxidation (IC50) at a certain time point.
Determination of Antioxidant Activity of Probucol in Whole Serum It has been reported that serum probucol levels are positively correlated with antiatherogenic activity in Watanabe rabbits. 8 To determine the bioavailability of probucol or to compare antioxidant efficacy between antioxidants using whole serum from antioxidant-treated animals, the following procedure is used. 1. Adjust the pH of a CuSO4 solution to pH 7.0 by adding 1 M Trisbase to a final Cu ~÷ concentration of 100 mM. 2. Add 5/xl of the CuSO4 solution to 100/zl of the serum to be tested, either from freshly prepared or frozen samples (-80°). The final concentration of Cu 2÷ is 4.76 mM. 3. Mix and incubate at 37°; stop the reaction at the desired time point by adding I ml of 20% TCA. 4. Follow steps 6-9 as described above.
510
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[51]
Typical antioxidant activities of serum samples (n = 6) obtained from 1% (w/w) cholesterol-fed rabbits given 0.5% (w/w) probucol or 0.5% (w/w) vitamin E are shown in Fig. 3. In this assay, probucol appears to be more effective in protecting whole serum from Cu2+-induced lipid peroxidation. Average ( - S . D . ) serum concentrations of probucol and vitamin E were 66.2 -+ 8 and 100.1 - 8/xg/ml, respectively.
A.
4hr
217
E 2
o m 16.8
E
4.7
#) o O
E ¢
B. 24 hr 281
273
:E
Control
Vitamin E
Probucol
FIG. 3. Inhibition of Cu2+-induced TBARS formation in whole serum isolated from 1% cholesterol-fed rabbits treated with 1% probucol or 1% vitamin E in the diet (w/w). Oxidation was allowed to proceed for 4 hr (A) and 24 hr (B). Control rabbits received feed containing 1% cholesterol only. Serum samples were obtained from a pool of six rabbits in each group.
[51]
511
ANALYSES OF PROBUCOL ACTIVITY IN SERUM
0.25
0.20
I I
0.15
I
E
E t-
tO
O 0.I0-
c'~
4
< I
<
?,
ai! 0.00
......
0
10
20
Time (min)
-
30 0
j.~o%__..~
--
10
20
30
T i m e (rain)
FIG. 4. Typical high-performance liquid chromatography profiles of serum extracts from control (left) and probucol-treated (right) rabbits. Samples were eluted from a Ct8 reversedphase column in acetonitrile/water (85 : 15, v/v) at a flow rate of 1.5 ml/min. Peaks 1, 2, 3, and 4 represent bisphenol, probucol, spiroquinone, and diphenoquinone, respectively.
Determination of Probucol and Metabolite Concentrations in Serum Animal studies have shown that 2 to 8% of probucol is absorbed after a single oral d o s e . 19'2° Absorption is enhanced when the drug is administered immediately after a meal. In healthy volunteers who received a single oral 3 g dose of probucol before a meal, after a meal, or in a fasting condition, peak plasma concentrations of 3.1 to 7.0, 19.2 to 23.7, and 2.0 to 3.9/~g/ml, respectively, were foundY Probucol is maintained in the circulation for a long period of time; an elimination half-life as long as 47 days has been reported. 22 Assays for probucol concentrations have t9 H. Choisy and H. Millart, Nouv. Presse Med. 9, 2981 (1980). 20 j. F. Heeg, M. F. Hiser, D. K. Satonin, and J. R. Rose, J. Pharm. Sci. 73, 1758 (1984). 21 j. F. Heeg and H. Tachizawa, Nouv. Press Med. 9, 2990 (1980). ~2 R. Fellin, A. Gasparotto, G. Valerio, M. R. Baiocchi, R. Padrini, S. Lamon, E. Vitale, G. Baggio, and G. Crepaldi, Atherosclerosis 59, 47 (1986).
512
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[51]
TABLE I PROBUCOL AND METABOLITE CONCENTRATIONS IN MOUSE SERUM Concentration (/~g/ml -+ S.D.) Day
N°
Probucol
Spiroquinone
Diphenoquinone
Bisphenol
7 14 21
4 4 3
5.6 -+ 1.5 4.9 + 2.4 4.7 + 1.0
0.55 -+ 0.13 0.53 + 0.29 0.57 + 0.21
0.12 -+ 0.03 0.11 - 0.05 0.18 - 0.02
0.65 -+ 0.13 0.53 -+ 0.17 0.50 -+ 0.17
Number of animals. Mice (CD-1) were fed with normal chow supplemented with 1% probucol.
been performed by gas chromatography-mass spectrometry 23 and highperformance liquid chromatography (HPLC) 7,8,24with good agreement between the methods. Because diphenoquinone, a metabolite of probucol, regenerates to another antioxidant, namely, bisphenol (Fig. 1), it is important to monitor these metabolites as well as probucol concentrations in serum. The following procedures have been developed in our laboratories. 1. Add 100/~1 of freshly prepared serum or plasma (or stored at - 80°) dropwise to 2 ml of an ether/ethanol mixture (3 : 1, v/v), while it is being vortexed in a 13 x 100 mm borosilicate glass tube. 2. Vortex for another 30 sec and then centrifuge at 4 ° for 10 min at 2000 rpm. Transfer the supernatant fraction to borosilicate glass tubes (13 x 100 mm) and dry under N2 gas. 3. Add 200 /~1 acetonitrile/hexane/0.1 M ammonium acetate (90 : 6.5 : 3.5, v/v/v), vortex, and inject 160/.d onto a HPLC system. 4. HPLC conditions are as follows: Deltapak C18 reversed-phase column (150 × 3.9 ram, 300 ,~; Waters, Milford, MA); mobile phase, acetonitrile/water (85 : 15, v/v); flow rate, 1.5 ml/min; detector, 990 photodiode array or monitor at 240 nm with single-channel, or at 240 nm and 420 nm (for diphenoquinone) with double-channel detector. 5. Prepare standard solutions by spiking probucol, spiroquinone, diphenoquinone, or bisphenol (Marion Merrell Dow, Cincinnati, OH) previously solubilized in DMSO into plasma (0-50 /~g/ml) or into acetonitrile/hexane/0.1 M ammonium acetate (90:6.5 : 3.5, v/v/v). 23 D. K. Satonin and J. E. Coutant, J. Chromatogr. 380, 401 (1986). 24 L. S. Elinder and G. Walldius, J. Lipid Res. 33, 131 (1992).
[52]
PBN AS ANTIOXIDANTFOR LDL
513
A typical HPLC analysis of probucol and its metabolites in probucoltreated rabbits is shown in Fig. 4. An example of the pharmacokinetic properties of probucol in mice is given in Table I. Most of the serum diphenoquinone is converted to bisphenol and seems to reach a steady state within 7 days when probucol was given in the diet (1%, w/w).
[52] a - P h e n y l N-tert-Butylnitrone as A n t i o x i d a n t for L o w Density Lipoproteins By DIANA M. LEE
Introduction There is increasing evidence that the initiation of atherosclerosis is related to free radical reactions, lipid peroxidation, and oxidative modification of low density lipoproteins (LDL). 1,2 Oxidized LDL (ox-LDL) is chemotactic for monocytes; 3 it may recruit monocytes into the subendothelial space. Ox-LDL is cytotoxic, 4 which may explain the endothelial damage occurring during atherogenesis. A prominent feature of atherosclerotic lesions is the cholesterol-loaded macrophage foam cell. 5 During the progression of atherosclerosis, circulating monocytes adhere to the endothelium, penetrate the vessel wall, differentiate into macrophages and become cholesterol-loaded foam cells. 6 LDL is believed to be the source of the cholesterol in foam cells. Macrophages ingest only small amounts of native LDL but may take up large amounts of ox-LDL via the scavenger receptor. This uptake is not down-regulated by the internalized cholesterol and thus leads to the loading of these cells with cholesterol and cholesteryl esters (CE) and to their transformation into cells with the characteristic properties of lipid-laden foam cells. In vitro oxidation of LDL can be mediated by all cells of the vascular system, namely, endothelial cells,
i D. Steinberg, S. Parthasarathy, T. E. Carew, J. C. Khoo, and J. L. Witztum, N. Engl. J. Med. 320, 915 (1989). 2 H. Esterbauer, G. Striegl, H. Puhl, S. Oberreither, M. Rotheneder, M. EI-Saadani, and G. Jiirgens, Ann. N.Y. Acad. Sci. 570, 254 (1989). 3 M. T. Quinn, S. Parathasarathy, L. G. Fong, and D. Steinberg, Proc. Natl. Acad. Sci. U.S.A. 84, 2995 (1987). 4 M. K. Cathcart, A. K. McNally, and G. M. Chisolm, J. Lipid Res. 32, 63 (1991). 5 A. M. Gown, T. Tsukada, and R. Ross, Am. J. Pathol. 125, 191 (1986). 6 R. G. Gerrity, Am. J. Pathol. 103, 181 (1981).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[52]
PBN AS ANTIOXIDANTFOR LDL
513
A typical HPLC analysis of probucol and its metabolites in probucoltreated rabbits is shown in Fig. 4. An example of the pharmacokinetic properties of probucol in mice is given in Table I. Most of the serum diphenoquinone is converted to bisphenol and seems to reach a steady state within 7 days when probucol was given in the diet (1%, w/w).
[52] a - P h e n y l N-tert-Butylnitrone as A n t i o x i d a n t for L o w Density Lipoproteins By DIANA M. LEE
Introduction There is increasing evidence that the initiation of atherosclerosis is related to free radical reactions, lipid peroxidation, and oxidative modification of low density lipoproteins (LDL). 1,2 Oxidized LDL (ox-LDL) is chemotactic for monocytes; 3 it may recruit monocytes into the subendothelial space. Ox-LDL is cytotoxic, 4 which may explain the endothelial damage occurring during atherogenesis. A prominent feature of atherosclerotic lesions is the cholesterol-loaded macrophage foam cell. 5 During the progression of atherosclerosis, circulating monocytes adhere to the endothelium, penetrate the vessel wall, differentiate into macrophages and become cholesterol-loaded foam cells. 6 LDL is believed to be the source of the cholesterol in foam cells. Macrophages ingest only small amounts of native LDL but may take up large amounts of ox-LDL via the scavenger receptor. This uptake is not down-regulated by the internalized cholesterol and thus leads to the loading of these cells with cholesterol and cholesteryl esters (CE) and to their transformation into cells with the characteristic properties of lipid-laden foam cells. In vitro oxidation of LDL can be mediated by all cells of the vascular system, namely, endothelial cells,
i D. Steinberg, S. Parthasarathy, T. E. Carew, J. C. Khoo, and J. L. Witztum, N. Engl. J. Med. 320, 915 (1989). 2 H. Esterbauer, G. Striegl, H. Puhl, S. Oberreither, M. Rotheneder, M. EI-Saadani, and G. Jiirgens, Ann. N.Y. Acad. Sci. 570, 254 (1989). 3 M. T. Quinn, S. Parathasarathy, L. G. Fong, and D. Steinberg, Proc. Natl. Acad. Sci. U.S.A. 84, 2995 (1987). 4 M. K. Cathcart, A. K. McNally, and G. M. Chisolm, J. Lipid Res. 32, 63 (1991). 5 A. M. Gown, T. Tsukada, and R. Ross, Am. J. Pathol. 125, 191 (1986). 6 R. G. Gerrity, Am. J. Pathol. 103, 181 (1981).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
514
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[52]
smooth muscle cells, and monocyte macrophages. 7-9 All three types of cells are present in atherosclerotic lesions. The oxidation of LDL in cells requires copper or iron ions as catalyst. Copper or iron in the absence of cells can also oxidize LDL. Based on studies using specific inhibitors, it was suggested that the enzyme responsible for oxidation of LDL is a 15-1ipoxygenase (LO) in activated monocytes, ~° endothelial cells, 1~ smooth muscle cells, H and mouse peritoneal macrophages, t2 Although Sparrow and Olszewski 13 suggested that 15-LO is not required for modification of LDL by cultured macrophages, others provided overwhelming evidence in support of 15LO being involved in atherogenesis) 4-18 Theoretically, antioxidants should be able to control lipid peroxidation and thus prevent atherosclerosis. However, in the past, use of dietary vitamin E as a therapeutic agent in most of the pathological conditions, including atherosclerosis, has been disappointing.19 Studies using probucol as an antioxidant demonstrated that the progression of the fatty streak or xanthoma was slowed down under conditions that had not lowered plasma cholesterol levels appreciably. 2°,2t These successful antioxidant experiments have revived new hope and interest in refocusing research on antioxidants and atherosclerosis. Now, since other antiatherogenic properties 7 T. Henriksen, E. M. Mahoney, and D. Steinberg, Proc. Natl. Acad. Sci. U.S.A. 78, 6499 (1981). a D. W. Morel, P. E. DiCorleto, and G. M. Chisolm, Arteriosclerosis (Dallas) 4, 357 (1984). 9 j. W. Heinecke, L. Balser, H. Rosen, and A. Chait, J. Clin. Invest. 77, 757 (1986). l0 A. K. McNally, G. M. Chisolm III, D. W. Morel, and M. K. Cathcart, J. Immunol. 145, 254 (1990). tt S. Parthasarathy, S. Wieland, and D. Steinberg, Proc. Natl. Acad. Sci. U.S.A. 88, 1046 (1989). 12 S. M. Rankin, S. Parthasarathy, and D. Steinberg, J. Lipid Res. 32, 449 (1991). 13 C. P. Sparrow and J. Olszewski, Proc. Natl. Acad. Sci. U.S.A. 89, 128 (1992). 14 S. Yl~-Herttuala, M. E. Rosenfeld, S. Parthasarathy, C. K. Glass, E. Sigal, J. L. Witztum, and P. Steinberg, Proc. Natl. Acad. Sci. U.S.A. 87, 6959 (1990). 15 S. YRi-Herttuala, M. E. Rosenfeld, S. Parthasarathy, E. Sigal, T. S~kioja, J. L. Witztum, and D. Steinberg, J. Clin. Invest. 87, 1146 (1991). 16 C. K. Derian and D. F. Lewis, Prostaglandins, Leukotrienes Essent. Fatty Acids 45, 49 (1992). 17 T. C. Simon, A. N. Makheja, and J. M. Bailey, Atherosclerosis (Shannon, Irel.) 75, 31 (1989). 18 M. Rosolowsky, J. R. Falck, J. T. Willerson, and W. B. Campbell, Circ. Res. 66, 608 0990). 19 p. M. Farrell, in "Vitamin E: A Comprehensive Treatise" (L. J. Machlin, ed.), p. 520. Dekker, New York, 1980. 20 T. E. Carew, D. C. Schwenke, and D. Steinberg, Proc. Natl. Acad. Sci. U.S.A. 84, 7725 (1987). 21 A° Yamamoto, Y. Matsuqawa, S. Yokoyama, T. Funahashi, T. Yamamura, and B. Kishino, Am. J. Cardiol. 57, 29H (1986).
[52]
PBN AS ANTIOXIDANTFOR LDL
515
have been found in association with p r o b u c o l , 22 it has become difficult to conclude that the effectiveness of probucol in the animal and human studies is due entirely to its antioxidant ability. The spin-trapping agent a-phenyl N-tert-butylnitrone (PBN) has been shown to offer protection by trapping free radicals during ischemia-reperfusion-mediated injury to the heart 23 and b r a i n 24 of experimental animals. Thus, it has been of considerable interest to establish whether it also exhibits antioxidant activity toward LDL by scavenging free radicals. Using an enzymatic system for LDL oxidation, we have undertaken an investigation on the possible capacity of PBN to inhibit the oxidation of LDL. This report describes the analyses of molecular species of neutral lipids in such a study. Procedure Isolation of Low Density Lipoproteins. A narrow density range of LDL with d values of 1.032-1.043 g/ml, representing the major peak of LDL, is isolated under N2 from fresh plasma using single-spin density gradient ultracentrifugation25 in the presence of 0.3 mM EDTA, 10 mM e-aminocaproic acid, penicillin G (500 units/ml), and streptomycin sulfate (50/xg/ ml). 26 The LDL is washed once to remove albumin35 Apolipoprotein B (apoB) accounts for 98% of the protein moiety. 27 All salt solutions and preservatives are saturated with N 2 prior to use. Enzymatic System. Soybean LO (SLO, type V, Sigma, St. Louis, MO) and bee venom phospholipase A 2 (PLA2) are used as oxidizing agents according to Sparrow et al., 28 with several modifications, in an attempt to mimic the physiological condition. The buffer system recommended by Sigma and Sparrow et al. 28 contains 50 mM borate, pH 9.0. However, we selected 0.15 M NaC1/5 mM Tris/1 mM CaC12, with or without preservatives containing 0.1 mM EDTA, pH 7.4, as a buffer system because in our experience borate at pH 9 acts as an oxidant and is capable of fragmenting apoB. In addition, high pH causes conformational change of L D L . 29 22 G. Franceschini, G. Chiesa, and C. R. Sirtori, Eur. J. Clin. Invest. 21, 384 (1991). 23 R. Bolli, B. S. Patei, M. O. Jeroudi, E. K. Lai, and P. B. McCay, J. Clin. Invest. 82, 476 (1988). 24 C. N. Oliver, P. E. Starke-Reed, E. R. Stadtman, G. J. Liu, J. M. Carney, and R. A. Floyd, Proc. Natl. Acad. Sci. U.S.A. 87, 5144 (1990). 25 D. M. Lee and D. Downs, J. Lipid Res. 23, 14 (1982). 26 D. M. Lee, A. J. Valente, W. H. Kuo, and H. Maeda, Biochim. Biophys. Acta 666, 133 0981). 27 D. M. Lee and P. Alaupovic, Biochim. Biophys. Acta 879, 126 (1986). 28 C. P. Sparrow, S. Parthasarathy, and D. Steinberg, J. Lipid Res. 29, 745 (1988). 29 S. Singh and D. M. Lee, Biochim. Biophys. Acta 876, 460 0986).
516
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[52]
Preparation ofa-Phenyl N-tert-Butylnitrone. Stock PBN solution (100 mM) is freshly prepared daily in a dark flask containing the same buffer system used for the enzymes. The buffer is purged with N2 for 30 min before addition of PBN. The flask is covered in an air-tight manner, and the PBN is dissolved gradually with the aid of a magnetic stirrer. Incubation System. The albumin-free LDL is dialyzed against the buffer system under N2 at 4 ° with six changes of buffer for 48 hr. The dialyzed LDL (0.5 mg protein/ml) is incubated with PLA 2 (3.3 units/ml) at 37°. After 2 hr of incubation, SLO (26,000 units/ml) is added. The tubes are loosely covered with tissue paper, and incubation is continued with gentle shaking at 37°. Aliquots are taken at time intervals for determination of the neutral lipid profile. The reaction is stopped by immersing the tubes in ice water and adding 3 mM EDTA plus 0.05% (w/v) reduced glutathione while air is replaced with argon. This experiment provides the time course of degradation of the molecular species of neutral lipids. To study the inhibition of oxidation, PBN is added to the incubation mixture of LDL/PLA2 at 0 hr at final concentrations of 0, 0.5, 1.0, 2.5, 5.0, and I0.0 mM. Incubation is carried out for 2 hr at 37°, and SLO added thereafter. The reaction is stopped after 24 hr, as described for the time study. Chemical Analyses. Protein is determined by the method of Lowry et al) ° The neutral lipid profile including unesterified cholesterol (UC), cholesteryl esters (CE), and triglycerides (TG) and their molecular species are determined by gas-liquid chromatography (GLC) according to the method described by Kuksis et al. 31 Samples are extracted under argon with haptane/isopropanol (3 : 7, v/v), containing cholesteryl butyrate as internal standard. The concentration of each molecular species of the neutral lipids of ox-LDL is individually expressed as the percentage of those in controls (native LDL, in the absence of enzymes or PBN). For determination of fatty acids in CE and TG of LDL, the lipid components are extracted under argon and separated by thin-layer chromatography (TLC) using the following solvent system: petroleum ether/diethyl ether/acetic acid, 80:20:1 (v/v) in a tank saturated with argon. The CE and TG spots on the TLC plate are scraped from the silica gel, then extracted and hydrolyzed under argon for fatty acid analyses using GLC. 3z 3o O. H. Lowry, N. J. Rosebrough, A. L. Farr, and R. J. Randall, J. Biol. Chem. 193, 265 (1951). 31 A. Kuksis, J. J. Myher, L. Marai, and K. Geher, J. Chromatogr. Sci. 13, 423 (1975). 32 W. J. McConathy, P. R. Blackett, and O. R. Kling, Clin. Chim. Acta 111, 153 (1981).
[52]
PBN AS ANTIOXIDANTFOR LDL
517
Results
Oxidative Effect on Molecular Species of Low Density Lipoprotein Cholesterol. Figure 1 shows a typical time course of SLO/PLA 2 oxidative degradation of UC and the molecular species of CE expressed on the basis of the fatty acyl chain length. The susceptibility of the LDL CE components to the enzymatic oxidation is in the order of CE-C20 > CEC18 > CE-C16 > UC. This trend was consistent among all eight normolipidemic subjects (with ad libitum diet) studied. UC was practically unaffected by the oxidative enzyme system, in contrast to the effect of Cu 2÷ (2.5 v,M)-catalyzed oxidation where a substantial amount of UC was modified (not shown). Table I shows the fatty acid composition of CE of native LDL from normolipidemic subjects. It is apparent that Ct8.2 is the predominant fatty acid followed by C18-1 and C16-0. Cl6-0 is the major fatty acid among the 16-carbon chain length fatty acids, whereas C20.4is the predominant fatty acid in the 20-carbon chain length fatty acids. The differences in unsaturation among C20, C~8, and C16 explain why the degradation rates are in the El tO0~L-I
"° I-° o
.
-"
•
s°I 60
lt. 0
z
40
n
20 0 0
I
4
I
,
I
i
I
,
i
8 12 16 20 INCUBATION TIME (hr)
i
24
FIG. 1. Time study for enzymatic oxidation of cholesteryl esters and unesteritied cholesterol in LDL. Native LDL containing 0.5 mg/ml of protein was incubated at 37° in 0.15 M NaCI/5 mM Tris/l mM CaC12, pH 7.4, in the presence of (El) PLA 2 (3.3 units/ml) or (E 2) SLO (26,000 units/ml). Aliquots were taken at time intervals for determination of neutral lipid profiles. Results were expressed as the percentage of controls (native LDL with no enzymes) for each molecular species. &, UC; 0 , CE-CI6; O, CE-C18; m, CE-C2o.
518
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[52]
TABLE I FATTY ACID COMPOSITION OF CHOLESTERYL ESTERS OF L o w DENSITY LIPOPROTEINS BEFORE AND AFTER ENZYMATIC OXIDATION
Fatty acid Ci4-o Ct6-o C16-I C18.0 CIs. 1 C18_2
Ci8. 3 C~3 C2o4
Total
Native L D L (%) 0.2 10.7 2.8 1.4 19.3 57.8 0.7
+- 0.3 a -+ 0.8 -+ 0.7 -+ 0.6 --- 2.2 --- 3.5 + 0.6 1.1 -+ 0.4 6.0 -+ 1.3 100
% Loss b
O x - L D L (%)
0 0 -35 0 -41 - 58 -82 -71 -79 - 48
0.5 20.8 3.5 2.7 22.1 47.1 0.3 0.6 2.4 100
a Mean -+ SD (n = 6 subjects). b Loss due to 24 hr of enzymatic oxidation is expressed as the percentage of each original fatty acid in CE. Cholesteryl esters o f L D L (d = 1.032-1.043 g/ml) were separated by TLC, extracted, hydrolyzed, and analyzed
by GLC as described in the text.
order of CE-C20 > CE-Cl8 > CE-CI6. Table I also shows the loss of each fatty acid in CE of LDL after 24 hr of incubation with PLA2/SLO. As expected, the loss is higher for the esterified fatty acid with higher unsaturation. The total loss of fatty acid from all CE is 48%. When expressed as a percentage of the remaining fatty acids in CE the composition is shown in the last column of Table I. Compared to the native LDL, this fatty acid composition of LDL-CE shows lower contents of highly unsaturated fatty acids and higher contents of saturated and monounsaturated fatty acids. These data are consistent with the neutral lipid composition which shows that ox-LDL has lower CE-C20 and CE-C~8 and higher UC and CE-C16 content than native LDL (Table II). Oxidative Effect on Molecular Species of Low Density Lipoprotein Triglycerides. Figure 2 shows a typical pattern of degradation rates of the molecular species of LDL TG in terms of the total fatty acyl chain length in each TG molecule. Here TG50 represents TG with two C16 plus one C18 fatty acids, TG52 denotes TG with two C18 plus one CI6, and TG54 denotes mainly TG with three Cl8 fatty acids. Whereas TG48 represents TG with three CI6 and TG56 represents TG with two CIS plus one C20, both TG48 and TG56 are present in trace amounts or undetectable in this LDL fraction. Again, the TG with fatty acids of longer chain lengths
[52]
PBN AS ANTIOXIDANT FOR LDL
519
T A B L E II TYPICAL COMPOSITION OF NEUTRAL LIPID MOLECULAR SPECIES OF LOW DENSITY LIPOPROTEINS BEFORE AND AFTER ENZYMATIC OXIDATION Species
Native L D L (%)
UC CE-CI6 CE-CI8 CE-C20 TG50 TG52 TG54 Neutral lipids
19.73 11.09 60.31 4.13 0.68 2.61 1.4
-+ 0.11 a _-x-0.18 -+ 0.42 4- 0.07 - 0.13 - 0.09 -+ 0.10 100
% Loss b
O x - L D L (%)
0 -25 -65 -74 -26 -40 -49 - 47
37.21 15.69 39.82 2.02 0.94 2.94 1.36 100
M e a n -+ SE, same L D L analyzed three times as controls. b L o s s due to 24 hr of enzymatic oxidation is e x p r e s s e d as the percentage of each original molecular species. a
showed a faster degradation rate than TG with shorter ones, so that the sequence of oxidation destruction was TG54 > TG52 > TG50. The fatty acid composition of TG in LDL is shown in Table III. In contrast to the fatty acid composition of LDL CE, the predominant fatty acid in LDL TG is Cl8-1 among the 18-carbon chain length species as well E
E2
.J I00 I---Z O
,
,,
B0
zw
70
~-
60
~ 50f 0
4
8
12
16
20
24
INCUBATIONTIME(hr] FIG. 2. Time study for e n z y m a t i c o x i d a t i o n o f triglycerides in L D L . Experimental conditions w e r e the same as described in Fig. 1. O, TG50; II, TG52; A, TG54.
520
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[52]
T A B L E III FATTY ACID COMPOSITION OF TRIGLYCERIDES OF NATIVE LOW DENSITY L1POPROTEINS F a t t y acid Cz4.o CI6-o C16-I Ct8.o Cls. 1 C18.2 Cls. 3 C20.3 C2e.4
Composition (%) 0.21 24.76 2.82 4.64 44.75 19.60 0.98 0.46 2.01
+ -+ -+ -+ -+ + -+ -+
0.17 a 2.23 0.78 0.57 2.05 4.54 0.15 0.33 0.25
a M e a n -+ SD (n = 4 subjects). Triglycerides of L D L (d = 1.032-1.043 g/ml) were separated by T L C , extracted, hydrolyzed, and analyzed by G L C as described in the text.
as among all fatty acids in TG. Clr-0 is the second major fatty acid, followed by C18.2. Thus, the TG50 consists mainly of two C~6-0 plus one C18-1, TG52 of two C18-1 plus one C16-0, and TG54 of three C~8_~and three C18.2. Therefore, the unsaturation in TG50, TG52, and TG54 is approximately 1, 2, and over 3, respectively. This unsaturation explains why the oxidative degradation rate is in the order of TG54 > TG52 > TG50. These data also explain why in the neutral lipid composition of ox-LDL, TG54 is lower and TG50 is higher than that in native LDL (Table II). This difference is more pronounced when the TG composition (separated from CE) of oxLDL is compared to that of native LDL. Thus, analysis of molecular species of the neutral lipid profile has provided a simpler and more accurate way to characterize the ox-LDL, instead of comparing the absolute fatty acid contents or their composition in CE of LDL. The decreased absolute contents of CE-C20, CE-C18, and TG54 or the increased percent contents of UC, CE-CI6, and TG50 in the composition of CE and TG, compared to native LDL, are characteristic of ox-LDL.
Effect of ot-Phenyl N-tert-Butylnitrone on Oxidative Modification of Cholesterol and Triglycerides of Low Density Lipoproteins. Figure 3 shows that the recovery of each molecular species of CE following enzymatic oxidation increased in a concentration-dependent fashion in the presence of PBN. When the activities of PLA 2 and SLO were reduced to one-fifth and one-half, respectively, to slow down the oxidation (not shown), the PBN inhibition responses for CE-C20 and CE-C~8 became
[52]
521
P B N AS ANTIOXIDANT FOR L D L ~"k . . . .
_J ~ ¢.r i-z
BO
~
40
-it,"
~
"J
°_~-~
".1~'"I
/"'1t"'~"
.If"
a.
20 0
0.o
i
2.5
.
.
.
.
i
5.o
. . . .
L.
7.5
,
,
i
~.o
PBN {raM) FIo. 3. Inhibition of ~xidation of cholesteryl esters and unesterified cholesterol in LDL by PBN. Various conceatations ~f PBN were incubated with the LDL/PLA2/SLO system at 37° under the same conditions as described in Fig. 1. The reaction was stopped after 24 hr. Aliquots were extracted immediately for neutral lipid analyses. Results were expressed as the percentoge of control LDL (native LDL with no enzymes and no PBN) for each molecular species. Symbols are the same as in Fig. 1.
curvilinear rather than near/y linear with increasing inhibition efficiency at lower PBN concentrations. Figure 4 shows that PBN also protects the LDL TG from oxidation. Under the same experimental conditions described in Fig. 3, the inhibition response was curvilinear for each TG molecular species. When PLAJ SLO activities were reduced to one-fifth and one-half, respectively, near 100% inhibition was reached at 5 mM PBN (not shown). Oxidation of LDL catalyzed by 2.5 ~M Cue+ demonstrated a much greater extent of damage than that by SLO/PLA2, in both the lipid and protein moieties (not shown). The PBN protective response was curvilinear for TG, but linear for CE (not shown). The I(]5o value for TG51) was I00 I~M PBN and that for TG54 was 200 p.M. Higher concentrations of PBN were required for other lipid components. Discussion
This study serves a 2-fold purpose. One, we have shown that the fatty acvl chain lengths of the molecular species of CE and TG in LDL reflect the unsaturation of the fatty acids in CE and TG: the longer the chain
522
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[52]
I00I !
f___--,
....
S--~
L.7~r.
._I l-Z r_3
fiOI +.m.,6~..~..~-"" -
ED Z ILl r~ EE ILl EL
i
40' 20 .
0.0
_
•
2.5
.n
5.0 PBN Ira,l,4}
,I
7.5
-~
--~--I
to.0
FiG. 4. Inhibition of oxidation of triglycerides in LDL by PBN. Experimental conditions were the same as described in Fig. 3. &, TGS0; 0 , TG52; II, TG54.
length, the higher the unsaturation, and the higher the susceptibility toward oxidation. This established relationship allows us to use neutral lipid profiles to replace fatty acid compositions for characterization of ox-LDL. The former approach is far superior as a routine method. It is fast, simple, accurate, precise, with high recovery, and has little chance of exposing samples to oxidation. In contrast, fatty acid analysis for each lipid component of LDL requires TLC separation of the components followed by scraping the gel, extraction, hydrolysis, and GLC analysis, which is very tedious, time-consuming, and may easily lead to exposure of samples to oxidation. Two, this study provides the chemical evidence that spintrapping agent PBN inhibits LDL from oxidative degradation by scavenging free radicals. Whether this phenomenon is a general characteristic of all spin-trapping agents remains to be explored in further experiments. By a different approach, Kalyanaraman et al. ~3 have also shown that PBN inhibits the oxidative modification of LDL caused by endothelial cells and Cu 2+ as judged from agarose electrophoretic mobility, macrophage degradation, and formation of thiobarbituric acid-reactive substances (TBARS). We have also measured lipid peroxidation by the TBARS method, but we found that PBN interacts with thiobarbituric acid, making the results unreliable. PBN also interferes with the absorbance at 234 nm 3~ B. Kalyanararnan, J. Joseph, and S. Parthasarathy, FEBS Lett. 280, 17 (1991).
[53]
B R A I N A N T I O X I D A N T A C T I V I T Y O F SPIN T R A P S
523
used to measure conjugated dienes, thus rendering that method invalid. It is worth noting that PBN also interferes with GLC analyses of fatty acids. However, PBN does not interfere with the GLC method of Kuksis et al. 31 which offers detailed information regarding the molecular species of CE and TG. The evidence presented in this study indicates that PBN is an effective antioxidant for LDL. The cardioprotective effects of PBN have previously been demonstrated. 23'34,35 PBN has been shown to reverse partially the reperfusion-induced damage in intact myocardium of dogs 23'34and to protect against reperfusion-induced arrhythmias in isolated heart models. 35 Taken together, results of these studies suggest that PBN or other spintrapping agents with similar chemical characteristics may be developed into antiatherogenic agents for use by humans. Acknowledgments The author thanks M.-m. Huang and C. Burden for skillfultechnical assistance, R. Whitmer for performingthe GLC analyses,and J. Pilcherfor typingthe manuscript. This work was supportedin part by the resourcesof the OklahomaMedicalResearchFoundation. 34R. Boll, M. O. Jeroudi, B. S. Patel, C. M. Dubose, E. K. Lai, R. Roberts, and P. B. McCay, Proc. Natl. Acad. Sci. U.S.A. 86, 4695 (1989). 35D. J. Hearse and A. Tosaki, Circ. Res. 60, 375 (1987).
[53] B r a i n A n t i o x i d a n t A c t i v i t y o f S p i n T r a p s in M o n g o l i a n G e r b i l s B y JOHN M. CARNEY and ROBERT A. FLOYD
Introduction Previous studies have demonstrated that there is a relationship between brain protein oxidation, reduction in marker enzyme activities, and behavioral dysfunction. 1 In an initial series of studies, we demonstrated that brain ischemia-reperfusion injury (IRI) results in significant increases in both salicylate hydroxylation and soluble brain protein oxidation, decreases in brain glutamine synthetase (glutamate-ammonia ligase) activ-
1 j. M. Carney, P. E. Starke-Reed, C. N. Oliver, R. W. Landum, M. S. Cheng, J. F. W u , a n d R . A . F l o y d , Proc. Natl. Acad. Sci. U.S.A. 88, 3636 (1991).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All fights of reproduction in any form reserved.
[53]
B R A I N A N T I O X I D A N T A C T I V I T Y O F SPIN T R A P S
523
used to measure conjugated dienes, thus rendering that method invalid. It is worth noting that PBN also interferes with GLC analyses of fatty acids. However, PBN does not interfere with the GLC method of Kuksis et al. 31 which offers detailed information regarding the molecular species of CE and TG. The evidence presented in this study indicates that PBN is an effective antioxidant for LDL. The cardioprotective effects of PBN have previously been demonstrated. 23'34,35 PBN has been shown to reverse partially the reperfusion-induced damage in intact myocardium of dogs 23'34and to protect against reperfusion-induced arrhythmias in isolated heart models. 35 Taken together, results of these studies suggest that PBN or other spintrapping agents with similar chemical characteristics may be developed into antiatherogenic agents for use by humans. Acknowledgments The author thanks M.-m. Huang and C. Burden for skillfultechnical assistance, R. Whitmer for performingthe GLC analyses,and J. Pilcherfor typingthe manuscript. This work was supportedin part by the resourcesof the OklahomaMedicalResearchFoundation. 34R. Boll, M. O. Jeroudi, B. S. Patel, C. M. Dubose, E. K. Lai, R. Roberts, and P. B. McCay, Proc. Natl. Acad. Sci. U.S.A. 86, 4695 (1989). 35D. J. Hearse and A. Tosaki, Circ. Res. 60, 375 (1987).
[53] B r a i n A n t i o x i d a n t A c t i v i t y o f S p i n T r a p s in M o n g o l i a n G e r b i l s B y JOHN M. CARNEY and ROBERT A. FLOYD
Introduction Previous studies have demonstrated that there is a relationship between brain protein oxidation, reduction in marker enzyme activities, and behavioral dysfunction. 1 In an initial series of studies, we demonstrated that brain ischemia-reperfusion injury (IRI) results in significant increases in both salicylate hydroxylation and soluble brain protein oxidation, decreases in brain glutamine synthetase (glutamate-ammonia ligase) activ-
1 j. M. Carney, P. E. Starke-Reed, C. N. Oliver, R. W. Landum, M. S. Cheng, J. F. W u , a n d R . A . F l o y d , Proc. Natl. Acad. Sci. U.S.A. 88, 3636 (1991).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All fights of reproduction in any form reserved.
524
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[53]
ity, and alterations in spontaneous exploratory activity. 2,3 The IRI-mediated changes in behavior are accentuated in aged gerbils. We have discovered that the changes in IRI-mediated oxidation and behavioral dysfunction could be prevented by pretreatment with phenyl tert-butylnitrone (PBN) and related spin-trapping c o m p o u n d s ) : Based on the results obtained with PBN in IRI-mediated oxidation, we determined if age-related alterations in brain function could be the result of progressive oxidative events, as had been proposed in the p a s t ) In a separate set of studies, we demonstrated that aged gerbils have a higher level of protein oxidation, compared to young adult gerbils. 1 As a result of these initial observations, the effects of daily PBN injections were evaluated on brain protein oxidation, glutamine synthetase activity, and radial arm maze performance. Radial arm maze performance is often used as an indicator of the status of short-term temporal-spatial memory in animals. 6 We present here more details of the PBN administration schedule as well as how the animals were tested in the behavior study protocol. Animals The subjects are young adult male gerbils (3-4 months of age) and aged, retired, male breeder gerbils (18-20 months of age) obtained from Tumblebrook Farms (West Brookfield, MA). Gerbils are housed three to a cage in standard rodent cages. Animals are maintained in the University of Kentucky central animal facility under a 12-hr light-dark cycle. All experiments are conducted during the light phase of the cycle. Food and water are available ad libitum throughout the day. Methods
Young adult male (3-4 months of age) and retired male breeder gerbils (18-20 months of age) are assigned to separate groups of 18 gerbils each. One group of young adult and aged animals is assigned to the vehicle (saline) control group, and the other groups receive PBN. Animals are given intraperitoneal injections twice daily (8:00 am and 8:00 pm) for a 2 W. Cao, J. M. Carney, A. Duchon, R. A. Floyd, and M. Chevion, Neurosci. Lett. 88, 233 (1988). 3 C. N. Oliver, P. E. Starke-Reed, E. R. Stadtman, G. J. Liu, J. M. Carney, and R. A. Floyd, Proc. Natl. Acad. Sci. U.S.A. 87, 5144 (1990). 4 R. A. Floyd, Science 254, 1597 (1991). D. Harman, Proc. Natl. Acad. Sci. U.S.A. 78, 7124 (1981). 6 D. S. Olton, J. T. Becker, and G. E. Handleman, Behav. Brain Sci. 2, 313 (1979).
[53]
BRAIN ANTIOXIDANT ACTIVITY OF SPIN TRAPS
I
525
I
FIG. 1. Radial arm maze testing apparatus for gerbils. Subjects were placed in the central start chamber (A) and allowed to acclimate for 3 min. At the end of the acclimation period, the doors were raised and the gerbil was free to explore the arms of the maze (B).
period of 14 consecutive days. It should be noted that PBN is light sensitive and is relatively stable at neutral pH. PBN obtained from Aldrich Chemical Co. (Milwaukee, WI) is dissolved freshly in neutral saline and administered at a dose of 32 mg/kg/injection (3.2 mg/ml). In subsequent studies, substantially lower doses have been used with comparable results. At the end of the 14 days, the animals are given an additional day of no injections to allow for the elimination of any residual PBN prior to testing. Previous studies with rodents had indicated that the half-life of PBN is between 2 and 3 hr. 7 Thus the extra day allowed for essentially all of the PBN to be eliminated in the period equal to approximately 10 half-lives. After the 24hr washout period, the gerbils are tested for radial arm maze performance. Radial Arm Maze Test
An eight-arm radial maze is used for testing patrolling behavior performance (Fig. 1). The gerbils are placed one at a time in the central compartment of the sunburst maze. When the doors to the arms are raised, each gerbil is free to explore the maze. Reentry into an arm more than once before exploring all eight arms is considered an error. Arm entry is registered electronically. Animals have 15 min to explore the maze. Normal young adult gerbils made between 4 and 5 errors, while aged gerbils made 9-11 errors. After 14 days of PBN treatment, the young adult gerbils made the same number of errors as the control group.1 In contrast, when aged gerbils that had received chronic PBN were tested, 7 G. Chen, T. M. Bray, E. G. Janzen, and P. B. McCay, Free Radical Res. Commun. 9, 317 (1990).
526
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[54]
they made significantly fewer errors. 1 In fact, the aged gerbils made the same number of errors as did the young adult gerbils that were 14 months younger. A parallel group of gerbils are posed in an identical manner and used for biochemical determinations during the period of PBN treatment. Chronic treatment resulted in a progressive decrease in the level of oxidized protein in the soluble protein fraction of the homogenate. 1 This progressive decrease in oxidized protein was mirrored by the concomitant increase in both glutamine synthetase and neutral protease activities in the aged gerbils. There was no change in these biochemical end points in the young adult gerbils treated with PBN in the same manner as the aged gerbils. 1 At the end of the 14 days, the PBN injections are terminated. A progressive return to the original level of protein oxidation was seen in the aged control gerbils and a progressive decrease in enzyme activities. Saline injections had no effect on these measures. Therefore, it appears there is a higher level of oxidative stress on the aged brain that can be modified, but the chronic stress apparently continued even in the face of chronic PBN administration. The intracellular site of generation of the oxidizing species remains to be identified and the defect characterized. One intriguing possibility is that damaged mitochondria may give rise to these oxidizing species. Acknowledgments Research was supported in part by National Institutes of Health Grants NS23307 and AG09690 and monies from the Glenn Foundation for Medical Research.
[54] A n t i o x i d a n t A c t i v i t y o f N i t e c a p o n e a n d Its A n a l o g O R - 1 2 4 6 : E f f e c t o f S t r u c t u r a l Modification on Antioxidant Action
By LUCIA MARCOCCI, YUICHIRO J. SUZUKI, MASAHIKO TSUCHIYA, and LESTER PACKER Introduction Nitecapone [3-(3,4-dihydroxy-5-nitrobenzylidene)-2,4-pentanedione] (Fig. 1), an effective peripherally acting inhibitor of catechol O-methylMETHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
526
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[54]
they made significantly fewer errors. 1 In fact, the aged gerbils made the same number of errors as did the young adult gerbils that were 14 months younger. A parallel group of gerbils are posed in an identical manner and used for biochemical determinations during the period of PBN treatment. Chronic treatment resulted in a progressive decrease in the level of oxidized protein in the soluble protein fraction of the homogenate. 1 This progressive decrease in oxidized protein was mirrored by the concomitant increase in both glutamine synthetase and neutral protease activities in the aged gerbils. There was no change in these biochemical end points in the young adult gerbils treated with PBN in the same manner as the aged gerbils. 1 At the end of the 14 days, the PBN injections are terminated. A progressive return to the original level of protein oxidation was seen in the aged control gerbils and a progressive decrease in enzyme activities. Saline injections had no effect on these measures. Therefore, it appears there is a higher level of oxidative stress on the aged brain that can be modified, but the chronic stress apparently continued even in the face of chronic PBN administration. The intracellular site of generation of the oxidizing species remains to be identified and the defect characterized. One intriguing possibility is that damaged mitochondria may give rise to these oxidizing species. Acknowledgments Research was supported in part by National Institutes of Health Grants NS23307 and AG09690 and monies from the Glenn Foundation for Medical Research.
[54] A n t i o x i d a n t A c t i v i t y o f N i t e c a p o n e a n d Its A n a l o g O R - 1 2 4 6 : E f f e c t o f S t r u c t u r a l Modification on Antioxidant Action
By LUCIA MARCOCCI, YUICHIRO J. SUZUKI, MASAHIKO TSUCHIYA, and LESTER PACKER Introduction Nitecapone [3-(3,4-dihydroxy-5-nitrobenzylidene)-2,4-pentanedione] (Fig. 1), an effective peripherally acting inhibitor of catechol O-methylMETHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[54]
ANTIOXIDANT ACTIVITYOF NITECAPONE AND OR-1246
527
Nitecapone O2N HO
H3 ttO
H3C---~
O O OR-1246
O2N ~~--~ 112 1tO ~ - - ~ C ItO
.CH3
1t 3 C - ~ O O
FIG. 1. Molecular structures of nitecapone and OR-1246.
transferase, 1'2 has been shown to be a promising antioxidant. Nitecapone scavenges superoxide, hydroxyl, and peroxyl radicals generated either in solution or in membranes, protects membranes from lipid peroxidation, recycles vitamin E through a reaction with ascorbate, and inhibits xanthine oxidase activity. 3,4 The compound has also been reported to form a reversible adduct with glutathione, and to protect this thiol from oxidative depletion. 5 Nitecapone exerts gastroprotective effects in various models of gastric and duodenal ulcers, 6-8 and it is being evaluated in a clinical study as an antiulcer drug. Although the stimulation of duodenal bicarbonate secretion mediated by nitecapone through the inhibition of catechol O-methyltransferase activity could explain the gastroprotective effect, 9 the antioxidant properties of the molecule could also contribute to this protection. In 1 R. BackstrOm, E. Honkanen, A. Pippuri, J. Pystynen, K. Heinola, E. Nissinen, I.-B. Lind6n, P. Mannisto, S. Kaakkola, and P. Photo, J. Med. Chem. 32, 841 (1989). z E. Nissinen, I.-B. Lind6n, E. Schultz, S, Kaakkola, P. Mannisto, and P. Photo, Eur. J. Pharmacol. 153, 263 (1988). 3 T. Metsii-Ketelii, E. Nissinen, T. Korkolainen, and I.-B. Lind6n, Dig. Dis. Sci. 35, 1037 (1990). 4 y. j. Suzuki, M. Tsuchiya, A. Safadi, V. E. Kagan, and L. Packer, Free Radical Biol. Med. 13, 517 (1992). 5 E. Nissinen, I.-B. Lind6n, and P. Photo, Free Radical Biol. Med. 9(S1), 19 (1990). 6 p. Aho, I.-B. Lind6n, E. Nissinen, and P. Photo, Dig. Dis. Sci. 33, 897 (1988). 7 p. Photo, P. Aho, and I.-B Lind6n, Eur. J. Pharmacol. 183, 314 (1990). 8 p. Aho and I.-B. Lind6n. Scand. J. Gastroenterol. 27, 134 (1992). 9 G. Flemstrom, B. Safsten, and G. Jedstedt, Gastroenterology 104, 825 (1993).
528
ANT1OXIDANT CHARACTERIZATION AND ASSAY
[54]
fact, a possible role o f reactive oxygen species has been reported in gastrointestinal damage. 10:1 It is now widely accepted that reactive oxygen species may play a role in the pathogenesis of several diseases (cancer, myocardial infarction, cataract) and aging) 2 As a consequence, nitecapone, by acting as an antioxidant, could be a useful therapeutic agent in a wide range of pathological conditions. Understanding how the structure of nitecapone relates to its antioxidant activities may enable improvements in its biological efficacy. However, determination of the relationship between the structure and the activity of an antioxidant in biological systems is difficult, because a complex set of information is necessary to cope with (1) the multiplicity of ways in which an antioxidant can operate and (2) the various elements that can modulate its chemical reactivity. A compound can exert antioxidant activity in various ways: it can directly scavenge various reactive oxygen species (superoxide, hydrogen peroxide, hydroxyl radical, hypochlorous acid, heme-associated ferryl species, peroxyl radical); it can inhibit formation of such species by reacting with promotors (iron, copper) or with generating enzymes (xanthine oxidase, NADPH-oxidase system, myeloperoxidase); or it can enhance the activity of other antioxidants. Furthermore, the mutal accessibility between an antioxidant and its target molecules as well as the availability of the antioxidant in its reactive chemical form are important in defining the antioxidant efficiency of a molecule. The relationships between structure and activity of antioxidants have been examined using homologs of vitamin E, 13'14 ubiquinol, 15 and dihydrolipoic acid.16 However, the complexity of free radical biology and chemistry precludes generalizations regarding how antioxidant structures relate to their activities. The double bond at C-3 of the pentanedione side chain gives nitecapone specific physicochemical properties. As outlined in Table I, OR-1246, a related compound which lacks the methylene double bond in the molecular structure [3-(3,4-dihydroxy-5-nitrobenzyl)-2,4-pentanedione] (Fig. 1), has a different UV-visible spectrum, a different pK a value, and a different l0 K. Kusterer, G. Pihan, and S. Szabo, Am. J. Physiol. 252, G811 (1987). II S. Arvidsson, Circ. Shock 16, 383 (1990). 12 B. Halliwell and J. M. C. Gutteridge, eds., " F r e e Radicals in Biology and Medicine." Oxford Univ. Press (Clarendon), Oxford, 1988. 13 V. E. Kagan, E. A. Serbinova, and L. Packer, Arch. Biochem. Biophys. 282, 221 (1990). 14 y . j. Suzuki, M. Tsuchiya, S. R. Wassail, Y. M. Choo, G. Govil, V. E. Kagan, and L. Packer, Biochemistry 32, 10692 (1993). 15 V. E. Kagan, E. A. Serbinova, G. M. Koynova, S. A. Kitanova, V. A. Tyurin, T. S. Stoytchev, P. J. Quinn, and L. Packer, Free Radical Biol. Med. 9, 117 (1990). 16 y . j. Suzuki, M. Tsuchiya, and L. Packer, Free Radical Res. Commun. 18, 115 (1993).
[54]
ANTIOXIDANT ACTIVITY OF NITECAPONE AND OR-1246
529
TABLE I PHYSICOCHEM1CAL CHARACTERISTICSOF NITECAPONE AND OR-1246 Property
Nitecapone
OR- 1246
Molecular weight Wavelength of maximum absorbance a (nm) Extinction coefficienta (M-I cm-~) pK a Log pb Log D c
265.22 380
267.22 444
24,200
2900
4.7 1.32 - 1.13
6.3 1.16 0.68
Measured in 20 mM phosphate buffer, pH 7.4. b p is the partition coefficient in octanol/100 mM HC1. c D is the partition coeffÉcient in octanol/67 mM phosphate buffer, pH 7.4.
partition coefficient in octanol/pH 7.4 buffer systems. Different antioxidant properties could result from these different physicochemical characteristics. In particular, as a consequence of the different electronic resonance structure of the two compounds, groups important in the reaction with radical species, such as the hydroxyl group on the aromatic ring, could have different reactivities. Furthermore, the different steric and electrostatic properties of nitecapone and OR-1246 could affect their interaction with membranes and proteins, thus causing the two molecules to have different antioxidant activities at these target sites. In this chapter, the methods used to assess the effects of the distinctive structural and physicochemical properties of nitecapone and OR-1246 on their respective antioxidant activities are described. We compare the two compounds in terms of (1) their reactions with superoxide in solution; (2) their reactions with peroxyl radicals generated both in solution and in membranes; (3) their effects on lipid peroxidation induced by peroxyl radicals generated in membranes; (4) their partition between membranes and aqueous phases; and (5) their interaction with an enzymatic source of reactive oxygen species (e.g., xanthine oxidase). Materials 2,2'-Azobis(2,4-dimethylvaleronitrile) (AMVN) and 2,2'-azobis(2amidinopropane) dihydrochloride (AAPH) are purchased from Polysciences, Inc. (Warrington, PA). cis-Parinaric acid is purchased from Mo-
530
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[54]
lecular Probes (Junction City, OR). Dioleoyl-L-a-phosphatidylcholine (DOPC), soybean L-a-phosphatidylcholine (soybean PC), butylhydroxytoluene (BHT), 2-thiobarbituric acid (TBA), xanthine, xanthine oxidase, and cytochrome c are from Sigma (St. Louis, MO). B-Phycoerythrin is a gift from Prof. A. N. Glazer (Department of Microbiology, University of California, Berkeley). Nitecapone and OR-1246 are from Orion-Farmos Pharmaceuticals (Espoo, Finland). All other reagents are reagent grade. Statistical Analysis Means and standard deviations are reported. Significant differences between values are calculated by using Student's t-test at p < 0.05. Scavenging Activity against Superoxide Anion Radical Superoxide scavenging activity by nitecapone and OR-1246 is determined by monitoring their effects on reduction of cytochrome c mediated by superoxide. The reduction can be monitored spectrophometrically by measuring the increase in absorbance of reduced cytochrome c at 550 nm. To avoid interference of the antioxidants with xanthine oxidase, the enzyme generally used to generate superoxide, we use a stable solution of superoxide prepared from alkaline dimethyl sulfoxide (DMSO).
Preparation of Reagents Superoxide Anion Solution. A stable solution of superoxide anion radical, as described by Hyland et al., 17 is prepared from alkaline DMSO. Mix I00 ml of DMSO with 1 ml of aqueous 0.5 mM NaOH, then incubate at 25 ° for 30 min. Oxidized Cytochrome c Solution. Dissolve 15/xM oxidized cytochrome c in 200 mM potassium phosphate buffer, pH 8.6, containing 100 txM EDTA. Ethanolic Solution of Nitecapone and 0R-1246. Dissolve nitecapone or OR-1246 in ethanol, in concentrations ranging from 1 to 10 mM. Assay Method Add 150/.d antioxidant solution and 1 ml oxidized cytochrome c solution to a silica cuvette. Add 0.5 ml superoxide solution and mix. Read the absorbance at 550 nm versus a blank prepared under similar conditions, except that the DMSO did not contain NaOH. Calculate the concentration 17 K. Hyland, E. Voisin, H. Banoun, and C. Auclair, Anal. Biochem. 135, 280 (1983).
[54]
ANTIOXIDANT ACTIVITY OF NITECAPONE AND O R - 1 2 4 6
531
20
15
10 0
O 0.0
• 0.3
Antioxidant
0.9
I 1.2
concentration
(mM)
0.6
Fir. 2. Effects of nitecapone and OR-1246 on reduction of cytochrome c by superoxide. Cytochrome c (15/~M) in 200 mM phosphate buffer containing 100/zM EDTA, pH 8.6, was exposed at 25° to superoxide solution, prepared from alkaline DMSO, in the presence of various concentrations of nitecapone (0) or OR-1246 (O). An experiment representative of four trials is reported.
of reduced cytochrome c by using an extinction coefficient of 21,000
M-~ cm-l. Results
When superoxide anion solution was added to cytochrome c, the protein was completely reduced. In the presence of the antioxidants, the reduction of cytochrome c was less extensive. The effect of the antioxidants was dose-dependent, and no difference was observed between nitecapone and OR-1246 (Fig. 2). In the presence of 200/zM antioxidant, 50% of cytochrome c was still reduced by superoxide; however, in the presence of 1 mM antioxidant, no reduction of the protein could be observed on addition of superoxide solution. The data thus indicate that nitecapone and OR-1246 are efficient scavengers of superoxide anion, and further that they are equally efficient. Scavenging Activity against Peroxyl Radicals Generated in Solution The quenching effects of nitecapone and OR-1246 on peroxyl radicals generated in aqueous solution are analyzed by the fluorescence assay described by Glazer. 18 The method is based on the loss of fluorescence 18 A. N. Glazer, this series, Vol. 186, p. 161.
532
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[54]
of a water-soluble protein, B-phycoerythrin, as consequence of chemical and conformational damage induced by reactive oxygen species. Compounds that scavenge peroxyl radicals will inhibit the decay of fluorescence of the protein. A water-soluble diazo initiator, AAPH, is used as a convenient source of aqueous peroxyl radicals. This compound is reported to produce a constant flux of peroxyl radicals on thermal decomposition) 9 Loss of fluorescence is monitored fluorometrically at 40 ° in a Perkin-Elmer (Norwalk, CT) MPF-44A fluorescence spectrophotometer with thermostatted cuvette holder and continuous magnetic stirring, using excitation at 540 nm and emission at 575 nm.
Preparation of Reagents B-Phycoerythrin. Dissolve 18 nM B-phycoerythrin in 75 mM sodium phosphate buffer, pH 7.0. Azo Initiator Solution. Dissolve 40 mM AAPH in 75 mM sodium phosphate buffer, pH 7.0. Store the solution on ice to prevent decomposition. Ethanolic Solutions of Nitecapone, 0R-1246, and Trolox. Dissolve 10 mM antioxidant in ethanol, then dilute in 75 mM sodium phosphate buffer, pH 7, to concentrations ranging from 30 to 300/~M. Assay Method Pipette 1.8 ml B-phycoerythrin solution into a silica fluorimetric cuvette and incubate at 40°C for 5 min. Add 200 /~1 AAPH solution and monitor the loss of fluorescence for about 15 min. Add 40/,d of antioxidant solution and monitor the loss of fluorescence for a further 45 min.
Results When AAPH was added to a solution of B-phycoerythrin at 40 °, a linear time-dependent decrease in the fluorescence characteristic of the protein was observed. On addition of nitecapone, OR-1246, or Trolox, a water-soluble vitamin E analog, the fluorescence decay of B-phycoerythrin induced by AAPH was arrested (Fig. 3) for a period of time that depended on the concentration of the antioxidant: 6.5, 12.6, or 19.1 min elapsed before the fluorescence of B-phycoerythrin began to decrease again in the presence of 1.5, 3.5, or 5.5/.tM nitecapone, respectively. The data indicate that the antioxidant reacted with the radical so efficiently that it competed with the protein and was consumed during the reaction. OR-1246 and Trolox showed the same effect as nitecapone. Trolox has 19 E. Niki, this series, Vol. 186, p. 100.
[54]
ANTIOXIDANT ACTIVITY OF NITECAPONE AND O R - 1 2 4 6
533
7O w
6'
o4 ~ 3 0
. 10
.
. 20
. . 30
40
50
I 60
Time (min)
FIG. 3. Effect of nitecapone and OR-1246 on peroxyl radicals generated in solution. A 5.6/zM solution of nitecapone (O) or OR-1246 (©) was added to a solution of 18 nM Bphycoerythrin exposed to 4 mM AAPH in 75 mM sodium phosphate buffer, pH 7.0, at 40°. The loss of fluorescence was monitored using excitation at 540 nm and emission at 575 nm. A similar curve was obtained in the presence of Trolox. The arrow indicates the addition of antioxidant. An experiment representative of four trials is reported.
been reported to interact with peroxyl radicals with a stoichiometry factor (moles of antioxidant able to trap 1 mole of peroxyl radical) of 2. 2°-22 By comparing the lag period for prevention of fluorescence loss by nitecapone and OR-1246 with that for prevention of fluorescence loss by Trolox, we calculate a stoichiometry factor of 2 for both nitecapone and OR-1246. Scavenging Activity against Peroxyl Radicals Generated in Membranes To assay the scavenging properties of nitecapone and OR-1246 on peroxyl radicals generated in membranes, we used the fluorimetric method of Tsuchiya e t al. 23 with a slight modification. The method is based on the decay of fluorescence of c i s - p a r i n a r i c acid, a hydrophobic reporting molecule, as a consequence of oxidation by peroxyl radicals generated in membranes. The hydrophilic azo initiator A M V N is used as a convenient source of peroxyl radicals. This compound, similar to A A P H , produces 20G. W. Burton, L. Hughes, and C. U. Ingold, J. Am. Chem. Soc. 105, 5950 (1983). 21L. R. C. Barclay, J. T. Locke, J. M. MacNeil, J. Van Kessel, G. W. Burton, and K. U. Ingold, J. Am. Chem. Soc. 106, 2479 (1984). 22y. Yamamoto, S. Haga, E. Niki, and Y. Kamiya, Bull. Chem. Soc. Jpn. 57, 1260 (1988). 23M. Tsuchiya, G. Scita, H. J. Freisleben, V. E. Kagan, and L. Packer, this series, Vol. 213, p. 460.
534
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[54]
a constant flux of peroxyl radicals on thermal decomposition. 19 Compounds that scavenge peroxyl radicals generated by AMVN in membranes will inhibit the decay of fluorescence of cis-parinaric acid. 23 DOPC liposomes are used as membrane systems. The loss of fluorescence of cis-parinaric acid is monitored fluorometrically at 40° in a PerkinElmer MPF-44A fluorescence spectrophotometer provided with continuous magnetic stirring and a thermostatted cuvette holder, using excitation at 328 and emission at 415 nm.
Preparation of Reagents Liposomes. Put 2 ml of a chloroform solution of DOPC (20 mg/ml) in a test tube and dry with a stream of N2. Add 4 ml of 20 mM Tris-HC1 buffer, pH 7.4. Vortex for 5 min and then sonicate for 10 min with a probe sonicator. Incubate at room temperature for 30 min. Dilute the liposome suspension by adding 16 ml of 20 mM Tris-HCl buffer, pH 7.4. Azo Initiator Solution. Dissolve 100 mM AMVN in ethanol and store in ice to prevent decomposition. cis-Parinaric Acid Solution. Dissolve 1 mM cis-parinaric acid in ethanol. Ethanolic Solution of Nitecapone and 0R-1246. Dissolve 10 mM antioxidants in ethanol, then dilute in 75 mM sodium phosphate buffer, pH 7, at concentrations ranging from 30 to 300/zM. Assay Method Add 3 ml liposome solution to a silica fluorimetric cuvette and warm at 40° for 5 min. Add 10/zl of AMVN solution and 40/~1 of antioxidant solution, then sonicate for 20 sec in a bath sonicator and incubate at 40° for an additional 5 min. Add 20/zl of cis-parinaric acid solution and monitor the loss of fluorescence by recording for 20 to 30 min.
Resul~ A time-dependent decrease in the fluorescence of cis-parinaric acid was observed in the presence of AMVN. Both nitecapone and OR-1246 inhibited the loss of fluorescence (Fig. 4). The value of V/Va, the ratio of the rate of fluorescence decay measured without antioxidant (V) to that measured with 5/xM, antioxidant (Va), was 1.50 -+ 0.091 and 2.21 -+ 0.039 for nitecapone and OR-1246, respectively. These values are significantly different from one another atp < 0.05, indicating better antioxidant properties for OR-1246 than nitecapone.
[54]
ANTIOXIDANT ACTIVITYOF NITECAPONEAND OR-1246
535
o ~o,
6
~4.
I 0
10
20
30
40
T i m e (min) FIG. 4. Effect of nitecapone and OR-1246 on peroxyl radicals generated in membranes.
cis-Parinaric acid (6.5/zM), added to a suspension of 2 mg/ml DOPC liposomes in 20 mM Tris-HCl buffer, pH 7.4, was exposed to 0.3 mM AMVN at 40° in the absence of antioxidant (A) or in the presence of 5 t~M nitecapone (0) or 5 /zM OR-1246 (O). An experiment representative of five trials is reported.
Effect on Induced Lipid Peroxidation: Spectrophotometric Assay of Thiobarbituric Acid-Reacting Substances The ability of nitecapone and OR-1246 to scavenge A M V N - g e n e r a t e d peroxyl radicals was also analyzed by measuring their effect on lipid peroxidation of m i c r o s o m e s e x p o s e d to A M V N . The spectrophotometric assay of thiobarbituric acid-reacting substances (TBARS) is used for measuring lipid peroxidation. 24 C o m p o u n d s that interact with reactive oxygen species will inhibit the formation T B A R S in m e m b r a n e systems exposed to oxidative stress. Rat liver m i c r o s o m e s are exposed to peroxyl radicals generated by the thermal decomposition of A M V N at 40 °. The formation of T B A R S is monitored spectrophotometrically at 532 nm.
