The Alkaloïds: chemistry and physiology, Volume 16

THE ALKALOIDS Chemistry and Physiology

VOLUME XVI

This Page Intentionally Left Blank

THE ALIMLOIDS Chemistry and Physiology Edited by

R. H. F. MANSKE Department of Chemistry, University of Waterloo Waterloo, Ontario, Canada

VOLUME XVI

1977

ACADEMIC PRESS

0

NEW YORK

0

SAN FRANCISCO

0

LONDON

A Subsidiary of Harcourt Brace Jovanovich, Publishers

COPYRIGHT 0 1977, BY ACADEMIC PRESS,INC. ALL RIGHTS RESERVED. NO PART O F THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING FROM THE PUBLISHER.

ACADEMIC PRESS, INC.

111 Fifth Avenue. New York, New York 10003

United Kingdom Edition published b y ACADEMIC PRESS, INC. (LONDON) LTD. 24/28 Oval Road, London NWl

Library of Congress Cataloging in Publication Data Manske, Richard Helrnuth Fred, The alkaloids. Vols. 8-16 edited by R. H. F. Manske. Includes bibliographical references. 1. Alkaloids. 2. Alkaloids-Physiological effect. I. Holrnes, Henry Lavergne, joint author. 11. Title: 1. Alkaloids. QV628 M288al Thru physiology. [DNLM: QD421.M3 547 '.I 2 50-5522 ISBN 0-12-469516-7

PRINTED IN THE UNITED STATES OF AMERICA

CONTENTS

LIST OF CONTRIBUTORS.. .................................................... PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . OF PREVIOUS VOLUMES.. .......................................... CONTENTS

ix xi xiii

Chapter 1. Plant Systematics and Alkaloids DAVIDS. SEIGLER

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Data to Be Utilized . . . . .......................................... UI. Application of the Data t iological Problems ........................ 1V. Alkaloids in Lower Vascular Plants and Gymnosperms . . . . . . . V. Alkaloids in the Angiosperms . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 3 8

73

Chapter 2. The Tropane Alkaloids ROBERTL. CLARKE

I. Introduction ... 11. New Tropane A1

.......... .......... ..............

...........

..........

..........

107

..........

153

..........

IX. Analytical Methods References .........................

...............

Chapter 3. Nuphar Alkaloids JERZY T. W R ~ B E L

I. Introduction . . . . . . . 11. 111. IV. V. VI.

.....

C,, Alkaloids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sulfur-Containing C,, Alkaloids . . . . . . . . . . . . . . . Mass Spectrometry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total Synthesis of C,, Nuphar Alkaloids . . . . . . . ............... Biosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

..... V

..............................

181

211 213

vi

CONTENTS

Chapter 4. Celestraceae Alkaloids ROGERM. SMITH I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

215

........... ... . . . . . . . 216 Acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 111. Structures of Esters of Nicot IV. Structures of Diesters of Substituted Nicotinic Acids, . . . . . . . . . . . . . . . . . . . . 227 11. Occurrence and Isolation . . .

t.......

V. Structures of Related Sesquiterpene ............................. VI. Biosynthesis . . . . . . . . . . . . . . . . . . . . . . VII. Biological Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

241 245 246 246

Chapter 5. The Bisbenzylisoquinoline A l k a l o i d s Occurrence, Structure, and Pharmacology M. P. CAVA,K. T. BUCK,and K. L. STUART I. 11. 111. IV. V. VI . VII . VIII.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structure Revisions . . . . . . . . . ......................... ............. New Alkaloids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Known Alkaloids from New Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . ......................... Methods and Techniques . . . ................... Pharmacology . . . . . . . . . . . . Bisbenzylisoquinoline Alkal ated by Molecular Weight.. . . . . . . ......... . . . . . . . . . . . . . . . . . . . . . . . . Appendix . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

250 251 257 297 297 300 301 304 312

Chapter 6. Syntheses of Bisbenzylisoquinoline Alkaloids MAURICESHAMMA and VASSILST. GEORGIEV 319 I. Introduction . . . . . . . . . . . . . . . . .......................... Dauricine-Type Alkaloids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320 Magnolamine-Type Alkaloids ............................ 336 Berbamine-Oxyacanthine-Type _ . . . . . . _ . . . . . 341 Thalicberine-Type Alkaloids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348

11. 111. IV. V. VI. VII. VIII. IX.

X. XI. XK XIII. XIV.

Trilobine-Isotrilobine-TypeA Menisarine-Type Alkaloids . . Tiliacorine-Type Alkaloids . . . . . . . . . . . . . . . . . . . . . . . . Liensinine-Type Alkaloids . . . Curine-Chondocurine-Type Alkaloids ......................... _................. Miscellaneous Syntheses . . . . . . . . . . . . . . . . . . . . . . . Syntheses Using Phenolic Oxidative Coupling . . . . . . . . . . . . . . . . . . . . . . . . . Synthesis Using Electrolytic Oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Use of Pentafluorophenyl Cop ............ . ...................

. . . . . . . . ...................

354 357 359 361 363 381 383 387 387 389

CONTENTS

vii

Chapter 7. The Hasubanan Alkaloids and TOSHIRO IBUKA YASUOINUSUSHI I. 11. III. IV. V. VI.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Occurrence and Physical Constants of the Hasubanan Structure Elucidations . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . Synthesis of the Hasubanan Skeleton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Synthesis of Hasubanan Alkaloids .......................... ... Biosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . .. . . ....................................

393 395 395 414 419 427 428

Chapter 8. The Monoterpene Alkaloids GEOFFREY A. CORDEU I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Isolation and Structure Elucidation of the Monoterpene Alkaloids . . . . . . 111. Biosynthesis and Biogenesis of the Monoterpene Alkaloids . . . . . . . . . . . . . IV. Pharmacology of the Monoterpene Alkaloids . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Summary ........................................................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

432 432 470 499 502 502

Chapter 9. Alkaloids Unclassified and of Unknown Structure R. H. F. MANSKE I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Plants and Their Contained Alkaloids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

511 511 551

SUBJECTINDEX... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

557

This Page Intentionally Left Blank

LIST OF CONTRIBUTORS Numbers in parentheses indicate the pages on which the authors’ contributions begin.

K. T. BUCK,Department of Chemistry, University of Pennsylvania, Philadelphia, Pennsylvania (249) M. P. CAVA,Department of Chemistry, University of Pennsylvania, Philadelphia, Pennsylvania (249) ROBERTL. CLARKE, Sterling-Winthrop Research Institute, Rensselaer, New York (83) GEOFFREY A. CORDELL, Department of Pharmacognosy and Pharmacology, College of Pharmacy, University of Illinois at the Medical Center, Chicago, Illinois (431) VASSILST. GEORGIEV, USV Pharmaceutical Corporation, Tuckahoe, New York (319) TOSHIRO IBUKA, Department of Pharmaceutical Sciences, Kyoto University, Sakyo-ku Kyoto, Japan (393) YASUOINUBUSHI, Department of Pharmaceutical Sciences, Kyoto University, Sakyo-ku Kyoto, Japan (393) R. H. F. MANSKE,Department of Chemistry, University of Waterloo, Waterloo, Ontario, Canada (511) DAVIDS. SEIGLER: Department of Botany, The University of Illinois, Urbana, Illinois (1) MAURICE SHAMMA, Department of Chemistry, The Pennsylvania State University, University Park, Pennsylvania (319) ROGERM. SMITH,School of Natural Resources, The University of the South Pacific, Suva, Fiji (215) K. L. STUART, Department of Chemistry, University of the West Indies, Kingston, Jamaica (249) JERZY T. WR~BEL, Department of Chemistry, University of Warsaw, Warsaw, Poland (181)

* Present address: Calle Peria 3166-9”A, Buenos Aires, Argentina.

ix

This Page Intentionally Left Blank

PREFACE The literature dealing with alkaloids shows no obvious signs of abatement. The classic methods of the organic chemist employed in structural determinations have evolved into spectral methods, and chemical reactions are involved largely in confirmatory and peripheral studies. Inasmuch as the spectral methods have become largely standardized we incline to limit the details in these volumes. Many new and already known alkaloids have been isolated from new and from previously examined sources. Novel syntheses are a prominent feature of recent publications. We attempt to review timely topics related to alkaloids.

R. H. F. MANSKE

x1

This Page Intentionally Left Blank

CONTENTS OF PREVIOUS VOLUMES

Contents of Volume I CHAPTER 1. Sources of Alkaloids and Their Isolation BY R . H . F . MANSKE . . . . . . . 2 . Alkaloids in the Plant BY W . 0. JAMES 3 . The Pyrrolidine Alkaloids BY LEOMARION . . . . . . 4 . Senecio Alkaloids BY NEISON J . LEONARD . . . . . . 5. The Pyridine Alkaloids BY LEOMARION . . . . . . 6. The Chemistry of the Tropane Alkaloids BY H . L . HOLMES . 7. The Strychnos Alkaloids BY H . L . HOLMES . . . . . .

. . .

. . . . . . .

.

.

. . . . . . . . .

1 15 91 107 165 271 375

Contents of Volume 11 8.1. 8.11. 9. 10. 11. 1 2. 13. 14. 15.

1 The Morphine Alkaloids I BY H . L . HOLMES . . . . . . . . The Morphine Alkaloids BY H . L . HOLMES AND (IN PART) GILBERT STORK161 Sinomenine BY H . L . HOLMES . . . . . . . . . . . . 219 Colchicine BY J . W . COOKAND J . D . LOUDON . . . . . . . . 261 Alkaloids of the Amaryllidaceae BY J . W . COOKAND J . D . LOUDON . . 331 Acridine Alkaloids BY J . R . PRICE . . . . . . . . . . . 353 The Indole Alkaloids BY LEOMARION . . . . . . . . . . 369 The Erythrina Alkaloids BY LEOMARION . . . . . . . . . 499 The Strychnos Alkaloids . Part 11BY H . L . HOLMES . . . . . . 513

Contents of Volume III 16. The Chemistry of the Cinchona Alkaloids BY RICHARD B . TURNER AND R. B. WOODWARD . . . . . . . . . . . . . . . 17. Quinoline Alkaloids Other than Those of Cinchona BY H . T . OPENSHAW 18. The Quinazoline Alkaloids BY H . T. OPENSHAW . . . . . . . 19. Lupine Alkaloids BY NELSON J . LEONARD . . . . . . . . . 20 . The Imidazole Alkaloids BY A . R . BATCERSBY AND H . T . OPENSHAW . . 21 . The Chemistry of Solanum and Veratrum Alkaloids BY V . PRELOG AND 0. JEGER . . . . . . . . . . . . . . . . . . 22 . P-Phenethylamines BY L . RETI . . . . . . . . . . . . 23. Ephreda Bases BY L . RETI . . . . . . . . . . . . . 24 . TheIpecac Alkaloids BY MAURICE-MARIE JANOT. . . . . . .

1 65 101 119 201

247 313 339 363

Contents of Volume N 25 . 26. 27. 28 .

The Biosynthesis of Isoquinolines BY R . H . F . MANSKE Simple Isoquinoline Alkaloids BY L. RETI . . . . Cactus Alkaloids BY L . RETI . . . . . . . . The Benzylisoquinoline Alkaloids BY ALFRED BURGER

xiii

. . . .

. . . .

. . . . . . . . . . . .

1 7 23 29

CONTENTS OF PREVIOUS VOLUMES

XiV

CHAPTER 29. The Protoberberine Alkaloids BY R . H . F . MANSKEAND WALTERR . ASHFORD . . . . . . . . . . . . . . . . . . . . . . . . . 30 . The Aporphine Alkaloids BY R . H . F. MANSKE 31 . The Protopine Alkaloids BY R. H . F. MANSKE . . . . . . . . 32. Phthalideisoquinoline Alkaloids BY JAROSLAV STANEK AND R. H . F . MANSKE . . . . . . . . . . . . . . . . . . 33 . Bisbenzylisoquinoline Alkaloids BY MARSHALL KULKA . . . . . 34 . The Cularine Alkaloids BY R. H . F . MANSKE . . . . . . . . 35 . a-Naphthaphenanthridine Alkaloids BY R . H . F . MANSKE . . . . 36 . The Erythrophleum Alkaloids BY G. DALMA . . . . . . . . 37 . The Aconitum and Delphinium Alkaloids BY E . S. STERN . . . .

77 119 147 167 199 249 253 265 275

Contents of Volume V 38 . 39. 40. 41 . 42 . 43 . 44 . 45 . 46 . 47 . 48 .

Narcotics and Analgesics BY HUGOKRUEGER . . . . . . . . Cardioactive Alkaloids BY E . L. MCCAWLEY . . . . . . . . Respiratory Stimulants BY MICHAEL J . DALLEMAGNE . . . . . Antimalarials B Y L. H . SCHMIDT. . . . . . . . . . . . Uterine Stimulants BY A . K . REYNOLDS . . . . . . . . . Alkaloids as Local Anesthetics BY THOMAS P. CARNEY . . . . . Pressor Alkaloids BY K . K . CHEN . . . . . . . . . . . Mydriatic Alkaloids BY H. R . ING . . . . . . . . . . . Curare-like Effects BY L . E . CRAIG . . . . . . . . . . . The Lycopodium Alkaloids BY R . H . F . MANSKE . . . . . . . Minor Alkaloids of Unknown Structure BY R . H . F . MANSKE . . .

1 79 109 141 163 211 229 243 265 259 301

Contents of Volume VI 1. 2. 3. 4. 5. 6.

7. 8. 9.

Alkaloids in the Plant BY K . MOTHES . . . . . . The Pyrrolidine Alkaloids BY LEOMARION . . . . . Senecio Alkaloids BY NELSON J. LEONARD . . . . . The Pyridine Alkaloids BY LEOMARION . . . . . The Tropane Alkaloids BY G . FODOR . . . . . . The Strychnos Alkaloids BY J . B . HENDRICKSON . . . The Morphine Alkaloids BY GILBERT STORK . . . . Colchicine and Related Compounds BY W . C . WILDMAN . Alkaloids of the Amaryllidaceae BY W. C . WILDMAN . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 31 35 123 145 179 219 247 289

Contents of Volume VII 10. 11 . 12. 13. 14. 15 . 16. 17 .

The Indole Alkaloids BY J . E . SAXTON . . . . . . . . . . The Erythrina Alkaloids BY V . BOEKELHEIDE. . . . . . . . Quinoline Alkaloids Other than Those of Cinchona BY H . T . OPENSHAW The Quinazoline Alkaloids BY H . T . OPENSHAW . . . . . . . Lupine Alkaloids BY NEWONJ . LEONARD . . . . . . . . . Steroid Alkaloids: The Holarrhena Group BY 0. JEGER AND V . PRELOG Steroid Alkaloids: The Solanum Group BY V . PRELQG AND 0. JEGER . . Steroid Alkaloids: Veratrum Group BY 0. JEGER AND V . PRELOG . .

1 201 229 247 253 319 343 363

CONTENTS OF PREVIOUS VOLUMES CHAFTER . . . . . . . . 18. The Ipecac Alkaloids BY R . H . F . MANSKE 19. Isoquinoline Alkaloids BY R . H . F . MANSKE . . . . . . . . 20. Phthalideisoquinoline Alkaloids BY JAROSLAV STAN~K . . . . . 21 . Bisbenzylisoquinoline Alkaloids BY MARSHALL KULKA . . . . . 22 . The Diterpenoid Alkaloids from Aconitum, Delphinium, and Garrya Species BY E . S. STERN. . . . . . . . . . . . . . 23 . The Lycopodium Alkaloids BY R . H . F . MANSKE . . . . . . . 24 . Minor Alkaloids of Unknown Structure BY R . H . F . MANSKE . . .

xv 419 423 433 439 473 505 509

Contents of Volume VIII 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

The Simple Bases BY J . E . SAXTON. . . . . . . . . . . Alkaloids of the Calabar Bean BY E . COXWORTH . . . . . . . The Carboline Alkaloids BY R . H . F . MANSKE . . . . . . . . The Quinazolinocarbolines BY R . H . F. MANSKE . . . . . . . Alkaloids of Mitragyna and Ourouparia Species B Y J . E . SAXTON. . Alkaloids of Gelsemium Species BY J . E . SAXTON . . . . . . . Alkaloids ofPicralima nitida BY J. E . SAXTON . . . . . . . . Alkaloids ofAlstonia Species BY J . E . SAXTON. . . . . . . . The Zboga and Voacanga Alkaloids BY W . I . TAYLOR . . . . . . The Chemistry of the 2,2'-Indolylquinuclidine Alkaloids BY W . I . TAYLOR The Pentaceras and the Eburnamine (HunteriabVicamine Alkaloids BY W . I . TAYLOR . . . . . . . . . . . . . . . . 12. The Vinca Alkaloids BY W . I . TAYLOR . . . . . . . . . . 13. RauwolfiaAlkaloids with Special Reference to the Chemistry of Reserpine

1 27 47 55 59 93 119 159 203 238

250 272

BYE . SCHLITTLER . . . . . . . . . . . . . . . 287 14. The Alkaloids ofdspidosperma, Diplorrhyncus,Kopsia, Ochrosia, Pleioc a r p , and Related Genera BY B . GILBERT . . . . . . . . 336 15. Alkaloids of Calabash Curare andStrychnos Species BY A . R . BATTERSBY . . . . . . . . . . . . . . 515 AND H . F . HODSON . 16. The Alkaloids of Calycanthaceae BY R . H . F. MANSKE . . . . . 581 17 . Strychnos Alkaloids BY G. F . SMITH . . . . . . . . . . . 592 18. Alkaloids ofHaplophyton cimicidum BY J . E . SAXTON . . . . . 673 19. The Alkaloids of Geissospermum Species BY R . H . F. MANSKEAND W . ASHLEY HARRISON . . . . . . . . . . . . . . . 679 20 . Alkaloids ofPsuedocinchona and Yohimbe BY R . H . F . MANSKE . . 694 AND A . HOFMA" . . . . . . 726 21 . The Ergot Alkaloids BY A . STOLL 22 . The Ajmaline-Sarpagine Alkaloids BY W . I . TAYLOR . . . . . . 789

Contents of Volume IX 1. 2. 3. 4.

1 The Aporphine Alkaloids BY MAURICE SHAMMA . . . . . . . 41 TheProtoberberine Alkaloids BY P . W . JEFFS . . . . . . . . Phthalideisoquinoline Alkaloids BY JAROSLAV S T A N ~ K . . . . . 117 Bisbenzylisoquinoline and Related Alkaloids BY M . CURCUMELLIRODWTAMO AND MARSHALL KULKA . . . . . . . . . . 133 5 . Lupine Alkaloids BY FERDINAND BOHLMANN AND DIETERSCHUMANN 175 6 . Quinoline Alkaloids Other than Those of Cinchona BY H. T . OPENSHAW 223

xvi

CONTENTS OF PREVIOUS VOLUMES

CHAPTER 7 . The Tropane Alkaloids BY G . FODOR . . . . . . . . . . 269 8 . Steroid Alkaloids: Alkaloids of Apocynaceae and Buxaceae BY V. ERN+ AND F . SORM. . . . . . . . . . . . . . . . . 305 9 . The Steroid Alkaloids: The Salamandra Group BY GERHARD HABERMEHL427 10. Nuphar Alkaloids BY J . T. WROBEL . . . . . . . . . . . 441 11. The Mesembrine Alkaloids BY A . POPELAK AND G. LETFENBAUER . . 467 12. The Erythrina Alkaloids BY RICHARD K . HILL. . . . . . . . . 483 13. Tylophora Alkaloids BY T . R . GOVINDACHARI. . . . . . . . 517 14. The Galbulimima Alkaloids BY E . RITCHIEAND W. C . TAYLOR . . . 529 15. The Stemona Alkaloids BY 0 . E . EDWARDS. . . . . . . . . 545

Contents of Volume X 1. Steroid Alkaloids: The Solanun Group BY KLAUSSCHRIEBER. . . 1 2 . The Steroid Alkaloids: The Veratrum Group BY S . MORRISKUPCHAN AND ARNOLDW.BY . . . . . . . . . . . . . . . . 193 3 . Erythrophleum Alkaloids BY ROBERT B . MORIN . . . . . . . 287 4 . The Lycopodium Alkaloids BY D . B . MACLEAN. . . . . . . . 306 5. Alkaloids of the Calabar Bean BY B. ROBINSON . . . . . . . 383 6. The Benzylisoquinoline Alkaloids BY VENANCIODEULOFEU,JORGE COMIN.AND MARCELO J . VERNENGO . . . . . . . . . . 402 7 . The Cularine Alkaloids BY R . H . F. MANSKE . . . . . . . . 463 8 . Papaveraceae Alkaloids BY R . H . F. MANSKE . . . . . . . . 467 9 . a-Naphthaphenanthridine Alkaloids BY R . H . F . MANSKE . . . . 485 10. The Simple Indole Bases BY J . E . SAXTON. . . . . . . . . 491 11. Alkaloids of Picralima nitida BY J . E . SAXTON . . . . . . . . 501 12. Alkaloids of M i t m g y m and Ourouparia Species BY J . E . SAXTON . . 52 1 13. Alkaloids Unclassified and of Unknown Structure BY R. H. F. MANSKE 545 14. The Taxus Alkaloids BY B . LYTHGOE . . . . . . . . . . 597

Contents of Volume XI 1. 2. 3. 4. 5. 6. 7. 8.

The Distribution of Indole Alkaloids in Plants BY V . SNIECKUS . . . I The Ajmaline-Sarpagine Alkaloids BY W . I . TAYLOR . . . . . . 41 The 2,2‘-Indolylquinuclidine Alkaloids BY W . I. TAYLOR . . . . . 73 The Iboga and Voacanga Alkaloids BY W . I . TAYLOR . . . . . . 79 The Vinca Alkaloids BY W. I . TAYLOR . . . . . . . . . . 99 The Eburnamine-Vincamine Alkaloids BY W. I . TAYLOR . . . . . 125 Yohimbirw and Related Alkaloids BY H . J . MONTEIRO . . . . . 145 Alkaloids of Calabash Curare and Strychnos Species BY A . R . BATTERSBY ANDH. F . HODSON . . . . . . . . . . . . . . . 189 9 . The Alkaloids of Aspidosperma, Ochrosia, Pleiocarpa, Melodinus, and Related Genera BY B . GILBERT . . . . . . . . . . . 205 10. The Amaryllidaceae Alkaloids BY W . C. WILDMAN . . . . . . 307 A N D B. A . PURSEY407 11. Colchicine and Related Compounds BY W . C. WILDMAN 12. The Pyridine Alkaloids BY W . A . AYERAND T . E . HABGOOD. . . . 459

CONTENTS OF PREVIOUS VOLUMES

xvii

Contents of Volume XI1 CHAFTER The Diterpene Alkaloids: General Introduction BY S. W. PELLETIER AND L. H. KEITH . . . . . . . . . . . . . . . . . . 1. Diterpene Alkaloids from Aconitum, Delphinium, and Garrya Species: The C,,-Diterpene Alkaloids BY S. W. PELLETIER AND L. H. KEITH 2. Diterpene Alkaloids from Aconitum, Delphinium, and Garrya Species: The Go-DiterpeneAlkaloids BY S. W. PELLETIER AND L. H. KEITH 3. Alkaloids ofAlstonia Species BY J. E. SAXTON. . . . . . . . 4. Senecio Alkaloids BY FRANK L."WARREN . . . . . . . . . 5. Papaveraceae Alkaloids BY F. SANTAVY . . . . . . . . . 6. Alkaloids Unclassified and of Unknown Structure BY R. H. F. MANSKE 7. The Forensic Chemistry of Alkaloids B Y E .G. C. CLARKE . . . .

xv 2 136 207 246 333 455 514

Contents of Volume XIII 1 The Morphine Alkaloids BY K. W. BENTLEY . . . . . . . . The Spirobenzylisoquinoline Alkaloids BY MAURICE SHAMMA . . . 165 The Ipecac Alkaloids BY A. BROSSI,S. TEITEL,AND G. V. PARRY. . . 189 Alkaloids of the Calabar Bean BY B. ROBINSON . . . . . . . 213 The Galbulirnima Alkaloids BY E. RITCHIEAND W. C. TAYLOR . . . 227 The Carbazole Alkaloids BY R. S. KAPIL . . . . . . . . . 273 Bisbenylisoquinoline and Related Alkaloids BY M. CURCUMELLI-RODC+ STAMO . . . . . . . . . . . . . . . . . . . 303 8. The Tropane Alkaloids BY G. FODOR . . . . . . . . . . . 351 9. Alkaloids Unclassified and of Unknown Structure BY R. H. F. MANSKE 397 1. 2. 3. 4. 5. 6. 7.