Preparation of Reagents Rat Liver Microsome Suspension. Prepare rat liver m i c r o s o m e s as reported b y H o g e b o o m . 25 Dilute the m i c r o s o m e suspension in 20 m M sodium p h o s p h a t e buffer, p H 7.4, to a final concentration of 2.5 mg protein/
24T. F. Slater, this series, Vol. 105, p. 283. 25G. H. Hogeboom, this series, Vol. 1, p. 16.
536
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[54]
ml. Microsomal proteins are assayed by the sodium dodecyl sulfate (SDS)-Lowry method according to Markwell. 26 Azo Initiator Solution. Dissolve 100 mM AMVN in ethanol and store in ice to prevent decomposition. Ethanolic Solutions of Nitecapone and 0R-1246. Dissolve nitecapone or OR-1246 in ethanol, in concentrations ranging from 1 to 10 mM. Thiobarbituric Acid Solution. Dissolve 0.67 mg of thiobarbituric acid in 100 ml water. Trichloroacetic Acid Solution. Add 30 ml trichloroacetic acid (TCA) to 70 ml water. Butylated Hydroxytoluene Solution. Dissolve 0.2 mM BHT in ethanol.
Incubation Conditions Put 1 ml of microsome suspension in a test tube. Add 10/zl antioxidant solution. Incubate at 40° for 15 min. Add 25/~1 AMVN solution. Incubate at 40° for 60 min.
Assay for Thiobarbituric Acid-Reactive Substances To 1 ml of the incubation mixture, add 1 ml TCA solution, 1 ml TBA solution, and 150/.d BHT. Heat the sample at 100° for 20 min. Cool the suspensions at room temperature and centrifuge at 20,000 g for 20 min. Remove the upper layer and measure the optical density at 532 nm against a water blank. Calculate the concentration of TBARS as concentrations of malondialdehyde (MDA), using an extinction coefficient of 1.56 × 105 M-1 cm-l.
Results A decrease in the formation of TBARS was observed in rat liver microsomes exposed to AMVN in the presence of nitecapone or OR-1246 (Fig. 5). The effect was dependent on the concentration of antioxidant; however, at all the concentrations used, OR-1246 inhibited TBARS formation more effectively than nitecapone (p < 0.05). Xanthine Oxidase Activity Compounds that interact with xanthine oxidase can affect the kinetics of the reaction of oxidation of xanthine to uric acid. The reaction can be monitored spectrophotometrically at 295 nm. 27 26 M. A. K. Markwell, S. M. Haas, L. L. Bieber, and N. E. Tolbert, Anal. Biochem. 87, 206 (1978). 27 I. Fridovich, in "Handbook of Methods for Oxygen Radical Research" (R. A. Greenwald, ed.), p. 51. CRC Press, Boca Raton, Florida, 1985.
[54]
ANTIOXIDANT ACTIVITY OF NITECAPONE AND OR-1246
537
--. 6
~
3 0
10 20 30 40 Antioxidant concentration (~M)
1 50
FIG. 5. Effect of nitecapone and OR-1246 on AMVN-induced lipid peroxidation in rat liver microsomes. Lipid peroxidation was measured as production of thiobarbituric acidreacting substances (TBARS) after the incubation of rat microsomes (2.5 mg protein/ml) with 2.5 mM AMVN for 1 hr at 40° in the presence of nitecapone (O) or OR-1246 (O). Average values from four experiments -+ SD are reported.
Preparation of Reagents Xanthine Oxidase Solution. Dilute xanthine oxidase to a final concentration of 250 m U / m l in 0.1 M p h o s p h a t e buffer, p H 7.4. Xanthine Solution. Weigh 1.74 mg of xanthine, add 10 ml water, then boil for 5 min to completely dissolve the powder. Ethanolic Solution of Nitecapone and 0R-1246. Dissolve nitecapone or OR-1246 in ethanol to create solutions at final concentrations between 1 and 10 mM. Assay Methods Pipette 0.9 ml of 0.1 M phosphate buffer, p H 7.4, into a quartz cuvette. Add 10 tzl antioxidant solution and 15/xl xanthine both to the sample and to the reference. Start the reaction by adding 10/xl of xanthine oxidase solution to the sample cuvettes. Record the kinetics of the reaction at 295 nm for 3 min at r o o m temperature.
Results N i t e c a p o n e has previously been found to be a competitive inhibitor of xanthine oxidase activity. 4 Although a dose-dependent inhibition of the
538
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[54]
90 0
60
!
0
50 Antioxidant
1OO
!
150
concentration
{pM}
FIG. 6. Effect of nitecapone and OR-1246 on xanthine oxidase activity. Xanthine oxidase activity (2.5 mU/ml) was measured at 295 nm by monitoring the forrnation of uric acid from 15 tiM xanthine at 25° in 100 mM phosphate buffer, pH 7.4, in the presence of various concentrations of nitecapone (0) or OR-1246 (O). Average values from three experiments + SD are reported.
enzyme was also observed in the presence of OR-1246, the effect was significantly weaker than that of nitecapone (p < 0.01). Xanthine oxidase activity was completely inhibited in the presence of 150 ftM nitecapone, whereas OR-1246 caused only 40% inhibition at the same concentration (Fig. 6). Partition between Aqueous and Membrane
Phase
Depending on its solubility, a substance distributes between the membrane and aqueous phases. The aqueous concentration of a substance with a hydrophobic character decreases when a hydrophobic system is added. The concentration of such a substance in the aqueous system can be measured before and after the addition of the hydrophobic system, and its concentration in the hydrophobic system can then be calculated by differencefls
Preparation of Reagents Liposome Preparation. Put 4 ml of an ethanolic lipid solution (20 mg/ ml) in a test tube and dry with a stream of N2. Add 4 ml of 20 mM sodium 28 A. Leo, C. Hansch, and D. Elkins, Chem. Rev. 71, 525 (1971).
[54]
ANTIOXIDANT ACTIVITY OF NITECAPONE AND O R - 1 2 4 6
539
phosphate buffer, pH 7.4, then vortex for 5 min. Sonicate for 10 min with a probe sonicator. Incubate at room temperature for 30 rain. Aqueous Solution of Nitecapone or 0R-1246. Weigh 20 mg of antioxidant and add 1 ml of 20 mM sodium phosphate buffer, pH 7.4. Vortex for 5 min, then filter with a 0.22 /.~m Millex-Gv filter unit (Millipore, Bedford, MA). Measure the optical density of the filtered solution at 380 nm for nitecapone and at 444 nm for OR-1246. Calculate the concentration of the antioxidant in solution by using the extinction coefficients reported in Table I. Dilute in 20 mM sodium phosphate buffer, pH 7.4, to create solutions at final concentrations between 5 and 100 p.M.
Assay Method Put I ml of liposome suspension in an Eppendorf tube. Centrifuge at 20,000 g for 15 min at room temperature, then remove the supernatant. Add 1 ml of antioxidant solution to the pellet. Resuspend the pellet with slow shaking for 10 min. Centrifuge again under the same conditions. Remove the supernatant and measure the optical density at 380 nm for nitecapone or at 444 nm for OR-1246. The supernatants from samples containing buffer without antioxidants are used as blanks.
Calculation To calculate the concentration of antioxidant in the membrane phase, subtract the antioxidant concentration in the recovered supernatant after centrifugation from the antioxidant concentration initially present in the solution added to the liposome pellet. The amount oflipids in the liposomes is calculated by using a molecular weight of 740 for phosphatidylcholine.
Results Nitecapone and OR-1246 distribute in soybean PC liposomes in a linear dose-dependent manner (Fig. 7). OR-1246 appeared more hydrophobic than nitecapone (p < 0.01). About 60% of OR-1246 or 30% of nitecapone added in solution could be recovered in the liposome pellet. Similar distributions were obtained with various membrane sources, including DOPC liposomes, dipalmitoylphosphatidylcholine (DPPC) liposomes, or rat liver microsomes (data not shown). Concluding Remarks No difference was observed between the antioxidant properties of nitecapone and OR-1246 against superoxide and peroxyl radicals gener-
540
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[54]
4
o't:l ~
2
0
• 0
2
4
6
h 8
Added a n t i o x i d a n t (nmol/mmol lipids) FIG. 7. Partition of nitecapone and OR-1246 in soybean PC liposomes. Soybean PCliposomes (20 mg lipids) were added to an aqueous solution of nitecapone (0) or OR-1246 (O) in 20 mM phosphate buffer, pH 7.4. The concentration of antioxidant in the liposomes was calculated as described in the text. Average values from three experiments -+ SD are reported.
ated in solution (Figs. 2 and 3). However, the two molecules have different antioxidant properties at the level of target sites such as membranes and proteins; OR-1246 scavenges peroxyl radicals in liposomes (Fig. 4) and protects microsomal membranes against lipid peroxidation (Fig. 5) better than does nitecapone. In contrast with the results in membranes, nitecapone inhibits xanthine oxidase more efficiently than does OR-1246 (Fig. 6). As can be calculated by using the Henderson-Hasselbalch equation and the values of pK a reported in Table I, nitecapone and OR-1246 are uncharged below pH < 2, whereas they are at least 90% in the dissociated anionic form at physiological pH. However, the different resonance structures that characterize nitecapone and OR-1246 could cause different distributions of the anionic charge over the molecules. Owing to the presence of the methylene double bond, the negative charge of nitecapone could be destabilized over the entire molecular structure, whereas the absence of the double bond in OR-1246 could cause the negative charge to be more localized over the aromatic ring than over the side chain. The smaller negative charge over the side chain of OR-1246 could enable this portion of the molecule to interact better with hydrophobic systems. The changing of electrostatic properties of nitecapone and OR-1246 with pH could explain their having the same partition coefficient in the octanol/HCl system, and the different partition coefficients we observed
[54]
ANTIOXIDANT ACTIVITY OF NITECAPONE AND OR-1246
541
in the octanol/buffer system at physiological pH (Table I). Furthermore, the more hydrophobic nature of OR-1246 at physiological pH could make access to the membrane compartment easier and thus explain the higher concentration of OR-1246 than nitecapone in membranes suspended in aqueous solution at pH 7.4 (Fig. 7). In heterogeneous membrane systems the efficiency of an antioxidant depends on its chemical reactivity, the uniformity of its distribution in the lipid bilayer, and the mobility of membrane lipids, which allows for mutual accessibility of lipid radicals and antioxidants. However, the concentration of an antioxidant in the membrane phase is also an important element in its rate of interaction with radicals. Nitecapone and OR-1246 also had different effects on xanthine oxidase activity. The different steric structure of nitecapone resulting from the presence of the double bond could favor its interaction with xanthine oxidase. However, the different inhibitory effects of nitecapone and OR1246 on xanthine oxidase activity could also be a consequence of the different charge distributions over the structure of the antioxidant molecules. Charge interaction between antioxidant and protein could modulate the binding of the antioxidant to sites of xanthine oxidase important for catalytic action. The differential interaction of nitecapone and OR-1246 with xanthine oxidase could be of physiological importance because xanthine oxidase is one of main sources of cellular superoxide, and the cellular concentration of this protein has been shown to increase under ischemia-reperfusion conditions. 29 The comparative study performed on nitecapone and OR-1246 demonstrated that one can manipulate the actions of an antioxidant by making a minimal structural modification. Alterations of function brought about by such a small structural change can affect the interactions of an antioxidant with the wide variety of molecules taking part in the complex set of processes involved in free radical generation and oxidative damage in biological systems. If one can manipulate the structure of an antioxidant, based on precise understanding of the consequences of the changes at various sites, then beneficial antioxidants can be developed. Acknowledgments Research was supported by the National Institutes of Health (CA 47597, CA 53812) and Orion Research Center, and was performed during the tenure of a research Fellowship from the American Heart Association, California Affiliate, to Y.J.S. We thank Dr. E. Nissinen and T. Lotta, Orion-Farmos Pharmaceuticals, for providing physicochemical data on nitecapone and OR-1246. 29 j. McCord, N. Engl. J. Med. 312, 159 (1985).
542
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[55]
[55] P e n i c i l l a m i n e as A n t i o x i d a n t
By MASAYUKI MIKI Introduction D-Penicillamine (2-amino-3-mercapto-3-methylbutanoic acid) is a trifunctional thiol amino acid.l The sulfhydryl group of penicillamine enables it to chelate divalent cations, particularly copper. It is widely used for the treatment of a variety of disorders including Wilson's disease,Z rheumatoid arthritis, 3 and heavy metal poisoning, 4 in which an involvement of freeradical reactions catalyzed by transition metals has been suspected. If hydroxyl radicals (.OH) are formed from hydrogen peroxide by low molecular mass metal complexes in free solution and then have to diffuse a short distance to attack a target, the protection by added .OH scavengers should be seen. 5 Otherwise superoxide dismutase and H202-removing enzymes must be effective by stopping superoxide anion (Oz~) and HEO2 from ever reaching the bound metal complexes. Nevertheless, the .OH formation is so site-specific and sequestered that almost any protection by scavenging is impossible .6 Protection here can be achieved by chelating agents that pull the metal ions away from sensitive sites and render them inactive. Wilson's disease is an inborn error of copper metabolism characterized by copper retention eventually leading to toxicity in multiple organs including the liver and brain7; therefore, it requires lifelong treatment. The management of Wilson's disease was revolutionized by the introduction of peniciUamine by Walshe in 1956. 2 Undesirable side effects do occur in 5 to 10% of patients, but D-penicillamine remains the standard treatment for the disease. It is also used as an anti-inflammatory drug, for instance, in the treatment of rheumatoid arthritis in which its mechanism of action is unknown but appears to be related to its copper-complexing abilities. 8 1 W. M. Weigert, H. O f f e r m a n n s , and P. Scherberich, Angew. Chem. 87, 374 (1975). 2 j. M. Walshe, Am. J. Med. 21, 487 (1956). 3 I. J. Jaffe, Arthritis Rheum. 8, 1064 (1965). 4 L. T. Z i m m e r and D. E. Carter, Life Sci. 23, 1025 (1978). 5 G. C o h e n , Photochem. Photobiol. 28, 669 (1978). 6 j. M. C. Gutteridge, this series, Vot. 186, p. 1. 7 D. M. Danks, in " T h e Metabolic Basis of Inherited D i s e a s e " (C. R. Scriver, A. L. Beaudet, W. S. Sly, and D. Valle, eds.), 6th Ed., Vol. 2, p. 1411. McGraw-Hill, N e w York, 1989. 8 j. R. J. Sorenson, Met. Ions Biol. Syst. 14, (1982).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994by AcademicPress, Inc. All rights of reproduction in any form reserved.
[55]
PENICILLAMINE AS ANTIOXIDANT
543
Unexpected Prooxidant Effect of Penicillamine in Presence of Copper Ions During episodes of fulminant hepatitis in Wilson's disease, hepatic copper is probably released in massive amounts. Such nonceruloplasmin copper levels in plasma are supposedly comparable to those seen in acute copper poisoning, and indeed hemolytic crises occur similarly in both cases.9 If copper ions released into the bloodstream catalyze the generation of free radicals, the consumption of radical-trapping plasma antioxidants should be a possible initial event. E x p e r i m e n t s . The total radical-trapping antioxidant power (TRAP) of the plasma of a patient is measured according to the method of Wayner et al. 1o The oxidation of linoleic acid in the presence of plasma is induced by 2,2'-azobis(2-amidinopropane) dihydrochloride (AAPH) at 37° in 20 mM phosphate buffer (pH 7.4) containing 1 mM EDTA and is monitored by oxygen consumption using an oxygen electrode. The oxidation of linoleic acid is initially inhibited by plasma antioxidants to form an induction period and then progresses rapidly. The duration of induction observed is quantified as an index of TRAP using the duration produced by a known quantity of Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid) with its stoichiometric factor, n equal to 2. The values obtained in the assay are expressed in terms of micromolar. Measurements showed low TRAP values in response to high levels of nonceruloplasmin copper (Fig. 1). 11 Serum nonceruloplasmin copper is detected using the phenanthroline assay, 12 which is based on the fact that a phenanthroline-copper ion complex can degrade DNA in the presence of a reducing agentand yields the thiobarbituric acid (TBA)-reactive chromogen. Loosely bound copper, like copper ions attached to albumin or to amino acids, is available for chelation by phenanthroline, but copper incorporated into ceruloplasmin is not available. Penicillamine does not affect the ability of copper ions to react with phenanthroline, so penicillamine-copper complexes should be measurable in the assay. 13 Our results also show that the low TRAP is attributable mainly to the depletion of urate in patient plasma. 11 Under the pathological condition which causes release of copper ions, thiols appear to switch from antioxidant to prooxidant activity, that is, 9 j. McC. Howell, S. R. Gooneratne, and J. M. Gawthome, Comp. Pathol. Bull. 16, 3 (1984). ~0D. D. M. Wayner, G. W. Burton, K. U. Ingold, L. R. C. Barclay, and S. J. Locke, Biochim. Biophys. Acta 924, 408 (1987). II M. Miki, unpublished data (1992). 12 j. M. C. Gutteridge, Biochem. J. 218, 983 (1984). 13 p. j. Evans, A. Bomford, and B. Halliwell, Free Radical Res. Commun. 7, 55 (1989).
544
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[55]
1200" 1000 q"
~ 800 ~. 6oo
F-4OO 2000
o
i
!
i
0 1 2 3 4 Nonceruloplasmin copper (I~M) FIG. 1. Relationship of plasma radical-trapping capacity to copper level in Wilson's disease. The radical-trapping capacity of plasma was examined by the TRAP assay. Nineteen healthy age-matched children were used as normal controls, showing TRAP values of 856 ___ 104 p,M (mean --- SD). Nonceruloplasmin copper was detected using the phenanthroline assay. Phenanthroline-detectable copper is not present in freshly prepared plasma from normal humans. 12,13The results show an inverse correlation of TRAP values with nonceruloplasmin copper levels (y = 901 - 158x, r = -0.796, p < 0.01).
they are oxidized and form H 2 0 2 in the presence of copper and oxygen. They also reduce copper which is reoxidized by reacting with H202 to generate .OH. Because of the dual role of thiols, copper-binding sites could serve as catalytic centers for repeated production of .OH. 14 To investigate this possibility, ghost membranes are prepared by the method of Burton e t al. 15 from erythrocytes of adult male Wistar rats, which are maintained on a vitamin E-deficient diet (supplied by Eisai Co., Tokyo, Japan).16 The ghosts obtained are suspended in 0.1 M phosphate buffer (pH 7.0) at a concentration of 1.5 mg protein/ml. The ghost suspension is mixed with CuSO4 and with additives, if necessary, and is incubated at 37° for 60 min. The final concentration of copper salt is 10/zM. An aliquot of the reaction mixture is combined with two parts TBA reagent [0.375% TBA in 15% trichloroacetic acid (TCA), to which HC1 is added at 0.25 N] and heated for 15 min in a boiling water bath. The absorbance t4 M. Chevion, Free Radical Biol. Med. 5, 27 (1988). t5 G. W. Burton, K. U. Ingold, and K. E. Thompson, Lipids 16, 946 (1981). 16 y . Takenaka, M. Miki, H. Yasuda, and M. Mino, Arch. Biochem. Biophys. 255, 344 0991).
[55]
PENICILLAMINE AS ANTIOXIDANT
545
TABLE I EFFECTS OF ADDITIVESa ON GHOST MEMBRANE OXIDATION CATALYZED BY COPPER IONS Additive SOD CP EDTA CySH Catalase Zn PCM TOC
Relative formation of TBA reactants b 105 144 10 334 118 70 404 34
_+ 10 -+ 23 + 9 -+ 56 -+ 15 + 31 _+ 93 - 12
Final concentrations of additives were as follows: superoxide dismutase (SOD), 20 units/ml; catalase, 4 ttg/ml; ceruloplasmin (CP), 120/zg/ml; ZnC12 (Zn), 100 /xM; EDTA, 1 mM; D-penicillamine (PCM), 1 mM; cysteine (CySH), 1 mM; and c~-tocopherol (TOC), 2.1 gM. Erythrocyte ghosts, containing atocopherol in the membranes, were obtained from rats after intramuscular injection of a-tocopherol micelles (50 mg/kg body weight). b Data are expressed as percentages as compared to the complete system without any additives and are means - standard deviations of three experiments.
of the supernatant is measured at 535 nm against a blank that contained all the reagents except copper (Table I).17 The oxidation of ghosts catalyzed by copper ions appears to be "sitespecific" and in close proximity to membrane components, explaining the inability of SOD and catalase to inhibit the formation of TBA reactants. The results also indicate what chelating agents should be used when copper ions are released massively into the bloodstream. EDTA renders copper ions less reactive in membrane lipid peroxidation. However, copper-penicillamine chelates are effective in catalyzing the oxidation of ghost membranes, which may be due to favorable alterations in the redox potential of copper. This acceleration effect of penicillamine may explain the initial aggravation of neurologic symptoms often observed with its use.18 17 M. Miki, H. Yasuda, T. Motoyama, T. Tanabe, M. Mino, and N. Taniguchi, in "Active Oxygen, Lipid Peroxides and Antioxidants, Proceedings of the 5th International Conference of Oxygen Radicals" (K. Yagi, E. Niki, and M. Kondo, eds.). Elsevier, Amsterdam, 1992. is G. J. Brewer, R. D. Dick, V. Yuzbasiyan-Gurkin, R. Tankanow, A. B. Young, and K. J. Kluin, Arch. Neurol. (Chicago) 48, 42 (1991).
546
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[55]
TABLE II COPPER RELEASE AND MEMBRANE OXIDATION DURING REACTION OF HYDROGEN PEROXIDE WITH HEPATIC CYTOSOLSa
Formation of TBA reactants c (absorbance at 532 n m × 103)
Phenanthroline-detectable copper b (nmol/g protein) Reaction
Cytosol A
Cytosol B
Complete WithoutH202 Without cytosol
1560 ___98 d 31 ± 11
20 - 17 12± 9
Cytosol A
Cytosol B
263 ± 19d 86 - 15 6 0 ± 10 42 ± 7 53 - 5
Cytosol fractions were obtained from livers of LEC rats (Cytosol A) and Wistar rats (Cytosol B). b Copper release when hepatic cytosols were oxidized by H202 was examined by the phenanthroline assay. The results are expressed in terms of nanomoles of copper ions per gram of protein in the reaction mixtures and are means - standard deviations of four experiments. c The oxidation of ghost membranes induced by added H202 in the presence of hepatic cytosols was estimated by the formation of TBA reactants. Data represent means ± standard deviations of five experiments. d Statistically different (p < 0.01) from data for Cytosol B. P e n i c i l l a m i n e E f f e c t A t t r i b u t e d to D e t o x i f i c a t i o n o f Copper Rather Than Decoppering T h e l o g i c f o r t h e u s e o f p e n i c i l l a m i n e in W i l s o n ' s d i s e a s e h a s b e e n b a s e d o n its e x p e c t e d p o t e n t i a l a s a c h e l a t o r to r e m o v e c o p p e r t h a t a c c u m u l a t e s in l i v e r a n d o t h e r t i s s u e s as a c o n s e q u e n c e o f t h e i m p a i r e d b i l i a r y excretion.'9 Penicillamine can chelate exogenous copper but not remove c o p p e r t h a t is a l r e a d y i n c o r p o r a t e d i n t o c y t o s o l i c p r o t e i n s . 2° T h e r e f o r e , patients may not be truly "decoppered" although penicillamine provokes a t r a n s i e n t r i s e in u r i n a r y c o p p e r e x c r e t i o n . 21 It h a s b e e n p r o p o s e d t h a t t h e l o n g - t e r m i m p r o v e m e n t t h a t is o b s e r v e d in p a t i e n t s w h o h a v e b e e n g i v e n p e n i c i l l a m i n e is d u e to i n c r e a s e d h e p a t i c l e v e l s o f m e t a l l o t h i o n e i n . 2 l M o r e o v e r , t h e r e is a r e p o r t t h a t t h e a d m i n i s t r a t i o n o f p e n i c i l l a m i n e to r a t s i n d u c e s h e p a t i c m e t a l l o t h i o n e i n s y n t h e s i s . 22 M e t a l l o t h i o n e i n s a r e c a p a b l e o f b i n d i n g c o p p e r a n d s t a b i l i z i n g it in a n o n t o x i c f o r m , 23 a n d d e c r e a s e s in m e t a l l o t h i o n e i n m a y l e a v e t h e p a t i e n t s 19v. Iyengar, G. J. Brewer, R. D. Dick, and C. Owyang, J. Lab. Clin. Med. 111, 267 (1988). 2oA. McQuaid and J. Mason, J. Inorg. Biochem. 41, 87 (1990). 21 I. H. Scheinberg, I. Sternlieb, M. Schilsky, and R. J. Stockert, Lancet 1, 95 0987). 22 H. E. Heilmaier, J. L. Jiang, H. Greim, P. Schramel, and K. H. Summer, Toxicology 42, 23 (1986). 23 I. Bremner, this series, Vol. 205, p. 25.
[55]
PENICILLAMINE AS ANTIOXIDANT
547
in a vulnerable state. This hypothesis would explain why a number of patients who experienced good long-term control with penicillamine but who elected to cease the treatment have died from rapidly progressing liver disease (more rapidly than one would expect toxic amounts of copper to reaccumulate). 24 The situation is compared to a "time bomb" in which high levels of copper are still present in the liver even though copper ions are stabilized by metallothionein. Copper-metallothionein would play the role of an oxidant if thiolate bonds in metallothionein are oxidized and copper ions are released from metaUothionein. Experiments. A mutant strain of Long-Evans rats (LEC, Long Evans Cinnamon) has been found to suffer a syndrome bearing close similarity to Wilson's diseasCS: (1) onset of acute hepatitis as sociated with hemolytic crisis at the age of 4 months; (2) Progressive accumulation of copper in the liver 50 times as great as that of Wistar rats; (3) Low level of serum ceruloplasmin and impaired biliary excretion of copper. It has been reported that hepatic levels of metallothionein are high in the patients and also in the LEC rats and that metallothionein contained the majority of cytosol copper. 26 Cytosol fractions are obtained from 10% liver homogenates in 0.1 M phosphate buffer (pH 7.0). Hepatic cytosols are treated with 1 mM sodium azide and added into a red blood cell (RBC) ghosts suspension (1.5 mg protein/ml). The oxidation of ghosts is initiated by the addition of H202 at a final concentration of 1 mM. The formation of TBA reactants in the mixtures is measured after 60 min of incubation at 37°. Additionally, copper r e k , s e is examined by the phenanthroline assay ~2 when hepatic cytosol is oxidized with H202 (Table II). 17 Our results show that hepatic cytosols of LEC rats can release copper ions in massive amounts when they are oxidized with H202 and thereby progressively promote the oxidation of ghost membranes. The biological implications are that the liver of LEC rats is significantly susceptible to oxidative stress because of its high content of copper metallothionein. This may also explain the clinical feature of Wilson's disease that stress such as infection sometimes triggers fulminant hepatitis with hemolytic crisis.7 Based on this view, thiomolybdates appear to be more appropriate drugs than penicillamine 18because they are much more effective in eliminating exogenous copper and are even capable of removing metallothionein-bound c o p p e r . 2°'27 24 j. M. Walshe and A. K. Dixon, Lancet 1, 845 (1986). 25 T. Okayasu, H. Tochimaru, T. Hyuga, T. Takahashi, Y. Takekoshi, Y. Li, Y. Togashi, N. Takeichi, N. Kasai, and S. Arashima, Pediatr. Res. 31, 253 (1992). 26 N. O. Nartey, J. V. Frei, and M. G. Cherian, Lab. Invest. 57, 397 (1987). 27 A. McQuaid, J. Lab. Clin. Med. 119, 744 (1992).
548
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[56]
[56] A n t i o x i d a n t A c t i o n o f L a z a r o i d s
By EDWARD D. HALL and JOHN M. MCCALL Introduction The role of oxygen free radical-induced lipid peroxidation in the acute pathophysiology of central nervous system (CNS) injury and ischemia (i.e., stroke) has been fairly well established) -3 Thus, efforts have been directed toward the discovery of effective lipid antioxidant compounds that can retard posttraumatic and postischemic neurodegeneration. The 21-aminosteroids, or lazaroids, are a novel series of compounds being developed for the acute treatment of traumatic or ischemic CNS injury; they have been specifically designed to localize within cell membranes and inhibit lipid peroxidation reactions. Morever, lipid peroxidation may play a role in chronic neurodegenerative disorders, and thu~ these novel lipid antioxidants may find utility in these areas as well. The 21-aminosteroids are the products of an effort to develop nonglucocorticoid steroids that duplicate the cerebroprotective pharmacology of synthetic glucocorticoid steroids. Extensive studies with the glucocorticoid steroid methylprednisolone (see Fig. 1) had indicated that large intravenous doses (30 mg/kg) could ameliorate many of the pathophysiological consequences of traumatic or ischemic injury in the CNS, and promote the functional recovery of experimentally injured animals by inhibiting posttraumatic lipid peroxidation. 4 The definition of this highdose, nonglucocorticoid antioxidant action led to the synthesis of a number of nonglucocorticoid steroid analogs of methylprednisolone (e.g., U-72099E; Fig. 1), which weakly inhibited lipid peroxidation in high concentrations and at high doses were active in models of experimental CNS trauma. 5 However, these became springboards for compounds that would be more potent and effective inhibitors of lipid peroxidation, with greater activity in experimental models of CNS trauma and ischemia. 6
l j. M. Braughler and E. D. Hall, Free Radical Biol. Med. 6, 289 (1989). 2 E. D. Hall and J. M. Braughler, Free Radical Biol. Med. 6, 303 (1989). 3 E. D. Hall and J. M. Braughler, in "Molecular and Cellular Approaches to the Treatment of Neurological Disease" (S. G. Waxman, ed.), p. 81, Raven, New York, 1993, 4 E. D. Hall, J. Neurosurg. 76, 13 (1992). 5 E. D. Hall, J. M. McCall, P. A. Yonkers, R. L. Chase, and J. M. Braughler, J. Pharmacol. Exp. Ther. 242, 137 (1987). 6 E. D. Hall, Ann. Neurol. 32, S137 (1992).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[56]
LAZAROIDS
549
METHYLPREDNISOLONESODIUMSUCCINATE CHz-O-C-CH2CH2-C-O"Na+ ] ii ii C=O 0 0 O~
O
H
CH3
U72099E-NON-GLUCOCORTICOID STEROID CH2-O-C-CHzCH2-C-O"Na+ I ii ii
c,a~ o
c=o o
C H s ~ ' '
'
o
• x H20
, & ].,,,L J 0-~ v 21-AMINOSTER01DS(LAZAROIDS) U74500A
U741~F (TIRILAZADMESYLATE)
c.~ CHz.-N
N,,-Ct
~=o ~ J
i ~ ~
"CHs
">-N-CH2CH s
"N=~
,N-C.~C..
CH2 CH3
0~
•
x HCI
FIG. 1. Chemicalstructures of the glucocorticoidmethylprednisolone,the nonglucorticoid steroid U-72099E, and the 21-aminosteroids U-74006FandU-74500A. The first compound in the 21-aminosteroid or lazaroid series was synthesized in 1985. U-74006F (Fig. 1; generic name tirilazad mesylate, trade name Freedox) has been selected for clinical development for the acute treatment of brain and spinal injury, subarachnoid hemorrhage, and stroke, and is currently involved in Phase III clinical trials in each of those disorders. In the present discussion, a main focus is on the known antioxidant mechanisms of U-74006F. Another 21-aminosteroid, U-74500A (Fig. 1), is actually a more potent inhibitor of iron-catalyzed lipid peroxidation than U-74006F, but it has not been chosen for development owing to pharmaceutical instability and rapid elimination in vivo.
Mechanisms of Lipid Peroxidation Inhibition U-74006F is a very lipophiliccompound (the log of the calculated octanol/water partitioncoefficientequals 8) that distributespreferentially
550
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[56]
to the lipid bilayer of cells. It appears that the compound exerts its effects through cooperative effects: a radical scavenging action (i.e., chemical antioxidant effect) and a physicochemical interaction with the cell membrane that serves to decrease membrane fluidity (i.e., membrane stabilization). Cytoprotection is governed partly by intrinsic reactivity toward free radicals, partly by location and orientation of U-74006F within the membrane, and partly by the ability of the compound to modify the physical properties of the lipid bilayer of the membrane. Activity in different in oitro and in vivo models probably reflect different balances between these actions. Antioxidant Effects in Membrane Systems
The 21-aminosteroids are potent inhibitors of lipid peroxidation in vitro. Using rat brain homogenates or purified rat brain synaptosomes as the lipid source, U-74006F and U-74500A potently inhibit iron-dependent lipid peroxidation, with an efficacy greatly surpassing that of the glucocorticoid steroid methylprednisolone. In a model that uses synaptic membranes prepared from rat brain as a lipid source and 200/zM ferrous chloride to initiate and catalyze the lipid peroxidation reactions, U-74006F inhibited lipid peroxidation with IC50 values ranging from 10 to 60/zM. 7 U-74006F also protects isolated liver microsomes from oxidative injury that is initiated by ferrous ammonium sulfate. The IC50 value is 3.8/.~M when the U-74006F is added in ethanol to a suspension of microsomes in Krebs buffer. Interestingly, when U-74006F is added as a lipid emulsion (triglyceride, phosphatidylcholine, drug, and water) rather than in ethanol, the IC50 value drops to below 0. I/zM. This illustrates one of the problems in testing very lipophilic compounds like U-74006F. When such compounds are added in organic solution to physiological buffers, they microprecipitate. Emulsion delivery is often a superior delivery technique for compounds of this class. Since many of the lipid peroxidation models involve initiation of oxidative injury by iron, the 21-aminosteroids have been mistakenly described as inhibitors that affect exclusively iron-dependent lipid peroxidation. However, we have also studied iron-free systems and shown lipid antioxidant effects of U-74006F and U-74500A. In addition, U-74006F has been shown effective in a model of lipid peroxidation that involved rat liver microsomes with initiation by cumene hydroperoxide. Free iron was removed with a chelation column, thus ensuring that the system was truly 7 j. M. Braughler, J. F. Pregenzer, R. L. Chase, L. A. Duncan, J. M. McCall, and E. J. Jacobsen, J. Biol. Chem. 262, 10438 (1987).
[56]
LAZAROIDS
551
iron-free. 8 U-74006F has further been demonstrated to inhibit diquatinduced lipid peroxidation in liver microsomes. 9 U-74006F and U-74500A have been reported to scavenge lipid peroxyl and phenoxy radicals in a methanol solution of linoleic acid in the presence of 2,2'-azobis(2,4-dimethylvaleronitrile) (AMVN) which induces peroxidation of the polyunsaturated linoleic acid, although U-74006F and U-74500A both possess slower rate constants in this environment than the prototypical peroxyl radical scavenger vitamin E. However, both compounds act to slow the oxidation of vitamin E during linoleic acid peroxidation and potentiate the antioxidant efficacy of vitamin E. ~° We further studied the lipid radical scavenging properties of the 21aminosteroids in three different models. H Model 1 involves a homogeneous methanolic solution of linoleic acid as the peroxidizable lipid substrate in methanol, with AMVN as the free radical initiator as previously described. 9 Model 2 involves multilamellar vesicles of dilinoleyllecithin with 2,2'-azobis(2-amidinoaminopropane), ABAP, as the water-soluble initiator. In Model 2, U-74006F or U-74500A was incorporated in the multilamellar vesicle as it was prepared. Both initiators are thermally activated and produce lipid free radicals at a constant and readily reproducible rate, thereby creating a steady-state kinetic system. Hydroperoxide formation was measured by high-performance liquid chromatography (HPLC) in Model I and by a xylenol orange color test for lipid hydroperoxides in Model 2; hydroperoxide LOOH formation was found to be linear for the time periods measured. The rates of hydroperoxide formation were proportional to the square root of the concentration of initiator and to the concentration of substrate. When U-74500A was added as the inhibitor, a transient decrease in the hydroperoxide production was observed, and during the same time period the compound was degraded in a first-order manner. Thus, in the homogeneous system, the inhibitor acts by scavenging lipid radicals, and its reactivity is about 30 times greater than that of linoleic acid. In Model 3, rat liver microsomes were treated with ferrous ammonium sulfate. This initiates an iron-mediated lipid peroxidation that is empirically described by the measure of malonyldialdehyde (MDA) that is formed. U-74006F was effective in all of these models. However, it was most effective when it was in the ordered environment of the lipid vesicle (Model 2) or the microsome (Model 3). U-74500A is actually a better 8 C. L. Bryan, R. A. Lawrence, E. D. Hall, and S. G. Jenkinson, FASEB J. 4, A630 (1990). 9 G. H. I. Wolfgang, R. A. Jolly, and T. W. Petry, Free Radical Biol. Med. 10, 403 (1991). 10j. M. Braughler and J. F. Pregenzer, Free Radical Biol. Med. 7, 125 (1989). ~! K. L. Linseman, B. S. Lutzke, J. M. McCall, and D. E. Epps, Toxicologist 13, 337 (1993).
552
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[56]
antioxidant than U-74006F. It has a lower oxidation potential, and this makes it a superior in vitro lipid peroxidation inhibitor. In addition to scavenging o f lipid peroxyl radicals, U-74006F also reacts with reactive oxygen species such as hydroxyl radicals generated during in vitro F e n t o n reactions (i.e., Fe 2+ + H202----> Fe 3+ + O H - + .OH). 12 In vivo studies employing the salicylate trapping method for measurement of hydroxyl radical have demonstrated that U-74006F administration decreases brain hydroxyl radical levels in models of concussive head injury in mice, 13 as well as global cerebral ischemia-reperfusion injury in gerbils. 12 This may be due to either direct scavenging of hydroxyl radical or a decrease in its injury-induced formation. A n t i o x i d a n t E f f e c t s in Whole Cells
The 21-aminosteroids have also been shown to inhibit lipid peroxidation in whole cells. F o r example, U-74500A inhibits copper-induced red cell lipid peroxidation. The compound is effective at concentrations as low as 1/xM. At 1/zM, it significantly reduces copper-induced and H202induced erythrocyte lipid peroxidation by 76.5 and 27.6%, respectively. The inhibition of erythrocyte lipid peroxidation was accompanied by an inhibition of hemolysis.14 U-74006F (5 ~M) has been shown to protect murine neocortical cell cultures that were exposed to 5 0 / x M ferric iron and 50/xM ferrous iron for 24 hr from neuronal degeneration. 15 The compound has also been reported to protect cultured murine spinal neurons from damage by 200/.tM ferrous iron. 16 U-74006F is also effective in an in vitro model for predicting the ability of a compound to prevent cell damage during periods of energy failure. Iodoacetic acid (IAA) was administered to cultured human astroglial cells (UC-11MG) at a concentration of 50 ~ M for 4 hr. This agent shuts down glycolysis and leads to subsequent irreversible breakdown of cellular membranes, and ultimately to cell death. During the first hours after addition, IAA rapidly depleted cellular levels of ATP and decreased active uptake of tritiated aminoisobutyric acid. Subsequent irreversible cellular injuries were characterized by the release of large amounts of free arachidonic acid into the extracellular medium, massive calcium influx, and leakage 12j. s. Althaus, C. W. Williams, P. K. Andrus, P. F. von Voigtlander, and E. D. Hall, Soc. Neurosci. Abstr. 17, 164 (1991). 13E. D. Hall, P. K. Andrus, and P. A. Yonkers, J. Neurochem. 60, 588 (1993). 14A. C. Fernandes, P. M. Filipe, and C. F. Manso, Eur. J. Pharmacol. 220, 211 (1992). 15H. Monyer, D. M. Hartley, and D. W. Choi, Neuron 5, 121 (1990). 16E. D. Hall, J. M. Braughler, P. A. Yonkers, S. L. Smith, K. L. Linseman, E. D. Means, H. M. Scherch, P. F. von Voigtlander, R. A. Lahti, and E. J. Jacobsen, J. Pharmacol. Exp. Ther. 258, 688 (1991).
[56]
LAzARoIos
553
of cytoplasmic contents (51Cr release). The appearance of 15-hydroxyeicosatetraenoic acid in membrane phospholipids and loss of cellular thiol groups indicated that the cell constituents were being assaulted by oxidative species. These manifestations oflAA-induced cell damage were inhibited by U-74006F. The IAA-induced release of tritiated arachidonic acid was inhibited with an IC50 value of 6/.~M. U-74006F was effective even when it was administered up to 1 hr after the onset of the metabolic insult.~7 In other work, U-74006F was also shown to decrease the release of arachidonic acid from cultured AtT-20 pituitary tumor cells triggered by exposure to either IAA or ferrous iron.~8 Physicochemical Effects on Membranes The 21-aminosteroids U-74006F and U-74500A also have potent stabilizing effects on cell membranes. As noted above, the compounds have a high affinity for the lipid bilayer because of their lipophilicity. Reflecting its membrane interaction, U-74006F has been shown to exert physicochemical effects on endothelial cell membranes. Bovine brain microvessel endothelial cells (BMECs) were labeled with diphenylhexatriene (DPH) fluorophores. Interactions with cell membranes were characterized by fluorescence anisotropy and fluorescence lifetimes. U-74500A and U-74006F preferentially altered the fluorescence anisotropy and lifetime parameters of the fluorescent DPH probe that distributed into the membranes throughout the BMECs. Little or no effect of the compounds were observed on the fluorescence parameters of the probe trimethylammonium-phenyl-DPH) that localized on the surface of BMEC plasma membranes. In contrast, cholesterol, used as a positive control, substantially altered the fluorescence parameters of BMECs labeled with either surface or membrane core probes. These experiments suggest that the 21-aminosteroids induce changes in the molecular packing order in membrane hydrophobic domains throughout the BMEC.19 Other research has also demonstrated physicochemical effects of the 21-aminosteroids on membranes. U-74006F and vitamin E were studied in bilayer lipid membranes with time-resolved fluorescence depolarization and angle-resolved fluorescence depolarization techniques, and by electron paramagnetic resonance utilizing probe molecules. 2° Lipid peroxida17 F. Sun, B. M. Taylor, and W. E. Fleming, FASEB J. 7, A658 (1992). 18 j. M. Braughler, R. L. Chase, G. L. Neff, P. A. Yonkers, J. S. Day, E. D. Hall, V. H. Sethy, and R. A. Lahti, J. Pharmacol. Exp. Ther. 44, 423 (1988). 19 K. L. Audus, F. L. Guillot, and J. M. Braughler, Free RadicaIBiol. Med. 11, 361 (1991). 2o G. Van Ginkel, J. M. Muller, F. Siemsen, A. A. van't Veld, L. J. Korstanje, M. A. M. van Zandvoort, M. L. Wratten, and A. Sevanian, J. Chem. Soc., Faraday Trans. 88, 1901 (1992).
554
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[56]
H~OIH /0 H
+1
/N~N CH2 V
O~'~"CH3
N -Ni~ Tk,Y
I~ N
©
FIG. 2. Membrane orientation of the 21-aminosteroid U-74006F.
tion products (oxidized fatty acids) strongly disorder the unsaturated lipid membranes they inhabit, but they do not affect the lipid dynamics. This is compatible with a model where the lipid hydroperoxy or hydroxy moieties reside closer to the polar head group region of the membrane lipids. U-74006F also has a disordering effect in the lipid systems, although the effects on dynamics vary depending on the surrounding lipids. Generally, U-74006F decreased dynamics (increased head group order). The decrease for U-74006F is consistent with the observations from additional work showing that the 21-aminosteroids are incorporated into the lipid bilayer where they occupy strictly defined positions and orientations. 2l As shown in Fig. 2, we hypothesize that U-74006F resides in the cell membrane, and that the piperazine nitrogen, which is largely protonated (i.e., positively charged) at physiologic pH, should orient with the head groups of the membrane bilayer by ionic interaction to the negatively charged phosphate-containing head groups. The steroid moiety, on the other hand, should localize within the hydrophobic core of the membrane. The pyrimidine amine of the molecule should help compress membrane phospholipid head groups. Indeed, head group viscosity in a lipid monolayer increases significantly with as little as 1.0 mol% (relative to lipid) of U-74006F (F. Kezdy et al., personal communication). 2[ j. S. Hinzmann, R. L. McKenna, T. S. Pierson, F. Han, F. Kezdy, and D. E. Epps, Chem. Phys. Lipids 62, 123 (1992).
[57]
ANTIOXIDANT
PROPERTIES
OF AMINOSALICYLATES
555
In addition to the chemical antioxidant properties of U-74006F (and U-74500A) described above, this "membrane stabilizing" action may help to inhibit the propagation of lipid peroxidation by restricting the movement of lipid peroxyl and alkoxyl radicals within the membrane. Thus, the 21aminosteroids block oxygen radical-induced lipid peroxidation apparently via a combination of chemical antioxidant (i.e., radical scavenging) and membrane stabilizing effects.
[57] A n t i o x i d a n t P r o p e r t i e s o f A m i n o s a l i c y l a t e s
By ALLEN M. MXLES and MATTHEW B. GRISHAM Introduction Ulcerative colitis (UC) is a recurrent inflammation of the colon and rectum that is characterized by rectal bleeding, diarrhea, fever, pain, anorexia, and weight loss. Active episodes of the disease are characterized by the extravasation and infiltration of large numbers of phagocytic leukocytes (neutrophils, monocytes, and macrophages) into the colonic mucosa. 1 This enhanced inflammatory infiltrate is accompanied by extensive mucosal injury including edema, crypt abscesses, loss of goblet cells, decreased production of mucus, erosions, and mucosal ulcerations. The apparent association between leukocyte infiltration and mucosal injury has led to the proposition that phagocytic leukocytes may mediate much of the pathophysiology associated with active disease. 2 There is a growing body of experimental data to suggest that the chronically inflamed intestine and/or colon may be subjected to considerable oxidative stress. 3'4 The most probable source of these oxidants are the phagocytic leukocytes since these cells are known to be present in large numbers in the inflamed
I R. H. Riddell, in "Inflammatory Bowel Disease" (J. Kirsner and R. G. Shorter, eds.), pp. 329-350. Lea & Febiger, Philadelphia, 1988. 2 M. B. Grisham and D. N. Granger, in "Current Topics in Gastroenterology" (R. MacDermott and W. Stenson, eds.), pp. 225-239. Elsevier, Amsterdam, 1992. 3 N. J. Simmonds, R. E. Allen, T. R. J. Stevens, R. N. M. Van Someren, D. R. Blake, and D. S. Rampton, Gastroenterology 103, 186 (1992). 4 A. Keshavarzian, S. Sedghi, J. Kanofsky, T. List, C. Robinson, C. Ibrahim, and D. Winship, Gastroenterology 103, 177 (1992).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[57]
ANTIOXIDANT
PROPERTIES
OF AMINOSALICYLATES
555
In addition to the chemical antioxidant properties of U-74006F (and U-74500A) described above, this "membrane stabilizing" action may help to inhibit the propagation of lipid peroxidation by restricting the movement of lipid peroxyl and alkoxyl radicals within the membrane. Thus, the 21aminosteroids block oxygen radical-induced lipid peroxidation apparently via a combination of chemical antioxidant (i.e., radical scavenging) and membrane stabilizing effects.
[57] A n t i o x i d a n t P r o p e r t i e s o f A m i n o s a l i c y l a t e s
By ALLEN M. MXLES and MATTHEW B. GRISHAM Introduction Ulcerative colitis (UC) is a recurrent inflammation of the colon and rectum that is characterized by rectal bleeding, diarrhea, fever, pain, anorexia, and weight loss. Active episodes of the disease are characterized by the extravasation and infiltration of large numbers of phagocytic leukocytes (neutrophils, monocytes, and macrophages) into the colonic mucosa. 1 This enhanced inflammatory infiltrate is accompanied by extensive mucosal injury including edema, crypt abscesses, loss of goblet cells, decreased production of mucus, erosions, and mucosal ulcerations. The apparent association between leukocyte infiltration and mucosal injury has led to the proposition that phagocytic leukocytes may mediate much of the pathophysiology associated with active disease. 2 There is a growing body of experimental data to suggest that the chronically inflamed intestine and/or colon may be subjected to considerable oxidative stress. 3'4 The most probable source of these oxidants are the phagocytic leukocytes since these cells are known to be present in large numbers in the inflamed
I R. H. Riddell, in "Inflammatory Bowel Disease" (J. Kirsner and R. G. Shorter, eds.), pp. 329-350. Lea & Febiger, Philadelphia, 1988. 2 M. B. Grisham and D. N. Granger, in "Current Topics in Gastroenterology" (R. MacDermott and W. Stenson, eds.), pp. 225-239. Elsevier, Amsterdam, 1992. 3 N. J. Simmonds, R. E. Allen, T. R. J. Stevens, R. N. M. Van Someren, D. R. Blake, and D. S. Rampton, Gastroenterology 103, 186 (1992). 4 A. Keshavarzian, S. Sedghi, J. Kanofsky, T. List, C. Robinson, C. Ibrahim, and D. Winship, Gastroenterology 103, 177 (1992).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
556
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[57]
-i:~
\""
H202 + C l -
Fe+3
• HOCI
:
0=+ OH-+OH"
(.. o-,/ o, NADPH oxidase
FIG. 1. Release of reactive oxygen metabolites by activated human neutrophils. Interaction of certain proinflammatory mediators (e.g., leukotriene B4, platelet-activating factor, or bacterial products) with specific receptors on the neutrophilic plasma membrane activates the membrane-associated enzyme NADPH oxidase. This oxidase reduces molecular oxygen (02) by one electron to yield the superoxide anion radical (027) which rapidly and spontaneously dismutates to yield hydrogen peroxide (H202). Superoxide may interact with H202 in the presence of trace amounts of iron (Fe 3+) to yield the highly reactive hydroxyl radical (.OH). Activated neutrophils also secrete the hemoprotein myeloperoxidase (MPO) into the extracellular space where it catalyzes the H202-dependent, two-electron oxidation of chloride (C1-) to yield the potent oxidizing and chlorinating agent hypochlorous acid (HOCI).
mucosa and are known to produce large amounts of reactive oxygen species in response to certain inflammatory stimuli (Fig. 1). 5,6 Oral administration of the drug sulfasalazine (SAZ) has proved effective in attenuating the inflammation and mucosal injury associated with 5 T. Yamada and M. B. Grisham, in "Inflammatory Bowel Disease: From Bench to Bedside" (F. Shanahan and S. Targan, eds.), pp. 133-150. Williams & Wilkens, Baltimore, 1994. 6 S. J. Klebanoff, in "Inflammatory--Basic Principles and Clinical Correlates" (J. I. Gallen, I. M. Goldstein, and R. Synderman, eds.), pp. 391-444. Raven, New York, 1985.
[57]
ANTIOXIDANT PROPERTIES OF AMINOSALICYLATES
557
ulcerative colitis. 7-9 Pharmacokinetic studies have demonstrated that SAZ passes unmodified through the upper gastrointestinal (GI) tract until it reaches the colon where it is metabolized by endogenous bacteria to yield sulfapyridine (SP) and 5-aminosalicylic acid (5-ASA). Several clinical studies have shown that 5-ASA is the pharmacologically active moiety of SAZ; however, its mechanism of action remains only speculative. 7-9 It has been suggested that 5-ASA may exert its anti-inflammatory activity in vivo by inhibiting prostaglandin synthase and/or lipoxygenase activities as well as by interfering with antibody synthesis. ~0-~4Although these proposed mechanisms of action of 5-ASA can be readily demonstrated in vitro, the concentrations required for significant inhibition are at least 10to 50-fold higher than those achieved in the colonic mucosal interstitium of a normal bowel perfused with clinically relevant concentrations of 5-ASA.~5 Our laboratory, as well as others, have demonstrated that 5-ASA is a very potent antioxidant and free radical scavenger at concentrations similar to those found within the colonic mucosal interstitium. 16-z4 Thus, we have proposed that the antioxidant and metal binding properties of 5-ASA are important mechanisms by which 5-ASA exerts its anti-inflammatory activity in rio0. 23'24 In this chapter, we have assembled a variety 7 A. H. Azad-Khan, J. Piris, and S. C. Truelove, Lancet 2, 892 (1977). 8 p. A. M. van Hees, J. H. Bakker, and J. H. M. van Tongeren, Gut 21, 632 (1980). 9 U. Klotz, K. Maier, C. Fischer, and K. Heinkel, N. Engl. J. Med. 303, 1499 (1980). 10 j. R. S. Hoult and P. K. Moore, Br. J. Pharmacol. 68, 719 (1980). 11 C. J. Hawley and S. C. Truelove, Gut 24, 213 (1983). 12 W. F. Stenson and E. Lobos, J. Clin. lnoest. 69, 494 (1982). 13 H. Allgayer and W. F. Stenson, Immunopharmacology 15, 39 (1988). t4 R. P. MacDermott, S. R. Schloemann, M. J. Bertovich, G. S. Nash, M. Peters, and W. F. Syenson, Gastroenterology 96, 442 (1989). 15 M. B. Grisham and D. N. Granger, Dig. Dis. Sci. 34, 573 (1989). 16 G. Carlin, R. Djursater, G. Smedegard, and B. Gerdin, Agents Actions 16, 377 (1985). 17 W. H. Betts, M. W. Whitehouse, L. G. Cleland, and B. Vernon-Roberts, J. Free Radical Biol. Med. 1, 273 0985). 18 O. I. Aruoma, M. Wasil, B. Halliwell, B. M. Hoey, and J. Butler, Biochem. Pharmacol. 36, 3739 (1987). 19 p. A. Carven, J. Pfanstiel, R. Saito, and F. R. DeRubertis, Gastroenterology 92, 1998 (1987). 2o I. Ahnfelt-Ronne and O. H. Nielson, Agents Actions 21, 1991 (1987). 21 B. J. Dull, K. Salata, A. Van Langenhove, and P. Goldman, Biochem. Pharmacol. 36, 2467 (1987). 22 I. Ahnfelt-Ronne, O. H. Nielson, A. Christensen, E. Langholz, and P. Riis, Gastroenterology 98, 1162 (1990). 23 M. B. Grisham, in "Inflammatory Bowel Disease: Current Status and Future Approach" (R. P. MacDermott, ed.), pp. 261-266. Elsevier, Amsterdam, 1988. 24 T. Yamada, C. Volkmer, and M. B. Grisham, Can. J. Gastroenterol. 4(7), 295 (1990).
558
ANTIOXlDANT CHARACTERIZATION AND ASSAY
[57]
of techniques that may be used to assess the antioxidant properties of several different anti-inflammatory drugs including the aminosalicylates. Antioxidant Properties of Aminosalicylates
Superoxide The most popular method for measuring the superoxide (02 :) scavenging properties of various compounds is the inhibition of O2--mediated reduction of cytochrome. 25In most cases this assay is entirely appropriate; however, it should be noted that some low molecular weight, easily oxidizable compounds will reduce hemoproteins like cytochrome c in a superoxide-independent mechanism, thereby introducing an artifactually high background rate of reduction. One such compound is 5-ASA (A. M. Miles and M. Grisham, unpublished observations, 1993). To assess the superoxide dismutase (SOD)-like activity of aminosalicylates and eliminate potentially interfering artifacts associated with the cytochrome c assay we have used a modification of the spectrophotometric assay of superoxide dismutase originally described by Marklund. 26 The assay directly measures the rate of spontaneous dismutation of potassium superoxide at 250 nm. 26 At alkaline pH and low 02: concentrations, this radical is stable enough to be studied using a laboratory spectrophotometer. The rate of dismutation of 02: can be monitored using the broad absorbance maximum (molar extinction coefficient = 2000) of 02: at 250 nm. 26 Using these conditions it is possible to observe the 02: scavenging properties of a variety of compounds including SOD and 5-ASA directly. Prior to the experiments all glassware is cleaned with potassium dichromic acid and rinsed extensively with deionized water (i.e., double-distilled and deionized by reversed osmosis). Three milliliters of filtered and degassed 50 mM 2-amino-2-methyl-l-propanol (AMP-HC1; pH 9.5; 25°) buffer containing 0.2 mM desferrioxamine and 20/zg/ml of catalase is added to a 1-cm quartz cuvette positioned in the sample holder of a recording spectrophotometer. Pulverized potassium superoxide (100 mg; KO2), prepared immediately before use, is then dissolved in 25 ml of icecold 50 mM NaOH containing 0.5 mM desferrioxamine. The flask is rapidly swirled and at exactly 10 sec after adding the KO2 to the flask, an aliquot (10/xl) of the KO2 solution is transferred to the tip of a Teflon-coated 25 I. Fridovich, in "Handbook of Methods for Oxygen Radical Research" (R. A. Greenwald, ed.), pp. 121-122. CRC Press, Boca Raton, Florida, 1985. 26 S. S. Marklund, in "Handbook of Methods for Oxygen Radical Research" (R. A. Greenwald, ed.), pp. 249-255. CRC Press, Boca Raton, Florida, 1985.
[57]
ANTIOXIDANT
PROPERTIES
559
OF AMINOSALICYLATES
spatula. The strip chart recorder is then started (300 mm/min) and the tip of the spatula containing the KO 2 solution is immediately inserted into the cuvette containing the AMP-HCI buffer and the contents mixed rapidly. The decrease in absorbance at 250 nm is then continuously recorded. We have found that 5-ASA directly interacts with 02 ~, causing the rapid decomposition of the radical (Fig. 2). The disappearance of O2" may occur by two possible pathways. One pathway would require that O2" is reduced by one electron by 5-ASA to give H202 . The other pathway could involve the one-electron oxidation of 02- to yield molecular oxygen. Because 5-ASA would acquire or lose one electron by either pathway, it would become, by definition, a free radical itself. We are currently investigating the mechanism of 02- decomposition by 5-ASA.
Hydrogen Peroxide There are a variety of methods available to measure the hydrogen decomposing activity of various substances. Virtually all these methods are based on the ability of a hemoprotein peroxidase (usually horseradish peroxidase) to catalyze the HzOz-dependent oxidation of an electron-donating detector molecule. In many cases the detector substrate is a leuko dye which, when oxidized, produces a chromogenic p e r o x i d e (H202)
~_o "~ C
°
1O0 80
\
u cO
oe
•
60
lo
•
27
20
& + 5-ASA
0
5
10
15 Time
20
25
30
(seconds)
FIG. 2, Decomposition of superoxide anion radical by 5-aminosalicylic acid (5-ASA) and SOD. Decomposition of superoxide (potassium superoxide) was assessed by measuring the decrease in absorbance at 250 nm (hmax for Oz ~) at pH 9.0 in the absence or presence of SOD (20 t~g/ml) or 5-ASA (0.25 mM).
560
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[57]
or fluorometric product that can easily be quantified. Unfortunately, we have found that certain aminosalicylates (including 5-ASA) are excellent substrates for peroxidase-catalyzed reactions and thus will compete with the detector compound for oxidation by the peroxidase (see section on Hemoprotein-Mediated Oxidative Reactions). These types of interfering reactions render this assay unsuitable for measuring the H202-decomposing activity of 5-ASA. To circumvent this problem we have used a modification of a method originally described by Heath and Tappe127 in which H 2 0 2 is detected using glutathione peroxidase-catalyzed oxidation of reduced glutathione (GSH). It is well known that the selenoprotein glutathione peroxidase has H20 2 +
2GSH--> H20 + GSSG
an absolute specificity for GSH as its electron donating substrate. 2a We have found that a variety of antioxidants will not interfere with the detection of H202 using this assay, unless of course it decomposes the oxidant! Briefly, the assay involves an initial incubation period in which H202 (50-200 ~M) is allowed to interact with the compound in question (e.g., 5-ASA) in l0 mM potassium phosphate buffer (pH 7.4) for 15-60 min at 37°. Following the incubation period, 1 mM GSH and 1 unit/ml of GSH peroxidase (bovine erythrocyte; Sigma, St. Louis, MO) is added to the reaction volume and incubated for an additional 5 min at 37 °. The tubes are then diluted 10-fold using 0.2 M Tris-HCl (pH 8.5), 1 mM 5,5'-dithiobis(2nitrobenzoic acid) (DTNB) added, the tubes incubated for 5 rain at 37 °, and the absorbances determined at 412 nm. DTNB is used to detect GSH via its interaction with GSH to yield 5-thio-2-nitrobenzoic acid (EM 13,600 at 412 nm). Because 2 mol of GSH are oxidized for every mole of H202 present, one may calculate the concentration of H202remaining following its interaction with potential H2OE-Scavenging compounds. Therefore, any compound capable of decomposing H202 will attenuate the oxidation of GSH. We have found that 5-ASA is unable to decompose significant amounts of H202 even when the ratio of 5-ASA to H 2 0 2 is 20 : 1.29
Hydroxyl Radicals It has been demonstrated that activated polymorphonuclear lymphocytes (PMNs) may produce relatively large amounts of hydroxyl radicals (. OH) in the presence of a metal catalyst such as iron (Fe 3÷) (Fig. 1). The source of the Fe is not known; however, it has been suggested to be 27 R. L. Heath and A. L. Tappel, Anal. Biochem. 76, 184 (1976). 28 A. Wendel, this series, Vol. 77, p. 325. 29 M. B. Grisham, Biochem. Pharm. 39, 2060 (1990).
[57]
ANTIOXIDANT PROPERTIES OF AMINOSALICYLATES
561
fenitin since it is well known that 0 2" will reductively release Fe from the protein core of ferritin. 3° More recent evidence suggests that activated PMNs may produce small amounts of • OH in the absence of a metal catalyst via the interaction between 02- and the hypochlorous acid (HOCI) generated from the myeloperoxidase-catalyzed oxidation of chloride (C1-). 31 In either case it is apparent that activated PMNs have the potential to produce significant quantities of the very reactive • OH. The hydroxyl radical is an extremely reactive species, reacting immediately with virtually all known biomolecules at diffusion-limited rates of reactions (-107-1010 M -1 s e c - l ) . Therefore, it is very short-lived and will react at the site where it is formed, that is, in a site-specific fashion. This reactivity dictates that the radical will be injurious to cells only if the metal catalyst is localized on biomolecules essential for survival. For example, DNA, membrane phospholipids, or proteins required for essential metabolic processes could be potential targets for such a metabolism. If, on the other hand, • OH is generated on biomolecules present in high concentrations (e.g., albumin, glucose) the physiological consequences of such degradative reactions may be minimal. 32 The iron-catalyzed, site-specific generation o f . OH may be measured using ferrous sulfate-mediated degradation of deoxyribose to yield thiobarbituric acid-reactive substances (TBARS) as described by Gutteridge. 33 It has been demonstrated that Fe 2÷, in the absence of an exogenous chelator, will associate with deoxyribose (DOR) where it autoxidizes to generate • OH on the surface of the carbohydrate (i.e., in a site-specific manner) 32,33Only those compounds capable of removing Fe 2÷ from deoxyDOR + Fe 2+ ~ DOR-Fe 2+ DOR-Fe 2+ + 02--> DOR-Fe 3+ + 02" 20 2v + 2H + --~ 2H202 + 0 2 DOR-Fe 2+ + H202 --* DOR-Fe 3+ + OH- + • OH ribose and rendering it poorly redox active or those compounds with H202 decomposing activity will be effective inhibitors of this system. 29'32'33 To generate • OH site specifically, each reaction volume (0.5 ml) contains the following compounds which are added in the following order: 2 mM deoxyribose, 0.1 mM ferrous sulfate, various concentrations of drug, and 10 mM potassium phosphate buffer (pH 7.4). Following a 30-min 30 p. Biemond, H. G. Van Eijk, A. J. G. Swaak, and J. F. Koster, J. Clin. Invest. 73, 1576 (1984). 31 C. L. Ramos, S. Pou, B. E. Britigan, M. S. Cohen, and G. M. Rosen, J. Biol. Chem. 267, 8307 (1992). 32 B. Halliwell and J. M. C. Gutteridge, Arch. Biochem. Biophys. 246, 501 0986). 33 j. M. C. Gutteridge, Biochem. J. 243, 709 (1987).
562
[57]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
incubation period at 37°, reactions are terminated by the addition of catalase (20/xg/ml). The TBARS are then quantified by the sequential addition of 0.5 ml trichloroacetic acid (2.8%) and 0.5 ml thiobarbituric acid (TBA) (1% in 0.05 N NaOH), and the tubes are heated at 100° for 15 min in a boiling water bath. The tubes are then cooled and the absorbance determined at 532 nm. 33Each absorbance value is corrected for nonspecific development that may occur in the absence of iron. Using this method to generate • OH we have found that 5-ASA was much more effective at inhibiting the formation of TBARS than was N-acetyl-ASA (NASA), SAZ, or SP (Fig. 3). Because we had already determined that 5-ASA did not decompose H202, we concluded that the mechanism by which 5-ASA inhibited this reaction was due to its ability to bind Fe and render it poorly redox active. Indeed, we subsequently demonstrated that 5-ASA but not NASA, SAZ, or SP chelates Fe 2+ or Fe3+. 29 To directly assess the ability of 5-ASA to chelate Fe 2+, ferrous sulfate (0.2 mM) was added to a solution of 5-ASA (1 mM) in 0.1 M NaCI (pH 7.4), and the absorbance spectrum was determined. The interaction
125[ 100 , ~ r
-m SP
•~
I'--.. 75
m
50
-&
SAZ
5 -ASA
!
25 0
1 O0
200
300
400
500
Drug Concentrotion (pJvI) FIG. 3. Effects of 5-aminosalicylic acid (5-ASA), N-acetyl-5-ASA (NASA), sulfasalazine (SAZ), and sulfapyridine (SP) on ferrous sulfate-mediated degradation of deoxyribose to yield thiobarbituric acid-reactive substances (TBARS). Each data point is the mean from at least four determinations and varied by less than ---5%. A mean absorbance value of 0.218 was achieved in the absence of drug and was designated as 100%. (Data derived from Grisham. 29)
[57]
ANTIOXIDANT PROPERTIES OF AMINOSALICYLATES
563
b e t w e e n Fe 2+ and 5-ASA produced a purple chromophore indicative of a charge transfer complex (Fig. 4). 29
Hydroxyl radicals may also be generated in free solution by exposing a detector molecule such as deoxyribose to a ferrous iron (Fe z+) chelate in the presence of H202: F e 2+ + H 2 0 2 ~
F e 3+ + O H -
+ • OH
In these experiments Fe 2+ is chelated to diethylenetriaminepentaacetic acid (DTPA), thereby preventing the metal from associating with deoxyribose. Thus, any" OH generated from the interaction between Fe2+-DTPA and HzOzwill have equal access to all components of the reaction volume including deoxyribose. Hydroxyl radical production is measured by quantifying the oxidative degradation of deoxyribose to yield TBARS as described by Gutteridge. 33
0,20-
0,15-
U C
0.1-
0.05-
b 0
400
s~o
e~o
7~o
Wavelength (nm)
FIG. 4. Absorbance spectrum of the Fe2+-5-ASA chelate in the absence or presence of deoxyribose. Ferrous sulfate (0.2 mM) was added to a solution containing 1 mM 5-ASA and 0.1 M NaC1, pH 7.4, in the absence or presence of 4 mM deoxyribose and the absorbance spectrum determined. (a) Fe 2÷, (b) 5-ASA, (c) Fe 2÷ plus 5-ASA, and (d) Fe 2÷ plus 5-ASA in the presence of deoxyribose. (Data derived from Grisham. 29)
564
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[57]
1001 g
60
~ ' ~ ~
NASA SP SAZ
,o
5-ASA
20 I 0
, 100
, 200
i 300
Drug C o n c e n t r a t i o n
i 400
i 500
OzM)
FIG. 5. Effects of 5-aminosalicylic acid (5-ASA), N-acetyl-5-ASA (NASA), sulfasalazine (SAZ), and sulfapyridine (SP) on the Fe2+-DPTA-mediated degradation of deoxyribose to yield thiobarbituric acid-reactive substances (TBARS). Each data point is the mean from four determinations and varied by less than ---5%. A mean absorbance value of 0.426 ~vas achieved in the absence of drug and was designated as 100%. (Data derived from Grisham. 29)
The FeZ+-DTPA complex is prepared by the method of Cohen 34 in which 1 mM ferrous sulfate is added to a solution containing 2 mM DTPA and 50 mM potassium phosphate buffer. 34 Each reaction (volume 0.5 ml) contains the following compounds which are added in the following order: 2 mM deoxyribose, 0.1 mM FeZ+-DTPA (0.1 mM Fe2+-0.2 mM DTPA), various concentrations of the aminosalicylate, and 10 mM potassium phosphate buffer (pH 7.4). Addition of 0.2 m M H202 is used to initiate the reaction. Following a 10-min incubation period at 37° reactions are terminated by the addition of 20/~g/ml catalase. The TBARS are then quantified as described above using the sequential addition of 0.5 ml trichloroacetic acid (2.8%) and 0.5 ml TBA (1% in 0.05 N NaOH). We have found that SAZ as well as 5-ASA, NASA, and SP are equally effective at inhibiting the • OH-mediated degradation of deoxyribose when • OH is generated in free solution (Fig. 5). 29 This is not at all surprising since virtually all phenolic compounds have been shown react with -OH at diffusion-limited rates. The data also illustrate two important points. First, the mechanism of action of 5-ASA cannot be due to its ability to G. Cohen, in "Handbook of Methods for Oxygen Radical Research" (R. A. Greenwald, ed.), pp. 55-64. CRC Press, Boca Raton, Florida, 1985.
[57]
ANTIOXIDANT PROPERTIES OF AMINOSALICYLATES
565
scavenge selectively the. OH since the pharmacology inactive metabolites NASA and SP are equally effective in this assay. Second, the data dispel the erroneous suggestion that certain compounds may act as specific "hydroxyl radical scavengers" since virtually all compounds react rapidly with • OH.
Peroxyl Radicals Hydroxyl radicals interact with certain carbohydrates, proteins, nucleotide bases, and lipids to produce peroxyl radicals (ROO .) as intermediates.35 Peroxyl radicals are slightly less reactive than. OH and thus possess "extended" half-lives of seconds instead of nanoseconds. These radicals would then be expected to react at site distant from those of. OH generation, thereby promoting the toxicity of" OH. Indeed, peroxyl radical intermediates have been shown to damage biomolecules. The best characterized example of peroxyl radical-mediated reactions is the peroxidation of polyunsaturated fatty acids (PUFA) initiated by • OH: •OH+LH--~HOH+L. L" + 02---> L 0 0 " L 0 0 " + LH --~ L 0 0 t t + L"
where • OH, LH, L., L O 0 . , and LOOH represent the hydroxyl radical, polyunsaturated lipid, lipid alkyl radical, lipid hydroperoxyl radical, and lipid hydroperoxide, respectively. Peroxyl radicals will also oxidize proteins, carbohydrates, and sulfhydryl components and hemolyze erythrocytes.3~ Because most free radical generators require the addition of potentially interfering cofactors and transition metals such as iron (or copper) to produce the free radicals we chose to use the thermolabile, peroxyl radical generator 2,2'-azobis(2-amidinopropane) dihydrochloride (AAPH). AAPH is a free radical initiator that decomposes unimolecularly without requirement for enzymes or biotransformation to yield nitrogen and carbon radicals. 37,3s The carbon-centered radicals react rapidly with molecular oxygen (02) to yield peroxyl radicals. The rate of decomposition of AAPtt ~5 R. B. Wilson, in "Oxidative Stress" (H. Sies, ed.), pp. 41-72. Academic Press, London and New York, 1985. 36 I. S. Sandu, K. Ware, and M. B. Grisham, Free Radical Res. Commun. 16, 111 (1992). ~7 E, Niki, this series, Vol. 186, p. 100. 38 E. Niki, E. Komuro, M. Takahashi, S. Urano, E. Ito, and K. Terao, J. Biol. Chem. 263, 19809 (1988).
566
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[57]
is determined primarily by temperature, and the rate of generation of radicals is virtually constant for the first few hours. The reaction sequence is as follows: A - - N ~ N - - A ~ A. + N 2 + A" A- + O 2 ~ A O O " AOO. + L H ~ AOOH + L . L . + 02 ~ LOO" LOO. + L H ~ LOOH + L . where A - - N ~ - - N - - A , A., AOO., and AOOH represent AAPH, alkyl radical, peroxyl radical, and hydroperoxide, respectively. Because peroxyl radical-mediated oxidative reactions are not initiated by Fe-catalyzed hydroxyl radical formation, an inhibitory effect by an antioxidant is due solely to its ability to scavenge the peroxyl radical. We have found that peroxyl radical-mediated lipid peroxidation represents a sensitive and simple method to assess the antioxidant activity of a variety of compounds such as 5-ASA. Lipid peroxidation is quantified by measuring the peroxyl radical-mediated formation of TBARS from a phospholipid substrate as described by Buege and Aust. 39 Folch fraction III phospholipid (brain extract; 80-85% phophatidyl serine) is obtained from Sigma. The phospholipid preparation is stored at - 2 0 ° as a 10% (w/v) solution in chloroform which contains 140 mM phospholipid phosphate. Aqueous solutions ofphospholipid (liposomes) are prepared by evaporating the chloroform from 100-/zl aliquots of the lipid preparation using oxygen-free nitrogen, adding 2 ml saline, and then sonicating the suspension for 10 sec under nitrogen. Briefly, AAPH (50 mM) is incubated in phosphate-buffered saline (PBS) containing phospholipid (brain extract; 0.7 mM phospholipid phosphate) for varying lengths of time at 37°. At 15 min intervals aliquots (0.5 ml) are removed from each tube and added to 1 ml of a solution containing 15% (w/v) trichloroacetic acid, 0.375% (w/v) thiobarbituric acid, and 0.25 N HC1. To prevent spurious lipid peroxidation during subsequent steps, 0.02% (w/v) butylated hydroxytoluene is added. Each mixture is then heated in a boiling water bath for 15 min; the tubes are allowed to cool and then are centrifuged at 8000 g for 5 rain to remove precipitate. The absorbance of each sample is then determined at 532 nm. We have found that 5-ASA but not SAZ, SP, NASA is capable of completely inhibiting peroxyl radical-mediated lipid peroxidation such that the concentration necessary for 5-ASA to inhibit peroxidation by 50% (ICs0) is equal to approximately 10/xM (Fig. 6). 24,36 39j. A. Buege and S. D. Aust, this series, Vol. 52, p. 303.
[57]
567
ANTIOXIDANT PROPERTIES OF AMINOSALICYLATES
^^ L _
tO .m
_,,._._.~v_
_-
SAZ
v
60
o
E
o
40
(/}
20
0
0
!
a
10
20
__
I,
30
40
5"0
Drug Concentrotion (/u,M) FIG. 6. Effects of 5-aminosalicylic acid (5-ASA), sulfasalazine (SAZ), N-acetyl-5-ASA (NASA), and sulfapyridine (SP) on peroxyl radical-mediated lipid peroxidation. Lipid (80% phosphatidylcholine) was incubated with the thermolabile free radical initiator 2,2'-azobis(2amidinopropane) dihydrochloride for 30 min at 37° in the absence of presence of SAZ or its metabolites. Lipid peroxidation was determined by measuring the formation of thiobarbituric acid-reactived substances (TBARS). (Data derived from Sandhu et al) 6)
We have also used the ability of aminosalicylates to inhibit peroxyl radical-mediated hemolysis of human erythrocytes to assess their antioxidant activity. Briefly, blood is collected from human volunteers by venipuncture. The erythrocytes are separated from leukocytes and platelets by dextran sedimentation, and the erythrocyte pellet is washed by centrifugation three times with PBS. The cells (2 × 109/ml) are incubated with 100 t~Ci sodium [SlCr] chromate for I hr at 37°. The cells are then washed three times with PBS and resuspended to 2 × l 0 9 cells/ml with Dulbecco's phosphate-buffered saline (DPBS). The erythrocytes are then diluted to 2 × 108/ml and incubated with varying concentrations of AAPH in DPBS for varying lengths of time at 37° with occasional swirling. At 15-rain intervals aliquots of cell suspension are removed and microcentrifuged, and the supernatant and pellet are counted for the presence of 51Cr to determine the extent of hemolysis. Interestingly, both 5-ASA and NASA are equally effective at inhibiting peroxyl radical-mediated hemolysis, whereas SAZ and SP are either much less effective or inactive (Fig. 7). 36 The data suggest that the mechanism by which peroxyl radicals mediated hemolysis may not be due solely to lipid peroxidation. In fact, we find very little evidence for lipid peroxidation in this system. 36
568
ANTIOXIDANT CHARACTERIZATION AND ASSAY
100~
[57]
_
90 t
70
'i
0
10
20
30 Drug
40
50
60
Concentration
70
80
90
100
(/~M)
FIG. 7. Effect of sulfasalazine and its metabolites on AAPH-induced hemolysis. Erythrocytes (2 × 108 cells/ml) were incubated with 50 mM AAPH in the absence or presence of anti-inflammatory drugs for 120 min at 37°. (@) 5-ASA, (A) 4-ASA, (11) N-acetyl-5-ASA, 01') sulfapyridine, and (T) sulfasalazine. Each data point represents the mean - SEM of duplicate determinations from at least three different donors. (Data derived from Sandu et al. 36)
Hemoprotein-Mediated Oxidative Reactions Work from our laboratory as well as that of others have demonstrated that hemoglobin (Hb) will interact with H2Oz or lipid hydroperoxides to yield a hemoprotein-associated oxidant that is capable of peroxidizing polyunsaturated fatty acids (PUFA), phospholipids, and cellular membranes. 4°-45 The interaction between methemoglobin (MetHb; Hb III) and H202 results in the two-electron oxidation of the hemoprotein to yield the radical form of ferrylhemoglobin (ferryltb; • Hb ~v) which is a potent oxidant. Inasmuch as active episodes of ulcerative colitis are accompanied 40 M. B. Grisham, J. Free Radical Biol. Med. 1, 227 (1985). 41 T. I. Yamada, C. Volkmer, and M. B. Grisham, Free Radical Biol. Med. 10, 41 (1991). 42 S. M. H. Sadrzadeh, E. Graf, S. S. Panter, P. E. Hallaway, and J. W. Eaton, J. Biol. Chem. 259, 14354 (1984). 43 j. Kanner and S. Harel, Arch. Biochem. Biophys. 237, 314 (1985). 44 M. B. Grisham and D. N. Granger, Dig. Dis. Sci. 33, Suppl., 6S (1988). 45 S. M. H. Sadrzadeh, D. K. Anderson, S. S. Panter, P. E. Hallaway, and J. W. Eaton, J. Clin. Invest. 79, 662 (1987).
[57]
569
ANTIOXIDANT PROPERTIES OF AMINOSALICYLATES I
,?
=~&'
'&"--
SAZ
4o _
0/ 0
,
,
,
l
20
40
60
80
•
1O0
Drug Concentration (b~M) FIG. 8. Effects of 5-ASA, NASA, SP, and SAZ on hemoglobin-catalyzed lipid peroxidation. Lipid was incubated in the absence or presence of varying amounts of SAZ or its metabolites with hemoglobin (0.02 mmol/liter) and HzO2 (0.1 mmol/liter) for 20 min at 37°. Lipid peroxidation was determined by measuring the formation of TBARS. (Data derived from Yamada et al. 41)
by subepithelial hemorrhage and hemolysis as well as mucosal lipid peroxidation, 2z'44 we have suggested that Hb-catalyzed reactions may play an important role in mediating and/or exacerbating inflammatory tissue injury. 41'44Indeed, the presence of Hb has been proposed to mediate and/or exacerbate oxidative injury in tissues such as the central nervous system, retina, kidney, and s y n o v i u m . 45-47 U s i n g Hb-catalyzed peroxidation of phospholipid as a model of oxidative degradation of membrane lipids, we assessed the ability of SAZ and its metabolites 5-ASA, NASA, and SP to inhibit the oxidative reaction. Hemoglobin-catalyzed lipid peroxidation is quantified using the method of Buege and Aust 39 in which TBARS are determined by measuring the absorbance at 532 nm. Human hemoglobin (predominantly HbIII; Sigma) is determined by the method of Tentori and Salvati 48in which cyanomethemoglobin is quantified. Briefly, each reaction (volume 0.5 ml) contains 20 mM potassium phosphate buffer (pH 7.4), 0.7 mM phospholipid phos46 S. Yoshino, D. R. Blake, S. Hewitt, C. Morris, and P. A. Bacon, Ann. Rheum. Dis. 44, 485 (1985). 47 G. J. Handelman and E. A. Dratz, Ado. Free Radical Biol. Med. 2, 1 (1986). 48 L. Tentori and A. M. Salvati, this series, Vol. 76, p. 707.
570
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[57]
120 10o
8O
o
~
A
~
a
SP
20 0
0
20
~
60
80
100
Drug Concentration (j~l)
FIG. 9. Effectsof 5-ASA, 4-ASA, NASA, and SP on myeloperoxidase(MPO) activity. HumanMPO activitywas assessed by measuringthe HzO2-dependentoxidationof 3,3',5,5'tetramethylbenzidinein the absence or presence of varyingconcentrationsof metabolites. (Data derived from von Ritter eta/. 49)
phate, Hb (12.5 tzM; 50/zM heme), varying concentrations of sulfasalazine or its metabolites, and 100/xM H202. Following a 15-min incubation at 37°, reactions are terminated by the addition of 1 ml of a solution containing 15% (w/v) trichloroacetic acid, 0.375% (v/v) thiobarbituric acid, and 0.25 N H C I . 39 To prevent spurious lipid peroxidation during subsequent steps, 0.02% butylated hydroxytoluene is added at this point. Each reaction mixture is then heated in a boiling water bath for 15 min; the tubes are removed and allowed to cool and then centrifuged at 8000 g for 5 rain to remove precipitate. The absorbance of each sample is then determined at 532 nm. 39 We found that 5-ASA and to a lesser extent NASA is effective at inhibiting Hb-catalyzed lipid peroxidation such that 5-ASA and NASA exhibited IC50 values of 50 and 125/~M, respectively (Fig. 8). 41 Neither SAZ nor SP were effective in this system. We have also found that 5-ASA is oxidized to a golden brown chromophore during its interaction with Hb and H202, suggesting that it is acting as an alternative substrate for the ferryl hemoprotein-mediated oxidation. 4'1 These are interesting data in view of a report demonstrating the presence of several different uncharacterized oxidation products of 5-ASA ob,tained from rectal dialyzates of patients with active ulcerative colitis being treated with 5-ASA. 22
[57]
ANTIOXIDANT PROPERTIES OF AMINOSALICYLATES
571
100"
C
o
5
O
Dapsone 0
0
• 50
N-AcelyI-S~ASA m 100pM
Drug Concenlration [..M]
FIG. 10. Hypochlorous acid (HOC1)-scavenging properties of sulfapyridine (SP), 5-ASA, 4-ASA, N-acetyl-5-ASA, and dapsone. All substances demonstrated a similar potential for scavenging HOC1. Each data point represents the mean for triplicate samples and did not vary by more than -+7%. (Data derived from von Ritter eta/. 49)
Another hemoprotein of interest in inflammation is myeloperoxidase (MPO), the enzyme released by activated PMNs at sites of inflammation. It is known that this enzyme combines with H202 to yield a potent hemoprotein-associated free radical (known as compound I) which oxidizes chloride by two electrons to yield the potent cytotoxic oxidant hypochlorous acid (HOCI) (Fig. 1). Using the H202-dependent oxidation of 3,3',5,5'tetramethylbenzidine at 655 nm to quantify MPO activity in the absence or presence of SAZ or its metabolites, we found that 5-ASA and its analog 4-ASA were very effective inhibitors of the catalytic activity of MPO (Fig. 9). 49 These aminosalicylates might act as alternative substrates for MPO in much the same way as they do with Hb-catalyzed oxidation reactions. In addition to showing inhibition of the catalytic activity of MPO, we, as well as others, have also demonstrated that 5-ASA is an exceedingly good scavenger of HOCI (Fig. 10). 18,49The concentration of HOC1 is determined by the method of Thomas e t al. 5° in which HOC1 is trapped by taurine to form taurine monochloramine. Taurine monochloramine is then quantified 49 C. von Ritter, M. B. Gfisham, and D. N. Granger, Gastroenterology 96, 811 (1989). 50 E. L. Thomas, M. B. Grisham, and M. M. Jefferson, J. Clin. Invest. 72, 444 (1983).
572
A N T I O X l D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[58]
by its ability to oxidize 2 mol of the yellow chromophore 5-thio-2-nitrobenzoic acid to 1 mol of the colorless 5,5'-dithiobis(2-nitrobenzoic acid). Conclusion Taken together the experimental data demonstrate that 5-ASA possesses potent antioxidant activity with the ability to scavenge a variety of free radicals and the ability to decompose neutrophilic oxidants (e.g., HOC1) and detoxify hemoprotein-associated oxidizing agents. Furthermore, 5-ASA has the additional property of being able to chelate iron and render it poorly redox active. Thus, the reason that 5-ASA may be so effective as an antiinflammatory agent in vivo may be due to its multitude of antioxidant properties. Acknowledgments Some of the work reported in this chapter was supported by a grant from the National Institutes of Health (DK43785; Project 6).
[58] A n t i o x i d a n t A c t i o n o f S t o b a d i n e By LUBICA
HORAKovA, H E L M U T SIES, and STEEN STEENKEN
Introduction Stobadine is a pyridoindole derivative synthesized in a search of new antiarrhythmic drugs) '2 It has been demonstrated that besides this effect it also reveals a-adrenolytic, antihistaminic, local anesthetic, and sedative effects. In comparison to its (+)-cis isomer it is less toxic and more powerful in diminishing epinephrine-induced arrhythmia. A protective, antihypoxic effect on myocardium and brain tissue has also been observed. The pharmacological effects of stobadine have been summarized. 3 The synthesis of stobadine is shown in Scheme 1. Treatment of ptolylhydrazine (1) with N-methyl-4-piperidone (II) yields 2,8-dimethyl2,3,4,5-tetrahydro-lH-pyrido[4,3-b]indole (liD. After catalytic hydrogenation of Ill, the (-+)-cis isomer (IV) is obtained. This isomer is recrystallized several times with (+)-dibenzoyltartaric acid to obtain the optically S. ~tolc, V. Bauer, L. Bene~, and M. Tich~, Czech. Patent 229,067 (1983). 2 S. ~tolc, V. Bauer, L. Bench, and M. Tich~, Swiss Patent 651,754 (1985). 3 L. Bene~ and S. ~tolc, Drugs Future 14, 135 (1989).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
572
A N T I O X l D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[58]
by its ability to oxidize 2 mol of the yellow chromophore 5-thio-2-nitrobenzoic acid to 1 mol of the colorless 5,5'-dithiobis(2-nitrobenzoic acid). Conclusion Taken together the experimental data demonstrate that 5-ASA possesses potent antioxidant activity with the ability to scavenge a variety of free radicals and the ability to decompose neutrophilic oxidants (e.g., HOC1) and detoxify hemoprotein-associated oxidizing agents. Furthermore, 5-ASA has the additional property of being able to chelate iron and render it poorly redox active. Thus, the reason that 5-ASA may be so effective as an antiinflammatory agent in vivo may be due to its multitude of antioxidant properties. Acknowledgments Some of the work reported in this chapter was supported by a grant from the National Institutes of Health (DK43785; Project 6).
[58] A n t i o x i d a n t A c t i o n o f S t o b a d i n e By LUBICA
HORAKovA, H E L M U T SIES, and STEEN STEENKEN
Introduction Stobadine is a pyridoindole derivative synthesized in a search of new antiarrhythmic drugs) '2 It has been demonstrated that besides this effect it also reveals a-adrenolytic, antihistaminic, local anesthetic, and sedative effects. In comparison to its (+)-cis isomer it is less toxic and more powerful in diminishing epinephrine-induced arrhythmia. A protective, antihypoxic effect on myocardium and brain tissue has also been observed. The pharmacological effects of stobadine have been summarized. 3 The synthesis of stobadine is shown in Scheme 1. Treatment of ptolylhydrazine (1) with N-methyl-4-piperidone (II) yields 2,8-dimethyl2,3,4,5-tetrahydro-lH-pyrido[4,3-b]indole (liD. After catalytic hydrogenation of Ill, the (-+)-cis isomer (IV) is obtained. This isomer is recrystallized several times with (+)-dibenzoyltartaric acid to obtain the optically S. ~tolc, V. Bauer, L. Bene~, and M. Tich~, Czech. Patent 229,067 (1983). 2 S. ~tolc, V. Bauer, L. Bench, and M. Tich~, Swiss Patent 651,754 (1985). 3 L. Bene~ and S. ~tolc, Drugs Future 14, 135 (1989).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[58]
573
ANTIOXIDANT ACTION OF STOBADINE
H3C~ + C?/CH3 ~t'~N~'NH2 O" v H
~.H 3 C ~ N ~ ' C H 3 H
(m)
H3C~N/CH3 (iv)
STOBADINE SCHEME 1
pure ( - ) - c i s enantiomer. The reaction with corresponding acids yields stobadine dihydrochloride or dipalmitate. Intensive studies of stobadine as an antioxidant were initiated after the finding of its antiarrhythmic cardioprotective and antihypoxic effects on myocardium.] As the participation of free radicals in the pathobiology of hypoxia (ischemia)-reperfusion injury has received considerable experimental support, the possibility of stobadine forming stable free radicals was investigated. Some Characteristics of Stobadine
For stobadine dihydrochloride, the pK a values for indole nitrogen and methyl nitrogen deprotonation are 3.2 and 8.7 as determined by spectrophotometric and potentiometric titration: The one-electron oxidation potential of stobadine at pH 7 (0.58 V/NHE) studied by pulse radiolysis 5 is lower than the potentials determined for biological target molecules such as purines and pyrimidines 6 as well as amino acids, 7,8 which indicates that stobadine can repair oxidized bases and amino acids by electron donation in biological systems. 4 M. ~tefek, L. Bene~, and V. Zelnik, Xenobiotica 19, 143 (1989). 5 S. Steenken, A. R. Sundquist, S. V. Jovanovi6, R. Crockett, and H. Sies, Chem. Res. Toxicol. 5, 355 (1992). 6 S. V. Jovanovi~ and M. G. Simic, Biochim. Biophys. Acta 1008, 39 (1989). 7 S. V. Jovanovi~, S. Steenken, and M. G. Simic, J. Phys. Chem. 95, 684 (1991). 8 M. R. De Felippis, C. P. Murthy, M. Faraggi, and M. H. Klapper, Biochemistry 28, 4847 (1989).
574
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[58]
Assay for Stobadine Several assay methods for stobadine in various biological matrices have been described: spectrofluorometric analysis of extracts from serum and urine,9 gas-liquid chromatography to study stobadine and metabolites in liver microsomes, 1° radiometric methods using radioactively labeled stobadine dihydrochloride to measure the unchanged drug by liquid scintillation counting, H and high-performance liquid chromatography (HPLC) to assess stobadine pharmacokinetics in dogs and humans ~2and to determine the concentration of stobadine in microsomes and liposomes.13 Antioxidant Properties of Stobadine
Radical Formation The ability of stobadine to form stable free radicals was first established in 1989.14 In the presence of oxygen, stobadine dissolved in ethanol generated free radicals by the action of y radiation. The electron spin resonance (ESR) spectrum reflected the presence of an NO. radical on the indole nitrogen. 14These findings were confirmed in a study where an ESR-observable radical of stobadine with a well-resolved hyperfine structure was detected on oxidation of stobadine with PbO 2 and tert-butyl hydroperoxide (tBuOOH). 15 Because no hyperfine splittings from the proton in the/3 position to the nitroxyl group could be detected, it was proposed that the intermediate radical II was converted by oxidation to nitrone III. After trapping a radical (possibly tBuOO, or tBuO-), adduct IV was formed (Scheme 2). It might be supposed that these steps of radical formation reflect one way by which stobadine scavenges reactive radicals. An alternative scavenging mechanism, involving electron transfer, was evaluated by using pulse radiolysis with optical detection, 5 showing the importance of the indole nitrogen. One-electron oxidation of stobadine with radicals such as C6H50., CC1302- , Br2 v, and HO. (reaction rate constants - 5 × 108-101°M - 1 sec-~) led to formation of the radical cation (absorbance maxima at 280 and 445 nm) which deprotonates from the indolic nitrogen (pKa 5.0) to give a nitrogen-centered radical (absorbance 9 V.Marko, Pharmazie 40, 192 (1985). l0 M. ~tefek and L. Bene~, J. Chromatogr. 415, 163 (1987). 11 V. ~ a s n ~ and M. ~tefek, Radioanal. Nucl. Chem. 111, 117 (1987). 12 L. ~olt6s, Z. K~llay, ~. Bezek, and V. Fedele~ovtt, BiopharmacoL Drug Dispos. 12, 29 (1991). 13 L'. Horfikovfi, K. Briviba, and H. Sies, Chem.-Biol. Interact. 83, 85 (1992). 14 H. Sz6csov~i and L. Bene~, Csl. Pharmazie 37, 121 (1989). 15 A. Sta~ko, K. Ondria~, V. Mi~fk, H. SzOcsovfi, and D. Gergel', Chem. Pap. 44, 493 (1990).
[58]
ANTIOXIDANT ACTION OF STOBADINE 9
H
I
H3C- .~S/ "A~ ~ r ~ b I A I -N-'CH3 ter¢-BuOOH H3Cv/~x ~ 7 ~ N ~ 3 PbO2 6 -i~I H 4
H
HY 'C ~N--CH3
H
~
575
N"-OH3
tert-
BuOOH
PbO 2
H
tert ~lauOO" or X
H 3 C ~ ~ x
N--CH3
O" IV
i
0 |1| SCHEME 2
maxima at 275, 335, and 410 nm), probably bearing a positive charge at the pyrido-nitrogen (Fig. 1).
Singlet Oxygen Quenching Stobadine is an effective quencher of ~O2 with an overall quenching rate constant of 1,3 × 108 M-1 sec-~, determined with the endoperoxide of 3,3'-(1,4-naphthylidene) dipropionate (NDPO2) as 10 2 source and by monitoring ~Ozphotoemission with a germanium diode.5 This rate constant is comparable to the constants for different tocopherol homologs (108 M sec-~) and is intermediate between the range of values determined for carotenoids (109 M -1 sec-1) 16 and thiols (106 M -1 sec-l), t7 Stobadine represents one of the most efficient water-soluble ~O2 quenchers reported to date.
Hydroxyl Radical Scavenging The ability of stobadine to scavenge the hydroxyl radical generated by a Fenton-type reaction was demonstrated using ESR spin trapping with 5,5-dimethyl-l-pyrroline N-oxide (DMPO). t8 The signal intensity of the DMPO-OH adduct reflected the relative O H radical concentration in the system. In the presence of 12 mM stobadine the height of the DMPO-OH signal was markedly decreased. i~p. Di Mascio, S. Kaiser, and H. Sies, Arch. Biochem. Biophys. 2"/4,532 (1989). 17T. P. A. Devasagayam,A. R. Sundquist, P. Di Mascio, S. Kaiser, and I-t. Sies, J. Photochem. Photobiol., B 9, 105 (1991). is K. Ondria~,V. Mi~fk,D. Gergel', and A. Stagko,Biochim. Biophys. Acza 111113,238 (1989).
ANTIOXlDANT CHARACTERIZATIONAND ASSAY
576
[58]
Me
Me
~
~N÷I H
-
/
I
-
I
I-I '
+H ÷ I
H
pKa
5.0
FIG. 1. Radical formation from stobadine.
The rate constant for the stobadine-'OH radical interaction was determined 19using two chemical methods for detecting "OH radicals: (I) deoxyribose oxidation to thiobarbituric acid (TBA)-reactive products and (2) ethylene production from 2-keto-4-methylthiobutyric acid (KMBA). Stobadine was found to be an efficient scavenger of "OH characterized by a second-order rate constant of approximately 1 × 101°M - 1 sec- ~ in both assays, in agreement with the directly measured value of 7 × 109 M - l sec-1.5 The potency of stobadine to prevent ethylene production from KMBA, characterized by the IC50 value, was compared for three different "OH generating systems, namely, a chemical, an enzymatic, and a membrane-bound enzymatic system. Similar IC50 values (0.74-0.93 mM) were obtained for the three systems studied, ~9 suggesting that stobadine is an efficient scavenger of "OH radicals produced not only free in solution but also in biological membranes.
Superoxide Radical Scavenging In the concentration range of 10 to 100/zM stobadine did not significantly scavenge superoxide anions generated in an enzymatic (xanthine/ xanthine oxidase) system, where superoxide anions were determined in vitro by spectrophotometric measurement of the reduction of ferricytochrome. 2° The rate constant for interaction of stobadine with superoxide was determined as approximately 103 M -1 s e c - l . 21 The IC50 value for inhibition ofpyrogallol autoxidation by stobadine was l0 mM, also indicating a low affinity of stobadine to superoxide; for comparison, the same ICs0 value for ascorbate is about 0.04 mM. 21 19M. ~tefek and L. Beneg, FEBS Lett. 294, 264 (1991). 20L'. Horfikov~, V. Uraz, O. Ondreji~kov~i,L'. LukoviC and I. Jur~nek, Biomed. Biochim. Acta 50, 1019 (1991). 21j. Humplov~f,M. ~tefek, L. Bene~, and L. Kri~anov~i, Cesk. Fysiol. 41, 80 (1992).
[58]
ANTIOXIDANT ACTION OF STOBADINE
577
As measured by superoxide-induced lucigenin-amplified chemiluminescence, the second-order rate constant for the reaction of stobadine with superoxide was estimated to be 7.5 x 102 M -1 sec-1. 22
Antioxidative Effect in Liposomes Stobadine exhibited its antioxidative effect in phosphatidylcholine liposomes, incubated under air at 50°, where lipid peroxidation was detected spectroscopically for conjugated diene and thiobarbituric acid product formation. 18 Several drugs were tested in different liposomal/drug molar ratios, and stobadine was found to be more effective than butylated hydroxytoluene (BHT). Stobadine reactivity with peroxyl radicals in liposomes was examined using a lipid-soluble azo initiator of peroxyl radicals, 2,2'-azobis(2,4-dimethylvaleronitrile) (AMVN). Stobadine exerted scavenging as evidenced by the inhibition of (1) cis-parinaric acid fluorescence decay (half-maximal effect at 20/zM) and (2) luminol-sensitized chemiluminescence (half-maximal effect at 33/zM). 22
Mechanisms of Antioxidant Action in Microsomes Stobadine effectively inhibited thiobarbituric acid-reactive product formation in a concentration-dependent manner, with an ICs0 value of 56 t~M in cumene hydroperoxide-induced lipid peroxidation in microsomes, and it also inhibited oxygen consumption. 23Nonenzymatic lipid peroxidation induced by ascorbate (0.5 mM)/ferrous ion (50/xM) in heat-denatured microsomes was also inhibited, with an ICs0 of 25/zM. Because stobadine did not prevent cumene hydroperoxide bioactivation 23and because superoxide anion radicals and hydroxyl radicals are not involved in cumene hydroperoxide-dependent lipid peroxidation, 24'25 the inhibitory effect of stobadine may be at least partially explained by its scavenging of alkoxyl (LO') and peroxyl (LOO') radicals. Stobadine was equally efficient in inhibiting microsomal lipid peroxidation induced by the lipid-soluble azo initiator of peroxyl radicals AMVN or the water-soluble 2,2'-azobis(2-aminopropane) hydrochloride (AAPH), azo initiators of peroxyl radicals with half-maximal effect at 17/zm. 22The stobadine octanol-water partition coefficient (log P of 0.57) explains its ability to quench peroxyl radicals in both lipid and aqueous phases. 22 V. E. Kagan, M. Tsuchiya, E. Serbinova, L. Packer, and H. Sies, Biochem. Pharmacol. 45, 393 (1993). 23 M. ~tefek, M., Masarykov~i, and L. Bene~, Pharmacol. Toxicol. 711, 407 (1992). 24 R. H. Weiss and R. W. Estabrook, Arch. Biochem. Biophys. 251, 348 (1986). 25 j. A. Thompson and N. P. Yumibe, Drug Metab. Rev. 20, 365 (1989).
578
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[58]
The antioxidant effect of stobadine was also examined with iron/ADP/ N A D P H as prooxidant in normal and vitamin E-deficient rat liver microsomes. 13 L o w level chemiluminescence was used as a sensitive method of monitoring lipid peroxidation. The duration of the lag phase is a measure of susceptibility to peroxidation, and stobadine at 5 /zM was found to double the lag phase. Stobadine at 5 / z M was without effect on vitamin E-deficient microsomes. Addition of N A D P H to microsomes activated the c y t o c h r o m e P-450 system to metabolize stobadine, z6 When liposomal peroxidation was started by adding F e S O 4 with the reductants N A D P H or ascorbate, stobadine was not consumed. It thus appears that in microsomes the antioxidant effect of stobadine depends on vitamin E, and stobadine is metabolized in the presence of N A D P H . In liposomes, stobadine reacted with lipid radicals, and N A D P H prevented the loss of stobadine. Antioxidant Effect in Mitochondria
The inhibitory effect of stobadine on FeZ+/ascorbate-induced lipid peroxidation was studied in brain mitochondria in comparison to the effect of B H T and nifedipine, z7 The ICs0 values for B H T , stobadine, and nifedipide were about 1,200, and 800/zM, respectively. Preventive Effect in lschemia and Reperfusion
Stobadine markedly delayed the onset of epinephrine-induced arrhythmia in laboratory animals, 28 decreased functional damage, and reduced the infarct size of isolated rat hearts subjected to a period of ischemia followed by reperfusion. 3'29 Isoproterenol could be a trigger of heart injury caused by free radicals.3° Administration of stobadine prevented a decrease in the content o f protein thiol groups and glutathione in hearts treated with high doses ofisoproterenol (30 mg/kg). 31 Moreover, stobadine also attenuated the increase in the content of malondialdehyde and decreased the activities of catalase and glutathione reductase as well as the ratio of reduced to oxidized glutathione 26M. ~tefek, L. Bene~, M. Jergelov~i, V. S~asmir, and L. Turi-Nagy, Xenobiotica 17, 1067 (1987). 27L'. Hor,'ikovfi,I. Jurfmek, and B. Bokn~ov~i, Biologia (Bratislava) 45, 313 (1990). 28A. Babulovfi, L. Buran, and L. Bene~, Farm. Obz. 54, 15 (1985). 29I. Gabauer, J. Styk, J. Slezfik,J. Okoli~finy,V. Holec, and L'. Bene~,Bratisl. Lek. Listy 85, 265 (1986). 30S. L. Jewet, L. J. Eddy, and P. Hochstein, Free Radical Biol. Med. 6, 185 (1989). 3l O. Ondreji~kov~, A. D~urba, J. Sedlfik, J. Tokfirov~i,T. Ma~i~kovfi, and L. Bene~, Biomed. Biochim. Acta 50, 1251 (1991).
[58]
ANTIOXIDANT ACTION OF STOBADINE
579
(GSH/GSSG) observed in heart mitochondria isolated from isoproterenoltreated animals. 31 A protective effect of stobadine was also demonstrated in several models of brain hypoxia followed by reoxygenation. 2°'a2'33 The dose employed in these experiments (2 mg/kg) is within the dose range typically used in testing the pharmacodynamic effects of stobadine in preclinical studies and is in the upper range of expected therapeutic doses. Stobadine decreased the level of conjugated dienes and TBA-reactive substances and maintained the level of total thiol groups after complete brain ischemia followed by incubation of brain cortex homogenates in a stream of wet nitrogen (hypoxia) and wet air (reoxygenation). 3z The ability of stobadine to prevent lipid peroxidation was tested also in vivo in incomplete rat cerebral ischemia models induced by 4 hr of ligation of the common carotid arteries with a subsequent 10 min of reperfusion. 2° The concentration of conjugated dienes and TBA-reactive substances significantly decreased in animals treated with therapeutic doses of stobadine (2 mg/kg), administered intravenously immediately before reperfusion or 10 rain after the onset of reperfusion. Stobadine was more effective than vitamin E, given in a dose of 30 mg/kg per day intramuscularly over 3 consecutive days prior to ischemia. Significant changes were found in the activities of antioxidative enzymes, namely, an increase in superoxide dismutase and a decrease in glutathione reductase activities in brain cortex samples. Stobadine prevented these changes. The density and affinity of a-adrenergic binding sites was investigated in the same experiments. Compared to the group of sham-operated animals, decreased density and increased affinity of [3H]dihydroergocryptine binding sites were found in cerebrocortical membranes of rats subjected to 4 hr of incomplete ischemia and 1 hr of reperfusion. Stobadine and vitamin E prevented these changes, indicating that oxygen free radicals might play a role in these processes. 34 The beneficial effect of stobadine was also demonstrated by higher survival of rats subjected to brain ischemia followed by reperfusion. 2° In the group of animals subjected to ischemia and reperfusion, 61% of the animals survived for 1 hr after reperfusion, 11.8% had seizures. After treatment with vitamin E or stobadine, 82 and 90% of the animals survived, respectively. Animals treated with either vitamin E or stobadine before ischemia and reperfusion were without seizures. 32 S. ~tolc and L. Horfikovfi, in " N e w Trends in Clinical Neuropharmacology" (D. Bartko, P. Tur~finy, and G. Stern, eds.), pp. 59-63. John Libbey, London, 1988. 33 L'. Horftkovfi, O. Ondreji6kovfi, V. Uraz, L'. Lukovi6, and I. Jur~nek, Experientia 48, 872 (1992). 34 Z. Kvaltfnovfi, L'. Lukovi~, and S. ~tolc, Neuropharmacology 38, 785 (1993).
580
ANTIOXlDANT CHARACTERIZATION AND ASSAY
[59]
Summary
The above-mentioned physicochemical, chemical, as well as pharmacological properties including successful results of Phase I and II clinical testing of stobadine as an antianginal agent permit us to consider this compound as potential drug in prevention and/or treatment of tissue injuries caused by oxidative stress.
[59] Nitroxides as A n t i o x i d a n t s
By MURALI C. KRISHNA and AMRAM SAMUN! Introduction Low molecular weight nitroxides are nonimmunogenic, cell-permeable, nontoxic 1stable radicals that readily partition among various cellular compartments. As paramagnetic species, detectable by electron paramagnetic resonance (EPR), nitroxides report on subtle changes in their chemical environment. 2 Consequently they have been predominantly used as biophysical markers to probe cellular metabolism, intracellular pH, oxygen level, molecular mobility of proteins and lipids, and membrane structure. 3-6 Additionally, nitroxides were proposed as contrast agents for nuclear magnetic resonance (NMR) imaging. 7,8These capacities of nitroxides have been extensively investigated, reported, and previously reviewed. 9 Being fully substituted in the ortho position, low molecular weight nitroxides are subject to steric hindrance that inhibits their dismutation, thus rendering them stable. However, nitroxides undergo chemical and cellular reduction to the corresponding one-electron reduced hydroxylamine, l° The hydroxylamine, on the other hand, is oxidized back to the nitroxide by various oxidants (chemical and cellular). The cellular reducI E. G. Ankel, C. S. Lai, L. E. Hopwood, and Z. Zivkovic, Life Sci. 40, 495 (1987). 2 L. J. Berliner, ed., "Spin Labeling," Vols. 1 and 2. Academic Press, New York, 1979. 3 H. M. Swartz, M. Sentjurc, and P. D. Morse II, Biochim. Biophys. Acta 888, 82 (1986). 4 j. Fuchs, W. H. Nitschmann, L. Packer, O. H. Hankovszky, and K. Hideg, Free Radical Res. Commun. 10, 315 (1990). 5 W. Froncisz, C. S. Lai, and J. S, Hyde, Proc. Natl. Acad. Sci. U.S.A. 82, 411 (1985). 6 j. F. Glockner, H. C. Chan, and H. M. Swartz, Magn. Reson. Med. 20, 123 (1991). 7 R. C. Brasch, Radiology 147, 781 (1983). 8 j. F. Keana and N. F. Van, Physiol. Chem. Phys. Med. N M R 16, 477 (1984). 9 A, Ionnone and A. Tomasi, Acta Pharm. Jugosl. 41, 277 (1991). l0 S. Belkin, R. J. Mehlhorn, K. Hideg, O. Hankovsky, and L. Packer, Arch. Biochem. Biophys. 256, 232 (1987).
METHODS 1N ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All fightsof reproductionin any formreserved.
580
ANTIOXlDANT CHARACTERIZATION AND ASSAY
[59]
Summary
The above-mentioned physicochemical, chemical, as well as pharmacological properties including successful results of Phase I and II clinical testing of stobadine as an antianginal agent permit us to consider this compound as potential drug in prevention and/or treatment of tissue injuries caused by oxidative stress.
[59] Nitroxides as A n t i o x i d a n t s
By MURALI C. KRISHNA and AMRAM SAMUN! Introduction Low molecular weight nitroxides are nonimmunogenic, cell-permeable, nontoxic 1stable radicals that readily partition among various cellular compartments. As paramagnetic species, detectable by electron paramagnetic resonance (EPR), nitroxides report on subtle changes in their chemical environment. 2 Consequently they have been predominantly used as biophysical markers to probe cellular metabolism, intracellular pH, oxygen level, molecular mobility of proteins and lipids, and membrane structure. 3-6 Additionally, nitroxides were proposed as contrast agents for nuclear magnetic resonance (NMR) imaging. 7,8These capacities of nitroxides have been extensively investigated, reported, and previously reviewed. 9 Being fully substituted in the ortho position, low molecular weight nitroxides are subject to steric hindrance that inhibits their dismutation, thus rendering them stable. However, nitroxides undergo chemical and cellular reduction to the corresponding one-electron reduced hydroxylamine, l° The hydroxylamine, on the other hand, is oxidized back to the nitroxide by various oxidants (chemical and cellular). The cellular reducI E. G. Ankel, C. S. Lai, L. E. Hopwood, and Z. Zivkovic, Life Sci. 40, 495 (1987). 2 L. J. Berliner, ed., "Spin Labeling," Vols. 1 and 2. Academic Press, New York, 1979. 3 H. M. Swartz, M. Sentjurc, and P. D. Morse II, Biochim. Biophys. Acta 888, 82 (1986). 4 j. Fuchs, W. H. Nitschmann, L. Packer, O. H. Hankovszky, and K. Hideg, Free Radical Res. Commun. 10, 315 (1990). 5 W. Froncisz, C. S. Lai, and J. S, Hyde, Proc. Natl. Acad. Sci. U.S.A. 82, 411 (1985). 6 j. F. Glockner, H. C. Chan, and H. M. Swartz, Magn. Reson. Med. 20, 123 (1991). 7 R. C. Brasch, Radiology 147, 781 (1983). 8 j. F. Keana and N. F. Van, Physiol. Chem. Phys. Med. N M R 16, 477 (1984). 9 A, Ionnone and A. Tomasi, Acta Pharm. Jugosl. 41, 277 (1991). l0 S. Belkin, R. J. Mehlhorn, K. Hideg, O. Hankovsky, and L. Packer, Arch. Biochem. Biophys. 256, 232 (1987).
METHODS 1N ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All fightsof reproductionin any formreserved.
[59]
581
NITROXIDES AS ANTIOXIDANTS TABLE I OXIDATION/REDUCTION MIDPOINT POTENTIALS OF NITROXIDE RADICALS
Nitroxide TEMPO TEMPOL TEMPAMINE TEMPONE
3-Carboxyproxyl 3-Aminomethylproxyl 3-Carbamoylproxyl 3-Cyanoproxyl 3-Carbamoyl-3-pyrroline OXANO C y c l o h e x a n e doxyl
Midpoint potential (mV versus normal hydrogen electrode)
Reversibility
722 810 826 913 792 853 861 976 966 960 900
+ + + + + + + + + -
tion of the nitroxide and the oxidation of the hydroxylamine depend on many factors which include the redox status of the cells and oxygen tension. The chemical reduction of nitroxides has also been shown to be determined by the substituent inductive effects, n Nitroxides can be classified into two types based on electrochemical behavior in cyclic voltammetric experiments. 12(1) The oxazolidine derivatives exhibit irreversible redox behavior in the regions of positive and negative potentials. (2) The piperidine and proxyl derivatives form a redox couple with their oxidized intermediate and exhibit a reversible behavior in the regions of positive potential. The midpoint redox potentials are listed in Table I. Less common has been the use of nitroxides as antioxidants. Initially, the participation of nitroxide radicals in one-electron redox reactions attracted research interest because such reactions affected their stability in cellular systems) ° Only a few attempts have been made to investigate their toxic effects 13 or potential role as antioxidants. 14'~ The protection of various biological systems by nitroxides, however, and particularly 11 S. Morris, G. S o s n o v s k y , B. Hui, C. O. Huber, N. U. Rao, and H. M. Swartz, J. Pharm. Sci. 80, 149 (1991). 12 M. C. Krishna, D. A. G r a h a m , A. Samuni, J. B. Mitchell, and A. Russo, Proc. Natl. Acad. Sci. U.S.A. 89, 5537 (1992). 13 H. Sies and R. Mehlhorn, Arch. Biochem. Biophys. 251, 393 (1986). 14 I. T. Brownlie and K. U. Ingold, Can. J. Chem. 45, 2427 (1967). 15 T. J. Weil, J. Van der Veen, and H. S. Olcot, Nature (London) 219, 168 (1968).
582
ANTIOXIDANT CHARACTERIZATIONAND ASSAY
[59]
their antioxidative activity have been rapidly established. 16-23In this chapter, the procedures adopted for applying and assaying the antioxidative activity o f nitroxides are presented.
Assay of Superoxide Dismutase Mimetic Activity Because the redox potential o f the nitroxide/hydroxylamine couple of oxazolidine derivatives, such as O X A N O , is about - 0 . 3 4 V , the system can oxidize superoxide anion ( 0 2 0 to O2.24 The hydroxylamine in turn can react with another 027 reducing it to H202. By this cyclic reaction, O X A N O is continuously restored and hence effectively dismutates O2 ~ yielding 02 and H202.
N - - O + 0 2-
O--~
/Z•
+
H+ ~ & ,
OH + 027 + H +
o-V/--OH
k2) 0 7 - - 0
+ 02
(1)
+ H20 2
(2)
Such superoxide dismutase (SOD) mimic activity of nitroxides is demonstrable by EPR, UV-visible spectrophotometry, and chemiluminescence techniques. 24 Unlike the case for native SOD, both k 1 and k2 depend on the p H and so does the ratio kl/k2. The individual reaction rate constants, however, cannot be derived from EPR measurements and should be determined by some independent technique such as the SOD-inhibitable 16U. A. Nilsson, L. I. Olsson, G. Carlin, and A. C. Bylund-Fellenius, J. Biol. Chem. 2,64, 11131 (1989). 17j. B. Mitchell, A. Samuni, M. C. Krishna, W. G. DeGraff, M. S. Ahn, U. Samuni, and A. Russo, Biochemistry 29, 2802 (1990). 18R. I. Zhdanov and P. G. Komarov, Free Radical Res. Commun. 9, 367 (1990). 19M. V. Bilenko, P. G. Komarov, A. A. Morgunov, and R. I. Zhdanov, Byull. Eksp. Biol. Med. 111, 500 (1991). 20j. An and A. W. Hsie, Mutat. Res. 270, 167 (1992). 21N. P. Konovalova, R. F. Diatchkovskaya, L. M. Volkova, and V. N. Varfolomeev, Anticancer Drugs 2, 591 (1991). = A. Samuni, D. Winkelsberg, A. Pinson, S. M. Hahn, J. B. Mitchell, and A. Russo, J. Clin. Invest. 87, 1526(1991). 23y. Miura, H. Utsumi, and A. Hamada, Arch. Biochem. Biophys. 300, 148(1993). 24A. Samuni, C. M. Krishna, P. Riesz, E. Finkeistein, and A. Russo, J. Biol. Chem. 263, 17921 (1988).
[59]
NITROXIDES AS ANTIOXIDANTS
583
reduction of ferricytochrome c (Cyt cnI). Piperidine nitroxide derivatives such as TEMPO, TEMPOL, or TEMPAMINE are oxidized by superoxide, but the oxidized intermediate, oxo-ammonium cation, is rapidly reduced by another O2 ~ as demonstrated for TEMPOL~2:
HO
- - O + 027 + 2H ÷
HO~~N~O
+ 027
~4 )
) HO
HOCN'--O
=O
+ H202
+ 0 2
(3)
(4)
Because k 3 is much less than k4, the nitroxide concentration is practically unaffected, although the net reaction is the same and the nitroxide is regenerated while 02 ~ is catalytically removed.12 Electron Paramagnetic Resonance The EPR experiment is performed by exposing approximately 10/zM nitroxide to 02- flux and following the residual nitroxide signal. A prolonged continuous flux of 02 ~ is achieved using 0.1 U/ml xanthine oxidase (EC 1.1.3.22, xanthine : oxygen oxidoreductase) and 1 mM hypoxanthine in 10 mM phosphate buffer at pH 7 containing 0.1 mM diethylenetriaminepentaacetic acid (DTPA). To ensure a steady rate of 027 formation, the reaction is carried out in a gas-permeable Teflon capillary (Zeus Industries, Raritan, NJ) of 0.8 mm inner diameter and 0.038 mm wall thickness. The tube containing the sample is folded twice, inserted in a quartz EPR tube opened at both ends, and placed in the EPR cavity. A constant oxygen concentration is maintained by flowing gas of the desired composition and temperature (measured by a thermocoupled probe set within the cavity) around the sample. Nitroxides do not react directly with H202, yet, in order to regenerate 02 from any accumulated H202, it is beneficial to add 100 U/ml catalase (SOD-free preparation from Boehringer, Mannheim, Germany). The intensity of the middle line of the nitroxide EPR spectra is monitored as a function of time. Spin loss occurring by reaction (1) is noted by a decrease of the initial signal (/initial) which subsequently reaches an intermediate equilibrium intensity. This decrease is attributed to the formation of the one-electron reduction product from the nitroxide, namely, the hydroxylamine. Later,
584
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[59]
following an accumulation of the hydroxylamine, reactions (1) and (2) balance one another, and the intensity (I) of the residual EPR signal becomes time-invariant. Since O2: is continuously removed through reactions (1) and (2) while H202 and molecular oxygen are generated, the reaction mechanism does not represent a genuine chemical equilibrium but rather a steady state where three ongoing processes affect the nitroxide concentration:
d[nitroxide]/dt = kE[hydroxylamine][O2-] + k_l[OE][hydroxylamine ] -
kl[nitroxide][O2:]
(5)
Therefore, in the presence of both oxygen and 02", a steady state is achievable reflected by a time-invariant ratio of [nitroxide]/[hydroxylamine]: [nitroxide]/[hydroxylamine] = kz/kl + k_l[OE]/kl[02-]
(6)
Under conditions where k_ 1[O2] <~ kl [ 0 2 7 ] the ratio [nitroxide] /[hydroxylamine] approaches k2/kl. Because
I/(Iinitia I
-
I) =
[nitroxide]/[hydroxylamine] = kz/kl
(7)
the relative intensity (I/Iinitial) of the residual signal reflects the steadystate concentration ratio [nitroxide]/([nitroxide] + [hydroxylamine]) and provides information about the reaction kinetics of catalysis. Ionizing radiation also serves as a source for a continuous flux of O2: radicals, where irradiation is performed with any steady state y source such as 6°C0 or 137Cs. When nitroxide in 10 mM phosphate buffer at pH 7 is y-irradiated in the presence of 0.1 M formate and oxygen, all the primary radicals are converted to 02: radicals. To prevent any direct reduction of nitroxide by formate radical, the solution should be saturated with 100% O2 and the nitroxide concentration kept as low as possible. After various radiation doses, samples are taken for EPR measurement, and the residual nitroxide signal is compared to the initial one, that is, nonirradiated sample. Reduction of FerricFtochrome c The SOD-inhibitable reduction of Cyt ¢III c a n be employed 25 to study the SOD mimic activity of nitroxides. The experimental procedure is identical to that employed in assaying the activity of any SOD preparation wherein the enzymatic system hypoxanthine/xanthine oxidase (HX/XO) is used as a biochemical source for O2:. Nitroxides are not substrates for 25 j. D. Crapo, J. M. McCord, and I. Fridovich, this series, Vol. 53, p. 382.
[59]
NITROXIDES AS ANTIOXIDANTS
585
XO and neither interfere with the production of Oz ~ nor react with Cyt C Ill. Because the reaction rate constants of O2 ~ with nitroxides are generally smaller by 3-5 orders of magnitude than that with native SOD, the nitroxide concentrations should be kept high enough to effectively compete with Cyt c m for O2 ~. Therefore, with 10-50/zM Cyt c m, the nitroxide concentration might range from 50 t~M up to 10 mM. As in the common assay for SOD activity, the reaction mixture should contain catalase to prevent oxidation of Cyt CII by H202 and DTPA to minimize XO inactivation by redox-active metals. The same procedure is applicable for studying the reaction of O2 ~ with the respective cyclic hydroxylamines. However, hydroxylamines are capable of reducing Cyt C III and interfering with the assay. Although the rate constants of reaction of hydroxylamines with Cyt c m are relatively low, it is recommended to account for the contribution of any reductive processes independent of O2 ~ by adding to the reference cuvette, which contains an excess (500 units) of SOD, anywhere from 50 tzM up to 10 mM of the cyclic hydroxylamine. Luminogenic Assay Alternatively, assay of the SOD mimic activity of nitroxides may be done monitoring the inhibition of Oz'-induced luminescence from lucigenin (Aldrich Chemical Co., Milwaukee, WI). A continuous flux of 02 ~ is generated using 5 mM HX and 0.3 U/ml XO in the presence of 250/zM lucigenin in aerated 50 mM buffered phosphate, pH 7.8, containing 50 /zM DTPA, and the chemiluminescence (CL) emitted is continuously monitored using an SLM-8000 spectrofluorimeter in the photon counting mode or any other luminometer. As long as neither HX nor 02 is depleted from the assay mixture, the 027 flux is practically constant and so is the monitored CL. On the addition of known amounts of nitroxide, which react with a fraction of the O2 ~ formed, the emitted light (cl) decreases accordingly. By plotting CL/cl versus [nitroxide] and knowing ko2_+lucigenin to be 103 M -~ sec -1, catalytic rate constant of the nitroxide with Oz ~ can be found: C L / c l = 1 + kl[nitroxide]/([lucigenin]ko2_+lucigenin)
(8)
The rate constants of O2 ~ dismutation by various nitroxides derived from spectrophotometry, luminescence, and EPR studies are given in Table II. Nitroxide-Metal Interaction Common antioxidants such as ascrobate, a-tocopherol, and reduced glutathione (GSH) act as mild reductants. In contrast, nitroxides can
586
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[59]
T A B L E II RATE CONSTANTS OF NITROXIDE REACTION WITH SUPEROXIDE
Nitroxide OXANO C y c l o h e x a n e doxyl TEMPO TEMPOL TEMPAMINE
Reaction rate c o n s t a n t s ( M - 1 s e c - l) 1.1 3.5 1.2 6.5 6.5
× × × × ×
103 102 105 104 104
function as oxidants. Predominantly, nitroxides protect targets from biological damage by oxidizing reduced transition metals 26 and consequently inhibiting their participation in Fenton reactions, z7 The deleterious nature of traces of labile redox-active metal ions coordinated to essential cellular sites is well documented. By reacting with H202 metal ions form peroxo complexes which, in turn, site-specifically yield hypervalent metals and/ or hydroxyl radical (.OH). Contrary to the reaction with oxygen or H20 2, metal oxidation by nitroxide is not accompanied by the production of any reactive secondary species. Such oxidation by nitroxide inhibits the reduced metal ion from participation in Fenton chemistry. To assay the nitroxide reaction with metals such as iron(II), EPR spectroscopy and UV-visible absorption spectrophotometry can be used. A major experimental limitation is the stability and solubility of iron ion under physiological conditions. The reaction is carried out under anoxic conditions to prevent any interference from 02, 027, or H202. There is no evidence of any nitroxide-metal complex, yet a ligand such as ADP or citrate should be added to maintain iron in solution while preserving its redox activity. It is anticipated, however, that various ligands would differently affect the metal reaction with the nitroxide. The reaction kinetics can be investigated by following the increase of absorbance at 353 nm resulting from Fe(III) formation and the decay of the nitroxide EPR signal. When either the nitroxide or the Fe(II) is maintained in large excess, both the appearance of OD353nm and the spin loss, respectively, obey pseudo-first-order kinetics. Figure 1 demonstrate a typical experiment performed with OXANO and DNA-bound Fe(II). By 26 U. L. Nilsson, G. Carlin, and A. C. Bylund-Fellenius, Chem.-Biol. Interact. 74, 325 (1990). 27 A. Samuni, D. Godinger, J. Aronovitch, A. R u s s o , and J. B. Mitchell, Biochemistry 30, 555 (1991).
[59]
NITROXIDES AS ANTIOXIDANTS 0.4
587 -1 °
m
0.2
8~
o
-3
0.0
~,
-4 0
100
200
Time (sec)
"5 -~
100
_= 3 5o
"2
r,1 0
.
0
.
.
.
|
100 200 T i m e (sec)
FIG. 1. Reaction between OXANO and DNA-Fe(II). OXANO was anoxically mixed with iron(II) in 50 mM 3-(N-morpholino)propanesulfonic acid (MOPS) buffer, pH 7, containing 0.1 mg/ml salmon DNA at 22°. The appearance of DNA-Fe(III) was spectrophotometrically monitored at 353 nm, whereas the spin loss of OXANO was monitored by following its EPR signal. (Top) Time dependence of AOD353nm ([~) and of ln(OD~ - ODt) ( ' ) resulting on mixing 1 mM OXANO with 0.1 mM Fe(II). (Bottom) Time dependence of EPR signal ([]) and In(EPR signal) (m) on mixing 50 ~ M OXANO with 1 mM Fe(II). (From Mitchell et al. 17 with permission.)
varying the concentration of reactant that is kept in excess and repeating the experiment, the second-order reaction rate constant for the oxidation of ferrous-DNA complex, k 9 c a n be calculated. DNA-Fe 2+ + R R ' N O + H+--~ DNA-Fe 3+ + RR'NOH
(9)
An alternative metal-nitroxide interaction involves ferryl species formed on the reaction of metmyoglobin (metMb), methemoglobin (metHb), or peroxidase with H202. In this reaction TEMPOL was found
588
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[59]
to act as catalase mimic and catalyze H202 dismutation. According to the proposed mechanism, heme-ferryl species oxidize nitroxides to the respective oxoammonium ion which in turn oxidizes heme-Fe(II) to heme-Fe(III). 28The assay of nitroxide catalase mimic activity is achieved by following polarographically the evolution of 02 under anoxia in the presence of H202, nitroxide, and metHb or metMb. Under aerobic conditions, however, nitroxide reportedly induces the formation of H202.29 Concomitant spin loss of the nitroxide was measurable by EPR spectroscopy. Such measurement, however, under common microwave power of 10-20 mW is prone to error, since changes in the 02 level cause artifactual changes in the EPR signal. This artifact occurs by virtue of both Heisenberg spin-spin interaction and the oxygen-enhanced spin-lattice relaxation, and should be corrected for. Reaction with Semiquinone Radicals Stable nitroxides react also with and detoxify semiquinones generated enzymatically. The experiment is carried out in anaerobic buffered phosphate, pH 7.4, containing 100 /~M DTPA, to minimize any interfering transition metal reactions, and 100 t~M TEMPOL. To this is added 1 mM NADPH and 2 nM cytochrome P-450 oxidoreductase (EC 1.6.2.4, NADPH-ferrihemoprotein reductase) to reductively activate 50/~M mitomycin C. 3° The reaction is performed anaerobically within the EPR cavity as detailed above, and the decrease in the TEMPOL EPR signal is followed as a function of time. The rate of nitroxide spin loss increases with the concentration of the drug, whereas in the absence of quinone the TEMPOL signal is not affected. This experiment demonstrates that nitroxides can react with and detoxify semiquinones within cells. Similar observations can be made when adriamycin, streptonigrin, or menadione are used. Intracellular Localization The distinction between extra- and intracellular localization of nitroxides is achievable using paramagnetic agents, such as chromium oxalate, which do not enter the cell and therefore broaden the lines of extracellular species. 3~ In a typical study, 2 x 107 V79 cells/ml were treated with 28R. J. Mehlhornand C. E. Swanson, Free Radical Res. Commun. 17, 157 (1992). 29E. E. Voest, F. E. van Fassen, A. B. van Asbeck, J. P. Neijt, and J. J. Marx, Biochim. Biophys. Acta 1136, 113 (1992). 30M. C. Krishna, W. DeGraff,S. Tamura, F. J. Gonzalez,A. Samuni,A. Russo, and J. B. Mitchell, Cancer Res. 51, 6622 (1991). 31C. S. Lai, W. Froncisz, and L. E. Hopwood,Biophys J. 52, 625 (1987).