Contents of Volume X N 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Steroid Alkaloids: The Veratrum and B w u s Groups BY J. TOMKO AND 1 2. VOTICKP . . . . . . . . . . . . . . . . . 83 Oxindole Alkaloids BY JASJIT S. BINDRA . . . . . . . . . Alkaloids of Mitragym and Related Genera BY J. E. SAXTON . . . 123 Alkaloids ofPicralima and Alstonia Species BY J . E. SAXTON . . . 157 The Cinchona Alkaloids BY M. R. USKOKOVIC AND G. GRETHE . . . 181 The Oxoaporphine Alkaloids BY MAURICE SHAMMA AND R. L. CASTENSON 225 Phenethylisoquinoline Alkaloids BY TETSUJIKAMETANIAND MASUO KOIZUMI . . . . . . . . . . . . . . . . . . 265 Elaeocarpus Alkaloids BY S. R. JOHNS AND J. A. LAMBERTON . . . 325 The Lycopodium Alkaloids BY D. B. MACLEAN . . . . . . . 347 TheCancentrine Alkaloids BY RUSSELLRODRIGO. . . . . . . 407 The Securinega Alkaloids BY V. SNIECKUS . . . . . . . . . 425 Alkaloids Unclassified and of Unknown Structure BY R. H. F. MANSKE 507

xviii

CONTENTS OF PREVIOUS VOLUMES

Contents

of

Volume X V

CHAPTER 1. The Ergot Alkaloids BY P. A. STADLER AND P. SWTZ . . . . . . 1 2. The Daphniphyllum Alkaloids BY SHOSUKE YAMAMURA AND YOGHIMASAHIRATA . . . . . . . . . . . . . . . . 41 3. The Amaryllidaceae A l k a l o i d s ~CIAUDIOFUGANTI ~ . . . . . . 83 AND E. U. KAUBMANN 165 4. The Cyclopeptide Alkaloids BY R. TSCHESCHE 5. The Pharmacology and Toxicology of the Papaveraceae Alkaloids BY V . PREININCER . . . . . . . . . . . . . . . 207 6. Alkaloids Unclassified and of Unknown Structure BY R. H. F. MANSKE 263

-CHAPTER

1-

PLANT SYSTEMATICS A N D ALKALOIDS DAVIDS. SEICILER The University of I ~ ~ i n o i s Urbana, Illinois

I. Introduction ........................................................ A. What Is Plant Systematics ? ....................................... B. Major Goals of Plant Systematics .................................. 11. Data to Be Utilized ................................................. A. Relationship of Chemical Data to Botanical Data .................... B. Rationale for Using Chemical D a t a . . ............................... C. Botanical and Chemical Literature ................................. D. Documentation of Plant Materials. . . . . . . . . . 111. Application of the Data to Biological Problems . A. Nature and Sources of Variation in Plants. .. B. Basic Pathways of Alkaloid Biosynthesis .... IV. Alkaloids in Lower Vascular Plants and Gymnos V. Alkaloids in the Angiosperms ......................................... A. Introduction ..................................................... B. The Magnoliopsida (Dicotyledonous Plants) .......................... ida (Monocotyledonous Plants) ...........................

......................................................

1

2 2 3 3 3 6 7

8 8 14 20 22 22 24 65 73

I. Introduction Many scientists, both chemical and biological, have sought to correlate chemical characters (i.e., the presence of certain types of compounds) with various botanical entities. I n the past, several factors have limited the success of such efforts, and it is only in recent years that such correlations have been applied to many plant groups. My purpose in this article is to review several of these earlier attempts as well as to examine current thinking in this area of endeavor. Several new ideas concerning the placement of selected plant groups within taxonomic systems will be discussed, and in addition, certain enigmatic problems that as yet cannot be clearly resolved will be posed as subjects for future investigation. As background t o these discussions, I will first describe the nature and goals of plant systematics t o provide the reader with the necessary perspective to understand the needs of that science.

2

DAVID S. SEIOLER

A. WHATIs PLANT SYSTEMATICS ? Systematics is the scientific study of the kinds and diversity of organisms and of the relationships between them ( 1 ) .I n former times, much systematic work was based on the examination of preserved herbarium specimens in an effort to describe and classify various plant taxa (a term indicating taxonomic entities of unspecified rank). These studies frequently involved an examination of the form and structural features of relatively small numbers of specimens. Although this approach is still viable in many tropical areas of the world where rich and unstudied floras are in immediate danger of destruction or extreme modification (2, 3), it is largely being supplanted by examination of larger numbers of plants from living populations in temperate areas of the world, where the floras are better known. By means of this latter method, often called biosystematics, one attempts to study as much of the biology of the plant as possible and utilize these data to clarify the taxonomic and evolutionary relationships of the taxa involved ( 4 ) . The information derived from both approaches is normally utilized in two ways: to prepare floras of a particular region (often a state or large natural geographic region) or to account for all the species within a given group-for example, a genus or a family, regardless of where the plants grow ( 5 ) . Although each of the above aspects of systematics assists in identification and location of plant materials this information may also be invaluable to workers in many other fields such as chemistry, ecology, forestry, horticulture, floriculture, genetics, agronomy, zoology, entomology, or pharmacognosy, because of its predictive nature. Despite the introduction of many new approaches and technological advances, the basic systems of taxonomy that have been used for the last two centuries have not changed radically nor are they likely to undergo substantial modification. Movement of certain groups within the systems has occurred frequently. I n this chapter, the system proposed by Cronquist ( 6 )will be used as a basis for discussion, although frequent reference will be made to a number of other contemporary systems. Several of these systems (at the level of family and above) have recently been compared by Becker in Radford et al. ( 5 ) , and reference to that work will prove useful in understanding many taxonomic problems that will be discussed. B. MAJORGOALSOF PLANT SYSTEMATICS I n summary, the principal goals of plant systematics are to (a) provide a convenient method of identifying, naming, and describing

1.

PLANT SYSTEMATICS

3

plant taxa, (b) provide an inventory of plant taxa via local, regional, and continental floras, and (c) provide a classification scheme that attempts to express natural or phylogenetic relationships and t o provide an understanding of evolutionary processes and relationships ( 5 ) .In the subsequent parts of this chapter, I will present and discuss ways in which chemical data and in particular alkaloid chemical data can be utilized in meeting these goals. 11. Data to Be Utilized

A. RELATIONSHIP OF CHEMICAL DATATO BOTANICAL DATA As both morphological and chemical features are determined by genetics, the structure of a molecule must be as much a character as any other (7). Further, all the “characters” of a plant must be related and self-consistent. Thus, it is scarcely surprising that new cytological, numerical, and chemical data have provided valuable complementary information about the placement of groups within the taxonomic system rather than upsetting the results of extensive morphological investigations. How did these two types of characters arise and how do they differ Z I n the course of evolution the fate of any change in the genetic material of an organism will in large part depend on the function of the products produced. For example, changes in respiratory proteins, such as cytochromes, are unlikely t o survive, whereas changes in the enzymes that produce alkaloids or other secondary metabolic products are more likely to persist. The evolution of morphological and chemical features of an organism must be interrelated, but significantly, the forces of natural selection do not have the same effect on each type of genetic expression. These differences in selection are very important from a systematic standpoint because evolution of chemical constituents differs from morphological evolution, making the examination of both morphological and chemical characters an extremely valuable approach to the study of evolutionary problems (8).Because the structure of any compound is determined by a series of biosynthetic steps, each of which is under differing selective forces, not only may the structure of the compound itself be useful, but the biochemical pathway by which i t has arisen may be of systematic significance. FOR USING CHEMICALDATA B. RATIONALE The two major groups of compounds that have been applied to t,axonomic problems involve basically different approaches and appear

4

DAVID S . SEIGLER

t o be useful in different manners. To date, these applications involve niacromolecules (in particular proteins) and micromolecules (mostly secondary metabolic compounds such as terpenes, flavonoids, alkaloids, cyanogenic and other glycosides, amino acids, and lipids of various types). When one utilizes macromolecules, he is examining the primary products of plant DNA and changes in amino acids within the protein reflect changes in the base sequence of the DNA. Initial studies of protein sequencing, especially those studies involving cytochrome C, indicate that this data provides valuable information about phylogeny and relationships a t the higher taxonomic categorical levels (families, orders, classes). Cytochrome c, which occurs in both animals and plants, has been sequenced in several species of animals (9). The fossil record for animals generally confirms information derived from these phylogenetic studies. The number of similarities in amino acids in particular positions in cytochrome c molecules from different animals makes it statistically improbable that they could have arisen from more than a single ancestral type with an ancestral cytochrorne c molecule. By tracing the differences in amino acid substitutions it is possible t o relate various groups of animals, as successive groups after a modification carry the changed cytochrome c molecule. I n plants, especially flowering plants, there is no extensive fossil record and much of the current knowledge of relationships and phylogeny in this group is based on extrapolation of studies of morphological data. To date, relatively few plant cytochromes have been studied, but in the few that have been investigated, it is apparent from the number of similarities of amino acid sequences that plant and animal G Y ~ O chromes are related. It is also evident that the sequences of amino acids in genera of the same family me more similar to each other than to those of other families and that families thought to be closely related by morphological evidence generally resemble each other more closely than less related families. The evolutionary history of plant groups, as well as of animals, appears t o be recorded in this and other proteins. Much recent work has established that micromolecular chemical data can also provide valuable insight into evolutionary processes ( 8 ) . Chemical studies of secondary products have proved useful in resolving many problems of specification and evolution but in contrast to protein sequencing data have generally been applied to the study of lower taxonomic categories, i.e., problems a t the species and genus level (10, 11). However, as will be pointed out, they may also be of value a t higher taxonomic levels. To understand how secondary compounds can be useful for the study

1. PLANT SYSTEMATICS

5

of systematic problems, it is necessary to consider how and why they arose. Plants have a multitude of proteinaceous materials, many of which have enzymatic functions. In primitive organisms these compounds were and are largely active in synthesizing primary metabolic components of cells. As these organisms evolved, genetic material and its derived proteins were duplicated and increased both in amount and in redundancy. Mutations occurred that subsequently produced changes in the proteins and their products. The forces of natural selection operated on all such products (12), selecting them for value to and compatibility with parental organisms and the ecological systems in which they occur. Many of these compounds were of a less critical nature than primary metabolites and were less widely distributed. Complications are introduced because one does not observe the primary gene products, but rather pools of compounds they produce, the concentrations of which are partially functions of the relative amounts and activities of enzymes, the availability of certain precursors, and compartmentalization and translocation with the cell ( 4 ) . Subsequent mutations may affect steps in a biosynthetic sequence that we observe as an accumulation or disappearance of an altered product. These mutations usually involve the loss, gain, blockage, or alteration of the specificity of an enzyme system. Loss of synthetic ability is presumably more common than gain or alteration, since it merely implies destruction or blocking of a process instead of setting up a new one ( 7 ) . This is partially confirmed by the observation that in several groups of species from the related genera Parthenium, Hymenoxys, and Ambrosia of the Compositae, more highly evolved members have simplified patterns of secondary compounds (13).A one-gene loss may also block an entire pathway. The determination of homologous origin of similar compounds in different taxonomic groups is one of the fundamental problems inherent in the taxonomic application of secondary compounds. Two taxa may synthesize or pool the same products by different pathways; therefore, the mere presence of a compound is not necessarily an indication of relationship; i.e., similarities in the chemistry of plant taxa (or morphological features) may reflect an evolutionary or phyletic similarity but may also be the result of convergent evolutionary processes ( 4 ) . With a knowledge of biosynthetic pathways of secondary compounds in plants, it should be possible to determine a t what point in a sequence divergence has occurred and what subsequent changes have come to pass (7). In reality this is rarely realized because of several factors; several classes of compounds do not appear to have specific structural requirements, whereas in others less variation can be tolerated. For example, most phenolic substances could serve as antioxidants or many

6

DAVID S. SEIGLER

lipid compounds for surface coatings as long as the necessary physical properties are met ; but attractants for specific pollinators or diterpenes with hormonal activities must be precisely synthesized (7). Many plant products arise by simple processes such as removal of activating groups (as phosphate or coenzyme A) or from oxidations, reductions, or methylations of easily modified groups (7). I n some cases the relative amounts of products produced may simply reflect the rates of two enzymes operating on a common precursor. Highly probable reactions, such as the introduction of an hydroxyl group ortho or para to an existing one in a phenol, occur frequently in nature. These types of changes are usually of only minor importance in considering the taxonomic significance of secondary compounds. Other reaction sequences are reversible or are controlled by feedback inhibition controls such that when a given compound disappears it disappears without a trace or causes accumulation of a compound far removed in the sequence. For example, polyketide chains, probably as coenzyme A esters, are rapidly reversible to their initial units unless some chemically irreversible stage is reached such as reduction or cyclization (7). In the fungus Penicillium islandicum which produces polyketide anthraquinones, mutation simply leads to the complete absence of these compounds. We have limited knowledge as to what pathways may be available in advanced plant groups as we can only see the products of those pathways that the plant utilizes a t a particular time. Several lines of work suggest that many plants are capable of carrying out complex reactions or reaction series but lack precursors or particular enzymes under normal situations. For example, when plants of Nicotiana are fed thebaine and certain other precursors of morphine they are able to perform several biosynthetic steps and produce morphine (14)which is not known to occur naturally in the genus. Interestingly, this conversion cannot be made by some species of Papaver, although other species of the genus contain thebaine and morphine. In assessing the importance of a particular change as an evolutionary step it is necessary to decide on the probability of its occurrence. As a general rule, the more difficult the reactions and the less available the building blocks or the more reaction steps required in a definite sequence to give rise to a compound, the rarer will be its convergent formation

(14). C . BOTANICAL AND CHEMICAL LITERATURE Many earlier publications were based on mass collections of materials, often gathered from large geographical areas and/or of uncertain origin.

1.

PLAXT SYSTEMATICS

7

Frequently, only the major constituents-those that were poisonous, crystallized readily, or had other easily detectable properties-were examined. These facts must be considered by those who intend to apply the information to a taxonomic problem. Another difficulty in utilizing chemical data from the literature is a lack of reliability of certain structure determinations and in particular the identification of plant products by such physical properties as gas-liquid chromatography retention time, paper and thin-layer chromatography R, values, color reactions, and spot tests. Misidentification of compounds by wet chemical methods is not uncommon in the older literature before advanced spectral methods became available and must always be considered. One of the most serious problems in utilizing literature data is that almost no chemical reports are supported by adequately vouchered plant materials. Proper vouchering records would make it possible to examine the original materials and allow comparison with other collections in order to ascertain whether (a) the material was correctly identified and (b) certain phenomena, such as hybridization, introgression, or subspecific variations exist. It would also permit subsequent workers to determine the presence of fungi, lichens, algae, insects, etc., that may be involved in the production of certain secondary compounds. If a small portion of the actual materials utilized for the research is also preserved, it would permit later analysis for foreign contaminants. I n other cases, careful perusal of the botanical literature will reveal that taxonomists have placed taxa of various rank incorrectly. These incorrect placements may range from questionable or aberrant species in a genus to the realignment of entire orders of plants. Chemical data can assist in resolving problems of this type, but they sometimes provide enigmatic results until sufficient information is available to allow a reassignment of the taxa involved. One must look carefully and critically a t all reported data to be sure both chemical and botanical portions of the work have been done and interpreted correctly before applying the data to a problem under investigation.

D. DOCUMENTATION OF PLANT MATERIALS As mentioned in the preceding section, many early reports of alkaloids and other secondary compounds are suspect because accurate techniques required for assignment of complex structures were not available. Nonetheless, the major problem in using these data for systematic studies is not the reliability of the chemical data but the identity of the plant materials that were examined (15).

8

DAVID S . SEIGLER

To document the materials used, the investigator should always have a competent person identify his plant materials and a portion should be dried or otherwise preserved as a voucher specimen so that further examination of the specimen is possible should it be desirable. The selected plant should be typical for the population and, when possible, should have mature reproductive organs. Full collection or acquisition data (data, location, collector, habitat, etc.) should be provided and the specimen deposited in a recognized herbarium. Taxonomists usually will be willing to assist with the necessary details of voucher specimen preparation. Most major universities have collections of dried plant specimens (a herbarium) that provide a wealth of data about the ranges, flowering t'imes, uses, soil preferences, and other information about particular species as well as preserving materials for future study or reinvestigation. I n publications describing chemical results, one should record the locations and dates of plant collections, the parts of the plants used in the study, the name of the herbarium where the voucher specimens are deposited, and the name of the taxonomist who identified the plants. With this information and with the possibility of comparing specimens collected a t other times with the original vouchers, later investigators can usually determine the relationship of the plants concerned to the original collection (15, 16).

111. Application of the Data to Biological Problems

A. NATUREAND SOURCES OF VARIATION IN PLANTS Until sensitive separation techniques (column, paper, thin-layer, and gas chromatography; countercurrent distribution; etc.) and sensitive methods of instrumental analysis (IR, NMR, UV, and mass spectrometry) became available, it was not feasible to undertake the analysis of secondary plant constituents from single plants of most species in naturally occurring populations. These new microtechniques permit the chemist or botanist to obtain chemical data from single plants rapidly, allowing the extension of the biosystematic approach to chemical as well as morphological characters. When phytochemical workers began to examine single plants, they were often frustrated by apparently uninterpretable variations of chemical constituents. Many of these investigators did not do adequate sampling, ignored the significance of these variations, and came to

1. PLANT SYSTEMATICS

9

conclusions based on a meager amount of data in comparison to what was actually needed. Recent combined chemical and morphological investigations have used this information more fully and proved that, instead of being troublesome, the study of chemical and morphological variation actually provides a key to the solution of many problems of biological speciation, hybridization, and introgression.* Relationship between plant taxa is established by “ summarizing ” the similarities between groups of organisms and contrasting their differences. We consider two plants to be closely related if they have many common characters and only distantly so (or at higher categorical levels) if the differences outweigh the similarities. In contrast to this, the name of the game in evolution is change and the ability to maintain variability. Few natural populations are without measurable variation; that is, plants from interbreeding groups that share a gene pool have phenotypic and genotypic differences that can be seen even by inexperienced observers. How do these variations arise and how are they maintained ! Each individual plant must possess the ability to respond to its environment, but this variation must remain within the limits set by the genetic makeup of the taxon (12, 1‘7). Thus, phenotypic expression is determined by both genotypic composition and reaction to a specific environment. Some characters are little changed by environment--e.g., leaf arrangement or floral structure-and these have been considered “good characters” or to be “genetically fixed.” Other characters are known to vary radically and are said to be “phenotypically plastic.” Examples of characters of this type are leaf shape, stem height, and time of flowering. The effects of environment are superimposed on and may obscure genotypic variability; further, it is the phenotype produced by both that is is exposed to the pressures of natural selection. Davis and Heywood ( 1 7 ) have listed a number of important physical factors in determining the appearance of a plant in nature. Among these are light, seasonal variation, elevational differences, terrestrial versus epiphytic state, photoperiodism, temperature, temperature periodic effects, water (heterophylly), wind, soil (e.g., halophytes), and biotic factors such as fungal and bacterial infection, ant habitation, galls, grazing and browsing, fire, and trampling. The population is considered by many to be the basic evolutionary * Introgression is the process by which the genes of one taxon are mixed with the genes of another by hybridization of the two taxa followed by backcrossing of the hybrid plants with either of the two parents. Even when hybrids are not significant in relative numbers, they can allow gene flow and mixing, producing increased variability of the two parental types.

10

DAVID S . SEIGLER

unit and when we discuss speciation and concomitant chemical change it is necessary to understand something of the nature of variation both within and among populations of a given taxon or group of taxa. Populational variations are a function of the variation of individual plants and of the common gene pool that they possess. Morphological and chemical features enable us to recognize the population, but they do not define it. It must also be remembered that the population is a dynamic entity. It changes in numbers of plants and, even in some perennials, in the particular individuals present in a given year. A population may occupy a much larger geographical area in some years than others. It may separate into two or several new populations under some conditions that may be maintained or later merge with the parental population. Taxonomic descriptions are sometimes based on a single plant specimen, which may not reflect the nature of the species or its populations. Several factors are important in determining genetic variation. Mutations usually produce a one-gene change, but these changes may have profound effects. Such changes as zygomorphic corollas t o actinomorphic corollas in Antirrhinum, the gamosepalous to polysepalous condition in Silene, spurred t o nonspurred flowers in Aquilegia, and annual to biennial condition in Atropa are all known to be controlled by one gene ( 1 7 ) . Most mutations affect several characteristics of the phenotype. Thus, a species may differ from another in several characters but still may be separated by only a one gene difference. Characters that have no selective advantage in themselves can become established through the secondary effects of genes that have been selected as valuable to the organisms for completely different reasons ( 1 7 ) .Certain genetic variants coexist in temporary or permanent equilibrium within a single population in a single spatial region in a phenomenon known as polymorphism ( 1 7 ) . Recombination of genetic variability in populations is largely determined by the breeding system. Cross-fertilized populations contain a large store of variability hidden in the form of recessive genes in the heterozygous condition. This variability serves as insurance in the presence of a constantly changing environment. I n sexual populations breeding tends to take place principally between neighboring individuals. I n summary, the three factors that largely control variation in populations are (a) external environmental modification, (b) mutation, and (c) genetic recombination ( 1 7 ) . Populations rarely stay the same over a period of time but are affected by the process of natural selection in a stabilizing, disruptive, or directional manner. Populations separated by geographical, ecological, or reproductive barriers will tend to differ-

1.

PLANT SYSTEMATICS

11

entiate into a series of populations that may have gradually accrued differences (clinal variation) or stepwise variations associated with ecological differences (ecotypic variation) (17). If the differences between populations increases sufficiently, and especially if reproductive barriers arise, these differentiating populations may be recognized as species. Stebbins ( 1 2 ) considers four major factors in speciation: (a) mutation, (b) genetic recombination, (c) natural selection, and (d) isolation. I n small, often peripheral populations, chance may play a greater role in speciation because the probability of loss of a particular character is greater; recessive genes are more likely to appear and become homozygous, and the genetic nature of the population may be determined by the “founders” or “survivors” of a period of catastrophic selection. These phenomena explain many of the variational patterns observed in the distribution and occurrence of secondary plant compounds, especially at the lower taxonomic ranks, and although they have mostly been examined by means of morphological characters, much evidence suggests that evolution and speciation may be studied or measured by chemical characters as well. I n the preceding discussion, variation of morphological characters has been considered. There is no reason t o think that variation in chemical characters has not occurred and is not maintained in a similar manner. I n contrast to morphological features, however, the specific structures and steps of biosynthetic pathways are easier to quantify and generally simpler in terms of genetic control (at least in principle). Secondary compounds are affected by environmental as well as genetic factors (18, 19). In a study of alkaloids of the genus Baptisia (Leguminosae), Cranmer studied the variation of lupine alkaloids during the development of individual plants in different populations of Baptisia leucophaea Nutt. ( 2 0 ) . Individual plants in each population exhibited considerable quantitative variation, while plants from different populations were similar at similar stages of development. However, there was striking variation in the specific alkaloids produced, the relative amounts of each, and in the total quantity of alkaloids present a t any given time in development. Nowacki encountered similar variation in lupine alkaloids in the genus Lupinus ( 2 1 ) . A number of workers have examined the genetics of alkaloid production by the study of hybrid plants (14, 21-25). These results indicate that the genetic mechanisms that control alkaloid synthesis are complex and that hybridization and introgression can produce significant variations in the alkaloid content of plants within a population. Many past workers have been unaware of natural hybridization and, because these plants are occasionally indistinguishable from the parental species,

12

DAVID S. SEIGLER

have not been able t o interpret the alkaloid patterns observed (14, 15). Hybridization and introgression in the genus Baptisia has been extensively studied by workers a t the University of Texas. Several populations that contained all possible hybrid combinations, plants derived from back-crossing these plants with the parental plants, and the parental plants were examined. The status of these plants was established by independent methods; subsequently the alkaloid chemistry was examined. The data indicated that the hybrid plants not only failed to exhibit the alkaloid chemistry of the parent species either singly or combined, but also showed some striking quantitative variation among individual hybrid plants. Mabry concludes that this variation is extremely useful and represents one of the best available techniques for detecting and documenting natural hybridization and introgression (26). Extensive variation can occur in the different parts of an individual plant ( 2 7 ) . Changes associated with the reproductive parts of a plant are often striking; these organs also exhibit the greatest amount of morphological change during a plant’s growth and development. Cranmer and co-workers (20, 28) observed that in Baptisia species alkaloids often showed greater variation between organs of plants from a single species than between the same organs for different species. The total yield of alkaloids from different organs was also shown t o vary significantly. The most thoroughly investigated plants in this regard are medicinally important ones such as Papaver somniferum L. and solanaceous plants of the genera Nicotiana, Atropa, Hyoscyamus, and Datum (27). At the present time our lack of knowledge of the specific enzymology of the synthesis of secondary metabolites prevents direct comparison of many of the pathways involved in various taxa. Examination and comparisons must frequently be restricted t o those systems ascertained t o be related by other reasoning, such as a knowledge of the structures of other compounds derived from and part of the biosynthetic pathways in the same and related species of plants. Secondary compounds have classically been viewed as waste or excretion products ( l a ) ,but a body of information is accumulating that suggests that many have important coevolutionary defensive and attractive roles (29-31) as well as primary metabolic importance (32-34). The forces of natural selection seldom operate on a single organism but on a total biological system. This is undoubtedly one reason convergence in the evolution of both morphological and chemical characters is observed. It is well known, for example, that certain habitats are occupied by

1.

13

PLANT SYSTEMATICS

plants that possess similar morphological features (12, 27, 35-38). It has not been definitely established, but it appears that various chemical components of plants can be seiected to produce convergence of chemical types. One example that confirms this possibility is that Ammodendron conollyi Bge., a legume native to Central Asia, contains the alkaloids ammodendrine (1)and sparteine (2), and another plant from

COCH, 1

that area, Anabasis aphylla L., a member of the Chenopodiaceae, contains similar alkaloids such as lupinine (3),aphyllin (a), and anabasin ( 5 ) .I n the legume, cadaverine (and hence lysine) serves as a precursor

0 3

4

5

for both types, whereas in Anabasis, the quinolizidine alkaloids are formed as in legumes but anabasine is derived from nicotinic acid as in Nicotiana. Thus, what might appear to be a close similarity is in reality an analogous route to the same compounds ( 1 4 ) . I n another example, three species of the genus Hymenoxys (Compositae), H . scuposa (DC.) K. F. Parker, H . acaulis (Pursh) K. F. Parker, and H . ivesianu (Greene) K. F. Parker, contain more than thirty flavonoids. The patterns of distribution of these compounds are correlated more strongly with population positions along an east-west gradient extending from Arizona to Texas than with the diagnostic morphological features of the species. The biochemical parallelism observed for populations of different species in the same region suggests the action of common selective forces (39). It has been observed that small, isolated island populations of mainland taxa usually have fewer and simpler compounds than their mainland ancestors. This may be because of lowered selection by predation or because island habitats have different environmental requirements (35).

14

DAVID S. SEIQLER

B. BASICPATHWAYS OF ALKALOID BIOSYNTHESIS In the preceding section we have surveyed some of the ways in which variation originates and is maintained in plants. A knowledge of these variations is extremely important in systematic studies a t the lower taxonomic levels (genus-species), but when one wishes to establish relationships a t higher ranks, e.g., at the family, order, and subclass level, it is necessary to survey as many taxa and individuals as possible to reduce the effects of these variations. That is, we need to know what morphological features are produced and what biosynthetic pathways exist in a particular group of taxa to compare them. This is made more difficult by our imperfect knowledge of biosynthetic pathways, but, by careful observation of their products, we can establish certain relationships. I n this chapter we will mostly consider the application of alkaloids to systematic problems. Other secondary compound data can prove equally usable and should also be considered in a complete study of the relationship of systematics and secondary compounds. I have necessarily addressed those problems for which alkaloid data appear to be most helpful or promising and have not pursued certain relationships that may be more clearly established by other chemical and morphological data. I n this section I will survey some of the fundamental and widespread pathways of alkaloid biosynthesis. Studies of many of these compounds have proven useful a t lower taxonomic ranks but, due to the widespread appearance and presumably simple biosynthetic origin, are not as valuable for delineating the higher categorical levels, although in a few cases compounds that appear to be very simple are observed to have limited distributions. The simplest alkaloids are several amines derived from common amino acids such as phenylalanine, tyrosine, histidine, tryptophan, lysine, ornithine, and anthranilic acid. Alkaloids containing simple aromatic moieties and some of their simply derived relations have been reviewed (40-46). These simple amines arise by decarboxylation of the corresponding amino acids, often with subsequent methylation, hydroxylation, and addition of other groups. They are widely distributed, and their presence is usually not of taxonomic significance at the higher taxonomic ranks. These compounds are important because they are frequently beginning points for the synthesis of more complex alkaloids. Phenylalanine gives rise to phenylethylamine (6) and the corresponding methylated compound (7),while tyrosine produces the corresponding compounds tyramine (8) and N-methyltyrosine (9). I n the Gramineae tyramine is converted to hordenine (lo),which is widespread

1.

15

PLANT SYSTEMATICS

7

6

8

in 1 is family ,ut not restricted to it. Tyrosine is also converteL to two other important intermediate compounds, dihydroxyphenylalanine (DOPA) (11) and its cyclic derivative, cycloDOPA (12). These compounds are especially common as intermediates in the synthesis of alkaloids of the benzylisoquinoline and betalaine types as well as alkaloids widely distributed in the Cactaceae (47, 48) (see Section V, B). I n the Rutaceae many of these simple aromatic compounds are converted to the corresponding amides, such as fagaramide (13) from

11

10

13

12

Fagara xanthoxyloides Lam. Although most gymnosperms do not contain distinctive alkaloids (with the notable exception of the Taxaceae and Cephalotaxaceae), the genus Ephedra (Ephedraceae), a group only distantly related to more common gymnosperms, contains methylated phenylethylamines such as 1-ephedrine (14) and d-pseudoephedrine (15), which are also characteristic of this group of plants but not restricted to it (49-52). CH3

I

HCNHCH,

I

HCOH

8 14

CH3

I

HC-NHCH,

I

HO-C-H

0 15

16

DAVID S. SEIGLER

The simple aliphatic compounds putrescine and cadaverine, derived from ornithine and lysine, respectively, are intermediates in the synthesis of many major groups of alkaloids and presumably occur in many plant groups but are seldom isolated and studied. Ornithine (or its successor N-methylputresine) gives rise to N-methylpyrrolidine via the reactions below (53). CHa-NH,

I

CHaNHCH3

I

CHa

CHa

CHa

+ CHS

I I

CHNH,

I

COaH

I I CHNH, I

CHaNHCH3

-con

I

CHa

I

CH,

__f

I

CH,-NH,

COaH

CHaNHCH3

A similar reaction series can produce the corresponding piperidine homolog from lysine. These compounds are easily alkylated by a number of compounds, for example, p-ketobutyric acid, to produce simple alkaloids such as hygrine (16) of the pyrrolidine type (43-55). I n a similar manner attack on an N-methylpiperidium cation yields

16

N-methylisopelletierine, an intermediate in the formation of characteristic alkaloids in the Punicaceae, Lythraceae, and Lycopodiaceae. Simple pyrrolidine and piperidine alkaloids are widespread among higher plants. Both groups may serve as substrates for additional alkylation reactions either internally to yield alkaloids such as tropine (17) and pseudopelletierine (18) or intermolecularly to yield more complex alkaloids. Pyrrolidine alkaloids are widespread, no doubt a reflection of the relatively small number of biosynthetic steps and chemical probability of their synthesis, but they are characteristically proliferated in a few families, such as the Solanaceae and Erythroxy-

1.

17

PLANT SYSTEMATICS

laceae and less commonly in others such as the Euphorbiaceae and Convolvulaceae and doubtfully in the Dioscoreaceae (49-52, 56, 57). Alkaloids of the piperidine type are more widely distributed. Many simply derived ones are found in the Crassulaceae, Punicaceae, and the Leguminosae, but they are also found in the Pinaceae, Euphorbiaceae, Chenopodiaceae, Equisetaceae, Piperaceae, Caricaceae, and Palmae.

17

18

Alkylation by phenylpyruvic acid may occur to produce other alkaloids characteristic of the Crassulaceae, such as sedamine (19) (53) and lobeline (20), found in the genus Lobelia of the Campanulaceae. Nicotinic acid may also alkylate the pyrrolidinium cation to produce compounds such as nicotine (21), one of the most widely distributed of all alkaloids (43, 50, 58). Many related compounds are found in the Solanaceae, especially in the genus Nicotiana. Anabasine (5) arises in Nicotiana by alkylation of the lysine-derived piperidinium cation. Coniine (22), the principal alkaloid of Conium (Umbelliferae), closely

20

0

22

18

DAVID S . SEIGLER

resembles intermediates in the synthesis of the isopelletierine alkaloids but has been demonstrated to be derived via a polyketide pathway (53, 59) from acetate precursors. This is a clear example of convergence in the types of compounds produced and it demonstrates why a knowledge of biosynthetic pathways is valuable in studies of phylogeny. Coniine has been reported from several other families (50).It would be especially interesting to determine the path of synthesis in each of these. Simple derivatives of tryptophan are also widely distributed in nature. Some, such as serotonin (23) and bufotenine (24), involve subsequent oxygenation. N,N-Dimethyltryptamine (25) and psilocybin (26) are widely known for their hallucinogenic properties. These compounds are more restricted in distribution than 23 and 24; 25 is 7H3

24

23

0-

I

HO-P=O

25

26

found in several families (50-52), but 26 appears t o be limited to fungi. Tryptamine and its derivatives serve as intermediates for many groups of alkaloids and by inference must occur in numerous plant taxa. Another group derived from tryptamine is the /?-carboline alkaloids,

--Q-,2&

Q - + . L O Z H ' N

N

H

H

OTJ

CH30 \

/N

H

H 27

1.

19

PLANT SYSTEMATICS

which occur in many plant families such as the Passifloraceae, Symplocaceae, Zygophyllaceae, Eleagnaceae, Malphigiaceae, Euphorbiaceae, and Loganiaceae. Many families which contain alkaloids of the /3-carboline type are otherwise devoid of alkaloids. Histamine (28) is widespread in higher plants, but only a few alkaloids derived from the parent amino acid histidine, such as pilocarpine (29)) are known otherwise. Alkaloids of this type are mostly restricted to the Rutaceae (Casimiroa and Pilocarpus) and certain groups of fungi.

28

29

Dimerization of intermediate compounds from ornithine and subsequent cyclization can lead to the basic skeleton of the pyrrolizidine alkaloids (53). Further elaboration of basic pyrrolizidine structures Ornithine + putrescino

HCO'

involves the type of oxidative process noted previously in relation to the biosynthesis of pyrrolidine and piperidine alkaloids. Pyrrolizidine alkaloids are usually esterified with mono or dibasic acids, many of which are unique to this series, e.g., heliosupine (30) and senecionine (31)(49-52, 60-64). Alkaloids of this type are found in several families CH3

H

H3C'foH HO--CCHOH--CHB

I c=o I

30

31

20

DAVID S. SEIGLER

b u t are characteristic of the Boraginaceae (several genera), the Compositae (tribe Senecioneae), and the Leguminosae (Crotalaria)(49-52, 60-64). Similar reactions with cadaverine, derived from lysine, produce lupin alkaloids such as lupinine (3). I n this instance the corresponding aldehyde may condense with another molecule of piperidine t o yield more complex compounds such as lamprobine (32),sparteine (Z), and matrine (33).Alkaloids of this type are best known from certain genera of the Leguminosae (28, 49-52, 65).

32

33

I n this section several fundamental pathways of alkaloids biosynthesis have been examined. We will make frequent reference t o these in the subsequent examination of a number of specific taxonomic problems because all have been observed to occur in many higher taxonomic groups.

IV. Alkaloids in Lower Vascular Plants and Gymnosperms Alkaloids are rarely found in lower plant groups. Algae, bryophytes, and ferns seldom contain compounds of this type. Among the lower vascular plants there are two notable exceptions; one is the genus Lycopodium, which contains complex alkaloids such as lycopodine (34) derived from lysine by means of precursors similar to those involved in the formation of pelletierine alkaloids in the Punicaceae (49-52, 66-69). The other exception is the genus Equisetum, which contains several alkaloids, such as palustrine (35). Nicotine (21) is also reported from Equisetum species. Although alkaloids are relatively uncommon among gymnosperms, simple compounds such as pinidine (36) are found in the Pinaceae and closely related families. The biosynthesis of compounds of this type has been previously outlined (Section 111, B). The Taxaceae (Taxales) (70) and Cephalotaxaceae (Cephalotaxales) (72, 72, 72a) contain alkaloids such as taxine (37),which is possibly

1.

PLANT SYSTEMATICS

21

34

of diterpine origin, and deoxyharringtonine (38),which are restricted to their respective families (and orders). The homoerythrina alkaloids of the Cephalotaxaceae are otherwise known only from the families Aquifoliaceae and Liliaceae (73, 7 4 ) . Both groups of alkaloids have antitumor activity and are extremely toxic.

nu

0

6H

1

31

OCH,

R = CH

CH-CHa-CH2C(OH)4H2COpMe

I co;

3 - ~

CH3 38

The presence of complex alkaloids in the Taxaceae and Cephalotaxaceae supports the separation of these orders from other gymnosperms. This separation has been suggested by several workers on both paleobotanical and morphological grounds (75-77). Although the fungi represent a distinct evolutionary line and are

22

DAVID S . SEIGLER

probably as distant from plants as they are from animals in evolutionary terms ( I ) , they do possess several interesting types of alkaloids. Many ofthese compounds, such as psilocybin (26), which is found mostly in the genera Psilocybe and Stropharia, are derived from simple amines which are also widespread in higher plants. Muscarine (40) is a hallucinogenic choline analog found in the fly mushroom, Amanita muscaria. Others, such as gliotoxin (39) from Trichoderma viride, are more

CH,OH 39

40

complex in structure. Many nitrogen-containing compounds from Fungi imperfecti, especially the genera Penicillium, Streptomyces, and Aspergillus have pronounced antibiotic activity; these have been reviewed elsewhere (49, 50, 78-80). Indole alkaloids of the ergot type are found in Claviceps and also in t'he angiospermous plant family Convolvulaceae (Section V, B).

V. Alkaloids in the Angiosperms A. INTRODUCTION Among the Angiosperms (flowering plants), Cronquist recognizes six subclasses of dicotyledonous and four subclasses of monocotyledonous plants ( 6 ) .Alkaloids are scarcely known from some of these, whereas in others they are common. Among the subclasses of Magnoliopsida (dicots)the Hamamelidae and Dilleniidae have few alkaloids-primarily simple bases and 8-carboline types that occur in many plant groups. Benzylisoquinoline alkaloids are characteristic of many orders of the subclasses Magnoliidae, although some tryptophan-derived bases are found in a small number of families which do not contain alkaloids of the benzylisoquinoline type. Diterpene alkaloids are found in several genera of the Ranunculaceae. The Caryophyllidae contain alkaloids derived from tyrosine and the corresponding dihydroxyphenylalanine (DOPA). Both simple types

1. PLANT SYSTEMATICS

23

and betalain pigments occur and their presence is characteristic of many families of the order. The situation is more complex in the subclass Rosidae, where families of some orders synthesize alkaloids and others do not. Those that produce significant numbers and types of alkaloids are the Rosales (Leguminosae and Crassulaceae), Myrtales (Lythraceae, Punicaceae), Proteales (Eleagnaceae), Cornales (Garryaceae, Alangiaceae), Euphorbiales (Buxaceae, Euphorbiaceae, Daphniphyllaceae, and Pandaceae), Celastrales (Celastraceae), Rhamnales (Rhamnaceae), Sapindales (Rutacae and Peganum of the Zygophyllaceae), Linales (Erythroxylaceae), and Umbellales (Conium of the Umbelliferae). There is little unity among the types of alkaloids produced by this group of plants. The extremely large and diverse family Leguminosae produces many types of alkaloids, among them are pyrrolizidine (Crotalaria), physostigmine (Physostigma), quinolizidine (several genera), Erythrina types (Erythrina),and Ormosia types (Ormosia). The Lythraceae produce an interesting type of quinolizidine alkaloids not known from other plants; the Punicaceae produce alkaloids similar to the better known tropane types; and the Garryaceae produce diterpene alkaloids, otherwise found principally in the Ranunculaceae. The Buxaceae contain alkaloids derived from triterpene skeletons. Euphorbiaceae is an extremely diverse family in terms of alkaloid types; in this regard, it is only rivalled by the Leguminosae and Rutaceae. Benzylisoquinoline, indole( ?), emetine( ? ), securinine, nicotine, polypeptide, Alchornea alkaloids, tropane, p-carboline, and simple bases are all known to occur within the family. The Daphniphyllaceae contain diterpene alkaloids of a unique type only known from this small family. The Pandaceae, Rhamnaceae, and Celastraceae contain alkaloids with attached polypeptide units. In the subclass Asteridae, many orders produce alkaloids. Among these are the Gentianales, Polemoniales (Solanaceae and Convolvulaceae), Lamiales (Boraginaceae), Campanulales (Campanulaceae), Rubiales (Rubiaceae), and Asterales (Compositae). The Gentianales and Rubiales are noted for prolific production of indole alkaloids and less for others of the tylophorine, monoterpene, and quinine type. The Solanaceae are known for the production of steroidal, tropane, and nicotine types, whereas a related family, the Convolvulaceae, produces both tropane and ergot alkaloids. The Boraginaceae and the tribe Senecioneae of the Compositae and Crotalaria, a genus of legumes, produce highly toxic alkaloids of the pyrrolizidine type. The genus Lobelia of the Campanulaceae synthesizes alkaloids of an unusual type restricted to that genus.

24

DAVID S. SEIGLER

B. THE MAGNOLIOPSIDA (DICOTYLEDONOUS PLANTS) 1. Introduction

The presence and phylogenetic significance of more advanced alkaloid groups in the various subclasses and orders of dicotyledonous plants (Magnoliopsida, sensu Cronquist) will now be examined. As the simple alkaloids previously discussed (Section 111, B) are of lesser significance from a systematic view, their presence will only be mentioned when appropriate, and numerous records of these compounds, which may be useful a t the lower categorical levels, will be omitted. The Caryophyllidae are probably the most primitive group and will be examined first, followed by the Magnoliidae and Rutaceae. The Hamamelidae, which do not contain alkaloids of complex structure, are omitted, as are all families of the Rosidae except for the few that contain alkaloids, i.e., the Leguminosae, Euphorbiaceae, Daphniphyllaceae, and Erythroxylaceae. Following this, a number of alkaloid types based on terpenoid structures will be examined. Most of these occur in families of the Asteridae, the most advanced subclass according to Cronquist, although some orders, such as the Cornales (sensu Cronquist), and a number of families of the Rosales possess the same iridoid compounds and certain of their alkaloidal derivatives. Members of the Nympheaceae (Magnoliidae, Sensu Cronquist) have sesquiterpene type alkaloids. The Garryaceae (Cornales, subclass Rosidae) and the genera Delphinum and Aconitum (Ranunculales, subclass Magnoliidae) as well as a few other isolated groups contain alkaloids based on a diterpene structure. The Apocynaceae (Holarrhena), the Buxaceae (Euphorbiales, subclass Rosidae), the Solanaceae, and many Liliaceous plants (of the Liliopsida) contain alkaloids based on steroidal and triterpenoid structures. Alkaloids based on tryptophan and monoterpene-iridoid structures and their distribution mostly in the families Apocynaceae, Loganiaceae, and Rubiaceae (all subclass Asteridae) will be reviewed. The relationship of alkaloid chemistry and systematics in several families of the Asteridae is then examined, e.g., the Solanaceae and the Convolvulaceae. The distribution of ergot alkaloids in the latter family and the fungal genus Claviceps is discussed. 2. The Caryophyllidae

The subclass Caryophyllidae is recognized by Cronquist as having 4 orders, 14 families, and about 11,000 species. Of these orders, the

Polygonales, Plumbaginales, and Batales are largely without alkaloids

1.

25

PLANT SYSTEMATICS

although harman, tetrahydroharman, and harmanine have been reported from a species of Calligonum of the Polygonaceae (50). I n contrast, alkaloids are widespread in most families of the Caryophyllales. They have been reported from the Aizoaceae (2500 species), Amaranthaceae (900 species), Basellaceae (20 species), Cactaceae (2000 species), Chenopodiaceae (1500 species), Didieraceae ( 9 species), Nyctaginaceae (300 species), Phytolaccaceae (150 species), and Portulaceae (500 species), but not from Caryophyllaceae (2000 species) and Molluginaceae (100 species). Because of the considerable controversy concerning the relationship of chemistry to the classification of this order, it has been studied more extensively than many others. Saponins are widely distributed through the order. They have been reported from the Aizoaceae, Molluginaceae, Amaranthaceae, Basellaceae, Cactaceae, Caryophyllaceae, Nyctaginaceae, and Phytolaccaceae. Many of these are based on triterpene aglycone skeletons (78, 81). Some species of the Chenopodiaceae contain a number of simple alkaloids derived from phenylalanine, tyrosine, tryptophane, ornithine, and lysine. Alkaloids derived from tyrosine are of particular interest because they are related to both benzylisoquinoline alkaloid precursors and precursors of the betalain pigments which are widespread in the order (37, 4 4 , 5 8 ) .Salsolin (41) is an example of an alkaloid of this type. Several relatively simple piperidine derivatives are found, as well as the '

41

alkaloid anabasine (5), which in this instance is structurally but not biosyntheticalIy related to nicotine. Lupinine (3) and other quinolizidine alkaloids are found in Anabasis aphylla. Alkaloids with structures similar to those derived from tyrosine above are widely distributed in Caetaceae (43, 49-52, 78, 81). One of these, mescaline (42), is widely known for its hallucinogenic properties. Others such as anhalidine (43) and anhalonidine (44) show similarity to

OCH, 42

OH 43

44

26

DAVID S. SEIGLER

certain precursors of benzylisoquinolinealkaloids. Other, more complex, alkaloids involving mevalonate units such as lophocerine (45) and dimerization of simple alkaloid units occur.

45

The genus Mesembryanthemum and related genera of the Aizoaceae contain alkaloids such as mesembrine (46), which are also derived from tyrosine (82).