[59]
NITROXIDES AS ANTIOXlDANTS
589
1 mM of various nitroxides. The EPR signal was monitored with and without 84 mM chromium oxalate, using 9.4 GHz, field modulation of 100 KHz, and 1 gauss modulation amplitude. In the presence of chromium oxalate, the EPR signal originating from extracellular nitroxide is broadened, whereas the intracellular species give rise to narrow resonance signals superimposed on the extracellular resonances. Knowing the volume fraction occupied by the' cells, the intensity of the narrow resonances can be directly correlated to the nitroxide intracellular concentration. Because the broadening increases with the intermolecular distances and chromium oxalate is negatively charged, the extent of broadening depends on the charge of the individual nitroxide. Cell Culture To study the protective effects of nitroxides against oxidative damage induced by various modalities, the clonogenic cell survival is a reliable end point. As part of our studies related to the protective effects of nitroxides on living cells, we employ a diverse group of cells ranging from prokaryotes to mammalian c e l l s . 17'22'27 Because of their rapid doubling time and high plating efficiency, we most routinely use Chinese hamster V79 lung fibroblasts, which are grown in sterile Ham's F 12 medium supplemented with 10% fetal calf serum with glutamine and without sodium bicarbonate (Hyclone Laboratories, Logan, UT), penicillin at 0.14 mg/ ml, and streptomycin at 0.2 mg/ml. Drug treatment or X-irradiation is performed in the presence or absence of varying concentrations (0.1-100 mM) of the particular nitroxide. In a typical experiment, 5 x 105 cells in 5 ml medium are plated into a 100-mm petri dish and incubated (95% air/ 5% CO 2, by volume) for 16 hr at 37°. Following cell adherence to the plates and exponential growth, 10 mM TEMPOL and 0-1.2 mM H202 are added. After 1 hr the cells are rinsed, treated with 0.03% trypsin, counted, and divided into dishes to be incubated for macroscopic colony formation. After 7 days the cells are fixed with methanol/acetic acid (3:1, v/v), stained with 0.3% crystal violet, rinsed, and air-dried, and the colonies containing over 50 cells are counted. Acknowledgments This research was partially supported by Grant 89-00124from the United States-Israel Binational Science Foundation (BSF), Jerusalem, Israel.
590
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
[60]
[60] Tamoxifen and Estrogens as Membrane Antioxidants: Comparison with Cholesterol
By
HELEN WISEMAN
Introduction T a m o x i f e n is a n a n t i c a n c e r d r u g that is w i d e l y u s e d in the t r e a t m e n t o f b r e a s t c a n c e r , with c o n s i d e r a b l e success~-4; it is n o w b e i n g a s s e s s e d in clinical trials as a p r o p h y l a c t i c a g e n t for this disease. 5-9 I n a d d i t i o n , t a m o x i f e n has a n u m b e r o f p o t e n t i a l h e a l t h b e n e f i t s such as p r o t e c t i o n a g a i n s t c a r d i o v a s c u l a r d i s e a s e s 1°-14 a n d a g a i n s t b o n e loss 15-17 a n d o s t e o a r thritis)8, ~9 T h e m o d e o f a c t i o n o f t a m o x i f e n w a s originally t h o u g h t to be solely t h r o u g h its a b i l i t y to a n t a g o n i z e the a c t i o n o f e s t r o g e n b y b i n d i n g to the I V. C. Jordan, J. Natl. Cancer Inst. 84, 231 (1992). 2 j. F. R. Robertson, I. O. Ellis, R. I. Nicholson, A. Robins, J. Beli, and R. W. Blamey, Breast Cancer Res. Treat. 20, 117 (1992). 3 D. Riley, M. Baum, J. Maclntyre, D. Berstock, A. McKinna, I. Jackson, J. R. C. Sainsbury, A. Wilson, T. Wheeler, and J. Dobie, Eur. J. Cancer 28A, 904 (1992). 4 M. Baum, J. Houghton, D. Riley, J. Maclntyre, D. Berstock, A. McKinna, I. Jackson, J. R. C. Sainsbury, A. Wilson, and T. Wheeler, Acta Oncol. 31, 251 (1992). 5 V. C. Jordan, Curr. Probl. Cancer 16, 129 (1992). 6 L. Bernstein, R. K. Ross, and B. E. Henderson, Am. J. Epidemiol. 135, 142 (1992). 7 C. Hamilton, in "Introducing New Treatments for Cancer Practical, Ethical and Legal Problems" (C. J. Williams, ed.), p. 315. Wiley, Chichester, 1992. 8 A. L. Jones and T. J. Powles, in "Introducing New Treatments for Cancer Practical, Ethical and Legal Problems" (C. J. Williams, ed.), p. 322. Wiley, Chichester, 1992. 9 R. R. Love, in "Introducing New Treatments for Cancer Practical, Ethical and Legal Problems" (C. J. Williams, ed.), p. 340. Wiley, Chichester, 1992. l0 D. V. Schapira, N. B. Kumar, and G. H. Lyman, Breast CancerRes. Treat. 17, 3 (1990). 11C. C. McDonald and H. J. Stewart, Br. Med. J. 303, 435 (1991). 12R. R. Love, D. A. Wiebe, R. A. Newcombe, L. Cameron, H. Leventhal, V. C. Jordan, J. Feyzi, and D. L. Demets, Ann. Intern. Med. 115, 860 (1991). 13j. A. Dewar, J. M. Horobin, P. E. Preece, R. Tavendale, H. Tunstallpedoe, and R. A. B. Wood, Br. J. Med. 305, 225 (1992). i4 R. R. Love, T. S. Surawicz, and E. C. Williams, Arch. Intern. Med. 152, 317 (1992). 15I. S. Fentiman, Z. Saad, M. Calefti, M. A. Chaudary, and I. Fogelman, Eur. J. Cancer 28, 684 (1992). 16R. R. Love, R. B. Mazess, H. B. Barden, S. Epstein, P. A. Newcombe, V. C. Jordan, P. P. Carbone, and D. L. Demet, N. Engl. J. Med. 326, 852 (1992). 17A. Goulding, E. Gold, and W. Fey, Bone Miner. 18, 143 (1992). 18C. L. Chander and F. M. Desa, Agents Actions 34, 282 (1991). I9 C. L. Tsai and T. K. Liu, Life Sci. 50, 1943 (1992).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
[60]
MEMBRANE ANTIOXIDANTS TAMOXIFEN AND ESTROGENS
591
estrogen receptor. However, the ability of tamoxifen to exert a growth inhibitory action in women with estrogen receptor-negative breast cancer has led to the recognition of a variety of other actions. These include an ability to inhibit the action of protein kinase C, 2°,2~ phospholipase A2 activity, 22 and calcium calmodulin-dependent cAMP phosphodiesterase. 23-25 Calmodulin binds tamoxifen in a calcium-dependent mannerZ6,27; however, tamoxifen inhibits the action of both calmodulin-dependent and calmodulin-independent Ca 2÷, Mg2+-ATPase] 8 Tamoxifen can also reduce plasma levels of insulin-like growth factor. 29'3° In addition, tamoxifen can induce cells surrounding the cancer cells to secrete a negative growth factor, namely, transforming growth factor/3. 31 The ability of tamoxifen to compete with estrogen for binding to the estrogen receptor is dependent on the similarity in the structural features of the two molecules. In addition, tamoxifen has been found to inhibit membrane lipid peroxidation, 32 and it is thought to act as a membrane antioxidant in a similar way to membrane cholesterol. 33 Furthermore, the estrogens estradiol, estriol, and estrone have been reported to inhibit lipid peroxidation in microsomal and liposomal systems. 34 This mechanism of membrane stabilization against lipid peroxidation by both cholesterol and tamoxifen is considered to be through a decrease in membrane fluidity. 35 20 E. Bignon, M. Pons, J. C. Dorc, J. Gilbert, T. Ojasoo, J. F. Miquel, and J. P. Raynaud, Biochem. Pharmacol. 42, 1373 (1991). 21 C. Borner and D. Fabbro, in Protein Kinase C: Current Concepts and Future Perspectives" (D. S. Lester and R. M. Epand, eds.), p. 297. Ellis Horwood, Chichester, 1992. 22 j. M. Fayard, S. Chanal, A. Fanidi, J. F. Pageaux, M. Lagarde, and C. Laugier, Eur. J. Pharmacol. 216, 127 (1992). 23 M. G. Rowlands, I. B. Parr, R. McCague, M. Jarman, and P. M. Goddard, Biochem. Pharmacol. 40, 283 (1990). 24 A. Fanidi, J. F. Pageaux, C. Courison, J. M. Fayard, and C. Laugier, Biol. Cell 72, 181 (1991). 25 A. Fanidi, C. Ahnadi, J. M. Fayard, J. F. Pageaux, and C. Laugier, J. Steroid Biochem. Mol. Biol. 41, 571 (1992). 26 M. Celeste, F. Lopes, M. Graca, P. Vale, and A. P. Carvalho, CancerRes. 50, 2753 (1990). 27 K. J. Edwards, C. A. Laughton, and S. Neidle, J. Med. Chem. 35, 2753 (1992). 28 j. O. Malva, M. C. F. Lopes, M. G. P. Vale, and A. P. Carvalho, Biochem. Pharmacol. 450, 1877 (1990). 29 M. N. Pollak, H. Huynh, and S. P. Lefebvre, Breast Cancer Res. Treat. 22, 91 (1992). 3o p. E. Lonning, K. Hall, A. Aakvaag, and E. A. Lien, Cancer Res. 52, 4719 (1992). 31 A. Butta, K. Maclennan, K. C. Flanders, N. P. M. Sacks, I. Smith, A. McKinna, M. Dowsett, L. M. Wakefield, M. B. Sporn, and M. Baum, Cancer Res. 52, 4261 (1992). 32 H. Wiseman, M. J. Laughton, H. R. V. Arnstein, M. Cannon, and B. Halliwell, FEBS Lett. 263, 192 (1990). 33 H. Wiseman, M. Cannon, H. R. V. Arnstein, and B. Halliwell, FEBS Lett. 274, 107 (1990). 34 K. Yagi and S. Komura, Biochem. Int. 13, 1051 (1986). 35 H. Wiseman, P. Quinn, and B. Halliwell, FEBS Lett. 330, 53 (1993).
592
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[60]
Tamoxifen inhibits yeast growth, 36 and its antioxidant action is shared by antifungal azoles such as ketoconazole. 37 The observed membrane antioxidant action of tamoxifen is now recognized, therefore, to relate to its diverse actions as an anticancer drug, an antifungal drug, 36 and as an anti-MDR (multidrug resistance) agent. 36,38 Cell membranes are thus an important target site for anticancer drugs, and the methods described here relate to the antioxidant action of tamoxifen and other compounds including the estrogen 17fl-estradiol. 32,33,37,39,4° Tamoxifen is metabolized in humans by cytochrome P-450 3A, 41 and the structures of the tamoxifen metabolites 4-hydroxytamoxifen and Ndesmethyltamoxifen 42 and the synthetic derivative 3-hydroxytamoxifen (droloxifene) are shown in Fig. 1.3-Hydroxytamoxifen has several advantages over tamoxifen including an increased ability to bind to the estrogen receptor, a shorter terminal elimination half-life, lower accumulation, and improved drug tolerability. 43-46We have expressed the molecular mimicry of cholesterol by tamoxifen, 4-hydroxytamoxifen, ergosterol, and 17/3estradiol, relative to cholesterol, in terms of a cholesterol coefficient (Cc) 47 and the increased antioxidant ability of the ergosterol-containing yeast membrane lipid fraction from tamoxifen-treated yeast in terms of a tamoxifen enhancement coefficient (TEC). 48 The relevance of the membrane antioxidant action of tamoxifen to its therapeutic use is discussed. Materials Tamoxifen, cholesterol, ergosterol, 17fl-estradiol, and ox brain phospholipids are from Sigma (Poole, Dorset, UK). 4-Hydroxytamoxifen, cis36 H. Wiseman, Trends Pharmacol. Sci. 15, 83 (1994). 37 H. Wiseman, C. Smith, H. R. V. Arnstein, B. Halliwell, and M. Cannon, Chem.-BioL Interact. 79, 229 (1991). 38 E. Berman, M. Adams, R. Duigou-Osterndorf, L. Godfrey, B. Clarkson, and M. Andreef, Blood 77, 818 (1991). 39 H. Wiseman, C. Smith, B. Halliwell, M. Cannon, H. R. V. Arnstein, and M. S. Lennard, Cancer Lett. 66, 61 (1992). 4o H. Wiseman, G. Paganga, C. Rice-Evans, and B. Halliwell, Biochem. J. 292, 635 (1993). 41 F. Jacolot, I. Simon, Y. Dreano, P. Beaune, C. Riche, and F. Berthou, Biochem. Pharmacol. 41, 1911 (1991). 42 E. A. Lien, E. Solheim, and P. M. Ueland, Cancer Res. 51, 4837 (1991). 43 U. Eppenberger, K. Wosikowski, and W. Kung, Am. J. Clin. Oncol. 14, $5 (1991). 44 H. J. Grill and K. Pollow, Am. J. Clin. Oncol. 14, $21 (1991). 45 I. Kawamura, T. Mizota, N. Kondo, K. Shimomura, and M. Kohsaka, Jpn. J. Pharmacol. 57, 215 (1991). 46 p. F. Bruning, Fur. J. Cancer 28A, 1404 (1992). 47 H. Wiseman, M. Cannon, H. R. V. Arnstein, and D. J. Barlow, Biochim. Biophys. Acta 1138, 197 (1992). H. Wiseman, M. Cannon, H. R. V. Arnstein, and B. Halliwell, Biochim. Biophys. Acta 1181, 201 (1993).
R2
(
T Ra
TAMOXIFEN RicH, R~OCH2CH2N(CHa)2,R~H 3-HYDROXYTAMOXIFEN RicH, R~OCH2CH2N(CHa)2,R~OH 4-HYDROXYTAMOXIFEN RI~-~-OH,R~---OCH2CH2N(CH3)2,R~H N-DESMETHYLTAMOXIFEN R~-H, R~OCH2CH2NHCHa, R~H
OH
17~ESTRADIOLR~=OH CH3 R~
Hi
ERGOSTEROLRI~-~)H,Rz=CH(CH3)CH:CHCH(CH3)CH(CH3)2 CHa R
CHOLESTEROLRI=OH,R2=CH(CH3)(CH2)3CH(CHa)2 Fl~. 1. Structures of tamoxifen, tamoxifen derivatives, 17fl-estradiol, ergosterol, and cholesterol, showing structural mimicry of the membrane sterols cholesterol and ergosterol by tamoxifen, related compounds, and 17fl-estradiol.
594
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[60]
tamoxifen, and N-demethyltamoxifen are from Zeneca Pharmaceuticals p.l.c. (Macclesfield, Cheshire, UK). Droloxifene (3-hydroxytamoxifen) is from Klinge Pharma (Munich, Germany). All other reagents are from BDH (Poole, Dorset, UK) and are of analar quality. Microbial growth media components are obtained from Difco (Detroit, MI). Methods
Membrane Antioxidant Action of Tamoxifen and 17fl-Estradiol in Liposomal and Microsomal Systems Microsomes are prepared from homogenates of liver from male Wistar rats by differential centrifugation techniques as described previously. 49 The microsomes are washed three times with phosphate buffer (20 mM KH2PO4-KOH, pH 7.4) and then resuspended in this buffer. The protein content of the microsomes is determined by the Lowry method, 5° and the microsomes are then stored frozen at - 2 0 ° for periods not exceeding 3 weeks. Ox brain phospholipid liposomes are prepared as described p r e v i o u s l y . 33'37'39 The phospholipid is added to phosphate-buffered saline, pH 7.4 (PBS: 140 mM NaC1, 2.7 mM KCI, 16 mM Na2HPO4, 2.9 mM KH2PO4) at a final concentration of 5 mg/ml followed by sonication in a bath sonicator at 0 ° and vortexing, both in the presence of glass beads (2.5-3.5 mm in diameter). The resulting liposomes are left in sealed, nitrogen-flushed bottles at 4 ° for I hr before use. The reaction mixtures (final volume of 1.0 ml) contains either microsomes (0.25 mg of microsomal protein) or liposomes (0.5 mg in 0. I ml of phosphate-buffered saline at pH 7.4); 0.5 ml of phosphate buffer, pH 7.4 is used for microsomal assays, and phosphate-buffered saline, pH 7.4, is used for liposomal assays. Five microliters of ethanol or test compound dissolved in ethanol is added. Freshly prepared aqueous solutions ofFeC13 (0.1 ml) and ascorbate (0.1 ml) are added to give a final concentration of 100 t~M of each. In some experiments on microsomal lipid peroxidation FeCI 3 (I00 tzM), ADP (1.7 mM), and NADPH (0.4 mM) are added to give the final concentrations stated. Freshly prepared ADP and FeC13 are premixed just before addition to the reaction mixture. Ascorbate or NADPH is added to start the reaction, and incubations are carried out for 20 min at 37°. The extent of lipid peroxidation is determined by the ~ G . J . Quinlan, B. Halliwell, C. P. Moorhouse, and J . M . C . Gutteridge, Biochim. Biophys. Acta 962, 196 (1988). 50 O. H. Lowry, N. J. Rosebrough, A. L. Farr, and R. J. Randall, J. Biol. Chem. 193, 265 (1951).
[60]
M E M B R A N E A N T I O X I D A N T S T A M O X I F E N A N D ESTROGENS
595
thiobarbituric acid (TBA) test. HCI [0.5 ml, 25% (v/v)] is added to each reaction mixture, followed by 0.5 ml of thiobarbituric acid [1% (w/v) in 50 mM sodium hydroxide] and heating for 30 min at 80 °. The chromogen is extracted with 2 ml of 1-butanol, and the A532 of the upper (organic) layer is measured. Tamoxifen, its derivatives, and 17/3-estradiol all inhibit lipid peroxidation in both microsomal and preformed liposomal systems. 32,37,39The test compounds are added to these systems dissolved in ethanol. Ethanol itself has no effect on peroxidation at the concentration used. The IC50 values of the compounds are shown in Tables I and II and indicate that 4-hydroxytamoxifen is a better inhibitor of microsomal lipid peroxidation in both the Fe(III)-ascorbate and Fe(III)-ADP/NADPH systems, and of liposomal peroxidation, than tamoxifen, 3-hydroxytamoxifen, and 17/3-estradiol. Cholesterol and ergosterol have no inhibitory effect on either microsomal or liposomal peroxidation.Tamoxifen and related compounds and 17/3estradiol (all at the IC50 concentrations) inhibit microsomal and liposomal lipid peroxidation by an approximately constant percentage over the time course investigated. Furthermore, there is no evidence for a "lag period" followed by an acceleration of peroxidation to the control rate that is often observed when chain-breaking antioxidants are added to a peroxidiz-
TABLE I INHIBITION OF MICROSOMAL LIPID PEROXIDATION BY TAMOXIFEN, TAMOXIFEN DERIVATIVES, AND 17fl-ESTRADIOL IC50 (/zM)
Compound/drug
Fe(III)-ascorbate system
Fe(III)-ADP/NADPH
system
Tamoxifen 4-Hydroxytamoxifen 3-Hydroxytamoxifen citrate Tamoxifen citrate 17fl-Estradiol
11 a
18 a
3a 13 b 17 b
3~ 11 a
11 b
9b
cis-Tamoxifen
16 C
23 c
18 c N . R . a'b'd N .R. a'd
25 ¢ N .R. a,b'd N .R. a'd
N-Desmethyltamoxifen Cholesterol Ergosterol a Wiseman e t al. 37 b Wiseman e t al. 39 c Wiseman e t al. 32 d N . R . , N o t reached.
19 b
596
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[60]
TABLE II INHIBITION OF PREFORMED LIPOSOMAL LIPID PEROXIDATION BY TAMOXIFEN, TAMOXIFEN DERIVATIVES, AND 17fl-ESTRADIOL
Compound/drug Tamoxifen 4-Hydroxytamoxifen 3-Hydroxytamoxifen citrate Tamoxifen citrate 17fl-Estradiol Cholesterol Ergosterol
IC5o values for preformed liposomal system (p~M) 630 9a 38 b
62b 11b N.R: -c N.R: .C
a Wiseman et al. 37 b Wiseman et al. 39 ¢ N.R., Not reached.
ing lipid s u b s t r a t e : ~ N o n e of the compounds tested interfere with the T B A test. N o inhibition is observed when the compounds are added to peroxidizing microsomes or liposomes at the same time as the TBA reagents instead of at the beginning of the incubation. Similar or enhanced inhibitory effects of these compounds on lipid peroxidation are obtained when B H T (butylated hydroxytoluene) is added at the end of the incubations to suppress peroxidation during the TBA test itself. M e m b r a n e Antioxidant Action o f Tamoxifen and 1 7fl-Estradiol: Comparison with Cholesterol
Using the preformed liposomal system described above, added cholesterol does not inhibit lipid peroxidation and thus has no antioxidant action. When introduced into liposomes during their preparation, however, cholesterol can exert a membrane antioxidant effect. Liposomes are prepared with and without the introduction of the compounds shown in Fig. 1, as described previously for cholesterol. 33,37,39,5zOx brain phospholipid is dissolved in chloroform to give a final concentration of 5 mg/ml, and a 0.8-ml aliquot is taken and mixed with an equal volume of cholesterol, ergosterol, tamoxifen, or 4-hydroxytamoxifen taken from stock solutions in chloroform or, in the case of 17fl-estradiol in m e t h a n o l - c h l o r o f o r m (1 : 2, v/v), to provide the concentration range required for each of the 51B. Halliwell, Free Radical Res. Commun. 9, 1 (1990). 52j. M. C. Gutteridge, Res. Commun. Chem. Pathol. Pharmacol. 22, 563 (1978).
[60]
MEMBRANE ANTIOXIDANTS TAMOXIFEN AND ESTROGENS
597
compounds. The organic solvents are then completely evaporated in a stream of nitrogen. Liposomes are prepared by the addition of 0.8 ml of phosphate-buffered saline at pH 7.4 followed by the sonication and vortexing procedure as described above. Reaction mixtures (final volume 1.0 ml) contain liposomes (0.5 mg in 0.1 ml of phosphate-buffered saline), phosphate-buffered saline, pH 7.4 (0.5 ml), and water (0.2 ml). Peroxidation using the Fe(III)/ascorbate system is carried out as described above, and incubations are allowed for 20 min at 37°. Membrane lipid peroxidation is measured by the TBA assay (see above). When introduced into the liposomal system 33'39tamoxifen inhibits lipid peroxidation to approximately the same extent as ergosterol. Cholesterol is a much less effective inhibitor than tamoxifen, whereas 4-hydroxytamoxifen and 17fl-estradiol are approximately equipotent. The latter two compounds are both more effective inhibitors of peroxidation than tamoxifen. This inhibitory ability can be expressed as the IC50 values derived for the compounds tested (see Table III). The results show that the concentration of tamoxifen or ergosterol that must be introduced into the membrane to achieve 50% inhibition of lipid peroxidation is approximately 15-fold less than that for cholesterol. Moreover, the IC50 values of 4hydroxytamoxifen and 17/3-estradiol are approximately 200-fold less than for cholesterol. The cholesterol coefficient (Cc) values (see Table IV) are calculated from the slopes of the approximately linear portions, at the low compound to phospholipid ratios, of the graphs that derive from the relationship between the ratio of percent inhibition of lipid peroxidation by the compound and percent inhibition of lipid peroxidation by cholesterol and the TABLE III INHIBITION OF LIPID PEROXIDATION BY TAMOX1FEN, TAMOXIFEN DERIVATIVES, AND 17/~-ESTRADIOL INTRODUCED INTO LIPOSOMES DURING LIPOSOMAL PREPARATION IC50 values introduced into liposomal system
(mM) Compound/drug
In liposome a
In reaction mixture
Cholesterol Tamoxifen Ergosterol 4-Hydroxytamoxifen 17/3-Estradiol
7.2 0.50 0.53 0.036 0.040
0.72 0.050 0.053 0.0036 0.0040
Wiseman e t al. 33
598
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[60]
TABLE IV CHOLESTEROL COEFFICIENTS FOR TAMOXIFEN, 4-HYDROXYTAMOXIFEN, 17fl-ESTRADIOL, ERGOSTEROL, AND CHOLESTEROL
Drug/compound
Cholesterol coefficient" (Co)
4-Hydroxytamoxifen Tamoxifen 17fl-Estradiol Cholesterol Ergosterol
147 29 127 0 32
a Wiseman
e t al. 47
compound to phospholipid ratio in ox brain phospholipid liposomes. 47 It is found that at high compound to phospholipid ratios tamoxifen, 4hydroxytamoxifen, and ergosterol are all very similar to cholesterol, whereas at low compound to phospholipid ratios tamoxifen and ergosterol are divergent from cholesterol, and 4-hydroxytamoxifen and 17fl-estradiol are highly divergent from cholesterol. The cholesterol coefficients show that both tamoxifen and ergosterol are not just similar to cholesterol in terms of inhibition of lipid peroxidation, but indeed are superior to it. This superiority is increased by approximately 5-fold with 4-hydroxytamoxifen and 17fl-estradiol.
Enhancement of Antioxidant Ability of Membrane Lipid Fraction Derived from Tamoxifen-Treated Yeast Saccharomyces cerevisiae (strain A224A) is grown in l-liter batches of culture in YEPD medium (10 g/liter yeast extract, 20 g/liter Bactopeptone, and 20 g/liter glucose) at 30° in an orbital shaker (240 r p m ) . 36'48 Cultures are treated with tamoxifen to give a final concentration of 30/.~M [at a final concentration of ethanol of 0.5% (v/v)] or ethanol added to give a final concentration of 0.5% (v/v). Tamoxifen (30/zM) inhibits growth by 80 + 2% (mean -+ standard deviation, where n = 8-12 tests) when growth inhibition is measured after 1.5 generation times of the control culture containing 0.5% (v/v) ethanol. 48 The cells are harvested at 8000 rpm for 15 min at 4°, and the cell pellets are resuspended in phosphate buffer (20 mM KHEPO4-KOH, pH 7.4) and then recentrifuged. This procedure is repeated twice. The cell pellets are disrupted using 4 passes of a French type press (at -15,000 psi) (APV, London, U.K.). Yeast microsomes are prepared from the disrupted cells by differential centrifugation techniques. 4s The microsomal protein content is determined by the Lowry procedure. 5°
[60]
MEMBRANE ANTIOXlDANTS TAMOXIFEN AND ESTROGENS
599
A lipid fraction is obtained from yeast microsomes (from untreated and tamoxifen-treated yeast cultures) by extraction with chloroform-methanol (2 : 1, v/v). 53 The microsomes are mixed with an equal volume of chloroform-methanol (2 : 1, v/v) and then centrifuged for 10 min at speed number 7 in a bench-top centrifuge at 20°. The lower organic layer is removed, and 1 ml of water is added. The tubes are vortexed and centrifuged as before, and the lower organic solvent layer is collected. This purification procedure is repeated twice. The solvent is evaporated under a nitrogen stream and the residue redissolved in chloroform and stored at - 7 0 °. Ox brain phospholipid is dissolved in chloroform to give a final concentration of 5 mg/ml. A 0.8-ml aliquot is taken and mixed with an equal volume of yeast lipid fraction from untreated or tamoxifen-treated yeast from stock solutions in chloroform to give the required amounts. 48 The organic solvents are then evaporated, in a stream of nitrogen. Liposomes are prepared by addition of 0.8 ml of phosphate-buffered saline at pH 7.4, followed by sonication and vortexing, as described above. The reaction mixture contains, in a final volume of 1 ml, liposomes (0.1 ml), phosphatebuffered saline, pH 7.4 (0.5 ml), and water (0.2 ml). Peroxidation using the Fe(III)/ascorbate system is carried out as described above, and incubations are allowed to proceed for 20 rain at 37°. Membrane lipid peroxidation is measured by the TBA assay (see above). The lipid fraction from untreated yeast is found to inhibit iron-dependent liposomal lipid peroxidation in a concentration-dependent manner when introduced into ox brain phospholipid liposomes during their preparation. However, the yeast lipid fraction from yeast that has been treated with tamoxifen again exerts a concentration-dependent inhibitory effect on lipid peroxidation but is more effective at inhibiting liposomal lipid peroxidation than that from untreated yeast. 48 When ratios are calculated for the inhibition of lipid peroxidation by tamoxifen-treated yeast lipid fraction relative to inhibition of lipid peroxidation by untreated yeast lipid fraction, and plotted against amounts of microsomal protein added, the slope of the linear portion of the graph has a value of 2.0. This value is designated the tamoxifen enhancement coefficient (TEC). 48
Discussion Application of the methods described above enabled the membrane antioxidant action of tamoxifen, tamoxifen derivatives, and 17/3-estradiol to be examined. These procedures also enabled calculation of two original 53 j. Folch, M. Lees, and G. H. Sloane-Stanley, J. Biol. Chem. 226, 497 (1957).
600
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[60]
parameters, namely, the cholesterol coefficient (Co) and the tamoxifen enhancement coefficient (TEC). The chemical structure of tamoxifen (see Fig. 1) indicates that it is unlikely to act as a chain-breaking antioxidant because it does not possess easily donatable hydrogen atoms. Furthermore, chain-breaking antioxidants usually introduce a lag period into the peroxidation time course. 51 This behavior was not observed for tamoxifen or for 4-hydroxytamoxifen, 3-hydroxytamoxifen, or 17/3-estradiol in either the microsomal or liposomal systems, even though 4-hydroxytamoxifen, 3-hydroxytamoxifen, and 17fl-estradiol each possess a potentially donatable hydrogen atom. The comparable effects of the compounds in the two microsomal systems (see Table I) also suggests that free radical intermediates are not being formed from the drugs because ascorbate might have re-reduced such intermediates, thus lowering the IC50 values observed in this system compared to the Fe(III)-ADP/NADPH system. The RBA (relative binding affinity) for the estrogen receptor of 4hydroxytamoxifen is similar to that of 17/3-estradiol and is 100-fold higher than that of tamoxifen, 54 whereas that of 3-hydroxytamoxifen is 10-fold higher than that of tamoxifen. 55 This is indicative of the importance of the ring position of the hydroxyl group on the tamoxifen molecule in relation to the structural mimicry of cholesterol and 17/3-estradiol that we have examined using computer molecular modeling techniques. 47 These differences in binding affinity are reflected in the chemical structures; thus, although tamoxifen, 4-hydroxytamoxifen, and 3-hydroxytamoxifen mimic the steroid nucleus of 17/3-estradiol, only 4-hydroxytamoxifen has a hydroxyl group in the same ring position as 17fl-estradiol. The greater binding affinity reported for 4-hydroxytamoxifen, 3-hydroxytamoxifen, and 17fl-estradiol compared to tamoxifen is reflected in their ability to inhibit lipid peroxidation more effectively than tamoxifen. This is reflected in the cholesterol coefficients of tamoxifen, 4-hydroxytamoxifen, and 17/3-estradiol. The demonstration that tamoxifen and 17/3-estradiol structurally mimic cholesterol, as indicated by the computer molecular modeling results and by the cholesterol coefficient values, suggests that these molecules might also mimic the natural membrane stabilizing actions of cholesterol itself, observed as inhibition of lipid peroxidation. Cholesterol is proposed to act in this way via interaction between its hydrophobic rings and the saturated, monosaturated, and polyunsaturated residues of phospholipid 54 B. J. A. Furr and V. C. Jordan, Pharmacol. Ther. 25, 127 (1984). 55 R. Loser, K. Seibel, H. D. Liehn, and H. J. Staab, Contrib. Oncol. 23, 64 (1986).
[60]
MEMBRANE ANTIOXIDANTS TAMOXIFEN AND ESTROGENS
601
f a t t y a c i d s . 33'35,36'37'39'47'48'52 Direct
evidence for the ability of tamoxifen and also 17fl-estradiol to decrease membrane fluidity in liposomes 35 and human breast cancer cells 56has been provided from steady-state polarization of fluorescence measurements with cells treated with tamoxifen or 17fl-estradiol. The membrane stabilizing action of tamoxifen through decreased membrane fluidity is likely to antagonize cell division in both cancer cells and yeast cells, and this effect is probably mediated through a second messenger system. 39'47 In addition, the reversal of MDR by tamoxifen could be achieved by inhibition of the active efflux pump action of the plasma membrane P-170 glycoprotein as a result of decreased membrane fluidity. 36'56It is of interest that 17fl-estradiol at high concentrations inhibits the growth of both the estrogen receptor-positive MCF-7 breast cancer cell line and the estrogen receptor-negative MDA-MB-436 line, 56 probably as a result of its membrane antioxidant action. However, it is unlikely that 17/3-estradiol could be useful as an anticancer agent because at physiological concentrations it stimulates the growth of estrogen receptor-positive cells. 57,58 17/3-Estradiol has been reported to stabilize human LDL (low density lipoprotein) against oxidative damage, 59 and similar findings have been reported for tamoxifen and 4-hydroxytamoxifen.4° It has been reported that microsomal membranes of Novikoff and Yoshida hepatoma cells have decreased levels of polyunsaturated fatty acids 6°-62 and thus decreased membrane fluidity and increased membrane stability against lipid peroxidation. In addition, plasma membranes from Morris hepatoma cells have a higher cholesterol to phospholipid ratio and, consequently, a decreased membrane fluidity compared with normal cells. 63 If these membrane changes are characteristic of cancer cells in general compared with normal cells, then the growth of cancer cells could be more easily inhibited by the membrane-modifying action of tamoxifen 56 R. Clarke, H. W. van den Berg, and R. F. Murphy, J. Natl. Cancerlnst. 82, 1702 (1990). 57 y . Furuya, N. Kohno, Y. Fujiwara, and Y. Saitoh, Cancer Res. 49, 6670 (1989). 58 R. L. Sutherland, C. S. L. Lee, R. S. Feldman, and E. A. Musgrove, J. Steroid Biochem. Mol. Biol. 41, 315 (1992). 59 L. A. Huber, E. Schemer, T. Poll, R. Ziegler, and H. A. Dresel, Free Radical Res. Commun. 8, 167 (1990). 6o K. H. Cheeseman, G. W. Burton, K. U. Ingold, and T. F. Slater, Toxicol. Pathol. 12, 236 0984). 61 K. H. Cheeseman, M. Collins, K. Proudfoot, T. F. Slater, G. W. Burton, A. C. Webb, and K. U. Ingold, Biochem. J. 235, 507 0986). 6~ K. H. Cheeseman, S. Emery, S. P. Maddix, T. F. Slater, G. W. Burton, and K. U. Ingold, Biochem. J. 250, 247 (1988). 63 T. Galeotti, S. BorreUo, G. Paiombini, L. Masotti, M. B. Ferrari, P. Cavatorta, A. Arcioni, C. Stremmenos, and C. Zannoni, FEBS Lett. 164, 169 (1984).
602
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[60]
than that of normal cells. Metastatic tumor cells have a higher plasma membrane fluidity than nonmetastatic cells64; therefore, decreased membrane fluidity may restore contact inhibition between such cells by increasing the rigidity of the cells at cell-junction contacts. This may oppose the metastasis of invasive epithelial malignancies of the breast that is aided by the generation of stromelysin-3 in the surrounding stromal cells. 65 The membrane antioxidant action of ergosterol was demonstrated by its ability to inhibit lipid peroxidation when introduced into phospholipid liposomes and by the ability of the yeast lipid fraction (containing ergosterol) to inhibit lipid peroxidation when introduced into phospholipid liposomes. Furthermore, the lipid fraction from yeast treated with tamoxifen was more effective at inhibiting lipid peroxidation than that from untreated yeast. The TEC value of 2.0 thus quantifies the effectiveness of tamoxifen in enhancing the membrane antioxidant action of the ergosterol-containing yeast lipid fraction. The mechanism of the anti-yeast ability of tamoxifen, which has potential clinical applications in the treatment of Candida infections,36 may involve a further decrease in membrane fluidity that is harmful to the yeast cells. This was indicated by increased membrane stabilization against lipid peroxidation by tamoxifen beyond that achieved by ergosterol and may be the cause of the observed inhibition of yeast cell growth. 36 The concept of the cell membrane as a target for cancer therapy is poorly developed, 36'66 and this conceptual approach has been extended here through consideration of the effect of membrane antioxidant action. The membrane antioxidant action of tamoxifen via decreased membrane fluidity has important implications for its use in the prevention and treatment of cancer, its potential use in the reversal of MDR, and in the treatment of fungal infections, and also in the other possible beneficial health effects of tamoxifen. 36
64 G. Tarboletti, L. Perin, and B. Bottazzi, Int. J. Cancer 44, 707 (1989). 65 p. Basset, J. P. Bellocq, C. Wolf, I. Stoll, P. Hutin, P. Limacher, O. L. Podhajer, M. P. Chenard, M. C. Rio, and P. Chambon, Nature (London) 348, 699 (1990). 66 H. Grunicke, Eur. J. Cancer 27, 281 (1991).
[61]
PLASMALOGENS
603
[61] Reactivity of Plasmalogens to Singlet Oxygen and Radicals By OLIVIERH. MORAND Plasmalogens P l a s m a l o g e n s a r e p h o s p h o l i p i d s b e a r i n g a v i n y l e t h e r s u b s t i t u e n t at the s n - 1 p o s i t i o n o f the g l y c e r o l b a c k b o n e . 1-3 T h e y a r e f o u n d to d i f f e r e n t e x t e n t s in m o s t m a m m a l i a n cells a n d t i s s u e s , p a r t i c u l a r l y in h u m a n b r a i n 4 a n d h e a r t 5'6 w h e r e t h e y r e p r e s e n t 2 0 - 3 0 % o f t h e p h o s p h o l i p i d s . Liver and lipoprotein phospholipids have comparatively lower levels o f p l a s m a l o g e n s . 7-9 T h e b i o s y n t h e s i s o f p l a s m a l o g e n s is p e c u l i a r in t h e s e n s e t h a t t h e first t w o e n z y m e s in the c a s c a d e , n a m e l y , d i h y d r o x y a c e t o n e p h o s p h a t e ( D H A P ) a c y l t r a n s f e r a s e ( E C 2.3.1.42, g l y c e r o n e - p h o s p h a t e a c y l t r a n s f e r a s e ) a n d a l k y l - D H A P s y n t h a s e ( E C 2.5.1.26, a l k y l g l y c e r o n e - p h o s p h a t e s y n t h a s e ) , a r e b o t h l o c a l i z e d in p e r o x i s o m a l m e m b r a n e s , ~°m w h e r e a s t h e o t h e r r e a c t i o n s a r e c a t a l y z e d b y e n z y m e s o f t h e e n d o p l a s m i c r e t i c u l u m . C o n s e q u e n t l y , p l a s m a l o g e n b i o s y n t h e s i s is d e f e c t i v e in a n u m b e r o f h u m a n p e r o x i s o m a l g e n e t i c d i s o r d e r s ) 2,~3 C u l t u r e d h u m a n cells d e r i v e d f r o m t h e s e d i s o r d e r s 14,~5 a n d a n i m a l cell I L. A. Horrocks and M. Sharma, in "Phospholipids" (J. N. Hawthorne and G. B. Ansell, eds.), p. 51. Elsevier, Amsterdam, 1982. 2 F. Snyder, in "Biochemistry of Lipids and Membranes" (D. E. Vance and J. E. Vance, eds.), p. 271, Benjamin-Cummings, Menlo Park, California, 1985. 3 T. Cursted, in "Ether Lipids, Biochemical and Biomedical Aspects" (H. K. Mangold and F. Paltauf, eds.), p. 1. Academic Press, New York, 1983. 4 R. W. Gross, Biochemistry 23, 158 (1984). 5 K. Owens, Biochem. J. 100, 354 (1966). 6 T. W. Scott, B. P. Setchell, and J. M. Bassett, Biochem. J. 104, 1040 (1967). 7 L. A. Horrocks, in "Ether Lipids: Chemistry and Biology" (F. Snyder, ed.), p. 177. Academic Press, New York, 1972. 8 V. P. Skipsky, in "Blood Lipids and Lipoproteins: Quantitation, Composition and Metabolism" (G. J. Nelson, ed.), p. 471. Wiley (Interscience), New York, 1972. 9 j. E. Vance, Biochim. Biophys. Acta 1045, 128 (1990). 10C. L. Jones and A. K. Hajra, Biochem. Biophys. Res. Commun. 76, 1138 (1977). tl A. K. Hajra and J. E. Bishop, Ann. N.Y. Acad. Sci. 386, 170 (1982). 12p. B. Lazarow and Y. Fujiki, Annu. Rev. Cell Biol. 1, 489 (1985). 13 H. W. Moser, Dev. Neurosci. 9, 1 (1987). 14H. S. A. Heymans, H. Van den Bosch, R. B. H. Schutgens, W. H. H. Tegelaers, J. U. Walther, J. Mfiller-Hiicker, and P. Borst, Eur. J. Pediatr. 142, 10 (1984). 15G. Schrakamp, R. B. H. Schutgens, R. J. A. Wanders, H. S. A. Heymans, J. M. Tager, and H. Van den Bosch, Biochim. Biophys. Acta 833, 170 (1985).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
604
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[61]
mutants 16-19 with primary lesions in peroxisomal functions proved to be very useful in investigating the biogenesis of peroxisomes and the function(s) of peroxisome-related metabolites. In particular, Chinese hamster ovary (CHO) cell mutants ~6were isolated that have peroxisomal DHAP acyltransferase and alkyl-DHAP synthase levels reduced 50and 5-fold, respectively, as compared to wild-type cells. These mutants exhibit a 10-fold reduction of plasmalogen content. Antioxidant Function for Plasmalogens The hypothesis that plasmalogens might have an antioxidant function originates from data showing that human2° and animal21 cell mutants deficient in plasmalogens are more susceptible to photosensitized oxidation than plasmalogen-rich wild-type cells. The peroxisomal defect in plasmalogen synthesis can be bypassed by supplementing the medium of the mutant cells with 1-O-hexadecyl-sn-glycerol~'Z2; it will restore not only plasmalogen levels but also nearly normal resistance to photosensitized oxidation, without restoring normal peroxisomes, z~ The correlate of these observations is that plasmalogens are very sensitive to activated oxygen species (e.g., singlet oxygen and radicals) and decompose.23-28 Plasmalogen breakdown products such as lysophospholipids, formic acid, and long-chain fatty aldehydes can be identified in wild-type CHO-K1 cells, but not in plasmalogen-deficient CHO cell mutants. 27 These products were found in the mutants only after restoration of normal plasmalogen levels following incubation with 1-O-hexadecyl-sn-glycerol.z7The protective effect ofplasmalogens against photooxidized killing of cells, together with the high
16 R. A. Zoeller and C. R. H, Raetz, Proc. Natl. Acad. Sci. U.S.A. 83, 5170 (1986). 17 T. Tsukamoto, S. Yokota, and Y. Fujiki, J. Cell Biol. 110, 651 (1990). 18 O. H. Morand, L. A. Allen, R. A. Zoeller, and C. R. H. Raetz, Biochim. Biophys. Acta 1034, 132 (1990). 19 R. A. Zoeller, S. Rangaswamy, H. Herscovitz, W. B. Rizzo, A. K. Hajra, A. K. Das, I-I. W. Moser, A. Moser, P. B. Lazarow, and M. J. Santos, J. Biol. Chem. 267, 8299 (1992). 20 G. Hoefler, E. Paschke, S. Hoefler, A. B. Moser, and H. W. Moser, J. Clin. Invest. 88, 1873 (1991). 2J R. A. Zoeller, O. H. Morand, and C. R. H. Zoeller, J. Biol. Chem. 263, 11590 (1988). 22 G. Schrakamp, C. G. Schalkwijk, R. B. H. Schutgens, R. J. A. Wanders, J. M. Tager, and H. Van den Bosch, J. Lipid Res. 29, 325 (1988). 23 E. Yavin and S. Gatt, Eur. J. Biochem. 25, 431 (1972). 24 E. Yavin and S. Gatt, Eur. J. Biochem. 25, 437 (1972). 25 W. N. Mariner, E. Nungesser, and T. A. Foglia, Lipids 21, 648 (1986). 26 T. A. Foglia, E. Nungesser, and W. N. Manner, Lipids 23, 430 (1988). 27 O. H. Morand, R. A. Zoeller, and C. R. H. Raetz, J. Biol. Chem. 263, 11597 (1988). 28 V. C. Anderson and D, H. Thompson, Biochim. Biophys. Acta 1109, 33 (1992).
[61]
PLASMALOGENS
605
reactivity of the vinyl ether moiety to activated oxygen species, is the basis for a scavengerlike antioxidant function.
Decomposition of Plasmalogens by Photosensitized Oxidation It is necessary to discuss briefly the chemical reactions possibly involved in the decomposition of plasmalogens induced by photosensitized oxidation. The exact composition of the different reactive oxygen species produced by photooxidation depends on the sensitizer and on experimental conditions. At least two types of chemistry can account for the decomposition of plasmalogens into lysophospholipids, formic acid, and longchain fatty aldehydesfl7 In type II chemistry, photosensitization in the presence of molecular oxygen generates singlet oxygen (oxygen in the lag state, 102), which in turn can react with cell membrane lipids. 10 2 can react with vinyl ether moieties 29-31 in general, and plasmalogens 27 in particular (e.g., plasmenylethanolamine, 1), to form a dioxetane intermediate (2) (Fig. 1). This unstable intermediate will decompose in the presence of water to 2-monoacylglycerophosphoethanolamine (3), formic acid (4), and pentadecanal (5). In type I chemistry,Z9.3° photosensitization of plasmalogens can also generate the allylic 1'-hydroperoxide intermediate 6 by an " e n e " reaction or by radical-mediated oxidation (Fig. 2), similarly to the photosensitized oxidation of 1-methoxycyclohexene.32The allylic l'-hydroperoxide 6 can form an alkoxyl radical derivative 7 by metal-assisted cleavage of the hydroperoxide. The alkoxyl radical derivative 7, can then generate diacylglycerophospholipid 8, A2-hexadecenal 9, and lysophospholipid 10, similarly to the decomposition of 10-hydroperoxyoctadeca-8-enoic a c i d . 33,34 This chapter describes several techniques to characterize and quantify plasmalogen decomposition in cells exposed to oxidative stress27: (1) photosensitized oxidation of cultured cells; (2) plasmalogen breakdown in [32p]Pi-labeled cells; (3) plasmenylethanolamine breakdown in [2-14C]ethanolamine-labeled cells; (4) formation of radioactive formic acid in [1-14C]hexadecanol-labeled cells; (5) formation of radioactive fatty aldehyde in [U-14C]hexadecanol-labeled cells; and (6) high-performance liquid 29 C. S. Foote, in "Free Radicals in Biology" (W. A. Pryor, ed.), Vol. 2, p. 85. Academic Press, New York, 1976. 30 A. A. Frimer, in "The Chemistry of Peroxides" (S. Patai, ed.), p. 201. Wiley, New York, 1983. 31 A. P. Schaap and K. A. Zaklika, in "Singlet Oxygen" (H. H. Wasserman and R. W. Murray, eds.), p. 173. Academic Press, New York, 1979. 32 p. D. Bartlett and A. A. Frimer, Heterocycles 11, 419 (1978). 33 R. Labeque and L. J. Marnett, Biochemistry 27, 7060 (1988). 34 R. Labeque and L. J. Marnett, Biochemistry 27, 7846 (1988).
606
[61]
ANTIOXIDANT CHARACTERIZATION AND ASSAY
O~pSO(CH2)2NH3 +
O..ptO(CH2)2NH3+
O.~RtO(CH2)2NH3+
-o" ~o .O~. ~0
-o" 'o
-o- 'o
> >
,o.....o 102
oO-
O~ O(CH2)2NH3+ .O'P\O
t.o.o
~.o. o A
o=~°
H20
H
IL
H H
O-
_4
H
H O= /
>
>
!
o
> > > >
> > > / >
FIG. 1. Decomposition of plasmenylethanolamine (1-alk-l'-enyl-2-acyl-sn-glycero-3phosphoethanolamine): formation of a dioxetane intermediate with singlet oxygen (type II chemistry 29'3°) and breakdown products. 1, Plasmenylethanolamine; 2, dioxetane intermediate; 3, 2-monoacylglycerophosphoethanolamine; 4, formic acid; 5, pentadecanal.
chromatography (HPLC) quantification and identification of fatty aldehyde moieties after dinitrophenylhydrazine (DNP) hydrazine derivatization.
Methods Method 1: Photosensit&ed Oxidation of Cultured Cells Principle. The method uses 12-(1'-pyrene)dodecanoic acid, a fluorescent fatty acid analog and also a photosensitizer. 21'27'35'36which is covalently incorporated into phospholipids and neutral lipids of cultured cells.21,37-39 Excitation of the pyrene moiety with long-wavelength UV light 35 S. T. Mosley, J. L. Goldstein, M. S. Brown, J. R. Falck, and R. G. W. Anderson, Proc. Natl. Acad. Sci. U.S.A. 78, 5717 (1981). 36 E. Fibach, O. Morand, and S. Gatt, J. Cell Sci. 85, 149 (1986). 37 O. H. Morand, E. Fibach, A. Dagan, and S. Gatt, Biochim. Biophys. Acta 711, 539 (1982). 38 j. Radom, R. Salvayre, T. Levade, and L. Douste-Blazy, Biochem. J. 269, 107 (1990). 39 j. Kasurinen and P. Somerharju, J. Biol. Chem. 267, 6563 (1992).
[61]
PLASMALOGENS
607 R'
'o
0% )(CH2)2NH3+
y_o_
.o.../o
R\
102
Oxidation
(
~. H
H
or
Radical
6
R"
> > > X
7 R"~ H-Abstraction
i
H k~ H
%. 10
H
R"
FIG. 2. Decomposition of plasmenylethanolamine: formation of allylic hydroperoxide and alkoxyl radical intermediates (type I chemistry29,3°) and breakdown products. R' is the remainder of the phospholipid moiety, and R" the remainder of the alkyl chain. 1, Plasmenylethanolamine; 6, allylic hydroperoxide; 7, alkoxyl radical; 8, diacylglycerophosphoethanolamine; 9, AZ-hexadecenal; 10, 2-monoacylglycerophosphoethanolamine. The allylic 1'-hydroperoxide intermediate 6 can be also converted to products 9 and 10 by a peroxidase.
(>300 nm) generates reactive oxygen species, including singlet oxygen and radicals. 4° Reagents and Materials 12-(l'-Pyrene)dodecanoic acid (Molecular Probes, Eugene, OR), 20 mM stock solution in dimethyl sulfoxide (DMSO), stored under nitrogen, in the dark, and at -20% diluted into growth medium Transilluminator (Ultraviolet Products Inc., San Gabriel, CA) equipped with two bulbs (15-W Sylvania F1ST8 Blacklight Blue) UV Intensity meter (Blak-Ray, Ultraviolet Products Inc.) Procedure. Cells are seeded, in the appropriate growth medium, in sterile plastic tissue culture dishes or in sterile glass tubes. Cells are incubated in the presence of 12-(1'-pyrene)dodecanoic acid together with 40 j. D. Spikes, in "The Science of Photobiology" (K. C. Smith, ed.), p. 87. Plenum, New York, 1977.
608
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[61]
a radioactive precursor for labeling phospholipids. 27 Prior to photosensitization, medium is aspirated and cells are washed with phosphate-buffered saline, pH 7.4 (PBS). Tissue culture dishes or glass tubes containing cells in PBS are placed above a Transilluminator equipped with UV light bulbs (>300 nm) and exposed to UV light for the appropriate amount of time. Cells must be placed at a distance of more than 2 cm from the UV source to assure homogeneous exposure of all samples, and exposure is calibrated with the UV intensity meter. Standard conditions of exposure are as follows: 10-20 W/m 2 at 365 nm for 5-30 rain. A glass plate must be intercalated between the tissue culture dish and the UV source to block any UV light below 300 nm.
Method 2: Plasmalogen Breakdown in [32p]Pi-Labeled Cells Principle. The following method is used to characterize the disappearance of 32p-plasmalogens in cultured Chinese hamster ovary (CHO-K1) cells under conditions of photooxidative stress. 27 Cells are preincubated with [32p]Pi to label all phospholipids, and with 12-(1 '-pyrene)dodecanoic acid to incorporate the photosensitizer into cell membrane lipids. After UV-irradiation, phospholipids are extracted and separated by two-dimensional thin-layer chromatography (TLC). Reagents and Materials Phosphorus-32 (orthophosphoric acid), approximately 9 Ci//zmol (Du P o n t - N e w England Nuclear, MA) 12-(1'-Pyrene)dodecanoic acid, 20 mM stock solution in DMSO, diluted at 10-I00/zM into growth medium Mercuric chloride, 10 mM in glacial acetic acid (heating helps solubilization) Silica gel 60 TLC plates (10 × 10 cm, E. Merck, Darmstadt, Germany) Solvent system A: chloroform/methanol/acetic acid/water (25 : 15 : 4 : 2, v/v) Solvent system B: chloroform/methanol/formic acid (65:25 : 10, v/v) Container to dry TLC plates in a stream of nitrogen Procedure. The CHO-KI cells are seeded in 1 ml of growth medium in sterile glass tubes ( - 2 x 104 cells/tube). One day after seeding, the medium is replaced with 0.9 ml of fresh medium containing [32p]Pi (200 /zCi/ml). The [32p]Pi can be also added at the time of seeding. After another day, each tube receives 0.1 ml of medium containing 12-(1'-pyrene)dodecanoic acid to provide a final concentration of the pyrene fatty acid of 1-10 t~M, and the tube is incubated for another 12-24 hr. Then, the medium is removed by aspiration, and cells are washed twice with PBS and placed in 1 ml of PBS.
[61]
PLASMALOGENS
609
Immediately after UV-irradiation (see Method 1) the reaction is quenched by addition of 2.5 ml of methanol and 1 ml of chloroform. After heating at 60°, samples receive 1.25 ml of PBS and 1.5 ml of chloroform to obtain a two-phase solvent system. 41 After vortexing and centrifuging, the lower organic phase is collected, and lipids are dried under nitrogen. Lipids are analyzed by two-dimensional thin-layer chromatography42 using I0 by 10 cm silica gel plates. Lipids solubilized in a small volume of chloroform/methanol (1 : 1, v/v) are spotted at the comer of the plate at a distance of 1.5 cm from both edges. Lipids are chromatographed in the first dimension in solvent system A to the top of the plate. Next, the plate must be carefully dried in a stream of nitrogen for 10 min. The plate is sprayed with 10 mM mercuric chloride in glacial acetic acid and allowed to dry thoroughly for 20 min in a stream of nitrogen. Treatment with mercuric chloride in acid will cleave the vinyl ether linkage of plasmalogens, 5 allowing for the separation of diacylphosphatidylcholine and diacylphosphatidylethanolamine from the plasmalogen-derived lysophospholipids (i.e., 2-monoacylglycerophosphocholine and 2-monoacylglycerophosphoethanolamine, respectively) in the second dimension.16'27 Ether phospholipids (e.g., plasmanylcholine and plasmanylethanolamine) are not hydrolyzed under these conditions. The plate is chromatographed in the second dimension in solvent system B to the top of the plate. After drying the plate, radioactive phospholipids are visualized by autoradiography, scraped off, and counted by liquid scintillation spectrometry. The percent distribution of each radioactive phospholipid is calculated. Comments. It is important that the photosensitization reaction is quenched rapidly. If cells cultured in suspension are used, they should be prepared in buffer (PBS, Tris, or HEPES) in glass tubes just before applying oxidative stress. Then the reaction is rapidly stopped by addition of methanol and chloroform. If cells usually growing on plastic are used (e.g., CHO-K1 cells) one possibility consists of seeding them instead in glass tubes, and replacing the growth medium with buffer just before applying oxidative stress and quenching with chloroform and methanol. Alternatively, the cells are seeded in conventional polystyrene tissue culture dishes. Then, the growth medium is replaced by buffer before applying oxidative stress, and reaction stopped with methanol alone, and cells scraped off and transferred to glass tubes prior to addition of chloroform. Figure 3 shows a typical distribution of phospholipids extracted from control and photosensitized 32p-labeled CHO-K1 cells. In control cells, ethanolamine-containing phospholipids were separated between diacyl41 E. G. Bligh and W. J. Dyer, Can. J. Biochem. Physiol. 37, 911 (1959). 42 j. D. Esko and C. R. H. Raetz, J. Biol. Chem. 255, 4474 (1980).
610
Control cells
E
G) >, (n 4)
>
[61]
A N T I O X I D A N T C H A R A C T E R I Z A T I O N A N D ASSAY
5 A
4e
Photosensitized cells
3
2
O (n A o4
6
....
/
....
4e
2
'lj,
11
1) Solvent system A
FIG. 3. Schematic representation of the separation of [32p]Pi-labeledphospholipids by two-dimensional thin-layer chromatography showing the disappearance of plasmalogens in photosensitized CHO-K1 cells.27O, origin; 1 and 2, sphingomyelin;3, phosphatidylcholine; 4, phosphatidylinositol; 5, phosphatidylserine; 6, lysophosphatidylethanolamine (derived from acidic cleavage of plasmenylethanolamine); 7, phosphatidylethanolamine.
phosphatidylethanolamine (spot 7, Fig. 3) and plasmenylethanolaminederived lysophosphatidylethanolamine (spot 6, Fig. 3). CHO-K1 cells contain little or no plasmenylcholine as evidenced by the absence of a second radioactive spot below diacylphosphatidylcholine (spot 3, Fig. 3). Photosensitization induces the selective disappearance of plasmenylethanolamine. No radioactive breakdown product derived from plasmenylethanolamine was found in this two-dimensional thin-layer chromatography, and the method described next can be used to identify the product that might have migrated with another phospholipid.
Method 3: Plasmenylethanolamine Breakdown in [2J4 C]Ethanolamine-Labeled Cells Principle. The following method is used to characterize (1) the disappearance of plasmenylethanolamine and (2) the formation of lysophosphatidylethanolamine in cultured cells under conditions of photooxidative stress. 27Cells are preincubated with [2-14C]ethanolamine to label plasmenylethanolamine, and with 12-(l'-pyrene)dodecanoic acid to incorporate the photosensitizer into cell membrane lipids. After UV-irradiation, radioactive phospholipids are extracted and separated by two-step one-dimensional thin-layer chromatography.
[61]
PLASMALOGENS
611
Reagents and Materials [2-14C]Ethanolamine, 50-60 mCi/mmol (Amersham Corp., Amersham, UK) 12-(1 '-Pyrene) dodecanoic acid, 20 mM stock solution in DMSO, then diluted at 10-100/xM into growth medium Silica gel 60 TLC plates (20 x 20 cm, E. Merck) Mercuric chloride, 10 mM in glacial acetic acid Solvent system A: chloroform/methanol/acetic acid/water (25 : 15 : 4 : 2, v/v) Container to dry TLC plates in a stream of nitrogen Procedure. The CHO-K1 cells are grown at the bottom of sterile glass tubes ( - 1 × 105 cells/tube) in 0.9 ml of medium containing approximately 0.2/~Ci/ml [2-14C]ethanolamine. After 1 day, each tube receives 0.1 ml of growth medium containing 12-(1'-pyrene)dodecanoic acid to provide a final concentration of the pyrene fatty acid of 1-10/zM and is incubated for another 12-24 hr. After the medium is removed by aspiration, the cells are washed twice with PBS and incubated in 1 ml of PBS. Immediately after UV-irradiation (see Method 1), the reaction is quenched by the addition of 2.5 ml of methanol and 1 ml of chloroform. After heating at 60°, samples receive 1.25 ml of PBS and 1.5 ml of chloroform to obtain a two-phase solvent system. After vortexing and centrifuging, the lower organic phase is collected, and lipids are evaporated to dryness under nitrogen. Lipids are analyzed by two-step one-dimensional thin-layer chromatography using 20 by 20 cm silica gel plates. Lipids solubilized in a small volume of chloroform/methanol (1 : 1, v/v) are spotted in 1-cm bands at a distance of 2 cm from the bottom edge of the plate and first chromatographed in solvent system A to a distance of 5 cm. Next, the plate must be carefully dried in a stream of nitrogen for 10 min. The plate is sprayed with 10 mM mercuric chloride in glacial acetic acid 5 and allowed to dry thoroughly for 20 min in a stream of nitrogen. The plate is rechromatographed in the same dimension using the same solvent system A to a distance of 15 cm. After drying the plate, the radioactive phospholipids are visualized by autoradiography, scraped off, and counted by liquid scintillation spectrometry. This two-step one-dimensional TLC procedure allows for the separation of diacylphosphatidyl[2-14C]ethanolamine (Rf -0.67), 2-monoacylglycerophospho[2-14C]ethanolamine derived from plasmenyl[2J4C]etha nolamine after acidic cleavage on the plate (Rf -0.48), and 2-monoacylglycerophospho[2-14C]ethanolamine generated by photooxidation and originally present in the extract (Rf -0.39). 27 The radioactive breakdown
612
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[61]
product 2-monoacylglycerophosphoethanolamine can be also characterized directly by one-step one-dimensional TLC using solvent systems such as chloroform/methanol/acetic acid/water (25 : 15 : 4 : 2, v/v) or chloroform/acetone/methanol/acetic acid/water (6 : 8 : 2 : 2 : 1, v/v) with a standard lysophospholipid running in parallel. Comments. This method can be partially transposed to characterize the decomposition of plasmenylcholine. CHO-K 1 cells contain little or no plasmenylcholine, but other cell types may have higher levels. In this case, [methylJ4C]choline will be used to label plasmenylcholine and diacylphosphatidylcholine. For identification the two-step two-dimensional system of Method 2 is preferable because 2-monoacylglycerophosphocholine, diacylphosphatidylcholine, and sphingomyelin will separate well. Lysophosphatidylcholine and sphingomyelin have low Rf values, and they separate rather poorly in conventional one-dimensional TLC systems.
Method 4: Formation of Radioactive Formic Acid in [1J4C]Hexadecanol-Labeled Cells Principle. [1-14C]Hexadecanol is used to label the vinyl ether-linked fatty chain of plasmalogens. The long-chain fatty alcohol will condense with acyI-DHAP to form alkyI-DHAP prior to acylation and subsequent polar head group transfer. 43 Carbon atom 1 of hexadecanol will be the source of radioactive formic acid produced from photooxidized plasmalogens (Fig. 1). After photosensitization, radioactive formic acid is characterized by its specific conversion to CO2 by formate dehydrogenase (EC 1.2.1.2) in the presence of NAD+. 44'45 Because the water-soluble formate can rapidly diffuse out of cells, both cell extracts and extracellular media must be analyzed. 27 Reagents and Materials [1J4C]Hexadecanol is synthesized46 from [1-14C]palmitic acid, 40-60 mCi/mmol (Du Pont-New England Nuclear) [14C]Formate, sodium salt, 40-60 mCi/mmol (Du Pont-New England Nuclear) 12-(1 '-Pyrene)dodecanoic acid, 20 mM stock solution in DMSO, then diluted at 9-90/zM into growth medium Formate dehydrogenase (from Pseudomonas oxalaticus, Sigma, St. Louis, MO), 7 units/ml in water 43 F. Paltauf, in " E t h e r Lipids, Biochemical and Biomedical Aspects" (H. K. Mangold and F. Paltaulf, eds.), p. 107. Academic Press, New York, 1983. J. R. Quayle, this series, Vol. 9, p. 360. 45 j. S. Blanchard and W. W. Cleland, Biochemistry 19, 3543 (1980). 46 p. A. Davis and A. K. Hajra, J. Biol. Chem. 254, 4760 (1979).
[61]
PLASMALOGENS
613
Formic acid, 0.4 mM in water NAD + (Sigma), 40 mM in water Methylbenzethonium hydroxide (Aldrich, Milwaukee, WI) Trichloroacetic acid (TCA), 80% (w/v) in water Glass vials with rubber stoppers and small plastic wells (insert) like those used for radioactive fatty acid/3-oxidation measurements Filter paper Disposable syringes (1-ml) Procedure. Dry radioactive [1-14C]hexadecanol (50/zCi) is dissolved in 0.2 ml of ethanol, added to 5 ml of growth medium, sonicated twice for 30 sec, and diluted to 25 ml with growth medium. CHO-K1 cells are seeded in 3 ml of medium at a density of 5 × 104 cells in 35-mm tissue culture dishes, and supplemented with 1 ml of the radioactive hexadecanol stock to get a final activity of 0.5 /zCi/ml ( - 9 /xM). After 2 days of incubation, cells receive 0.5 ml of medium containing 12-(l'-pyrene)dodecanoic acid to provide a final concentration of the pyrene fatty acid of 1-10/zM and are incubated further for 12-24 hr. The medium is removed by aspiration, and the cells are washed twice with PBS and incubated in 1.5 ml of PBS. Immediately after UV-irradiation (see Method 1), the buffer is collected, cells are washed twice with 1.5 ml of cold PBS, and the three PBS fractions are pooled. The pooled material defined as the extracellular medium is kept on ice. Cells are scraped off using a rubber policeman and transferred to glass tubes, subjected to three cycles of freezing and thawing, and kept on ice. Sonication can be also used to disrupt the cells. Both extracellular media and cells are centrifuged at 130,000 g for 60 min in a Beckman 50Ti rotor to pellet cell debris, and supernatants are kept on ice. For the formate dehydrogenase assay, 400/~1 of centrifuged extracellular medium or cell supernatant, 50/zl of 0.4 mM nonradioactive formate as carrier, 50/zl of 40 mM NAD +, and 400/xl of water are mixed in small glass vials on ice. Vials are closed with a rubber stopper, equipped with a small plastic well containing a filter paper soaked with 40/zl of methylbenzethonium hydroxide as the CO2 trapping agent. The mixture is warmed to 37° for 1 min, and the reaction is initiated by injecting 0.1 ml of formate dehydrogenase (7 units/ml) with a syringe. Final conditions are as follows: 20/~M cold formate, 2 mM NAD +, 0.7 units formate dehydrogenase/ml in 1 ml of 2.5-fold diluted PBS. The mixture is incubated at 37° for 30 min, and the reaction is stopped by injecting 0.1 ml of 80% TCA. The vials are kept in the water bath for another 1 hr to allow for the trapping of CO2. Next, filter papers are removed and counted by liquid scintillation spectrometry. [~4C]Formate diluted with cold formate is assayed in parallel
614
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[61]
for calibration. Between 50 and 90% of the radioactive formic acid produced from plasmalogens of photooxidized CHO-K1 cells is found in the extracellular medium. Method 5: Formation of Fatty Aldehydes in [UJ4C]Hexadecanol Labeled Cells Principle. Fatty aldehydes are breakdown products of photooxidized plasmalogens (Fig. 1 and 2). In the following method [U-14C]hexadecanol is used to label the vinyl ether-linked fatty chain of plasmalogens. 43 This vinyl ether-linked fatty chain will be the source of fatty aldehyde produced from photooxidized plasmalogens; it is extracted and identified according to its retention factor after separation by thin-layer chromatography. Its conversion to fatty acid in the presence of chromium trioxide 47 will also confirm its nature. Reagents and Materials [U-14C]Hexadecanol is synthesized46from [U-i4C]palmitic acid, >500 mCi/mmol (Du Pont-New England Nuclear) 12-(1'-Pyrene)dodecanoic acid, 20 mM stock solution in DMSO, then diluted at 10-100 ~ M into growth medium Silica gel 60 TLC plates (20 × 20 cm, E. Merck) Solvent system C: n-hexane/diethyl ether/acetic acid (80 : 20 : 1, v/v) Solvent system D: n-hexane/diethyl ether/acetic acid (90 : 10 : 1, v/v) Solvent system E: n-heptane/isopropyl ether/acetic acid (60:40:4, v/v) cis-Hexadecenal (Sigma or Aldrich) or another long-chain fatty aldehyde standard Chromium trioxide (Sigma), saturated solution in glacial acetic acid Sodium hydroxide, 3 N in water Concentrated HCI Procedure. [U-14C]Hexadecanol (10/zCi) is dissolved in 0.2 ml of ethanol, added to 5 ml of growth medium, and sonicated twice for 30 sec. CHO-K1 cells are seeded in 6 ml of medium in 100-mm plastic culture dishes at a density of 5 x 106 cells/dish. Each dish receives also [U-14C]hexadecanol in 1.2 ml of medium and 12-(l'-pyrene)dodecanoic acid at 10-100/xM in 0.8 ml of medium. Final conditions are as follows: 0.3/xM [U-14C]hexadecanol (0.3/xCi/ml) and 1-10/zM 12-(1 '-pyrene)dodecanoic acid. Cells are incubated for 12-24 hr, washed twice with PBS, and UV-irradiated in 4.5 ml of PBS for 5-30 min (see Method 1). The 47 G. M. Gray, in "Lipid Chromatographic Analysis" (G. V. Marinetti, ed.), Vol. 1, p. 401. Dekker, New York, 1967.
[61]
PLASMALOGENS
615
reaction is immediately quenched with 5 ml of methanol, and the cells are scraped off the dishes with a rubber policeman and transferred to glass tubes. After heating at 60 °, each tube receives 5 ml of chloroform prior to mixing and centrifuging. The lower organic phase is collected and evaporated to dryness under nitrogen, and the lipids are further analyzed. Lipids are separated first by thin-layer chromatography using solvent system C with standard lipids running in parallel. In this system the Rf values for fatty alcohol, fatty acid, triacyglycerol, fatty aldehyde, and cholesteryl ester are 0.09, 0.14, 0.35, 0.44, and 0.62, respectively. Phospholipids remain at the origin. Other minor lipids are usually seen. Radioactive lipids are visualized by autoradiography, and the band corresponding to long-chain fatty aldehyde is scraped off the plate and counted by liquid scintillation spectrometry. The lipids from the band can be also reextracted from the silica gel three times with 2 ml of chloroform/methanol (2: 1, v/v). After drying under nitrogen, the lipid is analyzed again by thin-layer chromatography in two different solvent systems D and E for comparison to a standard such as cis-hexadecenal running in parallel, cis-Hexadecenal is visualized on the plate with iodine vapors. For complete identification, the extracted lipid is tested for its ability to be oxidized to fatty acid in the presence of chromium trioxide. 47 The dry lipid is mixed with 1.5 ml of glacial acetic acid and heated at 40 ° prior to the addition of 0.5 ml of chromium trioxide solution. After 10 min at 40°, 3 ml of water is added, and free fatty acids are extracted twice with 2 ml of benzene. After evaporation under nitrogen the extract is mixed into 4 ml of methanol, 0.5 ml of 3 N NaOH, and impurities are extracted twice with 2 ml of n-hexane. The lower phase is acidified with 0.25 ml of concentrated HC1, and fatty acids are extracted three times with 2 ml of n-hexane. The extract is dried under nitrogen and analyzed by thin-layer chromatography using solvent system E with fatty aldehyde, fatty alcohol, and fatty acid standards running in parallel. Radioactive lipids are visualized by autoradiography. Comments. Long-chain fatty alcohols given to cells are for a good part converted to long-chain fatty acids by a fatty alcohol : NAD ÷ oxidoreductase (EC 1.1.1.192, long-chain-alcohol dehydrogenase).48 As a consequence, [UJ4C]palmitic acid can be recovered from [U-14C]hexadecanollabeled cells in its free form and esterified into complex liquids. In addition, long-chain fatty aldehydes can be readily converted to long-chain fatty acids by a fatty aldehyde dehydrogenase;49 this was evidenced by the 48 W. B. Rizzo, D. A. Craft, A. L. Dammann, and M. W. Phillips, J. Biol. Chem. 17412 (1987). 49 R. Lindhal and S. Evces, J. Biol. Chem. 259, 11991 (1984).
262,
616
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[61]
finding of pentadecanoic acid in photooxidized CHO-K1 cells, 27 supposedly derived from pentadecanal, a breakdown product of plasmalogens.
Method 6: High-Performance Liquid Chromatography of Fatty Aldehyde Moieties after Dinitrophenylhydrazine Derivatization Principle. The previous method to identify the formation of radioactive long-chain fatty aldehydes in photosensitized [U-14C]hexadecanol-labeled cells is simple and broad. It is, however, incomplete and does not provide accurate data on the nature of the fatty aldehydes produced from plasmalogen breakdown. It is possible to analyze the fatty acids derived from chromium trioxide treatment after another conversion to the corresponding phenacyl derivatives for characterization by HPLC, but unstable fatty aldehydes could be lost in this procedure. A more direct and reliable method consists of the derivatization of fatty aldehydes and plasmalogens with 2,4-dinitrophenylhydrazine (DNP hydrazine) producing the corresponding DNP hydrazine hydrazones.5° Plasmalogens and preexisting fatty aldehydes extracted from cells or tissues are first separated by silica gel chromatography, then derivatized with DNP hydrazine under acidic conditions, and finally subjected to HPLC for identification and quantification. The following procedure is adapted from published methods5°-5z with some modifications. Reagents and Materials 2,4-Dinitrophenylhydrazine (DNP hydrazine, Sigma) must be first repurified. A 4 mM DNP hydrazine solution is prepared in 1 N HCI, and kept at 70° for 30 min. After cooling, carbonyl impurities are extracted twice with two 3-ml portions of n-hexane; the nhexane remaining on top of the aqueous phase is evaporated under nitrogen. The DNP hydrazine reagent in 1 N HC1 is stored under nitrogen, in the dark, and at room temperature to be used within 24 hr. 12-(1 '-Pyrene)dodecanoic acid, 20 mM stock solution in DMSO, then diluted at 1-10/zM into growth medium Silica gel (200-400 mesh, Bio-Rad, Richmond, CA) HPLC 250 × 4.6 mm 5-~m Econosil-Cl8 column (Alltech) with precolumn 5o T. Huque, J. G. Brand, J. L. Rabinowitz, and F. F. Gavarron, Comp. Biochem. Physiol. B 86B, 135 (1987). 51 C. Pries and C. J. F. B6ttcher, Biochim. Biophys. Acta 98, 329 (1965). 52 H. Esterbauer, K. H. Cheeseman, M. U. Dianzani, G. Poli, and T. F. Slater, Biochem. J. 208, 129 (1982).
[61 ]
PLASMALOGENS
617
Mobile phase for HPLC: acetonitrile/water (95 : 5, v/v) Methanol, chloroform, and n-hexane (all carbonyl-free) cis-Hexadecenal (Sigma) Saturated and unsaturated long-chain fatty aldehyde standards, which can be synthesized by selective oxidation of the sulfanate esters of corresponding fatty acids 53 Procedure. The CHO-K1 cells are seeded in medium in 60-mm tissue culture dishes at a density of approximately 1 × 105 cells/well and incubated for 1 day. The next day, medium is replaced by 3 ml of medium containing 1-10 /zM of P12. After 12-24 hr of incubation the medium is removed, and cells are washed twice with PBS and UV-irradiated in 1 ml of PBS (see Method 1). The reaction is immediately quenched with 2.5 ml of methanol. Cells are scraped off the dishes with a rubber policeman, and the methanol/water suspension is transferred to glass tubes containing 1 ml of chloroform, mixed, and heated at 60 ° for 10 min. Next, samples receive 1.25 ml PBS and 1.5 ml of chloroform to obtain a two-phase solvent system. After vortexing and centrifuging, the lower organic phase is collected and lipids dried under nitrogen. Silica gel chromatography is used to separate phospholipids, including the plasmalogens, from the neutral lipids, including long-chain fatty aldehydes released during photooxidation. Small columns, containing about 0.2 g of silica gel in Pasteur pipettes, are first washed with three 2-ml portions of methanol and then with four 2-ml portions of chloroform. Commercial small silica gel columns (e.g., Supelco, Bellafonte, PA) can also be used. Lipids in 0.2 ml of chloroform are applied to the columns; neutral lipids are eluted with three 1-ml portions of chloroform, and polar lipids with four 1-ml portions of methanol. All fractions are evaporated to dryness under nitrogen in glass tubes with Teflon screw caps prior to derivatization with DNP hydrazine. Polar lipid fractions and neutral lipid fractions are redissolved in 0.5 ml of methanol, and 0.5 ml of DNP hydrazine reagent is added. Samples are vortexed and incubated at 70° for 30 min. After cooling, the DNP hydrazones are extracted twice with 1.5 ml of n-hexane. The pooled extracts are dried under nitrogen and redissolved in 20-50/~1 of acetonitrile/water (95 : 5, v/v) for HPLC analysis. Reversed-phase HPLC analyses are performed under isocratic conditions with acetonitrile/water (95 : 5, v/v) as the mobile phase at a flow rate of 2 ml/min with detection at 395 nm. Standard fatty aldehydes are derivatized and analyzed in parallel for identification and calibration. 53 V. Mahadevan, F. Phillips, and W. O. Lundberg, Lipids 1, 183 (1966).
618
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[61]
Comments. This method does not require radiolabeling, and it can be used with cells and tissues exposed to a variety of oxidative stress conditions. The DNP hydrazine derivatization of plasmalogens must be performed under acidic conditions, and it will be used to quantify the disappearance of plasmalogens. Depending on the type of cell or tissue, plasmalogens contain mostly C16 and C~8 vinyl ether-linked fatty chains, and possible long-chain fatty aldehyde breakdown products include pentadecanal and A2-hexadecenal, and heptadecanal and A2-0ctadecenal, respectively. The distribution of the fatty aldehydes produced at room temperature during photooxidation might be different from that obtained at 4°; instability of some of the fatty aldehydes could explain this difference. A higher sensitivity of detection will be obtained using dansyl hydrazine 54 for derivatization in place of DNP hydrazine; dansylhydrazones separated by HPLC are then detected by spectrofluorometry.
General Discussion Whether plasmalogens are a class of natural membrane antioxidants remains an open question. The analytical methods described in this chapter can be extended to other biological systems involving oxidative stress to examine the hypothetical decomposition of plasmalogens. Plasmalogen breakdown in tissues might escape detection when total fatty aldehydes are measured by acidic DNP hydrazine derivatization. Thus, it is important that neutral lipids are first separated from phospholipids as in Method 6, to allow for the detection of preexisting fatty aldehydes originating from oxidative decomposition ofplasmalogens. The identification ofpentadecanal and heptadecanal (Method 6), and of formic acid (Method 4) will point to the occurrence of singlet oxygen. The reactivity of plasmalogens to singlet oxygen and radicals is the basis for an antioxidant, scavenger function, provided that breakdown products are not cytotoxic and/or are rapidly metabolized. Interestingly, CHO-K1 cells can lose up to 30% of plasmenylethanolamine after photosensitization and remain viable,2~'27 pointing to plasmalogens as first-line membrane antioxidants. The cytotoxicity of photosensitization in plasmalogen-deficit cells supports the notion that other biochemical entities essential to cell integrity and functions are indeed protected from activated oxygen species when vinyl ether phospholipids are restored. If plasmalogens are scavengers of singlet oxygen it remains to be seen whether singlet oxygen is produced in defined biological systems causing plasmalogen decomposition. Natural photosensitization is limited to a few 54 j. M. Anderson, Anal. Biochem. 152, 146 0986)°
[61]
PLASMALOGENS
619
tissues such as the skin and eye, and it does not apply to tissues rich in plasmalogens such as heart, brain, muscle, and white cells. Although evidence has accumulated to show the occurrence of activated oxygen species such as superoxide anion in biological systems 55 the production of singlet oxygen in such systems needs further investigation. Kanofsky 56 reviewed several singlet oxygen-generating enzyme systems such as the peroxidase-hydrogen peroxide-halide system. One must be very careful when it comes to extrapolating these observations to living systems, and emphasize that most model enzyme systems producing singlet oxygen are using high, nonphysiological concentrations of hydroperoxides, hydrogen peroxide, and halides. The measurement of 1268 nm oxygen chemiluminescence 57'58 is an unequivocal method for the detection of singlet oxygen in biological systems and should help in sorting out these questions. It has already permitted the characterization of the production of singlet oxygen by stimulated intact human eosinophils 59via the myeloperoxidase-hydrogen peroxide-halide system. Whether stimulated eosinophils lose plasmalogens is not known. Stimulated polymorphonuclear leukocytes do produce fatty aldehydes, although these are not fully identified yet. 6° The oxidation of plasmalogens by radicals is more likely to occur in vivo because radicals are known to play a significant role in a number of biological situations. 55 These situations include ischemia-reperfusion injury in myocardial infarction and stroke, hyperoxygenation syndromes, oxidative-burst inflammatory disorders, lipoprotein oxidation, ionizing radiations, and others. Yavin and Gatt 23'24 have already demonstrated the oxygen-dependent cleavage of plasmenylcholine by rat brain homogenates, together with the production of lysophosphatidylcholine and longchain aldehydes. Loss of plasmalogens has been shown in a situation of ischemia-reperfusion injury in the traumatized cat spinal cord. 61 Older observations have confirmed the disappearance of plasmalogens in the infarcted area of human heart following lethal myocardial infarction. 62 It may well reflect ischemia-reperfusion injury of the myocardium63 causing the oxidative decomposition of plasmalogens; enzymatic hydrolysis of 55 C. E. Cross, B. Halliwell, E. T. Borish, W. A. Pryor, B. N. Ames, R. L. Saul, J. M. McCord, and D. Harman, Ann. Intern. Med. 107, 526 (1987). 56 j. R. Kanofsky, Chem.-Biol. Interact. 70, 1 (1989). 57 A. U. Kahn, J. Am. Chem. Soc. 105, 7195 (1983). 58 j. R. Kanofsky, J. Biol. Chem. 259, 5596 (1984). 59 j. R. Kanofsky, H. Hoogland, R. Wever, and S. J. Weiss, J. Biol. Chem. 263, 9692 (1988). 6o O. H. Morand and C. R. H. Raetz, unpublished results (1989). 61 p. Demediuk, R. D. Saunders, D. K. Anderson, E. D. Means, and L. A. Horrocks, Proc. Natl. Acad. Sci. U.S.A. 82, 7071 (1985). 62 K. A. Oster and P. Hope-Ross, J. Cardiol. 17, 83 (1966). 63 M. S. Sussman and G. B. Bukley, this series, Vol. 186, 711.
620
ANTIOXlDANT CHARACTERIZATION AND ASSAY
[62]
plasmalogens by a plasmalogen-selective phospholipase A 2 is also possible. 64 In conclusion, biological systems involving oxidative stress need to be reexamined to establish whether oxidative decomposition of plasmalogens takes place under physiological conditions. 64 D. A. Ford, S. L. Hazen, J. E. Saffitz, and R. W. Gross, J. Clin. Invest. 88, 331 (1991).
[62] A n t i o x i d a n t A c t i v i t y o f C a l c i u m C h a n n e l Blocking Drugs
By I. TONG MAK and WILLIAM B. WEGLICKI Introduction In common with most cardiovascular agents, the clinically used calcium channel blockers (nicardipine, nifedipine, verapamil, diltiazem) are amphiphilic in nature. Thus, in addition to their specific binding to protein receptors, these agents may readily partition into the phospholipid domain of cardiovascular membranes to various degrees according to their lipophilicity. Efforts from our laboratory have focused on the effects of such agents on the sensitivities of cardiac membranes and vascular cells to free radical injury. 1-7 At the membrane level, we have chosen the highly purified sarcolemmal membranes of ventricular myocytes as model membranes. Compared to other subcellular membranes, the sarcolemmal membranes were much more sensitive to free radical-mediated damage, 8 probably owing to the highly enriched phospholipid content. 8-1° To assess the extent of membrane lipid peroxidation, we chose to use the thiobarbituric acid (TBA) method because of its sensitivity and convenience. 1 I. T. Mak and W. B. Weglicki, Circ. Res. 63, 262 (1988). 2 I. T. Mak and W. B. Weglicki, Circ. Res. 66, 1449 (1990). 3 I. T. Mak, C. M. Arroyo, and W. B. Weglicki, Circ. Res. 65, 1151 (1989). 4 W. B. Weglicki, I. T. Mak, and M. G. Simic, J. Mol. Cell. Cardiol. 22, 1199 (1990). 5 I. T. Mak, A. M. Freedman, B. F. Dickens, and W. B. Weglicki, Biochem. Pharmacol. 40, 2169 (1990). 6 I. T. Mak, P. Boehme, and W. B. Weglicki, Circ. Res. 70, 1099 (1992). 7 I. T. Mak, J. H. Kramer, and W. B. Weglicki, Coronary Artery Dis. 3, 1095 (1992). s j. H. Kramer, I. T. Mak, and W. B. Weglicki, Circ. Res. 55, 120 (1984). 9 W. B. Weglicki, K. Owens, F. F. Kennett, A. Kessner, L. Harris, R. M. Wise, and G. V. Vahoouny, J. Biol. Chem. 255, 3605 (1980). l0 W. B. Weglicki, J. H. Kramer, I. T. Mak, B. F. Dickens, and T. M. Phillips, in "Isolated Adult Cardiomyocytes" (H. Piper and G. Isenberg, eds.), Vol. 1, p. 1. CRC Press, Boca Raton, Florida, 1988.
METHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
620
ANTIOXlDANT CHARACTERIZATION AND ASSAY
[62]
plasmalogens by a plasmalogen-selective phospholipase A 2 is also possible. 64 In conclusion, biological systems involving oxidative stress need to be reexamined to establish whether oxidative decomposition of plasmalogens takes place under physiological conditions. 64 D. A. Ford, S. L. Hazen, J. E. Saffitz, and R. W. Gross, J. Clin. Invest. 88, 331 (1991).
[62] A n t i o x i d a n t A c t i v i t y o f C a l c i u m C h a n n e l Blocking Drugs
By I. TONG MAK and WILLIAM B. WEGLICKI Introduction In common with most cardiovascular agents, the clinically used calcium channel blockers (nicardipine, nifedipine, verapamil, diltiazem) are amphiphilic in nature. Thus, in addition to their specific binding to protein receptors, these agents may readily partition into the phospholipid domain of cardiovascular membranes to various degrees according to their lipophilicity. Efforts from our laboratory have focused on the effects of such agents on the sensitivities of cardiac membranes and vascular cells to free radical injury. 1-7 At the membrane level, we have chosen the highly purified sarcolemmal membranes of ventricular myocytes as model membranes. Compared to other subcellular membranes, the sarcolemmal membranes were much more sensitive to free radical-mediated damage, 8 probably owing to the highly enriched phospholipid content. 8-1° To assess the extent of membrane lipid peroxidation, we chose to use the thiobarbituric acid (TBA) method because of its sensitivity and convenience. 1 I. T. Mak and W. B. Weglicki, Circ. Res. 63, 262 (1988). 2 I. T. Mak and W. B. Weglicki, Circ. Res. 66, 1449 (1990). 3 I. T. Mak, C. M. Arroyo, and W. B. Weglicki, Circ. Res. 65, 1151 (1989). 4 W. B. Weglicki, I. T. Mak, and M. G. Simic, J. Mol. Cell. Cardiol. 22, 1199 (1990). 5 I. T. Mak, A. M. Freedman, B. F. Dickens, and W. B. Weglicki, Biochem. Pharmacol. 40, 2169 (1990). 6 I. T. Mak, P. Boehme, and W. B. Weglicki, Circ. Res. 70, 1099 (1992). 7 I. T. Mak, J. H. Kramer, and W. B. Weglicki, Coronary Artery Dis. 3, 1095 (1992). s j. H. Kramer, I. T. Mak, and W. B. Weglicki, Circ. Res. 55, 120 (1984). 9 W. B. Weglicki, K. Owens, F. F. Kennett, A. Kessner, L. Harris, R. M. Wise, and G. V. Vahoouny, J. Biol. Chem. 255, 3605 (1980). l0 W. B. Weglicki, J. H. Kramer, I. T. Mak, B. F. Dickens, and T. M. Phillips, in "Isolated Adult Cardiomyocytes" (H. Piper and G. Isenberg, eds.), Vol. 1, p. 1. CRC Press, Boca Raton, Florida, 1988.
METHODS IN ENZYMOLOGY,VOL. 234
Copyright© 1994by AcademicPress, Inc. All rightsof reproductionin any formreserved.
[62]
ANTIOXlDANT PROPERTIES OF CALCIUM BLOCKERS
621
Increasing evidence has accumulated to suggest that lipid peroxidative processes are involved in the pathogenesis of vascular diseases.ll,12 Among those processes, oxidative injury of endothelial cells may represent a critical event in propagating atherogenesis. 11 The endothelial cells are potential targets of reactive oxygen radicals released from activated blood cells (e.g., neutrophils, macrophages, platelets) and oxidizable drugs and chemicals. Therefore, at the cellular level, we have used cultured endothelial cells to assess the cytoprotective effects of the calcium channel blockers against free radical-induced loss of glutathione and increased membrane permeability. Both parameters were relatively sensitive to the oxidative stress generated from a chemical oxygen-radical system. Cardiac Sarcolemmal Membrane Model
Sarcolemmal Preparation. Sarcolemmal membranes are isolated from adult canine ventricular myocytes. The isolation procedure has been described in detail elsewhere, s-l° Briefly, adult canine myocytes are isolated from ventricular tissue by enzymatic digestion with 0.05% collagenase. Following disruption of the myocytes by nitrogen cavitation (1000 psi, 30 min), the sarcolemmal membranes are enriched by differential and sucrose gradient centrifugation. The sarcolemmal fractions, which band between 21 and 26% sucrose, are about 80-fold enriched in the specific activity of the marker enzyme Na ÷, K+-ATPase over that of the myocyte homogenate.9'l° Generation of Free Radicals. The chemical system we use to generate oxygen-radicals consists of dihydroxyfumarate (DHF) and FeCI3-ADP. 13 Oxidation of D H F (Sigma Chemical Co., St. Louis, MO) in solution generates sustained levels of superoxide anions14'15; the rate of production is further promoted by the presence of metal chelates such as Fe-ADP. Hydroxyl radicals (.OH) are generated according to the following chemical reactions: DHF + 02--> .DHF + 02 ~ •DHF + 02 ~ diketosuccinate + O2 ~ 202: + 2H ÷ ~ H202 + 02 Fe3+-ADP + 02 ~ ~ Fe2+-ADP + 02 Fe2+-ADP + H202 --> -OH + Fe3+-ADP + O H 11 B. Hennig and C. K. Chow, Free Radical Biol. Med. 4, 99 (1988). 12 B. Halliwell, Br. J. Exp. Pathol. 70, 737 (1989). 13 I. T. Mak, H. P. Misra, and W. B. Weglicki, J. Biol. Chem. 258, 13733 (1983). t4 S. A. Goscin and I. Fridovich, Arch. Biochem. Biophys. 153, 778 (1972). t5 B. Halliwell, Biochem. J. 163, 441 (1977).
622
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[62]
Presumably, other iron-oxygen species (such as hypervalent iron complexes) may also be formed in Fenton-type reactionsl6; these species could be just as deleterious as .OH radicals. In our system, a concentrated solution of D H F is prepared in the incubation buffer; a brief period (<2 min) of heating in an 80° water bath is required to dissolve the D H F in solution. The dissolved D H F solution is placed on ice and adjusted to pH 7 by adding 1 N KOH. The F e - A D P chelate is prepared by mixing FeCI3 (1-2 mM) with a 10-fold higher concentration (10-20 mM) of ADP to achieve a final concentration ratio of F e 3+ to ADP of 1 : 10. Incubation and Measurement of Lipid Peroxidation. With the membrane system, we found that the TBA colorimetric method is quite satisfactory for assessing the membrane antioxidant activities of calcium blockers. Owing to the varying lipophilicity of the calcium channel blockers, stock solutions of the agents (1-20 mM) are prepared in absolute ethanol. Because of light sensitivity of the dihydropyridines, all experiments are conducted at minimal light. Sarcolemmal membranes ( - 5 0 /zg protein/0.50 ml) are mixed with 10/zl of each calcium-blocking drug and preincubated for 10 min, 37°, in a reaction buffer consisting of 120 mM KCI, 50 mM sucrose, and 10 mM potassium phosphate, pH 7.2; lipid peroxidation reactions are initiated by addition of freshly prepared F e - A D P (final concentrations: 0.025 mM FeCI 3 chelated by 0.25 mM ADP and DHF (0.83) mM). The rate of superoxide production is about 3 nmol/min/ml. At various times of incubation, the levels of TBA-reactive materials are determined by adding 0.5 ml of 0.5% TBA, 50 /.d of 10% trichloroacetis acid (TCA) and 10/xl of 1% butylated hydroxytoluene (BHT) to 0.5 ml of reaction mixture. The mixture is then vigorously vortexed and the color is developed by heating in a water bath at 80° for 30 min. After cooling on ice, 0.5 ml of 70% TCA is added to each sample which is then vortexed and centrifuged; the chromophore developed in the supernatant is measured at 532 nm. The TBA-reactive materials are estimated by using standards of malonaldehyde bis(dimethyl acetal) subjected to an identical heating procedure; the results are expressed as malondialdehyde (MDA) equivalents. The time course of lipid peroxidation in the sarcolemmal membranes, with or without added calcium blockers, is displayed in Fig. 1. The accumulation of reaction products in most samples appeared to be quasi-linear up to 20 min. Results in Fig. 1 indicate that 100/xM verapamil, nifedipine, 16 B. H. J. Bielski, Basic Life Sci. 49, 123 (1988).
[62]
ANTIOXIDANT PROPERTIES OF CALCIUM BLOCKERS A
623
e80
,&
E "6 E e-
60
/
.V
40
v
s'
C 0 s' j p f s sS
¢0
E
20
~w"
°°"
II. a
0 -10
0
10
20
30
Time of Incubation
40
50
(min)
FIG. 1. Time course of free radical-mediated lipid peroxidation in isolated sarcolemmal membranes in the absence and presence of various calcium channel blockers (100/zM each). Values are means of 3-8 determinations. O, R' control; A, + diltiazem; T, + verapamil; 0 , + nifedipine; II, + nicardipine.
and nicardipine provided varying degrees of significant inhibition throughout 45 min of incubation. The same level of diltiazem provided a significant inhibition at 20 min but not at 45 min. These time course studies led us to choose 20 min as the incubation time to compare the inhibitory potency of the calcium blockers. As represented in Fig. 2, all four calcium blockers exhibited concentration-dependent inhibition of sarcolemmal lipid peroxidation. The order of potency was nicardipine > nifedipine > verapamil > diltiazem; their respective values of ECs0 (in micromolar) were estimated to be 22.6 for nicardipine, 38 for nifedipine, 210 for verapamil, and 850 for diltiazem. At least 5 min of preincubation of the agents and membranes is required prior to the free radical reaction, suggesting that membrane-drug interactions are necessary to effect the subsequent results. To evaluate the initial association of the drugs with the membranes, the following experiments are designed according to Scheme 1. Samples for set A are incubated normally as described. For the duplicate set (set B), the sarcolemmal membranes with or without (controls) drug treatment (100/zM each, 10 min, 37°) are transferred to the airfuge tubes and centrifuged at 150,000g
624
[62]
ANTIOXIDANT CHARACTERIZATIONAND ASSAY Sarcolemmal membranes -+ drug (total volume 450/zl) SetA
~
,
10 min ~ S e t
+R. (50 ~1)
B
150,000~g, 10 min (Airfuge centrifugation)
J
Lipid peroxidation (20 min)
bSupernatant J removed Membrane pellets resuspended in 450/zl fresh buffer l
+R' (50/zl)
Lipid peroxidation (20 min) SCHEME I
for 10 min by using a Beckman Airfuge (Beckman Inst. Inc., Palo Alto, CA). The supernatants, containing some portion of drugs not associated with the membranes, are removed. The subsequent membranous pellets are resus¢-
.O
120
X O
100
._0. .-I c O
,~ ¢_=
E C 5 0 uM Nic 2 2 . 6 Nil 3 8 Ver Dilt
6O
i---'~.W ..... * /~-'"
Tj/',/'a"
206 850
~/,,,"
~,./"I
//~/"
,v
7,,///L
4O
. i . ~. L
1~mJ.. "
20
m_~.-~
V"'"
x-"- -f"
0 .3
.699
1.0
1.301
..'"~ .L
"~
.A" "
#
7
,
,
1.602
2.0
2,3
2.602
3.0
Log [uM] of Agents FIG. 2. Comparative inhibitory effects of the calcium channel blockers on sarcolemmal lipid peroxidation. The ECs0values (/zM)were as follows: (U) nicardipine, 22.6; (O) nifedipine, 38; (V) verapamil, 206; and (A) diltiazem, 850. (Results partially adapted from Mak and Weglicki,2 with permission.)
[62]
ANTIOXIDANT PROPERTIES OF CALCIUM BLOCKERS
625
TABLE I REQUIREMENT OF MEMBRANE PARTITIONING OF CALCIUM CHANNEL BLOCKERS FOR ANTIOXIDANT ACTIVITIES a M D A formation b
Conditions
Set A
Set B
Controls (R.) Agents (100 p.M) added: Nicardipine Nifedipine Verapamil Diltiazem
0.246
0.192
0.010 0.035 0.160 0.181
0.007 0.029 0.132 0.163
a Incubation conditions for set A were as described on p. 634. The conditions for set B were as described in Scheme 1; after the 10 min of membrane --- drug preincubation, the supernatants were removed by centrifugation and the membrane pellets were resuspended in fresh medium before addition of the free radical components. Other conditions were as described in Figs. 1 and 2. b (Absorbance at 532 nm)
pended in fresh buffer; the samples are then subjected to the free radical reaction for 20 min. Results summarized in Table I indicate that the calcium channel blockers in the samples of set B provided an extent of inhibition similar to those of set A. These results indicate that the calcium blockers mediate their antioxidant action in the lipid domain of the membranes. The overall results suggest that the calcium channel blockers provide lipophilic "chain-breaking" activity similar to that of a-tocopherol. Comments. The rate of DHF oxidation in solution can be followed by measuring the fall in absorbance at 300 nm.14 The corresponding rate of superoxide generation by the chemical oxygen radical system can be measured independently by the superoxide dismutase (SOD)-inhibitable reduction of cytochrome c (75/xM) according to the method of Kellogg and Fridovich/7 In the presence of Fe-ADP, we did not find that the addition of SOD (10 /zg/ml) would inhibit the rate of DHF oxidation. With electron spin resonance (ESR) spectroscopy using 5,5-dimethyl-1pyrolidine N-oxide (DMPO) as the spin trap, the formation of hydroxyl radicals was confirmed and the steady-state level could be monitored. 3'18 17 E. W. Kellogg and I. Fridovich, J. Biol. Chem. 252, 6721 (1977). i8 C.M. Arroyo, I. T. Mak, and W. B. Weglicki, Free Radical Res. C o m m u n . 5, 369 (1989).
626
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[62]
Under these conditions, none of the calcium blockers affected the rates of D H F oxidation and superoxide generation; none of the agents had any effect on the steady-state signal intensity of the D M P O - O H adducts. Despite some concerns about the specificity of the TBA method, ~9we found the method to be relatively reliable for the isolated membrane system; many of the intracellular components and metabolic systems that might influence the T B A - M D A chromophore formation were absent. To determine potential agent-mediated interference on the TBA reaction, experiments were performed in which the calcium blockers were added at the end of the 20-min, 37° incubation but before the 80 ° heating stage; the results indicated that none of the drugs interfered with the TBA assay. In a separate study, ~° we observed that the extent of free radical-induced sarcolemmal membrane MDA formation was accompanied by degradation of phospholipids.
Endothelial Cell Model
Cell Culture. Bovine aortic endothelial cells (GM 07372A) are obtained from the Coriell Institute for Medical Research (Camden, N J). The cells are cultured in Dulbecco's modified Eagle's medium supplemented with 15% calf serum. For subsequent subculturing, the cells are split 1 : 5 at confluence using trypsin-EDTA. Confluent plates are trypsinized by 0.05% trypsin in Hanks' balanced salt solution (HBSS) with 0.02% EDTA. The digestion is stopped by adding growth medium with serum. The cells are pelleted, washed twice at room temperature, and finally resuspended in the incubation buffer. Under these conditions, the isolated cells routinely displayed over 95% viability based on the trypan blue exclusion assay. Oxidative Incubation and Drug Treatment. Oxygen-radicals are generated from the D H F / F e - A D P system. The free radical generating system is prepared just before each experiment. Endothelial cells (1 x 106/ml) are resuspended gently in a buffer consisting of 10 mM glucose, 125 mM NaCI, 1.2 mM MgCI2, and 10 mM potassium phosphate, pH 7.2. Ten microliters of either calcium blocker (5-20 nmol) or BHT dissolved in ethanol is added to the 0.5 ml cell suspensions; 10 ttl of ethanol is added to the vehicle controls. All samples (with or without drug) are preincubated at 37° for 15 min in a shaking water bath. Finally, 25/xl of F e - A D P (final concentration of 50/.tM FeC13 chelated by 0.5 mM ADP) and 25 /zl of DHF (I .67 mM) are added to each sample to initiate the oxidative reaction. All experiments are performed in minimal light. The incubations are con19 D. R. Janero, Free RadicalBiol. Med. 9, 515 (1990).
[62]
ANTIOXIDANT PROPERTIES OF CALCIUM BLOCKERS
627
tinued up to 60 min; at various times, samples are assayed for thiols and viability. Assessments of Effects of Drugs on Losses of Cellular Glutathione and Viability. Because the oxygen radicals are generated in the extracellular space, presumably, the initial oxidative target is the plasmalemmal membrane. We therefore chose to use the increase in permeability to trypan blue as an index of membrane integrity. Cell aEiquots are mixed with an equal volume of 0.2% (w/v) trypan blue; the percentage of cells permeable to the dye is expressed as the percent loss of viability. The gEutathione system is considered critical in providing protection against oxidative stress in endothelial celEs.2° Cell aliquots for glutathione determinations are centrifuged at 200 g for 5 min; the cell pellets are resuspended in the incubation buffer containing 5% (w/v) 5-sulfosalicylic acid (SSA); cell disruption is achieved by sonication at 100 W (Fisher Sonic Dismembrator Model 300) for 45 sec. Total glutathione (GSH + ½ GSSG) in the cell supernatant fractions is determined by the enzymatic DTNB [5,5'-dithiobis(2-nitrobenzoic acid)]-GSSG reductase "cyclic method" originally described by Tietze 21 and modified by Anderson. 22 Briefly, pipette (i) 800/xE of working buffer (140 mM sodium phosphate, pH 7.5, 6.2 mM EDTA, with 245 ~g NADPH/ml added on the day of assay), (ii) 100/~1 of 6 mM DTNB solution, and (iii) 100 ~1 of sample of appropriate dilution into a cuvette. After mixing, the cuvettes are brought to 30° for 5 min; the assay is initiated by the final addition of 10/xl of GSSG reductase solution (133 U/mE). The rate of 5-thio-2-nitrobenzoic acid (TNB) formation, which is proportional to the total GSH content, is followed at 412 nm; the reaction is linear for at least 10 min. Standards of GSH in the range of 0. I-2 nmol containing the same amount of SSA as the samples are prepared similarly. For initial experiments, oxidized glutathione (GSSG), and protein thiols (protein SH) are determined. GSSG is measured by the above DTNB-GSSG reductase method with prior masking of GSH in the supernatant by 2% (v/v) 2-vinylpyridine22; the mixture is adjusted to pH 6-7 by adding 10/zE of triethanolamine per 200/zl sample volume. The GSSG standards are prepared in a solution containing 2-vinylpyridine and triethanolamine. Under alE conditions, we find that the level of GSSG is less than 5% of the total glutathione; therefore, for most studies only total
2o j. F. Jongkind, A. Verkerk, and R. G. A. Baggen, Free Radical Biol. Med. 7, 507 (1989). 21 F. Tietze, Anal. Biochern. 27, 502 (1969). 22 M. E. Anderson, in "Handbook of Methods for Oxygen Radical Research" (R. A. Greenwald, ed.), pp. 317-323. CRC Press, Boca Raton, FL, 1988.
628
[62]
ANTIOXIDANT CHARACTERIZATION AND ASSAY 0 -X-
...o.............ro ..........
-10 "~•,
-20 o a ¢= a: &=
-)(- -X-
•,
• T
\T ".~ • "~, -)(--,'+ L\ "~',,T
-30 -40
-X--~
..............
o
\
-50 -60
L
\"! . . . .
T
~
-70
r
]. -)(- -.N-
-80 -10
i
i
i
i
i
i
i
0
10
20
30
40
50
60
70
Time (Minutes) FIG. 3. Time course of losses caused by free radicals in total glutathione (V), protein thiols (©), and viability (&) in endothelial cells. (*p <~ 0.05, **p ~< 0.01 vs values at time zero.) (Reproduced with permission from Mak et al. 6)
glutathione is determined. Protein SH is measured according to the method of Di Monte e t al. 23 After disruption by sonication in the presence of SSA, the cell pellets (derived from 0.5 ml cell suspensions) are washed twice with 1 ml of 6% TCA and resuspended in 2 ml of 0.5 M Tris buffer, pH 7.6, containing 100 txM DTNB; after 20 min of incubation at 30°, the absorbance is measured at 412 nm. Total glutathione levels vary from 3.5 to 6 nmol/106 cells among preparations of endothelial cells. For the purpose of comparison, losses of total glutathione are expressed as percentages of controls at time zero. Figure 3 presents the time course for losses of endothelial glutathione, protein SH, and cellular viability in the presence of free radicals. On addition of the free radical system, total glutathione decreased rapidly for the initial 30 min but at a slower rate afterward; in close association, the loss of cell viability also occurred in a parallel manner. In contrast, the loss of protein SH was only moderate at 30 min of incubation. As summarized in Table II, pretreatment of the cells with each agent dose-dependently protected against the losses of GSH and viability at 30 min of incubation. The order of potency of cytoprotective effects is 23 D. Di Monte, D. Ross, G. Bellomo, L. Eklow, and S. Orrenius, Arch. Biochem. Biophys. 235, 334 (1984).
[62]
629
ANTIOXIDANT PROPERTIES OF CALCIUM BLOCKERS
TABLE II PROTECTIVE EFFECTS OF CALCIUMCHANNEL BLOCKERSAGAINSTFREE RADICAL-INDUCED LOSSES OF ENDOTHELIAL GLUTATHIONE AND VIABILITYa
Conditions Buffer controls R. (DHF + Fe-ADP) R. plus calcium blockers: Nicardipine Nifedipine Verapamil Diltiazem
a
Concentration (/xM)
20 5 20 5 20 5 20 5
GSH loss (%)
Cell death (%)
4 45
12 61
7 21 9 24 11 26 15 29
20 27 25 31 28 40 35 42
The cells were preincubated with each calcium blocker for 15 min before the addition of DHF plus F e - A D P (R.). After 30 min of incubation, the losses of glutathione and cell viability were determined. Values are means of 3-8 determinations. Results adapted from Mak e t a l . , 6 with permission.
nicardipine > nifedipine > verapamil > diltiazem; this appears to follow their membrane antiperoxidative activities.
Comparison of Pharmacologically Active and Inactive Enantiomers of Nicardipine. Of all the calcium blockers tested, nicardipine (racemic form) is the most effective cytoprotective agent. In an effort to further TABLE III ANTIOXIDANT PROPERTIES OF ISOMERS OF NICARDIPINE AND BUTYLATED HYDROXYTOLUENE IN ENDOTHELIAL CELLSa
Conditions Buffer control R. (DHF + Fe-ADP) R' plus agent: ( + )-Nicardipine ( - )-Nicardipine BHT
Concentration (~M)
20 5 20 5 5
GSH loss (%)
Cell death (%)
2 47
11 59
8 24 11 26 7
19 28 22 31 16
Incubation conditions were as described in Table II. Values are means of 4-9 determination. Results adapted from Mak et al., 7 with permission.
630
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[62]
distinguish the nature of this cytoprotection, we have examined the effects of the pharmacologically active ( + ) and inactive ( - ) isomers of nicardipine. Owing to an asymmetric carbon at position 4 of the dihydropyridine ring, nicardipine has two optical isomers, namely, (+)- and (-)-nicardipine; the ( + ) isomer is about 5-fold more potent than the ( - ) isomer as a vasodilator or calcium channel blocker. 24 Both (+)- and (-)-nicardipine are supplied by Syntex Research Laboratory (Palo Alto, CA). When the membrane antioxidant activities of these two compounds are studied using the sarcolemmal membranes, both isomers display identical dose-dependent inhibitory activities against lipid peroxidation; the estimated ECs0 values are 20.6/xM for ( + )-nicardipine and 22.8/xM for ( - )-nicardipine. For comparison, the ECs0 value for BHT is 5.2/xM. Therefore the antioxidant potency of either nicardipine isomer was about one-fourth that of BHT. Using the endothelial cell system, the protection of the nicardipine isomers against free radical-induced losses of GSH and viability is evaluated. As summarized in Table III, both nicardipine isomers (5 and 20 /zM) provide comparable, concentration-dependent protection against the peroxidative depletion of GSH and loss of cellular viability. These results indicate that the cytoprotective effects of the agents are due to intrinsic chemical antioxidant activity rather than calcium channel blocking ability. Comments. The initial rate of endothelial GSH loss is proportional to the concentration of the free radicals generated; doubling the concentration of the free radical components leads to a 50% loss of GSH in 15 min of incubation. Linear regression analysis indicates that the loss of cellular viability is correlated significantly with the decrease of GSH (r = 0.89, p < 0.001). Lipid peroxidation at the cell level cannot be assessed accurately (compared to the isolated membranes); however, we believe that the loss of endothelial GSH is due to its oxidation by increased levels of lipid peroxides. The calcium blocker-mediated effects appear to be secondary to their inhibition of peroxide formation in the membranes. Our results indicate that the dihydropyridine calcium blockers exhibit the greatest antioxidant potency, which is partly due to their higher membrane partitioning and, perhaps, to their more active redox chemistry. It remains to be determined whether such antioxidant properties contribute to their established anti-atherogenic effects, z5 Acknowledgments This research was supported by National Institutes of Health Grants PO 1-HL-38079 and ROI-HL-36418. 24 K. Iwatsuli, F. Iijima,and S. Chiba, Clin. Exp. PharmacoL Physiol. 11, 1 (1984). 25 p. D. Henry, J. Cardiovasc. Pharmacol. 15~ $6 (1991).
[63]
ESR OF ETOPOSIDE(VP-16) PHENOXYLRADICAL
631
[63] I n t e r a c t i o n s o f P h e n o x y l R a d i c a l o f A n t i t u m o r D r u g , E t o p o s i d e , w i t h R e d u c t a n t s in S o l u t i o n a n d in Cell a n d Nuclear H o m o g e n a t e s : Electron Spin R e s o n a n c e and High-Performance Liquid Chromatography
By T. G.
G A N T C H E V , J. E. VAN LIER, D. A. STOYANOVSKY, J. C . Y A L O W I C H , and VALERIAN E . K A G A N
Introduction Etoposide (VP-16) is an antitumor drug currently in use for the treatment of a number of human cancers, including testicular and small lung cancers and lymphoma.1 Mechanisms of cytotoxicity of VP-16 may involve DNA strand cleavage and/or topoisomerase II inhibition accompanied by formation of DNA-topoisomerase cross-links and DNA strand breaks. 2,3 The VP-16 molecule contains a hindered phenolic group (Fig. 1) which is crucial for its antitumor activity. It has been reported that several enzymatic systems (peroxidase, tyrosinase, prostaglandin synthase, cytochrome P-450), as well as exogenous sources of peroxyl radicals (e.g., azo initiators) activate VP-16 via one-electron oxidation of the phenolic group to yield reactive metabolites (quinones) capable of irreversible binding to macromolecular targets (DNA and proteins) and/or generation of hydroxyl radicals in the presence of metal catalysts. 4-6 An essential step in the process of VP-16 activation is the formation of its phenoxyl radical which can be further either converted to oxidation products or reduced by intracellular reductants to the initial phenolic form. 7'8 Obviously, this may be critical for enhancing or suppressing the cytotoxic effects of VP16 in tumor cells'or surrounding tissues. The formation of the transient VP-16 phenoxyl radical in the course of its oxidative/reductive conversions can be directly followed by electron spin resonance (ESR) spectroscopy. Together with simultaneous highI M. L. Slevin, Cancer (Philadelphia) 67, 319 (1991). 2 N. Osheroff, Pharmacol. Ther. 41, 223 (1989). 3 L. A. Zwelling, Cancer Metastasis Rev. 4, 263 (1985). 4 B. Sinha and M. Trush, Biochem. Pharrnacol. 32, 3495 (1983). 5 j. M. S. Van Maanen, J. Retel, H. M. J. de Vries, and H. M. Pinedo, J. Natl. Cancer Inst. 80, 1526 (1988). 6 N. Haim, J. Nemec, J. Roman, and B. Sinha, Cancer Res. 47, 5835 (1987). 7 N. Usui and B. K. Sinha, Free Radical Res. Commun. 10, 287 (1990). 8 j. C. Yalowich, D. A. Stoyanovsky, W. P. Allan, B. W. Day, and V. E. Kagan, Proc. Am. Assoc. Cancer Res. 34, 1779 (1993).
METHODS IN ENZYMOLOGY, VOL. 234
Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.
632
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[63]
H H3C"~ 0 HO
0 0
o
OH Fro. 1. Structural formula of etoposide (VP-16).
performance liquid chromatography (HPLC) measurements of the concentrations of VP-16 and its oxidation product(s), this may give detailed information on the interaction of VP-16, its intermediates, and oxidation product(s) with metabolizing enzymes and other intracellular and exogenous oxidants as well as with physiologically relevant reductants in model chemical and biological systems. Several examples below illustrate ESR detection of VP-16 phenoxyl radicals generated by different enzymatic and nonenzymatic systems as well by light-activated reactions.
Methods
Generation of VP-16 Phenoxyl Radicals by Enzymatic and Nonenzymatic Systems and Electron Spin Resonance Detection Generation of VP-16 Phenoxyl Radical by Tyrosinase and Peroxidase. Tyrosinase, a copper-containing enzyme widely distributed in biological systems, can hydroxylate phenols to form catechols and o-quinones. 9 Peroxidative attack on the phenolic moiety of VP-16 by tyrosinase in airsaturated phosphate-buffered aqueous solution (pH 7.4) results in generation of a typical phenoxyl free radical as detected by ESR spectroscopy 7,1° 9 S. H. Pomerantz, J. Biol. Chem. 241, 161 (1966). 10 B. Kalyanaraman, J. Nemec, and B. K. Sinha, Biochemistry 28, 4839 (1989).
[63]
ESR OF ETOPOSIDE(VP-16) PHENOXYLRADICAL
633
(Fig. 2A). The oxygen-centered phenoxyl radical exhibits a hyperfine structure similar to semiquinone radicals, but is relatively more stable and does not require additional metal ion or alkaline stabilization. In aqueous solution the twenty-line ESR spectrum is characterized by the following magnetic resonance parameters: g = 2.0048 -+ 0.0002 and hyperfine splitting arising from six identical methoxy protons (aHcH3 -----0.14 mT), two phenoxyl ring protons (ar~ng = 0.14 mT), one fl-proton (a H = 0.45 mT), and a y-proton (a H = 0.06 mT). Horseradish peroxidase generates the VP-16 phenoxyl radical of the same structure, but requires the presence of hydrogen peroxide 1° (Fig. 2B). Generation of VP-16 Phenoxyl Radical by Azo Initiators of Peroxyl Radicals. Azo compounds are known to decompose unimolecularly and nonezymatically to yield N2 and two identical carbon-centered radicals, which further react with molecular oxygen to form corresponding peroxyl radicals, ll Both hydrophilic [e.g., 2,2'-azobis(2-amidinopropane)dihydrochloride (AAPH)] and lipophilic [e.g., 2,2'-azobis(2,4-dimethylvaleronitrile) (AMVN)] azo initiators of peroxyl radicals can be used to induce oxidation in the aqueous phase or in the membranous phases, respectively.11 The practical advantage of these compounds as peroxyl radical initiators follows from their constant spontaneous decomposition rate (radical yield) at physiological conditions. Peroxyl (alkoxyl) radicals readily abstract the hydrogen atom from phenols to produce phenoxyl radicals.~2 Oxidation rates of VP-16 by AAPH in aqueous solution or by AMVN in dioleoylphosphatidylcholine liposomes at 37° are high enough to produce steady-state concentrations of the VP-16 phenoxyl radical easily detectable by ESR (Fig. 2C,D). AAPH and AMVN decomposition can be induced by UV-irradiation, causing decomposition of the azo bond (h -> 320 nm). Thus UV-irradiation of VP-16 in the presence of an azo initiator can also be used for generating ESR-detectable steady-state concentrations of the VP-16 phenoxyl radicals (not shown). Photosensitized Generation of VP-16 Phenoxyl Radicals. Metallophthalocyanines (MePc) and their sulfonated water-soluble derivatives [e.g., aluminum tetrasulfophthalocyanine (A1PcS4) or zinc tetrasulfophthalocyanine (ZnPcS4)] are red-light-absorbing photosensitizers with a promising application in the so-called photodynamic therapy of tumors.~3 Triplet-state excited phthalocyanines initiate photooxidation of biological substrates either via direct electron/hydrogen atom abstraction from subII E. Niki, this series, Vol. 186, p. 100. 12 G. W. Burton and K. U. Ingold, Acc. Chem. Res. 19, 194 (1986). is j. E. Van Lier and J. D. Spikes, in "Photosensitizing Compounds: Their Chemistry, Biology and Chemical Use" (G. Bock and S. Harnett, eds.), p. 17. Wiley, Chichester, 1989.
g=2.0048 Gain
lx10 2
A '
~
mT
B
lx103
, 1.1
J,I
C
D
E
z4
FIG. 2. ESR spectra of the VP-16 phenoxyl radical generated by enzymatic and nonenzymatic systems. Tyrosinase-induced VP-16 oxidation: 50 mM phosphate buffer (pH 7.4) containing 0.5 mM VP-16 and 160 U mushroom tyrosinase (A); peroxidase-induced VP-16 oxidation: horseradish peroxidase (8/zg/ml), H2O2 (2 mM), and VP-16 (0.2 mM) in acetate buffer (pH 5.0) (B); VP-16 oxidation induced by a water-soluble azo initiator 2,2'-azobis(2amidinopropane) dihydrochloride (AAPH, 250 raM) (C); VP-16 oxidation induced by a lipidsoluble azo initiator 2,2'-azobis(2,4-dimethylvaleronitrile)(AMVN, 60 mM) incorporated into dioleoylphosphatidylcholinetiposomes (10 mg/ml) (D); metallophthalocyanine-photosensitized oxidation of VP-16: aluminum tetrasulfophthalocyanine (A1PcS4, 0.1 mM), continuous irradiation with visible-light(~. -> 400 nm; P = 15 mW/cm2) of 1 mMVP-16 in phosphate buffer (pH 7.4) containing 2 mM sodium dodecyl sulfate (SDS) (E).
[63]
ESR OF ETOPOSIDE(VP-16) PHENOXYLRADICAL
,/
~1 t
635
0 min 5 min
~----- 30 min
~
45 min
'75 mirr~n '90 min 0
7 Retention time
i 14
FIG. 3. HPLC tracings of VP-16 and its oxidation product formed in the course of incubation with tyrosinase in phosphate buffer. Incubation conditions: VP-16 (0.5 raM) and tyrosinase (160 U) in 0.1 M phosphate buffer (pH 7.4 at 25°).
strate molecules or via generation of activated oxygen species: singlet oxygen (102) , superoxide anion (O20 and hydroxyl radicals (-OH). 14,15 Irradiation of air-saturated aqueous solutions containing AIPcS4 (or ZnPcS4) and VP-16 with visible light (h -> 400 nm) results in generation of the characteristic VP-16 phenoxyl radical as followed by the direct photoESR measurements (Fig. 2E). The reactive intermediate(s) involved in VP-16 oxidation by PcS4 remain unknown (possibilities include activated oxygen species and/or direct triplet state dye-ground state VP-16 interaction). 14 R. Langlois, H. Ali, N. Brasseur, R. J. Wagner, and J. E. van Lier, Photochem. Photobiol. 44, 117 (1986). 15 T. G. Gantchev, M. G. Kaltchev, and G. P. Gotchev, Int. J. Radiat. Biol. 60, 597 (1991).
636
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[63]
A
Gain
t(min)
4
_._._z
28
lx100 ' t 32
FIG. 4. ESR spectra of the tyrosinase-induced radicals in the presence of VP-16 and reductants (A) ascorbate and (B) reduced glutathione (GSH). Incubation conditions: 50 mM phosphate buffer (pH 7.4) containing 0.5 mM VP-16, 160 U mushroom tyrosinase, and 0.8 mM ascorbate (A) or 0.1 mM GSH (B). The prescan incubation time is shown in the right column; the receiver gain is shown in the left column.
(63]
E S R OF ETOPOSIDE (VP-16) PI-IENOXYL RADICAL
B Gain
t(min)
lx100
I mT -
-
-
-
-
. . . . . . . . .
.
,
12
2O
FIG. 4.
(continued)
637
638
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[63]
Reduction of VP-16 Phenoxyl Radical Studied by Electron Spin Resonance and Chromatography in Aqueous Solution Physiologically important reductants with appropriate redox potentials, such as reduced glutathione (GSH) and ascorbate are capable of donating electrons (hydrogen atoms) to phenoxyl radicals, thus regenerating phenols at the expense of their own oxidation. For hindered phenols, this regeneration process can be followed directly by ESR spectroscopy. In particular, reduction ofphenoxyl radicals of vitamin E and its homologs, the food preservatives butylated hydroxytoluene (BHT) and hydroxyanisole (BHA), and some phenolic drugs (e.g., probucol, acetaminophen) has been demonstrated by E S R . 16'17 Reduction of the VP-16 phenoxyl radical by intracellular reductants may prevent its activation to a cytotoxic species. We have used ESR spectroscopy to study interactions of the VP-16 phenoxyl radicals with some physiologically relevant reductants (reduced glutathione, ascorbate)/8 In these experiments, we use tyrosinase as the VP-16 phenoxyl radical generating system, which permits persistent detection of the characteristic ESR signal for 50-60 min. HPLC measurements show that incubation with tyrosinase resulted in a timedependent consumption of VP-16 and accumulation of its oxidation product(s) (Fig. 3). Thus, the effects of reductants should be conveniently evaluated both by ESR and HPLC measurements. Tyrosinase is not likely to cause secondry reactions, affecting the stability of the phenoxyl radical, as has been reported for horseradish peroxidase. 10In addition, tyrosinase is possibly one of the major VP-16 metabolizing enzymes in a number of malignant cells, especially in melanomas. 6'~9 When ascorbate is added to the incubation medium, the characteristic doublet of the semidehydroascorbyl radical could be observed in the ESR spectra unless ascorbate is consumed. This is followed by a rapid appearance and growth of the VP-16 phenoxyl radical ESR signal (Fig. 4A). Replacing ascorbate by reduced glutathione (100 /zM GSH) under the same conditions results in several "blank" ESR scans, followed by a less steep (relative to the case of ascorbate) growth of the VP-16 phenoxyl radical ESR signal (Fig. 4B). In both cases, in the presence of ascorbate or GSH, the buildup of the VP-16 phenoxyl radical ESR signal after the lag period reaches a saturation intensity level nearly the same as in the 16 V. E. Kagan, E. A. Serbinova, and L. Packer, Arch. Biochem. Biophys. 280, 33 (1990). 17 R. D. N. Rao, V. Fischer, and R. P. Mason, J. Biol. Chem. 265, 844 (1990). ~s D. Stoyanovsky, J. Yalowich, T. Gantchev, and V. Kagan, Biochem. Pharmacol., submitted for publication. 19 D. H. Kern, R. H. Shoemaker, S. U. Hildebrand-Zanki, and J. S. Driscoll, Cancer Res. 48, 5178 (1988).
[63]
ESR OF ETOPOSIDE(VP-16) PHENOXYL RADICAL
639
A [Asc],pM
t-.m c o~ £E CO UJ
j/
t(min)
I
I
0
I
10
20 [GSH],pM
(D
E UJ
~ 0
1,
210
t(min) I 30
FIG. 5. Direct ESR measurements of the kinetics of the VP-16 phenoxyl radical buildup in the presence of different concentrations of (A) ascorbate and (B) reduced glutathione. Conditions: Magnetic field lock at 336.3 mT; modulation amplitude 0.32 mT; and scan time 64 min. For more details, see text. absence of reductants. Quenching of the VP-16 phenoxyl radical E S R signal by reductants and its reconstitution after reductant depletion are reversible by m e a n s of repetitive addition of reductants to the solution after their consumption. In their oxidized form, neither d e h y d r o a s c o r b a t e nor G S S G causes a lag period in the a p p e a r a n c e of the VP-16 phenoxyl radical E S R signal (not shown). The lag periods produced by reductants
640
[63]
ANTIOXIDANT CHARACTERIZATION AND ASSAY 0.6-
VP-16 (control) z
.-
~
"r
I
~
x
T
~
~
+ Tyrosinase
¢.,
o
0.3.
~¢~
012'
tO
> 0.1 " 0.0
0
+ Tyrosinase 1~, ~ I
20
+Tyrosinase + GSH
10
8b
Time, rain FIG. 6. Effect of ascorbate and GSH on the kinetics of the tyrosinase-induced VP-16 oxidation as measured by HPLC. Concentrations of ascorbate and GSH were 0.8 and 0.15 mM, respectively.
during VP-16 oxidation can be conveniently recorded by kinetic measurements of the VP-16 phenoxyl radical ESR signal by locking the magnetic field. Typical kinetic scans obtained in the presence of different concentrations of ascorbate or GSH are presented in Fig. 5. The lag periods induced by GSH are longer than those induced by ascorbate under equivalent conditions. This is most likely due to a tyrosinase-catalyzed slow oxidation of ascorbate in the absence of VP-16. The HPLC tracings demonstrate that no decay of VP-16 is observed until ascorbate is completely consumed. After depletion of ascorbate the rate of VP-16 oxidation is the same as in the absence of ascorbate (Fig. 6). Similarly, in the presence of GSH, VP-16 oxidation proceeds with a lag period during which the concentration of VP-16 does not change (Fig. 6). The durations of lag periods for the appearance of the VP-16 phenoxyl radical signal in the ESR spectra in the presence of ascorbate or GSH are exactly the same as the lag periods observed for the onset of VP-16 oxidation in HPLC tracings. Neither ascorbate nor GSSG affects the time course of VP-16 oxidation by tyrosinase (data not shown).