CH, 46

The most widespread alkaloids of the order, however, are betalain pigments derived from L-DOPA (83).These red or yellow compounds have ultraviolet absorptions in the same ranges as anthocyanins and probably serve much the same function in plants of the Caryophyllales. The occurrence of the two classes of compounds is mutually exclusive; no known plant in a betalain-containing family has ever been shown to contain anthocyanins and vice versa (26, 83-87). The families Caryophyllaceae and Molluginaceae contain anthocyanins, a fact that has been used to suggest that they should be segregated into a closelyrelated but distinct order (87). The red-violet pigment of beets is betanin (47) whereas the related yellow pigment from the cactus

HO

/

47

$

C0.H

48

27

1. PLANT SYSTEMATICS

SCHEME 1

Opuntia ficus-indica Mill. is indicaxanthin (48). The first of these compounds arises via Scheme 1. Once formed, betanin may be converted t o other compounds via routes similar to those shown in Scheme 2. Based on both chemical and morphological evidence, Mabry considers that the " Centrospermae families " (the Caryophyllales without the Caryophyllaceae and Molluginaceae) were derived from a common ancestral line from some precursor of the angiosperms and that this major

48

SCHEME 2

28

DAVID S. SEIOLER

evolutionary line gives rise to two lines, one anthocyanin containing, the other betalain containing (87').The early evolutionary divergence of the Caryophyllales and Polygonales from other angiospermous lines is supported by protein sequencing data of Boulter (88).The similarity of cytochrome c amino acid sequences suggests that the Polygonaceae (Polygonales) and the Caryophyllales are more closely related to each other than either is to other plants that have been sequenced. The postulated early origin of the Centrospermae is also in accord with studies based on both morphological and chemical features by other workers (78, 89-92) but does not agree with the origin of this group as postulated by Cronquist ( 6 ) ,who suggests that it is derived from the Magnoliidae. Both this data and benzylisoquinoline alkaloid data suggest that the Magnoliidae are not ancestral to the other subclasses of Angiosperms, with the exception of the Rutaceae and a few other families. 3. The Magnoliidae

The subclass Magnoliidae as defined by Cronquist consists of 6 orders, 36 families, and more than 11,000 species, and in his view, they are the most primitive of the angiosperms (flowering plants), evolutionarily speaking. The Aristolochiales and Papaverales have not been included with the other four orders by many workers [see Becker's comparison of taxonomic systems in Radford et al. ( 5 ) , p. 6171 but were included by both Takhtajan (69) and Cronquist ( 6 ) principally on the basis of morphological characters. Before discussing the alkaloids and systematics of this large group, it will be helpful to consider major morphological features that separate the orders of the subclass as well as their major chemical constituents. The Magnoliales are all woody plants that possess specialized cells that contain essential oils. These oils are primarily of terpenoid and phenylpropanoid origin. The nature of numerous chemical constituents of the Magnoliales as well as other orders of the Magnoliidae have been reviewed (78, 81). Several families have scarcely been examined, and

LslERiDAE ROSlDLE

CARlOPHlLLlDlE

YAGNOLl IDLE

FIG.

1 . Subclasses of Magnoliopsida according to Cronquist (6).

1.

PLANT SYSTEMATICS

29

little can be said of the value of chemical characters for establishing their taxonomic position. Among these are the Amborellaceae (1 species), Austrobaileyaceae (2 species), Canellaceae ( 16-20 species), Degneriaceae ( 1 species), Schisandraceae (47 species), Trimeniaceae (7-1 5 species), and Winteraceae (95-120 species). When one compares the numbers of species in the remaining families, it is evident that a t least several species of the larger families have been examinedAnnonaceae (2100 species), Calycanthaceae ( 9 species), Eupomatiaceae (2 species), Hernandiaceae (50-65 species), Himantandraceae (2-3 species), Illiciaceae (42 species), Lauraceae (2000-2500 species), Magnoliaceae (215-230 species), and Monimiaceae (450 species). Members of the orders Piperales and Aristolochiales also have specialized oil cells, but in contrast to the Magnoliales are mostly herbaceous plants. The families of the small order Piperales, the Saururaceae (5-7 species), Piperaceae (1490-3000 species) (Cronquist accepts about 1500), and the Chloranthaceae (65-70 species) are generally low in alkaloid content but rich in compounds derived from phenylalanine or tyrosine metabolism via cinnamic acid and its relatives. The Aristolochiales, which consist of one family, the Aristolocbiaceae (600 species), are rich in compounds derived from the metabolism of cinnamic acid, p-coumaric acid, and their relatives but also contain many alkaloids. The Nympheales are aquatic plants that do not possess the oil glands typical of the three previously described orders. Some workers have considered the Nelumbonaceae to be sufficiently distinct so as to comprise a separate order, usually called the Nelumbonales ( 6 ) . Cronquist separates the Nelumbonaceae ( 2 species) from the Nympheaceae (65-93 species) (but retains both in his order Nympheales), largely on a basis of morphological characters, and the chemistry of these two groups has not been investigated with the exception of their alkaloids. The Ceratophyllaceae (4-1 0 species) has been little studied chemically. The Ranunculales also lack ethereal oil glands and most species of the order belong to three large families-the Ranunculaceae, Berberidaceae, and Menispermaceae. I n morphological features they are generally more advanced than the Magnoliales and are probably derived from them ( 6 ) . Chemical constituents from the three large families Ranunculaceae (800-2000 species), Berberidaceae (600-650 species), and Menispermaceae (350-425 species) have been studied extensively, but the remaining families of the order have been little examined. These are the

30

DAVID 5. SEIGLER

Circaeasteraceae ( 1 species), Lardizabalaceae (30-35 species), Coriariaceae (10-1 5 species), Corynocarpaceae (4 species), and Sabiaceae (90-1 60 species). The Papaverales consist of two families, the Papaveraceae and the Fumariaceae, which are advanced in many respects within the Magnoliidae. Cronquist considered the two families to be parallel groups that show different individual specializations a t least partly because of the absence of the latex system, which is well developed in the former family but missing in the later. These two medium-sized families have about 600 species ( 6 ) . Plants in these families excel in their ability to synthesize alkaloids of various types, but other constituents of the two families have not been examined to any great extent. Despite the widespread occurrence of compounds derived from phenylpropanoid metabolism and the almost ubiquitous presence of sizable quantities of terpenes within plants of the subclass, the presence of alkaloids derived from tyrosine and phenylalanine, namely those of the benzylisoquinoline type, more clearly defines the subclass. The general pathways leading to these benzylisoquinoline alkaloids have been reviewed (53, 93-98). This system arises from tyrosine (or phenylalanine?) in plants of the Magnoliidae by condensation of 3,4dihydroxyphenylethylamine and 3,4-dihydroxyphenylpyruvicacid and a subsequent Mannich condensation to yield norlaudanosoline (49) as the primary condensation product. This compound is subsequently methylated and desaturated to produce papaverine (50) in the opium poppy, Papaver somniferum (53, 93, 94). Methylation appears to occur after formation of the tetrahydrobenzylisoquinoline system but before dehydrogenation to papaverine. Norlaudanosine occurs with papaverine and also serves as an efficient precursor for its formation (53). Simple benzylisoquinoline alkaloids are known to occur in the Annonaceae, Hernandiaceae, Lauraceae, Magnoliaceae, Menispermaceae, Monimiaceae, Papaveraceae, Euphorbiaceae, Rhamnaceae, and Rutaceae (49-52). d-Reticuline (51), which is known to serve as an HO

HO HO

HO

CH30 49

CH30 60

51

31

1. PLANT SYSTEMATICS

intermediate in the biosynthesis of several more highly modified series of compounds is widely distributed and is known to occur in the Anonaceae, Hernandiaceae, Lauraceae, Monimiaceae, and Papaveraceae as well as the non-Magnoliidean family Rhamnaceae (49-52). Aporphine alkaloids [e.g., glaucine (53)and bulbocapnine (54)] have essentially the same distribution as simple benzylisoquinoline types (49-52) and arise by ortho-para coupling of compounds such as laudanosoline (52) (53, 94, 99-101) or where ortho-para coupling is not possible via the intermediacy of proaporphine compounds such as orientalinone (55) in the biosynthesis of isothebaine (56) in Papaver orientale L. (53,93,102).Aporphine alkaloids are known to occur in the CH,O

CH3

CH,O

CH,O

HO OCH, 53

OH 51

54

cH30 HO

56

32

DAVID S. SEIGLER

Berberidaceae, Ranunculaceae, Fumariaceae, Aristolochiaceae, Magnoliaceae, Lauraceae, Hernandiaceae, Monimiaceae, Menispermaceae, Nelumbonaceae, Papaveraceae, Symplocaceae, Euphorbiaceae, Rutaceae, and the Rhamnaceae. Morphine alkaloids, such as morphine (57), also arise by ortho-para coupling of compounds such as 1-reticuline (58) in the family Papaveraeeae (53,93,94,103-108).Certain intermediates in this pathway occur in other families, for example, salutaridine (59) in Croton salutaris Casar of the Euphorbiaceae.

OH 58

57

I n Cryptocarya bowiei (Hook.) Druce, an Australian member of the family Lauraceae, benzylisoquinoline precursors yield compounds with closure to the isoquinoline nitrogen such as cryptaustoline (60) (53,109). In the family Papaveraceae, various species of the genera Argemone and Eschscholtzia synthesize alkaloids from benzylisoquinoline pre-

HO

60

0 59

33

1. PLANT SYSTEMATICS

cursors with another type of closure. Representatives of these are Z-eschscholtzine (61) and Z-munitagine (62) (53, 93, 94, 96, 110). I n the closely related Fumariaceae, closure occurs to include an oxygen atom ring of cularine (63) (48, 93, 94, 103).

?H

62

61

,

OCH, 63

The genus Cocculus of the Menispermaceae synthesizes alkaloids of the Erythrina type. Alkaloids of this type are known t o arise in the genus Erythrina (Leguminosae) by complex rearrangements of benzylisoquinoline alkaloids such as N-norprotosinomenine (53, 93, 94, 111115). The N-methyl carbon atom of several benzylisoquinoline alkaloids is known to participate in formation of a " berberine bridge " in compounds such as berberine (64)(116,117).Although protoberberine alkaloids are known to occur in several families (Anonaceae, Ranunculaceae?, Aristolochiaceae, Magnoliaceae, and Menispermaceae), they are characteristic of the genus Berberis (Berberidaceae) and of the genera Corydalis and Dicentra of the Fumariaceae (49-52). Stylopine (65)in the

34

DAVID S. SEIGLER

65

66

latter two genera is converted to protopine (66)(118).The benzophenanthridine skeleton encountered in a number of alkaloids of the Papaveraceae is also derived from benzylisoquinoline precursors (48, 93, 94). Chelidonine (67)is an example of this type of alkaloid. Phthalideisoquinoline alkaloids, e.g., narcotine (68), are also found in the Papaveraceae and Fumariaceae with occasional occurrences in the Berberidaceae and Ranunculaceae (49, 53, 93, 94, 119). Coupling of benzylisoquinoline units occurs in an intermolecular as well as in an intramolecular fashion (53,93,94,120,121).The individual components are usually linked by one or two diphenyl ether bridges.

<SO) (:% OH

/

’

o

OCH,

0 67

OCH,

68

The distribution of compounds of this type is essentially the same as for the simple benzylisoquinoline units and aporphine alkaloids; they are found in the Menispermaceae, Lauraceae, Magnoliaceae, Monimiaceae, Hernandiaceae, Nelumbonaceae, Aristolochiaceae, and Ranunculaceae, with a questionable record from the Buxaceae (49-52). Aristolochic acid (69) occurs in the Aristolochiaceae and is often accompanied by aporphine alkaloids. Feeding studies have demonstrated that this naturally occurring nitro compound is probably derived from orientalinol (70) (94). Further, noradrenaline is incorporated into aristolochic acid with good incorporation rates, suggesting that 4-hydroxynorlaudanosoline is a precursor and that the 4-hydroxyl group is required for oxidation of the heterocyclic ring.

1. PLANT SYSTEMATICS

70

35

69

Many botanists agree that the orders of the Magnoliidae according to Cronquist are related and derived from common ancestors. This conclusion is largely based on morphological evidence, and chemical evidence i s considered supplemental, although in the subclass only the order Piperales and the order Nympheales (if one removes the Nelumbonaceae) lack either the simple benzylisoquinoline alkaloids or their more highly evolved derivatives. The Piperales are closely linked to other orders by the presence of many phenylpropanoid and terpenoid compounds as well as morphological features. The Nelumbonaceae are linked by the presence of benzylisoquinoline alkaloids to other orders of the subclass, but the other families of this order, especially the Nympheaceae, do not possess compounds of this type but rather alkaloids with a sesquiterpene skeleton. Because of the presence of ellagic acid and the absence of benzylisoquinoline alkaloids, Bate-Smith believes that the family Nymphaeaceae is completely out of place in this subclass (122),a view shared by some other workers (89-91). Pathways leading to benzylisoquinoline alkaloids are found in many (but not all) families of the remaining orders. Within these orders the presence of these types of alkaloids is observed because the plants that contain them descended from common ancestors and not because the pathways have evolved numerous times. The families Magnoliaceae, Annonaceae, Eupomatiaceae, Monimiaceae, Lauraceae, and Hernandiaceae of the Magnoliales contain benzylisoquinoline alkaloids. The families Himantandraceae, Myristicaceae, and Calycanthaceae contain alkaloids of other types, 71,26, and 72, respectively, and do not contain benzylisoquinoline alkaloids. At least one species of the Winteraceae contains alkaloids of an undetertermined type (123),whereas species of the Degeneraceae, Austrobaileyaceae, and Trimeniaceae have been tested and found not to contain alkaloids (78, 1234. The Lactoridaceae, Canellaceae, Illiciaceae, Schisandraceae, Amborrelaceae, and Gomortegaceae have apparently not been tested. The families Ranunculaceae, Berberidaceae, and

36

DAVID S. SEIQLER

Menispermaceae contain benzylisoquinoline alkaloids, while members of the Lardizabalaceae (123u, 123b), Corynocarpaceae (123a), and the Coriariaceae (123~-123c)have been tested and found not to contain alkaloids. The Sabiaceae and Circaeasteraceae have apparently not been tested. The families Aristolochiaceae (Aristolochiales)and the Papaveraceae and Fumariaceae (Papaverales) all contain benzylisoquinoline alkaloids as previously mentioned.

8"

HC



NO2

using certain electrophilic reagents (2). 5-Acetyl-deoxynupharidine was transformed to the 3-hydroxy-2-methylpyidylderivative (4) on heating with aqueous ammonia and ammonium chloride (2). Me I

Me 4

Polonovski transformation of ( + )-nupharidine carried out in a large excess of acetic anhydride resulted in A6-enamine( 5 ) (3).Hydrogenation of 5 resulted in ( - )deoxynupharidine and ( - )-7-epideoxynupharidine

3.

183

N U P H A R ALKALOIDS

Me

5

in a 7:1 ratio. Enamine 5 was transformed in two steps to (-)nupharamine (7)with 59y0overall yield ( 3 )(Eq. 1). Oxidation of 5 with osmium tetroxide-paraperiodic acid in pyridine-water-dioxane solution results in the formamidoketone (6)with 95y0yield. Thelatter compound, when refluxed in ether with large excess of methylmagnesium iodide yielded 7.

080

(1)

5----5-

a,lr ( - ) Nupharamine

7

6

A mechanistic interpretation of the Polonovski transformation ( 3 ) was attempted. The A6-enamine ( 5 ) was converted to deoxynupharidine-6p,7/3-d2(8)by catalytic addition of deuterium. The stereochemistry of the deut'erium atoms in 8 was based on the preferred cis catalytic hydrogenation of the a side of 5 and on NMR spectra. The C-6,equatorial hydrogen quartet (T = 7.30) of 3 appears as a singlet in 8, and the C-6g axial hydrogen quartet (T = 8.12) of 3 is absent in 8. From 8, a derivative corresponding t o 1 was prepared. The latter, treated under the Polonovski conditions, resulted in the corresponding A6-enamine (9) (Eq. 2 ) . The mass and NMR spectrometric studies Me

Me

8

9

184

JERZY T. W R ~ B E L

demonstrated that the hydrogen atom eliminated in the Polonovski transformation was the 6a-hydrogen. The oxidation of deoxynupharidine to nupharidine was found t o be almost three times faster than the oxidation of 7-epideoxynupharidine. This was explained in terms of oxidation of deoxynupharidine with inversion on nitrogen to give a cis-fused quinolizidine N-oxide (10) (Eq. 3). The cis-fused conformation of nupharidine was confirmed by H I

?

X-ray studies. I n view of the cis ring fusion in 1, the Polonovski transformation was considered to be a trans8 elimination; the mechanism would then involve the steps shown in Eq. 4.

Me

Me

Me

Me

( + )-Nupharidine was transformed to A3-dehydrodeoxynupharidine (11) using a modified Meisenheimer rearrangement ( 4 ) (Eq. 5 ) . Me

(5) l -

11

3.

185

N U P H A R ALKALOIDS

The mechanism was shown to involve the steps shown in Eq. 6.

Me

-

11

F OH

(6)

H

B. ABSOLUTE CONFIGURATION The absolute configuration of ( - )-deoxynupharidine and other C,, alkaloids (5) was questioned first by Turner et al. (6), who ascribed the R-configuration to the ( - )-a-methyladipic acid; the previously proposed absolute configuration of ( - )-deoxynupharidine 12 was predominantly based on the assumption that the ( - )-a-methyladipic acid obtained by oxidation of 3 has the S-configuration. Further work Me

12

13

by LaLonde et al. (7) on the synthesis of (-)-(R)-a-methyladipic acid supported this suggestion. The final proof was supplied by Oda and Koyama (8) in the form of an X-ray analysis. The above results indicate that the absolute configuration of (-)deoxynupharidine is represented by formula 13. This reassessment required a correction of the absolute configuration of other C,, alkaloids, e.g., dehydrodeoxynupharidine (14), nupharamine (15), anhydronupharamine (16), nuphamine (17), and 3-epinuphamine (18). The corrected absolute configurations for the above alkaloids are given by structures 14-18.

186

JERZY T. W R ~ B E L

h..

Me

Me

I

16

15

14

Me

18

C. NEWCOMPOUNDS 1. 7-Epideoxynupharidine (19)

This alkaloid was isolated by LaLonde et al. (9, 10) from Nuphar luteum Sibth. et Sm. subsp. variegatum. The structure was confirmed by IR and NMR spectra and hydrogenation of As-dehydrodeoxynupharidine ( 5 ) ,which produced deoxynupharidine (3)and the 7-epiisomer (19). Me

19

The NMR spectrum of 19 displayed methyl resonance doublets a t T (J = 3 and 5.4 Hz, respectively). I n comparison with NMR data for deoxynupharidine (3),the axial methyl groups with lower field signals and larger splittings and the equatorial methyl groups with higher field signals and smaller splittings can be correlated-a 9.08 and 9.26

3.

187

N U P H A R ALKALOIDS

phenomenon well-known in quinolizidine chemistry (IOU).The absolute configuration of 7-epideoxynupharidine represented by structure 19 follows correlation through 5 with deoxynupharidine (3). 2. Nuphenine (20) and Anhydronupharamine (24)

20

Nuphenine (20) was isolated first by Forrest et al. (11, l l a ) . Its molecular formula was determined as C,,H,,NO (mw = 233). The I R spectrum shows N-H (3310 cm-l), Bohlmann bands (2800 and 2730 cm-l), furan (1505, 880 cm-l); the NMR spectrum indicates the presence of a substituted double bond (multiplet at 4.88 7)Nuphenine can be hydrogenated either to a dihydro compound (21)or to hexahydro derivative (22) (Eq. 7 ) . The 4.88 signal is absent in the 20

22

21

NMR spectrum of 21, and the peak a t 8.3 r (6H,S) in nuphenine is shifted to 8.75 T (6H,d);this, together with the peaks at mle 164 (M-69) in the mass spectrum of 20 and at mle 168 in the spectrum of 22, confirms the presence of the (CH3)2C=CH-CHz(m/e 69) group in 20. Easy loss of this group suggests that it is located in the position alpha to nitrogen in the piperidine ring. Since H, is split by only one ring proton, the methyl group is assumed to be located on the adjacent carbon; the protons H, and H, with a coupling constant of 2.5 Hz must be in an axial-equatorial or equatorial-equatorial relation to one another (12). The presence of bands a t 2800 and 2730 cm-l in the I R spectrum of nuphenine was taken as evidence for t.wo hydrogens axial to the nitrogen atom. The proposed configuration ofnuphenine is shown in 23.

188

JERZY T. W R ~ B E L

~ b - k - ~ e \ /Me He /C=C\Me 23

Isomeric with nuphenine is anhydronupharamine (24) isolated by Arata et al. (13, 14) from Nuphar japonicum DC. It proved t o be identical with the dehydratation product of ( - )-nupharamine (15) and therefore its configuration should be as in 24.

24

3. Nuphamine (17)

17

The chemistry of this alkaloid was further studied and its configuration was related to deoxynupharidine (3) and nupharamine (15). The transformations in Eq. 8 have been effected. On the basis of Eq. 8, nuphamine is thought to have configuration 17. A study of the configuration around the double bond in nuphamine led to the conclusion that in the side chain the methyl group and hydrogen were in the trans position (15).This deduction is based on a general observation that in the X-CH,-C(CH,)=CH,-Y system a trans relationship between the methyl group and the vinyl proton results in a higher r value

3. 17

N U P H A R ALKALOIDS

Na2C03, C H d

189

24

( A T = 0.06-0.07) for the methyl protons than that observed for the cis isomer. Thus, the absolute configuration 27 of nuphamine (17) was established:

4. 3-Epinuphamine (28)

(C,,H2,N02)

The alkaloid was isolated by LaLonde et al. (16) from Nuphar luteum subsp. variegatum and was shown to have configuration 28. Its molecular formula was confirmed by mass spectroscopy. The IR and NMR spectra indicate the presence of a %fury1 group. Attachment of

190

JERZY T. W R ~ B E L

this group to the carbon a to nitrogen (C-6) was concluded from the presence of the proton (3.58 6) deshielded by the fury1 group and the nitrogen. The presence of OH and NH groups was established in the conversion of 28 to an N,O-dibenzoyl derivative. The presence of a

28

trisubstituted double bond was indicated by the I R and NMR spectra; the latter showed a hydroxymethyl group (3.93 6, 2H, broad singlet), a vinyl methyl group (1.65 6, 3H, broad singlet), and a methylene group. The trans stereochemistry of the double bond was based on the character of the vinyl proton signal in the NMR, as it was shown in nuphamine (15).Oxidation of 28 with MnOz resulted in an aldehyde (29), giving additional support to the proposed double bond stereochemistry. The

FYMe Me

29

UV spectrum of this aldehyde was in accord with known trans-2-methyl2-pentanal. a-Attachment of the side chain to nitrogen was consistent with the appearance of an ion at m/e 164 ( l O O ~ o )in the mass spectrum. The NMR spectrum showed the C-2 proton as a triplet of doublets, which could be explained as a coupling to the side chain methylene group and to a single proton. This implied substitution a t C-3 of a methyl group whose presence is indicated by a doublet a t 0.99 6. The substitution pattern in piperidine was determined by converting both the N,O-dibenzoyl derivative (30) and nuphenine benzamide to the aldehyde (32):

0 30 R = CH,OCOCeH,, R’ = CeH5C0 31 R = Me, R’ = CeH,CO

32

3.

191

N U P H A R ALKALOIDS

The presence of an axial methyl group at C-3 is implied by a doublet a t 0.99 6, which is a t a lower field than the resonance (0.91 6) displayed by the equatorial methyl of nuphamine. Other characteristics of NMR spectra are consistent with this assignment. 5 . Nupharolidine (33)(C,,H2,N02)

33

This alkaloid isolated from the rhizome of Nuphar luteum by Wr6bel and Iwanow ( I Y ) , was the first among the C,, alkaloids to be shown to have its hydroxyl group situated in the quinolizidine ring. The suggested structure of this alkaloid was based on spectroscopic correlation (IR, NMR, and mass spectra) with three other C,, basesdeoxynupharidine (3),castoramine (34), and nuphamine (17). The Me

34

R 1= CHaOH, Ra = H

crucial observations pertaining to the structure beside the transquinolizidine and a B-substituted furan ring indicated the presence of two

\ CH-CH, /

groups

(T

=

9.12 and 8.80; doublets),

\CH,-O& /

= 6.35, and 4.75,; IR, 3342 cm-l). The presence of two methyl groups, which appear as two doublets, ruled out the presence of a hydroxymethyl group and eliminated the possibility of C-1 and C-7 being the points of OH substitution. Since a strong signal a t mle 178 (fragment 35) was observed in the mass spectrum the presence of an OH group at C-6 position was also ruled out. (7

192

JERZY T. W R ~ B E L Me

35

The presence of the fragment 35 and of two others at mle 71 and 206 to which structure 36 and 37 were ascribed, respectively, point to C-9 as the location of the hydroxyl group. Thus, nupharolidine is thought to have structure 33.

37

36

m/e 206

m/e 71

6. Nupharolutine (38) (C,,H,,NO,)

Nupharolutine is another C,, alkaloid with a hydroxyl group. It was isolated and its structure was established by the Polish-Canadian group of workers (18).It is isomeric with nupharidine (1) and castoramine. Structure 38 for nupharolutine was based on spectroscopic and chemical data. Me

38

3.

NUPHAR ALKALOIDS

193

The IR spectrum shows the presence of an intermolecularly bonded hydroxyl group and a trans-quinolizidine system. Unsuccessful attempts at acetylation indicate the tertiary character of the hydroxyl. The NMR spectrum of the new alkaloid shows a doublet centered a t 0.92 and a singlet (3H) at 1.21 6. The singlet peak and its chemical shift

I I I I

are compatible with a -C!--C(CH3)OH-C--

I I

grouping in the molecule.

Other signals in the NMR spectrum were in accord with those observed for deoxynupharidine and indicated the presence of a p-substituted furan ring in the equatorial position (C-4-Haxialquartet 3.03 8 , J = 8.3 and 6.0 Hz). The final data for structure 38 were obtained from the mass spectrum. High resolution studies gave the composition of the ions observed, thereby giving further insight into the fragmentation process. The fragmentation is discussed later with that of other Nuphar alkaloids. Nupharolutine was correlated wiih deoxynupharidine (3) as in Eq. 9.

This sequence offers the final proof for the proposed structure and for the absolute configuration of nupharolutine. A dimeric compound related to nupharolutine was isolated by LaLonde et al. (19).Spectroscopic data indicate structure 39. This structure was confirmed by a synthesis beginning with dehydrodeoxynupharidine (14) (Eq. 10). Osmium tetroxide oxidation of 14 yielded diol 40, which wa,s transformed upon dehydration into 39, borohydride reduction of which generated a mixture of 41 and 42. Me

Me

39

194

JERZY T. W R ~ B E L

14

40

NaBH

I 39 b

R

Ri

2

Qr 41 42

Rl = O H , R, 5 H Rl = H RZ = OH

7. Epinupharamine (Epi-15) (C,,H,,NO,)

3-Epinupharamine (epi-15) was isolated by Forrest and Ray who established its structure. Its structure was proved on the basis of its spectra and by its synthesis from nuphenine (20). Mass spectrometry confirmed the molecular formula and the presence of the 3-methyl-3hydroxybutyl side chain (peak a t mle 164). The IR and NMR spectra

Epi - 15

showed the presence of the hydroxyl group (3575, 3150 em-, and T = 7.35) and the furan ring (IR, 1500, 1170, and 875 cm-l; NMR, 2.63 (2H), 3.57 (1H) T ; CH-CH, (ring) 9.03~dand a gem-CH, 8.83 T, 8.75 T). This assignment of the structure and stereochemistry was verified by the conversion of nuphenine (20) into a compound identical with the naturally occurring epi-15.

3.

N U P H A R ALKALOIDS

195

111. Sulfur-ContainingC,, Alkaloids

Thiobinupharidine (43) (C3,H,,N,02S)

"As

43

It was shown earlier (20, 21) that 43 is isomeric with neothiobinupharidine (44) and both 43 and 44 have almost the same characteristic structural pattern (quinolizidine, furan, -S-CH,-, two methyl groups, and similar pK, values). Extensive spectroscopic studies led to deduction of the structure and of the relative configuration of 43. The structure has been firmly established and the absolute configuration has been determined by a study of the crystal structure of thiobinupharidine dihydrobromide dihydrate (22). The structure of thiobinupharidine was established by Wr6bel and MacLean (22)by comparing the IR, NMR, and mass spectra with those previously obtained for neothiobinupharidine (44) (20, 21). The I R and NMR studies (23)of the alkaloid in question, of some model compounds, and of reduction products of biscarbinolamines led LaLonde to the same conclusion. Equimolecular solutions of 43 and 44 examined under the same conditions showed Bohlmann bands of nearly equal intensities. This indicates the presence of two trans-quinolizidine rings in 43. High-resolution mass measurements showed identical compositions of the major ions in the spectra of 43 and 44. The NMR spectra of the two alkaloids have been examined a t 220 MHz, and the anomalies of the earlier studies (20, 21) have been clarified. There is a signal of area 6 centered a t 6 0.91 ( J = 5 Hz) assignable to two CH-m, groups (compare 6 0.85, J = 5.5 Hz for 44 and 6 0.92, J = 5.6 Hz, for 3 as signals for the equatorial methyl groups). Observations concerning the furan proton are in accord with those made earlier (20, 21). I n the region 6 2.7-3.08, complex signals of area 4 appear that are attributed

196

JERZY T. WROBEL

to two protons in the furan ring (at C-4 and C-4') and to the two equatorial protons a t C-6 and C-6'. These assignments are made by analogy with the chemical shifts of the corresponding protons in 3. The spectrum of 43 also contains a well-defined AB pair of doublets centered a t 6 2.32 (J = 11.5 Hz) and attributed to the CH2-S group (compare with a singlet a t 6 2.67, W+ = 3 Hz, in the spectrum of 44). By analogy to the studies on model compounds (24) the absorption of the thiomethylene group suggests an equatorial conformation of the CH2-S with respect to the quinolizidine ring.

0 44

The equatorial linkage of the sulfur atom to the second ring was based on evidence presented by LaLonde (25) for the equatorial character of the C-7-S linkage in thionuphlutine A, which in turn was shown to be identical with thiobinupharidine. All the evidence indicates structure 43 for thiobinupharidine. It has been confirmed by an X-ray crystal structure determination of thiobinupharidine dihydrobromide dihydrate. The observed bond lengths are in good agreement with the accepted values. The only bond that exceeds the average value is that between C-17' and C-7'. The alkaloid has a pseudo-twofold axis. The nonpolar character of the S-containing ring and the inequivalence of S and C-17' destroy this element of symmetry. LaLonde et al. (23)provided further evidence consistent with structure 43. The 100 MHz NMR spectrum of thiobinupharidine determined in benzene shows the two C-4 protons as two overlapping quartets both with splittings of 1.5 and 10 Hz. Such a splitting pattern may be ascribed to an axial (3-4 proton rather than to an equatorial one. Evidence for the stereochemistry of the C-1 and C-1' methyl group comes from the direction of the solvent-induced shift of the C-1 methyl group observed in the NMR spectrum. The C-7 axial methyl group in deoxynupharidine is shifted downfield by 4.2 Hz and the C-1 equatorial methyl is shifted upfield by 5.0 Hz when deuterochloroform is replaced

3.

197

N U P H A R ALKALOIDS

by benzene. The same solvent change results in an upfield shift of 8 Hz for the methyl groups of 43. This demonstrates that both methyl groups in thiobinupharidine are equatorial. Extensive NMR studies allowed LaLonde et al. (23) to assign an equatorial sulfur bonded to the AB quinolizidine system and furthermore to suggest that the sulfur atom is involved in the reduction (NaBH, and NaBD,) of 6- and 6'-dihydroxyl derivatives of thiobinupharidine through a three-membered ring (25) (Eq. 11). S'

A+\

\S

S

+-LA
?/N 'H &

HBr. 130-135°C, 3 hours

/ \

0 \

OH 96

97

98

SCHEME 28

54

R = H R = CH3

~

356 MAURICE SHAMMA A N D VASSIL ST. GEORGIEV

6.

SYNTHESES OF BISBENZYLISOQUINOLINE

357

ALKALOIDS

104 1. HBr, HOAc, 100°C 2. HBr, 140-145°C

t 3. CHaNa

(+-1-95

SCHEME 29

( + )-isotrilobine (45).Finally, taking advantage of the known fact that in dilute acid ( + )-oxyacanthine (100) undergoes isomerization to ( - )-repandine (104),it was found possible to convert ( + )-oxyacanthine into natural ( + )-isotrilobine, Scheme 29 (46). Inubushi and co-workers have recently adapted their synthesis of ( + )-isotetrandrine and ( - )-phaeanthine to preparations of ( + )obaberine and ( -t)-trilobine (46a).

VII. Menisarine-Type Alkaloids The alkaloid ( + )-menisarine possesses the structure 105, which incorporates a diphenylenedioxy bridge, and an interesting synthesis of ( )-N-methyldihydromenisarine (107) has been achieved. The first stage of the synthesis concerned the preparation of the diamine 106, which was carried out via a double Ullmann, as shown in Scheme 30 (47, 48). The lower half (1) of the molecule was prepared using a Willgerodt reaction as per Scheme 31.

105

358

MAURICE SRAMMA AND VASSIL ST. OEORGIEV

cu, pyridine. A

Br

t

OCH,

OH

OH

OCH,

OCH,

106

SCHEME 30

Condensation of the diacid chloride of 1 with the diamide 106 a t high dilution, followed by Bischler-Napieralski ring closure, reduction, and Eschweiler-Clarke N-methylation furnished the desired racemic product 107, Scheme 32 (47, as), which was spectrally identical with the product derived from the reduction and N-methylation of natural ( + )-menisarine (105).

don

1. CH&OCI, AICI., 2. Dlmethyl GS. sulfate

0

II

'

~

0

0

c

~

c

HWillgerodt 3

6. 108

SYNTHESES OF BISBENZYLISOQUINOLINE ALKALOIDS

+ Diacid chloride of 1

359

+

2. NaBH,

107

SCHEME 32

VIII. Tiliacorine-Type Alkaloids ( + )-Tiliacorine and its diastereomer ( + )-tiliacorinine have been assigned structure 108 on the basis of extensive degradative studies ( 4 t h ) .These two alkaloids are unusual in having a biphenyl system in lieu of the usual diary1 linkage. A total synthesis of ( k )-O-methyltiliacorine (109) has been described in detail (49). Unsymmetrical Ullmann condensation of the bromophenols 110 and 111 yielded a mixture of three products from which the desired diester 112 was isolated by chromatography. Homologation and conversion to the diamine 113 was followed by condensation with the diacid chloride 114. The resulting bisamide 115 was converted to a mixture of ( f )-0methyltiliacorine and O-methyltiliacorinine by well established transformations. Careful chromatography of this mixture yielded ( f )-0methyltiliacorine, spectroscopically and chromatographically identical with material derived from the alkaloid. The diastereoisomeric ( f )-0methyltiliacorinine was obtained only in trace amounts, Scheme 33 (49).

360

MAIJRICE SHAMMA A N D VASSIL ST. GEORGIEV 1. K salts formation

C H 3 O O C ~ O C H a ~

'

B r D 2. 3. c Cu-bran=, Chromatography 0 diphenyl 0 ether, c A H

t

\

HO

OH

Br 110

C

111

H

3

0

0

C

vD C O O C H ,

1. LiAlH, 2. 3. S0Clz KCN 4. H.,

o\ 112

113

then,

113

+

COCl

__f

114

115

NYR)

6.

SYNTHESES OF BISBENZYLISOQUINOLINE

ALKALOIDS

361

1. CHJ 2. NaBH, 3. POC1. 4. H2, Pt

5. HCOH, HCOOH

t

108 R = H 109 R = CH,

SCHEME 33

IX. Liensinine-Type Alkaloids The alkaloid ( + )-liensinine (118) incorporates head-to-tail coupling through a diary1 ether linkage. A total synthesis of this alkaloid was achieved on the heels of the initial isolation and characterization reports. Ullmann condensation of ( - )-116 with ( - )-117 followed by hydrolysis gave the optically active alkaloid (50, 51). A synthesis of a diastereomeric mixture of liensinines, by a somewhat similar pathway, is also available ( 5 2 ) . The related alkaloid ( - )-isoliensinine (122)yields ( - )-O,O-dimethylisoliensinine (121)on treatment with diazomethane. Derivative 121 was synthesized by Ullmann condensation of ( - )-119 with ( - )-120 (53). Finally, optically active ( - )-isoliensinine (122) was obtained by the sequence in Eq. 1 ( 5 4 ) .Worthy of attention are the new conditions for the Ullmann condensation ( 5 2 , 5 4 )involving the use of copper powder, potassium carbonate, a small amount of potassium iodide, and dry pyridine heated to 155-160'c in a current of nitrogen. These conditions give better yields (about 15y0)than the usual Ullmann condensation. The newer base ( - )-neferine (123), related t o liensinine and isolienshine, was synthesized by a similar approach (Eq. 2 ) (55).

CH,O

HO

PhCH,O

Y~cH,\ /

\

+

I IBr I / I Y \ C H 3

PhCH,O

\

116

1. Ullmann 2. H,OB

cH30m

117

‘CH,

Hac,5:>: O

b \

F

O

O

I

OCH,

119

120

H

6.

SYNTHESES OF BISBENZYLISOQUINOLINE ALKALOIDS

363

PhCH,O

CH,O

CuO, K.COa. pyridine, A

\CH,

LY

CH30

HO CH30

0R

!

\

C

CH,

H

,

1 OCH,

OCH, 123

X. Curine-Chondocurine-Type Alkaloids

It was conclusively demonstrated in 1970 that the hitherto accepted structure for the alkaloid ( + )-tubocurarine,which had been represented as 124, was in error and that the correct structure is 125 (56, 57). This finding was of particular interest not only because of the importance of ( )-tubocurarine as a neuromuscular blocking agent, but also because of the fact that supposed total syntheses of the racemic di-0-methyl ether of tubocurarine iodide as well as of racemic tubocurarine iodide

+

364

MAURICE SHAMMA AND VASSIL ST. GEORGIEV

X0

c

H

3

0

m

m

,H

$4

I

\

0

124

125

had been claimed previously. A description of the synthetic work on tubocurarine follows. This description is complicated not only because of the above mentioned change in structural assignment, but also by the failure of the workers involved in the synthetic work in clearly differentiating between enantiomers, racemates, and diastereomers while comparing samples (58-62). As a preliminary attempt at the synthesis of the dimethyl ether of tubocurarine, the simple dimer 126 was constructed as described in Scheme 34. The product 126 was obtained as a mixture of two diastereomers from which the predominant racemate (mp 96-99°C) could be isolated (58). Essentially the same approach was utilized in the preparation of the so-called “di-0-methyl ether of tubocurarine iodide ’) (127), Scheme 35 (58, 59). The UV spectrum of one salt so obtained was apparently close t o or identical with the spectrum of an authentic sample of the di-0methyl ether of ( + )-tubocurarine iodide, and this finding was taken as proof of structure.* It must be pointed out, however, that most tetrahydrobenzylisoquinolines, as well as bisbenzylisoquinolines such as tubocurarine or its dimethyl ether, exhibit a maximum absorption near 280 nm, so that UV spectroscopy is not a reliable basis for comparison. Another criterion used was a mixture melting point between the

* There seems to be some confusion in the assignments of melting points of the final products. I n reference ( 5 8 ) , two supposed diastereomeric tubocurarine iodides were obtained (mp 131-135°C and 223-228°C). But in reference ( 5 9 ) , only one melting point was quoted [mp 257-268°C (ethanol)]. This latter material apparently gave no melting point depression with a sample of the natural salt (mp 262-264”C), even though no formal resolution was carried out on the synthetic material.

6.

365

SYNTHESES OF BISBENZYLISOQUINOLINE ALKALOIDS

CH,O

CH,O C

PhCH,O

3' H

HO

___, CH,O

,

O

'

P

"

N,

CH3 C H 3 0

r

~

3

'

+

CH,O

(636 ..jCx; 5'6

CH30\/

G

N

\

C

H

N\CH3

,

1.

Homoveratrylaminr

,

0

2. PO('I3

\ CH,

/ OCH,

CH,

OCH3

I

COOCH, OCH,

SCHEME 34

I . H,, Pt 2. HC'OH, HCOOH

cH30)3? Jy ' FOOH

+

HO

NHa

PhCH,O

PhCH,O

Ly

Cu, KOH, pyridine, 16O-18O0C

2. 209. HCI, A, 2 hours

z

OCH,Ph

OCH, OCH,

1. Zn, dil. HOAc, A, 1.5 hours 2. CHJ.CH30H

I "H 3 c \ : f 0 C H 3

H,C/ OCH,

OCH, 127

SCHEME 35

368 MAURICE SHAMMA A N D VASSIL ST. GEORGIEV

m

e

a 0, m

0

6. SYNTHESES OF BISBENZYLISOQUINOLINE ALKALOIDS

U

4

m c

9

369

w

t

6.

SYNTHESES OF BISBENZYLISOQUINOLINE

ALKALOIDS

371

naturally derived dextrarotatory di-0-methyl ether of tubocurarine iodide and the synthetic isomer, in which apparently no depression was observed. Such a comparison is, of course, invalid since (a) a racemate usually has a different melting point from that of a pure enantiomer, (b) melting points of bisbenzylisoquinoline salts are often unreliable and difficult t o reproduce, and (c) the structure assigned to ( + )-tubocurarine and its di-0-methyl ether was in error in the first place. A synthesis of the unsubstituted tubocurarine analog 129 is also known, Scheme 36 (63). The product proved t o be a mixture of two racemates, mp 225227°C and 121-124°C. As an extension of the synthetic work on the so-called “di-0-methyl ether of tubocurarine,” a preparation of the di-0-methyl ether of racemic chondodendrine (130) was carried out, Scheme 37 (64). A slightly different approach t o the so-called “di-0-methyl ether of tubocurarine” has also been recorded, Scheme 38 (60). The starting material was the diimine 131, which was known from previous work. Each of the two diastereomeric racemates of 132 gave two bismethiodides upon treatment with methyl iodide, a result that is somewhat difficult to rationalize; and one of these four isomeric salts, namely, that melting 257.5-259”c, was claimed to be identical with the dextrorotatory di-0-methyl ether iodide of natural ( + )-tubocurarine iodide. The criteria for comparison were simply closeness of UV spectra and melting points. A claim of a synthesis of a material assumed t o be identical with natural ( + )-tubocurarine iodide was put forward, even though an actual

1. Cu, K,COa,

CH,

2. Zn, HOAc

% Bismethiodide salts

131

133 SCHEME 38

372

MAURICE SHAMMA A N D VASSIL ST. GEORGIEV

CH30 K@ ‘0

, Cu, A

+

OCHaPh CH,O

Ac.0, pyridine

0

----.-+

A

OCHaPh

CHa

OCHaPh

I COOCH, CHa &:H

OCH, 133

C H d , NaOH, CH30H, A

3”’

OCH,Ph

OCH,

OCH,Ph

CH,

HNdo HN ’ Br

OCH,

184

OCH,

135

6.

SYNTHESES OF BISBENZYLISOQUINOLINE

ALKALOIDS

373

then,

IQ

134

H H,C' 3

c

,

IQ

l

H~ OCH,

136

SCHEME 39

separation of optical isomers had not been carried out. This synthesis is further obscured by the fact that two phenols corresponding t o structure 133were asserted t o exist, as well as two of the acetates 134 and two of the diamides 135.The final salt 136 was obtained as two compounds, one melting 257-260.5"C and the other 210-212°C. The former salt was claimed to be identical with (+)-tubocurarine iodide on the basis of UV spectral comparisons and identity of melting points ( !), Scheme 39 (61),even though no separation of enantiomers was performed. 1. A c p O 2. P0Cl3, C H C L A 3. H30@ 4. Cu, K.C03, pyridine, 150-1 80°C 5. Zn , AOAc

$H2

t

OCH,Ph HN&OH

HN&oH

/ OCH,

137

SCHEME 40

OCH,

374

MAURICE SHAMMA AND VASSIL ST. GEORGIEV

Finally, a synthesis of racemic so-called N , N ’-demethylchondodendrine ” (137)) erroneously assumed by the authors t o be identical with chondrofoline, has also been advanced and is described in Scheme 40 (62).Two products were obtained a t the conclusion of the sequence, and one of them was assumed to correspond t o chondrofoline on the basis of UV spectral comparisons and a negative Millon test. It was later shown by other workers that the correct structure for chondrofoline is 138 (65)) so that the claim of a synthesis of chondrofoline is unfounded (62). ((

H

3

c

,

: 0~

3

OCH, 138

I n other attempts a t the synthesis of tubocurarine-type bases, Ullmann condensation of the dibromotetrahydrobenzylisoquinoline 139 with the N-methylcoclaurine salt 140 was investigated but did not lead t o characterizable product (66).Studies of the efficient Ullmann condensation of phenols with aromatic halides substituted a t the ortho position(s)with nitro group(s)have been carried out and have culminated in the preparation of the imide 141 (67-69). 0,O-Dimethylcurine (143) was presumably obtained in the course of the previously described syntheses. But a more reliable preparation of this compound involves the Ullmann condensation between the levorotatory dibromotetrahydrobenzylisoquinoline 139 and the levorotatory diphenolic tetrahydrobenzylisoquinoline 142 (70). When the catalyst for the condensation consisted of cuprous chloride in the presence of potassium carbonate and pyridine and the conditions were heating a t 155-165’C for 24 hours, a small yield of optically active 0,O-dimethylcurine (143) together with a larger amount of 144 was obtained. When, however, the two starting tetrahydrobenzylisoquinolines were racemic rather than levorotatory and the catalyst was cupric oxide in pyridine

6.

SYNTHESES OF BISBENZYLISOQUINOLINE ALKALOIDS

375

141

heated a t 160-170°C for 50 hours, the products consisted of a small yield of a mixture of 0,O-dimethylcurines together with a mixture of tetrandrines and isotetrandrines (54), as well as a mixture of 144. Ullmann condensation of 2 moles of the racemic phenolic tetrahydrobenzylisoquinoline 145 followed by N-methylation yielded the hayatine analog 146 (2'1). Turning now to the structurally simpler alkaloid ( - )-cycleanbe (147), a promising route to its preparation appeared to be Ullmann condensation of 2 moles of 8-bromoarmepavine, since the alkaloid is symmetrical.

376

MAURICE SHAMMA A N D VASSIL ST. GEORGIEV

1. Cu, aq. NaOH, A 2. C H J

c1 OCH,

145

'o

OCH3 146

One such attempt using ( k )-8-bromoarmepavine (148) and the superior cupric oxide-potassium carbonate-pyridine catalyst gave some of the dimer 149 but none of the expected mixture of cycleanines (72). A fully authenticated first total synthesis of ( k )-cycleanine (147) involved as a first hurdle the synthesis of the amino acid 151 as well as that of its corresponding methyl ester 155 (73, 7 4 ) . The aldehyde 150 was condensed with nitromethane to give a yellow nitrostyrene. Catalytic hydrogenation over Adams catalyst in acetic acid then gave the required amino acid 151, Scheme 41. Furthermore, the methyl ester 155 of the acid 151 was synthesized by the following alternate route. 3,4-Dimethoxy-5-bromophenethylamine, prepared by the reduction of the nitrostyrene 152 under Clemmensen CH,O

CH30

:\CH3

+I$

::::q cH30 Br

6 44 CH,

CH,O

CH3

OH

H 3 c \ : M 0 C H 3

CH, OCH,

147

OCH, 148

I49

6.