Interactions of Endogenous Reductants with VP-16 Phenoxyl Radical in Presence of Cell and Nuclear Homogenates A delicate balance between oxidants (including oxidative enzymes) and reductants may predetermine the prevalence of oxidative over reductive
[63]
ESR OF ETOPOSIDE(VP-16) PHENOXYLRADICAL
~ 0i
.-~"" 2 , 10
3
"3
2b
641
t(min) l 30
4~
0
FIG. 7. ESR spectra of tyrosinase-induced radicals in the presence of VP-16 and K-562 cell nuclear homogenates. K-562 cell homogenates were prepared from (1) 0.3 x 106, (2) 0.6 x 106, and (3) 0.9 × 106, cells isolated from a culture at the mid-exponential growth; nuclear homogenates were prepared from I × 106 nuclei (4). ESR conditions were as in Fig. 5.
metabolism of VP-16 in specific intracellular compartments. Shifts in this balance may be at least partially responsible for the drug resistance of tumor cells via mechanism(s) involving reduction of the VP-16 phenoxyl radical. The redox balance toward the VP-16 phenoxyl radical can be evaluated by kinetic measurements of the tyrosinase-induced ESR signal of the VP-16 radical in the presence of cell homogenates. When tyrosinase and VP-16 are incubated in the presence of homogenates prepared from different numbers of K-562 human leukemia cells (Fig. 7, curves 1-3) or from nuclei isolated from these cells (curve 4, Fig. 7), the phenoxyl radical buildup to an ESR-detectable concentration follows a lag period with a subsequent relatively steep increase in the magnitude of the ESR signal. In nuclear homogenates the lag period is
10 tiM GSH
.__>,
~
.E
nuJ
I
k
I
I
0
10
20
30
r
FIG. 8. Effect of exogenous GSH added to K-562 cell homogenate after depletion of endogenous antioxidants on kinetics of the VP-16 phenoxyl radical ESR signal. K-562 cell homogenates were prepared from 0.5 × 106 cells; the GSH concentration was 10/.tM.
642
ANTIOXIDANT CHARACTERIZATION AND ASSAY
[63]
shorter as compared to the cell homogenate prepared from an equivalent number of cells. In K-562 cell and nuclear homogenates no semidehydroascorbyl radical ESR signals are detected during the lag period (preceding the appearance of the VP-16 phenoxyl radical ESR signal) as is observed in a model system on addition of ascorbate; hence, ascorbate is not likely to be the major intracellular reductant responsible for the VP-16 regeneration. The slow kinetics of the VP-16 phenoxyl radical ESR signal after a lag period are similar to those observed in a model system in the presence of GSH. Exogenously added GSH causes a rapid quenching and a slow reappearance of the VP-16 phenoxyl radical ESR signal if added to K-562 cell homogenates after the initial lag period is over (Fig. 8). Thus, it seems likely that reduced thiols may primarily contribute to the VP-16 regeneration in K-562 cell homogenates. Addition of ascorbate to K-562 cell or nuclear homogenates after completion of the lag period results in the same kinetics of quenching and abrupt reappearance of the VP-16 phenoxyl radical ESR signal after ascorbate oxidation as is observed in a model system (not shown). HPLC measurements reveal a lag period in a tyrosinase-catalyzed VP16 oxidation in the presence of K-562 cell or nuclear homogenates. They correspond to the lag periods for appearance of the VP- 16 phenoxyl radical ESR signal: the onset of the VP-16 oxidation coincides in time with the appearance of the VP-16 radical ESR signal. In conclusion, the procedure we have developed for measuring the duration of the lag periods for the VP-16 phenoxyl radical appearance in cell homogenates can be used for convenient evaluation of cell reductive capacity. Intracellular reductants may eliminate activation of VP-16 by oxidative metabolism via mechanisms involving reduction of the VP-16 phenoxyl radical. Deliberate depletion of intracellular reductants (e.g., by photosensitizers generating reactive oxygen species, azo initiators of peroxyl radicals, overexpression of oxidative enzymes) may be used in the future to enhance the cytotoxic effects of VP-16 in tumor cells.
AUTHOR INDEX
643
Author Index Numbers in parentheses are footnote reference numbers and indicate that an author's work is referred to although the name is not cited in the text.
A Aakvaag, A., 591 Abate, C., 164, 165(1, 2), 166(2, 7), 167, 167(2), 168(2, 16), 174(7), 224 Abbas, A. K., 141 Abbott, W. A., 493 Abdallah, M., 443 Abdallah, M. A., 441 Abe, K., 274 Aberg, F., 344 Aboujaoude, E. N., 16 Abrams, J., 491 Aburada, M., 415,416(31) Ackrell, B.A.C., 317, 354, 381 Adam, W., 79, 129 Adams, D. A., 49 Adams, G. E., 426 Adams, M., 592 Addis, P. B., 439 Adler, K. B., 257 Ae Shin, Y., 68 Afanas'ev, I. B., 421 Affany, A., 429 Aftonomes, B. T., 257 Aggarwal, B. B., 461 Aggarwar, B. B., 156 Ahn, B. W., 254 Ahn, M. S., 582, 587(17), 589(17) Ahnadi, C., 591 Ahnfelt-Ronne, I., 557, 569(22), 570(22) Ahnstrfm, G., 67, 78(18), 89, 94(9, 14), 96(9, 14), 100(9, 14), 102 Aho, P., 527 Ahokas, J., 477, 478(13) Akatsuka, N., 491 Akerboom, T.P.M., 367, 368(9), 370 Akeson, A. L., 507 Akman, S. A., 3 Alaupovic, P., 515 Albermann, K., 152, 154(4), 156(4)
Albiero, R., 401 Albroscheit, G., 415,416(28) Alcaraz, M. J., 421, 446, 449(15), 450(15), 453,453(14, 15), 454(14, 21) Alessio, H., 24, 31(27) Alfassi, Z. B., 428 Ali, H., 81,635 Alkner, N., 253,256(15) Allan, W. P., 631 Allen, A., 256 Allen, L. A., 604 Allen, R. E., 555 Allgayer, H., 557 Almeda, S., 411 Alperin, P. E., 24 Althaus, J. S., 552 Altmann, A., 149 Altmiller, D. H., 60, 62(6), 80 Altuvia, S., 217, 219 Ames, B. N., 16-17, 22, 24, 24(9), 25(25), 26(25), 27(9), 32(9), 59, 67, 79-80, 88, 117, 151, 156(1), 175(3), 176, 181(3), 187, 214, 216(18), 220,224, 253-254, 254(11), 255(11), 269, 273, 331, 344, 348(7), 375,406, 619 Amici, A., 254 Amstad, P., 191, 192(22) Amstad, P. A., 187, 191(19), 203(19) An, J., 582 Andersen, L., 421 Anderson, D. K., 568, 569(45), 619 Anderson, J. M., 618 Anderson, M. E., 492-493, 493(2, 3), 494, 494(16), 496(2, 21), 497, 497(2, 21), 498(21, 29), 499, 499(29), 500, 501(1), 502,502(1), 503,503(1, 3), 504, 504(3, 5), 627 Anderson, R.G.W., 606 Anderson, V. C., 604 Andersson, C. M., 479 Ando, Y., 341
644
AUTHOR INDEX
Andreef, M., 592 Andrus, P. K., 552 Andus, K. L., 553 Angel, P., 170 Angelov, D., 3 Ankel, E. G., 580 Antus, S., 422, 425(26) Aoki, K., 68 Applegate, L. A., 225 Arashima, S., 547 Arcioni, A., 601 Arichi, S., 446 Ariga, T., 421,423(22) Armstrong, D., 404 Armstrong, D. W., 394 Armstrong, R. N., 505 Arnit, B., 250, 251(6) Arnstein, H.R.V., 591-592, 592(32, 33), 594(33, 37, 39), 595(32, 37, 39), 596(33, 37, 39), 597(33, 39), 598(47, 48), 599(48), 600(47), 601(33, 37, 39, 47, 48) Aronovitch, J., 586, 589(27) Arroyo, C. M., 620, 625, 625(3) Arshad, M.A.Q., 283 Artz, W. E., 397 Arudi, R. L., 421 Aruoma, O. I., 3, 17, 67, 70(15), 73(15), 75(15), 76, 76(15), 489, 557, 571(18) Arvidsson, S., 528 Asada, K., 378, 383,459 Asahara, H., 123 Ascarelli, I., 424 Ascherio, A., 269 Asensi, M., 367, 371 Asmus, K. D., 422,426, 479 Asquith, T. N., 432,435 Astengo, M., 490 Astor, M. B., 494 Atherton, S. J., 68 Attaway, J. A., 416 Aubert, M. C., 397 Aubry, J. M., 385 Auclair, C., 530 Audic, A., 17, 80 Aufderheide, E., 89 Augeri, L., 35, 102, 110 Auld, D. S., 52 Auld, P.A.M., 492(9), 493, 494(9) Aust, S. D., 66, 67(2), 566, 569(39), 570(39) Ausubel, F. M., 221
Avron, M., 400 Awai, M., 341 Azad-Khan, A. H., 557
B Babulovfi, A., 578 Backer, U., 491 Backstr6m, R., 527 Bacon, B. R., 438 Bacon, K. B., 452 Bacon, P. A., 569 Badley, J. E., 194 Baeuerle, P. A., 151-152, 153(3), 154, 154(4), 155(3, 10), 156(4, 5), 157(3, 5, 10), 158(5), 159-160, 161(15), 162(3, 10), 163, 163(10), 217, 224 Baffet, G., 438 Baggen, R.G.A., 627 Baggio, G., 511 Bagnasco, M., 490 Bailey, J. M., 514 Bailly, V., 111, 113, 113(12) Baiocchi, M. R., 511 Bakac, A., 426 Baker, J. C., 375 Baker, P. F., 311,312(8), 313(8), 314(8) Baker, R. R., 256 Bakerian, R. H., 123 Bakker, J. H., 557 Balser, L., 514 Baltimore, D., 153, 159, 162(9), 163, 172 Bank, J. F., 123 Bannai, S., 137, 139(14, 15), 148 Banoun, H., 530 Baranes, L., 462 Barber, A. A., 280 Barclay, L.R.C., 269, 374,423,506, 533,543 Barden, H. B., 590 Barenholz, Y., 324 Barindelli, E., 487 Barja, G., 331 Barja da Quiroga, G., 332, 333(4), 334(4), 337(4) Barklay, J.R.C., 403 Barlow, D. J., 592,598(47), 600(47), 601(47) Bamhart, J. W., 505 Barnhart, R. L., 505, 506(4) Barr, R., 375
AUTHOR INDEX Bartels, K., 478 Barth, S. A., 462 Bartlett, P. D., 605 Barton, D., 414 Basset, P., 602 Bassett, J. M., 603 Bast, A., 362, 455,478,489 Bastide, P., 421 Basu-Modak, S., 224 Bates, A. L., 385 Battey, J. F., 226 Bauer, V., 572, 573(1) Bauernfeind, J. C., 388 Bauernfidd, J. C , 355 Baum, M., 590-591 Baumann, J., 411,421 Baumgarden, C., 253, 256(15) Baumstark-Khan, C., 88-89, 90(12), 91(12), 94(12), 96(12), I00, 100(12) Beaune, P., 592 Beck, Y., 250, 251(6) Becker, J. T., 524 Beckman, J. S., 475 Beckman, K. B., 22 Beecher, G. R., 396, 398 Beehler, B. C., 24, 36, 38(19) Behrend, R., 4 Behrens, G., 426 Behrens, W. A., 270, 271(12) Beier, H., 185 Beli, J., 590 Belkin, S., 580 Bell, S. D., 165 Bellocq, J. P., 602 Bellomo, G., 73, 77, 628 Bement, W. J., 438 Ben Amotz, A., 400 Benard, F., 81, 83(23) Ben Aziz, A., 424 Bendich, A., 4, 270 Beneg, L., 572-573, 573(1), 574, 576-578, 578(3), 579(31) Bennett, M.V.L., 236 Bennicelli, C., 490 Benninghoff, B., 135, 140(5), 142(5), 144(5) Be'nsasson, R. V., 480 Bentsath, A., 429 Berestecky, J. M., 236, 241(14) Berger, M., 3, 55, 80-82, 83(23, 25, 29), 85(25), 87, 87(25)
645
Berggren, M., 479, 487-489, 489(8) Bergsten, P., 337 Bergtold, D. S., 4, 24, 31(27), 131 Berliner, L. J., 580 Berman, E., 592 Bernard, G. R., 491 Bernhard, K., 390, 394(14) Bernier, J. L., 421,423(13), 425(13) Bernstein, L., 590 Berstock, D., 590 Berthou, F., 592 Bertovich, M. J., 557 Bertram, J. S., 235-237, 238(17), 239(8), 241(9, 14), 243(9) Bertrand, M., 423 Bessho, T., 24 Betts, W. H., 557 Beutler, B., 244 Beyer, E. C., 236 Bezek, 5., 574 Bhadra, S., 283 Biasin, M. R., 100 Biber, T.U.L., 458 Bidani, A., 253,256(9) Bieber, L. L., 536 Biedermann, J., 477 Bielski, B.H.J., 421,622 Biemond, P., 561,566(30) Bieri, J. G., 397 Bignon, E., 591 Bindoli, A., 425 Bing, D. H., 411 Birnboim, H. C., 67, 73(17), 77(17), 78(17), 89, 90(10, 13), 91(10, 13), 94(10, 13), 96(10, 13) Bischoff, F., 189 Bishop, G. A., 194 Bishop, J. E., 603 Bittner, D., 125 Blackett, P. R., 516 Blake, D. R., 555,569 Blakely, W. F., 3, 4(8), 5-6, 6(8), 9(32), 11(28) Blamey, R. W., 590 Blanchard, J. M., 231 Blanchard, J. S., 612 Blanchflower, W. J., 299 Blank, V., 161 Blasig, I. E., 422,423(34), 425(34) Blattner, F. R., 176, 216(8)
646
AUTHOR INDEX
Blavet, N., 475 Blazovics, A., 422,425(26) Bligh, E. G., 609 B1fcher, D., 100, 101(26) Block, G., 269 Blouin, F. A., 415,416(25) Blum, H., 185 Bockstette, M., 135 Boehme, P., 620, 628(6), 629(6) Boisnard, M., 438 Boiteux, S., 3, 10, 16, 16(14), 82, 117, 123, 125, 129-130 Bokn~ov~i, B., 578 Boli, R., 523 B61ker, M., 220 Bolle, A., 218 Bolli, R., 515, 523(23) Bolme, P., 487 Bolognesias, C., 96 Bolton, D. C., 254, 257 Bolton, P. H., 113 Boman, G., 491 Bomford, A., 543,544(13) Bonfanti, R., 296 Bongiorno, A., 401 Bonhomme, B., 462 Bonkowsky, H. L., 438 Bonner, W. M., 50 Bonomi, F., 457 Bonomo, R. P., 421 Boobis, A. R., 491 Boozer, C. E., 312,403 Borek, C., 252 Borel, J. P., 421 Borgstr6m, L., 487 Borish, E. T., 619 Borner, C., 591 Borrello, S., 601 Bors, W., 384, 420, 422, 424(24), 425, 425(24, 25), 426(25, 65), 427(25, 65), 428,428(24, 25), 429(65), 437 Borst, P., 603 Bothe, E., 426 Botnen, J., 278 Bottazzi, B., 602 B6ttcher, C.J.F., 616 Bourel, D., 438 Bourel, M., 438 Bourgeois, C., 295 Bourgeois, C. F., 299
Bourre, F., 45, 80, 119 Bourrier, M. J., 416 Boveris, A., 66, 437 Bowry, V. W., 354, 372 Box, H. B., 24, 36, 38(19) Boyer, V., 79 Bradford, M. M., 441 Bradley, K., 136 Brand, J. G., 616 Brankow, D. W., 237 Braquet, P., 421,423,462, 475 Brasch, R. C., 580 Brasseur, N., 81,635 Brattsand, R., 479 Braughler, J. M., 376, 402, 548, 550-553 Braun, A. G., 34 Braun, A. M., 384 Braunstein, S. N., 125 Bray, T. M., 525 Breimer, L. H., 34, 40, 235 Bremner, I., 546 Brent, R., 221 Brewer, G. J., 545-546, 547(18) Bridge, R. J., 493 Bridgeman, M.M.E., 488 Bridges, R. J., 493 Brigelius, R., 370 Briggs, M. R., 165 Brin, M., 327 Brissot, P., 437-438,441,443 Britigan, B. E., 561,566(31) Brittenham, G. M., 438 Briviba, K., 384, 574, 578(13) Brodskii, A. V., 421 Brodsky, M. H., 375 Brooks, G. A., 368.371(13) Brotherton, J. E., 435 Brown, D. M., 40, 44(27) Brown, E. D., 397 Brown, M. S., 606 Brown, R. A., 446 Brownlie, I. T., 581 Brownstein, B. H., 202 Brubacher, G., 295 Bruchelt, G., 337 Bruchhausen, F. V., 411 Bruener, B. A., 253, 254(11), 255(11) Bruice, T. W., 77 Brun, M., 416 Bruning, P. F., 592
AUTHOR INDEX Brunner, K. T., 135 Bruno, A., 415,416(19, 21), 418(19), 419(19) Bruno, G., 415,416(32) Bryan, C. L., 551 Bryant, P. E., 100, 101(26) Buchko, G. W., 79, 81-82, 83(25), 85(25), 87, 87(25) Buckpitt, A. R., 258 Budowski, P., 424 Buege, J. A., 566, 569(39), 570(39) Buettner, G. R., 70-71, 71(32), 375 Bujard, H., 166, 167(14) Bukley, G. B., 619 Bunnell, R. H., 295 Buran, L., 578 Burgunder, J. M., 488 Burk, R. F., 316 Burkitt, M. J., 66-67, 70, 70(16), 71, 71(16, 36), 72(38), 73, 73(16), 74(16), 75(16), 76, 76(16), 77(35) Burlakova, E. B., 345 Burlakova, Y. B., 356 Burr, J. A., 312-313,313(14), 315(12) Burtiss, J. L., 295 Burton, G. W., 269, 280, 286(1), 303-304, 310, 310(2), 316, 322,331,355-356, 362, 362(8), 372, 381(3), 403, 406, 506, 533, 543-544, 601,633 Burton, K., 35, 37(17), 40(17), 104 Busch, S. J., 505,506(4) Bush, K. M., 273,475 Butler, J., 68, 69(27), 489, 557, 571(18) Butler, L. G., 429,432,434-435,435(12, 15) Butta, A., 591 Bylund-Fellenius, A. C., 582, 586 C Cabrini, L., 344 Cadenas, E., 66, 79, 468, 476, 479(1), 480(1) Cadet, J., 3, 17, 55, 79-81, 81(12), 82, 83(23, 25, 29), 85(25), 87, 87(25) Cahill, D. S., 17 Caldarera, M., 373 Caldwell, K. A., 477-478, 47804) Caleffi, M., 590 Camerini-Otero, R. D., 125 Cameron, L., 590 Campbell, C. R., 162
647
Campbell, E., 496 Campbell, W. B., 514 Campion, J. P., 438 Canada, A. T., 424 Canney, A. H., 411 Cannon, M., 591-592, 592(32, 33), 594(33, 37, 39), 595(32, 37, 39), 596(33, 37, 39), 597(33, 39), 598(47, 48), 599(48), 600(47), 601(33, 37, 39, 47, 48) Cannone, P., 385 Cantilena, L. R., 393 Cantin, A. M., 256 Cantineau, R., 477 Cantley, J. C., Jr., 410 Cantoni, O., 77 Cao, W., 524 Carbone, P. P., 590 Carew, T. E., 505, 513-514 Carey, F., 446, 449(15), 450(15), 453,453(15) Carey, J., 153, 160(8), 165, 171(11) Carini, M., 415, 416(22) Carlin, G., 557, 582, 586 Carlsson, D. J., 384 Carney, J. M., 515, 523-524, 524(1), 5250), 526(1) Carri6re, I., 475 Carrillo, M.J.H., 321-322, 322(12), 323(14), 325(14), 326(14) Carter, D. E., 542 Carvalho, A. P., 591 Carven, P. A., 557 Catania, J., 82 Cathart, R., 331 Cathcart, K.N.S., 494 Cathcart, M. K., 513-514 Catignani, G. L., 394, 397 Cattabeni, F., 77 Catteau, J. P., 421,423(13), 425(13) Cavalieri, L. F., 4 Cavatorta, P., 601 Ceasrone, C. F., 96 Celeste, M., 591 Celotti, L., 100 Cerimele, B. J., 505 Cerottini, J.-C., 135 Cerutti, P., 77, 191, 192(22) Cerutti, P. A., 88, 187, 191(19), 203(19), 235 Cesarone, C. F., 489-490, 490(27) Ceva, P., 415, 416(29) Chait, A., 506, 514
648
AUTHOR INDEX
Chambon, P., 602 Chamulitrat, W., 70 Chan, B. C., 433 Chan, G. L., 35, 37(14), 103 Chan, H. C., 580 Chanal, S., 591 Chance, B., 66, 344 Chander, C. L., 590 Chandler, L. A., 396 Chapelat, M., 462 Chaplin, D. D., 202 Chase, R. L., 548, 550,553 Chatopadhyaya, R., 51 Chatterjee, A., 381 Chaudary, M. A., 590 Cheek, J. M., 258 Cheeseman, K. H., 601,616 Chen, C.-H., 166, 167(13) Chen, G., 525 Chen, Q., 16 Chen, S.-L., 68 Chen, W.-J., 505 Chenard, M. P., 602 Cheng, K. C., 17 Cheng, M. S., 523,524(1), 525(1), 526(1) Cherian, M. G., 547 Chevion, M., 524, 544 Chi, E. M., 505(8), 506, 509(8), 512(8) Chiba, S., 630 Chida, H., 322 Chiesa, G., 515 Ching, T.R.A.M.I.P., 423 Chipault, J. R., 300, 356 Chirico, S., 254 Chisolm, G. M., 513-514 Chiu, R., 170 Chiu, S., 3 Chiu, S.-m., 78 Choi, D. W., 552 Choisy, H., 511 Chomczynski, P., 193,227 Chong, S., 491 Choo, Y. M., 528 Chow, C. K., 621 Christ, B., 416, 417(40) Christensen, A., 557, 569(22), 570(22) Christiaens, L., 477 Christian, M. F., 214 Christman, M. F., 175(3), 176, 181(3) Chuang, S. E., 176, 216(8)
Chung, M.-H., 24, 80 Church, D. F., 252, 506 Churley, M., 237 Ciaccio, M., 401 Cilento, C., 79 Cillard, J., 421,437,441 Cillard, P., 421,437,441 Claeys, M., 416 Claiborne, A., 227 Clark, G. R., 68 Clarke, R., 601 Clarkson, B., 592 Claycamp, H. G., 19, 20(13) Clayton, D. A., 42 Cleland, L. G., 557 Cleland, W. W., 612 Clemetson, C.A.B., 421 Climent, I., 254 Clostre, F., 462,475 Clostre, P., 462 Clough, R. L., 314 Coassin, M., 425,478,479(20), 480(20) Cochrane, C. G., 77 Cogrel, P., 437 Cohen, G., 542, 564, 566(34) Cohen, L. A., 314 Cohen, M., 77 Cohen, M. S., 561,566(31) Cohen, N., 309 Cohen, R. M., 283 Cohen, S. S., 5 Colditz, G. A., 269 Collins, A. R., 77, 78(52) Collins, M., 601 Combs, G. F., 327 Conn, P. F., 385, 387(14) Connor, J. A., 236 Constantinescu, A., 363(34), 366 Cook, H. W., 481 Coon, M. J., 411 Cooney, R. V., 236-237, 238(17), 239(8), 241(9), 243(9) Cooper, M. J., 284 Coquerelle, T., 100 Corsaro, C., 421 Corthout, J., 416 Cosgrove, T. P., 273 Costa, R., 487 Costa de Oliveira, R., 121 Cotelle, N., 421,423(13), 425(13)
AUTHOR INDEX Cotgreave, I. A., 477-478,478(13), 479,482, 484, 485(5), 486, 486(5), 487, 487(7), 488-489, 489(8) Cotter, R. J., 57 Cotton, F. A., 69 Courison, C., 591 Coussio, J. F., 437 Coutant, J. E., 512 Cox, A. C., 327 Cox, D., 446 Craft, D. A., 615 Craft, N. E., 278, 397 Crain, P. F., 6-7, 7(30), 24 Cramp, W. A., 66, 78(4, 5), 79(4-6) Crane, F. L., 375 Crapo, J. D., 584 Crastes de Paulet, A., 423 Crawford, D., 191, 192(22) Crawford, D. R., 175, 181, 187, 191(19), 203(19) Crawford, J. M., 494 Crepaldi, G., 511 Crockett, R., 573,574(5), 575(5), 576(5) Cross, C. E., 252-254, 254(11), 255(11, 26), 256, 475, 619 Crystal, R. G., 256 Csallany, A. S., 439 Culcasi, M., 475 Cundy, K. C., 67, 117 Cunningham, R. P., 33, 37, 38(22), 40, 40(5), 41(28), 42(28), 102, 104(1), 105(1), 110(1), 123, 125 Curran, T., 163-164, 165(1,2, 8, 9), 166(2, 7, 9), 167, 167(2, 8), 168(2, 16), 170, 174(7, 9), 224 Cursted, T., 603 Cutler, R. G., 24, 31(27) Czapski, G., 77
D Dagan, A., 606 D'Agostini, F., 489, 490(27) Daikh, Y., 181 Dallner, G., 344 Dammann, A. L., 615 Damon, M., 423 D'Andrea, A. D., 34
649
Dani, C., 231 Daniel, V., 135-136, 139(6), 140(11), 492 Daniels, D. L., 176, 216(8) Danks, D. M., 542, 547(7) d'Arbigny, P., 475 Darley-Usmar, V. M., 423 Darnell, J. E., Sr., 199 Das, A. K., 604 Das, B., 252 Das, D. K., 410 Das, M., 411 Das, N. P., 401,421,437 Davies, D. S., 491 Davies, J.M.S., 175, 175(7), 176 Davies, K.J.A., 175, 175(4-7), 176, 254 Davies, M. J., 284, 423 Davis, K. A., 270 Davis, L. G., 226 Davis, P. A., 254, 255(26), 612, 614(46) Davis, W. B., 256 Davison, P. F., 98 Dawson, J., 488 Day, B. W., 631 Day, J. S., 553 DeB Butter, J., 337 De Bernardi di Valserra, M., 487 Debey, H. J., 145, 147(26) DeBoer, C. J., 147 Deby, C., 462 De Carro, L., 487 Decarroz, C., 55, 82, 83(29) Decuyper-Debergh, D., 121 Deeble, D. J., 3 Deelstra, H., 299 De Felippis, M. R., 573 DeFeudis, F. V., 462, 463(1) De Flora, S., 489-490, 490(27) Degan, P., 17, 24, 24(9), 26, 27(9), 32(9), 80 DeGraff, W., 588 DeGraff, W. G., 582, 587(17), 589(17) de la Harpe, J., 147 DeLange, R. J., 282 De Leenheer, A. P., 393 Della Loggia, A., 446 del Negro, P., 446 Delorenzo, O., 416 Demediuk, P., 619 Demet, D. L., 590 Demets, D. L., 590 DeMore, W. B., 262
650
AUTHOR INDEX
Demple, B., 123, 151, 175, 181,214(2), 217, 224 Denda, A., 24 Denny, R. W., 384 Dereu, N., 477, 480-481 Derguini, F., 394 Derian, C. K., 514 de Rooij, B. M., 478 DeRubertis, F. R., 557 Desa, F. M., 590 Desai, I. D., 295 Deschuytere, A., 299 Desvergne, B., 438 Dethmers, J. K., 493 Deugnier, Y., 438 Devary, Y., 224 Devasagayam, T.P.A., 45, 59, 80, 385, 386(18), 387(18, 24), 388,456, 575 de Vries, H.M.J., 631 Dewar, J. A., 590 Dhariwal, K. R., 270, 337 Dianzani, M. U., 616 Diatchkovskaya, R. F., 582 di Bilio, A. J., 421 Dibner, M. D., 226 Dick, R. D., 545-546, 547(18) Dickens, B. F., 620, 621(10), 626(10) DiCorleto, P. E., 514 Dietz, J. M., 397 Dignam, D., 159 Di Iorio, E. E., 469 Dikomey, E., 89, 93, 100, 101(5, 16, 28) Di Mascio, P., 45, 79-80, 116, 118(2, 3), 119, 120(3), 121, 129, 384-385, 385(12), 386(18), 387(12, 18), 575 Di Monte, D., 628 Dinarello, C. A., 141 Diplock, A. T., 295 Dirksen, M.-L., 6, 9(32) Distel, R. J., 170 Dixon, A. K., 547 Dizdaroglu, M., 3-4, 4(8), 5-6, 6(8, 21, 25), 7, 7(24), 8(24, 37), 9, 9(6, 26, 27, 31, 32), 10, 11(6, 25, 28), 13(21, 24), 14(21), 15(6, 21, 45), 16, 16(14), 17, 67, 70(15), 73(15), 75(15), 76(15), 82, 86, 117, 131 Djuric, Z., 5, 13(22) Djursater, R., 557 Doba, T., 322, 356 Dobberstein, B., 166, 167(14)
Dobie, J., 590 Doebeli, H., 167 Doelman, C.J.A., 489 Doetsch, P. W., 33-35, 37, 37(14), 38, 38(22), 40, 40(4, 5), 41(28), 42(8, 28), 102-104, 104(1, 3), 105(1), 110, ll0(l, 3), 112(15), ll3 Doghieri, P., 416 Doleiden, F. H., 384 Doly, M., 475 Donner, L. W., 337 Dorc, J. C., 591 Dorian, R. L., 497 Dormandy, T. L., 280(3), 281 Doroshow, J. H., 3 Dorozhko, A. I., 421 Dostalova, L., 296 Dougherty, M. H., 416 Douglas, K. T., 67, 73(14), 75(14) Doura, F., 4 Douste-Blazy, L., 423,606 Downs, D., 515 Dowsett, M., 591 Doyle, M. P., 469 Draper, H. H., 316 Dratz, E. A., 569 Dreano, Y., 592 Dresel, H. A., 601 Dri, P., 446 Drieu, K., 462-463,476 Drings, P., 135, 139(6) Driscoll, J. S., 638 Driskel, W. J., 397 Dr6ge, W., 135-136, 136(9), 137, 139, 139(6, 9, 16, 17), 140, 140(5, 11), 141(17), 142, 142(5, 8, 17), 143(17, 22), 144, 144(4, 5, 8, 16, 23), 145(7,24), 146(7, 8), 148, 148(9), 149, 149(7, 24), 152, 156(5), 157(5), 158(5), 492 Droy-Lefaix, M. T., 462, 474 Dubbelman, T.M.A.R., 69 Dubose, C. M., 523 Duchon, A., 524 Duck, M. V., 474 Duda, C. T., 328 Duhault, J., 423 Duigou-Osterndorf, R., 592 Duker, N. J., 131 Dulin, D., 311,313(7) Dull, B. J., 557
AUTHOR INDEX Duncan, L. A., 550 Dungworth, D. L., 257, 258(3) Dupuis, M., 462 Durand, R. E., 100 Durckheimer, W., 314 Duroux, E., 421 Dusting, G. J., 491 Dutton, H. J., 415 Duvall, T. R., 254, 257,475 Dyer, W. J., 609 Dypbukt, J. M., 76 Dzido, T. Z., 415,417(24) l)urba, A., 578, 579(31)
E Eakins, N. N., 437 Eaton, J. W., 568, 569(45) Ebashi, I., 341 Ebert, B., 422, 423(34), 425(34) Ebert, M. H., 108 Eck, H.-P., 135-136, 136(9), 137, 139(6, 9, 16, 17), 140, 140(5, 11), 141(17), 142, 142(5, 8, 17), 143(17), 144(5, 8, 16, 23), 146(8), 148(9), 492 Eckfeldt, J. H., 327 Edbauer-Nechamen, C. A., 175 Eddy, L. J., 578 Edelstein, S., 400 Edlund, P. O., 275 Edvardsson, K. A., 89, 94(9), 96(9), 100(9), 102 Edwards, J. C., 66, 79(6) Edwards, K. J., 591 Egan, D., 446 Eichhom, G. L., 68 Eisenberg, W. C., 257, 258(4) Eklow, L., 628 Eklund, A., 487, 488(12) Ela, S. W., 492 Elinder, L. S., 512 Elkins, D., 538 Ellis, I. O., 590 Ellison, E. G., 300 Ellman, A. L., 495 Ellman, G. L., 273,458, 502 EI-Saadani, M., 513 Elstner, E. F., 420 Elwell, J. H., 244
651
Emery, S., 601 Emslie, E. A., 200, 201(33) Endo, H., 282 Eneff, K. L., 3, 21, 59, 80 Engel, J., 250, 251(7) Engel, R. R., 284 Engemann, R., 462 Engers, H. D., 135 Englert, G., 390, 393, 394(14) Engman, L., 477-478, 478(13) Engmann, L., 477 Ennen, J., 137, 492 Epe, B., 79, 122-123, 125-126, 129-130 Epp, O., 478 Eppenberger, U., 592 Epps, D. E., 283,345,551,554 Epstein, S., 590 Erben-Russ, M., 422, 425(25), 426(25), 427(25), 428, 428(25) Erdman, J. W., 389, 397 Erickson, L. C., 128, 129(7) Erixon, K., 67, 78(18), 89, 94(14), 96(14), 100(14) Erjefalt, I., 253,256(15) Ernster, L., 344 Eskins, K., 415 Esko, J. D., 609 Espenson, J. H., 426 Essigmann, J. M., 17 Estabrook, R. W., 577 Esterbauer, H., 513,616 Estrela, J. M., 367 Etienne, A., 462 Evans, P. J., 411,424, 446, 453(11, 12), 543, 544(13) Evces, S., 615 Evers, M., 480 Ewertz, M., 24, 31(28) Ewig, R.A.G., 128, 129(7)
F Fabbro, D., 591 Facino, R. M., 415, 416(22) Fahey, R. B., 497 Fahey, R. C., 367 Fahrenholtz, S. R., 384 Fairweather, D. S., 82 Falck, J. R., 514, 606
652
AUTHOR INDEX
Fanger, M. W., 135 Fanidi, A., 591 Faraggi, M., 573 Fargnoli, J., 200, 204,205(34, 38), 208(38) Fariss, M. W., 367, 368(8), 369(8), 371(8) Farjanel, J., 438 Farmilo, A., 385 Farr, A. L., 516, 594, 598(50) Farr, S. B., 151, 156(1), 181, 217, 217(12), 224 Farrell, P. M., 514 F~itkenheuer, G., 135, 139(6) Faure, M., 423 Fausel, M., 476, 479(2), 482(2) Favier, A., 17, 80 Fayard, J. M., 591 Fedelo~ov,q, V., 574 Feelisch, M., 472 Feldman, R. S., 601 Feidstein, M., 262 Feletti, F., 487 Fellin, R., 511 Fells, G. A., 256 Fentiman, I. S., 590 Fernandes, A. C., 552 Ferrand, B., 438 Ferrfindiz, M. L., 446, 453,453(14), 454(14, 21) Ferrari, M. B., 601 Ferrari, R., 373 Ferraro, G. E., 437 Ferraro, P., 100 Ferrero, J. A., 371 Feskens, E.J.M., 454 Fewtrell, C.M.S., 411 Fey, W., 590 Feyzi, J., 590 Fibach, E., 606 Fiers, W., 244 Filipe, P. M., 552 Filipski, R. F., 327 Finkelstein, E., 582 Fischer, C., 557 Fischer, H., 477,481,482(33, 34) Fischer, V., 638 Fitchett, M., 73 Flamm, D. L., 264 Flanders, K. C., 591 Fleming, W. E., 553 Flemstrom, G., 527
Flenley, D. C., 488 Floyd, R. A., 3, 17, 21, 59-60, 62(6), 79-80, 87, 515, 523-524, 524(1), 525(1), 526(1) Fogelman, I., 590 Foglia, T. A., 604 Folch, J., 599 F61des-Papp, Z., 422 F61diak, G., 422, 425(26) Fong, K. L., 316 Fong, L. G., 513 Foote, C. S., 79, 314, 384, 605, 606(29), 607(29) Ford, D. A., 620 Ford, G. D., 458 Ford, P. C., 331 Forder, R. A., 446, 449(15), 450(15), 453, 453(15) Fornace, A. J., Jr., 128, 200, 204, 205(34, 38), 208(38) Forsmark, P., 344 Fort, P., 231 Forte, T., 254, 317, 363,373,467 Foster, D. O., 303, 310(2) Fraga, C. G., 24, 437 Franceschini, G., 515 Francis, F. J., 415 Franklin, R. M., 125 Franza, B. R., Jr., 164 Franzke, J., 93, 100, 101(16, 28) Frayer, W., 492(5), 493,494(5) Freedman, A. M., 620 Freeman, B. A., 256, 273,475 Frei, B., 253-254, 269-270, 273(10), 344, 348(7), 372 Frei, J. V., 547 Freisleben, H.-J., 371-372, 381(6), 401, 409(11), 460, 467, 533,534(23) Freistleben, H.-J., 358 Frelinger, J. A., 194 Frenkel, K., 3, 58 Fridovich, I., 3, 338, 357, 469, 536, 558, 566(25), 584, 621,625,625(14) Friedberg, E. C., 34, 37, 123 Friedman, C. A., 128, 129(7) Friedman, L. R., 78 Friedman, P. A., 411 Friedman, T., 40, 44(27) Filmer, A. A., 605, 606(30), 607(30) Fritsch, E. F., 48, 50(4), 117, 121(8), 170, 171(21), 193, 197(25), 202(25), 203(25),
AUTHOR INDEX 207(25), 213(25), 214(25), 226, 227(15), 228(15), 229(15), 230(15),232(15) Froncisz, W., 580, 588 Fuchs, J., 310, 317, 318(6), 580 Fuciarelli, A. F., 5, 9(26), 11(28) Fujiki, Y., 603-604 Fujimori, E., 384 Fujimoto, M., 52 Fujioka, S., 446 Fujisawa, H., 425 Fujiwara, Y., 601 Fukunaga, Y., 275 Fukuzawa, K., 322, 366 Funahashi, T., 514 Fung, H. L., 491 Furr, B.J.A., 600 Furukawa, M., 411,453 Furukawa, T., 371 Furuya, Y., 601
G Gabard, B., 484 Gabauer, I., 578 Gabe, E. J., 356 Gabor, M., 411 Gabriel, H., 479, 480(22), 481,482(33) Gabrielsson, J., 487 Gagne, E., 167, 168(16) Gaitonde, M. K., 148 Gajewski, E., 3, 5, 9(26, 27), 11(28), 16(14), 67, 70(15), 73(15), 75(15), 76(15), 82,117 Galeotti, T., 601 Gallas, H., 142, 144(23) GaUatin, W. M., 206 Galmozzi, M. R., 487 Galos, R., 189 Ganesan, A. K., 34, 127 Ganguly, T., 131 Gantchev, T., 638 Gantchev, T. G., 631,635 Gardana, C., 415,416(21, 22) Gard~s-Albert, M., 462 Garen, A., 52 Garland, D., 254 Garland, L. G., 423 Garner, M. M., 165, 171(10) Garry, P. J., 331 Gasc6, E., 371
653
Gasc6, M. A., 446 Gasparotto, A., 511 Gates, F. T., 123 Gatt, S., 604, 606, 619(23, 24) Gavarron, F. F., 616 Gawthorne, J. M., 543 Gaydou, E., 421,423(13), 425(13) Gedik, C. M., 77, 78(52) Geher, K., 516, 523(31) Geierstanger, B. H., 68, 70(23) Gentz, R., 164, 165(2), 166, 166(2), 167(2, 13, 14), 168(2) George, A. M., 66, 78(4, 5), 79(4-6) Gerber, G., 422 Gerber, L. E., 388 Gerdin, B., 557 Gergel, D., 574-575, 577(18) Gerlt, J. A., 113 Gerrard, J. M., 327 Gerrity, R. G., 513 Gesteland, R. F., 218 Getoff, N., 3 Ghizzi, A., 487 Giacomoni, P., 17, 80 Giannella, E., 424 Giannoni, P., 490 Gibson, D. D., 316 Gilbert, B. C., 67, 73 Gilbert, H. F., 367 Gilbert, H. S., 328 Gilbert, J., 591 Gilbert, W., 35, 38(18), 103 Gimeno, C. J., 24, 25(25), 26(25), 67, 117 Giovanniello, T. J., 53 Giovannucci, E., 269 Giulivi, C., 254 Glajch, J. L., 417 Glaze, W. H., 256 Glazer, A. N., 273,282,406, 460, 531 Glende, E. A., 404 Glinz, E., 302-303,307(3), 390, 394(14) Glockner, J. F., 580 Glogowski, J., 472 Gluzman, Y., 116 Gmiinder, H., 135-137, 139(16, 17), 140(5, 11), 141(17), 142, 142(5, 17), 143(17, 22), 144(5, 16, 23), 149, 492 Gober, K. H., 416, 417(40) Goddard, P. M., 591 Godfrey, L., 592
654
AUTHOR INDEX
Godinger, D., 586, 589(27) Goeddel, D. V., 244, 250, 251(8, 9) Gogolewski, M., 320 Gohil, K., 274, 368, 371(13) Goiffon, J. P., 416 Gold, E., 590 Goldman, D., 108 Goldman, P., 557 Goldman, R., 318 Goldstein, J. L., 606 Goldstein, M. S., 58 Goldstein, S., 77 Goli, M. B., 398 Gomperts, B. D., 411 Gonzalez, F. J., 588 Goodenough, D. A., 236 Goodwin, T. W., 388 Gooneratne, S. R., 543 Gopinathan, V., 284 Gordon, L. K., 42 Gorecki, M., 250, 251(6) Gorman, A. A., 384 Goscin, S. A., 621,625(14) Gossett, J., 37, 38(22), 102, 104(1), 105(1),
110(1) Gotchev, G. P., 635 Goth, S., 357, 358(22), 359(22) Goto, A., 425 Gotoh, N., 402 Gottlieb, R. A., 224 Goulding, A., 590 Goutier, R., 462 Gouyette, A., 123 Govil, G., 528 Gown, A. M., 513 Graca, M., 591 Graf, E., 477, 481,568 Graf, P., 476, 479(1), 480(1), 482 Grafstr6m, R., 487 Graham, D. A., 581,583(12) Grajewski, E., 17 Gramatica, P., 384 Graminski, G. F., 505 Granger, D. N., 256, 555, 557, 568, 569(44), 570(49), 571 GranstrOm, E., 449 Grant, R. D., 422 Gray, G. M., 614, 615(47) Grayer, R. J., 429 Graziani, Y., 411
Green, J., 355 Green, L. C., 472 Greenberg, J., 181 Greenberger, J. S., 181 Greenley, T. L., 423 Greenstein, J. P., 496 Gregolin, C., 317 Greiff, L., 253,256(15) Greim, H., 546 Griesenbach, U., 89, 90(12), 91(12), 94(12), 96(12), 100(12) Griffith, O. W., 492(8), 493,494(8), 496 Grill, H. J., 592 Griller, D., 422 Grimek-Ewig, R. A., 89, 90(8), 94(8) Grimm, S., 163 Grisham, M. B., 256, 555-558, 560, 561(29), 562(29), 563(29), 564(29), 565, 566(24, 29, 36), 567(36), 568, 568(36), 569(41, 44), 570(41, 49), 571 Groilman, A. P., 80 Grosch, W., 424 Gross, H. J., 185 Gross, J., 388,389(2), 390(2) Gross, R. W., 603, 620 Grossman, S., 424 Grunicke, H., 602 Gryglewski, R. J., 411,421,422(16), 42306) Guan, D. M., 439 Guardiola, B., 423 Guarnieri, C., 373 Gudej, J., 415, 417(24) Guguen, C., 438 Guguen-Guillouzo, C., 438 Guillaume, M., 477 Guillon, J. M., 462 Guillot, F. L., 553 Guillouzo, A., 438 Gulz, P.-G., 415, 416(23) Gunning, P., 250, 251(7) Gunter, E. W., 397 Guo, H., 236, 241(14) Gupta, R. C., 18 Gupta, S. P., 9 Gustafsson, B., 253, 256(15) Gutteridge, J.M.C., 3, 16, 66, 67(1), 80, 253, 256, 256(13), 269, 280(3), 281,372, 506, 528, 542-543, 544(12), 547(12), 561, 561(33), 563(33), 566(32, 33), 594,596, 601(52)
AUTHOR INDEX Gy6rgy, I., 422, 425(26) Gyorgy, P., 327 Gysel, D., 296
H Haagsman, H. P., 252 Haas, S. M., 536 Haenen, G.R.M.M., 362,455,478, 489 Haeuptle, M. T., 166, 167(14) Haga, S., 533 Hagaman, K. A., 505 Hagan, M. P., 3, 4(8), 6(8) Hagar, W. L., 257 Hagerman, A. E., 429, 431-432, 435, 435(12), 436, 436(20) Hagmaier, V., 337 Hahn, S. M., 582, 589(22) Haidle, C. W., 127 Haim, N., 631,638(6) Hajra, A. K., 603-604, 612, 614(46) Hakusui, H., 482 Halbrook, J. H., 175, 214(2) Halevy, O., 401 Hall, E. D., 548, 551-553 Hall, K., 591 Hallaway, P. E., 568, 569(45) Hallberg, A., 477, 479, 488 Halliwell, B., 3, 16-17, 66-67, 67(1), 70(15), 73(15), 75(15), 76, 76(15), 252-254, 254(11), 255(11, 26), 256, 256(13), 269, 273, 372, 411, 424, 446, 453(11-13), 459, 475, 489, 506, 528, 543, 544(13), 557, 561, 566(32), 571(18), 591-592, 592(32, 33), 594, 594(33, 37, 39), 595(32, 37, 39), 596, 596(33, 37, 39), 597(33, 39), 598(48), 599(48), 600(51), 601(33, 35, 37, 39, 40, 48), 619, 621 Hamada, A., 582 Hamano, M., 421,423(22) Hamberg, S., 141 Hamernyik, P., 335, 337(7) Hamilos, D. L., 135 Hamilton, C., 590 Hamilton, C. E., 403 Hamilton, C. S., 312 Hamilton, K. K., 33, 38, 40(4), 102-104, 104(3), ll0(3), 112(15), 113 Haming, W. J., 262
655
Hammes, G. G., 410 Hammond, G. S., 312, 403 Han, D., 321,361,363(26, 34), 366, 366(26), 383 t Han, F., 554 Han, J., 492(8), 493,494(8) Han, L.-P.B., 504 Hanahan, D., 120 Handelman, G. J., 381,398,569 Handleman, G. E., 524 Hankovszky, O. H., 580 Hanna, P. M., 69-70, 70(30) Hans, P., 462 Hansch, C., 538 Hansen, H.-J., 384 Hanusch, M., 394, 398(27) Hara, K., 491 Hara, S., 335, 416 Harada, T., 256 Haraikawa, K., 422, 425(40) Harborne, J. B., 420, 446 Harel, S., 568 H~iring, M., 130 Hariton, C., 429 Harm, W., 35, 36(15), 104 Harman, D., 524, 619 Harper, D. A., 5, 13(22) Harris, G., 66, 78(4), 79(4, 6) Harris, J., 17 Harris, L., 620, 621(9) Harrison, A. G., 57 Hart, B. A., 423 Hart, D. A., 135 Hart, L. E., 66, 78(4), 79(4, 6) Hartley, D. M., 552 Hartman, J. R., 250, 251(6) Hartmann, M., 136, 140(11), 492 Hartzell, W. O., 270 Haseltine, W. A., 34-35, 37(14), 42, 42(8), 103 Hashimoto, H., 393 Haslam, E., 431 Hatayana, K., 411 Hatch, G. E., 256 Hatefi, Y., 275 Haugan, J. A., 390 Haupt, R., 256 Havsteen, B., 446 Hawkes, W. C., 270, 271(13) Hawley, C. J., 557
656
AUTHOR INDEX
Hay, R. T., 153 Hazen, S. L., 620 Hearse, D. J., 523 Heath, R. L., 560,566(27) Heeg, J. F., 511 Hegler, J., 122, 125-126, 130 Heidelberger, C., 237 Heidemann, H. T., 462 Heijman, M.G.J., 426 Heilmaier, H. E., 546 Heimler, D., 415 Heinecke, J. W., 506, 514 Heinkel, K., 557 Heinola, K., 527 Heintz, N. H., 191 Heinz-Erian, P., 296 Heitzman, A.J.P., 426 Helbock, H. J., 24 H616ne, C., 79 Helland, D. E., 34, 40, 41(28), 42(8, 28), 79 Heller, W., 422, 424(24), 425, 425(24), 426(65), 427(65), 428, 428(24), 429(65), 437 Hemingway, R. W., 430, 433(3) Hemler, M. E., 481 Hems, R., 367, 370(7) Henderson, B. E., 590 Hengartner, U., 390, 394(14) Henglein, A., 426, 427(70) H6nichart, J. P., 421,423(13), 425(13) Henle, E. S., 51 Henner, W. D., 40, 41(28), 42(28) Hennig, B., 621 Hennighausen, L., 172 Henriksen, T., 514 Henry, P. D., 630 Hepp, M., 463 Hernanz, A., 331,335 Herrlich, P., 170 Herscovitz, H., 604 Hersey, A., 423 Hertog, M.G.L., 454 Herzenberg, L. A., 146, 492 Hess, D., 296 Heuvel, H.V.D., 416 Hewitt, S., 569 Heymans, H.S.A., 603 Hicks, K. B., 337 Hideg, K., 580 Hiermann, A., 415
Higashida, M., 410 Higgs, E. A., 475 Hildebrand-Zanki, S. U., 638 Hillis, W. E., 432 Hilton, J., 100 Hinzmann, J. S., 554 Hiraishi, H., 256 Hiramatsu, M., 401,409(9) Hiramitsu, T., 404 Hirono, I., 422,425(40) Hiser, M. F., 511 Hixson, C. S., 460 Ho, W.-T., 385 Hochstein, P., 331,578 Hochuli, E., 167 Hodnick, W. F., 424 Hoefler, G., 604 Hoefler, S., 604 Hoekstra, J. W., 469 Hoey, B. M., 489, 557, 571(18) Hofer, E., 199 Hoffman, M. Z., 284, 288(15) Hogeboom, G. H., 535 Hogsett, W. E., 21 Holbrook, N. J., 200, 204, 205(34, 38), 208(38) Holden, S. A., 494 Holdiness, M. R., 487 Holec, V., 578 Hollander, M. C., 200, 205(34) Hollman, P.C.H., 454 Hollunger, B. G., 344 Holm, B. A., 256 Holwitt, E., 3, 4(8), 6, 6(8), 9(32) Honkanen, E., 527 Hoogland, H., 619 Hope-Ross, P., 619 Hopwood, L. E., 580, 588 Hor¢tkov~i, L., 572, 574, 576, 578, 578(13), 579, 579(20) Horie, T., 453 Horiguchi, K., 24 Hornbrook, K. R., 316 Hornig, O., 337 Hornsey, S., 66, 78(4), 79(4) Horobin, J. M., 590 Horowitz, J. D., 491 Horrocks, L. A., 603,619 HorswiU, E. C., 312 Hosomi, M., 393
AUTHOR INDEX Hossain, M. Z., 236, 241(14) Hough, K., 491 Houghton, J., 590 Hoult, J.R.S., 411, 424, 443, 446, 449(15), 450(15), 452-453, 453(11-13), 557 Howeland, S. K., 328 Howell, J. McC., 543 Howes, T. W., 397 Hrstich, L. N., 433 Hsie, A. W., 582 Hu, M. L., 254, 475 Hua, X., 415,416(26) Huang, M. T., 411 Huang, R.-R.C., 477 Hubbard, R. C., 256 Huber, C. O., 581 Huber, L. A., 601 Huber, R., 478 Hubert, N., 443 Hudson, B. S., 375,460 Huff, D. L., 397 Hughes, L., 303,310(2), 533 Hughes, R. E., 332, 334(5) Huguet, A. I., 421 Hui, B., 581 Hull, W. E., 148 Humplov~, J., 576 Hunziker, F., 296 Huque, T., 616 Husain, S. R., 421 Hutchinson, F., 33, 35(2) Hutchison, M., 335, 337(7) Hutin, P., 602 Huynh, H., 591 Hyde, J. S., 580 Hyland, K., 530 Hyslop, P. A., 77 Hyuga, T., 547
I Ibrahim, C., 555 Ibrahimi, I., 166, 167(14) Ida, Y., 275 Ide, H., 34 Igarashi, O., 304 lijima, F., 630 Iizuka, K., 321, 356 Ikeno, H., 322
657
Ikenoya, S., 274 Ilan, Y., 426, 427(70) Imagawa, M., 170 Imai, K., 484 Imbra, R. J., 170 Imlay, J. A., 53 Inada, T., 321 Ingold, K. U., 269, 280, 286(1), 303-304, 310, 310(2), 312,316, 322,354-355,362, 362(8), 372, 374, 381(3), 403, 406, 422, 506, 533, 543-544, 581,601,633 Inoue, M., 338, 341-343 Inselmann, G., 462 Ionnone, A., 580 Ippendorf, H., 387(24), 388 Irvin, B., 269, 304 Ishi, K., 505,506(6) Ishibashi, E., 415,416(31) Ishii, T., 137, 139(15), 148 Ishikawa, H., 338 Ishizu, K., 366 Isler, O., 295, 388, 390(1), 392(1) Israel, A., 161 Itaka, Y., 312 Ito, E., 506, 565 Ito, Y., 415,416(26) Itoh, S., 351 Ivanov, S. A., 360 Ivanov, St. A., 320 Ivanov, V. I., 69 Ivey, K. J., 256 Iwata, T., 335 Iwatsuli, K., 630 Iyengar, V., 546 Izzotti, A., 489, 490(27)
J Jackson, I., 590 Jackson, J. H., 77 Jackson, R. L., 505, 505(7, 8), 506, 506(4), 507, 509(8), 512(7, 8) Jacob, R. A., 24 Jacobs, A., 438 Jacobs, L. W., 491 Jacobsen, E. J., 550, 552 Jacobson, F. S., 175(3), 176, 181(3) Jacolot, F., 592 Jaffe, I. J., 542
658
AUTHOR INDEX
Jaggy, H., 463 Jain, A., 492(5-7, 9), 493,494(5-7, 9) Janero, D. R., 626 Janeway, C. A., Jr., 141 Janssen, Y.M.W., 191 Janzen, E. G., 525 Jarman, M., 591 Jayaraman, J., 372 Jeanteur, P., 231 Jedstedt, G., 527 Jefferson, M. M., 571 Jego, P., 443 Jenkinson, S. G., 551 Jensen, C. D., 397 Jensen, G. L., 492-493,493(3) Jensen, N. H., 390 Jergelovfi, M., 578 Jeroudi, M. O., 515, 523, 523(23) Jevcak, J. J., 89, 90(10), 91(10), 94(10), 96(10) Jewet, S. L., 578 Jiang, J. L., 546 Jin, H. L., 394 Jin, R., 51 Jocelyn, P. C., 147 Johanson, K. J., 100 Johansson, M., 487 Johnson, A., 181 Johnson, A. W., 123 Johnson, D., 126 Johnson, F. F., 411 Johnson, G., 475 Johnson, M. C., 322 Johnson, P. F., 164 Johnsrud, L., 34 Jolly, R. A., 551 Jonat, C., 170 Jones, A., 478 Jones, A. L., 590 Jones, C. L., 603 Jones, D. A., 253,254(11), 255(11) Jones, T. J., 488 Jongkind, J. F., 627 Jordan, V. C., 590, 600 Joseph, J., 522 Joshi, V. C., 372 Jovanovi~, S. V., 573,574(5), 575(5), 576(5) Joyce, A., 316, 355 Julius, M. H., 146 Jurfinek, I., 576, 578-579, 579(20) Jfirgens, G., 513
K Kaakkola, S., 527 Kadiiska, M. B., 70 Kadle, R., 236, 241(14) Kadomatsu, K., 236 Kadonaga, J. T., 165, 172 Kagan, V., 321-322, 357-358, 358(22), 359(22), 363, 366(33), 382(25), 383,638 Kagan, V. E., 311, 316-318, 343-345, 346(5), 351, 353(10), 354, 361, 363, 363(26), 366(26), 371-373, 373(9), 381, 381(6), 382(9), 401,404,456, 460,467, 527-528, 533,534(23), 537(4), 577, 631, 638 Kagawa, T. F., 68, 70(23) K~gedahl, B., 484, 487 Kageyama, H., 181, 191(16) KAhlberg, M., 484 Kahmann, R., 220 Kahn, A. U., 619 Kaiser, S., 384, 385(12), 387(12), 575 Kakata, A., 220 Kakemi, M., 491 Kfillay, Z., 574 Kaitchev, M. G., 635 Kalyanaraman, B., 522, 632, 633(10), 638(10) Kamb, A., 244 Kamiya, Y., 317,356, 372,374, 382(10), 402, 404(12), 506, 533 Kan, M., 280(4), 281 Kaneda, H., 402, 480, 481(29) Kanematsu, S., 459 Kanner, J., 568 Kanofsky, J., 555 Kanofsky, J. R., 256, 619 Kanter, P. M., 89, 90(11), 91(11), 94(11), 96(11), 100(11) Kaplan, L. A., 389 Kappock, T.J.I., 237 Karimi-Booshehri, F., 87 Karin, M., 170, 224 Karlstr6m, O., 95 Kartha, V.N.R., 401 Kasai, H., 17, 24-25, 59, 61(1), 63(1), 64(1), 65(1), 80 Kasai, K., 80 Kasai, N., 547 Kashiwazaki, S., 282 Kass, G.E.N., 76
AUTHOR INDEX Kasurinen, J., 606 Katan, M. B., 454 Katayama, K., 274 Katcher, H. L., 40 Kato, A., 320, 321(4), 356 Kato, K., 311,313,313(6), 314, 314(6), 406 Kato-Jippo, T., 393 Kauf, H., 253,254(11), 255(11) Kaur, H., 254 Kawai, C., 505,506(6) Kawakami, A., 356, 372, 382(10), 402, 404(12) Kawamura, I., 592 Kawanishi, S., 70 Kaye, J., 141 Kaysen, K. L., 311-312, 312(8), 313(8, 14), 314(8), 315(11) Keana, J. F., 580 Kedes, L., 250, 251(7) Keenan, B. C., 82 Kehrer, J. P., 320(15), 328 Keilin, D., 469 Keller, H. E., 296 Kelley, D. G., 426 Kellogg, E. W., 625 Kennedy, T. A., 312, 315(11) Kennett, F. F., 620, 621(9) Keown, W. A., 162 Kern, D. H., 638 Kerppola, T. K., 164 Keshavarzian, A., 555 Keskinova, E., 3 Kessner, A., 620, 621(9) Keyse, S. M., 200, 201(33), 224-225 Kezdy, F., 554 Kezdy, F. J., 283, 345 Khachik, F., 396, 398 Khan, H., 66, 78(5), 79(5) Khan, S., 318,456 Khoo, J. C., 513 Khorana, H. G., 52 Khrapova, N. G., 345, 351(12), 356 Khwaja, S., 343,461 Kiesow, L. A., 53 Kikuchi, G., 225 Kikuchi, K., 181, 191(16) Kikuchi, S., 315,351 Kim, E. H., 175(4), 176 Kim, E. K., 422 Kim, M., 344, 348(7) Kim, P.M.H., 38, 112(15), 113
659
Kim, Y. K., 175 Kimmel, A. R., 197 Kimura, M., 422 Kindahl, H., 449 King, M. M., 294 Kingsbury, R., 219 Kingston, R. E., 221 Kinscherf, R., 139 Kirkkola, A.-L., 281,282(5) Kirkland, J. J., 417 Kishi, T., 275 Kishino, B., 514 Kissinger, C. M., 202 Kita, T., 505, 506(6) Kitagawa, S., 425 Kitanova, S. A., 344, 346(5), 528 Kittler, L., 79 Kitzler, J. W., 17, 24(9), 27(9), 32(9), 80 Kiyosawa, H., 24 Klapper, M. H., 573 Klebanoff, S. J., 556 Klein, P. D., 58 Kling, O. R., 516 Klotz, U., 557 Kluin, K. J., 545,547(18) Kogoma, T., 181,217, 217(12) Kohn, K. W., 89, 90(8), 94(8), 128, 129(7) Kohno, M., 422, 425(40) Kohno, N., 601 Kohsaka, M., 592 Kolachana, P., 17, 24(9), 27(9), 32(9), 80 Kolonel, L. N., 235 Komarov, P. G., 582 Komiya, T., 446 Komiyama, K., 320-321,322(12), 356, 360 Komura, S., 591 Komuro, E., 322, 348, 356, 383, 402, 506, 565 Kondo, N., 592 Konishi, Y., 24 Kono, Y., 378 Konovalova, N. P., 582 Koop, D. R., 411 Koppenol, W. H., 51,422 Korbut, R., 411 Korkolainen, T., 527 Kormann, A. W., 302-303,306, 310(4) Korn, T. S., 17, 24(9), 27(9), 32(9), 80 Kornitzer, D., 219 Korstanje, L. J., 553 Kosai, K., 4
660
AUTHOR INDEX
Kosower, E. M., 367, 497 Kosower, N. S., 367, 497 Koster, A. S., 367, 368(10) Koster, J. F., 561,566(30) Kostyuk, V. A., 421 Kourai, H., 453 Kourilsky, P., 161 Kow, Y. W., 34 Kowaluk, E., 491 Koyama, K., 338, 343 Koyama, Y., 390, 391(16), 392(16), 393, 393(16) Koynova, G. M., 344, 346(5), 528 Kozlov, Yu.P., 404 Kraemer, K. H., 35, 67, 73(13), 75(13), 121 Kramer, J. H., 620, 621(8, 10), 626(10), 629(7) Krasnovskii, A. A., Jr., 386, 387(22) Krebs, H. A., 367, 370(7) Krezowski, A. M., 327 Krinsky, N. I., 235, 372, 381,384, 393, 398 Krishna, M. C., 580-582, 583(12), 587(17), 588, 589(17) Krishnamurthy, S., 401 Kr/anov~i, L., 576 Krolikowska, M., 421,424(7) Kromhout, D., 454 Krstulovic, A. M., 397 Krueger, V.P.M., 8, 13(39) Ku, G., 507 Kubo, K., 34 Kucherlapati, R. S., 162 Kuchino, Y., 24 Kuchtina, Y. E., 356 Kuhn, D. C., 327 Ktihnau, J., 411,420 Kuhnlein, U., 53 Kuksis, A., 516, 523(31) Kulesz, M. M., 24 Kulesz-Martin, M. F., 36, 38(19) Kumar, N. B., 590 Kumazawa, T., 324, 326(24) Kume, N., 505, 506(6) Kung, W., 592 Kuninaka, A., 52 Kunitomo, R., 342 Kunitz, M., 52 Kunze, D., 256 Kuo, W. H., 515 Kurihara, K., 324, 326(22-24)
Kuroki, T., 181, 191, 191(16) Kurth, R., 137,492 Kurt-Jones, E. A., 141 Kusterer, K., 528 Kutnink, M. A., 270, 271(13) Kutty, R. K., 225 Kuypers, F. A., 345,373,382 Kvaltinov~l, Z., 579 Kyogoku, K., 411
L Labadie, R. P., 423 Labataille, P., 82 Labb6, R. F., 270, 335, 337(7) Labeque, R., 605 Lachman, L. B., 141 Ladenstein, R., 478 Laemmli, U. K., 66, 78(3), 108, 180, 242 Lafleur, M.V.M., 95 Lagarde, M., 591 Lahti, R. A., 552-553 Lai, C. S., 580, 588 Lai, E. K., 515, 523,523(23) Lakshman, M. R., 388 Lalonde, M., 303, 307(3) Lambert, C., 477 Lambert, H., 66, 79(6) Lamola, A. A., 384 Lamon, S., 511 LaMont, J. T., 256 Land, E. J., 68, 69(27) Landi, L., 344 LandoR, P. A., 257 Lands, W.E.M., 481 Landschulz, W. H., 164 Landum, R. W., 523,524(1), 525(1), 526(1) Lang, C. A., 368 Lang, J. K., 269, 274, 304 Langford, S. D., 253,256(9) Langholz, E., 557, 569(22), 570(22) Langlois, R., 81, 83(23), 635 Lanks, K., 147 Lanzer, M., 166, 167(14) Lardy, H., 126 Larson, R. A., 446 Larsson, K., 487,488(12) Larsson, S., 491 Lashley, D. W., 331
AUTHOR INDEX Laskawy, G., 424 Last, J. A., 254, 257-258, 264(12) Lau, J. M., 389 Lau, L. F., 224 Laughton, C. A., 591 Laughton, M. J., 411,424, 446, 453(11, 12), 591,592(32), 595(32) Laugier, C., 591 Laura, R., 411 Lauterberg, B., 488 Lautier, D., 225 Laval, J., 3, 10, 16, 16(14), 80, 82, 117, 123 Lawrence, R. A., 316, 551 Lazarow, P. B., 603-604 Leavitt, R., 252 Lebovitz, R. M., 159 Le Cam, A., 438 Lederer, F., 123 Le Doan, T., 79 Le Doucen, C., 423 Lee, C. R., 397 Lee, C. Y., 424 Lee, C.S.L., 601 Lee, D. M., 513,515 Lee, K., 33, 102-103, 104(1, 3), 105(1), llO(1, 3) Lee, M. S., 250,251(8) Lee, S.-H., 384 Lee, W., 172, 270, 335, 337(7) Lees, M., 599 Lefebvre, S. P., 591 Lefer, A. M., 475 Lehmann, J., 327 Leibowitz, M. E., 322 Lennard, M. S., 592, 594(39), 595(39), 596(39), 597(39), 601(39) Lenz, A. G., 254 Lenz, J., 33, 40(3), 104 Leo, A., 538 Leonard, W. J., 152 LeRosen, A. L., 390 Lescoat, G., 437-438, 441,443 Lesellier, E., 397 Lesser, R., 477,481(6) Levade, T., 606 Levanon, A., 250, 251(6) Leventhal, H., 590 Levin, J. D., 123,224 Levine, A., 189 Levine, M., 270, 337
661
Levine, R. L., 254 Levine, Y. K., 426 Levinthal, C., 52 Levy, E., 497, 498(29), 499(29) Levy, E. J., 492, 499-500, 501(1), 502, 502(1), 503, 503(1, 3), 504(3, 5) Lewis, C. D., 66, 78(3) Lewis, D. F., 514 Lewis, M., 331 Lewis, P. A., 484 Leyck, S., 481 Li, C., 258 Li, P., 415,416(26) Li, Y., 547 Liaaen-Jensen, S., 390 Liang, P., 207 Lichter, P., 241 Liebler, D. C., 311-312, 312(8), 313, 313(8, 14), 314(8), 315(11, 12) Liehn, H. D., 600 Lien, E. A., 591-592 Lightner, D. A., 273 Lim, B. P., 403 Lim, J.-S., 137, 139(17), 141(17), 142(17), 143(17) Limacher, P., 602 Lin, Y., 494 Lindahl, T., 34, 40, 95, 123 Lindan, C. P., 34 Lind6n, I.-B., 527 Lindhal, R., 615 Lindsay, D. A., 303,310(2) Linn, S., 51, 53, 123 Linseman, K. L., 551-552 Lion, Y., 462 Lippmann, M., 252 Lissi, E. A., 423 List, T., 555 Liu, F., 77 Liu, G. J., 515,524 Liu, G. T., 421 Liu, T. K., 590 Liuzzi, M., 54 Livrea, M. A., 401 Lloyd, R. S., 127 Lobos, E., 557 Locke, J. T., 533 Locke, S., 269, 280, 286(1) Locke, S. J., 269,423,506, 543 Lodola, E., 487
662
AUTHOR INDEX
Loeb, L. A., 17 Loewe, H., 422, 423(34), 425(34) Loewenstein, W. R., 236 Loft, S., 24, 31(28) L6hr, J. P., 481,482(34) Loman, H., 95 Lonchampt, M., 423 Longo, A., 487 Lonning, P. E., 591 Lopes, F., 591 Lopes, M.C.F., 591 L6pez-Torres, M., 332, 333(4), 334(4), 337(4) Lopresti, R. J., 309 LoSardo, J. E., 104 Loser, R., 600 Lou, Z. C., 416 Louie, S., 254, 475 Love, R. R., 590 Lowry, C. V., 175-176 Lowry, O. H., 516, 594, 598(50) Loyer, P., 438 Lualdi, P., 487 Lubon, H., 172 Luchette, C. A., 494 Luczynski, A., 320 Luethy, J. D., 200, 205(34) Luk, D., 164, 165(1, 2), 166(2), 167, 167(2), 168(2, 16) Lukovi~, L., 576, 579, 579(20) Lundberg, W. O., 617 Luo, Y., 51 Luongo, D. A., 5, 13(22) Lusby, W. R., 396, 398 Luscher, P., 225 Luts, A., 253,256(15) Lutzke, B. S., 551 Lyall, V., 458 Lyman, G. H., 590 M Macdermott, R. P., 557 MacDonald, H. R., 135 Machlin, L., 327 Machlin, L. J., 270, 310, 355 Ma~i~kov~t, 578, 579(31) Macintyre, J., 590 Mackenzie, C. G., 145, 147(26)
Mackenzie, J. B., 145, 147(26), 327 Maclennan, K., 591 MacNee, W., 488 MacNeil, J. M., 423,506, 533 Maddix, S. P., 601 Maddock, J., 484 Madere, R., 270, 271(12) Madzak, C., 116, 118(3), 120(3) Maeda, H., 515 Maes, D., 422 Maguire, J. J., 317, 351,354, 363, 381 Mahadevan, V., 617 Mahoney, E. M., 514 Maidt, L., 80 Maier, K., 557 Maines, M. D., 225 Maiorino, M., 317, 476, 478, 479(20), 480(4, 2O) Mak, I. T., 620-621,621(8, 10), 624(2), 625, 625(3), 626(10), 628(6), 629(6, 7) Makheja, A. N., 514 Makino, K., 220 Malek, T. R., 141 Malin, M. J., 431 Malva, J. O., 591 Manabe, M., 371 Manera, E., 415, 416(29) Manez, S., 421 Maniatis, T., 48, 50(4), 117, 121(8), 170, 171(21), 193, 197(25), 202(25), 203(25), 207(25), 213(25), 214(25), 226, 227(15), 228(15), 229(15), 230(15), 232(15) Manitto, P., 384 Miinnel, D., 152, 156(5), 157(5), 158(5) Mannisto, P., 527 Manohar, M., 191 Manoharan, M. A., 113 Manso, C. F., 552 Manwaring, J. D., 439 MaD, S.J.T., 505, 505(7, 8), 506, 509(8), 512(7, 8) Marai, L., 516, 523(31) Marchaj, A., 426 Marcinkiewicz, S., 355 Marcocci, L., 462, 526 Margison, G. P., 82 Margolis, S. A., 5, 9(27) Markland, S. S., 558,566(26) Marko, V., 574 Markwell, M.A.K., 536