SYNTHESES OF BISBENZYLISOQUINOLINE

ALKALOIDS

377

cH30vcH 4 *(I CH30

1. CH3NOa

"CH30 " " T N H ,

2. Ha, Pt, HOAc

CH,COOH

CH,COOH

150

151

SCHEME 41

conditions, was converted to the N-carbobenzoxy derivative 153. Ullmann condensation between 153 and methyl p-hydroxyphenylacetate afforded the product 154, and catalytic removal of the blocking group gave rise to the desired methyl ester 155, Scheme 42. The amino acid 151 was next protected as its N-carbobenzoxy derivative 156. Condensation between 155 and 156 furnished the amide 1. Zn/Hg, HCl c

H

3

0

T

v

"

0

2

2. Ph-CHP-O-C

' 40

c1

CH30 Br 152

CH,O H o ~ c H 2 - - C O O C H I .

cH30qT

CuO, K.CO3, pyridine

Br

t

OCH,Ph

153

CH30

CH&OOCH, 154

SCHEME 42

CH2COOCHS 155

378

=.I;

I

“s

0, M

I

0

G

MAURICE SHAMMA AND VASSIL ST. OEORGIEV

6. SYNTHESES OF BISBENZYLISOQUINOLINE ALKALOIDS m

+

379

380

MAURICE SHAMMA A N D VASSIL ST. GEORGIEV

157, which was converted t o the carboxylic acid 158. Esterification of 158 with p-nitrophenol and DCC was followed by treatment with hydro-

gen bromide to remove the carbobenzoxy group. The resulting amine hydrobromide 159 readily suffered cyclization t o the bisamide 160, and Fischler-Napieralski cyclization followed by reduction led to a mixture of tetrahydroisoquinolines. N-Methylation finally furnished a mixture

cH30v cH30 44 66

CH,O

Po

\

CH30

N\CH3

OCH,Ph

OCH,Ph

HN&z s”^ 3 H c

o & .

.

.

.

H .C \ N ) OCH,

163

164

OCHaPh

I

I

H,CLN&

CH, OCH, 165

6.

SYNTHESES OF BISBENZYLISOQUINOLINE ALKALOIDS

381

of products which generated ( t )-cycleanhe (147)after chromatography. Two other products obtained from the chromatographic separation were the dimers 161 and 162, Scheme 43. A later study in the cycleanine series demonstrated that BischlerNapieralski cyclization of the amide 163 proceeds in two directions to supply ultimately amines 164 and 165 (75). XI. Miscellaneous Syntheses The alkaloid aztequine was supposedly isolated from the leaves of yoloxochitl, Tabma mexicana Don. (Magnoliaceae) and was assigned structure 166 with no delineation of stereochemistry. This assignment is certainly in error, since in the same paper the unlikely claim was made that hydroiodic acid ruptured the diaryl ether linkage of the alkaloid without touching the methoxyl groups (?‘G).

I

I

OH

OH 166

Attempted syntheses of 166 either involve initial preparation of the diaryl ether corresponding to the two bottom rings, followed by further elaboration to construct the two tetrahydroisoquinoline units, or include an Ullmann condensation to bond together the two tetrahydrobenzylisoquinoline units (77-79). The bisbenzylisoquinolines 167, 168, and 169, which have no analogs in nature, have been synthesized through Ullmann condensation between 170 and 171 in the case of 167; 172 and 173 in the case of 168; and 174 and 175 in the case of 169 ( 8 0 , S l ) .

167

382

MAURICE SHAMMA AND VASSIL ST. GEORGIEV

168

169

170 R, = OH, R, = H 171 R1 = Br, R, = H 172 R, = H, Ra = OH 173 R, = H, R, = Br 174 R1 = H, RP = OH 175 R, = Br, R, = H

The dimer 176 has also been prepared in the course of a study of structural requirements for tumor-inhibitory activity among bisbenzylisoquinolines (13).

Lastly, an important related synthesis that should be a t least mentioned here in passing is that of the alkaloid ( + )-thalicarpine (177), which is an aporphine-benzylisoquinoline rather than a bisbenzylisoquinoline (82-84).

6.

383

SYNTHESES OF BISBENZYLISOQUINOLINE ALKALOIDS

XU. Syntheses Using Phenolic Oxidative Coupling Historically, significant attempts a t the phenolic oxidative coupling of tetrahydrobenzylisoquinoline free bases were reported as early as 1932, but they generated only dibenzopyrrocolines (85, 86). The first phenolic oxidative coupling leading to a bisbenzylisoquinoline was not reported until 1962, when it was shown that ferricyanide oxidation of the quaternary salt ( +_ )-magnocurarine iodide (178) at pH 10 yielded the dimer 180 in 1Sy0yield (87, 88).

-0‘

RO

178 R = H

179

R = CH3

XQ

0 # RO

OR

R = H 181 R = CH, 180

x@

384 MAURICE SHAMMA AND VASSIL ST. GEORGIEV

6.

SYNTHESES OF BISBENZYLISOQUINOLINE

ALKALOIDS

385

Similarly, ( f )-4’-O-methylmagnocurarine iodide (179) furnished the corresponding dimer 181, while ( )-armepavine methiodide, which has a methoxy group a t C-7 and a hydroxy a t C-4‘, could not be dimerized (87-89). I n a variation on this theme, and using the free base instead of the quaternary salt, it was demonstrated that ferricyanide oxidation of ( f )-4’-O-methyl-N-methylcoclaurine (182) in a two-phase system of chloroform-0.1 N sodium carbonate (pH 11.4) a t or below room temperature resulted in formation of the racemic diastereomers 183 and 184 in about 15% yield and separable by chromatography, Scheme 44 (901. It will be recalled that in an initial attempt it had been found that ( k )-armepavine methiodide did not dimerize a t room temperature. Reexamination of this oxidation under more severe conditions, namely, 0.1 N sodium carbonate solution and potassium ferricyanide on a steam bath or 1 N sodium hydroxide and silver nitrate a t room temperature, produced the carbon-carbon dimer 185 in about 15y0 yield (91,92).

185

In an atte.mpt to prepare the aporphine base ( f )-N-methylcaaverine (186) by phenolic oxidative coupling, the ferric chloride oxidation of racemic tetrahydrobenzylisoquinoline 187 was investigated. The products were the dienone 188 in 2.4y0 yield and the dimeric benzylisoquinoline 189 in 1.1% yield, Scheme 45 (93). A few studies have also been concerned with the enzymatic oxidation of tetrahydrobenzylisoquinolines. Oxidation of ( 5 )-N-norarmepavine (190) a t pH 6.5 with crude horseradish peroxidase and hydrogen peroxide yielded a complex mixture that included small yields of the isoquinolines 191, 192, and 193, Scheme 46 (94). Other investigations have dealt with the enzymatic oxidation of phenethyltetrahydroisoquinolines (95, 96).

c HO H 3 0 p N \ C H 3

CH3

aq. FeCl,, 30:40'C

+ H3C'

J&K l.

N

HO 188

187

186

SCHEME 45

CH3

'

\ 189

OH

6.

HO

SYNTHESES OF BISBENZYLISOQUINOLINE

ALKALOIDS

387

J 9 190

cH30p CH,O

191

HO OH

OH 192

198

SCHEME 46

XIII. Synthesis Using Electrolytic Oxidation The first preparation of a naturally occurring bisbenzylisoquinoline alkaloid, namely, dauricine, using an oxidative method occurred when the sodium salt of ( )-N-carbethoxy-N-norarmepavine (194) was subjected to electrolysis using tetramethylammonium perchlorate as the electrolyte, a graphite anode, and a platinum cathode (97). A mixture of the dimers 195 and 196 was obtained and separated. The dimer 196 then furnished a racemic and diastereomeric mixture of dauricines 3 following 0-benzylyation, reduction, and catalytic debenzylation. Such an electrolytic oxidative dimerization was unsuccessful when the nitrogen function was not protected, Scheme 47.

XIV. Use of Pentafluorophenyl Copper The most promising avenue to the bisbenzylisoquinolines presently appears to be via an improved Ullmann diary1 ether synthesis utilizing pentafluorophenyl copper in dry pyridine. Thus condensation of

388

MAURICE SHAMMA AND VASSIL ST. GEORGIEV

CaH5OOC /N

mz Electrolysis in wet

acetonitrile

b

O

e Nee

194

C2H,00C/N

CH,O

OCH,

195

+

196

then,

H3C’

1. 2. PhCH.CI, ImiAIH4 base

196

3. H., PdIC

NP

O

C

HOCH3 3

t

SCEEME 47

cH3 CH3O

6.

SYNTHESES OF BISBENZYLISOQUINOLINE ALKALOIDS

389

( + ) - 6'-bromolaudanosine (197)with ( + )-armepavine and pentafluoro-

phenyl copper in dry pyridine gave an impressive 53y0yield of the dimer 198, the S,S isomer of tetra-0-methylmagnolamine (98). Analogous condensations have also led to the preparation of aporphine-benzylisoquinoline dimers (98).

I

OCH, 198

REFERENCES

H. Kondo, Z. Narita, and S. Ueyo, Ber. 68, 519 (1935). T. Kametani and K. Fukumoto, Pet. Lett. 2771 (1964). T. Kametani and K. Fukumoto, J. Chem. SOC.6141 (1964). K. Fujitani, Y. Aoyagi, and Y. Masaki, Yakuqaku Zasshi 84, 1234 (1964). K . Fujitani, Y. Aoyagi, and Y. Masaki, Yakuqaku Zasshi 86, 654 (1966). T. Kametani, S. Takano, R. Yanase, C. Kibayashi, H. Ida, S. Kano, and K. Sakurai, Chem. Pharm. Bull. 14, 73 (1966). 7. M. Tomita, K. It6, and H. Yamaguchi, Chem. P h a m . Bull. 3, 449 (1955). 8. I. N. Gorbacheva, G. V. Bushbek, L. P. Varnakova, L. M. Shulov, and N. A. Preobrazhenskii, Zh. Obxh. Khim. 27, 2297 (1957). 9. F. D. Popp, H. W. Gibson, and A. C. Noble, J . Orq. Chem. 31, 2296 (1966). 9a. D. C. Smith and F. D. Popp, J . Heterocycl. Chem. 13, 573 (1976). 10. T. Kametani, K. Sakurai, and H. Iida, Yakugaku Zasshi 88, 1163 (1968). 11. K. Fujitani and Y. Masaki, Yakugaku Zasshi 86, 660 (1966). 1. 2. 3. 4. 5. 6.

390

MAURICE SHAMMA A N D VASSIL ST. GEORGIEV

12. M. Tomita and J. Niimi, Yakugaku Zasshi 79, 1019 (1959). 13. S. M. Kupchan and H. W. Altland, J. Med. Chem. 16, 913 (1973). 14. I. N. Gorbacheva, L. P. Varnakova, E. M. Kleiner, I. I. Chernova, and N. A. Preobrazhenskii, Zh. Obsch. Khim. 28, 167 (1957). 15. I. N. Gorbacheva, E. N. Tzvetkov, L. P. Varnakova, A. I. Gavrilova, and N. A. Preobrazhenskii, Zh. Obsch. Khim. 25, 1423 (1955). 16. T. Kametani, R. Yanase, S. Kano, and K. Sakurai, J. Heterocycl. Chem. 3,239 (1966). 17. T. Kametani, H. Iida, and K. Sakurai, Chem. Pharm. Bull. 16, 1623 (1968). 18. T. Kametani, H. Iida, and K. Sakurai, J. Chem. SOC.C 500 (1969). 19. T. Kametani, S. Takano, T. Kobari, H. Iida, and M. Shinbo, Chem. Pharm. Bull. 16, 1625 (1968). 20. M. Tomita, K. Fujitani, Y. Masaki, and Y. Okamoto, Chem. Pharm. Bull. 16, 70 (1968). 21. T. Kametani and F. Satoh, Chem. Pharm. Bull. 16, 773 (1968). 22. M. Tomita and K. It6, Yakugaku Zasshi 78, 103 (1958). 23. M. Tomita and K. It6, Yakugaku Zasshi 78, 605 (1958). 24. T. Kametani and H. Yagi, Tet. Lett. 953 (1965). 25. T. Kametani and H. Yagi, Chem. Pharm. Bull. 14, 78 (1966). 26. I. N. Gorbacheva, M. I. Lerner, G. G. Zapesochnaya, L. P. Varnakova, and N. A. Preobrazhenskii, Zh. Obsch. Khim. 27, 3353 (1957). 27. T. Kametani, H. Yagi, and S. Kaneda, Chem. Pharm. Bull. 14, 974 (1966). 28. J. Niimi, Yakugaku Zmshi 80, 123 (1960). 29. H. Kondo, H. Kataoka, and Y. Baka, Annu. Rep. I T S U U Lab. 5 , 59 (1954); C A 49, 14781e (1955). 30. H. Kondo, H. Kataoka, and K. Kigasawa, Annu. Rep. I T S U U Lab. 6, 46 (1955); C A 50, 10113a (1956). 31. W. M. Whaley, L. Starker, and M. Meadow, J. Org. Chem. 19, 833 (1954). 32. W. M. Whaley, L. N. Starker, and W. L. Dean, J. Org. Chem. 19, 1018 (1954). 33. W. M. Whaley, W. L. Dean, and L. N. Starker, J. Org. Chem. 19, 1020 (1954). 34. W. M. Whaley, M. Meadow, and W. L. Dean, J. Org. Chem. 19, 1022 (1954). 35. M. Tomita, K. Fujitani, and T. Kishimoto, Yakugcku Zasshi 82, 1148 (1962). 36. M. Tomita, Y. Inubushi, and Y. Masaki, Japanese Pat. 7121396; CA 76, 59845b (1972). 37. Y. Inubushi, Y. Masaki, S. Matsumoto, and F. Takami, J. Chem. SOC.C 1547 (1969). 37a. S. M. Kupchan, N. Yokoyama, and B. S. Thyagarajan, J. Pharm.Sci. 50,164 (1961). 38. M. Tomita, K. Fujitani, and Y. Aoyagi, Tet. Lett. 1201 (1967). 39. M. Tomita, K. Fujitani, Y. Aoyagi, andY. Kajita,Chem. Pharm. Bull. 16,217 (1968). 39s. T. Ibuka, T. Konoshima, and Y. Inubushi, Chem. Pharm. Bull. 23, 133 (1975). 39b. T. Kametani, H. Iida, S. Kano, S. Tanaka, K. Fukumoto, S. Shibuya, and H. Yagi, J. Heterocycl. Chem. 4, 85 (1967). 40. E. Fujita and A. Sumi, Chem. Pharm. Bull. 18, 2591 (1970). 41. E. Fujita, A. Sumi, and Y. Yoshimura, Chem. Pharm. Bull. 20, 368 (1972). 42. M. Tomita, Y. Inubushi, and M. Kozuka, Chem. Pharm. Bull. 1, 360 (1953). 43. Y. Inubushi, Chem. Pharm. Bull. 2, 1 (1954). 44. Y. Inubushi and M. Kozuka, Chem. Pharm. Bull. 2 , 215 (1954). 45. M. Tomita and H. Furukawa, Yakugaku Zasshi 83, 676 (1963). 46. M. Tomita and H. Furukawa, Yakugaku Zmshi 84, 1027 (1964). 468. Y. Inubushi, Y. Ito, Y. Masaki, and T. Ibuka, Tet. Lett. 2857 (1976). 47. M. Tomita, S. Ueda, and A. Teraoke, Tet. Lett. 635 (1962). 48. M. Tomita, S . Ueda, and A. Teraoka, Yakugaku Zmshi 83, 87 (1963).

6.

SYNTHESES O F BISBENZYLISOQUINOLINE ALKALOIDS

391

48a. M. Shamma, J. E. Foy, T. R. Govindachari, and N. Viswanathan, J . Org. Chem. 41, 1293 (1976). 49. B. Anjaneyulu, T. R. Govindachari, and N. Viswanathan, Tetrahedron 27,439 (1971). 50. Y.-Y. Hsieh, P.-C. Pan, W.-C. Chen, and Y.-S. Kao, Sci. Sin. 12, 2018 (1964); C A 62, 9184b (1965). 51. Y.-Y. Hsieh, P.-C. Pan, W.-C. Chen, and Y . 3 . Kao, Yao Hsueh Hweh Pao 13, 166 (1966); C A 65, 8979d (1966). 52. T. Kametani, S. Takano, K. Masuko, and F. Sasaki, Chem. Pharm. Bull. 14,67 (1966). 53. M. Tomita, H. Furukawa, T. H. Yang, and T. J. Lin, Tet. Lett. 2637 (1964). 54. T. Kametani, S. Takano, H. Iida, and M. Shinbo, J. Chem. SOC.C 298 (1969). 55. H. Furukawa, Yakugaku Zasshi 85, 335 (1965). 56. A. J. Everett, L. A. Lowe, and S. Wilkinson, Chem. Commun. 1020 (1970). 57. H. M. Sobell, T. D. Sakore, S. S. Tavale, F. G. Canepa, P. Pauling, and T. J. Petcher, Proc. Natl. Acad. Sci. U.S.A. 69, 2212 (1972). 58. L. V. Volkova, 0. N. Tolkachev, and N. A. Preobrazhenskii, Dokl. Akad. NaukSSSR 102, 521 (1955); C A 50, 4990i (1956). 59. 0. N. Tolkachev, V. G. Voronin, and N. A. Preobrazhenskii, Zh. Obshch. Khim. 29, 1192 (1958). 60. V. G. Voronin, 0. N. Tolkachev, and N. A. Preobrazhenskii, Izv. Vyssh. Uchebn. Zaved. Khim. Khim. Tekhnol. 5, 449 (1962); C A 59, 2877e (1963); and V. Voronk, 0. Tolkachev, A. Prokhorov, V. Chernova, and N. Preobrazhenskii, Khim. Geterotsikl. Soedin. 4, 606 (1969); CA 31, 79277p (1970). 61. V. G. Voronin, 0. N. Tolkachev, and N. A. Preobrazhenskii, Dokl. Akad. Nauk SSSR 122, 77 (1958); C A 53, 1345f (1959). 62. 0. N. Tolkachev, L. P. Kvashnina, and N. A. Preobrazhenskii, Zh. Obshch. Khim. 36, 1764 (1966). 63. E. N. Tzvetkov, I. N. Gorbacheva, and N. A. Preobrazhenskii, Zh. Obsch. Khim. 27, 3370 (1957). 64. V. I. Shvets, L. V. Volkova, and 0. N. Tolkachev, Izv. Vyssh. Uchebn. Zaved. Khim. Khim. Tekhnol. 5, 445 (1962); C A 59, 2876h (1963). 65. J. Baldas, I. R. C. Biek, Q. N. Porter, and M. J. Vernengo, Chem. Commun. 132 (1971). 66. H. Hellmann and W. Elser, Ann. 639, 77 (1961). 67. M. F. Grundon and H. J. H. Perry, J. Chem. SOC.3531 (1954). 68. J. R. Crowder, M. F. Grundon, and J. R. Lewis, J. Chem. SOC.2142 (1958). 69. M. F. Grundon, J. Chem. Soc. 3010 (1959). 70. T. Kametani, H. Iida, and K. Sakurai, J. Chem. SOC.C 1024 (1971). 71. K. P. Agarwal, S. Rakhit, S. Bhattarcharji, and M. M. Dhar, J.Sci. Ind. Res., Sect. B 19, 479 (1960); C A 55, 16585a (1961). 72. M. Tomita, Y. Aoyagi, Y. Sakatani, and K. Fujitani, Chem. Pharm. Bull. 15, 1996 (1967). 73. M. Tomita, K. Fujitani, and Y. Aoyagi, Tet. Lett. 4243 (1966). 74. M. Tomita, K. Fujitani, and Y. Aoyagi, Chem. Pharm. Bull. 16, 62 (1968). 75. M. Tomita, Y. Aoyagi, Y. Sakatani, and K. Fujitani, Chem. Pharm. Bull. 16, 56 (1968). 76. E. S. Pallares and E. M. Garza, Arch. Biochem. 16, 275 (1948). 77. T. Kametani, K. Fukumoto, and M. Ro, Yakugaku Zusshi 84, 532 (1964). 78. T. Kametani, M. Ro, and Y. Iwabuchi, Yakugaku Zasshi 85, 355 (1965). 79. T. Kametani, H. Iida, M. Shinbo, and T. Endo, Chem. Pharm. Bull. 16, 949 (1968). 80. J. Niimi, Yakugaku Zusshi 80, 451 (1960).

392

M A U R I C E SHAMMA A N D VASSIL ST. GEORGIEV

81. J. Niimi, Yakugaku Zasshi 80, 791 (1960). 82. S. M. Kupchan and A. J. Liepa, Chem. Commun. 599 (1971). 83. S. M. Kupchan, A. J. Liepa, V. Kameswaran, and K. Sempuku, J . Am. Chem. SOC. 95, 2995 (1973). 84. For other syntheses of aporphine-benzylisoquinolinealkaloids, see M. Tomita, H. Furukawa, S.-T. Lu, and S. M. Kupchan, Tet. Lett. 4309 (1965); Chem. Pharm. Bull. 15, 959 (1967); R. W. Doskotch, J. D. Phillipson, A. B. Ray, and J. L. Beal, Chem. Commum. 1083 (1969); J . Org. Chem. 36, 2409 (1971). 85. C. Schopf and K. Thierfelder, Ann. 497, 22 (1932). 86. R. Robinson and S. Sugasawa, J. Chem. SOC.789 (1932). 87. B. Franck, G. Blaschke, and G . Schlingloff, Tet. Lett. 439 (1962). 88. B. Franck and G. Blaschke, Ann. 668, 145 (1963). 89. B. Franck, G. Blaschke, and G. Schlingloff, Angew. Chem., Int. Ed. Engl. 3, 192 (1964). 90. M. Tomita, Y . Masaki, K . Fujitani, and Y. Sakatani, Chem. Pharm. BuZZ. 16, 688 (1968). 91. A. M. Choudhury, I. G. C. Coutts, A. K. Durbin, K. Schofield, and D. J. Humphreys, J. Chem. SOC.C 2070 (1969). 92. See also M. Tomita, Y. Masaki, and K. Fujitani, Chem. Pharm. BuZZ. 16, 257 (1968); M. Tomita, I(.Fujitani, Y. Masaku, and K.-H. Lee, ibid. 251. 93. T. Kametani and I. Noguchi, J . Chem. SOC.C 502 (1969). 94. Y. Inubushi, Y. Aoyagi, and M. Matsuo, Yet. Lett. 2363 (1969). 95. T. Kametani, S. Takano, and T. Kobari, J . Chem. SOC.C 9 (1969). 96. T. Kametani, H. Nemoto, T. Kobari, and S. Takano, J . HeterocycZ. Chem. 7, 181 (1970). 97. J. M. Bobbitt and R. C. Hallcher, Chem. Commun. 543 (1971). 98. M. P. Cava and A. Afzali, J. Org. Chem. 40, 1553 (1975).

-CHAPTER

7-

THE HASUBANAN ALKALOIDS YASUOINUBUSHI AND TOSHIRO IBUKA Kyoto University Sakyo.ku. Kyoto. Japan

I . Introduction ........................................................ I1. Occurrence and Physical Constants of Hasubanen Alkaloids .............. I11 Structure Elucidations ............................................... A . Mass Spectroscopy ............................................... B. Structures of Hasubenan Alkaloids ................................. I T. Synthesis of the Hasubanan Skeleton ................................. A. Synthesis via Ketolactones ........................................ B . Synthesis via'Ketonitriles ......................................... C. Synthesis via Cyclic Enrtmines ..................................... D . Synthesis via Spiroketone .......................................... E . Synthesis by Phenol Oxidation ..................................... V. Synthesis of Hasubanan Alkaloids ..................................... A . Cepharamine .................................................... B. Hasubanonine and Aknadilactem .................................. C. Metaphanine .................................................... V I . Biosynthesis ....................................................... References .........................................................

.

393 395 395 395 398 414 414 415 416 418 419 419 420 422 424 427 428

.

I Introduction Work on alkaloids of the hasubanan group up to 1970 have been reviewed in Volume XI11 of this treatise (1). I n the succeeding four years that are covered in the present review. significant advances in this field have been made in discovering thirteen new congeners and also in synthetic studies of the hasubanan skeleton and of this type of alkaloids. So far as we know. the occurrence of the hasubanan alkaloids has been noted in Stephania species only. and no alkaloid has been found in other species of Menispermaceae of special interest from the chemotaxonomical viewpoint.