AUTHOR INDEX Manner, W. N., 604 Marnett, L. J., 406, 605 Marsden, M., 488 Marsh, A. C., 396 Marsh, J. P., 191 Martelli, E. A., 415,416(32) Martensson, J., 492, 492(5-9), 493, 493(4), 494(4-9) Martin, A. M., 102 Martin, H.-D., 387(24), 388 Martin, J. S., 435 Martin, M. M., 435 Martino, V. S., 437 Marry, L., 231 Martz, B. L., 505 Marx, J. J., 588 Marzo, A., 415,416(32) Masarykovfi, M., 577 Masher, H., 484 Mashiko, S., 425 Maskos, Z., 422 Mason, J., 546, 547(20) Mason, K. E., 327 Mason, R. P., 66, 69-70, 70(30), 71, 71(32), 72(38), 424, 638 Masotti, L., 601 Massaro, E. J., 316 Matalon, S., 256 Mathews-Roth, M. M., 384 Matsubara, S., 236 Matsui, T., 311,313, 313(6), 314, 314(6) Matsumoto, S., 312-313 Matsuo, M., 312-313 Matsuqawa, Y., 514 Matsuzaki, Y., 415, 416(31) Matthews, J. R., 153 Mauer, S. I., 327 Maurer, R., 303,393 Manri, P., 415, 416(19, 22), 418(19), 41909) Mauri, P. L., 415, 416(21) Mautone, G., 487 Maxam, A. M., 35, 38(18), 103 Maxwell, S.R.J., 284 Mazess, R. B., 590 Mazumder, A., 77 Mazumder, S. C., 113 Mazur, P., 394 McCague, R., 591 McCall, J. M., 548, 550-551 McCann, J., 59
663
McCay, P. B., 294, 316, 362, 515, 523, 523(23), 525 McCloskey, J. A., 8-9, 13(39) McConathy, W. J., 516 McCord, J., 474, 541 McCord, J. M., 338, 584, 619 McDonagh, A. F., 273,406 McDonald, C. C., 590 McGowen, E. L., 331 McGrath, R. A., 89, 94(7) McHale, J. B., 355 Mclntosh, D. D., 505 McKenna, R., 283,345 McKenna, R. L., 554 McKinna, A., 590-591 McKinney, P. E., 487 McKnight, S. L., 164, 219 McLafferty, F. W., 9 McLean, L. R., 505 McManus, T. T., 252 McMillan, R. M., 454 McMiUin, D. R., 77 McMurray, C. H., 299 McNally, A. K., 513-514 McQuaid, A., 546-547, 547(20) Meadows, K. A., 77 Means, E. D., 552, 619 Medler, E. M., 491 Meduna, V., 393 Mehlhorn, R., 581 Mehlhorn, R. J., 322, 363,580, 588 Mehlman, M. A., 252 Meier, B., 152, 156(5), 157(5), 158(5) Meinkoth, J. L., 197 Meister, A., 367, 492, 492(5-9), 493,493(14), 494, 494(1, 4-9, 16), 496(2, 21), 497, 497(1, 2, 21), 498(21, 29), 499, 499(29), 500, 501(1), 502, 502(1), 503,503(1, 3), 504, 5O4(3, 5) Melander, B., 491 Melhorn, R. J., 317, 318(6), 351 Mello-Fihlo, A. C., 77 Mellors, A., 344, 362 Menck, C.F.M., 3, 45, 80, 115-116, 118(2, 3), 119, 120(3), 121 M6ndez, J., 446 Meneghini, R., 77 Meng, T., 416 Menzel, D. B., 252 Menzen, H., 456
664
AUTHOR INDEX
Merril, C. R., 108 Mertens, T., 135, 139(6) Messing, J., 40, 42(26) Metsii-Ketelfi, T., 281,282(5), 527 Meunier, M. T., 421 Meyer, K., 390, 394(14) Meyer, M., 154, 155(10), 157(10), 162(10), 163(10) Meyskens, F. L., 235 Miao, G., 164, 165(9), 166(9), 174(9) Michel, C., 384, 420, 422, 424(24), 425, 425(24), 426(65), 427(65), 428, 428(24), 429(65), 437 Michel, F., 423 Michielis, C., 373 Middleton, E., 446 Mihm, S., 135, 136(9), 137, 139, 139(9), 142, 142(8), 144(8, 23), 146(8), 148(9), 149, 492 Miki, M., 542-545, 547(17) Miles, A. M., 555, 558 Milhorak, A. T., 327 Mill, T., 311,313(7) Millart, H., 511 Miller, H., 47 Miller, K. W., 278 Miller, N. J., 279, 284 Miller-Eberhand, V., 411 Millest, A. J., 454 Milne, D. B., 278 Milne, L., 67, 70, 70(16), 71(16), 73(16), 74(16), 75(16), 76(16), 77(35) Milner, A. D., 284 Milosavljevic, E. B., 424 Min, D. B., 384 Minakami, S., 338 Minchenkova, L. E., 69 Mino, M., 321,544-545, 547(17) Miquel, J. F., 591 Mirabelli, F., 73 Mi~k, V., 574-575,577(18) Misra, H. P., 621 Mitchell, J. B., 581-582, 583(12), 586, 587(17), 588, 589(17, 22, 27) Mitchell, P., 172, 344 Miura, Y., 582 Miyake, N., 406 Miyamoto, T., 24 Miyata, A., 393
Miyauchi, Y., 342 Mizel, S. B., 141 Mizota, T., 592 Mizumoto, Y., 24 Mizuno, G. R., 300 Mizuta, E., 446 M6ckel, H., 426 Mokady, S., 400 Mold6us, P., 482, 484, 485(5), 486, 486(5), 487, 487(7), 488, 489(8), 490 Mold6us, P. M., 479 Mold6us, P. W., 489 Molnar, I., 416, 417(40) Monboisse, J. C., 421 Moncada, S., 475 Montesano, R., 26 Monti, D., 384 Monyer, H., 552 Moon, R. C., 235 Moore, D. D., 221 Moore, P. D., 118 Moore, P. K., 557 Moorhouse, C. P., 594 Mora, A., 421 Moran, E., 416, 446 Morand, O., 606 Morand, O. H., 603-604, 605(27), 606, 606(21, 27), 608(27), 609(27), 610(27), 611(27), 616(27), 618(21, 27), 619 Morehouse, L. A., 66, 67(2) Morel, D. W., 514 Morel, I., 437,441,443 Morgan, D. L., 257 Morgan, R. W., 175(3), 176, 181(3) Morgenstern, R., 477-478,478(13) Morgunov, A. A., 582 Morimoto, H., 351 Morino, Y., 341 Moroi, M., 491 Moroney, M. A., 411,424,443,446,449(15), 450, 450(15), 452(17), 453,453(12, 15) Morris, C., 569 Morris, S., 581 Morse, P. D. II, 580 Morton, M. R., 217 Moser, A., 604 Moser, A. B., 604 Moser, H. W., 603-604 Mosley, S. T., 606
AUTHOR INDEX Mossman, B. T., 191 Motchnik, P., 475 Motchnik, P. A., 24, 253, 254(11), 255(11), 269 Motoyama, T., 545,547(17) Mouret, J.-F., 17, 80, 82, 83(29), 87 Moustacchi, E., 79 Mower, R., 327 Muckel, C., 370 Mudd, J. B., 252 Mueller, M., 166, 167(14) Mueller, P. K., 262 Mukai, K., 315,351,366 Mukohato, H., 410 Miiller, A., 476, 479,479(1), 480(1, 22), 481, 482(33) Miiller, E., 123, 125 Muller, J. M., 553 Miiller, L., 456 Mfiller-Hiicker, J., 603 Mulloy, D., 475 Murahashi, S., 4 Muramatsu, T., 236 Murasecco-Suardi, P., 384 Murata, K., 24 Murphy, M. E., 328, 329(15), 384, 455 Murphy, R. F., 601 Murray, R.D.H., 446 Murray, T. K., 396 Murthy, C. P., 573 Musgrove, E. A., 601 Mustafa, M. G., 252 Mutoh, H., 256 M0tzel, P., 129 Myher, J. J., 516, 523(31)
N Nabel, G., 153, 162(9) Nackerdien, Z., 5, 6(25), 11(25) Nagano, Y., 505, 506(6) Naganuma, A., 493 Nagao, A., 403 Nagayama, K., 393 N/~her, H., 135, 139(6) Nakabayaski, S., 410 Nakabeppu, Y., 123 Nakae, D., 24
665
Nakagawa, S., 321 Nakahara, T., 321,322(12) Nakamura, K., 282 Nakamura, M., 335 Nakamura, T., 321,357 Nakane, S., 411 Nakanishi, K., 394 Nakano, M., 321,357 Narayanaswami, V., 479 Nartey, N. O., 547 Nash, G. S., 557 Nathan, C. F., 147 National Regional Council, 269 Naumov, V. V., 345, 351(12) Nauta, H., 426 Nebert, D. W., 200, 205(34) Neff, G. L., 553 Negre-Salvayre, A., 429 Neidle, S., 591 Neijt, J. P., 588 Nelis, H.J.C.F., 393 Nelson, D. P., 53 Nelson, J. H., 424 Nemec, J., 631-632,633(10), 638(6, 10) Netscher, T., 303,307(3) Neukom, C., 309 Newcombe, P. A., 590 Newcombe, R. A., 590 Newman, C. M., 491 Newton, G. L., 367,497 Ng, S. Y., 250, 251(7) Nguyen, T. D., 424 Nicholls, P., 469 Nicholson, B. J., 236, 241(14) Nicholson, R. I., 590 Nick, S., 135, 144(4) Nicotera, P., 67, 70(16), 71(16), 73(16), 74(16), 75(16), 76, 76(16) Nicotra, C., 401 Nielson, A. B., 390 Nielson, O. H., 557,569(22), 570(22) Nierenberg, D. W., 393 Nigro, R. G., 116, 118(2), 121 Niki, E., 311-312, 317, 322, 348, 356, 372, 374,382(10), 383,402,404(12), 423,460, 467, 480, 481(29), 506, 532-533, 534(19), 565,633 Nikula, K. J., 257-258, 258(3), 264(12) Nilsson, E., 488
666
AUTHORINDEX
Nilsson, U. A., 582 Nilsson, U. L., 586 Nishimura, M., 315 Nishimura, S., 17, 24-25, 59, 61(1), 63(1), 64(1), 65(I), 80 Nishiyama, J., 300 Nisonoff, A., 135 Nissinen, E., 527 Nitschmann, W. H., 580 Noack, E. A., 472 Nogala-Kalucka, M., 320 Noguchi, N., 402, 480, 481(29) Nomoto, T., 425 Nomura, H., 482 Nomura, T., 324, 326(22-24) Nordenbrand, K., 344 Norling, B., 344 Nose, K., 181, 191, 191(16) Nottenburg, C., 206 Novel, N., 462 Novikov, K. N., 404 Nungesser, E., 604 Nyberg, B., 95, 123
O Obendorf, M.S.W., 45, 59, 80 Oberley, L. W., 244 Oberlin, B., 296 Oberreither, S., 513 Oberritter, H., 337 Ocaktan, A., 443 Ocaktan, A. Z., 441 Ochial, M., 491 O'Connor, T. R., 123 Odin, F., 17, 80, 87 O'Donnell-Tormey, J., 147 O'Farrell, P. H., 181 Offermanns, H., 542 "Official Methods of Analysis" (Association of Official Analytical Chemists), 295 Ogasawara, T., 280(4), 281 Ogden, R. C., 49 Ogino, T., 341 Oguni, I., 425 Ohara, J., 141 Ohata, K., 453 Ohnishi, S., 416 Ohno, M., 491
Ojasoo, T., 591 Okamoto, T., 275 Okamura, M., 337 Okayasu, T., 547 O'Kennedy, R., 416, 446 Okenquist, S. A., 33, 40(3), 104 Oki, T., 321,357 Okoli6~iny, J., 578 Okuda, H., 446 Okudaira, H., 24 Olcot, H. S., 581 Oleinick, N. L., 3, 78 Olinski, R., 5, 6(25), 11(25) Olive, P. L., 100 Oliver, C. N., 254, 515, 523-524, 524(I), 525(1), 526(1) Oliveros, E., 384 Ol'khovskaya, I. P., 356 Ollis, W. D., 414 Olson, J. A., 235, 388 Olsson, B., 487 Olsson, L. I., 582 Olszewski, J., 514 Olton, D. S., 524 Omaye, S. T., 270, 271(13) Ondreji6v~i, O., 576, 578-579, 579(20, 31) OndriaL K., 574-575, 577(18) O'Neil, C. A., 394 O'Neill, C. A., 254 Ongun, A., 252 Ooshima, A., 505, 506(6) Oosting, R. S., 252 Ootsuyama, A., 24 Op den Kamp, J.A.F., 345, 373, 382 Oppenheim, A. B., 219 Orbell, J. D., 68 Oren, R., 250, 251(6) Orrenius, S., 67, 70(16), 71(16), 73, 73(16), 74(16), 75(16), 76, 76(16), 488, 628 Orunesu, M., 490 Osborne, D. J., 452 Osebold, J. W., 254, 257 Osheroff, N., 631 Oshino, N., 66 Oster, K. A., 619 Oszmianski, J., 424 Ota, S., 256 Otsuka, H., 446 Otter, R., 476, 479(2), 482(2) Ouchi, S., 366
AUTHOR INDEX Ouellet, R., 81, 83(23) Overvad, K., 24, 31(28) Owen, G. M., 331 Owens, K., 603,609(5), 611(5), 620, 621(9) Owyang, C., 546 Oyama, T., 415,416(31)
P Pace, D. M., 257 Pacht, E. R., 256 Pacifici, R. E., 175(5-7), 176 Packer, L., 156, 254, 255(26), 274, 310-311, 316-318, 318(6), 321-322, 343-344, 346(5), 351, 353(10), 354, 357-358, 358(22), 359(22), 361,363, 363(26, 34), 366, 366(26, 33), 368, 371,371(13), 372373, 373(9), 379(2), 381, 381(6), 382(9, 25), 383, 401, 409(9, 11), 454-456, 458(5), 459(8), 460-462,467, 526-528, 533,534(23), 537(4), 577, 580, 638 Padrini, R., 511 Padulo, G. A., 422 Paganga, G., 592, 601(40) Pagani, S., 457 Pageaux, J. F., 591 Pagonis, C., 423 Pallardo, F. V., 367 Pallard6, F. V., 371 Palma, L. A., 273 Palmer, R.M.J., 475 Palombini, G., 601 Palozza, P., 372 Paltauf, F., 612,614(43) Pan, Y.-C.E., 164, 165(9), 166(9), 174(9) Panalaks, T., 396 Pang, H., 9 Panter, S. S., 568, 569(45) Papathanasiou, M., 200, 205(34) Paramova, L. I., 386, 387(22) Parathasarathy, S., 513 Pardee, A. B., 207 Pardini, R. S., 424 Park, C. H., 438 Park, D., 403 Park, E. M., 17, 24, 24(9), 27(9), 32(9), 80 Park, J. W., 24, 67, 117 Parker, R. A., 505(8), 506, 509(8), 512(8)
667
Parker, R. S., 389 Parnham, M. J., 477, 481 Parr, I. B., 591 Parthasarathy, S., 513-515, 522 Paschke, E., 604 Pascoe, G. A., 328 Pascual, C., 425 Pasdeloup, N., 437,441,443 Pasquali, P., 344 Patel, B. S., 515,523,523(23) Patel, L., 164, 166(7), 174(7), 224 Paterson, M. C., 54 Patterson, B., 269 Pattison, T. S., 397 Paul, D. L., 236 Paul, W. E., 141 Pauling, L., 390 Paulsen, O., 487 Pavlotsky, N., 423 Payli, M., 421,443,446, 453(13) Pechard, M. R., 397 Pelle, E., 422 Penhoet, E. E., 53 Perdrix, L., 423 P6rez-Campo, R., 332, 333(4), 334(4), 337(4) Periasamy, M., 191 Perin, L., 602 Perruchot, T., 423 Perry-O'Keefe, H., 202 Persson, C.G.A., 253,256(15) Pesek, C. A., 392 Pessara, U., 137,492 Peters, M., 557 Peters, T., Jr., 53,482 Petry, T. W., 551 Petzoldt, D., 135-136, 139(6), 140(11) Pfander, H., 388 Pfanstiei, J., 557 Pflaum, M., 129-130 Phillips, F., 617 Phillips, M. W., 615 Phillips, T. M., 620, 621(10), 626(10) Photo, P., 527 Pickett, C. B., 217 Pidoux, M., 226 Piechaczyk, M., 231 Pierson, H. F., 397 Pierson, T. S., 554 Piestri, S., 475 Pieters, L.A.C., 416
668
AUTHORINDEX
Pietta, P., 415, 416(19, 21, 22, 29), 418(19), 419(19) Piette, J., 79, 121 Pihan, G., 528 Pincemail, J., 462 Pinedo, H. M., 631 Pinson, A., 582, 589(22) Pipkorn, U., 253,256(15) Pippurl, A., 527 Piretti, M. V., 416 Piris, J., 557 Plenevaux, A., 477 Plopper, C. G., 257-258, 258(3) Podhajer, O. L., 602 Pohl, C., 135, 139(6) Polgar, A., 390, 393 Poli, G., 616 Poll, T., 601 Pollack, S. E., 52 Pollak, M. N., 591 Pollow, K., 592 Polverelli, M., 6, 17, 80, 87 Pomerantz, S. H., 632 Pomeroy, J. J., 438 Pons, M., 591 Ponte, P., 250, 251(7) Poor, C. L., 389, 397 Popov, A., 320 Porter, L. J., 430, 433 Postlethwait, E. M., 253,256(9) Potapovitch, A. I., 421 Pottmeyer-Gerber, C., 135, 144(4) Pou, S., 77, 561,566(31) Poulsen, H. E., 24, 31(28) Poulsen, L. L., 482 Povirk, L. F., 129 Powell, R., 318, 456 Powles, T. J., 590 Powrie, F., 504 Powrie, R., 492, 493(2), 496(2), 497(2) Pradhan, D. S., 456 Preece, P. E., 590 Pregenzer, J. F., 376,402, 550-551 Price, M. L., 434-435, 435(15) Price, M. P., 432 Price, S., 80 Pries, C., 616 Prosser, E., 446 Proudfoot, K., 601 Proudrnan, K. E., 454
Priitz, W., 66, 67(7), 68(7), 69(7, 27), 73(7), 74(7) Priitz, W. A., 68, 73 Pryor, W. A., 252-253,506, 619 Przybyszewski, J., 24, 36, 38(19) Ptzoldt, D., 492 Puhl, H., 513 Pung, A., 237 Puppo, A., 421,422(20) Purl, R. N., 492, 493(1-3), 494(1), 496(2), 497(1, 2), 504 Pystynen, J., 527
Q Qimin, L., 416 Quackenbush, F. W., 396 Quayle, J. R., 612 Quigley, G. J., 68 Quilliam, M. A., 7 Quinlan, G. J., 594 Quinn, M. T., 513 Quinn, P. J., 322,344-345,346(5), 528,591, 601(35)
R Rabani, J., 426,427(70) Rabenstein, D. L., 498 Rabinowitz, J. L., 616 Race, M., 373 Radany, E. H., 34 Radford, I. R., 89 Radi, R., 273,475 Radom, J., 606 Raetz, C.R.H., 604, 605(27), 606(21, 27), 608(27), 609, 609(16, 27), 610(27), 611(27), 616(27), 618(21, 27), 619 Rahmsdorf, H. J., 170 Raisys, V. A., 335, 337(7) Raju, P. A., 492 Ramakrlshnan, N., 3 Ramasarma, T., 343 Ramos, C. L., 561,566(31) Rampton, D. S., 555 Ranalder, U. B., 302-303 Ranasarma, T., 372 Randall, R. J., 516, 594, 598(50)
AUTHOR INDEX Randoux, A., 421 Rangaswamy, S., 604 Ranney, H. M., 497 Ransom, J. A., 113 Rao, G.H.R., 327 Rao, N. U., 581 Rao, R.D.N., 638 Rapkin, L., 33, 40(3) Rasmussen, R. E., 257,258(5) Rasokat, H., 135, 139(6) Rattner, A., 438 Ratty, A. K., 421,437 Rauscher, F. J. III, 164, 165(2), 166(2, 7), 167(2), 168(2), 170, 174(7), 224 Rava, A., 415 Ravanat, J.-L., 79, 81-82, 83(23-25), 85(25), 87(25), 88(24) Raynaud, J. P., 591 Razandi, M., 256 Razell, W, E., 52 Reames, S. A., 393 Rechtin, A. E., 505(7), 506, 512(7) Recknagel, R. O., 404, 438 Reddy, J.M.C., 316 Reddy, P., 5, 9(27) Redegeld, F.A.M., 367, 368(10) Reed, C. J., 67, 73(14), 75(14) Reed, D., 328 Reed, D. J., 367, 368(8), 369(8), 371(8) Reif, D. W., 475 Remacle, J., 373 Renson, M., 477 Retel, J., 631 Rettenmaier, R., 294 Rettmer, R. L., 270 Revzin, A., 165, 171(10) Reznick, A. Z., 254, 255(26), 461 Reznikoff, C. A., 237 Riazzudin, S., 123 Ribeiro, D. T., 45, 80, 116, 118(3), 119, 120(3), 121 Rice, D. A., 299 Rice-Evans, C., 279,284, 592, 601(40) Richard, M.-J., 17, 80 Richardson, C. C., 42, 44(30) Riche, C., 592 Richelmi, P., 73 Richie, J. P., 368 Richter, C., 24 Rickard, R. C., 60, 62(6), 80
669
Riddell, R. H., 555 Rieber, P., 152, 153(3), 155(3), 157(3), 162(3), 217, 224 Riesz, P., 582 Rigo, M., 317 Riis, P., 557, 569(22), 570(22) Riley, D., 590 Riley, W. T., 35, 37(17), 40(17), 104 Rimm, E. B., 269 Rink, H., 89, 90(12), 91(12), 94(12), 96(12), 99, 100(12) Rio, M. C., 602 Rios, J. L., 421 Rippstein, A., 306 Riss, G., 302-303,306, 310(4) Rissel, M. Y., 438 Rizzardo, E., 422 Rizzo, W. B., 604, 615 Robak, J., 411,421,422(16), 423(16), 424(7) Robberson, D. L., 127 Robbins, C. T., 435,436(20) Roberfroid, M., 490 Roberts, R., 523 Robertson, J.F.R., 590 Robins, A., 590 Robinson, C., 452, 555 Rochette, L., 462 Rockholm, D. C., 435 Rodgers, M.A.J., 384-385 Roeder, R. G., 159, 167, 168(16) Roederer, M., 492 Roelofsen, B., 345, 373, 382 Rogers, M., 236, 241(14) Rogers, S. G., 123 Rojas, C., 332, 333(4), 334(4), 337(4) Roman, J., 631,638(6) Romanovsky, J. C., 262 Romay, C., 425 R6mer, A., 481,482(33, 34) Roosen, O., 4 Rose, C. S., 327 Rose, J. R., 511 Rosebrough, N. J., 516, 594, 598(50) Rosen, C. A., 166, 167(13) Rosen, G. M., 561,566(31) Rosen, H., 506, 514 Rosenfeld, M. E., 514 Rosenkrantz, H., 327 Rosolowsky, M., 514 Ross, D., 628
670
AUTHORINDEX
Ross, R., 513 Ross, R. K., 590 Ross, S., 189 Roth, E. E., Jr., 328 Roth, S., 135, 140(5), 142, 142(5), 144, 144(5, 23), 145(7, 24), 146(7), 149, 149(7, 24) Rotheneder, M., 513 Rouge6, M., 480 Rouseff, R. L., 416 Roved, A., 476, 478,479(20), 480(4, 20) Rowlands, M. G., 591 Roy, P. K., 411 Rfidiger, H., 130 Rumyantseva, G. V., 66 Rush, J. D., 51,422 Rushin, W. G., 397 Rushmore, T. H., 217 Rusnack, M. G., 331 Russett, M. D., 398 Russo, A., 581-582, 583(12), 586, 587(17), 588, 589(17, 22, 27) Rustow, B., 256 Rusznyak, S., 429 Rydberg, B., 89, 94(15), 96(15), 99(15), 100, 102 Ryle, A. P., 488 Ryter, S. W., 175(5), 176
S
Saad, Z., 590 Sabovljev, S. A., 66, 78(4), 79(4, 6) Sabrouty, S. E., 231 Sacchi, N., 193, 227 Sacks, N.P.M., 591 Sadrzadeh, S.M.H., 568, 569(45) Sfiez, J. C., 236 Safadi, A., 527, 537(4) Safayhi, H., 476, 479(2), 482(2) Saffitz, J. E., 620 Safsten, B., 527 Sage, E., 79 Sagfipanti, J.-L., 35, 67, 73(13), 75(13) Saiki, K., 390, 391(16), 392(16), 393(16) Salnsbury, J.R.C., 590 Saint-Jean, R., 385 St. John, T., 194, 206 Salto, M., 374 Saito, R., 557
Saito, T., 356, 372, 382(10), 402, 404(12) Saitoh, Y., 601 Sakiyama, S., 181, 191(16) Sakurai, H., 425 Salata, K., 557 Salditt, M., 125 Salim-Hanna, M., 423 Salin, M. L., 474 Salmenper~i, L., 296 Salmon, S. L., 175 Salser, W., 218 Saltzman, B. E., 262 Salvati, A. M., 569 Salvayre, R., 423,429, 606 Sambrook, J., 48, 50(4), 117, 121(8), 170, 171(21), 193, 197(25), 202(25), 203(25), 207(25), 213(25), 214(25), 226, 227(15), 228(15), 229(15), 230(15), 232(15) Sambucetti, L. C., 170 Samokyszyn, V. M., 406 Samuni, A., 580-582, 583(12), 586, 587(17), 588, 589(17, 22, 27) Samuni, U., 582, 587(17), 589(17) Sanadi, D. R., 455 Sander, L. C., 397 Sandu, I. S., 565, 566(36), 567(36), 568(36) Sandy, M. S., 479, 489 Santi, L., 96 Santos, M. J., 604 Saotome, H., 411 Saran, M., 384, 420, 422, 424(24), 425, 425(24, 25), 426(25, 65), 427(25, 65), 428, 428(24, 25), 429(65), 437 Sarasin, A., 45, 80, 115-116, 118(2, 3), 119, 120(3) Sarett, H. P., 491 S~irkioja, T., 514 Sarycheva, I. K., 356 Sasajima, M., 411 Sasaki, H., 415,416(31) Sastre, J., 367, 371 Satake, Y., 311,313(6), 314(6) Satoni, D. K., 512 Satonin, D. K., 511 Saucy, G., 309 Saul, R. L., 619 Saunders, R. D., 619 Saziki, R., 411 Scaiano, J. C., 422 Scala, J., 397
AUTHOR INDEX Scalia, M., 421 Scandura, O., 270 Scarabelli, L., 490 Scarpa, M. A., 317 ~asnfir, V., 574, 578 Schaap, A. P., 605 Schacher, A., 167 Schalch, W., 385, 387(14) Schalkwijk, C., 345,373 Schalkwijk, C. G., 604 Schapira, D. V., 590 Schaus, E. E., 270, 271(13) Schemer, E., 601 Scheinberg, I. H., 546 Scherberich, P., 542 Scherch, H. M., 552 Schiff, L. J., 257,258(4) Schillaci, M., 269, 304 Schilsky, M., 546 Schloemann, S. R., 557 Schmid, R., 303,307(3) Schmidt, H., 135, 144(4) Schmidt, K., 337 Schmitz, H. H., 389, 397 Schneider, G., 422 Schneider, J. A., 136 Schneider, J. E., 3, 21, 59, 80 Scholich, H., 455 Scholz, R. W., 316 Sch6neich, C., 479 Schrakamp, G., 603-604 Schram, K. H., 7, 9 Schramel, P., 546 Schrappe, M., 135, 139(6) Schraufstatter, I., 77 Schreck, R., 151-152, 153(3), 154, 154(4), 155(3, 10), 156(4, 5), 157(3, 5, 10), 158(5), 160, 161(15), 162(3, 10), 163(10), 217,224 Schreurs, W.H.P., 335, 337(9) Schrijver, J., 335, 337(9) Schroeder, W. A., 390 Schubert, M. P., 144 Schuchmann, H.-P., 67 Schuchmann, M. N., 3 Schiiep, W., 294, 296 Schuler, R. H., 428 Schulte-Frohlinde, D., 426 Schultz, E., 527 Schultz, W. A., 80
671
Schulz, H., 415,416(28) Schulz, W. A., 45, 59 Schuppe, I., 488 Schutgens, R.B.H., 603-604 Schwartz, H. S., 89, 90(11), 91(ll), 94(11), 96(11), 100(11) Schwartz, S. J., 394, 396-397 Schwarz, W., 388-389, 394, 398(10, 27) Schwenke, D. C., 505, 514 Schwiers, E., 331 Scita, G., 317-318, 358, 363, 372-373, 381(6), 401,460, 533,534(23) Scott, C. G., 309 Scott, T. W., 603 Scurlock, R., 480 Searls, R. L., 455 SeaweU, P. C., 34 Seawell, P. S., 127 Sechi, M., 344 Sedghi, S., 555 Sedlfik, J., 578, 579(31) Sedman, S. A., 108 Seegmiller, J. S., 136 Sefranka, J. A., 505 Seher, A., 320 Seher, V. A., 360 Seibel, K., 600 Seidman, J. G., 221 Seidman, M. M., 121 Seifart, K., 199 Sekaki, A. H., 462 Sekiguchi, M., 123 Sekiya, K., 446 Sellers, R. M., 284, 288(15) Sen, J. N., 312,403 Sen, R., 172 Sentjurc, M., 580 Serbinova, E., 321, 357, 358(22), 359(22), 363,366(33), 382(25), 383,456, 461,577 Serbinova, E. A., 317-318, 343-344, 346(5), 351, 353(10), 354, 361, 363, 363(26), 366(26), 372-373, 373(9), 381, 382(6), 528, 638 Sergent, O., 437, 441 Serra, D., 490 Sestili, P., 77 Setchell, B. P., 603 Sethy, V. H., 553 Sevanian, A., 175(4), 176, 553 Shaltiel, S., 254
672
AUTHOR INDEX
Sharma, M., 603 Sharma, O. P., 411 Sharp, P. A., 172 Sharp, R. J., 280(3), 281 Shedlofsky, S., 438 Sheffner, A. L., 491 Shen, B., 398 Shenk, T., 189 Shevach, E. M., 141 Shibahara, S., 225 Shibanuma, M., 181, 191, 191(16) Shibutani, S., 80 Shieh, J. J., 284, 288(15) Shigenaga, M. K., 16-17, 24, 24(9), 25(25), 26(25), 27(9), 32(9), 67, 80, 117 Shimamura, T., 393 Shimasaki, H., 402, 506 Shimizu, K., 402 Shimomura, K., 592 Shinagawa, H., 220 Shing Ho, P., 68, 70(23) Shinitzky, M., 324, 356 Shoemaker, R. H., 638 Shridi, F., 421,424(7) Shull, S., 191 Shvedova, A., 318,456 Shvedova, A. A., 404 Sichel, G., 421 Sicot, N., 423 Siemsen, F., 553 Sies, H., 3, 45, 59, 79-80, 116, 118(2, 3), 120(3), 121, 129, 338, 367, 368(9), 370, 384-385, 385(12), 386(18), 387(12, 18, 24), 388-389, 394, 398(10, 27), 455456, 468, 476, 479, 479(1), 480, 480(1, 22), 481,481(3), 482, 482(34), 572-574, 574(5), 575,575(5), 576(5), 577,578(13), 581 Sigal, E., 514 Sigman, D. S., 77 Silbert, L. S., 428 Sima, P., 256 Simic, M. G., 24, 31(27), 422, 573, 620 Simmon, V., 59 Simmonds, N. J., 555 Simmons, R. D., 475 Simon, I., 592 Simon, M., 438 Simon, T. C., 514 Simoni, R. D., 375,460
Simons, E. R., 423 Simpson, E., 146 Simpson, K. L., 388 Sinclair, P. R., 438 Singh, H., 172 Singh, S., 515 Singhal, R. K., 493,494(16) Sinha, B., 631,638(6) Sinha, B. K., 631-632, 632(7), 633(10), 638(10) Sinkina, Y. B., 356 Sirtori, C. R., 515 Sive, H. L., 206 Sj6din, K,, 488 Skipper, P. L., 472 Skipsky, V. P., 603 Sklan, D., 401 Sklar, L. A., 375,460 Slater, T. F., 437, 535, 601,616 Slevin, M. L., 631 Slezfik, J., 578 Sloane-Stanley, G. H., 599 Smedegard, G., 557 Smith, B. F., 256 Smith, C., 592, 594(37, 39), 595(37, 39), 596(37, 39), 597(39), 601(37, 39) Smith, D. L., 9 Smith, I., 591 Smith, J. C., 397-398 Smith, M. T., 479, 489 Smith, S. L., 552 Smith, W. E., 328 Smith, W. P., 422 Snodderly, D. M., 398 Snyder, F., 603, 604(2) Snyder, L. R., 415,417 Soares, J. H., 278 Soczewinski, E., 415,417(24) Solar, S., 3 Solheim, E., 592 Solomon, D. H., 422 ~olt6s, L., 574 Somerharju, P., 606 Sorenson, J.R.J., 542 Sosnovsky, G., 581 Sowell, A. L., 397 Sparrow, C. P., 514-515 Spector, A., 477 Speder, A., 490 Speek, A. J., 335, 337(9)
AUTHOR INDEX Speranza, G., 384 Spiegelman, B. M., 170 Spikes, J. D., 607, 633 Spiller, G. A., 397 Spinnewyn, B., 475 Spinnewyn, S., 462 Sporn, M. B., 591 Spray, D. C., 236 Staab, H. J., 600 Staal, F. J., 492 Stadtman, E. R., 254, 515,524 Stafford, H. A., 429 Stahl, W., 388-389, 394, 398(10, 27), 482 Staicup, A. M., 394 Stampfer, M. J., 269 Stanley, W. C., 368, 371(13) Starke-Reed, P. E., 515, 523-524, 524(1), 525(1), 526(1) Stag, A., 574-575, 577(18) Steenken, S., 3, 45, 59, 80, 83, 572-573, 574(5), 575(5), 576(5) ~tefek, M., 573-574, 576-578 Steighner, R., 129 Stein, B., 170 Stein, E. A., 389 Steinberg, D., 505, 513-515 Steiner, M., 327 Steinherz, R., 492(6), 493,494(6) Stenson, W. F., 557 Stern, G., 579 Sternlieb, I., 546 Stettmaier, K., 422, 428 Stevens, P. A., 256 Stevens, T.R.J., 555 Stewart, H. J., 590 Stivala, L., 73 Stocker, R., 254, 269-270, 273,273(10), 344, 354, 372,406 Stockert, R. J., 546 Stocks, J., 280(3), 281 Stoewe, R., 66, 67(7), 68(7), 69(7), 73(7), 74(7) ~tolc, S., 572, 573(1), 578(3), 579 Stole, E., 492(9), 493,494(9) Stoll, I., 602 Storz, G., 151, 156(1), 187, 214, 216(18), 217-218, 220, 224 Sttisser, R., 422 Stoyanovsky, D., 638 Stoyanovsky, D. A., 343,631
673
Stoytchev, T. S., 344, 346(5), 528 Strait, L. A., 504 Strauss, B. S., 118 Streitwieser, D., 59 Stremmenos, C., 601 Strickland, T., 506 Striegl, G., 513 Strong, L., 189 Struhl, K., 37, 40(21), 42(21), 110,221 Strumeyer, D. H., 431 Stueber, D., 166, 167(14) Stump, D. D., 328 Styk, J., 578 Subar, A., 269 Subbiah, M.T.R., 283 Subramanian, M., 456 Sugimoto, T., 256 Sugioka, K., 321,357 Sugita, Y., 137, 139(15) Suko, M., 24 Sumida, S., 317, 318(6), 322 Summer, K. H., 546 Sun, F., 553 Sun, W., 258, 264(12) Sundler, F., 253,256(15) Sundquist, A. R., 384-385,386(18), 387(18), 388-389, 394, 398(10, 27), 573, 574(5), 575,575(5), 576(5) Suprunchuk, T., 384 Surawicz, T. S., 590 Sussman, M. S., 619 Sutherland, M. W., 421 Sutherland, R. L., 601 Sutton, C. L., 77 Suzuki, N., 425 Suzuki, Y., 458 Suzuki, Y. I., 156 Suzuki, Y. J., 254, 255(26), 454-456, 458(5), 459(8), 461,526-528, 537(4) Svensson, C., 253, 256(15) Swaak, A.J.G., 561,566(30) Swain, T., 432 Swallow, A. J., 350 Swanson, C., 456 Swanson, C. E., 588 Swansson, C., 318 Swartz, H. M., 580-581 Sweeney, J. P., 396 Sweet, F., 257 Sweet, W. E., 257
674
AUTHOR INDEX
Swies, J., 411 Syenson, W. F., 557 Szabo, M. E., 475 Szabo, S., 528 Szent-Gy6rgyi, A., 429 Sz6csovfi, H., 574
T Tabor, S., 37, 40(21), 42, 42(21), 43(29), 110 Tachizawa, H., 511 Taller, J. M., 603-604 Tainsky, M., 189 Takada, M., 274 Takahama, U., 420 Takahashi, M., 322, 378, 383,506, 565 Takahashi, T., 547 Takamatsu, K., 366 Takamoto, M., 446 Takatsuki, K., 343 Takeda, S., 415, 416(31) Takegoshi, T., 482 Takei, H., 422 Takeichi, N., 547 Takekoshi, Y., 547 Takenaka, Y., 544 Takeshida, M., 80 Takii, T., 390, 391(16), 392(16), 393(16) Tamaka, A., 356 Tamura, K., 24 Tamura, S., 588 Tanabe, T., 545, 547(17) Tanaka, A., 320, 321(4) Tanaka, I., 411 Taniguchi, N., 545, 547(17) Tanimura, R., 317 Tankanow, R., 545, 547(18) Tarmenbaum, S. R., 472 Tano, K., 24 Tanooka, H., 17, 24 Tao, K., 220 Taoukis, P. S., 392 Tappel, A. L., 344, 355, 362, 477-478, 478(14), 560, 566(27) Tapper, D. P., 42 Tarboletti, G., 602
Tarkington, B. K., 254, 257-258, 264(12), 475 Tartaglia, L. A., 151, 156(1), 187, 216(18), 220, 224, 244, 250, 251(8, 9) Tatsuta, T., 321,356 Taub, I. A., 284, 288(15) Tauber, A. 1., 423 Tavendale, R., 590 Tavill, A. S., 438 Taylor, B. M., 553 Taylor, G. W., 491 Taylor, H. A., 505 Taylor, K. B., 454 Tchapla, A., 397 Tchou, J., 80 Tedeschi, R. E., 505 Teebor, G. W., 58 Teff, D., 219 Tegelaers, W.H.H., 603 Teicher, B. A., 494 Tentori, L., 569 T6oule, R., 6, 33 Terano, A., 256 Terao, J., 372, 403 Terao, K., 506, 565 Terekhova, S. F., 345 Terlinden, R., 481,482(33, 34) Tesoriere, L., 401 Tezuka, T., 422 Thiange, G., 477 Thirion, A., 462 Thomas, C., 66, 67(2) Thomas, C. E., 316, 506-507 Thomas, E. L., 571 Thomas, P. S., 196 Thompson, D. H., 604 Thompson, D.F.T., 401 Thompson, J. A., 577 Thompson, K. E., 544 Thor, H., 488 Thornally, P. K., 502 Thornes, R. D., 416, 446 Thorpe, G.H.G., 284 Tithe', 572, 573(1) Tiegs, G., 476, 479(2), 482(2) Tietze, F., 627 Tillotson, J. A., 331 Ting, S. V., 416 Tingey, D. T., 21
AUTHOR INDEX
Titani, K., 338 Tjian, R., 165 Tjonneland, A., 24, 31(28) Tochimaru, H., 547 Toda, S., 478 Togashi, Y., 547 Tokfirov~t, J., 578, 579(31) Tokumura, A., 322 Tolbert, N. E., 536 Toledano, M. B., 152, 218 ToUes, R. L., 325 Tomasi, A., 580 Tomomura, M., 236 Toney, J. H., 40, 41(28), 42(28) Torel, J., 421,437 Torresi, J., 491 Tosaki, A., 474, 523 Towsend, L. B., 9 Toyo6ka, T., 484 Trakshel, G. M., 225 Treutter, D., 415, 416(27) Tromvoukis, Y., 225 Trozzolo, A. M., 384 Truelove, S. C., 557 Trull, F. R., 273 Trumble, M. J., 316 Truscott, T. G., 385, 387(14) Trush, M., 631 Tsai, C. L., 590 Tsao, P. S., 475 Tsuchiya, J., 317, 356, 402 Tsuchiya, M., 357-358, 358(22), 359(22), 371-372, 381(6), 401, 454-456, 458(5), 459(8), 460-461, 467, 526-528, 533, 534(23), 537(4), 577 Tsuchiya, N., 321,356 Tsuiji, T., 24 Tsukada, T., 513 Tsukamoto, T., 604 Tsukatani, H., 322 Tsukayama, M., 453 Tsukida, K., 390, 391(16), 392(16), 393(16) Tsuneyoshi, H., 491 Tubaro, A., 446 Tuck, L. D., 504 Tunek, A., 488 Tunstallpedoe, H., 590 Tur~iny, P., 579 Turi-Nagy, L., 578
675
Tyler, K., 257, 258(4) Tyrrell, R. M., 224-226 Tyurin, V. A., 344, 346(5), 528
U Uchiyama, T., 150 Udardy, A., 199 Ueda, T., 304 Ueland, P. M., 592 Ueno, I., 422, 425(40) Ueno, Y., 311,313, 313(6), 314, 314(6), 406 Ulick, S., 327 Umezawa, I., 321,356 Underwood, M., 493 Urano, S., 351,506, 565 Uraz, V., 576, 579, 579(20) Ursini, F., 317,425,476,478,479(20), 480(4, 20) U.S. Code of Federal Regulations, 262 Usui, N., 631,632(7) Utsumi, H., 582 Utzinger, G. E., 504
V V~iclavincowi, V., 296 Vahoouny, G. V., 620, 621(9) Vairetti, M., 73 Vale, M.G.P., 591 Vale, P., 591 Valente, A. J., 515 Valenza, M., 401 Valerio, G., 511 van Asbeck, A. B., 588 van Fassen, F. E., 588 Van, N. F., 580 van Bennekom, W. P., 367, 368(10) Vance, J. E., 603 van den Berg, H. W., 601 van den Berg, J.J.M., 345, 373, 382 Van den Bosch, H., 603-604 Vander Jagt, D., 504 Van der Veen, J., 581 van der Vliet, A., 254 van der Zee, J., 69 Van de Vorst, A., 121
676
AUTHOR INDEX
van Dijk, H., 423 Vane, J. R., 411 Van Eijk, H. G., 561,566(30) Van Ginkel, G., 553 Van Golde, L.M.G., 252 Van Greevenbroek, M.M.J., 252 van Hees, P.A.M., 557 Van Kessel, J., 423, 506, 533 van Kuijk, F.J.G.M., 381 Van Langenhove, A., 557 van Lier, J. E., 55, 81-82, 83(23, 29), 631, 633, 635 Van Maanen, J.M.S., 631 Van Scoyoc, S., 434, 435(15) Van Sickle, W. A., 505(7), 506, 512(7) Van Someren, R.N.M., 555 van Steveninck, J., 69 van Tongeren, J.H.M., 557 van't Veld, A. A., 553 van Zandvoort, M.A.M., 553 Varfolomeev, V. N., 582 Varriale, A., 488 Vatassery, G. T., 327-328 Vecchi, M., 303,307(3), 309(1), 390, 393 Ventresca, G. P., 487 Verheof, J., 252 Verkerk, A., 627 Verly, W. G., 111, 113, 113(12) Vermeulen, N.P.E., 478 Vernon-Roberts, B., 557 Vesioli, O., 373 Vetter, W., 303,307(3) Videla, L. A., 423 Vieira, J., 40, 42(26) Vigny, P., 80, 81(12) Viguie, C., 368, 371(13) Vile, G. F., 401 Villar, A., 446 Villarejio, M. R., 167 Vifia, J., 367, 370(7), 371 Virelizier, J.-L., 153 Vistisen, K., 24, 31(28) Vitale, E., 511 Vliegenthart, J.F.G., 423 Vlietinck, A. J., 416 Voest, E. E., 588 Vogt, T., 415,416(23) Voisin, E., 530 Volkmer, C., 557, 566(24), 568, 569(41), 570(41)
Volkova, L. M., 582 von Bruchhausen, F., 421 von Ritter, C., 256, 570(49), 571 yon Sonntag, C., 3, 67 yon Voigtlander, P. F., 552 Voulalas, P. J., 164 Vuilleumier, J. P., 296
W Waggoner, A. S., 325 Wagner, D. A., 472 Wagner, G. R., 420 Wagner, J. R., 55, 81-82, 83(29) Wagner, R. J., 635 Wahl, G., 189 Wahl, G. M., 197 W~hlander, L., 491 Wakabayashi, T., 321 Wakasugi, N., 153 Wakefield, L. M., 591 Wakui, Y., 415,416(31) Wallace, D. M., 48, 193 Wallace, S. S., 33-34, 40, 40(3) Walldins, G., 512 Wallet, J. C., 421,423(13), 425(13) Walshe, J. M., 542, 547 Waither, J. U., 603 Walther, W., 302-303,307(3) Wanders, R.J.A., 603-604 Wang, A.H.-J., 68, 70(23) Wang, B. E., 421 Wang, C. H., 325 Wang, F., 164, 165(9), 166(9), 174(9) Wang, Y., 337 Wang, Y. M., 328,421 Ware, K., 565, 566(36), 567(36), 568(36) Warthesen, J. J., 392 Washko, P. W., 270, 337 Wasil, M., 557, 571(18) Wassail, S. R., 528 Wassen, J. B., 491 Watanabe, H., 137, 139(14), 321, 356, 478 Watanabe, N., 341 Watanabe-Kohno, S., 411,453 Watson, J. J., 17, 60, 62(6), 80 Watson, J. T., 8, 11, 11(38), 14(46) Watson, W. A., 487 Watterson, J. J., 434
AUTHOR INDEX Wayner, D.D.M., 269, 280, 286(1), 543 Weast, R. C., 374 Webb, A., 303-304, 310(2) Webb, A. C., 601 Weber, G., 309 Weber, R., 477 Wedner, H. J., 135 Wegher, B. J., 5, 11(28) Weglicki, W. B., 620-621, 621(8-10), 624(2), 625, 625(3), 626(10), 628(6), 629(6, 7) Wehr, C. M., 17, 24(9), 27(9), 32(9), 80 Weibezahn, K. F., 100 Weigert, W. M., 542 Weil, T. J., 581 Weimann, B. J., 294 Weiner, L. M., 66 Weinfeld, M., 54, 87 Weinges, K., 463 Weiser, H., 294, 303, 309(1), 310(4) Weiss, B., 123 Weiss, J. H., 202 Weiss, R., 477,481(6) Weiss, R. H., 577 Weiss, S. J., 619 Welankiwar, S., 393 Welch, R. W., 337 Wellman, R. B., 389 Wellner, V. P., 492, 493(3) Wells, J. V., 135 Wendel, A., 476, 478, 479(2), 482(2), 560, 566(28) Wenzel, D. G., 257 Werner, T., 387(24), 388 West, M., 17 West, M. S., 3, 21, 59, 80 Westmore, J. B., 7 Wever, R., 619 Wheeler, T., 590 Whitburn, K. D., 284, 288(15) Whitcutt, J. M., 257 White, E., 8, 13(39) White, G. P., 438 White, J. G., 327 Whitehead, T. P., 284 Whitehouse, M. W., 557 Whitfield, M. K., 256 Whittam, J. H., 397 Wiebe, D. A., 590 Wieland, S., 514
677
Wiese, A. G., 175(4, 7), 176, 181, 186(14) Wilbrandt, R., 390 Wild, C. P., 26 Wild, D., 126 Wiles, D. M., 384 Wilkins, L. R., 237 Wilkinson, F., 385 Wilkinson, G., 69 Willerson, J. T., 514 Wilier, W. C., 269 Williams, C. W., 552 Williams, E. C., 590 Williams, J., 189 Williams, R., 89, 94(7) Williams, V. M., 430 Willis, A. L., 327 Willson, R. L., 426 Wilpart, M., 490 Wilson, A., 590 Wilson, D. W., 257-258, 258(3), 264(12) Wilson, J. H., 119 Wilson, R. B., 565, 566(35) Wilson, S. R., 477 Wilson, T., 384 Winitz, M., 496 Winkelmann, J., 481 Winkelsberg, D., 582, 589(22) Winship, D., 555 Winter, R., 428 Winterbourn, C. C., 401 Winterle, J., 311,313(7) Wise, R. M., 620, 621(9) Wise, S. A., 278 Wiseman, H., 590-592, 592(32, 33), 594(33, 37, 39), 595(32, 37, 39), 596(33, 37, 39), 597(33, 39), 598(36, 47, 48), 599(48), 600(47), 601(33, 35-37, 39, 40, 47, 48), 602(36) Wishnok, J. S., 472 Wistort, P. M., 123 Witt, E., 318 Witting, L. A., 355,362(3) Witztum, J. L., 513 Wolbis, M., 421,424(7) Wolf, C., 602 Wolfgang, G.H.I., 551 Wollmer, P., 253,256(15) Wolohuis, J., 95 Wong, G.H.W., 244, 250, 251(8) Wong, P. K., 17, 21, 59-60, 62(6)
678
AUTHOR INDEX
Wong, P. T., 80 Wood, M. L., 17 Wood, R.A.B., 590 Woodward, A. J., 484 Wosikowski, K., 592 Wratten, M. L., 553 Wu, J. F., 523,524(1), 525(1), 526(1) Wu, R., 257-258, 264(12) Wurm, G., 411,421 Wyatt, G. R., 5 X Xanthoudakis, S., 163-164, 166(9), 167(8), 174(9) Xue, L., 3 Xue, L.-y., 78
165(8, 9),
Y Yagi, K., 508, 591 Yalowich, J., 638 Yalowich, J. C., 631 Yamada, T., 556-557,566(24) Yamada, T. I., 568,569(41), 570(41) Yamaguchi, M., 335 Yamaizumi, Z., 80 Yamamoto, A., 514 Yamamoto, F., 25 Yamamoto, H., 422 Yamamoto, K., 70 Yamamoto, S., 453 Yamamoto, S.H.T., 411 Yamamoto, T., 422 Yamamoto, Y., 254, 273,348, 356, 374-375, 402, 406, 480, 481(29), 506, 533 Yamamura, T., 514 Yamaoka, M., 320-321, 321(4, 5), 322, 322(12), 323(14), 325(14), 326(14), 356, 360 Yamashita, K., 123 Yamauchi, R., 311,313,313(6), 314, 314(6), 406 Yamazaki, S., 478 Yanagisawa, E., 415, 416(31) Yang, C. S., 278 Yaniseh-Perron, C., 40, 42(26) Yanishlieva-Maslarova, N., 320
Yasuda, H., 544-545, 547(17) Yasuda, K., 478 Yates, M. T., 505, 505(7, 8), 506, 509(8), 512(7, 8) Yatvin, M. B., 66, 78(5), 79(5, 6) Yavin, E., 604, 619(23, 24) Yee, B. G., 314 Yeo, H. C., 24, 79 Yin, Y., 189 Yin Foo, D.D.Y., 187, 191(19), 203(19) Yl~i-Herttuala, S., 514 Yodoi, J., 150 Yodoi, Y., 153 Yohannes, P., 102 Yokode, H., 505, 506(6) Yokota, S., 604 Yokoyama, C., 453 Yokoyama, S., 514 Yonei, S., 220 Yonkers, P. A., 548, 552-553 Yoshida, H., 505,506(6) Yoshida, T., 225 Yoshida, Y., 402, 480, 481(29) Yoshiji, H., 24 Yoshikawa, Y., 374 Yoshimoto, T., 453 Yoshimoto, Y. T., 411 Yoshino, H., 52 Yoshino, S., 569 Young, A. B., 545,547(18) Youngman, R. J., 88, 420 Yuki, H., 4 Yumibe, N. P., 577 Yuzbasiyan-Gurkin, V., 545, 547(18) Yuzuriha, T., 274 Z Zabel, U., 160, 161(15) Zabin, I., 167 Zaklika, K. A., 605 Zannoni, C., 601 Zarins, Z. M., 415, 416(25) Zaslavsky, Y. A., 345 Zbinden, I., 191, 192(22) Zblewski, R., 99 Zechmeister, L., 389-390, 390(12), 392(12), 393 Zee, Y. C., 257
AUTHOR INDEX Zelnik, V., 573 Zeng, L., 416 Zhang, L.-X., 235-236, 239(8), 241(9), 243(9) Zhang, R. Y., 416 Zhang, T. M., 421 Zhang, T. Y., 415,416(26) Zhdanov, R. I., 582 Zhee, Y. C., 254 Ziegler, D. M., 136, 482 Ziegler, R., 601
679
Ziegler-Heitbrock, L., 155 Zilli, C., 446 Zimmer, L. T., 542 Zitomer, R. S., 176 Zivkovic, Z., 580 Zoeller, R. A., 604, 605(27), 606(21, 27), 608(27), 609(16, 27), 610(27), 611(27), 616(27), 618(21, 27) Zucker, M. B., 328 Zucker, P. A., 477 Zwelling, L. A., 631
680
SUBJECT INDEX
Subject Index
A Absorption spectra antioxidant assay, 282, 284-290 /3-carotene cis isomers, 390-392 Ginkgo biloba extract hydroxyl radical effects, 466-467 superoxide radical effects, 464 Acetaminophen toxicity, N-acetylcysteine therapy, 490 N-Acetyl-5-aminosalicylic acid, antioxidant effects on hemoglobin-catalyzed lipid peroxidation, 569-570 on hydroxyl radical formation, 562-565 on myeloperoxidase, 571 on peroxyl radical-mediated lipid peroxidation, 566-567 N-Acetylcysteine anticarcinogenic properties, 489-490 antimutagenic properties, 501-502 antioxidant capacity, 291-293 antioxidant properties, 488-489 assay in biological systems, 483-487 derivatizing reagents, 484 metabolism, 488 pharmacokinetics in human, 487 substitution for cysteine, 144-145 therapeutic applications, 482-483,490492 N-Acetylglutathione diethyl ester, preparation, 503-504 N-Acetylglutathione monoethyl(glycyl) ester, preparation, 503 Activator protein-1 DNA binding activity, regulation, 163174 oligonucleotide probe, preparation, 170171 N-Acylglutathione derivatives, biological applications, 505 preparation, 503
Adipose tissue, a-tocopherol extraction, 305-306 Adult respiratory distress syndrome, Nacetylcysteine therapy, 491 Albumin assay, 291-293 bovine serum, precipitation of tannin, 436 Aldehydes, fatty, see Fatty aldehydes Alkaline elution assay damage profiles of DNA from mammalian cells, 128-129 DNA strand breaks, 88-91, 94-95 Alkaline phosphatase bacterial, post-Fenton reaction digestion of DNA, 54 hydrolysis of DNA, 6-7, 21 8a-Alkyldioxytocopherones, formation by oxidation of t~- tocopherol by azo initiators, 313-314 by peroxyl radicals, 311-312 Amino acids, crosslinked to DNA bases, GC-MS, 9-11 Amino acid sequencing, N-terminal, oxidant stress-induced proteins, 186-187 2-Amino-3-mercapto-3-methylbutanoic acid, see Penicillamine Aminosalicylates, antioxidant properties, 555-572 4-Aminosalicylic acid, effect on myeloperoxidase, 571 5-Aminosalicylic acid, antioxidant effects, 557 on hemoglobin-catalyzed lipid peroxidation, 569-570 hydrogen peroxide decomposition, 559560 on hydroxyl radical formation, 562-565 hypochlorous acid scavenging, 571-572 on myeloperoxidase, 571 on peroxyl radical-mediated lipid peroxidation, 566-567 superoxide radical scavenging, 558-559
SUBJECT INDEX
21-Aminosteroids, antioxidant action, 548555 Ampholytes, 183 Anthocyanidin, in assay of condensed tannins, 433-434 Antibodies, incubation, 188 Antioxidants assay, general principles, 402-403 lipid-soluble, assay, 274-279 plasma assay, 269-279 loss after ozone exposure, 254-255 total, in plasma and body fluids, assay, 279-293 analytical imprecision, 290-291 automated assay, 289 clinical applications, 293 manual assay, 290 treatment of cells, 156-157 water-soluble, assay, 270-273 Apurinic/Apyrimidinic lyase, 111-115 Arachidonic acid, in assay of eicosanoid metabolism, 450-452 ARDS, see Adult respiratory distress syndrome Ascorbate antioxidant capacity, 291-293 interaction with ~-tocopherol, 380 interaction with vitamin E phenoxyl radical, 319-320 reduction of 8a-substituted tocopherones to a- tocopherol, 315-316 Ascorbic acid assay, 270-272 serum, HPLC assay, 335-337 tissue, HPLC assay, 332-334 Ascorbyl radicals, in vivo assay, 338-343 Astroglial cells, lazaroid antioxidant effects, 552 Atherosclerosis, oxidized LDL role, 513514 Azide radicals, in generation of peroxyl radicals, 428 2,2'-Azinobis(3-ethylbenzothiazoline 6sulfonate) cation, in assay of antioxidants, 284-290 2,2'-Azobis(2-amidinopropane) dihydrochloride peroxyl radical generation, 282, 565566 VP-16 phenoxyl radical generation, 633
681
2,2'-Azobis(2-amidopropane) hydrochloride, peroxyl radical generation, 281 2,2'-Azobis(2,4-dimethylvaleronitrile) cis-parinaric acid fluorescence induced by antioxidant effects, 377-380 in assay of radical scavenging, 35736O time course and spectra, 376-377 peroxyl radical generation, 282 rate studies, 374, 375-376 c~-tocopherol oxidation, 313-314 VP-16 phenoxyl radical generation, 633
B Bacteria, see also Escherichia coil; Salmonella typhimurium
colonies lysis, 203 replicas, 202-203 DNA damage profiles, 126 Bacteriophage PM2, DNA damage profiles, 129 Benzaldehydes, substituted, in assay of condensed tannins, 434-435 Bilirubin, assay, 273,291-293 Binding assay, ethidium-based, copperdependent DNA oxidation, 73-75 Biotinylation, mRNA, 206 Bisbenzamide, fluorescence enhancement by cell DNA, 96-97 Bis(3,5-di-tert-butyl-4-hydroxyphenylthio)propane, see Probucol Bisphenol, serum, assay, 512-513 Blood, see also Plasma; Serum /3-carotene and lycopene geometrical isomers, separation, 397-398 collection, 328-329 oxidized glutathione, HPLC assay, 367371 Brain gerbil, spin trap antioxidant activity, 523-526 rat, a-tocopherol extraction, 305-306 Bronchitis, N-acetylcysteine therapy, 490491 C Calcium channels, drugs blocking, antioxidant activity, 620-630
682
SUBJECT INDEX
Capillary elution, for RNA transfer to filter, 197 Carbon monoxide, production from DOPA, in assay of antioxidants, 284 Carcinogenicity, inhibition by N-acetylcysteine, 489-490 B-Carotene antioxidant activity in membranes, 371383 assay, 274-279 cis isomers, absorption spectra, 390-. 392 effect on fluorescence decay of cisparinaric acid, 377-380 geometrical isomers, separation, 388400 in blood, 397-398 in fruits and vegetables, 396-397 in model systems, 393-394 in tissue, 400 interaction with a-tocopherol, 380 isomerization, induction, 392 Carotenoids delivery to target cells, 237-238 geometrical isomers, 390 quenching of singlet oxygen, 384-388 assay, 386 rate constants, 386-388 regulation of gap junctional communication and connexin expression, 235244 Cell cultures aortic endothelial cells, 626 CHO-9 cells, 89-90 fibroblasts C3H/10T1/2, 237 V79, 589 HeLa cells, 154 Jurkat cells, 154 ozone exposure, systems for, 257-265 preparation for DNA isolation, 18 rat hepatocytes, 438-439 Cell extracts, nuclear and whole, preparation, 158-160 Chemiluminescent assay antioxidants, 281-282, 284 connexin 43,243 peroxyl radical scavenging by Ginkgo biloba extract, 467-468 Chloroform, extraction of RNA, 218, 227228
CHO-9 cells cell culture, 89-90 X irradiation, 90 CHO-K1 cells [14C]ethanolamine-labeled, plasmenylethanolamine breakdown, 610-612 [14C]hexadecanol-labeled fatty aldehyde formation, 614-616 radioactive formic acid formation, 612-614 [32p]Pi-labeled, plasmalogen breakdown in, 608-610 Cholesterol and tamoxifen and estrogens, comparative antioxidant effects, 590-602 unesterified, in LDL, enzymatic oxidation, 517-518 effect of c~-phenyl N-tert-butylnitrone, 520-521 Cholesterol coefficients, 597-598 Cholesteryl esters, in LDL enzymatic oxidation, 517-518 effect of a-phenyl N-tert-butylnitrone, 520-521 fatty acid composition, 518 Chromanols, incorporation in membranes, 363 Chromanoxyl radicals electron spin resonance, 364-365 generation, 361-366 reduction, 365-366 Chromatography, see also specific techniques neutral lipid separation from phospholipids, 617 OxyR protein, 220-221 singlet oxygen DNA damage products, 79-88 yeast redoxyendonuclease on DEAE-cellulose, 105-106 on Mono S, 106-108 on phosphocellulose, 106 on Superose 6-Superose 12, 108-110 Cloning complementation technique, 213-214 in differential display techniques, 210211 in subtractive hybridization, 207-209 Colitis, ulcerative, 555-558 Colorimetric assay, lipid peroxidation in sarcolemmal membranes, 622-626
683
SUBJECT INDEX Complementation, cloning of genes identified by mutagenesis, 213-214 Connexin 43 analysis and quantitation, 244 detection, 242-243 electrophoresis, 242 expression, regulation by carotenoids, 235-244 membrane, isolation, 241-242 Western blot analysis, 241-244 Coomassie blue, in 2D electrophoresis, 185 Copper catalysis of ghost membrane oxidation, 544-545 Cu(I) binding to DNA, 69 preparation from Cu(lI), 69 stabilization of DNA, 68-69 Cu(II), destabilization of DNA, 68 detoxification in penicillamine therapy of Wilson's disease, 546-547 - D N A adducts hydroxyl radical formation dependent on, ESR detection, 70-73 overview, 66-67 physical properties, 68-69 DNA oxidation dependent on, assay, 73-75 endogenous, catalysis of DNA oxidation in intact cells, 75-78 hydroxyl radical formation dependent on, detection, 69-70 penicillamine in presence of, prooxidant effect, 543-545 role in DNA oxidation, probes of, 7778 COS-7 cells, introduction of IO2-damaged shuttle vectors, 119-120 Coumarins effects on cylooxygenase and lipoxygenase, 443-454 structures, 445 Culture media, bacterial, 8-oxoguanine assay, 30-31 Cyclooxygenase, eicosanoid generation, effects of flavonoids and coumarins, 443-454 Cysteine assay, lactate and pyrnvate effects, 147149 deficiency, effect on T cells, 142-144
extracellular, effect on intracellular glutathione levels, 140-145 intracellular, assay, 135-137 membrane transport systems, 137-139 requirement during T cell activation, 141-142 requirement of T lineage cells, lactate and pyruvate effects, 147-149 Cystine intracellular, assay, 135-137 membrane transport systems, 137-139 substitution for cysteine, 144-145 Cytochrome c, acylated derivative conjugated to poly(styrene-co-maleic acid) properties, 342-343 synthesis, 339-342 in in oioo assay of superoxide and vitamin C radicals, 338-343
D Decolorization assay, antioxidants, 284285 Dehydroascorbate serum, HPLC, 335-337 tissue, HPLC, 332-334 Denaturation DNA fluorometric analysis, 91-93, 95-99 kinetics, 97-98 RNA, 196 Deoxycytidine, HPLC, 22 Deoxycytidine monophosphate, products after Fe2+/H202 exposure, radiochromatogram, 56 Deoxyguanosine HPLC, 22 isotope effects, 58 products after Fe2+/H2Oz exposure, radiochromatogram, 55 2'-Deoxyguanosine, photosensitized reactions, 82-83 Deoxyribonuclease I footprinting, in mapping of OxyR binding, 222-223 hydrolysis of DNA, 6-7 post-Fenton reaction digestion of DNA, 53-54 Derivatization DNA, 7
684
SUBJECT INDEX
fatty aldehydes and plasmalogens with 2,4-dinitrophenylhydrazine, 616618 a-tocopherol to a-tocopherol methyl ether, 306-307 Dialysis, Fos and Jun renaturation during, 167 1,2-Dichloroethane, radical cations, scavenging by ebselen, 479 Diethylenetriaminepentaacetic acid, complex with Fe 2. hydroxyl radical generation, 563 preparation, 564 Differential display techniques, DNA, 200201,207-211 4,8-Dihydro-4-hydroxy-8-oxo-2'-deoxyguanosine, 4R* and 4S* diastereomers formation by 2'-deoxyguanosine photooxidation, 83 HPLC, 81, 83-85 mass spectrometry, 85-88 preparation, 81 and related base component, chromatography and mass spectrometry, 8688 Dihydrolipoic acid antioxidant properties, 458-460 detection, 458 generation from a-lipoic acid, 456-458 by lipoamide dehydrogenase, 456-457 by sodium borohydride, 457-458 homologs, structure, 455 peroxyl radical scavenging, fluorescence assay, 460 reaction with superoxide radicals, 458459 structure-antioxidant activity relationships, 454-461 7,8-Dihydro-8-oxo-2'-deoxyguanosine, 17, 80, 83 in biological fluids, analysis, 24-31 radiolabeled standards, synthesis, 25 sample preparation and isolation, 2631 in DNA, steady-state levels, 18-24 excretion by cells, 29-30 in plasma, recovery, 30 production in tissue culture, 28-29 in urine, recovery, 26-27
3-(3,4-Dihydroxy-5-nitrobenzylidene)-2,4pentanedione, s e e Nitecapone 3,4-Dihydroxyphenylalanine, CO production, 284 Diltiazem effect on loss of endothelial cell glutathione and viability, 627-629 inhibition of lipid peroxidation, 622-626 5,5-Dimethyl-l-pyrroline N-oxide, in detection of hydroxyl radical formation, 70, 71 2,4-Dinitrophenylhydrazine, fatty aldehydes derivatized with, HPLC, 616618 Dioleoylphosphatidylcholine, effect on cisparinaric acid fluorescence, 376 Diphenoquinone, serum, assay, 512-513 1,6-Diphenyl-1,3,5-hexatriene, fluorescence polarization, vitamin E-induced changes, 324-326 3,3'-Di-propylthiocarbocyanine, fluorescence, vitamin E-induced quenching, 324-326 Disodium 3,3'-(1,4-naphthylidene) dipropionate, singlet oxygen generation, 116 Dissolution, ~-carotene and lycopene, isomerization induced by, 392 1,4-Dithioerythritol, effect on serum ascorbic acid levels, 337 DNA acidic hydrolysis, 5-6 bacterial, damage profiles, 126 bacteriophage PM2, damage profiles, 129 bases crosslinked to amino acids, GC-MS, 9-11 free, GC-MS, 8 cell, enhancement of bisbenzarnide fluorescence, 96-97 cell-free, damage profiles, 125 cleavage products analysis, 38-39 redoxyendonuclease-generated, posttreatment, 43-44 complementary, labeling, 202 -copper adducts hydroxyl radical formation dependent on, ESR detection, 70-73