TABLE I PLANT SOURCE AND PHYSICAL PROPERTIES Plant species

Stephania abyssinica Walp.

Stephania cephalantha Hayata Stephania delavayi Diels Stephania hernandifolia Walp.

Stephania japonica Miers

Stephania sasakii Hayata

" Constants for methiodide.

Alkaloid Metaphanine Stephabyssine Stephaboline Prostephabyssine Stephavanine Cepharamine Delavaine 16-Oxodelavaine Aknadicine Aknadinine Hernandine Methylhernandine Hernandolinol Hernandifoline Hernandoline 3-0-Demethylhernandifoline Protostephabyssine Stephisoferuline Metaphanine Prome taphanine 16-Oxoprometaphanine Homostephanoline Hasubanonine 16-Oxohasubanonine Miersine Stephasunoline Stepham iersine Epistephamiersine Oxostephamiersine Aknadilactam Alrnadinine Constants for hydrobromide.

Formula

Melting point ("C)

233 178-180 186-188 196-198' 229-230 186-187 140-150 221-222 156 70

197-199 152-153 114-115 227-227.5 19G191 148-149 196-198" 133-135 232 207a 115 233 116-117 161 222 233 165 98 290 210-214

ralD

-21 (CHCl,) - 58.9 (CHCI,) +34.7 (MeOH) -105 (MeOH)" + 30 (pyridine) -248 (CHCI,) -240 (CHCI,) -180 (CHCI,) -200 (EtOH) -283 (EtOH) -33 (EtOH) +125 (EtOH) -97.9 (EtOH) -25 (EtOH)

-

-105 (MeOH)" +48 (MeOH) -41 (CHCI,) -32 (MeOH)& -52 (CHC1,) -247.8 (CHCI,) -219 (EtOH) - 105.2 (EtOH) 121.4 (CHCl,) +33 (CHCI,) + 64.1 (CHCI,) 88.3 (CHCI,) -189 (CHC1,) -183 (MeOH)b

+

+

Reference

3, 4 5 5 5 6

7 8 9 10-12 10-12 13 14 15 16 17 18 5 19 20-25 26, 27 28 29-31 32-37 38 1, 39 40, 41 40, 41 40, 41 40, 41 10, 42 43

7. HASUBANAN ALKALOIDS

395

The numbering system of the hasubanan skeleton (1) (2,3,4,5tetrahydro-3a,9b-butano-l H-benz[e]indole), which is used throughout this review, is that proposed by Tomita et al. in their earlier paper ( 2 , 32). 3

I

H 1

11. Occurrence and Physical Constants of the Hasubanan Alkaloids Table I gives a survey of the occurrence and physical constants of hasubanan alkaloids.

III. Structure Elucidations

A. MASS SPECTROSCOPY From the measurements of IR, UV, and NMR spectra, it is difficult t o determine the hasubanan skeleton of unknown alkaloids. The mass spectral feature, however, exhibits a very characteristic fragmentation pattern and therefore provides a rapid and convenient method for structure elucidation of hasubanan alkaloids, especially that of alkaloids obtained in small amounts (3, 5 , 1 3 , 1 6 , 40). 1 . Hasubanan Derivatives Possessing No Oxygen Function at C Ring

I n the mass spectra of 3,4-dimethoxy-N-methylhasubanan(2), 3-methoxy-4-hydroxy-N-methylhasubanan (3),and lO-oxo-3,4-dimethoxy-N-methylhasubanan (4), the most abundant and diagnostic peak appears at m/e M-56. The first rupture occurs in ring C to furnish an ion, a or e . The additional loss of methyl or hydrogen from the fragment ion a must give rise to ion b or c and the loss of a methoxyl radical from the ion c produces ions d and/or d'. The fragmentation pattern of these compounds is a primary breakdown path for all hasubanan alkaloids (44, 45).

396

YASUO INUBUSHI AND TOSHIRO IBUKA

&

;.19

OMe

OMe

rJ*

_3

+ /

fN

I

Me R=Me

2 3

a

R = H

I

I

Me

Me

m/e 245 ( R = Me) M-56

OMe

(?Me

+

d

m/e 213

OMe

I

I

/

I

I Me

Me

Me

&& b

m/e 230

__f

u: tN

I

m/e 244

c

0

d'

rn/e 213 OMe

+/

I

Me

Me

Me 4

e

m/E 259

e'

m/e 259

SCHEME 1

2. Alkaloids Possessing a Hemiketal or .a Ketal Ether Linkage between C-8-C-10: Metaphanine ( 5 ) and Stephamiersine (6)

The mass spectrum of metaphanine ( 5 ) (3, 20, 21, 44) exhibits the most abundant ion peak (a or a') a t m/e 245, which may arise from the intermediate f by homolysis of the C-5-C-13 bond and the associated hydrogen transfer from C - 5 or C-6 to C-10 or (2-13. The hydro,aen source

7.

397

HASUBANAN ALKALOIDS

and the mechanism of this hydrogen transfer, however, are uncertain. The ion a’ may derive further stabilization by the loss of hydrogen to a fragment ion g a t m l e 2 4 4 . Alternatively, the ion a t m/e 244 may also occur from the intermediate f through the simple homolysis of the C-5-C-13 bond, and the structure h could be proposed for this ion. Another significant ion, i is observed a t m/e 243, and this ion may be

....... 10

”H

0

-

/ \ +

I

Me

I

Me

f

5

a’ mle 245

1 I

I

Me

Me h m l e 244

i mle243

Me

R1 = H, R, = OMe 7 R1 = OMe, R2 = H 6

SCHEME 2

g

mle244

398

YASUO INUBUSHI AND TOSHIRO IBUKA

derived from the intermediate f by the loss of hydrogen and the associated C-5-C-13 bond fission (5, 40, 44, 45). The cleavage mode mentioned above is quite common for all metaphanine type alkaloids possessing an ether linkage between c-8and C-10, and a ketone function at (2-7. By contrast, the fragmentation of stephamiersine (6) and epistephamiersine (7)) which possess an ether linkage between C-8 and C-10 and a ketone function a t C-6, produces the most abundant ion, i, at mle 243 rather than an ion a' a t m/e 245. This difference may be of diagnostic significance, as it demonstrates the presence of a ketone function a t C-6 in metaphanine type alkaloids (4U). 3. Alkaloids Possessing an +Unsaturated Ketone Group a t C Ring: Iso-6-dehydrostephine (8) and Hasubanonine (129) Alkaloids such as hasubanonine (129), possessing an a,/?-unsaturated carbonyl group a t C ring, show a similar breakdown pathway as that of metaphanine and others. An important feature of the spectra of these alkaloids is that two intensive ion peaks are observed-one is an ion a or a' and the other is an ionj, which occurs by the loss of the ethanamine chain from the molecule. I n the case of isodehydrostephine (8)) the most abundant ion peak, j, was found a t mle 301 ( 6 ) . 0

1

+

I

H j

8

mle 301

SCHEME 3

B. STRUCTURES OF HASUBANAN ALKALOIDS 1. Stephisoferuline (9)

Stephisoferuline was isolated from Stephania hernandifolia, and the presence of four methoxyl groups, one secondary amino group, an a$-unsaturated ester moiety, and two phenolic hydroxyl groups was shown (19).A new hasubanan ester-ketal structure (9) was assigned to

7.

399

HASUBANAN ALKALOIDS

stephisoferuline on the following evidence. Hydrolysis of stephisoferuline afforded stephuline (10) and isoferulic acid. The former gave N-methylstephuline (11) on methylation, confirming the presence of a secondary amino group. Treatment of 10 with dilute hydrochloric acid led t o facile demethylation of acetalmethyl t o give 8-demethylstephuline (12) and oxidation of 10 with Jones’ reagent provided 6-dehydrostephuline (13). On the other hand, acetylation of 10, followed by acid treatment, resulted in the triacetyl derivative 14, and the downfield shift of the signal for the C-7 H (6 4.20) of 14 in its NMR spectrum compared with that of stephisoferuline (6 3.75) supported the assignment of the ketone function a t C-8.Hydrogenation of the triacetyl derivative 14 furnished the dihydrotriacetyl derivative 15, which on treatment with acetone dimethylacetal in the presence of p-toluenesulfonic acid afforded the rearrangement product 16. The compound 16 was identified with the base derived from aknadicine ( = 4-demethylnorhasubanonine) (17) ( 1 0 , 1 1 ) as follows. Reduction of 17 with NRH gave the C-6 epimeric alcohols 18. Acetylation of 18 gave the products that were converted t o the triacetyl derivative 16 by letting their chloroform solution stand. This chemical correlation established the planar structure of stephisoferuline, the stereochemistry a t C-8, C-10,C-13,and C-14, and the absolute configuration of the molecule. Since reduction of 6-dehydrostephuline (13) with NBH gave stephuline (10)solely, and the hydride attack from the a side of 13 was predictable from inspection of the molecular model, the /3 configuration of the C-6 hydroxyl group was suggested. On the other hand, chemical, spectral, and crystallographic examinations suggested the same configuration of five of the six asymmetric centers of stephisoferuline (9) with those of stephavanine (19) ( 6 ) .From the biogenetic analogy, the /3 equatorial configuration of the C-7methoxyl group of 9 was presumed (19). OMe

H 9

10 11 12

R2 R, = Me, R, = H R, = R, = M e R 1 = R, = H

400

YASUO INUBUSHI AND TOSHIRO IBUKA

Me0

’k I

H

AC

Ac 15

14

18

AcO

Me0

Me0 N

Me0 N

I

Ac 16

\

I

H 17

I

H 1s

2. Stephavanine (19)

Stephavanine was isolated from Stephania abyssinica grown in Eastern and Southern Africa, and the presence of one methylenedioxy, two hydroxyl, one secondary amino, and two methoxyl groups in its molecule was shown ( 6 ) .The mass spectrum of stephavanine revealed a diagnostic fragment ion k for hasubanan alkaloids at m/e 214 (44,45). Alkaline hydrolysis of stephavanine gave vanillic acid and stephine (20), and the 6,7-bistrirnethylsilyl ether 21 was derived from the latter. The NMR spectrum of 21 showed two unsplit aromatic proton signals, indicating the methylenedioxy group attached to C-2 and C-3 of an aromatic ring. Oxidation of 20 with Jones’ reagent provided 6-dehydrostephine (22), which on treatment with sodium hydroxide solution gave isodehydrostephine (8). Of six chiral centers of 19, the relative configurations of C-8, (2-10, C-13, and C-14 were inevitably established because of the cage ring system of the stephavanine moIecuIe. The /%axial configuration of the C-6 hydroxyl, which forms an ester linkage with vanillic acid, was deduced from the NMR spectral examination and the /I-equatorial configuration of the C-7 hydroxyl group was suggested

40 1

7. HASUBANAN ALKALOIDS

by the fact that oxidation of the diol20 provided selectively the monoketone 22. Thus, the structure 19 was assigned t o stephavanine (6). This conclusion was supported by X-ray crystallographic study of stephavanine hydrobromide ( 6 ) . Me0

H

O

G

RO

"H Me0 N

I

H 19

I

H 20 R = H 21 R = &(Me),

?"\

c# I

+/

"H

Me0 N

I

H

H 22

k

mle 214

3. Stephabyssine (23), Stephaboline (24), and Prostephabyssine (25)

Examination of basic constituents of Stephania abyssinica. collected in Ethiopia resulted in the isolation of three new phenolic hasubanan alkaloids-stephabyssine, stephaboline, and prostephabyssine ( 5 ) . Stephabyssine (23) had one N-methyl, one methoxyl, one saturated ketone, and two hydroxyl groups. The presence of a phenolic hydroxyl group with an unsubstituted para position was presumed by a positive color reaction with Gibbs' reagent. Methylation of 23 with methyl iodide in the presence of potassium carbonate provided O-methylstephabyssine, which was identified with metaphanine ( 5 ) (4,20-25, 46, 47). Thus, the structure of stephabyssine was established as 4-demethylmetaphanine (23).

402

YASUO INUBUSHI AND TOSHIRO IBUKA

OMe

OMe

J 3.. 3.. $ . H

0

HO

"H

HO N

"H HO N

I

I

I

I

Me

Me

23 R = H 5 R=Me

24

Stephaboline (24) was shown to possess one N-methyl, one methoxyl, and three hydroxyl groups ( 5 ) . The close relationship of stephaboline with stephabyssine (23)was indicated by similarities in their NMR spectra as well as positive reactions of each compound with ferric chloride and the Gibbs reagent. Since NBH reduction of stephabyssine gave stephaboline in a high yield, the structure of stephaboline was established except for the configuration of the C-7 hydroxyl group. The NMR spectrum of 24 exhibited a diffused multiplet a t S 4.4 assignable to the C-7 H. This signal changed to a pair of doublets (JAx= 5 Hz, J B X = 1 1 Hz) by treatment with D,O, and the magnitude of the coupling constant of J B X suggested the axial configuration of the C-7 H,thus confirming the equatorial configuration of the C-7 hydroxyl group (5). When treated with aqueous hydrochloric acid solution under mild conditions, prostephabyssine (25) gave stephabyssine (23)with loss of the elements of methanol in high yield. This facile hydrolysis demonstrated the presence of an enol methyl ether located at C-6-(2-7. Consequently, the structure 25 was assigned to prostephabyssine. Determination of the NMR spectra of prostephabyssine in a variety of solvents gave complex patterns indicative of the presence of the hemiketal25a and ketone 25b forms in equilibria similar to the solvent-dependent equilibria observed in prometaphanine (26,27). ?Me

?Me

OH HO N

I

I Me 258

25

SCHEME 4

Me 25b

7.

403

HASUBANAN ALKALOIDS

4. Stephamiersine (6), Epistephamiersine (7), Oxostephamiersine (26),

and Stephasunoline (28) Reinvestigations of basic constituents of Stephania japonica grown in Kagoshima Prefecture (the sourthern part of Japan) resulted in isolations of four new hasubanan alkaloids: stephamiersine, epistephamiersine, oxostephamiersine, and stephasunoline (40, 41). That the structures of these alkaloids were closely related to each other was presumed on the basis of their spectral data which are summarized in Table I1 and Table 111. TABLE I1 PHYSICAL CONSTANTSAND SPECTRAL DATAOF SOMEALKALOIDS FROM Stephania japonica Miers mp ("C)

A1kaloid Stephamiersine (6) Epistephamiersine (7) Oxostephamiersine(26) Stephasunoline (28)

165 98 290 233

[aID

(CHCl,) +33

+ 64.1 +88.3 $121.4

IR 3Y':7: (cm-')

W (nm)

1725 1735 1730, 1680 3550

286 286 286 286

AEtoH msx (6)

MS (mle) M+, base peak 389,243 389,243 403,257 377,245

2200 2300 2000 2000

TABLE I11 NMR SIGNALS OF SOME ALKALOIDS FROM Stephania japonica Miers4

Alkaloid

Aromatic protons (2H)

6 7 26 28

6.67 6.66 6.77 6.67

O

C-7-H 3.52 4.27 3.63 3.62

C-10-H 4.72 4.82 4.79 4.88

Methoxyl groups 3.92, 3.89, 3.92, 3.90,

3.82, 3.76, 3.83, 3.82,

3.34, 3.31 3.52, 3.45 3.33, 3.29 3.46

N-Methyl group 2.64 2.63 3.12 2.57

Chemical shifts are quoted in 6 values.

Equilibrium experiments of either stephamiersine (6) or epistephamiersine (7) with 1yosodium hydroxide solution gave an equilibrium mixture consisting of 6 and 7 in a 1:3 ratio. Consequently, 6 and 7 were epimers attributable to an asymmetric center adjacent to a carbonyl group, and 7 was thermodynamically more stable. Furthermore, permanganate oxidation of stephamiersine (6) gave the lactam, which was identified with oxostephamiersine (26). Reduction of epistephamiersine (7) with NBH provided dihydroepistephamiersine (27),* which on treatment

* Later. this compound waa obtained in nature from Stephuniu abyssinica by Dr. A. J. von Wyk.

404

YASUO INUBUSHI AND TOSHIRO IBUKA

with methanolic hydrochloric acid solution under mild conditions gave stephasunoline (28). This facile hydrolysis of dihydroepistephamiersine suggested the presence of the labile acetal methyl ether in its molecule. Thus, the chemical correlations among 6,7,26, and 28 were established. Acetolysis of stephamiersine (6) and epistephamiersine (7) provided 1,3-diacetoxy-2,5,6-trimethoxyphenanthrene (29) and 1,2,3-triacetoxy5,6-dimethoxyphenanthrene (30), respectively. On the other hand, acetolysis of dihydroepistephamiersine (27) gave the known 1-acetoxy2,5,6-trimethoxyphenanthrene (31) (26,27). On the occasion of acetolysis of morphinan and hasubanan series alkaloids, it is well known that a ketone function in the original molecule remains an acetoxyl group on the phenanthrene nucleus, and an alcoholic hydroxyl group is removed by dehydration in the course of the aromatization process ( 2 0 , 2 1 , 2 6 , 2 7 ,4 1 , 4 8 , 4 9 ) . From the structures of these phenanthrene derivatives derived from 6 , 7 , and 27, the positions of five of six oxygen functions were confirmed, and particularly, the C-3, C-4, and C-7 positions of three of four methoxyl groups and the C-6 position for an oxygen function in the original alkaloid molecule were established. I n the NMR spectra of 6 , 7 , and 28, a signal due to C-10 H appeared around 6 4.8 (doublet, J = 6.5 Hz). I n the spectrum of 7, the NOE [ I3y0 enchancement of the signal of this doublet ( 6 4.82)] was observed when irradiated a t the aromatic proton signal. The signals at 6 1.47 (doublet,J = 10.5 Hz) and 6 2.46 (double doublet, J = 10.5 and 6.5 Hz) were assigned to the C-9 methylene protons by the double resonance technique. From these assignments, it is obvious that an acetal ether linkage attaches to C-10. NBH reduction of oxostephamiersine (26) provided compound 33, which on treatment with perchloric acid-acetic anhydride gave compound 34. Oxoepistephamiersine (32) derived from epistephamiersine (7) by permanganate oxidation was reduced with NBH t o give compound 35, which on treatment with perchloric acid-acetic anhydride also afforded compound 34. Catalytic hydrogenation of 34 provided the conjugated ketone 36. On the other hand, NBH reduction of 16oxohasubanonine (37) (28, 38) gave epimeric alcohols (38), which on treatment with dilute mineral acid gave the same conjugated ketone 36. From these results, the skeletal structure and the attached positions of oxygen functions, C-6, C-7, C-8, and C-10 of oxostephamiersine (26) were established. The configurations of the C-7 OCH, group of these alkaloids were deduced from the NMR spectral experiments. I n the spectrum of stephamiersine (6), signals due t o the C-5 methylene protons were observed at 6 2.86 ( l H , double doublet, J = 11.5, 1.5 Hz) and 6 3.67

7.

HASUBANAN ALKALOIDS

405

( l H , doublet, J = 11.5 Hz). The long-range coupling between the C-7 H and one of two C-5 methylene protons ( 6 2.86) was observed by the homonuclear INDOR technique. On the other hand, the spectrum of epistephamiersine (7)revealed signals assignable to the C-5 methylene protons at 6 2.99 (IH, doublet, J = 11.5 Hz) and 6 3.18 (IH, doublet, J = 11.5 Hz) and the NOE (6.5y0enhancement) of the C-7 H signal was observed when irradiated the signal a t 6 3.18 but no signal enhancement was observed between the C-7 H and the signal at 6 2.99. From these findings, together with the equilibrium experiments previously discussed, the configuration of the C-7 OCH, was established to be aaxial in 6 and p-equatorial in 7. The configurations of C-6 OH and C-7 OCH, of stephasunoline (28) were also deduced from the NMR spectral examinations. The spectrum of stephasunoline exhibited signals assignable to the C-5 methylene protons a t 6 2.46 (IH, double doublet, J = 14.3, 2.4 Hz) and 6 2.82 ( l H , double doublet, J = 14.3, 3.8 Hz). When irradiated a t the signal appearing a t 6 2.46, the NOE (120J, enhancement) of the C-7 H signal (6 3.62, doublet, J = 3.9 Hz) was observed. This result, together with analysis of coupling constant values of the signals for four protons attached to C-5, c-6, and C-7, led to the conclusion that the C-7 OCH, group should be p-equatorial and the C-6 OH p-axial. Thus, the structure 28 was assigned to stephasunoline (40, 41). The planar structure of stephasunoline (28) is the same as that proposed for miersine (39)but the stereochemistry of C-6 OH and C-7 OCH, of the latter has not been established (1,39). 5 . 16-Oxohasubanonine (37)

This alkaloid was isolated from Stephania japonica and identified with 16-0xohasubanonine, which had been derived from hasubanonine by permanganate oxidation (28, 38). 6. 16-Oxoprometaphanine (40)

This alkaloid was isolated from Stephaniajaponica (28).On hydrolysis with dilute mineral acid 16-oxoprometaphanine gave known oxometaphanine (41) (50, 51) and compound 34, which had been derived from stephamiersine (6) and epistephamiersine ( 7 ) (40, 41). Acetylation of 16-oxoprometaphanine gave acetyl-16-oxoprometaphanine (42), which on treatment with dilute hydrochloric acid afforded compound 34. These chemical correlations and the NMR spectral examinations of 16-oxoprometaphacine and its transformation products supported the structure 40 for 16-oxoprometaphanine (28).

406

J 0-

- J t-+

4

YASUO INUBUSHI AND TOSHIRO IBUKA

On-

+

J 0-

i

0-

Ot...?

2

?

-J U

(0

-3 a

;bi O

i

El

m

w-

\

\ L

T

1

7. HASUBANAN ALKALOIDS

\

r"O - G p J -

H 0

B

407

408

YASUO INUBUSHI AND TOSHIRO IBUKA

?Me

?Me

Me0

I

I

Me 40s

Me 40b

40

SCHEME 6

@ 0

. .

HO Fi

I

Me 41

*

.H

‘0

&

Me0

OAc

O N

‘0

I

Me 42

7. Delavaine (43)

Delavaine was isolated from Stephania delavayi (8) and its IR spectrum exhibited bands a t 1670 cm- (cr,p-unsaturatedketone) and 1608 cm-l (C=C double bond) (8). Hydrolysis of the methylenedioxy group of delavaine with sulfuric acid and phloroglucinol gave the corresponding dihydroxy derivative 46, which on acetylation afforded the diacetyl derivative 47. The I R absorption of the ester carbonyl (17751780 c m - l ) in 47 showed the phenolic nature of the hydroxyls, from which it follows that the methylenedioxy group is attached to an aromatic ring. On the other hand, the NMR spectrum of delavaine exhibited two unsplit aromatic proton signals a t 6 6.41 and 6 6.64. Consequently, it is obvious that the methylenedioxy group is a t the C-2 and C-3 position of the aromatic ring. The Hofmann degradation of delavaine methiodide formed the methine base 44,which on acetolysis furnished the acetoxy-methoxy-phenanthrene derivative 45 (8), suggesting that delavaine belongs to the hasubanan alkaloids. In the NMR spectrum of delavaine, signals were present for N-methyl (6 2.49) and methoxyl(6 4.06 and 6 3.60) groups, and the C-5 methylene proton

7.

0

409

HASUBANAN ALKALOIDS

1

Me0 M eO 0& 0 M e 0/ N\

I

Me

Me

43

Me 44

Me

45

46 47

R = H R = Ac

signals appeared a t 6 2.46 (doublet, J = 16 Hz) and 6 3.00 (doublet, J = 16Hz). However, no C-9 H signal of the morphinan skeleton between 6 3.00 and 6 4.00 (52-55) was observed, thus demonstrating the hasubanan skeleton for delavaine. Consequently, structure 43 was proposed for delavaine (8), but no positive evidence regarding the stereochemistry of the ethanamine bridge is presented. 8. 16-Oxodelavaine (48)

16-Oxodelavaine was isolated from Stephunia delavayi grown in Transcaucasia (9).The UV spectrum of this alkaloid was similar to that of delavaine (8), and the IR spectrum showed bands for an a,/?unsaturated ketone (1686 cm-') and a lactam carbonyl (1670 cm-l) function. I n the NMR spectrum, signals were present for two isolated aromatic protons ( S 6.64, 1H, singlet and S 6.46, 1H, singlet),methylenedioxy ( 6 5.88, 2H, singlet), two methoxyl ( 6 4.10 and 3.66 each 3H and singlet), and an N-methyl (6 2.96, 3H, singlet) groups, and the c-5

410

YASUO INUBUSHI AND TOSHIRO IBUKA

methylene protons (6 2.90, lH, doublet, J = 16 Hz and 6 2.66, l H , doublet, J = 16 Hz). After various chemical, physicochemical, and spectral investigations, the structure 48 was proposed for 16-oxodelavaine (9).

Me 48

9. Hernandifoline (49) Hernandifoline was isolated from Stephania hernandifolia grown in t h e Black Sea littoral of Caucasia (16).The presence of four methoxyl, one secondary amino, two hydroxyl, and one a,p-unsaturated ester groups was shown. Methylation of hernandifoline (49) with methyl iodide afforded A'-methylhernandifoline (50), and alkaline hydrolysis of 49 gave a base (51) and hesperetic acid. The NMR spectrum of 51 revealed signals for two aromatic protons (6 6.49, 2H, singlet), C-10 H (6 4.76, doublet, J = 5.8 Hz), C-6 H (6 4.07, multiplet), C-7 H ( 6 3.62, doublet, J = 4.0 Hz), C-3 OCH, (6 3.67, singlet), C-8 OCH, (6 3.50, singlet), C-7 OCH, (6 3.38, singlet), C-5 methylene protons (6 3.04, 1H, quartet, J = 14.9, 3.5 Hz and 6 1.85, 1H, quartet, J = 14.9, 2.8 Hz), C-6 OH (6 2.13, l H , doublet,J = 10.0 Hz), C-9 methylene protons (6 2.34, 1H, quartet, J = 10.8, 5.8 Hz and 6 1.80, lH, doublet, J = 10.8 Hz). The mass spectrum of 51 showed the pattern characteristic for the hasubanan alkaloids (44),m/e 363 (M+), 217, and 216. Methylation of 51 with methyl iodide in methanol gave substance 52 and the further methylation of 52 with diazomethane gave compound 53. Following spectral investigations of the alkaloid and its degradative compounds, the structure of hernandifoline except the configuration at the C-7 OCH, group was proposed as 49 (16). This structure is the same as that proposed for stephisoferuline (9) (19),except for the configuration of the C-7 OCH, group. The reported melting points of hernandifoline (49) (227-227.5"C), the compound (51) (225-226"C), and the compound (52) (154-155°C)

7.

41 1

HASUBANAN ALKALOIDS

HO

M00Q7=~Lo@ / \ H

Me0

. .

Me0 N

I

R 49 50

R = H R = Me

'.H

Ho&

Me0

:

. *

*

.H

.

Me0

k

I

R2 5 1 R, = R, = H 52 R, = H, R, = Me 53 R, = R, = Me

differ from those of stephisoferuline (9) (133-135OC), stephuline (10) (223-225°C)) and N-methylstephuline (11) (126-128%), but there has been no report of direct comparisons of these alkaloids. 10. 3-0-Demethylhernandifoline (54)

3-0-Demethylhernandifoline was isolated from Stephania hernandifolia, and the presence of three hydroxyls, one secondary amino, and three methoxyl groups was shown (18). The IR spectrum exhibited bands for OH and NH a t 3560, 3440, and 3200-2700 cm-l, a carbonyl group at 1695 crn-l, and a conjugated double bond a t 1640 cm-l. I n the NMR spectrum signals were present for three methoxyls (6 3.89, 3.41, and 3.40), ortho-coupled aromatic protons (6 6.50, l H , doublet, J = 8.0 Hz and 6 6.60, 1H, doublet, J = 8.0 Hz), the C-5 methylene protons (6 2.02, l H , double doublet, J = 15.0, 2.3 Hz and 6 3.17, 1H, double doublet, J = 15.0, 4.1 Hz), C-6 H (6 5.40, l H , multiplet), C-7 H (6 3.74), and C-10 H (6 4.88, lH, doublet, J = 5.8 Hz). On alkaline hydrolysis, 3-0-demethylhernandifoline gave hesperetic acid and an amine (55))which gave an intense color reaction with ferric

I H 54

I

H 55

412

YASUO INUBUSHI AND TOSHIRO IBUKA

chloride characteristic for o-phenols. Methylation of 55 with methyl iodide, followed by treatment with diazomethane, furnished the N,O,O-trimethyl derivative, which was identical with compound 53 (16) derived from hernandifoline. From these chemical correlations, structure 54 was proposed for 3-0-demethylhernandifoline. 11. Hernandine (56)

Hernandine was isolated from Xtephania hernandifolia, and the presence of one N-methyl, two methoxyl, and three hydroxyl groups was suggested (13).The mass spectrum of this alkaloid revealed a characteristic fragment ion peak for hasubanan alkaloids a t m/e 231 (13, 44, 45). The NMR spectrum of hernandine showed signals for C-10 H (8 4232, OMe I

. :/J R20 N I

Me 56 or

R, = H, R, = Me R, = Me, Ra = H

1H, doublet, J = 6.2 Hz), C-9 methylene protons (8 1.51, l H , doublet, J = 10.8 Hz and S 2.85, 1H, double doublet, J = 10.8, 6.2 Hz), C-6 H ( 6 4.15, lH, multiplet), C-7 H (8 3.58, l H , doublet, J = 3.8 Hz), and C-5 methylene protons (6 3.09, l H , double doublet, J = 14.6, 3.5 Hz and 6 1.95, lH, double doublet, J = 14.6, 2.4 Hz). The axial configuration of C-6 OH was determined from the values of the spin-spin coupling between the C-5 methylene protons and c-6 H. From these results, structure 56 was proposed for hernandine (13),but the absolute

configuration of the ethanamine bridge, the configuration of the C-7 oxygen function, and the position of one of two methoxyl groups have not been definitely established. 12. Methylhernandine (57)

Methylhernandine was isolated from Stephania hernandifolia, and the presence of one N-methyl, two hydroxyl, and three methoxyl groups was suggested (14). On acetylation, methylhernandine gave diacetyl-

7.

HASUBANAN ALKALOIDS

413

methylhernandine, the IR spectrum of which showed carbonyl bands a t 1775 and 1730 cm-l, indicating that one of two hydroxyl groups is phenolic and the other alcoholic. I n the NMR spectrum of methylhernandine, signalswere present for C-5 methylene protons (S 1.93,lH, double doublet, J = 14.8, 2.9 Hz and 6 3.00, l H , double doublet, J = 14.8,

: :/

Me0 N

I

Me 57

3.4 Hz), C-6 H (6 4.05,lH, multiplet), C-6 OH (6 2.24, doublet, J = 9.8 Hz), C-7 H (6 3.62, l H , doublet, J = 4.1 Hz), C-10 H (6 4.81, l H , doublet, J = 6.2 Hz), and C-9 methylene protons (6 1.45,1H, doublet, J = 10.8 Hz and 6 2.63, lH, double doublet, J = 10.8, 6.2 Hz). Since

methylhernandine was identified with compound 52 (16) derived from hernandifoline (49) (16), structure 57 was proposed for methylhernandine (14). 1 3. Hernandolinol (58)

Hernandolinol was isolated from Stephunia hernandifolia grown in Caucasia, and the presence of one N-methyl, three methoxyl, and two hydroxyl groups was suggested. On Hofmann degradation, hernandolin01 gave the methine base (mp l14-115°C), which on acetolysis afforded the diacetoxydimethoxyphenanthrene derivative (mp 163164OC) (15). This methine base and phenanthrene derivative were OMe

Me

58

414

YASUO INUBUSHI AND TOSHIRO IBUKA

identified with the methine base and phenanthrene derivative derived from hernandoline, respectively (I?'),and hernandolinol was proved to be identical with the reduction product of hernandoline with sodium borohydride. Thus, structure 58, without stereochemical implications, was proposed for hernandolinol (15).

IV. Synthesis of the Hasubanan Skeleton The synthesis of the hasubanan skeleton has been undertaken in several laboratories with a remarkable degree of variability in the synthetic schemes. VIA KETOLACTONES A. STNTHESIS

Annelation reaction of the ketoester 59 with methyl vinyl ketone provided the ketolactone 60. Three methods available for introduction of the nitrogen atom into this ketolactone have been reported. The first method was reported by Inubushi et al. Treatment of the ketal lactone 61 from the ketolactone 60 with methylamine in the presence of methylamine hydrochloride gave the ketolactam 63 and the ketal amide 68 (56-58). Similarly, the ketoester 64provided the ketal lactone 66 and the ketolactam 67 via the ketolactone 65. The second method was developed by Evans et al. Reaction of the ketolactone 60 with methyl iodide in the presence of potassium carbonate gave the unsaturated ketoester 62, which on treatment with LAH-methylamine furnished the ketolactam 63 ( 5 9 , 6 0 ) .The last method was reported by Tahk et al. Reaction of the ketal lactone 61 with a large excess of methylamine gave the ketal amide 68, which was reduced with LAH to give the amino alcohol 69, acid R

R

59 R = H 64 R = OMe

R = H 65 R = OMe

60

R

61 R = H 66 R = OMe

415

7. HASUBANAN ALKALOIDS

I

Me 62

63 67

Me

R = H R = OMe

68 69

R = O R = Hz

I

Me 70

treatment of which afforded 7-0x0-N-methylhasubanan (70) (61, 62). The main disadvantage of these three methods was the low yield in the nitrogen introduction step.

B. SYNTHESIS VIA KETONITRILES This procedure consists in the Robinson annelation reaction of the ketonitrile (71 or 72) with methyl vinyl ketone. Treatment of the ketonitrile 71 with methyl vinyl ketone provided the separable stereoisomeric mixture 73. Treatment of the mixture with sodium alkoxide

4 Nc8 &

NC ’0

OH

N

Mo

71 72

R = H = OMe

R

73 74

I

OH

R = H R = OMe

%

H 75 76

R = H R = OMe

416

YASUO INUBUSHI AND TOSHIRO IBUKA

gave the ketolactam 75. Similarly, the ketonitrile 72 gave the ketolactam 76 (57, 58). This procedure is of practical value because of acceptable yields and simpler operations compared with the former methods.

C. SYNTHESIS VIA CYCLICENAMINES 1. Stork-Robinson Annelation Reaction

Synthesis of the key intermediate, the cyclic enamine 79, is analogous to that of 3-arylpyrroline in the mesembrine synthesis (63-65). Three methods available for synthesis of this intermediate have been developed. Reaction of /3-tertralone (77) with 1,2-dibrornoethane gave the spiroketone 78, which on treatment with methylamine furnished the cyclic enamine 79 (61).On the other hand, ketalization of the ketoester

Me

77 59

R = H R = CH,CO,Et

79

78

Me 80

81

59, followed by treatment with LAH-methylamine, afforded the ketal

amide 80. Successive treatments of 80 with LAH and aqueous acid solution provided the cyclic enamine 79 (59, 60). Further, reaction of p-tetralone (77) with excess methylamine, followed by treatment with titanium tetrachloride, yielded the enamine 81. When reacted with isopropylmagnesium chloride, this enamine gave the "bidentate" nucleophile which on treatment with bromochloroethane gave the cyclic enamine 79 (60). The cyclic enamine 79 thus synthesized was reacted with methyl vinyl ketone to yield 7-oxo-N-methylhasubanan (70) in a moderate yield (60-62).

417

7. HASUBANAN ALKALOIDS 2 . [4

+ 21 Cycloaddition and [2,3] Sigmatropic Rearrangement

A unique and elegant synthesis of hasubanan derivatives was reported by Evans et al. (66).Upon heating equimolar quantities of the sulfoxide 82 with the cyclic enamine 79,a diastereoisomer mixture of the sulfoxide 83 as well as some rearrangement amino alcohol 84 was obtained, indicating that [4 + 21 cycloaddition and [2,3] sigmatropic rearrangement were occurring consecutively. When heated with sodium sulfite, the unpurified reaction product from 79 and 82 afforded the

< R

. . C,H,-S

N

$ 1

0 Me 82

R = S-CSH,

83

J.

0 85

R

= C0,Me

Me

Me

84

86

desired amino alcohol 84. The evidence that 84 is a single isomer rather than an epimeric mixture was derived from its behavior on tlc, its cleanly resolved NMR spectrum, and the sharp melting range of the amine salt. The syn relationship between hydroxyl and nitrogen function was deduced from the observance of intramolecular hydrogen bonding in the IR spectrum. Similarly, addition of methyl pentadienoate to the cyclic enamine 79 was also found to afford the nicely crystalline tetracyclic ester 86 in 50% yield. Qualitatively, it appeared that the sulfoxide-substituted diene 82 was slightly less reactive than the estersubstituted diene 85 ( 6 6 ) .

418

YASUO INUBUSHI AND TOSHIRO IBUKA

D. SYNTHESIS VIA SPIROKETONE Another synthetic route for hasubanan derivatives was devised recently by the Bristol-Myers group. Alkylation of 7-methoxytetralone (87) with 1,4-dibromobutanein the presence of sodium hydride gave the

87

88

OMe I

R R

89 90

= CN = CH2NH,

OMe

(-yg I

.

Br

I H 91

92

spiroketone 88, which was transformed into the hydroxynitrile 89 by treatment with acetonitrile in the presence of n-butyllithium. LAH reduction of 89 furnished the amine 90, which on treatment with concentrated hydrochloric acid gave the amine 91. Treatment of 91 with (92) one equivalent of bromine yielded 3-methoxy-9-bromohasubanan in good yield (67, 74). A new synthetic method of dl-3-methoxy-N-methylhasubanan has been explored recently (75). Treatment of 91 with formalin in formic acid afforded dl-9,10-dehydro-3-methoxy-N-methylhasubanan, which was derived into dl-3-methoxy-N-methylhasubananby catalytic hydrogenation (75).

419

7. HASUBANAN ALKALOIDS

E. SYNTHESIS BY PHENOL OXIDATION Treatment of reticuline (93) with trifluoroacetic anhydride, followed by catalytic hydrogenation yielded the amide 94. Treatment of 94 with vanadium oxytrichloride gave rise, by phenol oxidation, to the dienone 95, which was transformed into the enone 96 by treatment with aqueous potassium carbonate solution. When reacted with methanolic hydrochloric acid, the enone 96 provided the cepharamine analog 97 (68). Me0

, 'COCF3

Me

HO

OMe

,

"")y u Ho

/

Md3(3 /

Me0

Me0

94

93

95

OMe

I

Me 96

OMe

I

Me 97

V. Synthesis of Hasubanan Alkaloids The syntheses of hasubanan alkaloids are of interest in connection with their pharmacological activities, since these alkaloids involve the structural unit of prafadol (98) (69),which is used as a potent analgesic. Hasubanan alkaloids are classified into three groups-the cepharamine, hasubanonine, and metaphanine types-on the basis of the oxidation stage at the B and C rings. The representative of each group has been synthesized from the common intermediate, the ketolactam 67, with an exception of one of two cepharamine syntheses.

420

YASUO INUBUSHI AND TOSHIRO IBUKA

A. CEPHARAMINE Methylation of the ketal99 derived from the ketolactam 67 (Section IV,A) with methyl iodide provided the ketal lactam 100. Since cepharamine (108) possesses a methoxyl group at C-3 and a hydroxyl group a t C-4, a partial demethylation step in the course of the synthetic route is required. When heated with potassium hydroxide and hydrazine in ethylene glycol ( 5 0 , 5 1 , 5 6 , 5 7 ) ,the ketal lactam 100 gave two kinds of phenols-101 (major) and 102 (minor). Acetylation of 101, followed by deketalization, provided the keto acetate 104, which on treatment with two equimolar quantities of bromine, followed in turn by heating with sodium acetate in acetic acid, gave an inseparable mixture consisting of the desired diketone 105 and an unidentified compound in a 1:1 ratio. Methylation of this mixture and separation of the reaction mixture furnished 16-oxocepharamine acetate (106) in a pure state. Reduction of 106 with LAH provided the epimeric alcohols (107), which on oxidation with DMSO-DCC-phosphoric acid (70) gave cepharamine (108) (57, 58, 62). OMe

OH

OMe

I

Me

I

I

R 99

9s

Me 101 R = H 103 R = AC

R = H

100 R = Me

& c

OMe

OH

?Me

\

I

Me

Me

Me

102

104

106

42 1

7. HASUBANAN ALKALOIDS OMe

OMe

OMe

I

Me

Me

Me

106

107

108

Another synthetic route to cepharamine utilizing photocyclization was designed. Heating of 2’-bromoreticuline (109) with trifluoroacetic anhydride, followed by catalytic hydrogenation, provided the dihydromethine 110. Irradiation of 110 with a mercury lamp in the presence of sodium hydroxide and sodium iodide gave the dienone 111. Hydrolysis of the amide function of 111 caused the Michael addition to yield an isomer of cepharamine. Transesterification of 112 with hydrochloric acid in methanol provided a mixture of cepharamine (108) and the starting material 112, from which cepharamine was isolated in a pure state (71). Me0 HO

H

HO

o

yJJ

d

Me0

Me0 109

& 110

?Me

M e00

Me0

I

COCF, 111

I

Me 112

422

YASUO INUBUSHI AND TOSHIRO IBUKA

B. HASUBANONINE AND AKNADILACTAM I n the synthesis of hasubanonine (129) from the ketolactam 67, introduction of two more oxygen functions a t the C-6 and C-8 positions are required. Oxidation of the ketolactam 67 with lead tetraacetate in the presence of boron trifluoride etherate gave three acetates-l13,114, and 115. In order to avoid the production of the undesired acetates 114 and 115, a lowering of the electron density of an aromatic ring was preferable. Thus, similar oxidation of the ketolactam 104 possessing an acetoxy group a t C-4 with lead tetraacetate was tried, and the acetoxyketone 116 was solely obtained in 65% yield. Treatment of 116 with two equimoIar quantities of bromine, followed by heating with sodium acetate, provided the enol acetate 117 and the bromoacetate 118 in a 1 O : l ratio, but the yield of 117 was rather poor. However, the acetoxyketone 116 was brominated with pyridinium hydrobromide perbromide, and the reaction product 119 was heated with sodium acetate to give solely the enol acetate 117. Partial hydrolysis of the enol acetate function of 117 provided the a-diketone 120, which was brominated to give

@ :

\

o&

.- f .

O H

R N

o&A(

. . AcO N

\O

Me 67

N/-0

O\

I

I Me

I R = H

Me 115

114

113 R = OAc

. .

RO

RO

RN

I

Me 104 R = H 116 R = OAc

I

Me 117 120

R = AC R =H

I

Me 118 R = AC 121 R = H

122 R = Me

&

7.

OMe

OMe

Br\ 0

. : . AcO N I

:fro

Br

Me0

Me0

I

119

Me

123

124

?Me

OM0

0

Me0

Me0

Me0

125 R = Me 127 R = H

I

Me

OMe

I

OMe

0

Me

Me

423

HASUBANAN ALKALOIDS

I

Me 126 R = Me 128 R = H

I

Me 129

the bromoketone 121 in high yield. Methylation of 121 with diazomethane furnished compound 122, which was heated in an aqueous sulfuric acid according to the Hesse’s procedure to produce predominantly the p-diketone 123 together with the compound 124. The p-diketone 123 was methylated with diazomethane, and silica gel chromatographic separation gave 16-oxohasubanonine (125) and its isomer (126) from the earlier eluate in a 1 :1 ratio, and continued elution provided aknadilactam (127) and its isomer (128) in a 1:1 ratio. On the other hand, permanganate oxidation of hasubanonine produced optically active 16-oxohasubanonine (28,38),a sample of which was proved to be identical with that of the synthetic one (125) except in optical rotation. Since LAH reduction of 16-oxohasubanonine followed by oxidation with activated manganese dioxide regenerated hasubanonine, the synthesis of 16-oxohasubanonine is equivalent to the complete synthesis of hasubanonine (129) (38, 72).

424

YASUO INUBUSHI AND TOSHIRO IBUKA

C. METAPHANINE The ketolactam 67 was also chosen as the starting material for the metaphanine synthesis. Since the introduction of an oxygen function a t the C-S position of 67 had been established during the synthesis of hasubanonine, the major problems are the stereoselective introduction of the C-10 hydroxyl group trans to the ethanamine bridge and the selective reduction of the lactam carbonyl group when both the lactam carbonyl and the hemiketal ring are present. Oxidation of 100 and 130 with chromic anhydride-acetic acid gave lo-0x0 compounds 131 and 132, respectively, but the yields were rather poor and irregular. The synthetic intermediate that had been utilized for the hasubanonine synthesis was converted to its ketal derivative 133.Chromic anhydride oxidation of 133 provided the 10-0x0 ketal lactam 134 in high yield. Hydrolysis of the acetoxyl groups of 134,followed by methylation with diazomethane, produced 10-0x0 compound 135.For the purpose of the hemiketal formation between C-8 and C-10, the relative configuration of the hydroxyl group derived from C-10 0x0 group must be trans to the ethanamine bridge. (The terms “cis” and “trans” in this section are

H

100 131

Me R = H, R =0

Me 130 132

R R

Me

= H, =0

133 134

R = H, R =0

“OR

Me 135

Me 136 R = H 143 R = Ac

Me 137

7.

425

HASUBANAN ALKALOIDS

OMe

OMo

OMe

I

Me

Me

138 139 140

R R R

=H = AC = THP

141 142

R = Ao

Me 144

R = THP

used to express the relative configuration of the C-10 hydroxyl group to the ethanamine bridge.) Reduction of 135 with various metal hydrides was tried, but the major product was the undesired cis C-10 hydroxyl compound 136, although the cis-trans ratio varied depending on solvents and metal hydrides used. Catalytic hydrogenation of the C-10 0x0 compounds 135 and 137 was unfruitful. Next, reduction of 135 with sodium in various alcohols was examined, and in this case, the yield of the trans isomer was superior to that of the cis isomer. However, the total yield was rather poor. Finally, reduction of 135 was successfully carred out by the Meerwein-Varley procedure to give the trans C-10 hydroxyl compound 138 in an excellent yield. After the hydroxyl group at C-10 of 138 was protected as an acetoxyl group or a pyranyl ether group, the acetate 139 or the pyranyl ether 140 was oxidized to produce the C-8 0x0 compound 141 or 142, Removal of the protected group afforded 16-oxometaphanine. Jones’ oxidation of the cis C-10 acetoxyl compound 143 gave the ketoacetate 144, which on treatment with aqueous sodium carbonate solution produced the C-10 0x0 compound 135. This rearrangement was assumed to be caused by an intramolecular 1,4 hydride shift from C-10 to C-S of compound 144. I n order to demonstrate this mechanism, 135 was converted to the deuterated cis C-10 hydroxyl compound 145, which gave the deuterated cis C-10 acetate 146 by acetylation. Jones’ oxidation of 146 gave the C-S 0x0 compound 147, which on treatment with aqueous sodium carbonate solution produced quantitatively the C-10 0x0 compound possessing deuterium at C-S with the ,l3 configuration, as indicated by the mechanism shown in 148. Thus, validity of the 1,4-hydride shift was verified, and the stereochemistry of the C-10 oxygen functions, which was based on the NMR spectral analyses, was chemically established.

426

YASUO INUBUSHI AND TOSHIRO IBUKA

"OR

I Me 145 R = H 146 R = Ac

Me

Me 147

148

"H

Me 149

150 155

Me R =0 R =S

OTHP

Me 151

.'H HO N

I

Me

Me