SUBJECT INDEX
overview, 66-67 physical properties, 68-69 damage profiles, GC/MS/SIM acquisition, 131 derivatization, 7 differential display, 200-201,207-211 differential hybridization, 200-204 7,8-dihydro-8-oxo-2'-deoxyguanosine in, steady-state levels, 18-24 enzymatic hydrolysis, 6-7, 20-21 ethanol precipitation, 20 Fenton reaction-mediated damage, detection, 51-58 Fos binding, dependence on Cys redox state, 163-174 isolation, 18-21 Jun binding, dependence on Cys redox state, 163-174 L1210 leukemia cell, damage profiles, 129-130 mammalian cell, damage profiles, 128129 mitochondrial, damage profiles, 125-126 mobility shift, 221 nuclear, oxidative damage biomarkers in, HPLC assays, 16-33 oxidation copper-dependent, assay, 73-75 in intact cells, catalysis by endogenous copper, 75-78 role of copper and iron, probes of, 7778 oxidative damage biomarkers, HPLC assays, 16-33 endonuclease fingerprinting, 122-131 eukaryotic enzymes recognizing, detection and characterization, 33-44 GC-MS analysis, 3-15 by reactive oxygen species, assay, 4850 by singlet oxygen assay, shuttle vector for, 115-122 electrophoretic assay, 117-118 product analysis, 79-88 OxyR binding DNase I footprinting assay, 222-223 gel retardation assay, 221 phenol extraction, 19-20 plasmid, s e e Plasmids
685
post-Fenton reaction digestion to nucleosides, 53-54 preparation for HPLC, 18-21 strand breaks alkaline elution assay, 90-91 dose-effect curves, 93 fluorescence assay, 91-93, 95-99 induction, 90, 99-100 rejoining, 90 rejoining curves, 93-94 relaxation assay, 127-128 repair, 100-102 strand cleavage, mode of, determination, 42-44 substrates incubation with eukaryotic enzymes, 37 osmium tetroxide-damaged, preparation, 40, 42 ultraviolet-damaged, preparation, 37 subtractive hybridization, 200, 204-207 supercoiled, endonuclease-sensitive modifications, relaxation assay, 127-128 thymine glycol-containing, redoxyendonuclease digestion, 40-41, 42-43 ultraviolet radiation effects, 36-37 unwinding fluorometric analysis, 91-93, 95-99 kinetics, 97-98 DNA N-glycosylase, 34 DNase I, s e e Deoxyribonuclease I DOPA, s e e 3,4-Dihydroxyphenylalanine Dot blotting, heine oxygenase 1 transcription assay, 225-226 DTPA, s e e Diethylenetriaminepentaacetic acid Dye microinjection assay, gap junctional communication, 238-239
E Ebselen biological effects, 481 disposition, 481-482 glutathione peroxidase-like activity detection, 479 reaction scheme for, 477-479 metabolism, 481-482
686
SUBJECT INDEX
protective effect against lipid peroxidation, 480-481 radical scavenging, 479-480 singlet oxygen quenching, 480 synthesis, 477 Eicosanoids, generation by leukocytes inhibition by coumarins, 445 by flavonoids, 444 radiochromatographic assay, 450-452 radioimmunoassay, 449-450 stimulation, 449 Electroblotting, oxidant stress-induced proteins, 188 Electron spin resonance chromanoxyl radicals, 364-365 copper-dependent hydroxyl radical formation, 69-70 copper-DNA adduct-dependent hydroxyl radical formation, 70-73 nitroxides, 583-584 vitamin E phenoxyl radicals, 316-320, 351-353 VP-16 phenoxyl radicals detection, 632-637 reduction reactions in aqueous solution, 638-640 in cell and nuclear homogenates, 641-642 Electrophoretic mobility shift assay, 153 N F - K B activity, 160-162 Elution alkaline, see Alkaline elution assay capillary, for RNA transfer to filter, 197 End-labeling analysis, plasmid DNA strand breaks, 50 Endonucleases fingerprinting of oxidative DNA damage, 122-131 preparations, 124-125 supercoiled DNA modifications sensitive to, assay, 127-128 Endothelial cells aortic culture, 626 loss of glutathione and viability, effects of calcium channel blockers, 627-629 oxidative incubation, 626-627
brain microvessel, membrane physicochemical properties, effects of lazaroids, 553 Epoxy-8a-hydroperoxytocopherones, formation, 312 Epoxytocopherones, preparation, 313-314 Epoxy-a-tocopherylquinones, formation, 312 EPR, see Electron spin resonance Erythrocytes antioxidant effects of lazaroids, 552 ghosts chromanoxyl radical generation, 363 membranes, copper ion-catalyzed oxidation, 544-545, 547 glutathione diethyl ester transport into, 500-501 isolation, 328-329 saponification, 329 tocopherols and tocopherolquinone, assay, 327-331 Escherichia coil
culture media, 8-oxoguanine assay, 3031 expression and purification of recombinant Fos and Jun, 166-169 oxyR-cont r ol l e d regulon, 217 OxyR purification, 220-221 in screening of mutation-containing plasmids, 120-121 shuttle vector rescue, 120 ESR, see Electron spin resonance Estradiol-17/3, membrane antioxidant activity comparison with cholesterol, 596-598 in liposomal and microsomal systems, 594-596 Estrogen receptors, binding affinity of tamoxifen and tamoxifen metabolites, 600 Estrogens, membrane antioxidant activities, 590-602 Ethanol, precipitation of DNA, 20 Ethanolamine, ~4C-labeled, labeling of plasmenylethanolamine, 610-611 Ethidium bromide, in assay of copperdependent DNA oxidation, 73-75 N-Ethylmaleimide, quenching of reduced glutathione, 368-369
SUBJECT INDEX Etoposide, phenoxyl radicals ESR detection, 632-637 generation by azo initiators of peroxyl radicals, 633 by peroxidase, 632-633 photosensitized generation, 633-635 by tyrosinase, 632-633 interactions with reductants, ESR and HPLC studies in aqueous solution, 638-640 in cell and nuclear homogenates, 640642 Exonuclease, hydrolysis of DNA, 6-7 Extraction DNA with phenol, 19-20 flavonoids, 413-415 primary transcripts, 199-200 RNA, 193-196 with acid phenol and chloroform, 218 with guanidinium thiocyanate, t93-194 guanidinium thiocyanate-phenolchloroform method, 227-228 poly(A) + mRNA preparation, 194-196
F Fast protein liquid chromatography, yeast redoxyendonuclease on Mono S, 106-108 on Superose 6-Superose 12, 108-110 Fatty aldehydes 2,4-dinitrophenylhydrazine-derivatized, HPLC, 616-618 formation in [U-~4C]hexadecanol-labeled cells, 614-616 Fenton reaction DNA damage mediated by, 51-58 hydroxyl radical generation, 422 Ferric nitrilotriacetate, 438,440 Ferricytochrome c, reduction by nitroxides, 584-585 Ferrylmyoglobin radicals, generation of 2,2'-azinobis(3-ethylbenzothiazoline 6-sulfonate) radical cation, 284-285 Fibroblasts Chinese hamster lung, V79 culture, 589
687
intracellular nitroxide localization, 588-589 human dermal 7,8-dihydro-8-oxo-2'-deoxyguanosine, 8-oxoguanine, and 8oxoguanosine levels in vitro, 28-29 glutathione diethyl ester transport into, 500-501 hydrogen peroxide treatment, 227 UVA irradiation, 226-227 murine, C3H/10T1/2, culture, 237 Flavonoids antioxidant mechanisms, 421 chemistry, 412-413 effects on cylooxygenase and lipoxygenase, 443-454 extraction, 413-415 55Fe mobilization from hepatocytes, 443 functions in mammalian cells, 411 from Ginkgo biloba leaves, HPLC, 417418 HPLC, 413-420 oligomers, assay, 429-437 reactions with oxygen radicals, 420-429 structures, 412-413,445 tannin-related, assay, 429-437 Fluorescence bisbenzamide, enhancement by cell DNA, 96-97 cis-parinaric acid in assay of antioxidants, 283 azo initiator-induced decay antioxidant effects, 377-380 in assay of radical scavenging, 358360 time course and spectra, 376-377 effect of dioleoylphosphatidylcholine, 376 in liposomes, measurement, 374-375 quenching and polarization measurement, 323-324 vitamin E partitioning effects, 325-326 Fluorescence assay antioxidants, 282-284 peroxyl radical scavenging by dihydrolipoic acid, 460 by nitecapone and OR-1246
688
SUBJECT INDEX
in membranes, 533-535 in solution, 531-533 Fluorometric analysis, DNA unwinding, 91-93, 95-99 Footprinting, DNase I, in mapping of OxyR binding, 222-223 Formate dehydrogenase, in assay of formic acid formation from plasmalogens, 613-614 Formic acid, radioactive, formation in [l-J4C]hexadecanol-labeled cells, 612614 N-Formylglutathione, preparation, 504 N-Formylglutathione monoethyl(glycyl) ester, preparation, 504 Freedox, see U74006F Free radical reductase, 316-317 G Gap junctions communication dye microinjection assay, 238-239 regulation by carotenoids, 235-244 plaques, immunofluorescent detection, 239-241 Gas chromatography, capillary c~-tocopherol methyl ether stereoisomers, 307-309 all-rac-a-tocopherol stereoisomers, 302310 Gas chromatography-mass spectrometry acquisition of DNA damage profiles, 131 chemical determination of oxidative DNA damage, 3-15 4,8-dihydro-4-hydroxy-8-oxo-2'-deoxyguanosine and related base component, 86-88 DNA base-amino acid crosslinks, 9-11 free bases, 8 instrumentation, 7-8 nucleosides, 8-9 reference compounds, 4-5 with selected ion monitoring, low annlyre concentration measurements, 11-15 c~-tocopherol stereoisomers, 310 Gel electrophoresis, see also Isoelectric focusing assay of DNA damage by 1Oz, 117-118 connexins, 242
heme oxygenase 1 mRNA, 225, 229-230 monitoring of RNA quality, 246-247 Northern blot, mRNA, 196-199 oxidant stress-induced proteins, 179-181 oxidative DNA damage assay, 48-50 two-dimensional, oxidant stress-induced proteins, 181-187 Gel retardation assay Fos and Jun redox state, 171-174 OxyR binding to DNA, 221 Genes complementation cloning, 213-214 expression during oxidative stress, assessment comprehensive survey strategies, 177 DNA techniques, 200-211 in eukaryotes and prokaryotes, 177179 genetic strategies, 211-216 overview, 175-177 protein techniques, 179-191 RNA techniques, 191-200 supF, 119
mutations, 120-121 Ginkgo biloba extract
antioxidant action, 462-475 composition, 463 effect on xanthine oxidase, 469 flavonoids from, HPLC, 413-418 hydroxyl radical scavenging, 465-467 nitric oxide scavenging, 469-473 biological consequences, 475 nitrite-based detection, 472-473 oxyhemoglobin-based analysis, 469471 peroxyl radical scavenging, 467-468 superoxide radical scavenging, 464-465 without terpenes hydroxyl radical scavenging, 465-467 superoxide radical scavenging, 464465 Glutamate, extracellular, effect on lymphocyte function, 139-140 Glutamine syntbetase, effect of a-phenyl N-tert-butylnitrone, 526 Glutathione assay, 291-293 delivery by glutathione diethyl ester, 499-500 by glutathione monoesters, 492-493
SUBJECT INDEX depletion, effect on T cells, 142-144 intracellular levels assay, 135-137 effect of extracellular cysteine, 140145 oxidized, in blood, HPLC assay, 367371 reduced, quenching, 368-369 substitution for cysteine, 145 Glutathione diester, conversion to monoester in murine plasma, 500 Glutathione diethyl ester preparation, 501 transport into human cells, 500-501 Glutathione monoesters formation from diester in murine plasma, 500 free, preparation, 496 as glutathione delivery agents, 492-493 high-performance liquid chromatography, 497-498 metal impurities, 498-499 oxidation in vitro and in vivo, 499 preparation and use, 492-499 recrystallization, 496 thin-layer chromatography, 497 Glutathione monoethyl ester hemihydrosulfate, preparation, 495-496 Glutathione monoethyl ester hydrochloride, preparation, 494-495 Glutathione peroxidase, mimicry by ebselen, 476-482 N-Glycosylase, 33-34 activity of yeast redoxyendonuclease, 110-111 Gradient gels, in 2D electrophoresis, 184185 Guanidinium thiocyanate, extraction of RNA, 193-194, 227-228 H Heine oxygenase 1 agents inducing, 224-225 mRNA, accumulation assay selection, 225-226 Northern analysis, 229-235 oxidative stress-induced increase, 233-235 transient enhancement, 224-235
689
Hemoglobin, lipid peroxidation catalyzed by, effects of aminosalicylates, 568570 Hepatocytes, see also Liver iron-loaded, 55Fe mobilization, 443 rat, isolation and culture, 438-439 Hexadecanoi, 14C-labeled, labeling of plasmalogens, 612-614 Hexane, a-tocopherol and a-tocotrienol radical scavenging in, 357-359 High-performance liquid chromatography N-acetylcysteine, 485-487 bilirubin, 273 /3-carotene geometrical isomers, 393-400 in blood, 397-398 in fruits and vegetables, 396-397 in tissue, 400 dehydroascorbate from serum, 335-337 from tissues, 332-334 4,8-dihydro-4-hydroxy-8-oxo-2'-deoxyguanosine 4R* and 4S* diastereomers, 81, 83-85 DNA hydrolysates, 21-24 fatty aldehyde moieties after 2,4-dinitrophenylhydrazine derivatization, 616-618 flavonoids, 413-420, 416 glutathione monoesters, 497-498 8-hydroxyguanine nucleosides, 63-65 immunoaffinity-purified biological fluids, 31 lipid hydroperoxides, 375 lipid-soluble antioxidants, 276-279 lycopene geometrical isomers, 393-400 in blood, 397-398 in fruits and vegetables, 396-397 in tissue, 400 malondialdehyde, 439, 441 nucleosides produced by post-Fenton reaction digestion, 54-58 oxidative DNA damage biomarkers, 1633 oxidized glutathione in blood, 367-371 probucol and metabolites in serum, 511513 tocopherol, 329-330 ct-tocopherol methyl ether stereoisomers, 307 tocopherolquinone, 329-330
690
SUBJECT INDEX
all-rac-c~-tocopherol stereoisomers, 302310 uric acid from serum, 335-337 from tissues, 332-334 vitamin C from serum, 335-337 from tissues, 332-334 vitamin E homologs, 294-302 fluorescence detection, 299 reproducibility, 300 reversed-phase experiments, 298 straight-phase experiments, 298 VP-16 phenoxyl radical reduction in aqueous solution, 638-640 in cell and nuclear homogenates, 642 water-soluble antioxidants, 271-272 Horseradish peroxidase, radical generation, 284 Human immunodeficiency virus, type 1 long terminal repeat, in assay of NF-KB, 162 replication, inhibition by N-acetylcysteine, 492 Hybridization differential, 200-204 in Northern analysis of heme oxygenase 1 mRNA, 230-232 for Northern blot electrophoresis, 197199 in one-day Northern blotting, 248-249 subtractive, 200, 204-207 in transcriptional run-on assay, 200 Hydrogen peroxide decomposition by 5-aminosalicylic acid, 559-560 DNA damage induced by, detection, 5158 reaction with hepatic cytosol, associated copper release and membrane oxidation, 546-547 role in lymphocyte activation, 145-147 treatment of cells, 155-156 treatment of fibroblasts, 227 Hydrolysis acidic, DNA, 5-6 enzymatic, DNA, 6-7, 20-21 8a-Hydroperoxytocopherone, formation, 312
8-Hydroxy-2'-deoxyadenosine HPLC with elecrochemical detection, 63-65 synthesis by methylene blue and light, 62, 65 ultraviolet spectra, 62 8-Hydroxyguanosine HPLC with elecrochemical detection, 63-65 photochemical synthesis, 59-65 by methylene blue and light, 60-61, 65 Udenfriend system, 61-62, 65 ultraviolet spectra, 62 Hydroxyl radicals copper-dependent formation, detection, 69-70 copper-DNA adduct-dependent formation, ESR detection, 70-73 DNA-bound copper-catalyzed formation, 66-67 generation copper-binding sites as catalytic centers for, 544-545 by dihydrofumarate oxidation in presence of Fe-ADP, 621-622,625626 Fenton system, 422-423 by neutrophils, 560-561 site-specific methods, 561-562 inhibition by aminosalicylates, 562-565 interaction with N-acetylcysteine, 489 reaction with aliphatic structures, 426 scavenging by Ginkgo biloba extract, 465-467 by stobadine, 575-576 site-specific reactivity, 561 8a-Hydroxytocopherone, formation, 312 Hypochlorous acid interaction with N-acetylcysteine, 489 scavenging by 5-aminosalicylic acid, 571-572
Immunoaffinity columns, 26-28 Immunofluorescent microscopy, gap junctional plaques, 239-241 Immunoprecipitation, oxidant stressinduced proteins, 188-191
SUBJECT INDEX Iron chelated to DTPA hydroxyl radical generation, 563 preparation, 564 chelation, role in flavonoid antioxidant action, 437-443 DNA damage induced by, detection, 5158 55Fe, mobilization from hepatocytes, 443 interactions with nitroxides, 586-588 reaction with 5-aminosalicylic acid, 562563 role in DNA oxidation, probes of, 77-78 Ischemia, stobadine effects, 578-579 Isoelectric focusing, see also Gel electrophoresis gel used in, application to SDS gel, 184 oxidant stress-induced proteins, 181-184 Isomerization,/3-carotene and lycopene, induction, 392 Isotope effects, in HPLC of deoxyguanosine, 58
K Kinetics competition, in generation of oxygen radicals, 424-425 dihydrolipoic acid reaction with superoxide radicals, 458-459 DNA unwinding, 97-98 peroxyl radical scavenging by retinoids, 408-410 all-trans-retinol reaction with linoleic acid-derived peroxyl radicals, 404408
L Lactate, effect on cysteine requirement of T lineage cells and on cysteine assay, 147-149 Lazaroids, antioxidant action, 548-555 LDL, see Lipoproteins, low-density Leaf, Ginkgo biloba, flavonoids from extraction, 413-415 HPLC, 417-418
691
Leukocytes eicosanoid generation inhibition by coumarins, 445 by flavonoids, 444 radiochromatographic assay, 450452 radioimmunoassay, 449-450 stimulation, 449 mixed, preparation from rat peritoneal cavity, 447-448 polymorphonuclear, suspensions, preparation, 448-449 Light, induction of fl-carotene and lycopene isomerization, 392 Linoleic acid, peroxyl radicals derived from generation, 402-403 reaction with all-trans-retinol, 404-408 Linoleic acid methyl ester, oxidation, 402404 assays, 403-404 inhibition by all-trans-retinol, 404-405 by vitamin analogs, 408-410 Lipid hydroperoxides formed by cis-parinaric acid peroxidation, assay, 375 production during fluorescence decay of cis-parinaric acid, 376-377 Lipid peroxidation ebselen effects, 480-481 hemoglobin-catalyzed, aminosalicylate effects, 568-570 in hepatic microsomes, effects of ubiquinol and vitamin E, 345-348 iron-induced, analysis, 439-442 peroxyl radical-induced aminosalicylate effects, 565-566 nitecapone and OR-1246 effects, 535536 in sarcolemmal membranes, assay, 622626 Lipids neutral, separation from phospholipids, 617 peroxidation in hepatic microsomes, effects of ubiquinol and vitamin E, 345-348
692
SUBJECT INDEX
yeast membrane, antioxidant ability after tamoxifen treatment, 598-599 Lipoamide dehydrogenase, dihydrolipoic acid formation with, 456-457 a-Lipoic acid characteristics, 454-455 dihydrolipoic acid generation from, 456458 NMR spectra, 458 solution, preparation, 456 Lipoperoxyl radicals, scavenging by vitamin A and analogs, 401-410 Lipoproteins, low-density cholesterol molecular species, enzymatic oxidation, 517-518 effect of a-phenyl N-tert-butylnitrone, 520-521 chromanoxyl radical generation in, 363 isolation, 515 oxidized, role in atherosclerosis, 513514 probucol antioxidant activity in, assay, 508-509 triglyceride molecular species, enzymatic oxidation, 518-520 effect of a-phenyl N-tert-butylnitrone, 520-521 vitamin E phenoxyl radical in, 319-320 Liposomes 17/3-estradioi antioxidant effects, 594596 cis-parinaric acid-incorporated fluorescence, measurement, 374-375 model of antioxidant activities in membranes, 371-383 stobadine antioxidative effects, 577 tamoxifen antioxidant effects, 594-596 ct-tocopherol and a-tocotrienol radical scavenging in, 359-360 and water, partitioning of nitecapone and OR-1246, 538-539 Lipoxygenase eicosanoid generation, effects of flavonoids and coumarins, 443-454 oxygen radical generation, 423-424 and phospholipase A2, oxidative effects on LDL molecular species, 517521 Liquid chromatography /~-carotene geometrical isomers, 393
carotenoid isomers in fruits and vegetables, 396 lycopene geometrical isomers, 393 Liver, see also Hepatocytes cytosol, reaction with hydrogen peroxide, associated copper release and membrane oxidation, 546-547 homogenates, preparation, 297 tx-tocopherol extraction, 305-306 vitamin E homologs, HPLC, 294-302 Lucifer Yellow CH, 238-239 Luminogenic assay, nitroxides, 585 Lycopene assay, 274-279 geometrical isomers, separation, 388-400 in blood, 397-398 in fruits and vegetables, 396-397 in model systems, 393-394 in tissue, 400 isomerization, induction, 392 Lymphocytes, T, see T cells Lysophosphatidylethanolamine, formation in [2-14C]ethanolamine-labeled cells, 610-612
M Malondialdehyde, HPLC assay, 439, 441 Mass spectrometry, see also Gas chromatography-mass spectrometry 4,8-dihydro-4-hydroxy-8-oxo-2'-deoxyguanosine, 85-88 singlet oxygen DNA damage products, 79-88 Membranes /3-carotene antioxidant activity, 371-383 chromanol incorporation, 363 connexin 43, isolation, 241-242 cysteine and cystine transport systems, 137-139 dihydrolipoic acid scavenging of peroxyl radicals, 460 erythrocyte ghost, copper ion-catalyzed oxidation, 544-545 estrogen antioxidant effects, 590-602 comparison with cholesterol, 596-598 in liposomal and microsomal systems, 594-596 fluidity, effects of tamoxifen and estrogens, 600-602
SUBJECT INDEX lazaroid antioxidant effects, 550-552 peroxyl radical scavenging by nitecapone and OR-1246, 533-535 physicochemical properties, effects of lazaroids, 553-554 sarcolemmal lipid peroxidation, assay, 622-626 preparation, 621 tamoxifen antioxidant effects, 590-602 comparison with cholesterol, 596-598 in liposomal and microsomal systems, 594-596 ~-tocopherol antioxidant activity, 360361, 371-383 c~-tocotrienol antioxidant activity, 360361 tocotrienol restriction within, 320-327 ubiquinol antioxidant activity, 371-383 vitamin E distribution to, associated fluorescence properties, 325-326 and water, partitioning of nitecapone and OR-1246, 538-539 yeast, lipid fraction, antioxidant ability after tamoxifen treatment, 598-599 2-Mercaptoethanol, substitution for cysteine, 145 Metallothionein, role in penicillamine therapy of Wilson's disease, 546-547 Metals impurities in glutathione monoester preparations, 498-499 interactions with nitroxides, 585-588 Methylene blue in 8-hydroxy-2'-deoxyadenosine synthesis, 62, 65 in 8-hydroxyguanosine synthesis, 60-61, 65 Methyl linoleate, s e e Linoleic acid methyl ester Metmyoglobin in assay of antioxidants, 285-290 purification, 287-288 Microsomes chromanoxyl radical generation, 363 17/3-estradiol antioxidant effects, 594596 hepatic, lipid peroxidation, effects of ubiquinol and vitamin E, 345-348 stobadine antioxidative effects, 577-578 tamoxifen antioxidant effects, 594-596
693
yeast, lipid fraction, antioxidant ability after tamoxifen treatment, 598-599 Mitochondria chromanoxyl radical generation, 363 DNA, damage profiles, 125-126 stobadine antioxidative effects, 578 Monobromobimane, 485-487 Monocytes, treatment with hydrogen peroxide, 155 Mononuclear cells, glutathione diethyl ester transport into, 500-501 Mutagenicity, inhibition by N-acetylcysteine, 489-490 Mutations oxidative damage resistance detection in higher eukaryotes, 215216 hypersensitivity-inducing, 211-214 resistant phenotype-inducing, 214-215 pleiotropic assessment, 213 causes, 215 s u p F gene analysis, 121 screening, 120-121 Myeloperoxidase, effects of aminosalicylates, 571-572 Myoglobin, in assay of antioxidants, 285289
N NADPH oxidase, oxygen radical generation, 423 3,3'-(1,4-Naphthylidene) dipropionate endoperoxide, 385-386 NDGA, s e e Nordihydroguaiaretic acid Neocuproine, probe of copper and iron roles in DNA oxidation, 77-78 Neurons, antioxidant effects of lazaroids, 552 Neutrophils human, suspensions, preparation, 448449 hydroxyl radical generation, 560-561 Nicardipine effect on loss of endothelial cell glutathione and viability, 627-629 inhibition of lipid peroxidation, 622-626
694
SUBJECT I N D E X
pharmacologically active and inactive enantiomers, comparison, 629-630 Nifedipine effect on loss of endothelial cell glutathione and viability, 627-629 inhibition of lipid peroxidation, 622-626 Nitecapone antioxidant activity, effect of structural modification, 526-541 effect on peroxyl radical-induced lipid peroxidation, 535-536 effect on xanthine oxidase, 536-538 gastroprotective effects, 527-528 partition between aqueous and membrane phases, 538-539 peroxyl radical scavenging in membranes, 533-535 in solution, 531-533 physicochemical properties, 528-529 superoxide radical scavenging, 530-531 Nitrate tolerance, N-acetylcysteine therapy, 491 Nitric oxide, scavenging by Ginkgo biloba extract, 469-473 biological consequences, 475 nitrite-based detection, 472-473 oxyhemoglobin-based analysis, 469-471 Nitrite, in assay of nitric oxide scavenging by Ginkgo biloba extract, 472-473 Nitroxides antioxidant properties, 580-589 electron paramagnetic resonance, 583584 interaction with metals, 585-588 intracellular localization, 588-589 luminogenic assay, 585 midpoint redox potentials, 581 reaction with semiquinone radicals, 588 reduction of ferricytochrome c, 584585 superoxide dismutase mimetic activity, 582-583 Nordihydroguaiaretic acid inhibition of Mn superoxide dismutase mRNA induction, 250-251 reduction of 8a-substituted tocopherones to c~-tocopherol, 315-316 Northern blotting heme oxygenase 1 mRNA, 225-226, 229-235
capillary blot procedure, 230 quantification, 232-235 mRNA, 196-199 one-day, for mRNA detection, 244-252 reverse, 203-204 Nuclear magnetic resonance dihydrolipoic acid, 458 a-lipoic acid, 458 Nuclease P1 hydrolysis of DNA, 20-21 post-Fenton reaction digestion of DNA, 54 Nuclei homogenates, VP-16 phenoxyl radicaltyrosinase interaction in, 640-642 isolation, 199 Nucleic acids, precipitation, 193 Nucleosides GC-MS, 8-9 mixtures, HPLC, 54-58 production by post-Fenton reaction digestion of DNA, 53-54 Nucleotides, 32P-labeled, labeling of overlapping oligonucleotides, 249-250
O Oligo(dT)-cellulose, preparation, 247 Oligonucleotides AP-I probe, preparation, 170-171 overlapping, labeling, 249-250 OR-1246 antioxidant activity, 526-541 effect on peroxyl radical-induced lipid peroxidation, 535-536 effect on xanthine oxidase, 536-538 partition between aqueous and membrane phases, 538-539 peroxyl radical scavenging in membranes, 533-535 in solution, 531-533 physicochemical properties, 528-529 superoxide radical scavenging, 530-531 Osmium tetroxide, DNA substrates damaged by, preparation, 40, 42 Ovarian cells, glutathione diethyl ester transport into, 500-501 Oxidation, see also Photooxidation cerebral proteins, effect of a-phenyl Ntert-butylnitrone, 526
SUBJECT INDEX DNA assay, 73-75 in intact cells, catalysis by endogenous copper, 75-78 role of copper and iron, probes of, 7778 ghost membranes, catalysis by copper ions, 544-545,547 glutathione monoesters in vitro and in vivo, 499 low-density lipoprotein molecular species, 517-520 effect of c~-phenyl N-tert-butylnitrone, 520-521 cis-parinaric acid, effects of ubiquinol and vitamin E, 345 phospholipids, inhibition by tocotrienols and tocopherols, 321 photosensitized cultured cells, 606-608 plasmalogens, 605-606 a-tocopherol by peroxyl radicals, 310313 Oxidative stress damage and repair pathways, 176, 178 eukaryotic cells, heme oxygenase 1 as marker, 224-235 gene expression during, assessment comprehensive survey strategies, 177 DNA techniques, 200-211 in eukaryotes and prokaryotes, 177179 genetic strategies, 211-216 overview, 175-177 protein techniques, 179-191 RNA techniques, 191-200 8-Oxo-2'-deoxyguanosine, HPLC assays, 16-33 8-Oxoguanine in bacterial media, assay, 30-31 in biological fluids, analysis, 24-31 radiolabeled standards, 25 sample preparation and isolation, 2631 excretion by cells, 29-30 HPLC assays, 16-33 production in tissue culture, 28-29 8-Oxoguanine endonuclease, detection, 38-39
695
8-Oxoguanosine in biological fluids, analysis, 24-31 radiolabeled standards, 25 sample preparation and isolation, 2631 excretion by cells, 29-30 production in tissue culture, 28-29 2-Oxo-4-thiazolidine carboxylate, substitution for cysteine, 145 Oxycarotenoids, quenching of singlet oxygen, 386-388 Oxygen activated, treatment of plasmids, 48 reactive species enzymes metabolizing, overexpression, 157-158 plasmid strand breaks induced by, localization, 45-51 singlet DNA damage products formed by, analysis, 79-88 generation, 79, 116, 385 induced DNA damage and mutagenicity, assay, shuttle vector for, 115122 quenching by carotenoids, 384-388 by ebselen, 480 by stobadine, 575 reactivity to plasmalogens, 603-620 treatment of cells, 156 uptake, in assay of antioxidants, 281 Oxygen radicals generation, 422-425 reactions with flavonoids, 420-429 Oxyhemoglobin, in assay of nitric oxide scavenging by Ginkgo biloba extract, 469-471 Ozone biomolecular damage, evaluation in vitro, 252-256 experimental design, 253-255 extrapolation to in vivo situation, 256 contact with respiratory tract lining fluids, 252-253 effects on plasma, 254-255 exposure of cultured cells and tissues, 257-265 large system, 261-262
696
SUBJECT I N D E X
small system, 258-260 uniformity, 262-265 monitoring, 262
P
cis-Parinaric acid fluorescence in assay of antioxidants, 283 azo initiator-induced decay antioxidant effects, 377-380 in assay of radical scavenging, 358360 time course and spectra, 376-377 effect of dioleoylphosphatidylcholine, 376 in liposomes, measurement, 374-375 oxidation, effects of ubiquinol and vitamin E, 345 peroxidation, lipid hydroperoxides formed from, assay, 375 PDTC, inhibition of NF-KB activation, 157 Penicillamine antioxidant effects, 542-547 prooxidant effect in presence of copper ions, 543-545 therapy of Wilson's disease, 542 copper detoxification role, 546-547 2,2,5,7,8-Pentamethylchroman-6-ol, 313 Peritoneal cavity, leukocyte preparation, 447-448 Peritonitis, induction, 447-448 Peroxidase, generation of VP-16 phenoxyl radicals, 632-633 Peroxidation, cis-parinaric acid, lipid hydroperoxides formed from, assay, 375 Peroxisomes, plasmalogen biosynthesis, 603-604 Peroxyl radicals generation, 423,427-428 in aqueous phase, 281-282 by 2,2'-azobis(2-amidinopropane) dihydrochloride, 565-566 by 2,2'-azobis(2,4-dimethylvaleronitrile), 374, 375-376 in lipid phase, 282 interactions with a-tocopherol and c~tocotrienol
in hexane, 357-359 in liposomes, 359-360 linoleic acid-derived, reaction with alltrans-retinol, 404-408 lipid peroxidation aminosalicylate effects, 565-566 nitecapone and OR-1246 effects, 535536 oxidation of a-tocopherol, 310-313 reactions with flavonoids, rate constants, 426-428 reactivity with ubiquinol and vitamin E, 344-345 scavenging by dihydrolipoic acid, fluorescence assay, 460 by ebselen, 479-480 by Ginkgo biloba extract, 467-468 by nitecapone and OR-1246 in membranes, 533-535 in solution, 531-533 by retinoids, kinetics, 408-410 pH, effect on HPLC of ascorbic and uric acids, 333 Phagemids, subtractive hybridization based on, 204-205 Pharmacokinetics, N-acetylcysteine in human, 487 Phase partitioning, nitecapone and OR1246, 538-539 l, 10-Phenanthrolene, probe of copper and iron roles in DNA oxidation, 77-78 Phenol DNA extraction, 19-20 RNA extraction, 218, 227-228 Phenolics general assays, 432-433 protein precipitable, assay, 435-436 Phenoxyl radicals, etoposide ESR detection, 632-637 generation by azo initiators, 633 by peroxidase, 632-633 photosensitized generation, 633-635 by tyrosinase, 632-633 interactions with reductants, ESR and HPLC studies in aqueous solution, 638-640 in cell and nuclear homogenates, 640642
SUBJECT INDEX 2-Phenyl- 1,2-benzisoselenazol-3(2H)-one, see Ebselen ~-Phenyl N - t e r t - b u t y l n i t r o n e antioxidant activity in brain, 523-526 antioxidant for low-density lipoproteins, 513-523 effect on oxidation of LDL molecular species, 520-521 preparation, 516 o-Phenylenediamine, radical generation, 282 Phosphodiesterase, post-Fenton reaction digestion of DNA, 54 Phospholipase A2, and lipoxygenase, oxidative effects on LDL molecular species, 517-521 Phospholipids oxidation, inhibition by tocotrienols and tocopherols, 321 separation from neutral lipids, 617 Photochemistry, 8-hydroxyguanine nucleoside synthesis, 59-65 Photoemission assay, singlet oxygen quenching by carotenoids, 386 Photolysis, oxygen radical generation, 422 Photooxidation, 2'-deoxyguanosine to 4,8dihydro- 4-hydroxy-8-oxo-2'-deoxyguanosine, 83 Phycoerythrin assay, antioxidants, 282283 Plasma, see also Blood; Serum 7,8-dihydro-8-oxo-2'-deoxyguanosine recovery, 30 glutathione diester conversion to monoester, 500 lipid-soluble antioxidants, 274-279 model for respiratory tract lining fluids, 253 ozone exposure, 254-255 ct-tocopherol extraction, 305-306 all-rac-c~-tocopherol stereoisomers, separation, 302-310 total antiooxidant status, 279-293 vitamin E homologs, HPLC, 294-302 water-soluble antioxidants, assay, 270273 Plasmalogens antioxidant function, 604-605 breakdown in [32p]Pi-labeled cells, 608610
697
decomposition by photosensitized oxidation, 605-606 derivatization with 2,4-dinitrophenylhydrazine, 616-618 fatty aldehyde formation in [U-~4C]hexadecanol-labeled cells, 614-616 formic acid formation in [l-t4C]hexade canol-labeled cells, 612-614 peroxisomal biosynthesis, 603-604 reactivity to singlet oxygen and radicals, 603-620 Plasmenylethanolamine, breakdown in [2~4C]ethanolamine- labeled cells, 610612 Plasmids, see also Shuttle vectors oxidative damage, assay, 48-50 pDS56, in expression of recombinant Fos and Jun, 166-169 redoxyendonuclease substrates, generation, 103-104 strand breaks, localization, 45-51 treatment with activated oxygen, 48 Platelets isolation, 328-329 saponification, 329 tocopherols and tocopherolquinone, assay, 327-331 Polydeoxyadenosine, products after Fe2÷/ H202 exposure, UV absorbance profile, 57 Polyenes, synthetic, quenching of singlet oxygen, 387-388 Polymerase chain reaction in differential display techniques, 209210 subtractive hybridization based on, 204205 Poly(styrene-co-maleic acid), conjugates of cytochrome c and superoxide dismutase, in in vivo assay of superoxide and vitamin C radicals, 338-343 Precipitation, see also Immunoprecipitation DNA with ethanol, 20 nucleic acids, 193 yeast redoxyendonuclease, 105-106 Primer extension assay, OxyR regulon, 219-220 Probucol antioxidant activity, assay, 506-510
698
SUBJECT INDEX
in low-density tipoproteins, 508-509 in whole serum, 509-510 metabolic pathway, 506 and metabolites, assay in serum, 511513 Proteins cerebral, oxidation, effect of a-phenyl N-tert-butylnitrone, 526 digestion, 18-19 Fos recombinant bacterial expression and purification, 166-169 properties, 164 redox-dependent DNA binding activity, 163-174 redox state, assay, 171-174 reduction by cellular proteins, 165-166 renaturation during dialysis, 167 Jun recombinant bacterial expression and purification, 166-169 properties, 164 redox-dependent DNA binding activity, 163-174 redox state, assay, 171-174 reduction by cellular proteins, 165-166 renaturation during dialysis, 167 oxidant stress-induced amino N-terminal microsequencing, 186-187 gel electrophoresis on polyacrylamide, 179-181 two-dimensional techniques, 181187 identification by pulse labeling, 185186 immunoprecipitation, 188-191 Western blotting, 187-188 OxyR activation of transcription, in vitro assay, 223 binding of DNA DNase I footprinting, 222-223 gel retardation assay, 221 purification, 220-221 thiol groups, assay, 273-274 Prussian blue assay, phenolics, 432-433
12-(1'-Pyrene)dodecanoic acid, in photosensitized oxidation of cultured cells, 606-608 Pyrimidine dimer endonuclease, yeast, detection, 38-39 Pyruvate, effect on cysteine requirement of T lineage cells and on cysteine assay, 147-149
R Radial arm maze test, gerbils, effect of aphenyl N-tert-butylnitrone, 525-526 Radial diffusion assay, tannin, 436-437 Radiochromatographic assay, eicosanoid metabolism, 450-452 Radiofluorography, 189-191 Radioimmunoassay, eicosanoid release from leukocytes, 449-450 Radiolabeling cDNA probes, 202 cell extracts for immunoprecipitation studies, 189 overlapping oligonucleotides, 249-250 phospholipids with [32p]Pi, 608 plasmalogens with [1-~4C]hexadecanol, 612-614 plasmenylethanolamine with [2J4C] ethanolamine, 610-611 pulse, oxidant stress-induced proteins, 185-186 Radiolysis, 422 pulse, oxygen radical generation, 424425 Recrystallization, glutathione monoesters, 496 Redox factor-l, reduction of Fos and Jun, 165-166 Redoxyendonucleases comparison, 40-42 digestion of thymine glycol-containing DNA, 40-41, 42-43 DNA cleavage products generated by, posttreatment, 43-44 substrates, generation, 103-104 yeast apurinic/apyrimidinic lysase activity, 111-115 assay, 103-105
SUBJECT INDEX detection, 38-39 N-glycosylase activity, 110-111 properties, 110 purification, 105-110 substrate specificity, 110-115 Relaxation assay DNA strand breaks, 127-128 endonuclease-sensitive modifications in supercoiled DNA, 127-128 Renaturation, Fos and Jun during dialysis, 167 Reperfusion, stobadine effects, 578-579 Respiratory tract, ozone exposure epithelial cells or explants, in vitro systems, 257-265 exposure uniformity, 262-265 large system, 261-262 small system, 258-260 lining fluids, 252-253 extrapolation of in vitro results to in vioo situation, 256 plasma as model, 253 Retinoids, peroxyl radical scavenging, 408-410 Retinol, see Vitamin A Ribonuclease, inactivation during RNA extraction, 193 RNA denaturation, 196 denatured, transfer to filter, 197 digestion, 18-19 extraction, 193-196 guanidinium thiocyanate-phenolchloroform method, 227-228 hybridization, 197-199, 200 isolation, 218 messenger biotinylation, 206 detection by one-day Northern blotting, 244-252 heme oxygenase 1 Northern analysis, 229-235 oxidative stress-induced increase, 233-235 transient enhancement, 224-235 Mn-superoxide dismutase, induction by TNF, 244-252 Northern blot electrophoresis, 196199
699
poly(A) ÷, preparation, 194-196 oxidant-modulated, basic study techniques, 191-193 poly(A) ÷, isolation, 247-248 primer extension assay, 219-220 quality, electrophoretic monitoring, 246247 total cellular, isolation, 227-228 cytoplasmic, isolation, 245-246 preparation, 193-194 transcriptional run-on assay, 199-200 Runoff transcription assay, heme oxygenase 1,225 S Saccharomyces cerevisiae
crude cell extracts, preparation, 105 proteins recognizing oxidative DNA damage, identification and characterization, 36-39 tamoxifen-treated, membrane lipid fraction antioxidant ability, 598-599 Salmonella typhimurium, oxyR-controlled
regulon, 217 Saponification, red cells and platelets, 329 Semiquinone radicals, reaction with nitroxides, 588 Serum, see also Blood; Plasma bisphenol, assay, 512-513 dehydroascorbate, assay, 335-337 diphenoquinone, assay, 512-513 preparation for HPLC of vitamin E homologs, 297 probucol antioxidant activity, 509-510 assay, 511-513 spiroquinone, assay, 512-513 uric acid, assay, 335-337 vitamin C, assay, 335-337 Shuttle vectors, for assay of ~O2-induced DNA damage and mutagenicity, 115122 rescue into Eseherichia coli, 120 transfection of mammalian cells, 119120 treatment with ~O2, 116-119 Silver staining, 185
700
SUBJECT INDEX
Sodium borohydride, dihydrolipoic acid synthesis with, 457-458 Sodium dodecyl sulfate gel, application of isoelectric focusing gel, 184 Spectrophotometric assay dihydrolipoic acid, 458 heme oxygenase 1 transcription, 225 hydrogen peroxide decomposition by 5aminosalicylic acid, 560 nitecapone and OR-1246 effects on peroxyl radical-induced lipid peroxidation, 535-536 scavenging of superoxide radical, 530531 nitric oxide scavenging by Ginkgo biloba extract, 469-473 ozone, 262-263 phenolics, 432-433 protein thiols, 273-274 superoxide scavenging by aminosalicylates, 558-559 xanthine oxidase, 469, 536-538 Spin traps, antioxidant activity in brain, 523-526 Spiroquinone, serum, assay, 512-513 Spleen exonuclease, hydrolysis of DNA, 6-7 Stobadine antioxidative effects, 572-580 in liposomes, 577 in microsomes, 577-578 in mitochondria, 578 assay, 574 chemical properties, 573 effects in ischemia and reperfusion, 578579 hydroxyl radical scavenging, 575-576 quenching of singlet oxygen, 575 radical formation, 574-575 superoxide radical scavenging, 576-577 synthesis, 572-573 Sulfapyridine, antioxidant effects on hemoglobin-catalyzed lipid peroxidation, 569-570 on hydroxyl radical formation, 562-565 on myeloperoxidase, 571 on peroxyl radical-mediated lipid peroxidation, 566-567 Sulfasalazine, antioxidant effects, 556557
on hemoglobin-catalyzed lipid peroxidation, 569-570 on hydroxyl radical formation, 562-565 on peroxyl radical-mediated lipid peroxidation, 566-567 Supercritical fluid chromatography, r-carotene geometrical isomers, 397 Superoxide dismutase conjugated to poly(styrene-co-maleic acid), in in vivo assay of superoxide and vitamin C radicals, 338-343 mimetic activity of nitroxides, 582-583 MnmRNA, induction by TNF, 244-252 overexpression, effect on NF-rB activation, 158 Superoxide radicals generation, 422 by dihydrofumarate oxidation in presence of Fe-ADP, 621-622, 625626 enzymatic, 424 reactions with N-acetylcysteine, 489 with dihydrolipoic acid, 458-459 with ct-tocopherol and c~-tocotrienol, 357 role in lymphocyte activation, 145-147 scavenging by aminosalicylates, 558-559 by Ginkgo biloba extract, 464-465 by nitecapone and OR-1246, 530-531 by stobadine, 576-577 treatment of cells, 156 in oioo assay, 338-343
T Tamoxifen membrane antioxidant activity, 590-602 comparison with cholesterol, 596-598 in liposomal and microsomal systems, 594-596 metabolites, 592 yeast treated with, membrane lipid fraction from, antioxidant ability, 598-599 Tannins, condensed assay, 429-437
SUBJECT INDEX biological sources, 431 extraction, 431 functional group assays, 433-435 protein precipitation-based assays, 435437 purification, 431-432 structure, 429-431 T cells activation cysteine requirement, 141-142, 147149 role of reactive oxygen intermediates, 145-147 function extracellular glutamate effects, 139140 intracellular cysteine effects, 140-145 glutathione diethyl ester transport into, 500-501 Molt-4 cysteine, cystine, and methionine levels, 137 glutathione levels, 136-137 subsets, effects of cysteine deficiency and glutathione depletion, 142-144 Tetrahydrofuran, in delivery of carotenoids to target cells, 237-238 Thin-layer chromatography fatty acids formed from plasmalogens, 615-616 glutathione monoesters, 497 two-dimensional plasmalogen breakdown products, 609-610 plasmenylethanolamine breakdown products, 611-612 Thiobarbituric acid-reactive substances assay aminosalicylate antioxidant properties, 561-563,566,569 antioxidants, 283 17fl-estradiol membrane antioxidant action, 594-596 lipid peroxidation nitecapone and OR-1246 effects, 535536 in sarcolemmal membranes, 622-626 probucol antioxidant activity in low-density lipoproteins, 508-509 in whole serum, 509-510
701
tamoxifen membrane antioxidant action, 594-596 Thiols, protein, assay, 273-274 THP-1 cells, cysteine, cystine, methionine, and glutathione levels, 137 Thymidine, HPLC, 22 Thymine glycol, DNA containing, redoxyendonuclease digestion, 40-41, 42-43 Tirilazad mesylate, s e e U74006F Tissues /3-carotene and lycopene geometrical isomers, separation, 400 dehydroascorbate, uric acid, and vitamin C, assay, 332-334 ozone exposure, systems for, 257-265 all-rac-a-tocopherol stereoisomers, separation, 302-310 TNF, s e e Tumor necrosis factor c~-Tocopherol, s e e Vitamin E 3,-Tocopherol, 274-279 a-Tocopherol methyl ether formation from ct-tocopherol, 306-307 stereoisomers capillary gas chromatography, 307-309 chiral phase HPLC, 307 Tocopherolquinone, platelet and red blood cell, assay, 327-331 Tocopherols antioxidative activity in heterogeneous system, 321-323 in homogeneous system, 320-321 HPLC, 294-302 inhibition of phospholipid oxidation, 321 platelet, assay, 327-331 red blood cell assay, 327-331 lability, 331 Tocopherones, 8a-substituted, reduction to a-tocopherol, 315-316 Tocopheroxyl radicals generation, 310 reactions, 310-311 a-Tocotrienol antioxidant properties in membranes, 360-361 overview, 354-357 interaction with superoxide radicals, 357 radical scavenging activity in hexane, 357-359 in liposomes, 359-360
702
SUBJECT INDEX
recycling efficiency, 361-366 Tocotrienols antioxidative activity in heterogeneous system, 320-327 in homogeneous system, 320-321 HPLC, 294-302 inhibition of phospholipid oxidation, 321 Transactivation assay, NF-KB activity, 162 Transcription, activation by OxyR, in vitro assay, 223 Transcriptional run-on assay, oxidantmodulated RNA, 199-200 Transcription factors NF-KB activation, effect of Mn-superoxide dismutase overexpression, 158 electrophoretic mobility shift assay, 160-162 regulation in vitro and in vivo, 152153 transactivation assays, 162 oxyR, 151-152 soxRS, 151-152 Transfection, mammalian cells with I02damaged shuttle vectors, 119-120 TRAP assay, antioxidants, 280-282 Triglycerides, LDL-associated, molecular species enzymatic oxidation, 518-520 fatty acid composition, 519-520 Tumor cells breast carcinoma, NF-KB activation, effect of Mn- superoxide dismutase overexpression, 158 HeLa culture conditions, 154 hydrogen peroxide treatment, 155 leukemia K-562, homogenates, VP-16 phenoxyl radical-tyrosinase interactions in, 640-642 L1210, DNA damage profiles, 129-130 U937 cysteine, cystine, and methionine levels, 137 glutathione levels, 136-137 pituitary, antioxidant effects of lazaroids, 553 T lymphoma Jurkat culture conditions, 154
hydrogen peroxide treatment, 155 L5178Y, cysteine, cystine, methionine, and glutathione levels, 137 Tumor necrosis factor, induction of Mnsuperoxide dismutase mRNA, 244-. 252 Tumor necrosis factor receptors, type 1, activation, Mn-superoxide dismutase mRNA induction via, 244-252 Tyrosinase generation of VP-16 phenoxyl radicals, 632-633 interactions with VP-16 phenoxyl radicals in cell and nuclear homogenates, 640-642
U U-74006F, inhibition of lipid peroxidation, 549-554 in membrane systems, 550-552 physicochemical effects on membranes, 553-554 in whole cells, 552-553 U-74500A, inhibition of lipid peroxidation, 549-554 in membrane systems, 550-552 physicochemical effects on membranes, 553 -554 in whole cells, 552-553 Ubiquinol antioxidant activity assay, 343-354 mechanisms, 344 in membranes, 371-383 assay, 274-279 effect on fluorescence decay of cisparinaric acid, 377-380 effect on lipid peroxidation in hepatic microsomes, 345-348 in vivo studies, 346-348 effect on cis-parinaric acid oxidation, 345 interaction with a-tocopherol, 380 prevention of vitamin E oxidation, 348351 reactivity toward peroxyl radicals, 344345 reduction of vitamin E phenoxyl radical, 351-354
SUBJECT INDEX Ubiquinones antioxidant function assay, 343-354 mechanisms, 344 assay, 274-279 Udenfriend system, 8-hydroxyguanosine synthesis, 61-62, 65 Ulcerative colitis, 555-558 Ultraviolet ozone analyzer, 262 Ultraviolet radiation DNA substrates damaged by, preparation, 37 effects on DNA, 36-37 generation of vitamin E phenoxyl radical, 318-319 UVA, human fibroblast exposure, 226227 Ultraviolet spectra, 8-hydroxyguanine nucleosides, 62 Urate, assay, 291-293 Uric acid assay, 270-272 serum, HPLC, 335-337 tissue, HPLC, 332-334 Urine 7,8-dihydro-8-oxo-2'-deoxyguanosine isolation, 26-27 HPLC with electrochemical detection, 31 8-oxoguanine isolation, 26-27 8-oxoguanosine isolation, 26-27
V Vanillin, in assay of condensed tannins, 434-435 Verapamil effect on loss of endothelial cell glutathione and viability, 627-629 inhibition of lipid peroxidation, 622-626 Vitamin A and analogs, lipoperoxyl radical scavenging in homogeneous solution, 401-410 reaction with linoleic acid-derived peroxyl radicals, 404-408 Vitamin C, s e e Ascorbic acid Vitamin E, s e e a l s o Tocopherols; Tocotrienols
703
antioxidant properties in membranes, 360-361,371-383 overview, 354-357 assay, 274-279, 291-293 derivatization to a-tocopherol methyl ether, 306-307 effect on fluorescence decay of cisparinaric acid, 377-380 effect on lipid peroxidation in hepatic microsomes, 345-348 in v i v o studies, 346-348 effect on cis-parinaric acid oxidation, 345 extraction from plasma and tissue, 305306 formation by reduction of 8a-substituted tocopherones, 315-316 homologs HPLC, 294-302 overview, 294-296 structures, 295 interaction with antioxidants, 380 interaction with superoxide radicals, 357 occurrence in nature, 294 oxidation by peroxyl radicals, 310-313 phenoxyl radicals ESR studies, 316-320 interaction with reductants, 319-320 reduction by ubiquinol, 351-354 UV-induced generation, 318-319 radical scavenging activity comparison to ubiquinol, 344-348 in hexane, 357-359 in liposomes, 359-360 reactivity toward peroxyl radicals, 344345 recycling efficiency, 361-366 regeneration, assessment, 316-320 stereoisomers, separation, 302-310 ubiquinol-dependent regeneration, 348354 VP-16, s e e Etoposide
W Water, and membrane phases, partitioning of nitecapone and OR-1246, 538-539 Western blotting connexin proteins, 241-244 oxidant stress-induced proteins, 187-188
704
SUBJECT INDEX
Wilson's disease, penicillamine therapy, 542-547 X Xanthine oxidase effect of Ginkgo biloba extract, 469 effects of nitecapone and OR-1246, 536538
oxygen radical generation, 423 Xanthophylls, quenching of singlet oxygen, 386-388 X irradiation, CHO-9 cells, 90
Y Yeast, see Saccharomyces cerevisiae